DATA FOR BETTER

GOVERNANCE
BUILDING GOVERNMENT ANALYTICS ECOSYSTEMS
IN LATIN AMERICA AND THE CARIBBEAN
Juan Francisco Santini, Flavia Sacco Capurro, Daniel Rogger, Timothy Lundy,
Galileu Kim, Jorge de León Miranda, Serena Cocciolo, and Chiara Casanova
Data for Better
Governance

This book, along with any associated content or subsequent updates, can be accessed at
https://hdl.handle.net/10986/42413.
 https://reproducibility.worldbank.org

A reproducibility package is available for this book in

the Reproducible Research Repository at
https://reproducibility.worldbank.org/index.php/catalog/209.

Scan to learn about the

Government Analytics collection.
Data for Better
Governance
Building Government Analytics
Ecosystems in Latin America
and the Caribbean

J UA N F R A NC I SC O SANTI NI ,
F L AV I A SACC O CAPUR R O,
DA NI E L R OGGER , TI M OTHY LUN DY,
GA L I L E U K I M , JOR GE DE L EÓN MIRAN DA,
S E R E NA C O C C I OLO, AND C HI ARA CASAN OVA
© 2024 International Bank for Reconstruction and Development / The World Bank
1818 H Street NW, Washington, DC 20433
Telephone: 202-473-1000; Internet: www.worldbank.org
Some rights reserved
1 2 3 4 27 26 25 24
This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and
conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive
Directors, or the governments they represent.
The World Bank does not guarantee the accuracy, completeness, or currency of the data included in this work and
does not assume responsibility for any errors, omissions, or discrepancies in the information, or liability with respect
to the use of or failure to use the information, methods, processes, or conclusions set forth. The boundaries, colors,
denominations, links/footnotes, and other information shown in this work do not imply any judgment on the part of
The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. The
citation of works authored by others does not mean The World Bank endorses the views expressed by those authors or
the content of their works.
Nothing herein shall constitute or be construed or considered to be a limitation upon or waiver of the privileges and
immunities of The World Bank, all of which are specifically reserved.
Rights and Permissions

This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http://
creativecommons.org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy,
distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions:
Attribution—Please cite the work as follows: Santini, Juan Francisco, Flavia Sacco Capurro, Daniel Rogger, Timothy
Lundy, Galileu Kim, Jorge de León Miranda, Serena Cocciolo, and Chiara Casanova. 2024. Data for Better Governance:
Building Government Analytics Ecosystems in Latin America and the Caribbean. Washington, DC: World Bank.
doi:10.1596/978-1-4648-2159-2. License: Creative Commons Attribution CC BY 3.0 IGO
Translations—If you create a translation of this work, please add the following disclaimer along with the attribution:
This translation was not created by The World Bank and should not be considered an official World Bank translation.
The World Bank shall not be liable for any content or error in this translation.
Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution:
This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole
responsibility of the author or authors of the adaptation and are not endorsed by The World Bank.
Third-party content—The World Bank does not necessarily own each component of the content contained within
the work. The World Bank therefore does not warrant that the use of any third-party-owned individual component
or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from
such infringement rests solely with you. If you wish to re-use a component of the work, it is your responsibility to
determine whether permission is needed for that re-use and to obtain permission from the copyright owner. Examples of
components can include, but are not limited to, tables, figures, or images.
All queries on rights and licenses should be addressed to World Bank Publications, The World Bank, 1818 H Street NW,
Washington, DC 20433, USA; e-mail: [email protected].
ISBN (paper): 978-1-4648-2159-2
ISBN (electronic): 978-1-4648-2178-3
DOI: 10.1596/978-1-4648-2159-2
Cover image: © Pixel Matrix / Adobe Stock. Used with the permission of Pixel Matrix / Adobe Stock. Further permission
required for reuse.
Cover design: Bill Pragluski, Critical Stages, LLC.
Library of Congress Control Number: 2024950453
Contents

Forewordix
Acknowledgmentsxi
About This Collection xiii
About the Authors xv
Main Messages  xvii
Executive Summary xxi
Abbreviationsxxv

Chapter 1 The Case for Government Analytics in

Latin America and the Caribbean 1
Introduction1
Key Messages 4
The Government Analytics Ecosystem 5
Regional Opportunities and Challenges for Government Analytics 6
A Regional Assessment and Policy Recommendations to Strengthen
Government Analytics10
Notes15
References16
Chapter 2 A Conceptual Framework for Government Analytics 19
Introduction19
How Does Government Analytics Improve Government Functioning? 21
What Enabling Conditions Support Government Analytics? 29
Notes37
References37
Chapter 3 Case Studies of Government Analytics in
Latin America and the Caribbean 39
Introduction39
Case Study 3.1: Using Descriptive Analytics to Reduce Missed
Medical Appointments (Chile) 40
v
 Case Study 3.2: Using Diagnostic Analytics of Human Resources Data
to Explore the Gender Pay Gap (Colombia) 42
Case Study 3.3: Data Analytics for Citizen Accountability—The Case of
reAcción (Paraguay)44
Case Study 3.4: Building an Integrated Human Resources Management
Platform for Strategic Workforce Planning (Uruguay) 45
Case Study 3.5: Analyzing Tax Data to Increase Corporate
Tax Collection (Ecuador)47
Case Study 3.6: Using Predictive Analytics to Prevent School
Dropouts (Guatemala)49
Case Study 3.7: Using Tax Data to Boost Procurement Efficiency (Brazil) 50
Case Study 3.8: Improving Government Efficiency through a
Centralized Analytics Unit (Colombia) 51
Case Study 3.9: Establishing a Decentralized Analytics Unit
to Improve the Efficiency of Tax Collection (Peru) 54
Case Study 3.10: Datathons as a Tool for Promoting Data Analytics
(Multiple Countries)55
Case Study 3.11: The Power of Data Integration to Strengthen the
Integrity of Public Procurement (Brazil) 56
Case Study 3.12: Supporting the Implementation of Framework
Agreements through Data Analytics (Uruguay) 57
Notes58
References59
Chapter 4 A Regional Assessment of Government Analytics in
Latin America and the Caribbean 61
Introduction61
Survey Methodology 63
What Does Government Analytics Look Like in Latin America
and the Caribbean? 64
What Analytical Capabilities Do Governments in the Region Possess? 70
What Is the State of Data Infrastructure in the Region? 74
Notes82
References82
Chapter 5 Policy Recommendations to Strengthen Government
Analytics in Latin America and the Caribbean 83
Introduction83
Policy Recommendations 84
Conclusion89
References90

vi Data for Better Governance

APPENDIXES
Appendix A Survey Methodology 91
Appendix B Types of Management Information Systems and Their Functions 97

BOXES
1.1 The Government Analytics Handbook. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Why Administrative Data? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

FIGURES
1.1 Central Government Revenue and Spending Composition
in Latin America and the Caribbean, 2010–22. . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.2 Coverage of Key Management Information Systems in
Latin America and the Caribbean and Worldwide . . . . . . . . . . . . . . . . . . . . . . . . 9
2.1 A Conceptual Framework for Government Analytics . . . . . . . . . . . . . . . . . . . . 20
2.2 Cumulative Fiscal Savings in Three Wage Bill Policy Scenarios,
Brazil, 2019–30 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.3 Enabling Conditions for Government Analytics. . . . . . . . . . . . . . . . . . . . . . . . . 29
3.1 Using Descriptive Analytics to Reduce Missed Medical
Appointments, Chile. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
3.2 Building an Integrated Human Resources Management
Platform for Strategic Workforce Planning, Uruguay. . . . . . . . . . . . . . . . . . . . . 47
3.3 Using Predictive Analytics to Prevent School Dropouts, Guatemala . . . . . . . 50
3.4 Improving Government Efficiency through a Centralized
Analytics Unit, Colombia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
4.1 Types of Analytics, by Type of MIS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
4.2 Types of Analytics, by Country and Type of MIS. . . . . . . . . . . . . . . . . . . . . . . . . 65
4.3 Data Elements Used for Analytics, by Type of MIS . . . . . . . . . . . . . . . . . . . . . . 66
4.4 Applications of Analytical Products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
4.5 Applications of Analytical Products, by Type of MIS . . . . . . . . . . . . . . . . . . . . . 69
4.6 Kinds of Analytical Production, by Type of MIS. . . . . . . . . . . . . . . . . . . . . . . . . . 70
4.7 Career Tracks for and Training and Assessments on Analytics. . . . . . . . . . . . 71
4.8 Opportunities for Internal Funding and Collaboration with Academics,
Nonprofits, or Multilateral Organizations on Analytics Projects,
by Type of MIS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
4.9 Drivers of Collaboration on Analytics with Academics,
Nonprofits, or Multilateral Organizations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73

Contents vii
 4.10 Fully Digitalized Systems, by Type of MIS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
4.11 Fully Digitalized Systems, by Country and Type of MIS. . . . . . . . . . . . . . . . . . . 75
4.12 Formal Data Access Protocols, by Type of MIS . . . . . . . . . . . . . . . . . . . . . . . . . 78
4.13 Formal Data Access Protocols, by Country and Type of MIS. . . . . . . . . . . . . . 78
4.14 Data Quality Controls in Place, by Type of MIS. . . . . . . . . . . . . . . . . . . . . . . . . . 80

MAP
4.1 Countries in Latin America and the Caribbean That Participated in
the Survey. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

TABLES
1.1 Summary of Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.2 Key Findings and Policy Recommendations for Strengthening
Government Analytics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
A.1 Countries and Organizations That Participated in the Survey 92
A.2 Countries That Participated in the Survey, by Questionnaire 95

viii Data for Better Governance

Foreword

Improving the quality of governance is critical to virtually all the agendas in Latin
America and the Caribbean (LAC), whether they are providing better and more
responsive social services, accelerating growth, ensuring public safety, or supporting
lagging areas. Further, high debt service levels and the absence of public fiscal space put
a premium on using available resources as efficiently as possible.

The LAC region is well positioned to make progress on these challenges. It has invested
heavily over the past 20 years in building information systems to underpin government
business. The data stored and tracked in digital management information systems in
the region now cover an average of 79 percent of total government revenue and grants
and at least 40 percent of the total spending of the central government.

Embedded in every byte of these data is an opportunity to strengthen the quality of

the region’s governance. Each byte acts not just as a record of government business but
also as a diagnostic of its strengths and weaknesses. The repurposing of government
administrative data toward strengthening public administration is what the World
Bank has termed “government analytics.” In line with transformations of private sector
companies based on the value inherent in large collections of data, the public sector
has an opportunity to change its own production function for better service delivery
and more effective government.

This repurposing of government administrative data toward a diagnostic purpose is

challenging, and it requires that data systems be made accessible to analysts and that
analysts have the capabilities to deliver useful insights from these data. This requires the
right staff in the right places, a culture of evidence-based policy making, and national
commitment over time.

This report explores the extent to which LAC governments are indeed analyzing their
administrative data to strengthen the functioning of public administration. The picture
painted is that government analytics is being done in LAC, but its potential remains
largely untapped. There is little longer-term strategic vision, training is ad hoc and
disparate, and data systems could be coordinated much better.

ix
 As the examples provided in The Government Analytics Handbook (2023) show, the
impact of realizing our regional potential for undertaking government analytics in
terms of performance and efficiency is huge, leading to hundreds of millions of dollars
in additional government revenue, better personnel management, and immediate
improvements in the targeting of resources to schools and more responsive health
provision. This volume in the Government Analytics collection provides a road map
to realize that potential and a link between academic studies and practical policy
measures. We have much to do.

William F. Maloney
Chief Economist, Latin America and the Caribbean Region
Former Chief Economist, Equitable Growth, Finance, and
Institutions Vice Presidency
The World Bank

x Data for Better Governance

Acknowledgments

This report is a product of the Bureaucracy Lab, a partnership between the World
Bank’s Governance Global Practice (GGP) and Development Impact (DECDI) Group.
It was produced in collaboration with the GGP—in particular, the Governance Team in
the Latin America and the Caribbean Region (LCR); the Public Institutions Data and
Analytics Global Unit; and the Public Administration Global Unit. The report is funded
by the Office of the Chief Economist for LCR, with the objective of assessing and
providing guidance on how to strengthen the use of analytics to governments in the
region. It was produced by a core team consisting of Chiara Casanova, Serena Cocciolo,
Jorge de León Miranda, Galileu Kim, Timothy Lundy, Daniel Rogger, Flavia Sacco
Capurro, and Juan Francisco Santini. We would like to thank a former consultant at the
Bureaucracy Lab, Anna Paula Matos, who contributed to this report in its early stages.

The report was produced under the overall guidance of William F. Maloney
(chief economist, LCR), Arianna Legovini (director, DECDI), Arturo Herrera
(global director, GGP), Oscar Calvo-Gonzalez (regional director, LCR), and Chiara
Bronchi (practice manager, GGP Public Institutions Data and Analytics Global Unit).
We are grateful to Eric Arias (economist), Elizabeth Grandio (senior procurement
specialist), Zahid Hasnain (lead governance specialist), Silvana Kostenbaum (senior
public sector specialist), Alberto Leyton (practice manager, GGP), Bernard Myers
(senior public sector specialist), Diana Parra Silva (senior public sector specialist),
Francesca Recanatini (lead economist), Ruben Ruano Chinchilla (senior public sector
specialist), Luciano Wuerzius (senior procurement specialist), and the Governance
and Procurement teams in LCR for their overall guidance and support. We thank
Cem Dener (lead public sector specialist), Hunt La Cascia (senior public sector
specialist), Hubert Nii-Aponsah (consultant), and the GovTech team for their generous
guidance in structuring the report and implementing a regional survey on government
analytics. We also extend our gratitude to the three reviewers of the report’s concept
note—Adrian Fozzard (practice manager), Sebastian Galiani (professor, University
of Maryland), and Daniel Ortega (lead governance specialist)—for their insightful
comments and valuable guidance throughout the writing of this report. We thank the
three peer reviewers for the final version of the report—Cem Dener, Julia Michal Clark,

xi
 and Daniel Ortega—for their helpful comments. Last, we would like to thank Mariano
Lafuente, Carlos Pimenta, Alejandro Rasteletti, and David Rivera at the Inter-American
Development Bank for sharing their experience developing analytics in Latin America
and the Caribbean, as well as Charlotte van Ooijen and Luanna Roncaratti for sharing
their expertise on digital government.

Finally, we would like to thank the government officials who were partners in
the design and creation of the evidence presented in this report. More than
100 government officials from 20 countries coordinated the completion of the
regional survey on government analytics in Latin America and the Caribbean, and
their contributions were essential to the completion of the report. In particular, we
would like to thank Charlene Laing (The Bahamas), Nicola Callender (Barbados),
Maria Pech and Alexia Peralta (Belize), Khantuta Muruchi and Francisco Belmonte
(Bolivia), Ciro Avelino and Ronnie Dilli (Brazil), Jose Inostroza and Rafael Hernández
(Chile), Manuela Serrano (Colombia), Erick Mora (Costa Rica), Jermaine Jean-Pierre
(Dominica), Alejandra Perez and Edwin Rodriguez (Dominican Republic), Juan Yepez
(Ecuador), Hugo Forkel and Daniel de León (Guatemala), Heidy Alachán (Honduras),
Gary Campbell and Anika Shurrleworth (Jamaica), Evelyn Rodriguez (Panama), Laura
Salinas (Paraguay), Darwin Quispe (Peru), Caswallon Duncan (St. Vincent and the
Grenadines), Shelley-Ann Clarke and Natasha Ottley (Trinidad and Tobago), and
Ignacio Velazco (Uruguay).

xii Data for Better Governance

About This Collection

This report is part of a World Bank collection examining how analytics using
government microdata is revolutionizing public administration throughout the
world. The collection is based on The Government Analytics Handbook (2023),
a comprehensive guide to using data to understand and improve government. The
collection encompasses practical guides and resources for policy makers and public
officials around the world seeking to improve government functioning by better using
their administrative and survey data. It includes The Government Analytics Handbook
and associated tools as well as region-specific reports, data, and approaches for
practitioners seeking a deeper understanding of government analytics.

The general principle of the Handbook is as follows: governments across the world
make thousands of personnel management decisions, procure millions of goods
and services, and execute billions of processes each day. They are data rich. And
yet there is little systematic practice to date that capitalizes on these data to make
public administrations work better. This means that governments are missing out on
data insights to save billions in procurement expenditures, recruit better talent into
government, and identify sources of corruption—to name just a few.

The Handbook seeks to change that. It presents frontier evidence and practitioner
insights on how to leverage data to make governments work better. Covering a range of
microdata sources—such as administrative data and public servant surveys—as well as
tools and resources for undertaking analytics, it transforms the ability of governments
to take a data-informed approach to diagnose and improve how public organizations
work.

Throughout this report, gold callout balloons will

be used to point out chapters in The Government
Analytics Handbook where you can learn more
about each topic the report discusses.

xiii
About the Authors

Chiara Casanova is a marketing and communications specialist in the World Bank’s

Development Impact Group. She is an experienced strategic communications
professional specializing in social impact and development projects, with a background
in public affairs and corporate communications. Her extensive background lies
in developing and implementing integrated communication strategies for private,
nonprofit, and international development clients. She holds a master’s degree in public
relations and corporate communications from Georgetown University.

Serena Cocciolo is an economist in the World Bank’s Governance Global Practice for
the Middle East and North Africa and a member of the Bureaucracy Lab. Her work
has focused on strengthening public administrations, public procurement, institutions,
and governance. She joined the World Bank after completing her dissertation on
community participation in development projects and field applications in the water
and education sector. She holds a PhD in economics from Stockholm University.

Jorge de León Miranda is a research and project management professional with

more than 10 years of international development work focused on public sector
management, public financial management, fiscal policy, macroeconomics, and
education. His projects have covered several countries in Latin America and the
Caribbean, Africa, the Middle East, and Central Asia. He has also held positions at
Ernst and Young, the Inter-American Development Bank, the International Monetary
Fund, and the Ministry of Economy of Guatemala. He holds an MA in international
economics and finance from the Johns Hopkins University School of Advanced
International Studies.

Galileu Kim is a public sector specialist in the Public Institutions Data and Analytics
Global Unit in the Governance Global Practice of the World Bank. His work focuses
on how governments can leverage data and analytics to improve public administration,
with a methodological focus on large-scale administrative data. He is a member of the
World Bank’s Bureaucracy Lab and a former researcher in the Development Impact
Group. He is a research fellow in Brazil’s National School of Administration. He holds a
PhD in political science from Princeton University.

xv
 Timothy Lundy is an editor and writing consultant in the World Bank’s Development
Impact Group. He holds a PhD in English and comparative literature from Columbia
University.

Daniel Rogger is a senior economist in the World Bank’s Development Impact Group.
He manages the group’s Governance and Institution Building unit and is colead of the
World Bank’s Bureaucracy Lab, a collaboration between the group and the Governance
Global Practice that aims to bridge research and policy to strengthen public
administration. His research focuses on the organization of the delivery of public
goods. He is a cofounder of the Worldwide Bureaucracy Indicators, Global Survey of
Public Servants, and Microdata and Evidence for Government Action initiatives. He
was a PhD scholar at the Institute for Fiscal Studies, where he is now an international
research fellow. He holds a PhD in economics from University College London.

Flavia Sacco Capurro is a data coordinator in the World Bank’s Development Impact
Group working on issues related to public sector employment and reforms, government
analytics, and gender in the public sector. Before joining the group, she was a
consultant in the Poverty and Equity Global Practice in the Latin America and the
Caribbean Region, working on projects related to poverty analysis, inequality, gender,
social development, and fiscal policy. Previously, she worked for the government of
Paraguay in the Ministry of Economic and Social Planning. She holds a master’s degree
in public policy from the University of Chicago.

Juan Francisco Santini is a researcher in the Governance and Institution Building

unit of the World Bank’s Development Impact Group, a member of the World
Bank’s Bureaucracy Lab, and a Visiting Research Fellow at Brazil’s National School
of Public Administration. Before joining the World Bank, he was a Research Fellow
at Innovations for Poverty Action and worked for the Inter-American Development
Bank, among other institutions. His research and policy interests lie in development
economics, particularly its intersection with public economics and public management.
He holds a PhD in economics from the Pontifical Catholic University of Rio de Janeiro,
Brazil.

xvi Data for Better Governance

Main Messages

Governments in Latin America and the Caribbean face complex development

challenges, from slowing economic growth and inflation to persistent institutional
weaknesses across the public sector. Strengthening government functioning to respond
effectively to these challenges requires a detailed understanding of what is failing and
where. It requires diagnostics based on granular, real-time information and the ability
for every public servant to take action.

Governments already have the data they need to meet the challenges they face.
They have been producing and collecting data in great detail and for many years. But
they need to be able to analyze the data they collect in the course of their everyday
operations to inform managerial decisions across every government function. For
instance, data on recruitment practices can inform human resources management
decisions, data on payment delays can improve procurement processes, and data on
taxpayer compliance can help design tax instruments. By not taking advantage of the
data that are now being collected in digital systems all over the region, governments are
leaving substantial amounts of money on the table and diminishing their impact on the
lives of citizens.

Data for Better Governance: Building Government Analytics Ecosystems in Latin

America and the Caribbean provides a conceptual framework and empirical evidence
for evaluating how governments use administrative data to improve the functioning
of the state, and it highlights opportunities for improvement. Governments in Latin
America and the Caribbean are world leaders in the digitalization of government
work—but they do not systematically use the data they have collected to diagnose and
improve their functioning. This represents a missed opportunity.

The report offers governments in this region and beyond a road map for government
analytics: repurposing government data to improve the efficiency and effectiveness of
each aspect of government functioning. It identifies the critical enabling conditions
for government analytics—data infrastructure and analytical capabilities—and it
offers strategies for strengthening them. The report draws on data from a survey of
government officials from 20 countries in Latin America and the Caribbean who are
experts in core government functions and their respective management information

xvii
 systems, as well as 12 case studies of analytical initiatives, to present the following key
findings and recommendations.

EXPANDING THE USE OF GOVERNMENT ANALYTICS IN

LATIN AMERICA AND THE CARIBBEAN
The use of analytics in Latin America and the Caribbean varies across government
functions and policy areas. There is also significant heterogeneity in the systems
and practices used within countries. This suggests that governments lack a systematic
approach to analytics for the whole of public administration. This also presents an
opportunity for cross-fertilization, allowing countries to share their experiences and
adopt successful practices demonstrated by their peers.

Governments in the region can use advanced analytics more extensively to address
complex development challenges. Governments predominantly use administrative
data to produce descriptive analytics for operational and transactional purposes, but
they miss out on opportunities to use advanced analytics to improve decision-making,
design more effective and efficient public policies, and strengthen public sector
functioning and service delivery.

The impact of government analytics is already being felt in the region. For example,
Ecuador and Peru have collected millions of dollars in additional tax revenue by
analyzing transactional and external data to detect evasion and better allocate
resources for enforcement. Guatemala has improved education services by analyzing
student-level data to identify and support at-risk students, reducing dropout rates
by 9 percent for students at a pivotal moment in their education. These examples
demonstrate that governments can use analytics to improve many different aspects
of their operations, capitalizing on the wealth of data contained in their management
information systems.

BUILDING ENABLING CONDITIONS FOR

GOVERNMENT ANALYTICS
Governments must complete the digitalization of their management information
systems. Two-thirds of the experts surveyed reported that their systems are not
fully digitalized. Levels of digitalization vary substantially by government function:
information systems for taxation and public financial management exhibit a relatively
high degree of digitalization, while health management information systems have the
lowest degree of digitalization. Incomplete digitalization results in data infrastructure
problems that prevent governments from fully leveraging their administrative data to
improve policy design and implementation.

High-quality, integrated data are essential to analytics, but data fragmentation

and isolated information systems constrain digitalization. Outdated or inadequate

xviii Data for Better Governance

data infrastructure limits the quality and accessibility of administrative data,
making it difficult for organizations to use these data for analytics and policy making.
When management information systems are fragmented, various subsystems operate
independently, making it virtually impossible to create a centralized, fully digitalized
system. Consequently, data sharing and interoperability between the different systems
are restricted, the quality of data available for analytics is compromised, and the
available data do not always align with strategic priorities and policy needs.

Information systems need mechanisms to ensure the quality of the data they
contain. According to the World Bank’s GovTech Maturity Index, only 25 percent
of countries in Latin America and the Caribbean have implemented a data quality
framework, which is lower than the implementation rate in other regions. When data
quality controls are not integrated into an information system, each team using the
system’s data must conduct its own quality control process, resulting in inefficiencies.
Moreover, the lack of systematic controls can undermine the accuracy, reliability, and
replicability of analytics.

Improving data accessibility and system interoperability should be a government-

wide effort. According to the GovTech Maturity Index, less than 35 percent of
countries in Latin America and the Caribbean have implemented a government
interoperability framework that allows for efficient and secure information exchange
between government systems and organizations. Sharing information enables new and
innovative analytical applications: Brazil has repurposed tax data to boost procurement
efficiency, reducing purchase prices by about 13 percent. If governments establish the
foundational infrastructure for interoperability, they can advance their overall potential
for using data for analytics and policy making enormously.

Data systems are not enough; public administration must change, too. Governments
can systematize the use of analytics in decision-making by incorporating dedicated
analytics units into the organizational structure of public administration. In
Colombia, for instance, a dedicated analytics unit has supported quality-of-life
improvements, information sharing, and public service delivery for the entire city of
Bogotá, thanks to its centralized organizational model. On the other hand, relying on
part-time teams for analytical tasks can lead to suboptimal outcomes because these
teams might prioritize immediate operational needs over broader analytical objectives.

INTEGRATING DATA ANALYTICS INTO

DECISION-MAKING PROCESSES
Analytics is a tool, not a replacement for decision-makers or experience. Analytics
cannot replace decision-makers, design policies, or substitute for the experience
and knowledge of policy makers and public servants. But officials must be given the
ability to use the evidence provided by analytics to guide key decisions, identify gaps
and weaknesses, and refine the design and implementation of policy interventions.
Training to support decision-makers and data analysts in using government analytics

Main Messages xix

 is fragmented and weak across Latin America and the Caribbean. This can be
improved rapidly.

Governments in the region need more structured efforts to attract and retain
skilled data analysts at all levels of seniority. Offering career tracks for data analysts
is critical to building a robust government analytics ecosystem. Governments in the
region face a shortage of structured career development opportunities in analytics: only
12 percent of governments have a dedicated career track for data analysts.

Finally, governments must foster a culture of evidence-based decision-

making, which is essential to using government analytics to transform public
administration. Analytics is more impactful when policy makers ask for the evidence
they need and use it to shape policies and inform critical decisions. The region’s
governments require a deeper culture of evidence-based decision-making. Knowledge
exchange on best practices and successful experiences can increase decision-makers’
awareness of the benefits of grounding policy decisions in data and evidence.
Improving decision-makers’ knowledge and ability to comprehend data analytics
results—and identify their limitations—is crucial for effectively integrating analytical
findings into decision-making processes.

Latin America and the Caribbean—and the rest of the world—has the data it needs to
improve government. Now is the time to use them.

xx Data for Better Governance

Executive Summary

INTRODUCTION
Governments in Latin America and the Caribbean face complex development
challenges. The region is at substantial risk from slowing economic growth,
inflation, and fiscal pressures, and its public sector is constrained by institutional
weaknesses, including inefficient public spending and service delivery. Core functions
of government, including human resources management and public financial
management, require substantial improvements in their efficiency and effectiveness.
Strengthening government functioning requires actionable diagnostics, based
on granular, real-time information: for instance, data on recruitment practices to
inform human resources management decisions, data on payment delays to improve
procurement processes, and data on taxpayer compliance to design tax instruments.
Governments can use administrative data that are specific to government functions to
inform managerial decisions to improve those functions.

Governments in Latin America and the Caribbean are well positioned to leverage data
for better government. The region is a global pioneer in establishing management
information systems (MISs): specialized systems that collect administrative data on
specific functions of government, including human resources, procurement, service
delivery, taxes, and more. These information systems gather granular, real-time data
on core government functions, and these data can serve as a rich source of information
on challenges and opportunities in these functions. However, the administrative
data recorded in these information systems are often underutilized because of
inadequate data quality and accessibility, as well as limited analytical capabilities within
governments. By addressing these constraints, governments can use administrative
data to improve the efficiency and effectiveness of each aspect of government
functioning.

A reproducibility package is available for this book in the Reproducible Research Repository at
https://reproducibility.worldbank.org/index.php/catalog/209.

xxi
 THIS REPORT
This report outlines how governments in Latin America and the Caribbean can enable
the use of their administrative data to strengthen government functioning, a practice
the report refers to as government analytics. It provides a conceptual framework that
governments can use to assess how they employ administrative data for analytics and
identify areas for improvement. The framework can be applied to different information
systems and their corresponding functions (for instance, education and health services,
procurement, and taxation), and it provides detailed insights and recommendations
to improve the use of data for each core function. For example, a government may be
using its tax data to their full analytical potential while underutilizing human resources
data for personnel management.

The report argues that governments can better leverage their data to improve their
functioning by focusing on the enabling conditions for analytics: data infrastructure
and analytical capabilities. For each type of information system, governments should
build data infrastructure that ensures data are accurate, high-coverage, and accessible.
As a whole, governments can strengthen their analytical capabilities by setting up
dedicated analytical units and career tracks for data analysts. Government analytics and
its enabling conditions are part of an ecosystem: by improving the enabling conditions
for analytics, governments create an environment in which administrative data can be
used to strengthen their functioning, improving effectiveness and efficiency in human
resources management, procurement, taxation, and more.

Guided by the conceptual framework, the report provides evidence on the

government analytics landscape in Latin America and the Caribbean. First,
it presents 12 case studies documenting how governments in the region have
leveraged their administrative data to improve specific aspects of government
functioning. These case studies cover a wide set of countries, information systems,
and government functions, highlighting the significant value and impact of
government analytics. For instance, Ecuador and Peru have collected millions of
dollars in additional tax revenue by analyzing tax data to detect evasion and better
allocate resources for enforcement. Guatemala has improved education services,
reducing student dropout rates by analyzing education data to identify and support
at-risk students. The case studies illustrate that governments can use analytics
to enhance distinct dimensions of their functioning, drawing on the richness of
administrative data.

