Open data

study
New technologies
Becky Hogge

For more information contact:

Transparency &
Accountability Initiative
c/o Open Society Foundation
4th floor, Cambridge House
100 Cambridge Grove
London, W6 0LE, UK
Tel: +44 (0)20 7031 0200
www.transparency-initiative.org

Copyright 2010. Creative Commons License.

This work is licensed under a Creative Commons Attribution
3.0 Licence: http://creativecommons.org/licenses/by/3.0/us/

TAI New Technology /Open data study 3

Contents

Introduction 4
The UK data.gov.uk

The US data.gov

The three-tiered approach

Civil Society

The push from the middle

12

Top-level drive

15

Utility not Users the absent driver?

17

International perspectives

10

18

The relationship between FOI and Open Data

19

Questions over impact

21

Data characteristics

21

The three-tiered approach: top level drivers

25

The three-tiered approach: the middle layer

27

The three-tiered approach: civil society

29

Donors: a fourth tier?

30

Issues for developing and middle income country users

32

Complementary strategies for open data

33

An aside: advice for funders

34

Avenues for further research

35

An open data strategy checklist

36

Status of FOI

37

Current data collection activities

37

Potential end-users

37

Tier 1: Civil society

37

Tier 2: The middle layer

37

Tier 3: The top-level

38

Tier 4: Donor drivers

38

Conclusion 39
Annexes 41

Annexes I: Methodology

42

Annexes II: List of interviewees

43

Bibliography 44
Acknowledgements 45
About the author

45

4 TAI New Technology /Open data study

Introduction

There are substantial social and economic gains to be made from opening government data to the
public. The combination of geographic, budget, demographic, services, education and other data,
publicly available in an open format on the web, promises to improve services as well as create future
economic growth.
This approach has been recently pioneered by governments
in the United States and the United Kingdom (with the
launch of two web portals www.data.gov and www.
data.gov.uk respectively) inspired in part by applications
developed by grassroots civil society organisations (CSOs)
ranging from maps of bicycle accidents to sites breaking
down how and where tax money is spent. In the UK, the
data.gov.uk initiative was spearheaded by Tim Berners-Lee,
the inventor of the World Wide Web.
This research, commissioned by a consortium of funders
and NGOs under the umbrella of the Transparency and
Accountability Initiative, seeks to explore the feasibility of
applying this approach to open data in relevant middle
income and developing countries. Its aim is to identify
the strategies used in the US and UK contexts with a view
to building a set of criteria to guide the selection of pilot
countries, which in turn suggests a template strategy to
open government data.
The report finds that in both the US and UK, a three-tiered
drive was at play. The three groups of actors who were
crucial to the projects success were:
Civil society, and in particular a small and
motivated group of civic hackers ;1
An engaged and well-resourced middle layer
of skilled government bureaucrats; and
A top-level mandate, motivated by either an outside
force (in the case of the UK) or a refreshed political
administration hungry for change (in the US).

1 For

a definition of civic hackers, see the Civil society section

of this report.

As Tim Berners-Lee observed in interview,It has to start

at the top, it has to start in the middle and it has to start
at the bottom. The conclusion to this report strengthens
that assertion, and warns those attempting to mirror the
successes of the UK and US projects not to neglect any
of these three layers of influence.
Based on these findings, and on interviews conducted with
a selection of domain and region experts to refine these
observations for a developing and middle-income country
context (where a fourth tier of potential drivers towards
open data in the shape of international aid donors
is identified) the report presents a list of criteria to be
considered when selecting a pilot country in order to test
this strategy. This checklist includes questions on:
t he status of Freedom of Information law
and data collection activities in the country;
t he capacity of civil society, the media and other
end-users to access and make use of the data;
t he attitude of high and mid-level political operators
and government officials to opening up data;
the role of multilateral and bilateral donors.

TAI New Technology /Open data study 5

The UK - data.co.uk

6 TAI New Technology /Open data study

data.gov.uk is a UK government web portal providing

access to data gathered and maintained by UK
government and related public agencies.