Next, the report presents quantitative data from an original survey of government
officials conducted in 20 countries and across six types of information systems, offering
the first comprehensive picture of government analytics and its enabling conditions
in Latin America and the Caribbean. In each country, the survey questionnaires were
answered by public servants from institutions responsible for managing different
information systems. The respondents assessed the extent to which institutions in
core functions of government use administrative data to improve their functioning.
The regional survey reveals significant variation in the use of government analytics,

xxii Data for Better Governance

the state of data infrastructure, and the strength of analytical capabilities across
different governments and their core functions.

Governments in Latin America and the Caribbean predominantly use descriptive

analytics. Their use of diagnostic and predictive analytics is uneven and is mostly
concentrated on taxation and service delivery. They do not substantially use diagnostic
and predictive analytics for cross-cutting government functions, such as human
resources management, procurement, and public financial management. In sum,
governments in the region primarily use analytics to strengthen the revenue side of
their operations (for instance, taxation) rather than the expenditure side (for instance,
procurement). Across information systems, governments primarily apply analytics
for monitoring, with less focus on policy evaluation and design. This means that
governments underutilize their administrative data for more complex types of analytics
that could help them diagnose and address the causes of inefficiencies. By expanding
the use of different types of analytics for all their functions, governments can enhance
the efficiency and effectiveness of public administration more broadly.

The survey also reveals opportunities for institutional reform. Some information
systems in the region are only partially digitalized, limiting the quality and
comprehensiveness of the administrative data they generate. For example, health
information systems are rarely fully digitalized, owing to the use of paper records.
Information systems also exhibit significant gaps in data governance. In education,
health, and public financial management, data accessibility is restricted by a lack of
formalized access protocols. In addition, whereas data quality controls are widespread
for education and tax systems, procurement, public financial management, and
health information systems lag behind. Interoperability among information systems
and data sharing among government organizations are similarly limited. The survey
also underscores the challenges governments face in developing their analytical
capabilities. Many lack structured career tracks for data analysts, targeted training in
analytical skills, and adequate funding for analytical projects. These constraints limit
governments’ capacity to expand the use of data to improve their functioning.

Finally, based on the survey and case studies and guided by the conceptual framework,
the report presents a set of policy recommendations. First, governments should
identify evidence gaps for policy making and create awareness of how administrative
data could help fill these gaps, boosting demand for government analytics. They should
aim to strengthen their use of descriptive analytics and move toward diagnostic and
predictive analytics applications to inform policy evaluation and design. Governments
should also invest further in the enabling conditions for government analytics. To
build analytical capabilities, governments should develop targeted data analytics
training programs and set up institutional pathways for applying analytical skills. This
can be done by establishing career tracks for analysts, setting up dedicated analytics
units, allocating a stable budget for training and analytical products, and fostering
collaborations with academic partners. To strengthen data infrastructure, governments
should institute systematic data quality controls, develop comprehensive and accessible
data inventories, enhance data connectivity and information system interoperability,
and establish protocols for maintaining and updating their e-government systems.

Executive Summary xxiii

 This report is part of a series of World Bank publications focusing on government
analytics, including The Government Analytics Handbook (Rogger and Schuster 2023).
It is designed as a guide for policy makers and senior public officials who want to
improve government functioning by using their administrative data to its full potential.
As such, the report does not extensively cover the technical details of implementation,
but it provides additional references for practitioners seeking a deeper understanding
of the technical requirements for government analytics. By using this report’s
conceptual framework and survey as a starting point, governments can assess their use
of government analytics and assign priority levels to actions to build a more efficient
and effective public administration.

REFERENCE
Rogger, Daniel, and Christian Schuster, eds. 2023. The Government Analytics Handbook: Leveraging Data to
Strengthen Public Administration. Washington, DC: World Bank. https://doi.org/10.1596/978-1-4648-1957-5.

xxiv Data for Better Governance

Abbreviations

COVID-19 coronavirus disease 2019

CV curriculum vitae
DECDI Development Impact Group [World Bank organizational unit]
DNCP Dirección Nacional de Contrataciones Públicas (National Public
Procurement Agency) [Paraguay]
EdMIS education management information system
ENTRE Estrategia Nacional para la Transición Exitosa (National Strategy for
Successful Transition) [Guatemala]
FONACIDE Fondo Nacional de Inversión Pública y Desarrollo (National Public
Investment and Development Fund) [Paraguay]
GDP gross domestic product
GGP Governance Global Practice [World Bank organizational unit]
GRAS Governance Risk Assessment System
HealthMIS health management information system
HRMIS human resources management information system
LCR Latin America and the Caribbean Region [World Bank operational
unit]
MINEDUC Ministerio de Educación (Ministry of Education) [Guatemala]
MINSAL Ministerio de Salud (Ministry of Health) [Chile]
MIS management information system
NSO national statistical office
ONSC Oficina Nacional del Servicio Civil (National Civil Service Office)
[Uruguay]
PFMIS public financial management information system
SEFAZ/RS Secretaria da Fazenda do Estado do Rio Grande do Sul (Secretariat
of Finance of the State of Rio Grande do Sul) [Brazil]
SRI Servicio de Rentas Internas (Internal Revenue Service) [Ecuador]

xxv
 SUNAT Superintendencia Nacional de Aduanas y de Administración
Tributaria (National Superintendency of Customs and Tax
Administration) [Peru]
TaxMIS tax management information system
UCD Unidad de Científicos de Datos (Data Scientists Unit) [Colombia]
VAT value added tax

All dollar amounts are US dollars unless otherwise specified.

xxvi Data for Better Governance

CHAPTER 1

The Case for

Government Analytics in
Latin America and the
Caribbean

INTRODUCTION
Countries in Latin America and the Caribbean face challenges on several fronts,
demanding more efficient and responsive governments that engender confidence
among their citizens.

Governments in the region are constrained in making necessary investments

as a result of increased deficits and debt arising from the COVID-19 pandemic,
pointing to the need for these governments to become more effective and efficient
(IMF 2021; Maloney et al. 2024). They have already demonstrated that they can
achieve this: for example, relative to 30 years ago, increases in the professionalization
and efficiency of ministries of finance and central banks have burnished the region’s
global reputation for macroeconomic stability. Other core functions of government
need to follow. Estimates point to waste of about 4 percent of GDP, or 17 ­percent of
all public spending, on account of inadequate procurement practices, misdirected
transfers, and poorly designed human resources management (Izquierdo, Pessino,
and Vuletin 2018; World Bank 2022b). In some countries in the region, the estimated
waste is even higher.

Recent social unrest in the region also reveals deep popular dissatisfaction with
state performance, in areas ranging from pensions to education to antipoverty
policies. Citizens’ distrust in the government and public institutions, as well as
their perceptions of embedded corruption in the state, make them reluctant to
vote in favor of large infrastructure projects or contribute with taxes. Doubts about
what government institutions offer lead many small entrepreneurs to remain
unregistered and informal. Citizens in Latin America and the Caribbean don’t just

A reproducibility package is available for this book in the Reproducible Research Repository at
https://reproducibility.worldbank.org/index.php/catalog/209.

1
 have low “tax morale”: they have low faith in governance in general (Keefer and
Scartascini 2022).

These demands for effective governance will only become more acute. To adapt to
climate change and take advantage of opportunities offered by the green transition and
the realignment of the global economy after the COVID-19 pandemic, governments
must be able to mobilize more resources, plan strategically, and implement plans
efficiently. Demands for a more activist state—one that stimulates innovation and
midwifes structural change—have arisen after the disappointing growth rates of the
last two decades, which have called into question the tenets of the neoliberal growth
model (refer, for example, to Hausmann and Rodrik 2006; Mazzucato and Rodrik
2023). However, many countries have moved to minimize state intervention in the
economy, given its poor track record in conceiving and implementing industrial
policies in previous decades. Without improvements in state capacity, monitoring, and
transparency, governments that engage in a new round of experimental policies will
only wind up disappointed again.

It is therefore paramount that governments in Latin America and the Caribbean

improve their core functions, including human resources management,
procurement, taxation, public financial management, and more. Broad-based
improvements in government functioning are fundamental to improving the region’s
development prospects and must be at the top of national policy agendas. To provide
policy tools and support for these agendas, the World Bank produced The Government
Analytics Handbook (Rogger and Schuster 2023), which consolidates cutting-edge
academic and policy evidence on using data analytics to measure government
performance and on designing policies to improve it (box 1.1). One lesson from the
Handbook is that although progress in some areas will remain challenging, there are
also relatively quick wins. For instance, improving the performance of the worst-
performing procurement officers could reduce a government’s annual procurement
expenditures by 12 ­percent (Best, Hjort, and Szakonyi 2023).

This report aims to provide a road map so governments in Latin America and the
Caribbean can bring these tools to the front line of policy and use them to enhance
their functioning. Governments in the region are well placed to use data to improve
their core functions, owing to successive investments in technologies that are
foundational for analytics. The region is a global pioneer in establishing management
information systems (MISs), which record and store large amounts of data on specific
aspects of government functioning—such as taxation, public finance, and human
resources management—and make them available. These administrative data are
large reservoirs of underutilized evidence and information. By analyzing these data,
governments can assess the effectiveness and efficiency of public administration and
understand where it can be strengthened. However, because of weaknesses in their
analog complements (for instance, data analysts and data access protocols), these data
are rarely used for evidence-based policy making (World Bank 2016, 2023). This means
governments in Latin America and the Caribbean can do much more with the data
they already have to improve their functioning, address development challenges, and
build citizens’ trust.

2 Data for Better Governance

 Box 1.1 The Government Analytics Handbook

This report builds on The Government Analytics Handbook (Rogger and

Schuster 2023), a comprehensive guide to using government data to strengthen
public administration. Governments across the world make thousands of
personnel management decisions, procure millions of goods and services, and
execute billions of processes each day. When these functions are recorded in
management information systems, governments are data rich. Yet there is limited
systematic use of the resulting data to strengthen public administration. This
means that governments are missing out on opportunities to save money on
procurement expenditures, recruit better talent into government, and identify
sources of corruption.

The Handbook provides comprehensive and in-depth guidance to help

governments develop a data-informed approach to government analytics. It brings
together evidence and insights from various countries and practitioners into a
one-of-a-kind resource covering a range of topics, including survey design, the
strategic use of human resources management data, and ethical considerations
for data use. It also considers a variety of data sources, including different types of
administrative and survey data. In this way, the Handbook presents frontier evidence
and practitioner insights on how to leverage data to make governments work better.

This report applies lessons from the Handbook to Latin America and the
Caribbean and draws on relevant examples from the region to help policy makers
and senior public officials use government analytics to enhance the effectiveness
and efficiency of their public administrations. It focuses on administrative data,
because these data are accessible at low cost to most government organizations
(refer to box 1.2).

Many governments in the region are already capitalizing on the opportunity to use
administrative data to inform policy making. For example, the government in one
Brazilian state has reduced procurement prices by 13 ­percent by using tax data to
calculate market reference prices for procurement agents. Chile has saved hundreds of
millions of dollars by using health data to design a program to reduce missed medical
appointments and improve health behaviors among patients with chronic conditions,
and Colombia has saved the lives of thousands of newborns and expectant mothers by
tracking health risks. Ecuador and Peru have collected millions of dollars in additional
tax revenue by using tax data to detect evasion and better allocate resources for
enforcement. And Guatemala has reduced the dropout rate for students entering lower
secondary school by 9 ­percent by using education data to identify and support at-risk
students. Through these analytical initiatives, governments have drawn on administrative
data they already have to enhance policy design, strengthen service delivery, and reduce
inefficiency and waste.1 This report aims to support initiatives like these by offering
guidance on how to build and sustain a robust government analytics ecosystem that
facilitates the use of data to improve the functioning of public administration.

THE CASE FOR GOVERNMENT ANALYTICS IN LAC 3

 KEY MESSAGES
The first key message of this report is that governments in Latin America and
the Caribbean can enhance the design and implementation of public policy by
using government analytics more broadly and intensively than they currently
do. Governments’ use of administrative data for analytics varies across different
functions and policy areas. The report shows that the revenue side of government
operations often makes more extensive use of data to inform its actions. For
instance, governments in the region often use advanced analytics of tax data to
improve tax compliance. On the expenditure side, however—such as in human
resources management and procurement—there is significant room for growth.
For instance, governments in the region rely primarily on descriptive analytics of
personnel headcount to manage public employees, and they rarely use advanced
analytics to forecast public administration’s wage bill or to improve the efficiency of
public procurement and secure competitive prices. Likewise, there are significant
opportunities to develop the use of advanced analytics for service delivery—for
example, to forecast student enrollment.

The second key message is that governments must foster two essential enabling
conditions to make their administrative data useful for policy making. The first
is high-quality, accessible, and integrated data infrastructure. Improving data
quality controls and accessibility should be a government-wide effort, enabling
economies of scale. Likewise, ensuring that different information systems across
the government are interoperable and connecting their data can multiply the
potential insights and applications of government analytics. The second is analytical
capabilities, such as funding opportunities, dedicated analytical units, and career
tracks for data analysts. Government-wide strategies and institutionalized funding
streams enable a more harmonized, cohesive approach to developing analytical
initiatives, which can facilitate cross-pollination, learning, and collaboration across
government organizations.

The final key message is that governments must integrate data analytics into their
decision-making processes to use it effectively for policy making and reform.
Government analytics is not a solution in itself: it is a tool to provide policy makers
with accurate and relevant evidence to guide key decisions, identify gaps and
weaknesses, and refine the design and implementation of policy interventions
(Rogger and Schuster 2023; World Bank 2021). Analytics cannot replace decision-
makers, design policies, or substitute for the experience and knowledge of policy
makers and public servants. To be useful, analytics must be problem oriented, demand
driven, and responsive to the specific needs of decision-makers, because strengthening
the public sector requires more than data and measurement (Bridges and Woolcock
2023). In addition to creating the enabling conditions for analytics, governments must
foster a culture of evidence-based decision-making within public administration
to ensure that analytics delivers on its potential to enhance the design and
implementation of public policy.

4 Data for Better Governance

 Throughout this report, gold callout balloons will
be used to point out chapters in The Government
Analytics Handbook where you can learn more
about each topic the report discusses.

THE GOVERNMENT ANALYTICS ECOSYSTEM

Government analytics can help make public administration more efficient by meeting
the needs and objectives of policy makers and informing specific policies and critical
decisions. Governments can use data to support evidence-based decision-making in
three major ways, as outlined in the World Bank’s World Development Report 2021:
Data for Better Lives (World Bank 2021). First, they can analyze data to improve policy
making and service delivery by increasing access to government services and ensuring
that policies are grounded in the best available evidence. Second, they can use data
to help allocate scarce resources through targeting and monitoring. Finally, they can
publicize data to foster transparency and increase their own accountability for using
public resources effectively and efficiently (World Bank 2021). Government analytics
thus aligns with a broader World Bank agenda on the use of data and analytics to
promote development.2

Chapter 2 of this report presents a conceptual framework that highlights the different
uses of government analytics and how to strengthen its enabling conditions. As the
World Bank’s World Development Report 2016: Digital Dividends (World Bank 2016)
highlights, digital technologies require analog complements to achieve their potential.
Similarly, analytics requires two essential enabling conditions to be an effective policy-
making tool: the infrastructure to produce and use high-quality data and the analytical
capabilities to leverage these data. This report uses the metaphor of an ecosystem
to refer to the interdependence of government analytics applications and their
enabling conditions. Just as diverse organisms require an environment that supports
them, government analytics is enabled by robust data infrastructure and analytical
capabilities.

The conceptual framework first considers how government analytics provides evidence
to address policy-making challenges and meet decision-makers’ needs. Governments
face a wide variety of policy-making challenges that demand diverse evidence and
thus diverse analytical products. There are many types of analytical products, each one
drawing on different types of administrative data to respond to different policy needs.
There can be analytics on taxation or procurement, on human resources or public debt
data. For this reason, the conceptual framework describes in detail the different types of
analytics governments undertake, how the insights generated by analytics are applied,
and how analytics is produced and distributed among public servants and citizens.

THE CASE FOR GOVERNMENT ANALYTICS IN LAC 5

 Even within a single government function, different challenges might demand different
types of analytics. For instance, a government struggling with procurement delays
might analyze procurement data to understand the extent of these delays and what
causes them. On the other hand, a government considering establishing a centralized
payment system might analyze procurement data to predict the effects of the new
system. Public administration faces a diverse set of challenges, so the solutions that
government analytics provides to policy makers should be diverse as well.

The conceptual framework then describes two enabling conditions that help
government analytics yield solutions to policy-making challenges. The first is data
infrastructure. Government analytics depends on high-quality data supported by
robust data governance and access protocols to ensure they are secure, accurate, and
widely used. If the quality of data infrastructure deteriorates, so do the quality and
diversity of analytics. The second is analytical capabilities, at both the individual and
organizational levels: dedicated career tracks and training for data analysts, funding
opportunities, dedicated analytics units, and a comprehensive analytics strategy.
These enabling conditions strengthen each other: improvements to data infrastructure
provide incentives for investments in analytical capabilities by making administrative
data more accessible and easier to use, and stronger analytical capabilities provide
incentives for investments in improving data infrastructure, because staff and resources
are available to use those data for analytics. Furthermore, analytical applications can
help strengthen these enabling conditions: for example, by identifying data gaps and
motivating public servants to develop their skills through training.

The conceptual framework also structures the analysis of qualitative and quantitative
evidence on government analytics—along with the policy recommendations derived
from this evidence—presented in the rest of the report. Chapter 3 presents 12 case
studies of analytical initiatives from across Latin America and the Caribbean, covering
a wide range of countries and government functions. Then, chapter 4 presents the
results of an original survey administered in 20 countries to public servants who are
experts in digital government, administrative data, and analytics. Finally, chapter 5
concludes with policy recommendations to help governments at all levels of analytics
maturity apply the conceptual framework to develop government analytics and meet
the policy-making challenges they face.

REGIONAL OPPORTUNITIES AND CHALLENGES FOR

GOVERNMENT ANALYTICS
Governments in Latin America and the Caribbean are well positioned to develop
a government analytics ecosystem to support evidence-based policy making and
improve their functioning. Countries in the region have invested significant resources
in building MISs: integrated data frameworks that collect, store, and manage
administrative data across government functions (World Bank 2022a). This report
focuses on six MISs that are key to government functioning:

6 Data for Better Governance

● Human resources management information system (HRMIS)
● Procurement management information system (e-Procurement)
● Public financial management information system (PFMIS)
● Tax management information system (TaxMIS)
● Education management information system (EdMIS)
● Health management information system (HealthMIS).
Together, these MISs cover government functions that account for an average of
79 ­percent of total revenue and grants and at least 40 percent of total spending at the
central government level in Latin America and the Caribbean (figure 1.1). This makes
them a vital source of data for governments.3 (Box 1.2 explains why this report focuses
on administrative data, and appendix B describes each MIS in more detail.)

Since the 1980s, governments in Latin America and the Caribbean have made
substantial progress in implementing information systems to digitalize, automate, and
simplify core government functions. These investments have significantly expanded
the availability of digital administrative data across the essential government functions
considered in this report (panel a of figure 1.2).4 By 2022, every country in the region
had both a PFMIS and a TaxMIS. Additionally, 91 ­percent of the countries had an
HRMIS, and 84 ­percent had an e-Procurement system (World Bank 2022a). The
extensive adoption of digital MISs makes the region a global leader in information
system coverage (panel b of figure 1.2).

FIGURE 1.1 Central Government Revenue and Spending Composition

in Latin America and the Caribbean, 2010–22

a. Revenue b. Spending
Average % of GDP Average % of GDP
30 30 27.0
23.8 23.8 24.8
21.4 21.5 4.9
4.5 4.2
20 5.6 20
4.6 4.3
8.1 9.0
7.2

10 10 2.8 3.2 3.6

16.8 17.2 18.2
6.5 6.8 6.9
2.8 2.5 2.6
0 0
2010–15 2016–20 2021–22 2010–15 2016–20 2021–22
Nontax revenue and grants Capital expenditures
Tax revenue Subsidies and other current transfers
Total revenue and grants Purchases of goods and services
Wages and salaries
Other current and capital expenditures
Total expenditure and lending
minus repayments

Source: Original figure for this publication, based on data from CEPALSTAT
(https://statistics.cepal.org/portal/cepalstat/index.html?lang=en).

THE CASE FOR GOVERNMENT ANALYTICS IN LAC 7

 Box 1.2 Why Administrative Data?

This report focuses on government analytics using administrative data: information routinely
collected and stored by government organizations in management information systems to
facilitate program administration and public service delivery. For instance, administrative
data include government records of births and deaths, student enrollment, procurement
contract values, tax revenues, and the wages of the workforce, among many other records
of how public administration translates policy into practice each day.

Administrative data typically offer more frequent and more geographically comprehensive
insights into individuals and organizations than are possible with survey data. Because
administrative data are collected for administrative purposes, these data are also often less
costly than other types of data (Statistics Canada, n.d.; UK Office for Statistics Regulation,
n.d.; US Census Bureau, n.d.). Repurposing administrative data for analytics turns them into a
valuable resource for improving the design, implementation, management, and assessment
of programs, thereby improving government functioning.

Administrative data can also be enhanced and complemented by survey data, which can
offer insights into challenges not captured by routinely collected data. However, unlike
administrative data, survey data are not readily available to government agencies. This
report focuses on government analytics using administrative data, because these data are a
powerful but underutilized resource and are accessible to most government organizations.

Administrative data can also improve the operations of government organizations involved in
survey data collection, such as national statistical offices (NSOs), creating valuable synergies
for the public sector (Rivas and Crowley 2018). Using administrative data from sectors such
as tax, health, and education, NSOs can significantly improve the compilation and accuracy
of important national statistics. Tax data can support the creation of indexes of economic
activity. Similarly, health and education records can be used to verify expenditure data.

NSOs’ expertise in data management and governance can likewise add substantial value
to administrative data. NSOs can assist in improving data quality through quality controls
and ensuring data standardization across the public sector, which in turn facilitates
interoperability between data systems. Collaborations between organizations responsible
for administrative data and those responsible for survey data thus have great potential to
enhance information quality and strengthen government analytics.

The transition from analog to digital MISs creates an opportunity for governments in
Latin America and the Caribbean to leverage existing administrative data to improve
policy design and implementation. Digital MISs generate an unprecedented amount
of data about how public administration works, facilitating measurement at a granular
level and in real time. Governments in the region have also built international networks
to support the development and scale-up of analytical solutions, such as the Inter-
American Network of Digital Government. Investments in government technology,
like citizen portals and open data, have strengthened core government systems in the
region, enabling the region to make significant progress toward a more data-driven
public sector (World Bank 2023).

8 Data for Better Governance

Despite high MIS coverage, governments in Latin America and the Caribbean
must address institutional constraints to fully leverage their administrative data
for government analytics. Governments in the region still contend with partial
digitalization and fragmentation of their MISs and poor data governance. The
challenges of implementing MIS reforms have resulted in a mixture of tools and
technologies with varying degrees of effectiveness, leading in turn to information silos
and inconsistent data standards. As a result, interoperability among different systems
is limited, the quality of data available for analytics is compromised, and the available
data do not necessarily align with strategic priorities and policy needs. Governments
also face challenges in recruiting and retaining the skilled analysts needed to analyze
data and produce useful analytical applications.

FIGURE 1.2 Coverage of Key Management Information Systems in Latin America and
the Caribbean and Worldwide

a. Latin America and the Caribbean, 1984–2022

% of countries
100 100
100
90 91
84
80
70
60
50
40
30
20
10
0
84
86

88
90

92

94
96

98
00

02

04
06

08

10

12

14
16

18
20

22
20

20
20

20
20
19

19
19

20
19

19

20
19

19

20
20

20
19

20
20

PFMIS TaxMIS HRMIS e-Procurement

b. Worldwide, 2022

PFMIS

TaxMIS

HRMIS

e-Procurement

0 20 40 60 80 100
% of countries
SAR ECA LAC MNA AFR EAP
Source: World Bank 2022a.
Note: AFR = Africa; EAP = East Asia and Pacific; ECA = Europe and Central Asia; e-Procurement = procurement
management information system; HRMIS = human resources management information system; LAC = Latin
America and the Caribbean; MNA = Middle East and North Africa; PFMIS = public financial management
information system; SAR = South Asia; TaxMIS = tax management information system.

THE CASE FOR GOVERNMENT ANALYTICS IN LAC 9

 A REGIONAL ASSESSMENT AND POLICY
RECOMMENDATIONS TO STRENGTHEN GOVERNMENT
ANALYTICS
This report presents evidence and practical guidance for governments in Latin
America and the Caribbean on using their administrative data to improve government
functioning. Chapter 3 offers a series of qualitative case studies that provide firsthand
accounts of the opportunities and challenges governments in the region have
encountered while repurposing administrative data to solve governance challenges
through analytics, and the results they have achieved. Table 1.1 summarizes each
case study. Then, chapter 4 presents quantitative evidence from an original survey
conducted in 20 countries in the region. Public servants in each country who are
experts in specific information systems and core government functions responded
to questionnaires on the country’s government analytics ecosystem. By presenting a
variety of analytical experiences in the case studies and survey, this report can help
public servants learn from one another’s challenges and success stories.

But this report is more than an assessment: it is a practical guide for governments
that want to make data analytics a larger part of their policy-making tool kit. Just as a
gardening guide explains what environment various plants need to thrive, this report
offers recommendations for creating a resilient ecosystem to support diverse analytics
applications that can meet a wide variety of policy-making needs. Chapter 5 offers a
set of policy recommendations targeting each element of the conceptual framework,
based on the findings in the regional assessment. The recommendations are tailored to
different levels of information system and analytics maturity so that all governments
can find practical next steps for developing analytics for evidence-based policy making.
Table 1.2 summarizes these key findings and policy recommendations.

To put these recommendations into practice, governments must be aware of broader

political economy challenges. The political context that shapes policy making and
the prevailing culture within public administration have a significant impact on the
extent to which policies and decisions are informed by evidence. For instance, leaders
in public administration may view the accountability promoted by data analytics with
discomfort, perceiving it as a threat to their autonomy. Leaders may also make their
personal agendas a priority, especially in environments in which electoral competition
does not provide incentives for evidence-based policy change. Even when leaders
want to make evidence-based policy decisions, the political process might demand
quick decisions, without adequate time to gather data, conduct analytics, and produce
evidence. Government analytics alone cannot address these constraints.

Using government analytics to improve government functioning, address development

challenges, and ultimately build citizens’ trust thus requires going beyond the enabling
conditions outlined in this report. It also demands continuous support from policy
makers and a cultural shift toward embracing evidence-based policy making. This
report aims to guide policy makers interested in analytics by offering practical

10 Data for Better Governance

guidance on how to support it institutionally, as well as concrete examples of analytical
applications that have successfully enhanced policy design and implementation. The
hope is that it offers evidence and inspiration to public servants and policy makers
who want to use data to make government work better—in their own countries and for
Latin America and the Caribbean as a whole.

TABLE 1.1 Summary of Case Studies

Country Case study Summary

Chile 3.1: Using The Ministry of Health of Chile analyzed HealthMIS data to understand
Descriptive which patients were missing medical appointments—at a cost of
Analytics to Reduce $180 million annually—and to design a program to reduce missed
Missed Medical appointments and increase primary care use. Implementing a reminder
Appointments system and remedying gaps in administrative records decreased
missed appointments by 6 ­percent and increased the use of primary
care services by about 10 ­percent among patients with diabetes and
hypertension. This case study demonstrates the impact relatively simple
descriptive analytics can have when it is based on rich, comprehensive
data, as well as the importance of data quality and how analytical
projects can help implement quality controls and improve administrative
records.
Colombia 3.2: Using The Data Scientists Unit of the National Planning Department of
Diagnostic Colombia analyzed HRMIS data to study the question of whether
Analytics of Human there is a gender pay gap among some public sector workers. It
Resources Data to found a gender pay gap of approximately 6 ­percent and informed the
Explore the Gender authorities that oversee gender policies of gaps in compliance with
Pay Gap Colombia’s gender antidiscrimination law. This case study demonstrates
that good policy evaluation produces evidence describing not just what
is happening but also why it is happening, as well as the critical role a
dedicated analytics unit can play in generating that evidence.
Paraguay 3.3: Data Analytics Leveraging open procurement data, reAcción Paraguay built an open
for Citizen platform to monitor the allocation of millions of dollars of public funds
Accountability: The distributed through a school infrastructure program. The platform’s
Case of reAcción insights prompted a response from the government and improved
the targeting of resources to schools with the greatest need by
400 ­percent. This case study demonstrates how open and accessible
government data can empower civic mobilization, enable civil society
to uncover mismanagement of public funds, and improve government
transparency and accountability and the allocation of public resources.
It also demonstrates how the impact of government analytics is
amplified when data from different sources are connected.
Uruguay 3.4: Building The National Civil Service Office of Uruguay is developing a human
an Integrated resources management platform that integrates administrative data
Human Resources from multiple government organizations. The platform will transform
Management public sector recruitment and management and create the enabling
Platform for conditions for future analytics. This case study demonstrates the
Strategic Workforce impact government analytics can have when data are connected; the
Planning powerful applications of descriptive analytics; and the importance of
monitoring the composition of the public sector workforce, the skills and
competencies within it, and future human capital needs.

(continues on next page)

THE CASE FOR GOVERNMENT ANALYTICS IN LAC 11

TABLE 1.1 Summary of Case Studies (continued)

Country Case study Summary

Ecuador 3.5: Analyzing Tax The Internal Revenue Service of Ecuador, in collaboration with a
Data to Increase team of researchers, evaluated the integration of third-party data
Corporate Tax with the country’s TaxMIS to combat tax evasion. The evaluation
Collection showed that using third-party data in tax collection could generate
hundreds of millions of dollars in revenue if complemented by effective
enforcement. This case study demonstrates how administrative data
can be used for rigorous policy evaluation, the impact of connecting
multiple administrative records and nonadministrative databases, and
the benefits for public administration of partnerships with external
researchers and data analysts.
Guatemala 3.6: Using Predictive In Guatemala, approximately 40 ­percent of sixth graders drop out
Analytics to Prevent of school before reaching ninth grade. To tackle this issue, the
School Dropouts Ministry of Education, in collaboration with the World Bank, piloted
an early warning system to identify students at risk of dropping out
that leveraged predictive analytics based on recent improvements
to student-level data. This innovative program achieved a 9 ­percent
decrease in the dropout rate during the transition from primary to
lower secondary school. This case study demonstrates how analytics
can support policy evaluation and design and test the effectiveness of
policies before scale-up, as well as the importance of high-quality data
to enable analytics to have an impact.
Brazil 3.7: Using Tax The government of the state of Rio Grande do Sul in Brazil used tax
Data to Boost data to enhance the efficiency of procurement processes, saving
Procurement 4 ­percent on its annual total expenditure for specific products. Since
Efficiency 2008, Brazil has required registered firms that are subject to state tax
on the circulation of goods and services to issue digital invoices for their
transactions. These data were used to calculate market reference prices
for setting tendering parameters. When procurement teams had access
to these analytics, purchase prices were reduced by 13.2 ­percent. This
case study shows how government data can be used for analytics
across functions and how simple statistics, such as the calculation of
reference prices, can be leveraged to improve procurement.
Colombia 3.8: Improving The municipal government of Bogotá, Colombia, created a centralized
Government analytics unit, Ágata, to generate insights across management
Efficiency through information systems and government functions to meet the needs of
a Centralized the city’s stakeholders. Among many projects, Ágata has leveraged
Analytics Unit electronic health records and predictive analytics to identify real-time
health risks for pregnant women and newborns, saving thousands of
lives. It has also integrated public financial management data across
organizations, enabling them to visualize and avoid duplicating the
services they provide to citizens, enhancing resource allocation. This
case study demonstrates the value that a centralized analytics unit can
bring to public administration.

(continues on next page)

12 Data for Better Governance

TABLE 1.1 Summary of Case Studies (continued)

Country Case study Summary

Peru 3.9: Establishing In Peru, tax compliance is a challenge for tax authorities. To address
a Decentralized this challenge, the National Superintendency of Customs and Tax
Analytics Unit Administration created its own analytics unit. This unit has undertaken
to Improve the numerous analytical projects to strengthen tax and customs control
Efficiency of Tax and compliance, yielding millions of dollars in additional revenue and
Collection serving as a key element of the superintendency’s strategy to improve
the efficiency of tax administration. This case study demonstrates how
decentralized analytics units can develop sophisticated predictive
analytics, leading to significant improvements in tax revenue collection.
Multiple 3.10: Datathons as a Many governments in Latin America and the Caribbean have
countries Tool for Promoting established platforms or hosted events for sharing best practices,
Data Analytics tools, and success stories related to analytics. At these “datathons,”
public servants and members of the community are given access
to government data, learn about a relevant policy question, and
collaborate to produce analytics that can be used to inform policy
making. This case study demonstrates that governments benefit from
engaging with data analysts inside and outside the public sector
to gather ideas for analytical products and strengthen recruitment
channels.
Brazil 3.11: The Power of In 2022, the World Bank piloted the Governance Risk Assessment
Data Integration System (GRAS) in Brazil. GRAS uses algorithms to identify patterns
to Strengthen the associated with public procurement fraud and corruption, targeting
Integrity of Public public suppliers, contracting agencies, and individual actors. A pillar
Procurement of GRAS is its robust data integration framework, which aggregates,
links, and analyzes large volumes of public procurement data, including
detailed microdata on firms and individuals. During the pilot, GRAS
identified about 800 firms that were awarded contracts despite legal
sanctions and assisted federal police in uncovering networks of
shell companies and money-laundering operations. This case study
demonstrates how public sector corruption can be detected and
reduced by integrating government data, applying descriptive and
predictive analytics, and triangulating data.
Uruguay 3.12: Supporting In 2019, the World Bank conducted a review of Uruguay’s public
the Implementation procurement data that identified framework agreements as a strategy
of Framework for improving the performance of the procurement system. During this
Agreements review’s implementation phase, analytics was useful for identifying
through Data items that could be purchased through framework agreements. This
Analytics case study demonstrates the impact of descriptive and diagnostic
analytics to identify suitable interventions and support their design
during the implementation stage of a program.

Source: Original table for this publication.

Note: HealthMIS = health management information system; HRMIS = human resources management information
system; TaxMIS = tax management information system.

THE CASE FOR GOVERNMENT ANALYTICS IN LAC 13

TABLE 1.2 Key Findings and Policy Recommendations for Strengthening
Government Analytics

a.  ow can governments move toward a more strategic approach to

H
government analytics to generate evidence for decision-making?

Key findings Policy recommendations

● When government analytics is fit for ● Strengthen the culture for evidence-based policy making and
a particular purpose and tailored the demand for analytics within public administration. Review
to specific policy needs, it can how data are used in policy making. Define priorities for reform
enhance decision-making and help and the evidence needed to guide reforms. Build demand for
achieve development outcomes. analytics through workshops, conferences, and data-sharing
● Governments in Latin America programs. Foster interagency partnerships and implement data
and the Caribbean mainly use governance and management strategies.
administrative data for descriptive ● Define the strategic use of government data. Review how
analytics; diagnostic and predictive administrative data analytics can inform decision-making in
analytics remain underutilized. specific policy areas. Strategize regarding broader uses of
● Governments in the region data and consider how connected data can benefit multiple
mainly use analytical products for organizations.
monitoring and accountability and, ● Streamline descriptive analytics and develop diagnostic and
to a lesser extent, policy evaluation predictive analytics. Create the conditions needed to use
and design. administrative data effectively for monitoring and accountability
● Governments in the region mainly by defining a framework, establishing a reporting process, and
use management information creating a dissemination tool. Move toward diagnostic and
systems (MISs) for operational and predictive analytics to inform policy evaluation and design.
transactional purposes, missing
opportunities to use administrative
data and advanced analytics to
improve decision-making.

b. How can governments build analytical capabilities for individuals and

organizations?

Key findings Policy recommendations

● Although governments in Latin ● Assess existing analytical capabilities. Regularly assess public
America and the Caribbean sector analytical capabilities to identify capacity-building needs.
acknowledge the importance of Develop plans to align workforce skills with those needed for
analytical skills, they lack structured strategic government analytics.
career tracks for data analysts and ● Build analytical capabilities strategically. Offer data analytics
proficiency evaluations for public training tailored to the specific needs of decision-makers,
servants. analysts, and subject area experts. Develop a framework
● Funding for the development of connecting skills development to practical applications.
analytical projects is limited in the ● Establish a dedicated career track for data analysts to attract
region. and retain talent.
● Creating dedicated analytics units ● Institutionalize government analytics. Establish dedicated
and collaborating with academic or analytics units and secure a budget for supporting analytical
international organizations can be capabilities and products.
effective strategies for addressing ● Develop external collaborations. Forge external collaborations
capacity-building challenges and with academia, nonprofits, and multilateral organizations to
streamlining the use of analytics in address budget constraints and skills shortages and to facilitate
decision-making. the coproduction of analytical products through effective data-
sharing frameworks.

(continues on next page)

14 Data for Better Governance

TABLE 1.2 Key Findings and Policy Recommendations for Strengthening
Government Analytics (continued)

c. How can governments strengthen data infrastructure?

Key findings Policy recommendations

● Despite progress in digitalization, ● Assess the quality, completeness, and timeliness of
many MISs in Latin America and the government data and define steps for improvement. Regularly
Caribbean are not comprehensive assess and update MISs and examine broader data ecosystems
enough or are inadequately to prevent silos and strengthen MIS effectiveness.
updated, with considerable ● Establish regular, systematic, and automated data quality
variation. controls. Embed data quality controls within MIS frameworks
● Measures for controlling data to ensure accuracy and reliability. Establish a dedicated team
quality are not widely implemented responsible for this.
in MISs in the region, and the ● Establish a data inventory. A data inventory should include
manual, sporadic nature of data available data types, storage locations, and field definitions.
cleaning poses challenges for the This simplifies project development and ensures proper data
accuracy, reliability, promptness, life cycle management.
and replicability of analytics. ● Connect government data. Establish protocols for data sharing
● MIS interoperability is limited, and between government agencies to connect and integrate
data sharing between government administrative data effectively.
agencies is ad hoc, preventing ● Plan for MIS maintenance and updates. Create the conditions
governments from fully unlocking for MIS updates to continually leverage the newest technological
the potential of digitalized MISs for advancements.
analytical projects.

Source: Original table for this publication.

NOTES
1. Refer to table 1.1 and chapter 3 for detailed information on these and other analytical initiatives by
governments in the region.
2. For instance, through the Global Evaluation Initiative, the World Bank is promoting global efforts to
strengthen country systems for monitoring, evaluation, and use of evidence for decision-making. This
includes the establishment of Regional Centers for Learning on Evaluation and Results in countries in
Latin America and the Caribbean such as Brazil and Chile (GEI, n.d.-a, n.d.-b).
3. The report recognizes that other MISs are important for the effective functioning of public
administration. However, it focuses on these six types of MISs because they are widely adopted in the
region and represent the core functions of the executive branch of the state.
4. The World Bank has been a crucial partner for countries in Latin America and the Caribbean,
supporting the design and implementation of initiatives to adopt or modernize MISs. These efforts
have been particularly focused on public financial management information systems, because of the
importance of modern systems for policy implementation and service delivery. By 2011, for example,
45 ­percent of the World Bank’s projects in Latin America and the Caribbean included key investments
in information and communication technology for financial management (Dener, Watkins,
and Dorotinsky 2011; Pimenta and Pessoa 2016). These efforts reflect the World Bank’s ongoing
commitment to support governments in fostering data-driven decision-making through their MISs
and in leveraging innovations in information and communication technology to improve policy
implementation and service delivery.

THE CASE FOR GOVERNMENT ANALYTICS IN LAC 15

 REFERENCES
Best, Michael Carlos, Jonas Hjort, and David Szakonyi. 2023. “Individuals and Organizations as
Sources of State Effectiveness.” American Economic Review 113 (8): 2121–67. https://doi.org/10.1257​
/aer.20191598.
Bridges, Kate, and Michael Woolcock. 2023. “Measuring What Matters: Principles for a Balanced Data
Suite That Prioritizes Problem Solving and Learning.” In The Government Analytics Handbook:
Leveraging Data to Strengthen Public Administration, edited by Daniel Rogger and Christian Schuster,
chap. 4. Washington, DC: World Bank. https://doi.org/10.1596/978-1-4648-1957-5.
Dener, Cem, Joanna Alexandra Watkins, and William Leslie Dorotinsky. 2011. Financial Management
Information Systems: 25 Years of World Bank Experience on What Works and What Doesn’t.
Washington, DC: World Bank. https://documents.worldbank.org/curated/en/485641468139212120​
/Financial-management​-information-systems-25-years-of​-World-Bank-experience-on-what-works​
-and-what-doesnt.
GEI (Global Evaluation Initiative). n.d.-a. “The CLEAR Network.” GEI, World Bank. Accessed May 8,
2024. https://www​.globalevaluationinitiative.org/clear-network.
GEI (Global Evaluation Initiative). n.d.-b. “Who We Are.” GEI, World Bank. Accessed May 8, 2024.
https://www​.globalevaluationinitiative.org/who-we-are.
Hausmann, Ricardo, and Dani Rodrik. 2006. “Doomed to Choose: Industrial Policy as Predicament.”
Working paper, Center for International Development, Harvard University, Cambridge, MA.
https://drodrik.scholar​.harvard.edu/publications/doomed-choose-industrial-policy-predicament.
IMF (International Monetary Fund). 2021. Regional Economic Outlook: Western Hemisphere; A Long and
Winding Road to Recovery. World Economic and Financial Surveys. Washington, DC: International
Monetary Fund. https://www.imf.org/en/Publications/REO/WH/Issues/2021/10/21/Regional​
-Economic​-Outlook-October-2021-Western-Hemisphere.
Izquierdo, Alejandro, Carola Pessino, and Guillermo Vuletin, eds. 2018. Better Spending for Better Lives:
How Latin America and the Caribbean Can Do More with Less. Washington, DC: Inter-American
Development Bank. https://dx.doi.org/10.18235/0001217-es.
Keefer, Philip, and Carlos Scartascini, eds. 2022. Trust: The Key to Social Cohesion and Growth in Latin
America and the Caribbean. Washington, DC: Inter-American Development Bank. http://dx.doi.org​
/10.18235/0003911.
Maloney, William, Pablo Garriga, Marcela Meléndez, Raúl Morales, Charl Jooste, James Sampi, Jorge
Thompson Araujo, and Ekaterina Vostroknutova. 2024. Competition: The Missing Ingredient
for Growth? Latin America and the Caribbean Economic Review, April 2024. Washington, DC:
World Bank. http://hdl.handle.net​/10986/41230.
Mazzucato, Mariana, and Dani Rodrik. 2023. “Industrial Policy with Conditionalities: A Taxonomy and
Sample Cases.” Working Paper 2023/07, Institute for Innovation and Public Purpose, University
College London. https://www.ucl.ac.uk/bartlett/public-purpose/wp2023-07.
Pimenta, Carlos, and Mario Pessoa, eds. 2016. Public Financial Management in Latin America: The Key to
Efficiency and Transparency. Washington, DC: International Monetary Fund. https://doi.org/10.18235​
/0000083.
Rivas, Lisbeth, and Joe Crowley. 2018. “Using Administrative Data to Enhance Policymaking in
Developing Countries: Tax Data and the National Accounts.” IMF Working Paper 18/175, International
Monetary Fund, Washington, DC. https://www.imf.org/en/Publications/WP/Issues/2018/08/02/Using​
-Administrative-Data-to-Enhance-Policymaking-in-Developing-Countries-Tax-Data-and-the-46054.
Rogger, Daniel, and Christian Schuster, eds. 2023. The Government Analytics Handbook: Leveraging
Data to Strengthen Public Administration. Washington, DC: World Bank. https://doi.org/10.1596​
/978-1-4648-1957-5.
Statistics Canada. n.d. “Administrative Data.” Accessed March 15, 2024. https://www.statcan.gc.ca/en/our​
-data/where/administrative-data.

16 Data for Better Governance

UK Office for Statistics Regulation. n.d. “Administrative Data (Part 1).” Accessed March 15, 2024. https://
osr.statisticsauthority.gov.uk/guidance/administrative-data-and-official-statistics/quality-assurance​
-of-administrative-data-case-examples/administrative-data-part-1/.
US Census Bureau. n.d. “Administrative Data.” Accessed March 15, 2024. https://www.census.gov/topics​
/research/guidance/restricted-use-microdata/administrative-data.html.
World Bank. 2016. World Development Report 2016: Digital Dividends. Washington, DC: World Bank.
https://doi.org/10.1596/978-1-4648-0671-1.
World Bank. 2021. World Development Report 2021: Data for Better Lives. Washington, DC: World Bank.
https://www.worldbank.org/en/publication/wdr2021.
World Bank. 2022a. GovTech Maturity Index, 2022 Update: Trends in Public Sector Digital Transformation.
Equitable Growth, Finance and Institutions Insight—Governance. Washington, DC: World Bank.
https://hdl.handle.net/10986/38499.
World Bank. 2022b. New Approaches to Closing the Fiscal Gap. LAC Semiannual Update. Washington, DC:
World Bank. https://hdl.handle.net/10986/38093.
World Bank. 2023. Interoperability: Towards a Data-Driven Public Sector. Equitable Growth, Finance and
Institutions Insight—Governance. Washington, DC: World Bank. https://doi.org/10.1596/38520.

THE CASE FOR GOVERNMENT ANALYTICS IN LAC 17

CHAPTER 2

A Conceptual
Framework for
Government Analytics

INTRODUCTION
Governments in Latin America and the Caribbean can deepen their use of government
analytics to design and implement better public policy. To understand how to do
so, practitioners first need to examine the broader context that shapes analytical
initiatives. This chapter introduces a conceptual framework for assessing government
analytics that describes how data and analytics can be used to improve government
outcomes through evidence-based policy making and maps the enabling conditions for
government analytics (figure 2.1).

Countries and government organizations in Latin America and the Caribbean vary
considerably in the extent to which they analyze their administrative data, the types
of analytics they undertake and the methodologies they employ, and the ways they
apply analytics to address policy-making challenges. For this reason, the conceptual
framework presented in this chapter is designed to be generalizable to different
contexts. The conceptual framework informs this report’s description of analytical
initiatives in chapter 3, its empirical overview of analytics in the region in chapter 4,
and the policy recommendations in chapter 5.

By assessing government analytics holistically, policy makers in the region can reflect
on how they are using their administrative data and identify necessary steps to unlock
its full potential. To make the conceptual framework more concrete and intuitive,
this chapter points to real-world examples of analytical initiatives from the chapter 3
case studies. Online appendix C also includes a simulated example that puts readers
in the position of an analyst looking for insights in human resources data. The case
studies and the simulated example provide an accessible overview of government
analytics and its enabling conditions, but readers who are interested in the more
technical details of implementation are advised to consult The Government Analytics

A reproducibility package is available for this book in the Reproducible Research Repository at
https://reproducibility.worldbank.org/index.php/catalog/209.

19
 FIGURE 2.1 A Conceptual Framework for Government Analytics

Outcomes
Fiscal
sustainability
Effective public Trust in public
service delivery institutions

Applications

• Monitoring and accountability

• Government transparency
• Policy evaluation and design

Types Distribution

• Descriptive
• Consumption
• Diagnostic
• Production
• Predictive

Analytics

Quality Individual

Accessibility Organizational

Infrastructure Capabilities

Source: Original figure for this publication.

Handbook (Rogger and Schuster 2023), which provides in-depth analytical guidance
on specific government functions and information systems, such as human resources
management, procurement, and public financial management.

In designing the conceptual framework, the chapter draws on evidence from

research and policy reports, as well as survey instruments that measure governments’
digital readiness (for instance, the Organisation for Economic Co-operation and
Development’s Survey on Digital Government, the United Nations’ E-Government
Survey, the World Bank’s GovTech Maturity Index, and the Global Evaluation
Initiative’s Monitoring and Evaluation Systems Analysis tool). In this way, the chapter

20 Data for Better Governance

connects the conceptual and empirical analysis in this report to a wider policy agenda
on the use of government technology in public administration and on government
technology enablers and their analog complements, which serve as enabling factors for
digital technologies (World Bank 2016, 2020, 2021, 2022b). The conceptual framework
presented in this chapter makes government analytics concrete and actionable, offering
governments a road map for applying its concepts in practice.

Government analytics should be developed and applied strategically, in response to real

problems, demands, and needs. The first half of this chapter considers how analytics
operates: what types of analytics governments can undertake, how analytical insights
can be used to guide policy and decision-making, and how analytics is produced
and consumed within public administration. The second half lays out the enabling
conditions for producing and using analytics: the analytical capabilities and data
infrastructure that make analytics of administrative data possible and ensure it can be
used to address policy-making challenges across public administration. Creating these
conditions is essential to making analytics part of the decision-making culture in public
administration and deepening evidence-based policy making.

HOW DOES GOVERNMENT ANALYTICS IMPROVE

GOVERNMENT FUNCTIONING?
To understand how governments use data to improve their functioning, this chapter
first identifies the types of analytics that governments conduct, how governments apply
analytics in their work, and how analytics is distributed among stakeholders within
public administration and beyond.

What Types of Analytics Do Governments Conduct?

Government analytics can give

rise to many types of initiatives.
• Descriptive
Analytics can help public • Diagnostic
servants target a reminder • Predictive
program to reduce missed
medical appointments (case
study 3.1) or help procurement
officers do their work more effectively (case study 3.7). It might even ground the
creation of a “smart city” (case study 3.8). These initiatives range widely in their
complexity and goals, but data experts typically categorize data analytics into three
types: descriptive, diagnostic, and predictive (refer, for instance, to Cote 2021).1

Descriptive Analytics

Descriptive analytics is a foundational type of analytics. In the context of

government analytics, it refers to the use of data to describe the public sector

A Conceptual Framework for Government Analytics 21

 and illustrate broader trends. It might also summarize and present data to make analytical
insights accessible to decision-makers. In general, descriptive analytics answers questions
like the following:

● What happened? (In procurement, for instance: What did the government purchase?)
● When and how often did it happen? (For instance: How are government purchases
distributed throughout the year?)
● Where did it happen? (For instance: How are purchases distributed across regions?)
● What are the principal characteristics or features of the data? (For instance: What is
the average price paid for government purchases by product type?)
Descriptive analytics of administrative data involves the application of statistical
techniques, including disaggregated percentages and ratios, simple correlations
illustrating the relationship between variables, and cross-tabulations enabling the
comparison of variables. Data visualization tools can also be used to communicate the
results of the analysis to policy makers and the public through dashboards, graphs,
and maps.

Descriptive analytics is an essential first step that makes analytical insights and evidence
available to policy makers, laying the foundation for evidence-based policy making.
In Chile, for example, the Ministry of Health designed a program that reduced missed
medical appointments and increased primary care use by about 10 percent among its
target population by using descriptive analytics to better understand the characteristics
and needs of the patients who were missing appointments the most (case study 3.1).

Diagnostic Analytics

Diagnostic analytics builds on descriptive analytics by trying to understand not just

what is happening but why it is happening. Whereas descriptive analytics identifies
and illuminates trends in the public sector, diagnostic analytics determines the causes
behind these trends using statistical models. In general, diagnostic analytics answers
questions like the following:

● Why did this happen? (For instance: Why is competition low in public
procurement?)
● What factors influenced this result? (For instance: What influences competition in
public procurement?)
Diagnostic analytics builds on descriptive analytics by enabling public officials to
identify causal mechanisms that explain the trends they observe in data. For example,
in Colombia, the Data Scientists Unit of the National Planning Department analyzed
data from human resources management information systems (HRMISs) using a
combination of descriptive and diagnostic analytics to understand the causes of wage
differentials in the public sector and discovered a gender pay gap of about 6 percent
among temporary workers. This enabled the department to assess de facto compliance
with Colombia’s gender antidiscrimination law and inform the relevant national
authority so it could act on the findings (case study 3.2).

22 Data for Better Governance

Predictive Analytics

Predictive analytics is the most complex type of analytics. It uses historical data and
statistical modeling techniques, informed by an understanding of causal mechanisms,
to make predictions about what will happen in the future and how individuals or
organizations are likely to respond to actions and interventions. These predictions
enable governments to be proactive and apply a deeper understanding of future
consequences to present-day decisions. In general, predictive analytics answers
questions like the following:

● What is likely to happen in the future? (For instance: How much is the government
likely to spend in each sector in the future?)
● What is the probability that a specific event will occur? (For instance: How likely is
it that a contract will be renegotiated?)
Predictive analytics builds on descriptive and diagnostic analytics because it combines
a data-driven picture of the current state of the public sector (provided by descriptive
analytics) with a deep understanding of the causal mechanisms driving trends
(provided by diagnostic analytics) to make informed predictions.

For example, the federal government of Brazil and the World Bank used predictive
analytics to project the government wage bill under different policy scenarios.
These projections considered headcount growth, salary progression, and other
factors to simulate wage bill growth while modeling the effects of reducing
wage growth, hiring freezes, and other policies (figure 2.2). These projections
informed a redesign of the federal government’s wage policy, providing a clear
justification for wage adjustments to avert a fiscal crisis (Tavares, Ortega Nieto,
and Woodhouse 2023).

FIGURE 2.2 Cumulative Fiscal Savings in Three Wage Bill Policy

Scenarios, Brazil, 2019–30

R$, billions (2017 price levels)

40
35
30
25
20
15
10
5
0
19

20

21

22

23

24

25

26

27

28

29

30
20
20

20
20

20

20

20
20

20

20

20
20

Scenario A Scenario B Scenario C

Source: Tavares, Ortega Nieto, and Woodhouse 2023, based on Brazilian government
data from 2019.

A Conceptual Framework for Government Analytics 23

 For examples of different types of
analytics using administrative data, refer to
chapters 10–12, 14, and 15 of The Government
Analytics Handbook.

Classifying different analytical initiatives as descriptive, diagnostic, or predictive

helps practitioners better understand how each type of analytics complements the
others. For example, results from descriptive analytics can highlight the need for more
in-depth analysis through diagnostic or predictive analytics. Likewise, a well-designed
dashboard that helps users visualize and interpret trends in the data (descriptive
analytics) could be enhanced with features that help evaluate the possible outcomes
of different scenarios (predictive analytics). Governments require more sophisticated
capabilities and infrastructure to undertake more advanced analytics. Nevertheless,
descriptive analytics may be sufficient for questions that require straightforward
measurement to answer, and it also provides the high-quality foundation required for
more complex analytical tasks. For instance, descriptive analytics enabled the youth
anticorruption organization reAcción Paraguay to track government spending on
schools, increasing government accountability to citizens (case study 3.3).

Governments should recognize the types of analytics they can already produce
as well as identify next steps to build their capacity and infrastructure for more
complex analytics, without losing sight of the kinds of questions that will be useful
to policy makers. Each type of analytics serves the overarching goal of transforming
raw data into actionable information, each type of analytics depends on the others,
and each type of analytics can be applied to strengthen government functioning.

How Do Governments Apply Analytics?

After the types of analytics a government

• Monitoring and accountability
is undertaking have been assessed, • Government transparency
the focus can turn to how it applies • Policy evaluation and design
analytics to enhance policy design and
implementation. The report’s conceptual
framework identifies three general
applications of government analytics: to
enable monitoring and accountability, to
increase government transparency, and
to aid in policy evaluation and design.

Monitoring and Accountability

The first application of government analytics considered here is to enable monitoring

and accountability. Analytics can be used to report on the status quo within public
administration, monitor trends, track progress toward goals, and hold public officials

24 Data for Better Governance

accountable. For example, analytics can be used to design tools aimed at identifying
fraud and corruption. In procurement, this could mean implementing a system that
flags irregularities in bidding contracts (case study 3.11).2 As another example, in
Uruguay, the National Civil Service Office is developing an online human resources
management platform that will enable strategic workforce planning by making
information on the composition of the public sector workforce, the skills and
competencies within it, and future human capital needs more accessible to decision-
makers (case study 3.4).

To successfully use analytics for monitoring and accountability, governments

must be mindful of communication. Using analytics for monitoring can lead to
pervasive distrust among public servants, particularly if governments do not clearly
communicate how data will be used. Careless communication can foster the perception
that analytics is a tool for control rather than a means to foster accountability and
efficiency, hindering the successful introduction of analytical innovations in the public
sector.

For examples of how analytics can be

used for monitoring and accountability
to reduce corruption, refer to chapter 8 of
The Government Analytics Handbook.

Government Transparency

The second application of government analytics this section considers is to increase

government transparency. When citizens have access to actionable information about
how government operates—for instance, through a dashboard or portal that tracks
service delivery or public investment in different regions of a country—they are
empowered to hold the government accountable, identify policy issues, communicate
their priorities, and advocate for change.

The organization reAcción Paraguay demonstrates how citizen mobilization,

empowered by publicly available data from the country’s e-Procurement system,
brought to light mismanagement of public funds, helped the government improve the
allocation of public resources, and generated momentum to make the implementation
and management of public programs more transparent and accountable to the public
(case study 3.3). Using analytics to increase government transparency aligns closely
with policy goals regarding open government, open data, and citizen engagement
(World Bank 2022a).

Policy Evaluation and Design

The final application of government analytics considered here is to guide policy

evaluation and design. Using analytics for policy evaluation is a growing agenda with
wide-ranging applications across different areas of government (refer, for example,

A Conceptual Framework for Government Analytics 25

 to Legovini and Jones 2021). External actors can also contribute to these applications.
Granting research institutions and universities access to administrative data not only
increases transparency but can also fuel a research agenda to help policy makers
investigate the effects of existing policies, evaluate their cost-effectiveness, and gain a
clearer understanding of the mechanisms underlying their effects.

Administrative data analytics can be a low-cost option for policy evaluation, potentially
overcoming the lack of baseline data for long-running programs (Barca et al. 2023). For
example, researchers working with the tax authority in Ecuador analyzed tax data to
evaluate a policy intervention in which firms were notified about revenue discrepancies
in their tax reports, demonstrating under what conditions the policy was effective in
increasing corporate tax collection (case study 3.5).

Government analytics can also guide the development and design of new policies.
Analytics can be integrated into the design of programs to help governments leverage
existing resources efficiently, ensuring that programs are cost-effective before they are
implemented more widely. In Guatemala, the Ministry of Education piloted a program
incorporating analytics to improve education service delivery. By using predictive
analytics to produce lists of students at risk of dropping out of school, the ministry
enabled school staff to identify and support these students. An impact evaluation of the
program measured its effectiveness in a few schools within the country before it was
scaled up nationwide (case study 3.6).

How Is Government Analytics Distributed?

Even if government analytics

is highly advanced and its
goals are clearly defined, it will • Consumption
• Production
not contribute to improving
government functioning unless it
is actively used by policy makers
and other stakeholders. Assessing
the impact of analytics requires understanding the extent of its consumption and
production, including identifying current and potential users of analytical products.
To have an impact, analytics must be produced in response to a specific policy need
identified by its consumers.

Consumption

The first factor considered here is who consumes analytical products in the public sector
and why. Understanding analytics consumption makes it possible to identify whether
the knowledge being generated is relevant to the people who use it—and thus whether
it is useful for shaping critical decisions. For instance, analytics on primary school
students’ learning trajectories might help an early childhood development agency
refine its pedagogical decisions, but it might not be relevant (especially in the short
run) to a higher education agency.

26 Data for Better Governance

At the same time, different government agents might use the same analytical products
for different purposes. For example, politicians might use analytics to design more
effective laws and policies, while public sector managers might use it to assess the
performance of their units. Understanding who uses an analytical product and
adapting the product in response to their feedback is crucial to ensure that analytics
remains relevant to the needs of policy makers. Likewise, taking a broad view of the
potential consumers of analytical products can help governments understand the need
for evidence among different decision-makers across public administration, enabling
them to identify opportunities for analytical solutions and better target and implement
those solutions. Analytics has more impact when policy makers ask for the evidence
they need and use it to shape policies and inform critical decisions, which requires a
deeper culture of evidence-based decision-making.

By understanding and responding to existing operational needs, analytics producers

may even discover new uses for existing analytics. For example, the tax team at the
Secretariat of Finance of Rio Grande do Sul state in Brazil developed an algorithm that
uses tax data from electronic business-to-business invoices to estimate market reference
prices. This algorithm initially served a descriptive function within the tax unit. But its
value increased dramatically when it was shared with procurement teams, who were able
to use the market reference price information to enhance their negotiation strategies,
leading to a significant reduction—13.2 percent—in the price of procured goods (case
study 3.7). When analytical products are shared and adapted by producers to meet the
needs of consumers, new types and applications of analytics can be uncovered.

For an outline of how policy makers can

consume and disseminate analytical
products, refer to chapters 25 and 26 of
The Government Analytics Handbook.