In June 2009 the then Prime Minister, Gordon Brown,

announced that Tim Berners-Lee the man credited
with inventing the World Wide Web was joining the
Government as an expert advisor.2 Since 1999, Berners-Lee
has been developing the idea of a semantic web based
on linked data facilitated by metadata standards. In March
2009, Berners-Lee had made a call at the influential TED
conference for data-gathering bodies around the world to
release Raw Data Now!.3
For several years previous to Browns announcement, the UK
Government had been under pressure from civil society
and in particular a vocal group of civic hackers responsible
for the development of grassroots political engagement
websites to release its data openly and in machinereadable formats. This pressure sometimes exerted
through traditional advocacy, but also exerted through
demonstrator projects had led to internal policy reviews
and initiatives such as the Power of Information Review
(co-written by mySocietys Director Tom Steinberg) and
the Power of Information Taskforce (championed by then
Cabinet Office minister Tom Watson MP and led by Richard
Allan, a former Member of Parliament and at the time a
policy advisor to Cisco systems).

This situation eventually changed in early 2010. data.gov.

uk was launched in beta (invitation only) on 30 September
2009, with an official, open launch in January 2010. On 1
April 2010, the Ordnance Survey announced that it would
make crucial data assets openly available through data.gov.
uk7. It is unclear whether the change of heart on Trading
Funds derived more from mounting evidence including
internal Government development of a database to cater to
the UK census or from simple political will. Nonetheless,
the announcement was seen as a victory for the open data
community, and the UK is now witnessing a flourishing of
postcode-related campaigning and political engagement
sites,(in particular of sites geared towards the May 2010
General Election).
As of April 2010, data.gov.uk listed 3,241 datasets and hosted
49 derived applications (although many more applications
which make use of the data and particularly the geospatial
data are hosted elsewhere on the web). Departments
are not compelled to publish data via the portal; rather,
it is presented as an aspirational choice for government
departments wishing to open their data to the possibilities of
data mash-ups and the semantic web.

These initiatives succeeded in fostering collaboration

between civil servants in the middle layer of government
administration and civic hackers of many different stripes,
for example in regular GovDataBarCamps, and through
competitions such as Show Us a Better Way, which attracted
around 500 entries, including the budget tracking website
WhereDoesMyMoneyGo?4 However, there was strong
resistance from a small number of Government agencies, and
in particular the Governments cartographic mapping agency,
the Ordnance Survey, which was funded from revenues from
the commercial exploitation of public data (so-called Trading
Funds). The Power of Information Review had recommended
a detailed cost-benefit analysis of the Trading Fund model,
which was commissioned jointly by HM Treasury and the
then Department for Business, Enterprise and Regulatory
Reform, and published in February 2008.5Despite that study
finding strongly in favour of abandoning the Trading Fund
model6, and despite the Power of Information Taskforce
recommending reform on the back of its findings, positions
against re-examining the Trading Fund model within
government remained entrenched.

2 Brown, 2009.
3 Berners-Lee, 2009.
4 See: http://www.wheredoesmymoneygo.org/
5 Newbury, Bently and Pollock, 2008.

6 The

study found that in most cases, although making data

available without charge would create additional expenses for

government, the benefits to society of moving to a marginalcost regime effectively zero cost for digital data outweighed
the costs of short-term cost-recovery models.
7 BBC News, 2010.

TAI New Technology /Open data study 7

The US - data.gov

8 TAI New Technology /Open data study

data.gov is a US government web portal providing the public with access to federal governmentcreated datasets. It was launched in 2009, both to allow citizen feedback and new ideas enabling
transparency, participation and collaboration between state and citizen and to increase efficiency
among government agencies. Most US government agencies already work to codified information
dissemination requirements, and data.gov is conceived as a tool to aid their mission delivery.