Production

The second element examined here is how often agencies produce analytics. Even high-
quality, much-needed analytics, if only produced intermittently or when it is already
too late to inform policy decisions, will not be useful to decision-makers. For example,
imagine an agency that produces stellar predictive analytics about the evolution of the
government’s wage bill—but does so only when it is already too late to inform budget
discussions.

Three different kinds of analytics production can be identified. Analytics can be

produced ad hoc, in the form of single exercises in response to one-off requests. It can
also be produced on a more regular basis in a bureaucratic fashion, to comply with
existing laws and regulations. Finally, analytics can be produced strategically, in the
form of continuous analytical products that respond to an overall strategic plan to meet
organizational needs and respond to real-time demands.

A Conceptual Framework for Government Analytics 27

 A strategic approach to producing analytics is essential to getting useful information
into the hands of decision-makers when they need it most. If agencies produce
analytics on a regular schedule, it is easier to disseminate information to those who
need it and integrate evidence into existing decision-making processes. Strategically
producing analytics also helps decision-makers think proactively about the evidence
they need to guide policies and reforms, and it can even help them identify problems
or opportunities they would not otherwise notice, making it easier to build toward
more complex analytics. For example, continually monitoring health indicators
enabled the Ministry of Health of Chile to discover that the patients who need regular
medical checkups most—those with chronic conditions—were missing primary
care appointments more frequently, a problem they had not previously identified.
This insight enabled the ministry to take appropriate action to address the issue
(case study 3.1).

How Do the Types, Applications, and Distribution of Analytics Interact?

What analytics looks like needs to be broken down in this level of detail because it
is possible to pursue each of the applications of government analytics—enabling
monitoring and accountability, increasing government transparency, and aiding policy
evaluation and design—using descriptive, diagnostic, and predictive types of analytics.
It is also possible to pursue them at different scales of production and consumption:
an analytical project might be targeted at a specific government agency or aim to
reach a government-wide audience, including public sector managers and politicians.
For example, consider a dashboard tracking data regarding service delivery. This is
a descriptive analytics application, but it might be applied differently by different
stakeholders. A government agency might use the dashboard to monitor its own
efficiency, whereas a watchdog organization might use it to evaluate government
transparency, and policy makers might consult it when making budget decisions. For
this reason, practitioners need to consider the types, applications, and distribution of
government analytics to describe it holistically.

These independent dimensions of analytics also interact. Governments should tailor

the types of analytics they produce to the operational needs of its current and potential
users. Likewise, they should consider how different types and applications of analytics
can build on one another, especially when analytical products are regularly produced,
shared, and improved. However, understanding how different types of analytics are
produced and consumed within public administration and applied to address different
policy-making challenges illustrates only a piece of the overall government analytics
ecosystem. Analytics also interacts with its enabling conditions: the underlying factors
that make analytical projects possible and help them succeed in improving government
functioning. The following section discusses two essential enabling conditions for
government analytics: analytical capabilities and data infrastructure.

28 Data for Better Governance

WHAT ENABLING CONDITIONS SUPPORT
GOVERNMENT ANALYTICS?
An understanding of what analytics looks like in public administration—including
what types of analytics are produced and how they are applied—provides a starting
point for a deeper exploration of the enabling conditions for analytics. What factors
make government analytics possible and enable it to have an impact with respect
to improving decision-making and public policy? By defining and identifying these
factors, governments can map opportunities to build capacity in the directions outlined
in the previous section—toward more diverse types of analytics, more intensive
and strategic production of analytics, and more extensive application of analytics
by decision-makers—all while ensuring that analytics remains problem driven and
relevant to policy makers’ needs.

The discussion here first examines the analytical capabilities governments need,
focusing on strategies and actions at both the organizational and individual levels to
promote and sustain the use of analytics. Second, it considers the data infrastructure
that ensures data are high in quality and readily accessible to analytics teams and
decision-makers (figure 2.3).

FIGURE 2.3 Enabling Conditions for Government Analytics

Analytics
Different types of analytics can be applied
in a variety of ways to improve government
functioning.

High-quality, Individual skills and

accessible data from organizational
management structures and
information systems incentives help in the
offer raw material for development and
analytics. use of analytics.

Capabilities
Infrastructure

Source: Original figure for this publication.

A Conceptual Framework for Government Analytics 29

 What Analytical Capabilities Does the Government Possess?

The underlying capabilities required for data analytics and evidence-based policy
making include not just individual skills but also the institutional structures and
incentives that support the development and use of analytical insights. Analytical
capabilities can thus be categorized as either organizational or individual.

Capabilities

Individual

Organizational

Organizational Capabilities

Structures, incentives, and strategies at the organizational level play a critical role in
promoting and sustaining the use of analytics as a key decision-making tool. Dedicated
analytics units, incentives to support analytical projects, and strategies for collaborating
with external actors create conditions in which decision-makers have easy access to
high-quality analytics when they need it. Governments that proactively invest in their
analytical capabilities are more likely to use data effectively overall (Pew Charitable
Trusts 2018).

Dedicated analytics units are instrumental in promoting the production of analytics

and ensuring that analytics insights align closely with operational needs, as discussed
in the previous section. These units not only centralize expertise but also foster a more
cohesive and agile approach to using analytical products to tackle policy challenges.
Without them, analytics efforts tend to be ad hoc, making it challenging to strategically
and consistently produce analytics that meet organizational demands and the needs of
decision-makers.

Analytics units can be organized in a centralized manner, with a single, dedicated

analytics unit in the government responsible for producing analytics across diverse
government functions and information systems in response to the needs and demands
of different agencies. For example, the Ágata data analytics unit in Bogotá, Colombia,

30 Data for Better Governance

is responsible for generating insights across different policy areas, including health and
public finance, to support Bogotá’s transformation into a “smart city” (case study 3.8).

Alternatively, a government might operate under a decentralized model in which

individual agencies invest in developing their own analytical teams to meet their
own needs. For example, in Peru, the National Superintendency of Customs and
Tax Administration has its own analytics unit. This unit has undertaken numerous
analytical projects to improve tax and customs control and compliance, serving
as a key element of the superintendency’s strategy for improving the efficiency of
tax administration (case study 3.9). While centralized analytics units can achieve
economies of scale by enabling larger and more diverse teams, decentralized units
may offer sector-specific knowledge relevant to better targeting of analytical products.
However they are organized, analytics units allow governments to institutionalize
analytical innovation and experimentation.

Governments can further encourage the use of data analytics through incentives,
including recognition programs, and by allocating funding to analytics. Internal
funding opportunities for analytical projects can motivate agencies and public servants
to explore innovative applications of data and analytics. Governments can also
establish platforms or host events for sharing best practices, tools, and success stories,
underscoring the significance of using analytics in enhancing outcomes. For example,
governments in Ecuador and other countries in Latin America and the Caribbean
are hosting special events for data analytics enthusiasts. At these “datathons,” public
servants and members of the public are given access to administrative data, which
they use to identify relevant policy questions and produce analytics that can be used to
inform policy making and improve government functioning (case study 3.10).

Strategies for collaborating with external actors also play a pivotal role in fostering a
supportive environment for analytics. These strategies may include partnerships with
researchers, academic institutions, nonprofits, or multilateral organizations. Among
other initiatives, governments can collaborate with such organizations to obtain
technical support for specific analytical projects, enhance data management practices,
engage in analytical capacity-building activities, and integrate new analytical methods
and products into their operations.

For guidance on how organizational

reforms can enable government
analytics, refer to chapter 3 of The
Government Analytics Handbook.

Individual Capabilities

To build individual capabilities for analytics, governments must consider how to

assess and strengthen the analytical skills of public servants, as well as how to recruit
and retain data analysts. A mixture of subject area and analytical expertise is often

A Conceptual Framework for Government Analytics 31

 necessary to ensure that analytical products will be relevant to decision-makers in a
particular policy area (Bhupatiraju et al. 2023). Fostering this combination of expertise
is a management challenge, however, especially in light of increasing job specialization.

In some contexts, it may be possible to train subject area experts, such as tax
specialists, to conduct analytics on data in their field. Another approach is to create
working groups (or units) that bring together subject area experts and data analysts
to collaborate on analytical products. In this case, subject area experts must be able to
articulate the evidence they need, understand analytical insights, and apply analytics to
guide their decision-making. Which of these two approaches is better depends on an
administration’s existing skills and organizational capabilities, especially whether it has
dedicated analytics units.

To choose an approach, governments also need to assess the analytical skills of public
servants more broadly. Assessment tools like exams, focus groups, and surveys can help
determine which analytical skills public servants have or need to develop. Governments
must invest in ongoing training to support these skills, because the process of building
analytical capacity requires time, while personnel turnover poses a challenge to the
sustainability of increased capabilities.

Governments should also evaluate their recruitment practices and career advancement
opportunities for data analysts, which determine whether they are adequately staffed
with public servants capable of using data for analytics. Governments can signal their
commitment to analytics and provide incentives for skilled data analysts to join and
remain in the public sector by establishing a dedicated career track for data analysts.
Having a dedicated career track also places public administration in a stronger position
to capitalize on the potential of administrative data and emerging technologies.
Research suggests, for example, that government organizations that consider technical
or data skills key attributes when hiring public servants are better able to integrate and
use new technologies to improve their operational effectiveness (Lember, Kattel, and
Tõnurist 2018).

In addition to building analytical skills and recruiting and retaining data analysts,
governments should consider initiatives to build decision-makers’ capacity to use
analytical products. Improving decision-makers’ knowledge and ability to comprehend
data analytics results—and identify their limitations—is crucial for effectively
integrating analytical findings into decision-making processes (refer, for example, to
Hjort et al. 2021; Mehmood, Naseer, and Chen 2024).

What Is the State of the Government’s Data Infrastructure?

Governments’ organizational and individual capabilities to undertake, manage,

and apply data analytics and its insights are important, but analytics will not lead to
better outcomes unless the data on which it is built are high in quality and accessible
(US Commission on Evidence-Based Policymaking 2017).3 Effective data management
includes the creation and establishment of protocols to guarantee data quality,
accessibility, interoperability, and relevance, along with government agencies to oversee

32 Data for Better Governance

these protocols. It is also crucial to ensure that data are used ethically and transparently.
Laws and regulations ensuring citizens control their own data can foster transparency
and increase the demand for government analytics among public servants and citizens
(Attard et al. 2015; World Bank 2021).

Infrastructure

Quality

Accessibility

Data Quality

High-quality data are accurate (they represent reality as closely as possible), complete
(they are not missing any critical information), consistent (they adhere to the same
standards and formats across records and over time), and reliable (repeated analyses of
the same data points produce similar results).

Establishing robust, systematic data quality controls is the first step toward harnessing
administrative data for analytics. Management information systems typically function
as central repositories for transactional data: they collect and process the daily
operations of government organizations. For example, an e-Procurement system
registers contracts for government purchases, and a public financial management
information system (PFMIS) collects data about budget allocations to programs. Since
administrative data are not usually collected and compiled with analytical purposes
in mind, guaranteeing the quality of these data and their readiness for analytics is
particularly important to ensure high-quality analytical products.

The processes for preparing administrative data for analytics demand investments
in technical infrastructure, including software and hardware, as well as skilled
personnel. The critical task of “cleaning” data, for instance, includes identifying
and correcting duplicates, corrupt observations, inaccuracies, and inconsistencies.
It is also important to ensure that data adhere to standardized formats to facilitate
harmonization and interoperability across various data sets and information systems.4

A Conceptual Framework for Government Analytics 33

 Finally, anonymization or pseudonymization techniques are necessary to meet data
protection and privacy standards.5

Linking or integrating data from different sources through common data identifiers
is another way to increase data relevance, make data more usable, and increase the
impact of analytical products. For example, the World Bank, in collaboration with
subnational governments in Brazil, developed an innovative, data-driven tool to
improve the detection of fraud, corruption, and collusion in public procurement. This
tool, known as the Governance Risk Assessment System, depends on the integration of
diverse data sources, including e-Procurement systems and HRMISs, with corporate
and shareholder information. The risk assessment system proved instrumental in
investigating procurement irregularities in Brazil (case study 3.11).

For principles and guidance on

strengthening data infrastructure, especially
that for data from human resources
management information systems and
public financial management information
systems, refer to chapters 9 and 11 of
The Government Analytics Handbook.

Data Accessibility

High-quality data must be accessible and promptly available to have an impact on

government functioning. Data accessibility means that public servants can regularly
access data owned by their organizations, that data and analytical products are readily
shared within government, and that data are made available to the public, all while
adhering to protocols for confidentiality and ethics.6

When data are accessible and readily available, they enable analysts, policy makers,
and researchers to harness a wealth of information that can drive informed decisions.
For example, the Secretariat of Finance of Rio Grande do Sul state in Brazil was
able to design an algorithm to guide procurement based on administrative data that
had originally been collected for taxation. The integration of data across different
information systems created the conditions for collaboration across units, resulting in
an innovative analytical product that decreased procurement prices by 13.2 percent
(case study 3.7).

Administrative data are valuable not only for internal analytics but also for external
sharing to promote transparency, transforming data into a public asset. In this
context, the open data movement has been pivotal in facilitating public access to data
for transparency and civic engagement. Paraguay, for instance, was the first country
in the world to create a portal adhering to the Open Government Partnership’s
Open Contracting Data Standard for publication and access to data on government
contracts and public tenders (World Bank 2018). Data from this portal empowered

34 Data for Better Governance

the organization reAcción Paraguay to mobilize citizens and help the government
improve the allocation of public resources (case study 3.3). When data are accessible
to the public, transparency, accountability, and public trust are all enhanced because
government actions and policies can be examined and validated.

These potential applications of administrative data are hampered by accessibility

challenges. Inconsistent procedures and legal frameworks for data access and
sharing (both within government and with outside parties), coupled with insufficient
information about available data sets and their elements, hinder collaboration and
innovation. When valuable data remain siloed within government agencies, it is
difficult to cross-reference information and capture the full picture of a particular
policy issue.

One way to tackle these challenges is to create a standardized process for analysts
and policy makers to use to access administrative data. This entails, among other
things, improving how data are managed and documented. For instance, establishing
a comprehensive data inventory that outlines the available data sets and elements
within information systems ensures that analysts and policy makers understand what
information exists. In addition, detailed, searchable metadata documents can help users
understand data characteristics, such as time frames, measurement units, and coverage.
This enables users to assess the utility of the data for addressing policy questions before
accessing them, so users can determine whether gaining access is worthwhile. Finally,
a formal, well-documented, and easy-to-follow protocol should be put in place to
regulate data access and security. By emphasizing data accessibility and simplifying
protocols for their use, governments can markedly improve their analytical products.

How Do Analytical Capabilities and Data Infrastructure Interact?

The enabling conditions for government analytics—the individual and organizational

capabilities and underlying data infrastructure—are the essential ingredients for every
analytical project. But they are also the factors that make public administration a
fruitful environment for evidence-based policy making in general: the kind of place
where public servants are aware of the data that governments produce and work
with, empowered to identify and undertake new analytical projects, and skilled at
using analytical insights to make decisions, implement policies, and deliver goods
and services with greater effectiveness and efficiency. Investment in analytical
capabilities and data infrastructure can ensure public administration offers this kind of
environment.

A Conceptual Framework for Government Analytics 35

 These enabling conditions also strengthen one another. The quality and accessibility
of data infrastructure are strengthened by the skills of personnel who prepare data
for analytics and the organizations that oversee quality and accessibility protocols.
Likewise, stronger data infrastructure that links or integrates data from different
sources makes higher-quality data accessible to organizations and individuals, enabling
them to further develop their skills and undertake more complex analytics.

However, governments do not need to wait for the perfect set of enabling conditions
to start undertaking analytics with administrative data. In fact, as this chapter has
shown, good analytical projects not only produce decision-making insights but also
help strengthen the enabling conditions for analytics within public administration.
Analytical projects can help reveal weaknesses in the infrastructure for the data
underlying the analysis, enabling improvements to be made. At the same time,
undertaking an analytical project can help organizations identify the analytical skills
their personnel need to develop, as well as the organizational structures and incentives
that would make future projects more effective. When data are regularly produced by
dedicated teams and consumed by decision-makers who offer feedback, next steps to
further strengthen the enabling conditions for analytics are easier to identify. The next
chapter looks at how some analytical initiatives in Latin America and the Caribbean
have both drawn on and strengthened data infrastructure and analytical capabilities
within public administration, all while improving government functioning across
various policy areas.

36 Data for Better Governance

NOTES
1. Data experts commonly acknowledge a fourth type, referred to as prescriptive analytics. However,
given the overlap in definition with the other three types, this report does not treat it as a separate
category.
2. Beyond its use for monitoring and accountability in specific sectors or policy areas, government
analytics can also be applied to monitoring systems (commonly known as “monitoring and evaluation”)
at the national level. Centralized monitoring systems track progress toward the achievement of
medium- or long-term objectives established in national strategies. Government data represent a rich
source of information for national monitoring systems, especially for measuring how the functioning
of public administration contributes to the achievement of broader development objectives.
3. Given the availability of high-quality consulting materials on data infrastructure, this section of
the conceptual framework offers only a broad view of this important enabling condition. Notes
provide references to companion pieces that provide the necessary details for readers seeking a more
comprehensive exploration of the topics discussed.
4. For insights on systems interoperability and data harmonization, refer to Interoperability: Towards a
Data-Driven Public Sector (World Bank 2023).
5. It is important to note that even if administrative data are of high quality, they may not have an
architecture that allows policy makers or researchers to effectively use them. This issue is particularly
significant with regard to personally identifiable information, as data files often need to be encrypted
both in transit and at rest to protect privacy, imposing additional constraints on how the data are
managed and used. Sharing encrypted data across agencies can be problematic if there are no
standardized encryption protocols or agreements on data access and use. Data governance authorities
should ensure that there are clear rules and procedures that facilitate analytics while protecting privacy.
For insights on data governance, refer to chapters 6 and 8 of the World Development Report 2021: Data
for Better Lives (World Bank 2021).
6. For insights on institutions, laws, and regulations to improve data accessibility, refer to the GovTech
Maturity Index, 2022 Update: Trends in Public Sector Digital Transformation (World Bank 2022a).

REFERENCES
Attard, Judie, Fabrizio Orlandi, Simon Scerri, and Sören Auer. 2015. “A Systematic Review of Open
Government Data Initiatives.” Government Information Quarterly 32 (4): 399–418. https://doi​
.org/10.1016/j​.giq.2015.07.006.
Barca, Valentina, Madhumitha Hebbar, Charles Knox-Vydamov, and Ida Brzezinska. 2023. “We Have the
Data, Let’s Use It Better: Pushing the Boundaries of Social Protection Administrative Data Analysis
and Use.” Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ) GmbH, Bonn, Germany.
https://socialprotection.org/sites/default/files/publications_files/GIZ-%20Use%20Admin%20Data%20
Social%20Protection.pdf.
Bhupatiraju, Sandeep, Daniel Chen, Slava Jankin, Galileu Kim, Maximilian Kupi, and Manuel Ramos
Maqueda. 2023. “Government Analytics Using Machine Learning.” In The Government Analytics
Handbook: Leveraging Data to Strengthen Public Administration, edited by Daniel Rogger and Christian
Schuster, chap. 16. Washington, DC: World Bank. https://doi.org/10.1596/978-1-4648-1957-5.
Cote, Catherine. 2021. “4 Types of Data Analytics to Improve Decision-Making.” Business Insights (blog),
October 19, 2021. https://online.hbs.edu/blog/post/types-of-data-analysis.
Hjort, Jonas, Diana Moreira, Gautam Rao, and Juan Francisco Santini. 2021. “How Research Affects Policy:
Experimental Evidence from 2,150 Brazilian Municipalities.” American Economic Review 111 (5):
1442–80. https://doi.org/10.1257/aer.20190830.

A Conceptual Framework for Government Analytics 37

 Legovini, Arianna, and Maria Ruth Jones. 2021. “Administrative Data in Research at the World Bank:
The Case of Development Impact Evaluation (DIME).” In Handbook on Using Administrative
Data for Research and Evidence-Based Policy, edited by Shawn Cole, Iqbal Dhaliwal, Anja
Sautmann, and Lars Vilhuber, chap. 14. Cambridge, MA: Abdul Latif Jameel Poverty Action Lab.
https://admindatahandbook.mit.edu/book​/v1.0-rc6/dime.html.
Lember, Veiko, Rainer Kattel, and Piret Tõnurist. 2018. “Technological Capacity in the Public Sector:
The Case of Estonia.” International Review of Administrative Sciences 84 (2): 214–30. https://doi​
.org/10.1177/0020852317735164.
Mehmood, Sultan, Shaheen Naseer, and Daniel L. Chen. 2024. “Training Policymakers in Econometrics.”
Unpublished manuscript, last modified January 2024. https://users.nber.org/~dlchen/papers/Training​
_Policy_Makers_in_Econometrics.pdf.
Pew Charitable Trusts. 2018. How States Use Data to Inform Decisions: A National Review
of the Use of Administrative Data to Improve State Decision-Making. Philadelphia: Pew
Charitable Trusts. https://www​.pewtrusts.org/en/research-and-analysis/reports/2018/02​
/how-states-use-data-to-inform-decisions.
Rogger, Daniel, and Christian Schuster, eds. 2023. The Government Analytics Handbook:
Leveraging Data to Strengthen Public Administration. Washington, DC: World Bank. https://doi​
.org/10.1596/978-1-4648-1957-5.
Tavares, Rafael Alves de Albuquerque, Daniel Ortega Nieto, and Eleanor Florence Woodhouse. 2023.
“Government Analytics Using Human Resources and Payroll Data.” In The Government Analytics
Handbook: Leveraging Data to Strengthen Public Administration, edited by Daniel Rogger and Christian
Schuster, chap. 10. Washington, DC: World Bank. https://doi.org/10.1596/978-1-4648-1957-5_ch10.
US Commission on Evidence-Based Policymaking. 2017. The Promise of Evidence-Based Policymaking:
Report of the Commission on Evidence-Based Policymaking. Report to the President of the United States,
Speaker of the House, and President of the Senate. https://www2.census.gov.mcas.ms/adrm/fesac/2017​
-12-15/Abraham​-CEP-final-report.pdf.
World Bank. 2016. World Development Report 2016: Digital Dividends. Washington, DC: World Bank.
https://doi​.org/10.1596/978-1-4648-0671-1.
World Bank. 2018. “Paraguay: Systematic Country Diagnostic.” World Bank, Washington, DC.
https://documents1.worldbank.org/curated/en/827731530819395899/pdf/Paraguay-SCD-06292018.pdf.
World Bank. 2020. Digital Government Readiness Assessment (DGRA) Toolkit V.31: Guidelines for Task
Teams. Version 3.0 (April 2020). Washington, DC: World Bank.
World Bank. 2021. World Development Report 2021: Data for Better Lives. Washington, DC: World Bank.
https://www.worldbank.org/en/publication/wdr2021.
World Bank. 2022a. GovTech Maturity Index, 2022 Update: Trends in Public Sector Digital Transformation.
Equitable Growth, Finance and Institutions Insight—Governance. Washington, DC: World Bank.
https://hdl​.handle.net/10986/38499.
World Bank. 2022b. New Approaches to Closing the Fiscal Gap. LAC Semiannual Update. Washington, DC:
World Bank. https://hdl.handle.net/10986/38093.
World Bank. 2023. Interoperability: Towards a Data-Driven Public Sector. Equitable Growth, Finance and
Institutions Insight—Governance. Washington, DC: World Bank. https://doi.org/10.1596/38520.

38 Data for Better Governance

CHAPTER 3

Case Studies of
Government Analytics
in Latin America and
the Caribbean

INTRODUCTION
The conceptual framework presented in the previous chapter offers a high-level view
of the different elements that make up the government analytics ecosystem, which
practitioners can use to understand how analytical initiatives succeed in enhancing
policy design and implementation. This chapter addresses this question by offering
detailed descriptions of 12 case studies of analytical initiatives in Latin America and the
Caribbean, covering various countries and policy areas. Each case study provides an
example of how government and citizen organizations have used administrative data
to generate evidence for decision-making and the significant impact of this evidence
on how policy challenges are addressed. The case studies also consider how enabling
conditions (data infrastructure and analytical capabilities) have shaped these initiatives.
For example, case study 3.9 describes how the tax authority in Peru created an analytics
unit (organizational capabilities) to develop predictive analytics about tax evasion
(types and applications of analytics).

The case studies represent just a small portion of the diverse analytical work that
public servants in the region have conducted within public administration. They
were identified through a mixture of desk review and interviews with government
officials and practitioners from multilateral organizations. Cases were then selected for
inclusion in the report based on their potential to illustrate analytical concepts, policy
objectives, challenges, and lessons. Each case study emphasizes evidence that is linked
to the conceptual framework. However, not all elements of the conceptual framework
are relevant to every case, so just the relevant evidence for selected elements and case
studies is highlighted.

A reproducibility package is available for this book in the Reproducible Research Repository at
https://reproducibility.worldbank.org/index.php/catalog/209.

39
 The case studies demonstrate the wide variety of policy goals that can be pursued
through government analytics. Analytics can inform decisions at different stages
of program implementation: it can identify gaps and areas for improvements (for
example, case study 3.2), test potential interventions before scale-up (for example, case
study 3.6), and help define the implementation details of new policies (for example,
case study 3.12). Analytical projects do not have to be complex to have an impact on
policy making, but they do take time to mature and often develop through an iterative,
trial-and-error approach. Governments should account for the time cycles associated
with such an approach and nurture an enabling ecosystem to help analytical initiatives
grow. In particular, dedicated analytics units and external collaborations can be
effective strategies to leverage analytical capabilities. Connecting data from different
information systems also enables data to be applied outside the government function
that generated them, making a strong case for data accessibility across government
organizations.

Finally, the 12 case studies presented in this chapter show that government analytics
can be developed in any sector or ministry and at any level of government. There
is no one-size-fits-all approach: a decentralized, ministry-specific analytics unit or
a centralized, cross-cutting one can enable the use of data for analytics. The right
approach to government analytics depends on context: policy-making challenges,
decision-making needs, and strategic goals.

CASE STUDY 3.1: USING DESCRIPTIVE

ANALYTICS TO REDUCE MISSED
MEDICAL APPOINTMENTS (CHILE)
Missed appointments are a persistent problem for health care systems—and a
costly one. In Chile, patients miss an estimated 10–20 percent of all scheduled
appointments, resulting in an annual loss of approximately $180 million (Boone
et al. 2022; CNEP 2022). Patients with chronic conditions that require continual
monitoring and treatment—like hypertension and type 2 diabetes—are at
particular risk if they miss appointments, which can increase mortality rates and
health care costs.

To reduce missed medical appointments, Chile’s Ministerio de Salud (Ministry of

Health, or MINSAL) used descriptive analytics to better understand how missed
appointments affect the country’s health care system and develop a program in
response. MINSAL drew on Chile’s advanced health management information system
(HealthMIS), which includes digital patient records with data about demographics,

40 Data for Better Governance

contact information, health visits, disease prevalence, and test results. In particular,
MINSAL compared patient-level data on missed appointments with data on disease
prevalence. By analyzing these data, MINSAL was able to better understand which
patients missed appointments the most and formulate a policy tailored to this
population’s needs.

MINSAL discovered that patients with chronic conditions—approximately

60 percent of Chile’s population (Margozzini and Passi 2018)—accounted for a
significant fraction of missed appointments, and it responded by developing a
program targeting them, the Mensajería para la Gestión de Citas en Pacientes
Crónicos (Critical Care Appointment Management Program). Through this
program, patients with chronic conditions received automated reminders by text
message a few days before a scheduled appointment, with options to confirm,
reschedule, or cancel the appointment. By encouraging patients to communicate
with health clinics, the program aimed not just to reduce missed appointments
but also to help clinics manage scheduling and allocate human resources more
efficiently.

MINSAL’s program effectively reduced missed appointments by 6 percent and

increased the probability of using primary care services by 8.7 percent for patients
with type 2 diabetes and by 10.7 percent for patients with hypertension (Boone
et al. 2022, 2023; CNEP 2022). The program also improved health behaviors: it
led patients to better adhere to prescribed medications and decreased in-hospital
mortality rates.

In developing the program, MINSAL also discovered new information about

the causes of missed appointments. While implementing the reminder system,
MINSAL found missing and incorrect contact information in electronic records
that accounted for about 20 percent of missed appointments (CNEP 2022). The
discovery underscores the importance of implementing systematic data quality
controls in information systems.

MINSAL’s initiative illustrates the impact that relatively simple descriptive

analytics can have when it is based on rich, comprehensive data (figure 3.1).
By providing an accurate picture of health care challenges in Chile, descriptive
analytics helped MINSAL pinpoint where problems were most acute and
formulate a targeted policy to address them. It also revealed problems with the
accuracy and completeness of the underlying data, enabling corrections to be
made. In this way, MINSAL’s program not only encouraged the patients who
most need primary care services to use them but also strengthened the overall
government analytics ecosystem in Chile.

Case Studies of Government Analytics in LAC 41

FIGURE 3.1 Using Descriptive Analytics to Reduce Missed Medical
Appointments, Chile

Applications
Comprehensive monitoring of health care
system; designing program to reduce missed
medical appointments

Types

Descriptive

Analytics
Discover and correct
missing contact information

Infrastructure Capabilities

Stream of funds for analytics; decision-makers'

Integrated and complete data
readiness to utilize analytical products

Source: Original figure for this publication.

CASE STUDY 3.2: USING DIAGNOSTIC ANALYTICS

OF HUMAN RESOURCES DATA TO EXPLORE
THE GENDER PAY GAP (COLOMBIA)
Colombia’s Ley 1496/11 (Law 1496 of 2011) mandates that women and men receive
equal pay and compensation for work of equal value. But is the law followed in
practice?

42 Data for Better Governance

The Unidad de Científicos de Datos (Data Scientists Unit, or UCD) of Colombia’s
Departamento Nacional de Planeación (National Planning Department) used a
combination of descriptive and diagnostic analytics to determine whether there was a
gender pay gap in the country among public sector workers with temporary contracts.
Established in 2017, the UCD is a dedicated analytics unit that aims to harness the
potential of administrative data to support the design, monitoring, and evaluation of
public policies in Colombia.

To study pay differences among public sector workers, the UCD gained access to
human resources management information system (HRMIS) data through data-
sharing agreements with the Departamento Administrativo de la Función Pública
(Administrative Department of the Public Service). The UCD used three data tables
from the HRMIS that provided comprehensive information on public sector workers
from 2012 to 2019, including the following:

● Personal information, such as name, gender, and date of birth (at the worker level)
● Education information (at the worker level)
● Information about the nature and financial value of contracts (at the contract
level).
After integrating these data tables, the UCD conducted data quality checks
to prepare the data for analytics. This involved eliminating duplicate entries,
rectifying records with missing information or outliers, and resolving data
validation errors.

The UCD began its study with descriptive analytics, by comparing average
earnings between men and women across different years. This descriptive
comparison offered valuable insights, but it did not account for other factors that
might influence workers’ pay, such as their work experience or education. To
pinpoint the role of gender in pay differences, the UCD continued the study with
diagnostic analytics. It used econometric models and hypothesis testing to compare
workers who shared similar characteristics, including education, contract details,
geographic region, and experience, isolating gender as the sole difference between
them.

The UCD’s study revealed a gender pay gap of 5–6 percent among temporary
workers in the public sector (UCD 2019), and the National Planning Department
informed the authorities that oversee gender policies of these findings. By
combining different types of analytics, the UCD not only mapped pay differences
among public sector workers (descriptive analytics) but also verified that gender
accounted for part of those differences (diagnostic analytics). The UCD’s analytical
work demonstrates that good policy evaluation produces evidence describing not
just what is happening but also why it is happening. It also offers an example of
the critical role a dedicated analytics unit can play in generating that evidence and
supporting better decision-making.

Case Studies of Government Analytics in LAC 43

 CASE STUDY 3.3: DATA ANALYTICS FOR
CITIZEN ACCOUNTABILITY—THE CASE OF
reACCIÓN (PARAGUAY)
In recent years, Paraguay has pledged to use revenue from electricity sales to fund
local school infrastructure through the Fondo Nacional de Inversión Pública y
Desarrollo (National Public Investment and Development Fund, or FONACIDE),
established in 2012. However, as a result of local patronage and influence, funds have
been channeled toward politically favored schools rather than the marginalized ones
for which they were originally intended. In response, reAcción Paraguay, a recognized
youth anticorruption organization, has analyzed open procurement data to ensure
FONACIDE funds reach the intended schools, enhancing government transparency
and strengthening citizen participation in the allocation of public resources (reAcción
Paraguay, n.d.).

Paraguay has a rich legacy of data analytics in procurement and a resilient

infrastructure for publishing open data. In addition to its efforts to make informed
decisions based on timely, comprehensive analytics, the Dirección Nacional de
Contrataciones Públicas (National Public Procurement Agency, or DNCP) has been
committed to the recommendations of the Open Government Partnership since
2012. In 2014, the DNCP introduced a user-friendly, real-time open contracting
portal with comprehensive information on tenders and contracts, machine-readable
data, and an application programming interface for innovative data integration.
Through this portal, Paraguay became the first country in the world to adopt the Open
Contracting Data Standard for publication of and easy access to data on government
contracts and public tenders (World Bank 2018).1 The DNCP’s commitment to
publishing open data on public contracts has empowered journalists, activists, and
civil society organizations to monitor public procurement and identify potential
irregularities.

reAcción Paraguay has used the DNCP’s open data portal to track the allocation
of FONACIDE funds by developing a platform called FOCO that integrates these
open procurement data with data from other official sources (like the Ministerio
de Educación [Ministry of Education]) and data acquired during visits to schools
(reAcción Paraguay 2023). FOCO had its start in a government-organized hackathon, a
program to promote citizen-government interaction through technology (refer to other
examples of this kind of initiative in case study 3.10). FOCO enables users to track
infrastructure spending for eligible FONACIDE-funded schools by presenting data on
the following:

● Infrastructure needs by year and school

● Schools’ rankings on the Ministry of Education’s priority list
● Schools that have received disbursements and the amounts of these disbursements

44 Data for Better Governance

● Categories of spending that were covered by the disbursements (for example,
classrooms, toilets, or furniture)
● Companies contracted for school improvements and the amounts of the contracts
● Results from inspections of target schools and FONACIDE-funded improvements.
FOCO has brought to light mismanagement of FONACIDE funds, increasing the
fund’s transparency and helping improve its administration, management, and
integrity. For example, reAcción’s monitoring led to a remarkable 400 percent increase
in the accurate allocation of FONACIDE resources from 2015 to 2017 in Ciudad del
Este, Paraguay’s second-largest city (Riveros García 2019). As a result of reAcción’s
efforts, the DNCP was persuaded to reform FONACIDE’s national procurement
process. In 2020, reAcción began working directly with Paraguay’s government to
further develop data collection and analytics to make school infrastructure funding
based on evidence a priority, and it continues to work with the DNCP and the Ministry
of Education to monitor FONACIDE.

The example of reAcción Paraguay demonstrates that the impact of government

analytics is amplified when data from different sources are connected. By building
on open and accessible government data, a grassroots initiative was able to empower
civic mobilization, help civil society uncover mismanagement of public funds, increase
transparency and accountability, and improve the allocation of public resources.

CASE STUDY 3.4: BUILDING AN INTEGRATED HUMAN

RESOURCES MANAGEMENT PLATFORM FOR
STRATEGIC WORKFORCE PLANNING (URUGUAY)
Planning for the future of public administration requires data on the composition
of the public sector workforce, the range of skills and competencies it encompasses,
and future human capital requirements. The government of Uruguay and its Oficina
Nacional del Servicio Civil (National Civil Service Office, or ONSC) are developing an
innovative online platform for human resources management that makes these data
available to public servants, policy makers, and the public to optimize the use of public
resources; foster the enabling conditions for future analytics; and make government
more open, transparent, and accountable to all (ONSC, n.d.).

The new platform is a collaborative effort that integrates and analyzes data
from the information systems of multiple government organizations. As the
organization leading this effort, the ONSC has been tasked with integrating,
securing, and disseminating the data received from other organizations,
including transforming the data to ensure they are consistent, comparable,
accessible, and comprehensible to the public. The platform will introduce new
features and descriptive analytics to provide timely, high-quality information to

Case Studies of Government Analytics in LAC 45

 inform decision-making about public sector personnel, enhance recruitment,
foster inclusivity in public employment for minority groups (including Afro-
descendant people, people with disabilities, victims of violence, and transgender
people), bolster government transparency and accountability, and streamline
processes that support the rights and responsibilities of public workers.

To accomplish these goals, the platform will transform the landscape of public
sector recruitment. One component, the “Digital CV,” merges data from
Uruguay Concursa, the existing public sector recruitment portal, with artificial
intelligence to streamline recruitment, ensuring that public positions are filled
with the most qualified candidates. The Digital CV will allow applicants and
current employees to update information about their education, experience,
career history, and skills for use in future selection processes. By providing a
dynamic view of every job position and its holder—including each role’s purpose,
responsibilities, and skills—the Digital CV aims not just to improve recruitment
but also to maintain a secure, organized database of jobs, systematically
categorized by occupational categories, that can be used for strategic workforce
planning and monitoring.

The platform will also enhance human resources management, especially performance
evaluation. It will continuously update public servants’ digital records with their
achievements and competencies by integrating performance data. This will significantly
reduce the time and effort involved in promotions and hiring. The platform will also
enable users to view information about colleagues’ and supervisors’ performance,
receive timely feedback, and set individual objectives, streamlining the talent
management workflow.

Finally, the platform will provide analytics units, such as the ONSC’s Laboratorio
de Innovación y Observatorio de la Función Pública (Innovation Laboratory
and Public Service Observatory), with a new source of integrated data. Among
other things, analysts will be able to study the types of public servants who are
poised to retire or move across positions, as well as the types of jobs the public
administration has difficulty filling, to identify where extra resources will be
essential for sustaining service delivery. For example, the Innovation Laboratory
and Public Service Observatory has found that positions for specialized public
servants are harder to fill than other types of positions and that the salary
offered plays a crucial role in filling these positions (ONSC 2022a, 2022b,
2022c). From recruitment to human resources management to analytics,
Uruguay’s new platform demonstrates the impact government data can have
when different administrative records are connected, the powerful applications
that can be built using descriptive analytics techniques, and the importance of
monitoring and analyzing the public sector workforce for strategic planning
(figure 3.2).

46 Data for Better Governance

FIGURE 3.2 Building an Integrated Human Resources Management Platform for
Strategic Workforce Planning, Uruguay

Applications

Government transparency; monitoring and

accountability for strategic workforce planning

Types

Descriptive

Analytics
Improve data accessibility
for analytics

Infrastructure Capabilities

Integrated data; interoperable management

Analytics unit; skills assessment tools
information systems; data quality controls

Source: Original figure for this publication.

CASE STUDY 3.5: ANALYZING TAX DATA TO INCREASE

CORPORATE TAX COLLECTION (ECUADOR)
Tax evasion is a major concern for governments because it hampers tax collection
and disrupts the allocation of resources within an economy. It can also lead
to dependence on economically inefficient tax instruments. Traditionally, tax
authorities have audited taxpayers to validate their declarations. Recently, they

Case Studies of Government Analytics in LAC 47

 have also begun comparing their declarations with third-party data, such
as employer-provided salary information or reports from firms’ trading
partners. This has become an essential tax collection strategy in high-income
countries. Advancements in information technology and information systems
have made third-party data more accessible and easier to use by integrating
them with tax management information system (TaxMIS) data. These
improvements hold promise for boosting tax collection efforts, especially in
developing economies (Carrillo, Pomeranz, and Singhal 2017; Pomeranz and
Vila-Belda 2019).

Ecuador’s Servicio de Rentas Internas (Internal Revenue Service, or SRI), in

collaboration with a team of researchers, evaluated the first systematic, large-scale
application of third-party data to support Ecuador’s tax enforcement efforts. Under
this initiative, firms were notified when the revenue figures they had reported did
not match third-party data from corporate tax returns detailing the value of their
activities. Firms were urged to file amended returns based on the tax authority’s
access to these data.

To pinpoint discrepancies between the revenue figures reported by firms and those
estimated by third parties, the SRI cross-referenced data from value added tax (VAT)
return forms against multiple sources, including credit card sales, customs data, and
financial institutions’ records. The SRI’s TaxMIS was instrumental in facilitating the
efficient storage and retrieval of digital VAT information and corporate tax returns.
The SRI’s notifications to firms not only highlighted discrepancies in revenue between
a firm’s declaration and third-party data but also included the SRI’s own revenue
calculations.

Researchers estimated the effects of these notifications by analyzing changes

in the firms’ returns before and after receiving the notifications. They found
that responding firms significantly adjusted their reported revenue, aligning it
almost perfectly with the amount highlighted in the notifications. However, the
researchers also found that firms adjusted reported costs, offsetting part of the
potential increases in tax collection, which were estimated to be in the hundreds
of millions of dollars annually. Understanding these responses is crucial for
evaluating the effectiveness of third-party reporting, as its impact may be limited
by enforcement constraints faced by tax authorities and offsetting adjustments
made by taxpayers beyond just reported revenues.

Ecuador’s initiative demonstrates that TaxMIS data (at the taxpayer or transaction
level) can be useful beyond recording transactions. Tax authorities can also
use these data to analyze taxpayer behavior, evaluate responses to policy and
administration changes, and design optimal taxation policies (Brockmeyer
2019). More broadly, the initiative shows how administrative data can be used
for rigorous policy evaluation, especially when multiple administrative records
and nonadministrative databases are connected and external researchers and data
analysts partner with public servants.

48 Data for Better Governance

CASE STUDY 3.6: USING PREDICTIVE ANALYTICS TO
PREVENT SCHOOL DROPOUTS (GUATEMALA)
Students who drop out of school earn less than peers who complete their education,
and they face other socioeconomic challenges as well. Dropouts can be prevented
through early warning systems that identify at-risk students and enable school staff
members to take proactive measures and ensure success for all students. In Guatemala,
where nearly 40 percent of sixth graders drop out of school before ninth grade, the
Ministerio de Educación (Ministry of Education, or MINEDUC) has used analytics to
build an effective early warning system (Haimovich, Vazquez, and Adelman 2021).

MINEDUC’s analytics were enabled by the significant strides it has made in recent
years to enhance its education management information system (EdMIS) and increase
the availability of student-level data. These data—which concern families, attendance,
and test scores—can be used to identify at-risk students. Importantly, MINEDUC
also introduced unique student identifiers for all students in primary and secondary
schools, which allow it to track the educational progress of Guatemalan students over
time (Montes 2022).

In 2017, the government of Guatemala, in collaboration with the World Bank,

developed the Estrategia Nacional para la Transición Exitosa (National Strategy
for Successful Transition, or ENTRE), a program focusing on the transition from
primary to lower secondary school. ENTRE used predictive analytics to develop an
early warning system to identify students at risk of dropping out of school during this
transition. MINEDUC leveraged student-level data from its EdMIS, linear regression
models, and a simple algorithm to produce a list of students at risk of dropping out.
After these students were identified, MINEDUC provided school principals with a list
of students to target for additional support and offered teachers additional training
in preventing dropouts. In its pilot phase, ENTRE was implemented in 4,000 public
primary schools, representing 17 percent of Guatemala’s primary schools.

ENTRE reduced the dropout rate in the transition from primary to lower secondary
school by 9 percent within its first year, underscoring the program’s effectiveness
(Haimovich, Vazquez, and Adelman 2021). Moreover, ENTRE was designed to
be scalable, since it could be implemented using primarily MINEDUC’s existing
administrative data structures, management systems, and personnel, with limited
additional costs.

ENTRE illustrates how predictive analytics can offer a highly cost-effective method
for building an early warning system to prevent dropouts, and the pilot has since been
expanded into a nationwide program (MINEDUC 2019). By investing in the quality
and availability of administrative data, Guatemala created the enabling conditions for
applying predictive analytics, and by designing and implementing a program with
analytics in mind, it was able to measure the program’s effectiveness before scaling it up
(figure 3.3).

Case Studies of Government Analytics in LAC 49

FIGURE 3.3 Using Predictive Analytics to Prevent School Dropouts, Guatemala

Applications

Designing and evaluating early warning

system to reduce school dropouts

Types

Diagnostic; predictive

Analytics

Infrastructure Capabilities

EdMIS with student-level data; data integration

External collaborations
(through unique student identifiers)

Source: Original figure for this publication.

Note: EdMIS = education management information system.

CASE STUDY 3.7: USING TAX DATA TO BOOST

PROCUREMENT EFFICIENCY (BRAZIL)
Public procurement can be made less efficient and less competitive by asymmetry of
information between procurement officers and sellers. When procurement officers
lack access to accurate market reference prices, their negotiating power is undermined,
particularly in regard to products with limited providers. This can lead to higher
purchase prices (Grennan and Swanson 2020). To address this challenge, the Secretaria
da Fazenda do Estado do Rio Grande do Sul (Secretariat of Finance of the State of Rio
Grande do Sul, or SEFAZ/RS) in Brazil introduced an innovative solution: a pricing
algorithm that leverages electronic invoicing data from local business-to-business
transactions to calculate market reference prices for pharmaceutical products.

50 Data for Better Governance

To create this algorithm, SEFAZ/RS took advantage of existing data that had been
integrated across data repositories. Since 2008, Brazil has required registered firms that
are subject to state tax on the circulation of goods and services to issue digital invoices
for their transactions. Over time, this transition to digital invoicing has produced a
large body of transactions data. Although these transactions data were collected for
taxation purposes, SEFAZ/RS saw that they could also be used to calculate market
reference prices for procured products. If procurement officers had access to analytics
based on these repurposed data, they would be better positioned to prepare tender
documents and negotiate more competitive prices for products.

SEFAZ/RS designed an algorithm to guide the procurement of pharmaceutical

products. The algorithm derived market reference prices from transactions data in
three steps:

1. Identifying the product linked to a specific digital invoice

2. Breaking down the product into appropriate units for pricing
3. Calculating the reference price per unit.
In this way, SEFAZ/RS provided procurement officers with a price range for specific
products derived from the distribution of past business-to-business transactions in the
database.

When procurement officers were provided with market reference prices before they set
tendering parameters, purchase prices were reduced (Martinez-Carrasco, Conceição,
and Dezolt 2023). Products that had high initial unit prices, a limited number of
suppliers, and a limited number of public buyers saw the most substantial price
reductions. On average, the final prices of these products decreased by 13.2 percent,
resulting in savings of approximately 4 percent on the average annual total expenditure
for pharmaceutical products. Providing procurement officers with market reference
prices thus appears to be an effective way to boost government efficiency in Brazil.

SEFAZ/RS’s pricing algorithm demonstrates how data can be used for analytics
across government functions. SEFAZ/RS conducted predictive analytics resulting
in an analytical product that helps public servants do their jobs more effectively. By
repurposing tax data, drawing on integrated data repositories, collaborating across
units, and effectively integrating analytics into procurement officers’ purchase process,
SEFAZ/RS was able to establish a well-rooted analytics program with room to grow.

CASE STUDY 3.8: IMPROVING GOVERNMENT

EFFICIENCY THROUGH A CENTRALIZED ANALYTICS
UNIT (COLOMBIA)
At the end of the 2010s, Bogotá, Colombia, embarked on a journey to transform
itself into a “smart city”—a city that is more sustainable, livable, and efficient for
its inhabitants—by embracing smart technologies and data analytics. Guided by
this vision, in 2020, the government of Bogotá established Ágata, a data analytics

Case Studies of Government Analytics in LAC 51

 agency within the government (Alcaldía Mayor de Bogotá, Colombia 2020). Ágata
is constituted as a simplified joint stock company in which different city government
organizations and companies are partners. As a mixed-capital company, it funds its
own operations, decreasing its reliance on public financing and the prevailing political
will to advance data analytics.

Ágata functions as a centralized analytics unit for the city of Bogotá. It is responsible
for generating insights across various information systems and government functions
to address the needs of the city’s stakeholders. With its current staff of 66 employees,
Ágata produces an array of analytical products using descriptive, diagnostic, and
predictive analytics, as well as artificial intelligence tools. Its centralized organizational
model promotes economies of scale, allowing it to recruit employees with specialized
analytical skills and develop a team with a holistic view of public administration, the
challenges it faces, and the available data, reducing information silos. For the local
government, the agency aims to improve decision-making by creating analytical
products using data from multiple organizations. For citizens, it aims to elevate the
quality of life, streamline access to public services, and promote the adoption and use of
digital innovations.

From its inception, Ágata’s analytical work has had an impact on diverse sectors, and
the agency has expanded analytical capabilities in both the public and private sectors
(Riaño and Delgado 2024). For example, Ágata played an important role during the
COVID-19 pandemic by creating dashboards to track the spread of the virus across
the city and monitor hospital occupancy rates. Using Ágata’s analytics, the city could
track changes in population mobility, design targeted interventions to stop the virus’s
spread, and ensure that the health care system could effectively manage its capacity.
Ágata’s work in the health sector has continued beyond the pandemic. Leveraging
electronic health records and predictive analytics, it tracks pregnant women and
newborns in real time to identify health risks. By encouraging preventive health
measures, Ágata’s analytics could save the lives of 870 newborns and 22 expectant
mothers annually.

Ágata’s projects have also had an impact on public financial management.

For example, Ágata developed a platform that offers immediate access to
comprehensive citizen data, including contact details and a record of services
provided by the city (Riaño and Delgado 2024). By cleaning, standardizing, and
integrating data from diverse sources, this platform aims to reduce the time
government organizations need to prepare and use citizen data. By ensuring
up-to-date contact information and cross-referencing each citizen’s service and
financial aid records, the platform aims to create efficiencies in resource allocation
and reduce irregularities in service delivery.

Unless it makes proactive efforts to comprehend the needs of government

organizations, a centralized analytics unit may become isolated from the broader
structure and operations of public administration. This isolation can diminish the
production and use of analytical products or make them less relevant and timely
for decision-making. To mitigate this risk, Ágata conducts workshops with various

52 Data for Better Governance

organizations to understand their needs, assess how analytics could meet those needs,
and determine which data could be used. It also hosts workshops with national
government bodies and other local governments to expand its services and advance
analytics throughout the public sector.

Ágata illustrates how a centralized analytics unit sustained by diverse sources of

funding can advance the analytical journey of an entire government (figure 3.4).
By addressing the demands of different government functions, from health to
public finance, and by remaining responsive to the needs of policy makers, Ágata
has dismantled information silos, enriching the Bogotá government’s analytical
perspective.

FIGURE 3.4 Improving Government Efficiency through a Centralized Analytics Unit,

Colombia

Applications
Citywide program evaluation and design;
monitoring and accountability;
government transparency

Types
Descriptive; diagnostic;
predictive

Analytics
Improve data accessibility
for analytics

Infrastructure Capabilities

Interoperable management information Dedicated analytics unit;

systems; data quality controls; data inventories workshops to advance analytics agenda

Source: Original figure for this publication.

Case Studies of Government Analytics in LAC 53

 CASE STUDY 3.9: ESTABLISHING A DECENTRALIZED
ANALYTICS UNIT TO IMPROVE THE EFFICIENCY OF
TAX COLLECTION (PERU)
Peru’s fiscal situation deteriorated starting in 2014, with tax revenue decreasing
from 18.2 percent of GDP (average 2005–14) to 16.4 percent of GDP (average
2015–21). These levels are below the average for Latin American and Caribbean
countries (21.4 percent) and the member countries of the Organisation for Economic
Co-operation and Development (33.5 percent) in the same period (2015–21).
In addition to economic factors (such as a decline in mining prices and an economic
slowdown in major trading partners), one factor contributing to the decrease in
tax revenues was the difficulty faced by the Superintendencia Nacional de Aduanas
y de Administración Tributaria (National Superintendency of Customs and Tax
Administration, or SUNAT) in identifying tax evasion and underreporting of taxable
activities (Luján Taipe and Sánchez Castro 2022).

SUNAT faced significant challenges in its effort to detect tax evasion and improve tax
reporting. Traditional processes were complex and slow, and the systems and databases
required technological upgrades to meet current needs. To address these challenges,
since 2019, SUNAT has initiated a strategic transformation—including creating a
dedicated analytics unit—to better understand and predict taxpayer behavior.

SUNAT’s application of data analytics is part of a broader set of reforms to digitalize

revenue collection, improve efficiency, and foster a more user-friendly approach.
SUNAT improved taxpayer services by reviewing and simplifying responses to their
inquiries and streamlined the processing of electronic invoices to provide timely
access to transactional data (enhancing voluntary tax compliance). It also developed
and automated its TaxMIS, applying a set of detailed definitions and specifications
for cross-referencing information using key information sources such as electronic
purchase records, electronic sales records, electronic payment receipts, and tax
forms. These broader reforms enabled analysis based on TaxMIS data.

In 2019, SUNAT established a dedicated analytics unit to enhance tax and customs
control and compliance. The unit consisted of five data scientists who developed
analytics to detect evasion and optimize audits using TaxMIS data in combination with
external data sources. Among the first projects developed by the unit to improve tax
compliance were the following:

● Using big data techniques such as web scraping, the unit developed an algorithm to
determine whether e-commerce sellers were registered in the taxpayer registry and
discovered that 57 percent were not. SUNAT sent a text message to these sellers,
resulting in the registration of 320 new sellers and an increase of about $1 million in
sales declarations.
● Using text mining, the unit developed an automated alert system to prevent
­taxpayers from deducting expenses unrelated to their business. The unit’s analysis
revealed that, on average, taxpayers did not record 64 out of every 100 invoices

54 Data for Better Governance

 that the alert system had flagged as personal expenses, protecting them from fines
and audits.
● The unit developed a taxpayer segmentation model to help SUNAT focus its
­collection strategies (Larios, Azuero, and Rojas 2018; SUNAT 2019).
Since 2019, the analytics unit has continued to grow and to develop new predictive
analytics to help SUNAT assess the risk of tax noncompliance. For example, the
unit developed a model to determine the risk that taxpayers lack the operational
capacity to carry out a declared economic activity, helping to prevent the generation
of improper tax credits or deductible expenses. This model has generated $15 million,
and SUNAT has scheduled operational capacity verifications with 1,780 taxpayers
flagged by the model. The unit also developed a model to optimize the allocation of
SUNAT’s resources that identifies high-risk taxpayers and assigns them a priority
level based on the unit’s operational capacity and the probability and impact of
noncompliance (SUNAT 2022). These models have become substantial components of
risk management at SUNAT.

SUNAT’s strategic transformation demonstrates the impact that a dedicated analytics

unit can have within an organization, especially when analytics is part of a broader
reform strategy. By producing analytics closely aligned with operational needs, SUNAT’s
analytics unit has supported the agency’s transformation and Peru’s fiscal sustainability.

CASE STUDY 3.10: DATATHONS AS A TOOL FOR

PROMOTING DATA ANALYTICS (MULTIPLE COUNTRIES)
In recent years, “datathons” have become an important tool within the data science
community for promoting the use of analytics and encouraging the sharing of best
practices, tools, and success stories. This type of event has recently gained traction
in the public sector. At a public sector datathon, public servants and members of the
public are given access to government data, learn about a relevant policy question, and
collaborate to produce analytics that can be used to inform policy making. Participants
work to uncover patterns and trends in the data, demonstrating how analytics can be
used to conceptualize or address problems.

Public sector datathons can benefit governments even beyond the analytical insights
produced during the events. When datathons are open to both public servants and the
wider community, they help governments discover analytical talent, creating a pool
of candidates for future recruitment. Datathons also engage individuals with diverse
skills to tackle policy challenges in a collaborative and informal environment, which
can foster the development of innovative ideas and encourage participants to apply
their newfound insights within their own workplaces. Moreover, datathons signal an
organizational commitment to analytics, which can make public sector data analysts
more satisfied with their jobs and improve retention.

Numerous countries in Latin America and the Caribbean have adopted datathons as a
strategy to encourage analytical experimentation. For instance, in 2022, Ecuador’s Servicio

Case Studies of Government Analytics in LAC 55

 Nacional de Contratación Pública (National Public Procurement Service) organized
a datathon, Rompiendo la Contratación Pública (Disrupting Public Procurement), to
discover innovative applications of data from the e-Procurement system (SERCOP 2022).
The event, which spanned four days, was open to both public servants and the public.
Participants developed analytical products, including interactive visualizations and
statistical models, to understand the factors influencing local purchasing, the role of public
procurement in boosting the local economy, and sectors lacking gender diversity among
suppliers. They also devised alert systems to flag conflicts of interest in public purchases.

Similarly, in 2023, Mexico’s Secretaría Ejecutiva del Sistema Nacional Anticorrupción

(Executive Secretariat of the National Anti-Corruption System) organized a
datathon focused on applying data in the fight against corruption. Participants were
encouraged to develop algorithms using e-Procurement data to detect behavioral
patterns associated with administrative misconduct among public officials and private
organizations, including improper hiring practices, influence peddling, and collusion
in procurement (SESNA 2023).

In Brazil, meanwhile, the Escola Nacional de Administração Pública (National School

of Public Administration) organized a datathon focused on studying racial inequalities
in the public service. Leveraging HRMIS data, participating teams studied the areas
of public policy where Black public servants are employed, the dynamics affecting the
appointment of Black people into leadership roles in the public service, and the main
barriers to Black people’s access to public service positions (ENAP 2023).

As these examples show, datathons represent a valuable tool for fostering the
development of analytical products across different government functions. They
provide a platform through which both public servants and external experts can share
and learn best practices, cultivate talent, and build up organizational identification and
commitment. Datathons also highlight the significance of government analytics for
identifying issues within public administration and proposing solutions, showcasing
the need for more and better analytics to improve public policy.

CASE STUDY 3.11: THE POWER OF DATA

INTEGRATION TO STRENGTHEN THE INTEGRITY
OF PUBLIC PROCUREMENT (BRAZIL)
Corruption significantly undermines development, and public procurement is
exceptionally vulnerable, leading to considerable waste of public resources. Corruption
represents an estimated 8 percent of the global value of procurement contracts,
reaching some $880 billion lost yearly (Bosio 2021), and current methods for detecting
and investigating corruption in procurement are insufficient relative to the problem’s
vast scale. Approaches that rely on manual analysis, follow-up investigations of specific
complaints, and anecdotal evidence are inefficient, because they demand extensive
human and financial resources. But advancements in data collection, digitalization,
and transparency within the public sector have paved the way for data-driven
anticorruption strategies.

56 Data for Better Governance

In this context, the World Bank, in partnership with subnational governments in Brazil,
developed the Governance Risk Assessment System (GRAS). This cutting-edge tool
harnesses analytics to detect fraud, corruption, and collusion in public procurement.
GRAS uses algorithms to identify risk patterns associated with public procurement
fraud and corruption, targeting public suppliers, contracting agencies, and individual
actors. A pillar of GRAS is its robust data integration framework, which aggregates,
links, and analyzes large volumes of public procurement data, including detailed
microdata on firms and individuals (World Bank 2023). This integration framework
makes it easier to identify risk patterns that could not be identified by analyzing a single
data set.

GRAS encompasses 60 red flags and examines risk patterns across four dimensions.
The first dimension focuses on the procurement cycle, and GRAS flags tendering
processes that lack public announcements or in which a single bid is submitted.
The second focuses on interbidder collusion, such as that which can occur when
competitors have a common shareholder or their behavior suggests coordinated
bidding. The third focuses on supplier characteristics, including registration in tax
haven jurisdictions, unusually high profitability, and engaging in multiple economic
activities. Finally, the fourth dimension focuses on the political connections of
suppliers, such as those implied by their contributions to electoral campaigns or
political parties and by their having public officials or politicians as shareholders.

In 2022, the World Bank piloted GRAS in the states of Mato Grosso and Rio de
Janeiro and the municipalities of São Paulo and Porto Alegre. During the pilot, GRAS
identified about 800 firms that had been awarded contracts despite being under legal
sanctions, about 4,000 firms that had been awarded contracts despite being owned by
public servants or having other political connections, and about 1,000 firms that had
won bids against competitors with a common shareholder (World Bank 2023). During
the pilot, GRAS also supported corruption investigations by public prosecutors’ offices.
In one participating state, GRAS-supported analyses helped the federal police uncover
networks of shell companies and a money-laundering operation (World Bank 2023).

These findings underscore the pervasive risks of corruption and conflicts of interest
within public procurement. GRAS demonstrates how public sector corruption can
be detected and reduced by integrating government data, applying descriptive and
predictive analytics, and triangulating data, presenting a promising model for using
analytics to foster integrity and transparency in public procurement.

CASE STUDY 3.12: SUPPORTING THE

IMPLEMENTATION OF FRAMEWORK AGREEMENTS
THROUGH DATA ANALYTICS (URUGUAY)
Public procurement covers diverse purchases, and the best purchasing strategies vary
across different categories of product. One purchasing strategy is to establish framework
agreements, umbrella agreements that establish the terms—including price, quality, and
quantity—under which purchasing organizations can award contracts for a specific

Case Studies of Government Analytics in LAC 57

 product to preapproved suppliers. For example, a country’s central procurement unit
might establish a framework agreement with an information technology company
to purchase computers for the country’s entire public administration for five years.
Framework agreements are efficient purchasing strategies for products that represent
low supply risks and that are purchased in large quantities and across many procuring
organizations. Analytics can help define efficient purchasing strategies for different
products, including by helping prepare framework agreements.

In 2019, a World Bank review of Uruguay’s public procurement data identified

consolidation of purchases as the most significant potential source of savings and
pointed to framework agreements as the most effective instrument for implementing
the required strategy (Cocciolo, Samaddar, and Fazekas 2023).2 Procurement
regulations in Uruguay already covered the use of framework agreements for common
goods and services, making framework agreements an easy strategy to implement
without changes in laws and regulations. In 2021, the World Bank worked with
Uruguay’s Agencia Reguladora de Compras Estatales (Regulatory Agency for Public
Procurement) and the Unidad Centralizada de Adquisiciones (Central Procurement
Unit) within the Ministry of Economy and Finance to build the organizations’
capacity to generate and manage framework agreements and develop pilot framework
agreements for goods and services with the greatest savings potential (Cocciolo,
Samaddar, and Fazekas 2023).

During this review’s implementation phase, analytics was useful for identifying items
that could be purchased through framework agreements. With the support of the
World Bank, the agency developed an algorithm that considers the total procurement
volume of a product per year, the number of procuring organizations purchasing it,
and the number of purchases and procurement procedures per year. The algorithm
has been useful for identifying and giving priority to sectors and products with the
largest potential for efficiency gains and for targeting the next steps in the preparation
of framework agreements (for example, consolidation of purchases and market
analysis). Uruguay’s experience demonstrates the impact of descriptive and diagnostic
analytics for identifying suitable interventions and supporting their design during the
implementation stage of a program.

NOTES
1. More information about the standard can be found on the website of the Open Contracting
Partnership at https://standard.open-contracting.org/latest/en/.
2. “Consolidation of purchases” refers to the practice of grouping purchases of the same product or
service to reduce duplication. Purchases can be consolidated within one procuring organization or
across procuring organizations by aggregating fragmented purchases into larger contracts. The goal is
to improve efficiency by streamlining procurement processes, reducing procedural costs, stimulating
competition, and exploiting economies of scale.

58 Data for Better Governance

REFERENCES
Alcaldía Mayor de Bogotá, Colombia. 2020. “Ágata, la nueva agencia analítica de datos que hará de
Bogotá líder global en transparencia e inteligencia digital para sus ciudadanos.” Noticias, December 14, 2020.
https://www.sdp.gov.co​/n­ oticias/agata​-la-nueva-agencia-analitica-de-datos-hara-de-bogota-lider-global​
-transparencia-e-inteligencia.
Boone, Claire E., Pablo Celhay, Paul Gertler, and Tadeja Gracner. 2023. “Encouraging Preventative Care
to Manage Chronic Disease at Scale.” NBER Working Paper 31643, National Bureau of Economic
Research, Cambridge, MA. https://doi.org/10.3386/w31643.
Boone, Claire E., Pablo Celhay, Paul Gertler, Tadeja Gracner, and Josefina Rodriguez. 2022. “How
Scheduling Systems with Automated Appointment Reminders Improve Health Clinic Efficiency.”
Journal of Health Economics 82: 102598. https://doi.org/10.1016/j.jhealeco.2022.102598.
Bosio, Erica. 2021. “Reducing Corruption in Public Procurement.” Let’s Talk Development (blog), August
17, 2021. https://blogs.worldbank.org/en/developmenttalk/reducing-corruption-public-procurement.
Brockmeyer, Anne. 2019. “Working with Administrative Tax Data: A How-to-Get-Started Guide.”
Macroeconomics, Trade & Investment (MTI) Practice Note 7, World Bank, Washington, DC. https://www​
.annebrockmeyer.com/uploads/1/2/1/4/121485108/mti_practice_note_on_admin_tax_data.pdf.
Carrillo, Paul, Dina Pomeranz, and Monica Singhal. 2017. “Dodging the Taxman: Firm Misreporting and
Limits to Tax Enforcement.” American Economic Journal: Applied Economics 9 (2): 144–64. https://doi​
.org/10.1257​/a­ pp.20140495.
CNEP (Comisión Nacional de Evaluación y Productividad). 2022. Eficiencia en la Gestión de Atención
Primaria de Salud (APS). Santiago, Chile: Comisión Nacional de Evaluación y Productividad.
https://cnep.cl/estudios​-finalizados/eficiencia-en-gestion-de-atencion-primaria-de-la-salud/.
Cocciolo, Serena, Suxhmita Samaddar, and Mihaly Fazekas. 2023. “Government Analytics Using
Procurement Data.” In The Government Analytics Handbook: Leveraging Data to Strengthen Public
Administration, edited by Daniel Rogger and Christian Schuster, chap. 12. Washington, DC:
World Bank. https://doi.org​/10.1596/978-1-4648-1957-5.
ENAP (Escola Nacional de Administração Pública). 2023. “DATATHON Desigualdades Raciais no Serviço
Público.” Notice 119/2023, Process 04600.002764/2023-26, Brasília, Brazil. https://repositorio.enap.gov​
.br​/b
­ itstream/1/7779/1/Edital%20Datathon%20Desigualdades%20Raciais.pdf.
Grennan, Matthew, and Ashley Swanson. 2020. “Transparency and Negotiated Prices: The Value of
Information in Hospital-Supplier Bargaining.” Journal of Political Economy 128 (4): 1234–68.
https://doi.org/10.1086/705329.
Haimovich, Francisco, Emmanuel Vazquez, and Melissa Adelman. 2021. “Scalable Early Warning Systems
for School Dropout Prevention: Evidence from a 4,000-School Randomized Controlled Trial.” Policy
Research Working Paper 9685, World Bank, Washington, DC. https://documents.worldbank.org​
/en/publication/documents-reports/documentdetail/983591622568486300/scalable-early-warning​
-systems-for-school-dropout-prevention-evidence-from-a-4-000-school-randomized-controlled-trial.
Larios, José, Rodrigo Azuero, and Agnes Rojas. 2018. “Utilizando big data para combatir la evasión.”
Recaudando Bienestar (blog), August 29, 2018. https://blogs.iadb.org/gestion-fiscal/es/utilizando-big​
-data​-para-combatir-la-evasion/.
Luján Taipe, Jhonatan Erick, and Lucía Carmen Sánchez Castro. 2022. “Implementando la Transformación
Digital en SUNAT: acercándonos al contribuyente.” MBA thesis, Universidad de Piura. https://hdl​
.handle​.net/11042/5525.
Margozzini, Paula, and Álvaro Passi. 2018. “Encuesta Nacional de Salud, ENS 2016–2017: Un aporte a la
planificación sanitaria y políticas públicas en Chile.” Ars Medica: Revista de Ciencias Médicas 43 (1):
30–34. https://doi.org/10.11565/arsmed.v43i1.1354.
Martinez-Carrasco, Jose, Otavia Conceição, and Ana Luzia Dezolt. 2023. “More Information, Lower
Price? Access to Market-Based Reference Prices and Gains in Public Procurement Efficiency.”
Working Paper IDB-WP-1441, Inter-American Development Bank, Washington, DC. http://dx.doi​
.org/10.18235/0004794.

Case Studies of Government Analytics in LAC 59

 MINEDUC (Ministerio de Educación [de Guatemala]). 2019. Guía para promover la transición
exitosa de sexto primaria a primer grado del ciclo básico. 2nd ed. Guatemala City: MINEDUC.
https://cnbguatemala.org/wiki/Gu%C3%ADa_para_promover_la_transici%C3%B3n_exitosa_de​
_sexto​_Primaria_a_primer_grado_del_Ciclo_B%C3%A1sico.
Montes, Nancy. 2022. Usos de los sistemas de información en el planeamiento y gestión de políticas
educativas en América Latina: Informe regional. Buenos Aires, Argentina: Oficina para América
Latina y el Caribe del Instituto Internacional de Planeamiento de la Educación de la Organización
de las Naciones Unidas para la Educación, la Ciencia y la Cultura. https://unesdoc.unesco.org​
/ark:/48223/pf0000385439.locale=en.
ONSC (Oficina Nacional del Servicio Civil). 2022a. “Análisis de los concursos desiertos de la
Administración Central, 2011–2022.” Documento de Trabajo 03/2022. ONSC, Montevideo, Uruguay.
https://www.gub.uy​/o ­ ficina-nacional-servicio-civil/comunicacion/publicaciones/analisis-concursos​
-desiertos​-administracion​-central-2011-2022.
ONSC (Oficina Nacional del Servicio Civil). 2022b. “Avances en la Plataforma de Gestión Humana del
Estado.” Noticias, August 4, 2022. ONSC, Montevideo, Uruguay. https://www.gub.uy/oficina-nacional​
-servicio-civil​/c­ omunicacion/noticias/avances-plataforma-gestion-humana-del-estado.
ONSC (Oficina Nacional del Servicio Civil). 2022c. “Nueva Plataforma de Gestión Humana del Estado
GHE.uy.” ONSC tv. Streamed live on November 23, 2022. YouTube video, 1:05:25. https://www​
.youtube.com​/w ­ atch?v=u51tU2mxR_E.
ONSC (Oficina Nacional del Servicio Civil). n.d. “Proyecto estratégico: Plataforma de Gestión Humana
del Estado (ghe.uy).” Accessed February 28, 2024. https://www.gub.uy/oficina-nacional-servicio-civil​
/politicas​-y-gestion/proyectos/plataforma-gestion-humana-del-estado-gheuy.
Pomeranz, Dina, and José Vila-Belda. 2019. “Taking State-Capacity Research to the Field: Insights from
Collaborations with Tax Authorities.” Annual Review of Economics 11: 755–81. https://doi.org/10.1146​
/­annurev​-economics-080218-030312.
reAcción Paraguay. 2023. “FOCO.” Last modified February 19, 2023. https://foco.reaccion.org.py/.
reAcción Paraguay. n.d. “reAcción.” Accessed July 17, 2024. https://reaccion.org.py/.
Riaño, Manuel, and Alejandro Delgado. 2024. Yo, IA Innovación gubernamental: La era de la inteligencia
artificial. Barcelona: Bronce.
Riveros García, David. 2019. “Politics, Technology, and Accountability: The Transparency Façade of Open
Government Data Reforms in Paraguay.” eJournal of eDemocracy and Open Government (JeDEM)
11 (2): 60–93. https://doi.org/10.29379/jedem.v11i2.537.
SERCOP (Servicio Nacional de Contratación Pública). 2022. “Datatón–Rompiendo la Contratación
Pública.” SERCOP, Quito, Ecuador. https://dataton.compraspublicas.gob.ec/wp-content/uploads​
/2022/04/Metodologia​-Evento-Dataton_vf-2.pdf.
SESNA (Secretaría Ejecutiva del Sistema Nacional Anticorrupción). 2023. “Datatón Anticorrupción
2023: Convocatoria.” SESNA, Mexico City. https://dataton2023.plataformadigitalnacional.org/files​
/­datatonanticorrupcion2023-basesv1_3.pdf.
SUNAT (Superintendencia Nacional de Aduanas y de Administración Tributaria). 2019. “Memoria
Anual 2019: Resumen Ejecutivo.” SUNAT, Lima, Peru. https://www.sunat.gob.pe/cuentassunat​
/planestrategico/memoria​/e­ jecutiva/memoriaEje-2019.pdf.
SUNAT (Superintendencia Nacional de Aduanas y de Administración Tributaria). 2022. Memoria
Institucional 2022. Lima, Peru: SUNAT. https://www.sunat.gob.pe/cuentassunat/planestrategico​
/memoria​/memoria2022.pdf.
UCD (Unidad de Científicos de Datos). 2019. “Dirección de Desarrollo Digital.” Departamento Nacional
de Planeación, Bogotá, Colombia. https://colaboracion.dnp.gov.co/CDT/Desarrollo%20Digital/UCD​
/­Proyectos/2019_24_Brecha_salarial/Brecha_salarial_Informe.pdf.
World Bank. 2018. “Paraguay: Systematic Country Diagnostic.” World Bank, Washington, DC.
https://documents1.worldbank.org/curated/en/827731530819395899/pdf/Paraguay-SCD-06292018.pdf.
World Bank. 2023. Governance Risk Assessment System (GRAS): Advanced Data Analytics for Detecting
Fraud, Corruption, and Collusion in Public Expenditures. Equitable Growth, Finance and Institutions
Insight—Governance. Washington, DC: World Bank. https://openknowledge.worldbank.org/handle​
/10986/40640.

60 Data for Better Governance

CHAPTER 4

A Regional Assessment
of Government Analytics
in Latin America and the
Caribbean

INTRODUCTION
This chapter offers a comprehensive assessment of government analytics and its enabling
conditions in Latin America and the Caribbean, based on data from an original survey
conducted in 20 countries (map 4.1). In each country, public servants who are experts in
core government functions and their respective management information systems (MISs)
were interviewed, making it possible to assess how different kinds of administrative data are
used to address policy-making challenges. For this reason, most of the analysis is presented
at the information system level.1 The survey targeted MISs associated with core government
functions: education (education management information system, or EdMIS), health (health
management information system, or HealthMIS), human resources (human resources
management information system, or HRMIS), procurement (e-Procurement), public
finance (public financial management information system, or PFMIS), and taxation (tax
management information system, or TaxMIS). Country-level data on government initiatives
for cultivating the analytical capabilities of public servants were also collected. Survey
data are analyzed according to the conceptual framework introduced in chapter 2: first,
the chapter describes what government analytics looks like in the region, then it analyzes
governments’ organizational and individual capabilities, as well as their data infrastructures.

The regional assessment highlights the ways governments across Latin America and the
Caribbean are already leveraging administrative data for analytics as well as policy areas
where analytics can help address further challenges. It shows that governments in the region
predominantly use administrative data to produce descriptive analytics for operational and
transactional purposes, like monitoring and accountability. This means governments are
missing out on opportunities to use advanced analytics to improve decision-making, design
more effective and efficient public policies, and strengthen public sector functioning and
service delivery.
A reproducibility package is available for this book in the Reproducible Research Repository at
https://reproducibility.worldbank.org/index.php/catalog/209.
61
MAP 4.1 Countries in Latin America and the Caribbean That Participated in the Survey

The Bahamas

Colombia
Ecuador

Brazil
Peru
Dominican Bolivia
Republic
Belize
Dominica
Honduras Jamaica Chile
Paraguay
St. Vincent and Barbados
Guatemala Costa Rica the Grenadines
Trinidad and
Panama Tobago
Uruguay

Source: Original map for this publication.

Strengthening the enabling conditions for government analytics is key to capitalizing

on these opportunities. For now, governments in Latin America and the Caribbean
face significant challenges in building analytical capabilities within government
organizations. Although they acknowledge the importance of analytical skills and in
some cases offer training programs for analytics, governments lack structured career
tracks for data analysts, proficiency evaluations on data analytics for public servants,
funding for analytical projects, and an overall strategy for moving public administration
toward evidence-based policy making. But some governments are also pursuing two
promising strategies for addressing limited analytical capabilities: collaborating with
academic or international organizations and establishing dedicated analytics units.

Governments in the region also face data infrastructure challenges that prevent them
from fully unlocking the transformative potential of digital MISs and catalyzing the
value of the investments they have already made. As the case studies in chapter 3
demonstrate, analytics can have the biggest impact on policy making when government
data from different sources are combined. Despite the progress they have made in
digitalization, many governments in the region still contend with MISs that are limited
in data comprehensiveness and functionality, fragmented systems, informal data access
protocols, and limited quality control measures. Governments need to improve the
functionality, interoperability, and data quality of MISs so these systems can better
support informed decision-making.

Finally, the regional assessment reveals that analytics, capabilities, and infrastructure
vary significantly, not only across countries but also across government functions,
suggesting that governments lack a systematic approach to analytics for the whole

62 Data for Better Governance

of public administration. Although analytical products have the greatest impact
when they respond to specific policy needs, governments can avoid fragmented and
duplicated efforts by developing government-wide enabling ecosystems for analytics.
As the survey results show, this kind of approach is especially needed to ensure the
interoperability of MISs, build analytical capabilities, and create a culture for evidence-
based decision-making in public administration. By breaking down how the different
elements of the conceptual framework vary across MISs, this chapter offers evidence to
help governments identify opportunities and weaknesses for themselves and determine
how to develop enabling ecosystems within their administrations.

Throughout this chapter, the voices of the

“ public servant experts that were surveyed are

highlighted in blue speech bubbles like this one.

SURVEY METHODOLOGY
As noted earlier, the Government Analytics Survey in Latin America and the Caribbean
was conducted in 20 countries. It consisted of seven questionnaires (presented in
online appendix D). Six individual MIS-level questionnaires focused separately on
each type of MIS (EdMIS, HealthMIS, HRMIS, e-Procurement, PFMIS, and TaxMIS).
One additional country-level questionnaire, the “capabilities questionnaire,” focused
on government initiatives for cultivating the analytical capabilities of public servants.
More than 100 government officials responded to the questionnaires, providing
insights based on their specific areas of expertise.

The capabilities questionnaire was typically completed by the country’s digital

government agency (n = 16 countries). MIS questionnaires were completed by
different public sector organizations, depending on the government function
involved. The EdMIS questionnaire was typically completed by the country’s ministry
of education (n = 13 countries), the e-Procurement questionnaire by the country’s
procurement agency (n = 11 countries), the HealthMIS questionnaire by the ministry
of health (n = 14 countries), the HRMIS questionnaire by the ministry of civil
service (n = 14 countries), the PFMIS questionnaire by the ministry of finance or
economy (n = 16 countries), and the TaxMIS questionnaire by the government’s tax
authority (n = 17 countries). For further information on the survey methodology,
refer to appendix A.

A Regional Assessment of Government Analytics IN LAC 63

 WHAT DOES GOVERNMENT ANALYTICS LOOK LIKE IN
LATIN AMERICA AND THE CARIBBEAN?
What Types of Analytics Do Governments Conduct?

Governments in Latin America and the Caribbean primarily

use administrative data for descriptive analytics. In the region,
96 percent of MISs are used for descriptive analytics, and
this trend is consistent across all the government functions
considered in the survey (figure 4.1). As noted in the conceptual
framework in chapter 2, descriptive analytics allows government
organizations to understand the current state of their operations
and summarize data for tracking and reporting. Diagnostic analytics is the second-
most-used type of analytics (55 percent of MISs) in the region, and predictive analytics
is employed a little less often (50 percent of MISs).

FIGURE 4.1 Types of Analytics, by Type of MIS

a. HRMIS b. e-Procurement c. PFMIS

Descriptive 100 Descriptive 100 Descriptive 86

Diagnostic 29 Diagnostic 40 Diagnostic 64

Predictive 21 Predictive 30 Predictive 36

0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100

Percent Percent Percent
d. TaxMIS e. EdMIS f. HealthMIS

Descriptive 94 Descriptive 100 Descriptive 100

Diagnostic 65 Diagnostic 54 Diagnostic 75

Predictive 82 Predictive 54 Predictive 67

0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100

Percent Percent Percent

Source: Government Analytics Survey in Latin America and the Caribbean.

Note: The figure shows, by type of MIS, the proportion of systems in the region used for each type of
analytics. The results for each type of MIS do not total 100 percent because the question allowed for
multiple responses. N = 85 country-MISs, distributed as follows: EdMIS questionnaire (n = 13 countries);
e-Procurement questionnaire (n = 11 countries); HealthMIS questionnaire (n = 14 countries); HRMIS questionnaire
(n = 14 countries); PFMIS questionnaire (n = 16 countries); TaxMIS questionnaire (n = 17 countries). EdMIS =
education management information system; e-Procurement = procurement management information system;
HealthMIS = health management information system; HRMIS = human resources management information
system; MIS = management information system; PFMIS = public financial management information system;
TaxMIS = tax management information system.

64 Data for Better Governance

 “Efforts are being made to incorporate predictive

“ models in health care; however, the main problem

is that there are no systems in place to generate
these models.”
– HealthMIS expert

The types of analytics governments undertake differ across MISs and among countries.
Descriptive analytics is the most common, as illustrated in figure 4.2. Diagnostic
analytics alone is the least common: only in one country is the PFMIS used for
diagnostic analytics alone. Data from TaxMISs, EdMISs, and HealthMISs are used
for more complex types of analytics (combinations of descriptive, diagnostic, and
predictive analytics) in numerous countries, suggesting a more sophisticated approach
to analytics in these government functions. For example, predictive analytics most
often draws on tax data (primarily to generate tax revenue forecasts) and health data
(for instance, to predict the likelihood of disease spread and support planning for
prevention campaigns). Data from HRMISs, on the other hand, are primarily used
for less complex descriptive analytics. These patterns suggest that governments in the
region do not yet use their administrative data and analytics tools to their full potential.
More steps can be taken to use administrative data, especially those from HRMISs,
PFMISs, and e-Procurement systems, for strategic applications based on diagnostic
and predictive analytics: for instance, to identify the root causes of performance gaps,
support forecasting, and inform policy design.

FIGURE 4.2 Types of Analytics, by Country and Type of MIS

Country 1
Country 2
South Country 3
America
Country 4
Country 5
Country 6
Central
America Country 7
and the Country 8
Caribbean Country 9
HRMIS e-Procurement PFMIS TaxMIS EdMIS HealthMIS
Type of analytics: Descriptive Diagnostic Descriptive, diagnostic Descriptive, diagnostic, predictive

Source: Government Analytics Survey in Latin America and the Caribbean.

Note: The figure shows the types of analytics conducted across MISs and countries in the region. The data
presented include only countries that completed all six MIS questionnaires. N = 54 country-MISs. EdMIS =
education management information system; e-Procurement = procurement management information system;
HealthMIS = health management information system; HRMIS = human resources management information
system; MIS = management information system; PFMIS = public financial management information system;
TaxMIS = tax management information system.

A Regional Assessment of Government Analytics IN LAC 65

 Just as not all MISs are used for all types of analytics, not all data elements within
MISs are used equally. Although most available data within each type of MIS seem to
be used for some type of analytics, the specific data elements that are used vary, and
distinct patterns emerge across each type of MIS (figure 4.3). For instance, within
HRMISs, data on employment status and personnel characteristics are extensively

FIGURE 4.3 Data Elements Used for Analytics, by Type of MIS

a. HRMIS b. e-Procurement c. PFMIS

Employment 100 Contract Revenue 100

status price 100 information
Personnel 93 Contract Budgeting 93
characteristics 90
information
Wages and 79 Item Cash 93
benefits 90 information
information
Career Bid Expenditures
management 57 80 64
information
Skills and 50 Entity 80 Public debt 64
talent information
Attitudes and 7 Complaint 50 Asset 36
motivations information information
0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100
Percent Percent Percent
d. TaxMIS e. EdMIS f. HealthMIS

Revenue 100 Infrastructure 92 Health 100

services
Taxation and 100 Teacher 85 Health 100
levies information facilities
Taxpayer 94 Human 85 Vaccination 100
information resources rates

Tax returns 88 Student Disease 100

information 85 surveillance

Compliance 82 Academic 77 Demographics 92

records
Taxpayer
finance 82 Curriculum 46 Health 83
information outcomes
0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100
Percent Percent Percent

Source: Government Analytics Survey in Latin America and the Caribbean.

Note: The figure shows, by type of MIS, the proportion of systems in the region that uses each type of data
element for analytics. The data presented include only countries with the specified data elements in their MISs.
The results for each type of MIS do not total 100 percent because the question allowed for multiple responses.
N = 85 country-MISs, distributed as follows: EdMIS questionnaire (n = 13 countries); e-Procurement questionnaire
(n = 11 countries); HealthMIS questionnaire (n = 14 countries); HRMIS questionnaire (n = 14 countries); PFMIS
questionnaire (n = 16 countries); TaxMIS questionnaire (n = 17 countries). EdMIS = education management
information system; e-Procurement = procurement management information system; HealthMIS =
health management information system; HRMIS = human resources management information system;
MIS = management information system; PFMIS = public financial management information system; TaxMIS = tax
management information system.

66 Data for Better Governance

 “Although we cover the six data elements described,

“ it should be clarified that they are found in

different systems.”
– EdMIS expert

used for analytics (100 and 93 percent of HRMISs, respectively), mainly for reporting
purposes. Data on employees’ skills and talents and on their attitudes and motivations
are comparatively underutilized (50 and 7 percent of HRMISs, respectively). Likewise,
within e-Procurement systems, data on complaints are the least used (50 percent of
e-Procurement systems), whereas data on contract prices are extensively used (100
percent of e-Procurement systems). Within PFMISs, data on state-owned assets
are highly underutilized for analytics (36 percent of PFMISs), whereas revenue
information and budgeting data are extensively used (100 percent and 93 percent
of PFMISs, respectively). Finally, within EdMISs, curriculum data and students’
academic records are the least used for analytics (46 and 77 percent of EdMISs,
respectively).

Why do governments use some administrative data elements for analytics more than
others? This trend points toward limitations in the enabling conditions for government
analytics. Data elements may not be readily accessible owing to limited data sharing
among government organizations and weak interoperability among information
systems. Even if data elements are accessible, infrastructure limitations may constrain
their use. For example, human resources data on employees’ attitudes and motivations
usually come from surveys of public servants, which are often not integrated into
HRMISs more broadly. Similarly, complaints data are often not integrated into standard
e-Procurement systems and instead require integration with a separate complaints
database. These challenges are described further in the discussion of data infrastructure
later in the chapter.

How Do Governments Apply Analytics?

Government organizations in Latin America and the Caribbean mainly use analytical
products based on administrative data for monitoring and accountability and to
increase government transparency to citizens. More than 93 percent of respondents
to the survey identified monitoring as the main application of analytics, 81 percent
identified citizen transparency, and 79 percent mentioned accountability (figure 4.4).
These findings align with the widespread adoption of control dashboards and open
data initiatives across the region.

A Regional Assessment of Government Analytics IN LAC 67

FIGURE 4.4 Applications of Analytical Products

Monitoring 93
Transparency
81
toward citizens
Accountability 79

Policy design 74

Policy evaluation 73

0 10 20 30 40 50 60 70 80 90 100
Percent

Source: Government Analytics Survey in Latin America and the Caribbean.

Note: The figure shows the proportion of MISs in the region that are used for each application. The results
do not total 100 percent because the question allowed for multiple responses. N = 85 country-MISs.
MISs = management information systems.

Governments in the region use administrative data analytics less frequently for policy
evaluation and design; however, for some government functions, this difference
is not substantial (figure 4.5). The widespread use of analytics for monitoring and
accountability represents a commendable first step toward a more accountable public
administration. Likewise, the more limited use of analytics for policy evaluation
and design represents an opportunity to use administrative data more strategically
to improve government functioning by expanding the scope of these applications of
analytics.

How Is Government Analytics Distributed?

As noted in the conceptual framework in chapter 2, analytics production can be

ad hoc, bureaucratic, or strategic. Most respondents to the survey described analytics
production in their organizations as strategic, meaning that their organizations
produce analytics continuously in response to an overall strategic plan. The resulting
analytical products often take the form of data dashboards for control purposes.
However, not all MIS analytics production follows this pattern. HRMIS analytics
production is primarily bureaucratic, reflecting a focus on regulatory compliance rather
than strategic insights (figure 4.6).

68 Data for Better Governance

FIGURE 4.5 Applications of Analytical Products, by Type of MIS

a. HRMIS b. e-Procurement c. PFMIS

Monitoring 79 Monitoring 90 Monitoring 93

Accountability 71 Accountability 70 Accountability 86

Transparency 50 Transparency 100 Transparency 93

toward citizens toward citizens toward citizens
Policy 64 Policy 80 Policy 71
evaluation evaluation evaluation
Policy design 57 Policy design 70 Policy design 64

0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100

Percent Percent Percent
d. TaxMIS e. EdMIS f. HealthMIS

Monitoring 100 Monitoring 92 Monitoring 100

Accountability Accountability Accountability
88 67 85
Transparency Transparency Transparency
toward citizens 82 toward citizens 75 toward citizens 92
Policy Policy Policy
evaluation 65 evaluation 75 evaluation 85

Policy design 82 Policy design 92 Policy design 77

0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100

Percent Percent Percent

Source: Government Analytics Survey in Latin America and the Caribbean.

Note: The figure shows, by type of MIS, the proportion of systems in the region that are used for each
application. The results for each type of MIS do not total 100 percent because the question allowed for
multiple responses. N = 85 country-MISs, distributed as follows: EdMIS questionnaire (n = 13 countries);
e-Procurement questionnaire (n = 11 countries); HealthMIS questionnaire (n = 14 countries); HRMIS
questionnaire (n = 14 countries); PFMIS questionnaire (n = 16 countries); TaxMIS questionnaire (n = 17
countries). EdMIS = education management information system; e-Procurement = procurement management
information system; HealthMIS = health management information system; HRMIS = human resources
management information system; MIS = management information system; PFMIS = public financial
management information system; TaxMIS = tax management information system.

A Regional Assessment of Government Analytics IN LAC 69

FIGURE 4.6 Kinds of Analytical Production, by Type of MIS

Percent
80 73 75

70
62
60 56 54
50 46

40 33
31 31 31
30 23 25
20
20 15
11
7 8
10
0
0
HRMIS e-Procurement PFMIS TaxMIS EdMIS HealthMIS
Ad hoc Bureaucratic Strategic

Source: Government Analytics Survey in Latin America and the Caribbean.

Note: The figure shows, by type of MIS, the proportion of systems in the region that produce analytical products
in an ad hoc, bureaucratic, or strategic manner. N = 85 country-MISs, distributed as follows: EdMIS questionnaire
(n = 13 countries); e-Procurement questionnaire (n = 11 countries); HealthMIS questionnaire (n = 14 countries);
HRMIS questionnaire (n = 14 countries); PFMIS questionnaire (n = 16 countries); TaxMIS questionnaire (n = 17
countries). EdMIS = education management information system; e-Procurement = procurement management
information system; HealthMIS = health management information system; HRMIS = human resources
management information system; MIS = management information system; PFMIS = public financial management
information system; TaxMIS = tax management information system.

WHAT ANALYTICAL CAPABILITIES DO GOVERNMENTS

IN THE REGION POSSESS?
As described in the conceptual framework in chapter 2,
analytical capabilities—including adequate staffing, career and
skill development, and funding opportunities—are an essential
enabling condition for government analytics. The results of
the survey suggest that governments in Latin America and the
Caribbean have significant room to develop their analytical
capabilities.

Governments in the region face a shortage of structured career development

opportunities in analytics, which are essential for attracting and retaining skilled data
analysts in the public sector. Only 12 percent of governments have a dedicated career
track for data analysts (figure 4.7). For this reason, many digital MIS projects in the
region suffer a progressive loss of human resources to other projects that offer better
career prospects or higher salaries (Porrúa et al. 2021). When governments lack career
incentives for data analysts, they can significantly undermine the investments they have
made in recruiting and training these skilled personnel (OECD 2014).

70 Data for Better Governance

FIGURE 4.7 Career Tracks for and Training and Assessments on Analytics

Percent
70
62
60

50

40

30 25

20
12
10

0
Is there a career Are there Are there trainings
track for data assessments of data related to data
analytics? analytics skills? analytics?

Source: Government Analytics Survey in Latin America and the Caribbean.

Note: The figure shows the proportion of surveyed countries in Latin America and the Caribbean
for which the answer to each question is yes. n = 16 countries.

Governments in Latin America and the Caribbean also offer few options for capacity
assessment and data analytics training (figure 4.7), and the options that do exist are
not closely integrated with a strategic plan or work program on government analytics.
Of the countries surveyed, 62 percent report that they offer training programs to
strengthen the analytical capabilities of the public sector workforce. The range of
these training options is wide, encompassing different formats, durations, content,
and objectives. However, only 25 percent of the countries surveyed assess their staffs’
proficiency in data analytics. In addition, existing training programs were described
by the digital government experts surveyed as sporadic and sometimes as lacking a
cohesive structure and a clear pathway for applying newly acquired skills in existing
work programs. To be effective, capacity-building programs should be designed based
on an assessment of the analytical skills that already exist in the public sector workforce
in relation to medium- and long-term needs for analytical expertise. Governments that
do not assess the skills public servants have and compare them with strategic goals risk
offering training that does not reflect the skills public servants need.

“Data analyst positions exist throughout the

“ administration, but there is no specifically designed

career progression for these roles.”
– Digital government expert

A Regional Assessment of Government Analytics IN LAC 71

 As these findings underscore, governments in the region need more structured efforts
to cultivate analytical expertise within government organizations, as well as a more
strategic approach to establish analytics production as a core work stream in public
administration. Organizations need stable, dedicated funding to produce analytical
products, establish workflows that closely connect policy needs to analytics production,
and build analytical capabilities within government organizations. However, many
organizations face significant challenges in securing consistent funding for analytical
initiatives, because resources are limited and priorities often shift rapidly. Funding for
analytics must be balanced against competing needs in the general government budget.

To mitigate these challenges, organizations should secure funding for analytics through
careful consideration, strategic planning, and a long-term vision. Only 33 percent of
surveyed experts reported that their governments offer internal funding opportunities
for analytical initiatives. As figure 4.8 illustrates, internal funding opportunities appear
more widespread for analytical projects related to education and health (46 and
43 percent of MISs, respectively).

FIGURE 4.8 Opportunities for Internal Funding and Collaboration with Academics,
Nonprofits, or Multilateral Organizations on Analytics Projects, by Type of MIS

Percent
70
62
60

50 46
43 43
40 36 36 35 35
31
30
21 21
19
20

10

0
HRMIS e-Procurement PFMIS TaxMIS EdMIS HealthMIS
Internal funding Collaboration with external partners

Source: Government Analytics Survey in Latin America and the Caribbean.

Note: The figure shows, by type of MIS, the proportion of systems in the region that have internal funding
opportunities available for developing analytical products or that have collaborations with external partners on
data analytics. N = 85 country-MISs, distributed as follows: EdMIS questionnaire (n = 13 countries); e-Procurement
questionnaire (n = 11 countries); HealthMIS questionnaire (n = 14 countries); HRMIS questionnaire (n = 14
countries); PFMIS questionnaire (n = 16 countries); TaxMIS questionnaire (n = 17 countries). EdMIS = education
management information system; e-Procurement = procurement management information system; HealthMIS =
health management information system; HRMIS = human resources management information system; MIS =
management information system; PFMIS = public financial management information system; TaxMIS = tax
management information system.

72 Data for Better Governance

 “Public budgets are limited, so funding gaps are

“ identified to request alignment and support through

international cooperation.”
– PFMIS expert

Support from international organizations and donors can help fill these gaps. The
primary driver of collaboration on data analytics with external partners is a need for
technical assistance (figure 4.9), suggesting a strategic opportunity for government
organizations to leverage external expertise to enhance analytics, particularly where
internal resources and capabilities are limited. However, the same government
functions that are more likely to receive internal funding are also more likely
to collaborate strategically on analytical projects with academics, nonprofits, or
multilateral organizations. Internal funding opportunities and external collaborations
for analytical projects based on HRMIS, e-Procurement, PFMIS, or TaxMIS data
appear to be very limited (less than 36 percent of respondents reported them), as
shown in figure 4.8.

FIGURE 4.9 Drivers of Collaboration on Analytics with Academics, Nonprofits, or

Multilateral Organizations

Technical assistance 91

Data access 68

Innovation in data analytics 50

Financial support 48

0 10 20 30 40 50 60 70 80 90 100
Percent

Source: Government Analytics Survey in Latin America and the Caribbean.

Note: The figure shows the proportion of MISs in the region for which each described category is a driver of
collaboration on data analytics with external partners. The results do not total 100 percent because the question
allowed for multiple responses. N = 85 country-MISs. MIS = management information system.

“We collaborate on request, but not as part of a

“ strategy.”
– TaxMIS expert

A Regional Assessment of Government Analytics IN LAC 73

 “The institutional budget does not include a specific

“ item for research or data analysis; however, it does

include the payment of salaries for the institutional staff,
which means that there is indirect budgetary support.”
– PFMIS expert

Governments in the region can also systematize the use of analytics in decision-
making by incorporating dedicated analytics units into the organizational structure of
public administration. Analytics units can help the government attract data analysts;
establish a consistent, long-term analytical work stream; and secure regular funding
through the payroll of dedicated staff. Nearly 80 percent of the experts surveyed
reported that their governments have specialized units tasked with producing analytics
using administrative data. However, many respondents noted that these units are not
exclusively dedicated to analytics, as they split their working time between analytical
tasks and operational functions as needed. Only a few respondents mentioned an
innovation lab or a dedicated analytics team that focuses full-time on producing
analytics. Relying on part-time teams for analytical tasks can lead to suboptimal
outcomes because these teams might assign immediate operational needs a higher
priority than broader analytical objectives, missing opportunities for innovation,
optimization, and long-term planning for evidence-based policy making.

WHAT IS THE STATE OF DATA INFRASTRUCTURE IN

THE REGION?
As described in the conceptual framework in chapter 2, data
infrastructure is also a key enabling condition for government
analytics. Outdated or inadequate data infrastructure limits data
quality and accessibility, making it difficult for organizations to
use administrative data for analytics and policy making. Data
infrastructure problems hinder the development of analytics in
Latin America and the Caribbean: although the region is a global pioneer in establishing
MISs (as discussed in chapter 1), 61 percent of the MIS experts surveyed reported that
their MISs are not fully digitalized.
Levels of MIS digitalization vary substantially by government function (figure 4.10).
TaxMISs and PFMISs exhibit a relatively high degree of digitalization (71 and 50 percent
of these types of MIS, respectively, are fully digitalized), whereas HealthMISs have the
lowest degree of digitalization. In the health sector, only 8 percent of respondents reported
that their HealthMIS is fully digitalized. According to MIS experts, some information is
still recorded in Word documents, PDFs, and Excel spreadsheets. Levels of digitalization
also vary significantly across different types of MIS within a single country (figure 4.11).
Although some countries have more fully digitalized MISs than others, TaxMISs are
consistently the most advanced.
74 Data for Better Governance
FIGURE 4.10 Fully Digitalized Systems, by Type of MIS

TaxMIS 71

PFMIS 50

EdMIS 46

e-Procurement 30

HRMIS 21

HealthMIS 8

0 10 20 30 40 50 60 70 80
Percent

Source: Government Analytics Survey in Latin America and the Caribbean.

Note: The figure shows, by type of MIS, the proportion of systems in the region that are fully digitalized.
N = 85 country-MISs, distributed as follows: EdMIS questionnaire (n = 13 countries); e-Procurement
questionnaire (n = 11 countries); HealthMIS questionnaire (n = 14 countries); HRMIS questionnaire (n = 14
countries); PFMIS questionnaire (n = 16 countries); TaxMIS questionnaire (n = 17 countries). EdMIS = education
management information system; e-Procurement = procurement management information system; HealthMIS
= health management information system; HRMIS = human resources management information system;
MIS = management information system; PFMIS = public financial management information system; TaxMIS = tax
management information system.

FIGURE 4.11 Fully Digitalized Systems, by Country and Type of MIS

Country 1
Country 2
South
Country 3
America
Country 4
Country 5
Country 6
Central
America Country 7
and the Country 8
Caribbean Country 9
HRMIS e-Procurement PFMIS TaxMIS EdMIS HealthMIS
Fully digitalized: Yes No

Source: Government Analytics Survey in Latin America and the Caribbean.

Note: The figure shows the level of system digitalization across MISs and countries in the region. The data
presented include only countries that completed all six MIS questionnaires. n = 54 country-MISs. EdMIS =
education management information system; e-Procurement = procurement management information system;
HealthMIS = health management information system; HRMIS = human resources management information
system; MIS = management information system; PFMIS = public financial management information system;
TaxMIS = tax management information system.

A Regional Assessment of Government Analytics IN LAC 75

 “e-Procurement is partially digitalized: Word documents,

“ PDFs, images, Excel, and CSV [comma-separated values]

files are still used.”
– e-Procurement expert

Why is digitalization limited to some government functions? MIS experts surveyed

indicated that data fragmentation and isolated systems constrain digitalization: for
example, when health records are kept in physical form, steps in the procurement cycle
happen outside the government’s e-Procurement system, or payroll information and
data on performance evaluations and career trajectories are kept in separate systems.
Fragmentation might also occur if different ministries use different procurement
systems, or if different hospitals have different portals to record information on services
provided. When systems are fragmented, various subsystems operate independently,
and some are digitalized but others are not. This makes it virtually impossible to create
a centralized, fully digitalized MIS with detailed, accurate, and updated information,
preventing governments from effectively leveraging administrative data.

Fragmentation also means government organizations are unable to access and use
each other’s administrative data, preventing them from combining data from multiple
sources to generate new evidence. Data sharing between government organizations
is constrained by the inadequacy of formal access protocols for administrative data.
These protocols serve two purposes: they help uphold trust in the agency responsible
for producing the data, and they ensure uninterrupted access to the data over time.

“The Ministry of Public Health has

“ 103 information systems, most of which are

not interoperable.”
– HealthMIS expert

“The MIS is composed of various modules for

managing the human talent subsystems; however,
some of these modules to date are still not fully
automated or digitalized.”
– HRMIS expert

76 Data for Better Governance

 “There is no single protocol; each collaboration

“ derives from an agreement that regulates the rules

of relationship, and we have policies and standards
for the use and handling of data.”
– HealthMIS expert

“There is a verbal system in place [for data access],

but no formal written procedure.”
– HealthMIS expert

Although they do exist—57 percent of MIS experts reported that there are formal
protocols governing access to administrative data—many of these protocols appear
to be ad hoc arrangements between the organization that manages an MIS and those
seeking data access. These ad hoc agreements often lack explicit and standardized
guidelines and requirements for and restrictions on data sharing.

According to some MIS experts, data accessibility is facilitated by laws governing open
data initiatives. These laws, however, are typically designed with external consumers
in mind, not internal government actors. Consequently, data provided through
these initiatives may not be at a level of granularity sufficient to generate valuable
insights about the functioning of public administration. In contrast, according to the
Organisation for Economic Co-operation and Development’s Digital Government
Index, 96 percent of the organization’s member countries outside Latin America and
the Caribbean report having an explicit formal requirement for data sharing between
government organizations (OECD 2019).2 Without more formal protocols in Latin
America and the Caribbean, data may be accessible only to those agencies that have the
political leverage to negotiate for access.

However, just as is the case with analytical applications and capabilities, formal access
protocols vary across different government functions. As figure 4.12 illustrates, access
protocols seem to be more prevalent for HRMISs (86 percent), e-Procurement systems
(73 percent), and TaxMISs (71 percent), whereas they are less common for EdMISs and
HealthMISs (38 percent and 14 percent, respectively). This pattern also holds within
countries: even if most MISs in a country have formal protocols for managing and
sharing data, HealthMISs and EdMISs are exceptions (figure 4.13). This variation may
be due to different regulatory frameworks, levels of data fragmentation, organizational
priorities, or levels of stakeholder engagement across the agencies managing these
information systems. For HealthMISs, the absence of formal access protocols may also
be caused by their low degree of digitalization, as noted earlier.

A Regional Assessment of Government Analytics IN LAC 77

FIGURE 4.12 Formal Data Access Protocols, by Type of MIS

HRMIS 86

e-Procurement 73

TaxMIS 71

PFMIS 62

EdMIS 38

HealthMIS 14

0 10 20 30 40 50 60 70 80 90
Percent

Source: Government Analytics Survey in Latin America and the Caribbean.

Note: The figure shows, by type of MIS, the proportion of systems in the region that have formal protocols for
accessing system data. N = 85 country-MISs, distributed as follows: EdMIS questionnaire (n = 13 countries);
e-Procurement questionnaire (n = 11 countries); HealthMIS questionnaire (n = 14 countries); HRMIS questionnaire
(n = 14 countries); PFMIS questionnaire (n = 16 countries); TaxMIS questionnaire (n = 17 countries). EdMIS =
education management information system; e-Procurement = procurement management information system;
HealthMIS = health management information system; HRMIS = human resources management information
system; MIS = management information system; PFMIS = public financial management information system;
TaxMIS = tax management information system.

FIGURE 4.13 Formal Data Access Protocols, by Country and Type of MIS

Country 1
Country 2
South
Country 3
America
Country 4
Country 5
Country 6
Central
America Country 7
and the Country 8
Caribbean Country 9
HRMIS e-Procurement PFMIS TaxMIS EdMIS HealthMIS
Data access protocols: Yes No

Source: Government Analytics Survey in Latin America and the Caribbean.

Note: The figure shows the existence of formal protocols for accessing system data across MISs and countries
in the region. The data presented include only countries that completed all six MIS questionnaires. n = 54
country-MISs. EdMIS = education management information system; e-Procurement = procurement management
information system; HealthMIS = health management information system; HRMIS = human resources
management information system; MIS = management information system; PFMIS = public financial management
information system; TaxMIS = tax management information system.

78 Data for Better Governance

 “The HealthMIS is only partially digitalized due to the

“ existence of some physical forms that are not integrated

into the system.”
– HealthMIS expert

Data accessibility is also limited when organizations lack systematized information

on the data available in information systems. According to more than half (58
percent) of MIS experts surveyed, comprehensive data inventories, which list all
available data elements within an MIS, do not exist for the systems they work with.
Without data inventories, organizations struggle to determine what types of data are
available and where they are stored, which makes it more complicated for them to
conceive and implement new analytical projects. Additionally, without comprehensive
documentation, organizations cannot properly manage data throughout their life
cycles, from creation and storage to archiving or deletion. Furthermore, only 15 percent
of MISs measure how often their data are accessed. More robust and standardized
approaches to tracking administrative data access are important to identify users and
frequency of use and optimize data utilization by organizations.

In addition to being comprehensive and accessible, data must also be high in quality,
so MISs need mechanisms to ensure the quality of the data they contain. According
to the majority (65 percent) of MIS experts surveyed, data quality controls are
implemented in the systems they work with to ensure the accuracy, consistency,
completeness, reliability, and overall quality of administrative data. These controls
seem to be particularly prevalent for TaxMISs (82 percent) and EdMISs (85 percent),
but they are less frequent for HealthMISs (43 percent) (figure 4.14). However, these

“Inventories exist, but they do not frame or

“ contain all the data.”

– HealthMIS expert

“There is no organized record of the available data.”

– PFMIS expert

“There are usage metrics for the platform in

general, but not specifically regarding data access.”
– TaxMIS expert

A Regional Assessment of Government Analytics IN LAC 79

 MIS experts also noted that data-cleaning processes are often infrequent and must
be performed manually, outside the systems. Indeed, according to the World Bank’s
GovTech Maturity Index, only 25 percent of countries in the region have implemented
a data quality framework, which is lower than the implementation rate in other regions
(World Bank 2022).3

FIGURE 4.14 Data Quality Controls in Place, by Type of MIS

EdMIS 85

TaxMIS 82

HRMIS 64

PFMIS 62

e-Procurement 45

HealthMIS 43

0 10 20 30 40 50 60 70 80 90
Percent

Source: Government Analytics Survey in Latin America and the Caribbean.

Note: The figure shows, by type of MIS, the proportion of systems in the region that have data quality controls
in place. N = 85 country-MISs, distributed as follows: EdMIS questionnaire (n = 13 countries); e-Procurement
questionnaire (n = 11 countries); HealthMIS questionnaire (n = 14 countries); HRMIS questionnaire (n = 14
countries); PFMIS questionnaire (n = 16 countries); TaxMIS questionnaire (n = 17 countries). EdMIS = education
management information system; e-Procurement = procurement management information system;
HealthMIS = health management information system; HRMIS = human resources management information
system; MIS = management information system; PFMIS = public financial management information system;
TaxMIS = tax management information system.

This lack of systematic data quality controls can pose problems for organizations
seeking to leverage administrative data for analytics. When data quality controls are not
integrated into an MIS, each team accessing the data needs to perform its own quality
controls, duplicating other teams’ efforts. This results in a highly inefficient quality
control process. Moreover, if quality issues, such as erroneous data input or incomplete
data fields, remain, they can significantly undermine the accuracy, reliability, and
replicability of analytics.

“There are some quality control checks and auditing

“ processes in place; however, additional controls are

necessary moving forward to ensure the optimum
quality of data.”
– HealthMIS expert

80 Data for Better Governance

 “Although there are no data-cleaning processes, work

“ is being done to improve the data and their quality

by means of validation upon entry and correction
requests.”
– TaxMIS expert

“There are quality controls in the transactional

system that provides data for e-Procurement, but
they are not sufficient for adequate data quality.”
– e-Procurement expert

“Data are reviewed on a regular basis, but it is labor-

intensive with the lack of appropriate tools; hence,
more can be done.”
– HRMIS expert

All the data infrastructure limitations mentioned here—incomplete digitalization; data

fragmentation; and the lack of formal access protocols, comprehensive data inventories,
and systematic data quality controls—weaken the interoperability of government
information systems. If governments established the foundational infrastructure for
interoperability, they could thus advance their overall potential for using data for
analytics and policy making enormously. According to the World Bank’s GovTech
Maturity Index, less than 35 percent of countries in Latin America and the Caribbean
have implemented a government interoperability framework that allows efficient and
secure information exchange between government systems and organizations (World
Bank 2022).4 Governments should transition from a fragmented, siloed approach to a
comprehensive, whole-of-government approach to ensure responsive, effective public
administration. This means that rather than requiring individual organizations to
collect, store, and analyze data in isolation to pursue organization-specific objectives,
governments should encourage collaboration among organizations spanning
different levels and sectors. Such a collaborative approach can foster synergy and help
organizations achieve shared objectives across diverse domains.

Overall, the regional assessment presented in this chapter reveals the enormous
potential for governments in Latin America and the Caribbean to use their existing
administrative data to enhance policy design and implementation. By enhancing
their analytical capabilities (through building dedicated analytical career tracks
and dedicated analytics units) and by strengthening their data infrastructure
(through enhancing data quality controls and strengthening MIS interoperability),
governments can build on their existing use of descriptive analytics for monitoring
and accountability. Diagnostic and predictive analytics, if developed strategically and

A Regional Assessment of Government Analytics IN LAC 81

 in response to policy needs, can help advance evidence-based policy making and
strengthen government functioning in the region. The final chapter of this report lays
out policy recommendations to show how governments can strengthen their overall
government analytics ecosystem to pursue these goals.

NOTES
1. All analyses in this chapter are presented at aggregate levels. Individual countries are not identified
owing to data confidentiality agreements for the survey. Keeping countries anonymous and informing
respondents of this approach beforehand likely reduced social desirability bias in their responses.
2. The 2019 Digital Government Index question reads as follows: “Does your country have an explicit
formal requirement for public sector organizations to share the data they produce with other public
sector organizations?”
3. Only Africa (excluding North Africa) has a lower implementation rate than Latin America and
the Caribbean, with just 13 percent of countries having implemented a data quality framework.
In comparison, 51 percent of countries in Europe and Central Asia, 50 percent of countries in South
Asia, 33 percent of countries in the Middle East and North Africa, and 30 percent of countries in East
Asia and the Pacific have implemented data quality frameworks (World Bank 2022).
4. The implementation rate for data interoperability frameworks in Latin America and the Caribbean is
significantly lower than those in South Asia (75 percent) and Europe and Central Asia (74 percent),
but it is higher than those in the Middle East and North Africa (33 percent), East Asia and Pacific
(27 percent), and Africa (23 percent) (World Bank 2022).

REFERENCES
OECD (Organisation for Economic Co-operation and Development). 2014. “Recommendation of the
Council on Digital Government Strategies.” Adopted July 14, 2014. https://legalinstruments.oecd.org​
/en/instruments/OECD-LEGAL-0406.
OECD (Organisation for Economic Co-operation and Development). 2019. “Digital Government
Index: 2019 Results.” OECD Public Governance Policy Paper 3, OECD, Paris. https://doi​
.org/10.1787/4de9f5bb-en.
Porrúa, Miguel, Mariano Lafuente, Benjamin Roseth, Laura Ripani, Edgardo Mosqueira, Angela Reyes,
Javier Fuenzalida, Francisco Suárez, and Rodrigo Salas. 2021. Digital Transformation and Public
Employment: The Future of Government Work. Washington, DC: Inter-American Development Bank.
https://doi.org/10.18235/0003245.
World Bank. 2022. GovTech Maturity Index, 2022 Update: Trends in Public Sector Digital Transformation.
Equitable Growth, Finance and Institutions Insight—Governance. Washington, DC: World Bank.
https://hdl.handle.net/10986/38499.

82 Data for Better Governance

CHAPTER 5

Policy Recommendations
to Strengthen
Government Analytics
in Latin America and
the Caribbean

INTRODUCTION
This report has offered an overview and analysis of how governments in Latin America
and the Caribbean can use data analytics to address development challenges. It has
argued that governments in the region can take advantage of significant advances in
management information system (MIS) coverage to develop analytics strategically
and apply them to policy design, implementation, and evaluation, especially by
strengthening the enabling conditions for these advances: analytical capabilities
and data infrastructure. The report’s conceptual framework (chapter 2), case studies
(chapter 3), and regional assessment (chapter 4) offer governments guidance in
developing a government analytics ecosystem in their context and inspiration for
the many ways analytics can be applied to drive evidence-based policy making and
improve government functioning.

This chapter offers targeted policy recommendations to help governments put the
other chapters of this report into practice by developing and using data analytics to
further their own policy objectives. The policy recommendations reflect the structure
of the conceptual framework (figure 2.1). First, the chapter discusses how to move
toward a more strategic approach to government analytics to generate evidence
for decision-making. Then, it presents recommendations for building analytical
capabilities, and finally, it looks at ways to strengthen data infrastructure.

A reproducibility package is available for this book in the Reproducible Research Repository at
https://reproducibility.worldbank.org/index.php/catalog/209.

83
 These policy recommendations can be used by government organizations that are
just beginning to leverage their administrative data as well as those that are already
far along their analytics journey. As the regional assessment in chapter 4 shows,
MIS digitalization, data access protocols, and quality controls can vary widely across
government functions within a single country. For this reason, this chapter breaks
most of the recommendations down into strategic steps of increasing complexity.
Organizations can approach these steps in order, and they can use them to identify the
appropriate next step in their own contexts.

Identifying next steps is important because developing a strong government

analytics ecosystem is an iterative process. When governments push data
collection and analytics to the limit, they can put pressure on public servants,
resulting in perverse incentives that limit the power of analytics to inform
decision-making. Under pressure, organizations might have incentives to focus
only on measuring what is easiest to quantify, resulting in a partial or distorted
understanding of phenomena and taking attention away from analyzing their
foundational causes. Public servants might even face pressure to falsify or
misrepresent data, threatening the health of the analytics ecosystem as a whole
(Bridges and Woolcock 2023).

One way that governments can avoid creating these distortionary incentives is by
approaching analytics strategically and incrementally, while paying attention to
how analytical initiatives depend on and transform enabling conditions within the
analytics ecosystem. By focusing on targeted, gradual reforms, governments do not
sacrifice the significant impact of analytics, because even slightly better analytics can
lead to large improvements in government functioning. As the case studies in this
report have shown, government organizations that take small, initial steps toward
analytics—like organizing a datathon or developing descriptive analytics about health
appointments—can still make a large impact.

POLICY RECOMMENDATIONS
How Can Governments Move toward a More Strategic Approach to Analytics
to Generate Evidence for Decision-Making?

As outlined in chapter 4, administrative data in Latin America and the Caribbean

are predominantly used for descriptive analytics. Diagnostic and predictive
analytics remain largely underutilized, hindering governments’ ability to
effectively anticipate and respond to policy challenges. Furthermore, analytical
products are primarily employed for monitoring and accountability, with less
emphasis on policy evaluation and design. Relying on administrative data for
operational and transactional purposes often results in missed opportunities to
leverage advanced analytics for more strategic decision-making. These findings
underscore the need for a more strategic approach to government analytics, which
is reflected in the following recommendations.

84 Data for Better Governance

Strengthen the culture for evidence-based policy making and the demand for analytics
within public administration.

1. Review whether data and evidence are being used to define policies and strategies
in each government function. (For detailed information on analytical initiatives
and decision-making in human resources, public finance, and procurement, refer
to chapters 10–12 of The Government Analytics Handbook [Tavares, Ortega Nieto,
and Woodhouse 2023; Piatti-Fünfkirchen, Brumby, and Hashim 2023; Cocciolo,
Samaddar, and Fazekas 2023].)

2. Strengthen the link between analytics applications and decision-making by

systematically defining the priorities for reform in each government function and
the evidence that would be needed to guide reform in that function. (For example:
What could be improved about procurement? What would governments need to
know to make those improvements?)

3. Build demand for data analytics through workshops, conferences, and programs
that create awareness of effective applications of analytics, establish networks
and partnerships, and encourage data sharing among government organizations.
(For more information on data sharing, refer, for example, to Welch, Feeney, and
Park 2016.)

4. Strengthen partnerships among organizations and coordinate public stakeholders’

interests (refer, for example, to Allard et al. 2018).

5. Once demand for data analytics has been established, design and implement an
institutional data governance strategy and a data management strategy. These
strategic documents are essential to enable government organizations to effectively
manage their data assets; create internal protocols, processes, roles, and policies;
mitigate risks; and strengthen a culture that supports evidence-based decision-
making. (For more details on data governance, refer to chapters 6 and 8 of the
World Development Report 2021: Data for Better Lives [World Bank 2021].)

Define the strategic use of government data.

1. Review how data from each MIS could be used to inform decision-making in the
related government function. (For example: How can e-Procurement data be used
to strengthen procurement?)
2. Strategize about the broader uses of data from each MIS across organizations,
especially if they are connected with other government data. (For example: What
can e-Procurement data be used for besides procurement? Which organizations
are currently using procurement data and how? Which organizations would benefit
from using procurement data and how?)
Streamline descriptive analytics and develop diagnostic and predictive analytics.

1. Create the conditions for effective and efficient use of administrative data for
monitoring and accountability. Define a monitoring and accountability framework,
establish a process for creating regular reports based on administrative data, and
create an interactive dashboard for control and reporting purposes.

Policy Recommendations to Strengthen Government Analytics IN LAC 85

 2. Move toward using diagnostic and predictive analytics to inform policy evaluation
and design. Develop more advanced analytical products, such as (for procurement)
red-flag systems, bid-rigging screening tools, or government risk assessment
systems (case study 3.11). Progressing from descriptive to diagnostic and predictive
analytics can unleash great potential from government data, but it requires political
support to shift how policy decisions are made as well as strong foundations in
data infrastructure and analytical capabilities. Predictive analytics also requires
data analysts with more advanced technical skills, partnerships with external
institutions, or both.

How Can Governments Build Analytical Capabilities for Individuals

and Organizations?

The report identifies a critical need to strengthen analytical capabilities in both

individuals and organizations within the public sector of Latin America and
the Caribbean. Although governments in the region recognize the importance
of analytical skills, there is a significant gap in structured career tracks for data
analysts and a lack of systematic proficiency evaluations in data analytics for public
servants. This deficiency, coupled with limited funding for analytical projects,
constrains the development and implementation of comprehensive analytics
strategies. The findings suggest that establishing dedicated analytics units and
fostering collaborations with academic and international organizations can play
a pivotal role in addressing these capacity-building challenges. These strategies
are essential for enhancing the strategic use of data analytics in government
decision-making.

Assess existing analytical capabilities.

1. Assess analytical capabilities in the public sector regularly, especially in dedicated

analytics units (if they exist). Use these assessments to identify capacity-building
needs and shape strategic workforce planning.
2. Align the analytical capabilities of the public sector workforce with the skills and
expertise needed for the strategic production and use of government analytics.
Build analytical capabilities strategically.

1. Offer training programs in data analytics and develop initiatives to strengthen

public servants’ capacity to use data and evidence to make key decisions.
2. Develop a coherent framework for connecting skills development to practical
applications, integrating training and capacity building into an overall strategy to
move public administration toward evidence-based policy making. Approaching data
analytics training programs strategically can help ensure that they have a cohesive
structure and offer clear pathways for trainees to apply newly acquired skills within
existing work programs.
3. For decision-makers, build capacity to identify knowledge gaps and evidence needs
and apply analytical insights.

86 Data for Better Governance

4. For data analysts, complement training programs with sector-specific technical
training when necessary. It is essential that analytics teams understand the context
and implications of analytics in specific policy areas.
5. For subject area experts, build capacity to identify knowledge gaps, design
analytical initiatives, interpret analytical results, and put results into context.
Establish a dedicated career track for analysts.

1. Establish a dedicated career track for data analysts in the public sector to generate
structured career development opportunities in analytics, which are essential for
attracting and retaining skilled data analysts.
2. Build relationships with academic institutions to create opportunities for attracting and
fostering specialized talent through internships, fellowships, and educational leave with
pay for public officials. (For more information, refer to chapter 6 of Porrúa et al. 2021.)
Establish dedicated analytics units.

1. Establish a unit within the organizational structure of public administration whose

mandate is to use government data for monitoring, evaluation, reporting, policy
design, and research. Consider whether a centralized unit, which provides analytics
for the entire administration, or decentralized units, which focus on specific
government functions within a single organization, are better in the given context.
2. Leverage analytics units to attract data analysts; establish a consistent, long-term
analytical work stream; and secure regular funding through the payroll of dedicated
staff.
Secure a budget to support analytical capabilities and products.

A dedicated budget is essential for supporting analytical products (such as dashboards,

reports, and research) and capacity building.

1. Offer internal funding opportunities for analytical initiatives.

2. Leverage analytics units and external collaborations to secure resources for data
analytics and mitigate budget constraints.
3. Share funding for analytics teams with other public institutions through formal
or informal arrangements to reduce costs and promote data interoperability and
knowledge exchange.
4. Develop a strategic plan and a long-term vision to make data analytics a priority
within the government budget.
Develop external collaborations to support all the recommendations in this chapter.

1. Develop strategies to collaborate on analytics and research with academics,

nonprofits, foundations, or multilateral organizations. External partners can help
mitigate budget constraints. They can also address a lack of analytical skills and
expertise in public administration by helping to design and implement analytical
products. Collaborations require a clear framework for data sharing and the
coproduction of analytical products.

Policy Recommendations to Strengthen Government Analytics IN LAC 87

 2. Consider partnerships with the educational sector to build capabilities. Globally,
universities are launching initiatives to address a lack of analytical skills in the labor
market. Governments, as major employers, can play a crucial role in these efforts.
For instance, Uruguay has forged a productive partnership with the Universidad
de la República, resulting in the creation of an ongoing education course, “Data
Quality Management for Digital Government.” This course is specifically designed
to help public servants working on public sector digital transformation develop
specialized analytical skills.

How Can Governments Strengthen Data Infrastructure?

The report highlights significant challenges in the data infrastructure of Latin America
and the Caribbean, despite advances in digitalization. Many MISs are insufficiently
comprehensive or irregularly updated, resulting in significant variations in data
quality. There is a notable lack of systematic data quality control measures, and data
cleaning is often manual and sporadic. This situation hampers the accuracy, reliability,
promptness, and replicability of data analytics. Additionally, limited interoperability
among MISs and the ad hoc nature of data sharing among government organizations
prevent the full potential of digitalized MISs from being realized in analytical projects.
These issues underscore the critical need for improved data governance and robust
infrastructure to support effective government analytics in the region.

Assess the quality, completeness, and timeliness of government data and define steps for
improvement.

Regularly assessing, reviewing, and updating MISs is essential to catalyze the value of
their data.

1. Ensure that MISs are fully digitalized. This requires transitioning away from manually
recording information in Word documents, PDFs, or Excel spreadsheets. Because of
resource constraints, certain data elements may need to be given priority over others.
2. Determine whether administrative data are correct, complete, timely, and available
for analytics, and identify any obstacles to data quality.
3. When an MIS is part of a data ecosystem with a joint policy objective (for instance,
the public financial management information ecosystem, which, in practice,
includes several institutions and MISs), examine data quality, completeness, and
timeliness in relation to the overall data ecosystem. Assess the quality of all MISs in
the ecosystem to avoid information silos and strengthen less effective MISs.
Establish regular, systematic, and automated data quality controls.

Data quality issues, such as erroneous data input or incomplete data fields, can
significantly undermine the accuracy, reliability, and replicability of analytics.

1. Integrate data quality controls (such as data cleaning, coverage, and harmonization)
into MISs to make quality control more efficient and reliable.
2. Establish dedicated teams responsible for overseeing data quality controls.

88 Data for Better Governance

Establish a data inventory.

A data inventory helps users determine what types of data are available in an MIS,
where they are stored, and the exact definition of data fields. This makes it much easier
to develop and implement new analytical projects. A data inventory is also necessary
for the proper management of data throughout their life cycle, from creation and
storage to archiving or deletion.

Connect government data.

1. Establish protocols for data sharing between government agencies.

Establishing explicit, formal protocols for sharing administrative data
is a key step toward connecting and integrating government data and
producing more advanced analytical products. (For information on
institutions, laws, and regulations to enhance data accessibility, refer to
World Bank 2022.)
2. Integrate MISs to unlock the full potential of fully digitalized systems for
analytics. (For details on data integration and systems interoperability, refer
to World Bank 2023.)

Plan for MIS maintenance and updates.

Establish protocols for MIS maintenance and updates to ensure that MISs continually
leverage the newest technological advances.

CONCLUSION
The policy recommendations that conclude this report present strategic steps
that governments in Latin America and the Caribbean can take to promote the
development of a government analytics ecosystem. By strengthening their analytical
capabilities and data infrastructure holistically, governments can unlock the full
potential of their administrative data while remaining responsive to decision-makers’
needs for evidence in different policy areas. Government analytics that supports
the knowledge and experience of policy makers and managers can be a key tool for
creating a culture of evidence-based policy making within public administration, with
positive impacts on fiscal sustainability, public service delivery, and citizens’ trust in
public institutions.

Over the past few decades, governments in Latin America and the Caribbean have
built a strong foundation for data analytics and have experimented with different
analytical applications. By drawing on the conceptual framework, case studies,
regional assessment, and policy recommendations in this report, governments can
further advance their use of administrative data for analytics. Government analytics
initiatives can provide policy makers with the evidence needed to make informed
decisions, address governance and development challenges, and build a more
efficient and effective public administration.

Policy Recommendations to Strengthen Government Analytics IN LAC 89

 REFERENCES
Allard, Scott W., Emily R. Wiegand, Colleen Schlecht, A. Rupa Datta, Robert M. Goerge, and Elizabeth
Weigensberg. 2018. “State Agencies’ Use of Administrative Data for Improved Practice: Needs,
Challenges, and Opportunities.” Public Administration Review 78 (2): 240–50. https://doi.org/10.1111​
/puar.12883.
Bridges, Kate, and Michael Woolcock. 2023. “Measuring What Matters: Principles for a Balanced Data
Suite That Prioritizes Problem Solving and Learning.” In The Government Analytics Handbook:
Leveraging Data to Strengthen Public Administration, edited by Daniel Rogger and Christian Schuster,
chap. 4. Washington, DC: World Bank. https://doi.org/10.1596/978-1-4648-1957-5.
Cocciolo, Serena, Suxhmita Samaddar, and Mihaly Fazekas. 2023. “Government Analytics Using
Procurement Data.” In The Government Analytics Handbook: Leveraging Data to Strengthen Public
Administration, edited by Daniel Rogger and Christian Schuster, chap. 12. Washington, DC:
World Bank. https://doi​.org/10.1596/978-1-4648-1957-5.
Piatti-Fünfkirchen, James Brumby, and Ali Hashim. 2023. “Government Analytics Using Expenditure
Data.” In The Government Analytics Handbook: Leveraging Data to Strengthen Public Administration,
edited by Daniel Rogger and Christian Schuster, chap. 10. Washington, DC: World Bank. https://doi​
.org/10.1596/978-1-4648-1957-5.
Porrúa, Miguel, Mariano Lafuente, Benjamin Roseth, Laura Ripani, Edgardo Mosqueira, Angela Reyes,
Javier Fuenzalida, Francisco Suárez, and Rodrigo Salas. 2021. Digital Transformation and Public
Employment: The Future of Government Work. Washington, DC: Inter-American Development Bank.
https://doi​.org/10.18235/0003245.
Tavares, Rafael Alves de Albuquerque, Daniel Ortega Nieto, and Eleanor Florence Woodhouse. 2023.
“Government Analytics Using Human Resources and Payroll Data.” In The Government Analytics
Handbook: Leveraging Data to Strengthen Public Administration, edited by Daniel Rogger and Christian
Schuster, chap. 10. Washington, DC: World Bank. https://doi.org/10.1596/978-1-4648-1957-5_ch10.
Welch, Eric W., Mary K. Feeney, and Chul Hyun Park. 2016. “Determinants of Data Sharing in U.S.
City Governments.” Government Information Quarterly 33 (3): 393–403. https://doi.org/10.1016/j​
.giq.2016.07.002.
World Bank. 2021. World Development Report 2021: Data for Better Lives. Washington, DC: World Bank.
https://www.worldbank.org/en/publication/wdr2021.
World Bank. 2022. GovTech Maturity Index, 2022 Update: Trends in Public Sector Digital Transformation.
Equitable Growth, Finance and Institutions Insight—Governance. Washington, DC: World Bank.
https://hdl​.handle.net/10986/38499.
World Bank. 2023. Interoperability: Towards a Data-Driven Public Sector. Equitable Growth, Finance and
Institutions Insight—Governance. Washington, DC: World Bank. https://doi.org/10.1596/38520.

90 Data for Better Governance

APPENDIX A

Survey Methodology

The questionnaires used in the government analytics survey conducted for this
report were designed following the conceptual framework outlined in chapter 2 and
benefited from the input of different teams within the World Bank’s Governance Global
Practice, as well as feedback from government and international organization experts.
Each questionnaire for a particular type of management information system (MIS)
consisted of 20 questions and was uniformly structured, with identical questions but
response options tailored to that type of MIS, including yes/no and multiple-choice
formats. Respondents were also asked to elaborate on their responses, provide detailed
explanations, share relevant documentation, and offer additional comments to support
their answers. The capabilities questionnaire consisted of four questions.

The survey team administered the questionnaires online via the SurveyCTO platform,
ensuring that only one response per MIS per country was recorded. The team’s
outreach strategy involved multiple stages. Initially, the team reached out to the digital
government authority in each country to identify focal points. These digital authorities
were invited via an email that introduced the survey and requested the appointment
of a focal point for further communication. The team then invited these focal points
to an information session about the survey. The focal points were then responsible for
coordinating with relevant MIS experts within their governments to collect responses
to the MIS-level questionnaires. To streamline the process, the team conducted a
thorough review of MIS experts across all 32 countries in the region and shared the
information with the focal points to help them coordinate data collection efforts.

Data collection took place from November 2023 to March 2024. Throughout this
period, the team hosted multiple information sessions with representatives from
21 countries in the region. These sessions aimed to clarify the survey’s objectives,
offer detailed guidance on navigating and completing the survey, and address any
questions or concerns raised by government officials. In total, 20 countries participated
in the survey. The team received 85 responses to the MIS-level questionnaires, and
16 countries responded to the capabilities questionnaire. The 85 responses to the
MIS-level questionnaires were distributed as follows: 13 countries responded to the
education management information system (EdMIS) questionnaire, 11 countries to
the e-Procurement questionnaire, 14 countries to the health management information
system (HealthMIS) questionnaire, 14 countries to the human resources management

91
 information system (HRMIS) questionnaire, 16 countries to the public financial
management information system (PFMIS) questionnaire, and 17 countries to the tax
management information system (TaxMIS) questionnaire. A detailed breakdown of
respondents by country and organization is provided in table A.1, and table A.2 lists
the participating countries by questionnaire.

The team meticulously reviewed all responses to ensure data consistency,

comprehensiveness, and comparability while identifying areas needing further
evidence or documentation. During this process, the team followed up with each
country for additional information, clarification, and examples to better contextualize
its responses.

TABLE A.1 Countries and Organizations That Participated in the Survey

Country Participating organizations, by questionnaire

Bahamas, The Capabilities: Department of Information and Communication Technology
Barbados Capabilities: Ministry of Industry, Innovation, Science and Technology
HRMIS: Ministry of the Public Service
TaxMIS: Barbados Revenue Authority
Belize Capabilities: Ministry of Public Utilities, Energy, Logistics and E-Governance
EdMIS: Ministry of Education, Culture, Science and Technology
HealthMIS: Ministry of Health and Wellness
HRMIS: Ministry of the Public Service, Constitutional and Political Reform and Religious
Affairs
PFMIS: Ministry of Finance
TaxMIS: Belize Tax Service, Ministry of Finance
Bolivia Capabilities: Agencia de Gobierno Electrónico y Tecnologías de Información y
Comunicación
EdMIS: Ministerio de Educación
e-Procurement: Ministerio de Economía y Finanzas Públicas
HealthMIS: Ministerio de Salud y Deportes
HRMIS: Ministerio de Economía y Finanzas Públicas
PFMIS: Ministerio de Economía y Finanzas Públicas
TaxMIS: Servicio de Impuestos Nacionales
Brazil Capabilities: Ministério da Gestão e da Inovação em Serviços Públicos
PFMIS: Secretaria do Tesouro Nacional
TaxMIS: Secretaria Especial da Receita Federal
Chile Capabilities: Secretaría de Gobierno Digital
EdMIS: Ministerio de Educación
e-Procurement: ChileCompra
HealthMIS: Ministerio de Salud
HRMIS: Servicio Civil
PFMIS: Ministerio de Hacienda
TaxMIS: Servicio de Impuestos Internos
Colombia PFMIS: Ministerio de Hacienda
TaxMIS: Dirección de Impuestos y Aduanas Nacionales

(continues on next page)

92 Data for Better Governance

TABLE A.1 Countries and Organizations That Participated in the Survey (continued)

Country Participating organizations, by questionnaire

Costa Rica Capabilities: Ministerio de Ciencia, Innovación, Tecnología y Telecomunicaciones
EdMIS: Ministerio de Educación Pública
e-Procurement: Ministerio de Hacienda
HealthMIS: Ministerio de Salud
HRMIS: Dirección General de Servicio Civil
PFMIS: Ministerio de Hacienda
TaxMIS: Ministerio de Hacienda
Dominica Capabilities: Caribbean Digital Transformation Project
EdMIS: Ministry of Education, Human Resource Planning, Vocational Training and National
Excellence
e-Procurement: Government of the Commonwealth of Dominica
HealthMIS: Dominica Hospitals Authority
HRMIS: Establishment, Personnel and Training Department
PFMIS: Government of the Commonwealth of Dominica
TaxMIS: Government of the Commonwealth of Dominica
Dominican HRMIS: Ministerio de Administración Publica
Republic PFMIS: Ministerio de Hacienda
Ecuador Capabilities: Ministerio de Telecomunicaciones y de la Sociedad de la Información
EdMIS: Ministerio de Educación
e-Procurement: Servicio Nacional de Contratación Pública
HealthMIS: Ministerio de Salud Pública
HRMIS: Ministerio del Trabajo
PFMIS: Ministerio de Economía y Finanzas
TaxMIS: Servicio de Rentas Internas
Guatemala Capabilities: Comisión Presidencial de Gobierno Abierto y Electrónico
EdMIS: Ministerio de Educación
e-Procurement: Ministerio de Finanzas Publicas
HealthMIS: Ministerio de Salud Pública y Asistencia Social
HRMIS: Oficina Nacional de Servicio Civil
PFMIS: Ministerio de Finanzas Publicas
TaxMIS: Superintendencia de Administración Tributaria
Honduras Capabilities: Oficina de la Presidencia de la República
EdMIS: Secretaría de Educación
e-Procurement: Oficina Normativa de Contratación y Adquisiciones del Estado
HealthMIS: Secretaría de Salud
HRMIS: Administración Nacional de Servicio Civil
PFMIS: Secretaría de Finanzas
TaxMIS: Administración Aduanas
Jamaica Capabilities: eGov Jamaica Limited
e-Procurement: Public Procurement Commission
HealthMIS: Ministry of Health and Wellness
TaxMIS: Tax Administration

(continues on next page)

Survey Methodology 93
TABLE A.1 Countries and Organizations That Participated in the Survey (continued)

Country Participating organizations, by questionnaire

Panama Capabilities: Autoridad Nacional para la Innovación Gubernamental
EdMIS: Ministerio de Educación
e-Procurement: Dirección General de Contrataciones Públicas
HealthMIS: Ministerio de Salud
HRMIS: Ministerio de la Presidencia
PFMIS: Ministerio de Economía y Finanzas
TaxMIS: Ministerio de Economía y Finanzas
Paraguay Capabilities: Ministerio de Tecnologías de Información y Comunicación
EdMIS: Ministerio de Educación y Ciencias
e-Procurement: Dirección Nacional de Contrataciones Publicas
HealthMIS: Ministerio de Salud Pública y Bienestar Social
HRMIS: Viceministerio de Capital Humano y Gestión Organizacional
PFMIS: Ministerio de Economía y Finanzas
TaxMIS: Dirección Nacional de Ingresos Tributarios
Peru Capabilities: Secretaría de Gobierno y Transformación Digital
EdMIS: Ministerio de Educación
PFMIS: Ministerio de Economía y Finanzas
TaxMIS: Superintendencia Nacional de Aduanas y de Administración Tributaria
St. Vincent and HealthMIS: Ministry of Health, Wellness and the Environment
the Grenadines
Trinidad and Capabilities: Ministry of Digital Transformation
Tobago EdMIS: Ministry of Education
HealthMIS: Ministry of Health
HRMIS: Service Commissions Department
PFMIS: Ministry of Finance
TaxMIS: Inland Revenue Division
Uruguay Capabilities: Agencia de Gobierno Electrónico y Sociedad de la Información y del
Conocimiento
EdMIS: Administración Nacional de Educación Pública
e-Procurement: Agencia Reguladora de Compras Estatales
HealthMIS: Ministerio de Salud Pública
HRMIS: Oficina Nacional del Servicio Civil
PFMIS: Ministerio de Economía y Finanzas
TaxMIS: Dirección General Impositiva

Source: Original table for this publication.

Note: The table lists the primary organizations involved in completing each questionnaire, but additional
organizations also contributed. EdMIS = education management information system; e-Procurement =
procurement management information system; HealthMIS = health management information system; HRMIS =
human resources management information system; PFMIS = public financial management information system;
TaxMIS = tax management information system.

94 Data for Better Governance

TABLE A.2 Countries That Participated in the Survey, by Questionnaire

Questionnaire Participating countries

Capabilities The Bahamas, Belize, Bolivia, Brazil, Chile, Costa Rica, Dominica, Ecuador, Guatemala,
Honduras, Jamaica, Panama, Paraguay, Peru, Trinidad and Tobago, Uruguay
EdMIS Belize, Bolivia, Chile, Costa Rica, Dominica, Ecuador, Guatemala, Honduras, Panama,
Paraguay, Peru, Trinidad and Tobago, Uruguay
e-Procurement Bolivia, Chile, Costa Rica, Dominica, Ecuador, Guatemala, Honduras, Jamaica, Panama,
Paraguay, Uruguay
HealthMIS Belize, Bolivia, Chile, Costa Rica, Dominica, Ecuador, Guatemala, Honduras, Jamaica,
Panama, Paraguay, St. Vincent and the Grenadines, Trinidad and Tobago, Uruguay
HRMIS Barbados, Belize, Bolivia, Chile, Costa Rica, Dominica, Dominican Republic, Ecuador,
Guatemala, Honduras, Panama, Paraguay, Trinidad and Tobago, Uruguay
PFMIS Belize, Bolivia, Brazil, Chile, Colombia, Costa Rica, Dominica, Dominican Republic,
Ecuador, Guatemala, Honduras, Panama, Paraguay, Peru, Trinidad and Tobago, Uruguay
TaxMIS Barbados, Belize, Bolivia, Brazil, Chile, Colombia, Costa Rica, Dominica, Ecuador,
Guatemala, Honduras, Jamaica, Panama, Paraguay, Peru, Trinidad and Tobago, Uruguay

Source: Original table for this publication.

Note: The table lists the countries that responded to each questionnaire. EdMIS = education management information
system; e-Procurement = procurement management information system; HealthMIS = health management information
system; HRMIS = human resources management information system; PFMIS = public financial management information
system; TaxMIS = tax management information system.

Survey Methodology 95
APPENDIX B

Types of Management
Information Systems
and Their Functions

Human Resources Management Information System

An HRMIS is a platform designed to streamline and automate the management

of human resources within government organizations. It typically covers various
human resources functions, such as personnel administration, workforce planning,
compensation management, career development and training, attendance tracking,
recruitment and deployment, mobility, and performance management (Cortázar
Velarde, Lafuente, and Sanginés 2014; Nunberg 2021). The main objectives of an
HRMIS are to ensure compliance with existing payroll rules, enhance the productivity
of public servants, and ensure efficiency in human resources functions such as
recruitment, rotation, and dismissals (Farooq and Kim 2023).

Public Financial Management Information System

A PFMIS is a platform designed to streamline and automate financial operations within

budget and treasury units. It typically covers various budget execution functions, such
as accounts payable and receivable, commitment and cash management, financial
reporting, budget formulation, auditing, debt management, and public investment
management. The main objective of a PFMIS is to provide relevant, timely, and reliable
financial information to plan, execute, and monitor public finances (Pimenta and
Pessoa 2016). Additionally, this type of MIS contributes to the effective and efficient
allocation of public resources, fosters fiscal transparency, and links public financial
management processes to government functions (Dorotinsky and Watkins 2013;
Pimenta and Pessoa 2016).

Procurement Management Information System

An e-Procurement system is a platform designed to streamline the procurement

process for government organizations, enhancing the efficiency, competition,

97
 and transparency of procurement activities (OECD, n.d.). It typically covers
various dimensions of the procurement cycle, such as budget planning and tender
preparation; tendering, bidding process, and bid evaluation; contract award and
signing; contract execution and monitoring; and logistics (Cocciolo, Samaddar, and
Fazekas 2023; UNOPS 2021). The main objective of an e-Procurement system is to
enhance transparency, efficiency, and competition by providing key stakeholders
with timely, accurate, and accessible procurement information to optimize resource
allocation, mitigate risks, and ensure compliance with procurement regulations
(World Bank 2011).

Tax Management Information System

A TaxMIS is a platform designed to streamline and automate the management,

accounting, and reporting of tax-related tasks and transactions. It typically covers
various taxation functions, such as taxpayer registration, return processing,
tax payment handling and assessment, management of appeals and objections,
tax collection and refunds, registration cancellations, issuance of tax clearance
certificates, and cross-institutional reporting (Yoon, Chaithanya, and Kong 2014).
The main objectives of a TaxMIS are to provide accurate, reliable tax data to
strengthen the core business processes of the tax administration, as well as to control
compliance and enhance tax collection enforcement mechanisms (Junquera-Varela
and Lucas-Mas 2024).

Education Management Information System

An EdMIS is a platform that enables the public sector to collect, monitor, manage,
analyze, and disseminate information about educational inputs, processes, and
outcomes (Abdul-Hamid 2017). It typically covers various educational planning,
monitoring, and evaluation processes, including student enrollment, attendance
tracking, academic performance analysis, curriculum management, teacher
deployment, facilities and materials management, financial resource management,
and school infrastructure maintenance (Abdul-Hamid 2017). The main objective of an
EdMIS is to provide systematic, relevant, timely, and accurate information to enable
governments to examine and strengthen the performance of their education systems
(Porta and Arcia 2011).

Health Management Information System

A HealthMIS is a platform designed to streamline and automate the management and

delivery of health care services within public health systems. It typically covers various
dimensions of the health care system, including individual patient records, disease
surveillance, services delivered, health outcomes, health care utilization patterns, health
facility management, community-level records, pharmaceutical inventory, and health

98 Data for Better Governance

personnel management (Lippeveld et al. 2019). The main objectives of a HealthMIS
are to generate information at all levels of the health system to enhance health service
delivery and optimally allocate resources (PAHO 2021).

REFERENCES
Abdul-Hamid, Husein. 2017. “What Is an Education Management Information System and Who Uses It?”
In Data for Learning: Building a Smart Education Data System, 7–16. Washington, DC: World Bank.
https://doi​.org/10.1596/978-1-4648-1099-2.
Cocciolo, Serena, Suxhmita Samaddar, and Mihaly Fazekas. 2023. “Government Analytics Using
Procurement Data.” In The Government Analytics Handbook: Leveraging Data to Strengthen Public
Administration, edited by Daniel Rogger and Christian Schuster, chap. 12. Washington, DC:
World Bank. https://doi​.org/10.1596/978-1-4648-1957-5.
Cortázar Velarde, Juan Carlos, Mariano Lafuente, and Mario Sanginés, eds. 2014. Serving Citizens:
A Decade of Civil Service Reforms in Latin America (2004–13). Washington, DC: Inter-American
Development Bank. https://publications.iadb.org/en/serving-citizens-decade-civil-service-reforms​
-latin-america-2004-13.
Dorotinsky, William, and Joanna Watkins. 2013. “Government Financial Management Information
Systems.” In The International Handbook of Public Financial Management, edited by Richard Allen,
Richard Hemming, and Barry H. Potter, chap. 36. New York: Palgrave Macmillan. https://doi.org​
/10.1057/9781137315304_37.
Farooq, Khuram, and Galileu Kim. 2023. “Creating Data Infrastructures for Government Analytics.”
In The Government Analytics Handbook: Leveraging Data to Strengthen Public Administration,
edited by Daniel Rogger and Christian Schuster, chap. 9. Washington, DC: World Bank. https://doi​
.org/10.1596/978-1-4648-1957-5.
Junquera-Varela, Raúl Félix, and Cristian Óliver Lucas-Mas, eds. 2024. Revenue Administration Handbook.
Washington, DC: World Bank. https://doi.org/10.1596/978-1-4648-2053-3.
Lippeveld, Theo, Tariq Azim, David Boone, Vikas Dwivedi, Michael Edwards, and Carla AbouZahr. 2019.
“Health Management Information Systems: Backbone of the Health System.” In The Palgrave Handbook
of Global Health Data Methods for Policy and Practice, edited by Sarah B. Macfarlane and Carla
AbouZahr, 165–81. London: Palgrave Macmillan. https://doi.org/10.1057/978-1-137-54984-6_9.
Nunberg, Barbara. 2021. “Strengthening Subnational Human Resource Management Systems: A
Primer. Equitable Growth, Finance and Institutions Insight.” World Bank, Washington, DC. https://
documents.worldbank.org/pt/publication/documents-reports/documentdetail/099225502022348809​
/p1754490dbaaa300f0a2b60af27d2e3508c.
OECD (Organisation for Economic Co-operation and Development). n.d. “Methodology for Assessing
Procurement Systems.” Accessed March 30, 2024. https://www.oecd.org/gov/public-procurement​
/methodology​-assessing-procurement/.
PAHO (Pan American Health Organization). 2021. Eight Guiding Principles of Digital Transformation of
the Health Sector: A Call to Pan American Action. Washington, DC: PAHO. https://iris.paho.org/handle​
/10665.2/54256.
Pimenta, Carlos, and Mario Pessoa, eds. 2016. Public Financial Management in Latin America: The
Key to Efficiency and Transparency. Washington, DC: International Monetary Fund. https://doi​
.org/10.18235/0000083.
Porta, Emilio, and Gustavo Arcia. 2011. “Improving Information Systems for Planning and Policy
Dialogue: The SABER EMIS Assessment Tool.” World Bank, Washington, DC. https://www.academia​
.edu/41964853/Improving_Information_Systems_for_Planning_and_Policy_Dialogue_The_SABER​
_EMIS_Assessment_Tool_SABER_System_Assessment_and_Benchmarking_for_Education_Results.

Types of Management Information Systems and Their Functions 99

 UNOPS (United Nations Office for Project Services). 2021. UNOPS Procurement Manual. Revision 7,
last modified July 1, 2021. Copenhagen, Denmark: UNOPS. https://content.unops.org/service-Line​
-Documents​/Procurement/UNOPS-Procurement-Manual-2021_EN.pdf.
World Bank. 2011. e-Procurement Reference Guide. Washington, DC: World Bank. https://documents1​
.worldbank.org/curated/ar/751741468163749785/pdf/882160WP0Box380ementReferenceGuide.pdf.
Yoon, Seok Yong, Chava Chaithanya, and Dongsung Kong. 2014. Tool Kit for Tax Administration
Management Information System. Mandaluyong, Philippines: Asian Development Bank. https://www​
.adb.org/sites/default​/files/publication/150133/tool-kit-tax-administration-management-information​
-system.pdf.

100 Data for Better Governance

 ECO-AUDIT
Environmental Benefits Statement

The World Bank Group is committed to reducing its environmental

footprint. In support of this commitment, we leverage electronic
publishing options and print-on-demand technology, which is located
in regional hubs worldwide. Together, these initiatives enable print runs
to be lowered and shipping distances decreased, resulting in reduced
paper consumption, chemical use, greenhouse gas emissions, and waste.
We follow the recommended standards for paper use set by the Green
Press Initiative. The majority of our books are printed on Forest Stewardship
Council (FSC)–certified paper, with nearly all containing 50–100 percent
recycled content. The recycled fiber in our book paper is either unbleached
or bleached using totally chlorine-free (TCF), processed chlorine–free
(PCF), or enhanced elemental ­chlorine–free (EECF) processes.
More information about the Bank’s environmental philosophy can
be found at http://www.worldbank.org/corporateresponsibility.
Governments in the Latin America and the Caribbean (LAC) region face significant developmental
and institutional challenges, such as slowing growth, fiscal constraints, and inefficiencies in
the public sector. At the same time, governments have invested significantly in government
technologies (GovTech), making LAC a global pioneer in management information systems (MISs).
This investment creates an opportunity for governments to leverage MIS data to strengthen the
functioning of government and achieve development goals—that is, government analytics.

This report provides a conceptual framework to assess and provide guidance on the regional
government analytics agenda and how to harvest the benefits of GovTech investments.
It examines how government analytics can inform policy making and improve accountability
and efficiency, drawing on survey data and successful applications of government analytics.
The report also explores the enabling conditions for government analytics—data infrastructure
and analytical capabilities—and how to strengthen them. Finally, it provides practical guidance on
how to develop a holistic government analytics agenda.

Data for Better Governance: Building Government Analytics Ecosystems in Latin America and
the Caribbean is part of the Government Analytics collection, which began with The Government
Analytics Handbook (2023). This growing series features frontier evidence and expert insights on
how to leverage data to improve government performance.

https://reproducibility.worldbank.org

A reproducibility package is available for this book in the Reproducible Research

Repository at https://reproducibility.worldbank.org/index.php/catalog/209.

ISBN 978-1-4648-2159-2

SKU 212159