Like the UK, the US also exhibited a small and influential

community of civic hackers, working independently
and most often without financial reward, to repurpose
government-created datasets and present them online in
more accessible and enriched formats. For example, in 2004
a student called Josh Tauberer launched GovTrack.us, which
repurposes publicly available data about key activities
of the US Congress and publishes it in an accessible,
searchable form. Perhaps in response to this activity, the
Chief Information Officers (CIOs) in key states in the US
most notably the District of Columbia began to release
state-level datasets through data portals. It was the CTO for
the District of Columbia, Vivek Kundra, who was hired by
Barack Obama in March 2009 to act as the Federal CIO.
One of the first memorandums signed by President Obama
when he entered the White House was on openness and
transparency. As a Senator, Obamas headline achievements
included the CoburnObama Transparency Act, which
established USAspending.gov, a search engine on federal
spending, showing that the future President got the
possibilities early on. Following a period of research and
development, data.gov was launched on 21 May 2009.
It initially contained 76 datasets from 11 government
agencies. By late April 2010 that number had risen to 1,284
datasets from 170 government and related public agencies.8
Fears that not enough data was being channelled through
data.gov led to Obama issuing a decree on 8 December
2009 that each government agency should post at least
three high-value datasets by a given date.9 Indeed, after
data.gov.uk launched in January 2010, US commentators
began comparing US efforts unfavourably, complaining that
the UK site had three times as many useful datasets as the
US site, even though it lagged six months behind.10

8 For

the purposes of comparing the UK and US efforts, only

datasets that fall within the Raw Data Catalog in data.gov

have been counted (see below).
9 Executive

Office of the President, Office of Management
and Budget, 2009.

The data.gov website is a curated web portal sitting on

top of three separate, searchable data catalogues (Raw
Data Catalog, Tool Catalog and Geodata Catalog),11 each
of whose entries include a metadata template based on
the Dublin Core standard (Dublin Core Metadata Element
Set, Version 1.1). Although data.gov does not host derived
applications on its site, it does record the number of times
datasets have been downloaded: for example, in the week
beginning 19 April 2010, it reported that datasets across all
three catalogues had been downloaded over 25,000 times
in the previous week.
Unlike data.gov.uk, data.gov does not generally employ
data formats associated with semantic web development.
However, it does include integration with the emerging
semantic web as one of its future goals, and is in the
process of commissioning pilot projects around semantic
web development.

10 Kirkpatrick, 2010.
11 As
 noted above, for the purposes of comparing the UK and

US efforts, only datasets that fall within the Raw Data Catalog
have been counted.

TAI New Technology /Open data study 9

The three-tiered
approach

10 TAI New Technology /Open data study

It has to start at the top, it has to start in the middle

and it has to start at the bottom.
Tim Berners-Lee

At least three constituencies have been involved in the

push for data.gov and data.gov.uk. These constituencies
are civil society; the middle layer of professional public
administrators; and top-level political leaders. It is arguable
that without each one of these constituencies driving
progress, we would not be where we are today.
Civil society, as will be elaborated below, was of a very
particular kind, and provided bottom-up pressure for
change through traditional advocacy, but also through
demonstrator projects such as TheyWorkForYou.com (UK)
and GovTrack.us (US). The middle layer of civil servants
and state and federal administrators were also motivated
to open government data, spotting opportunities for
efficiency and for greater recognition of the work they did.
Without their expertise, top-level political leaders could
not have been confident that plans for data portals could
realistically be realised. Finally, political leaders could
squeeze out institutional inertia, and give what looked like
highly technical projects some democratic grounding.

Civil Society
Its not like mySociety were the only people reusing
data, but we were virtually the only people reusing
it in a sphere that meant that politicians and policy
people paid any attention to who we were.
Tom Steinberg

TheyWorkForYou.com was launched in the UK in June 2004,

having been developed by a loosely-affiliated group of
volunteers who thought it should be really easy for people
to keep tabs on their elected MPs, and their unelected Peers,
and comment on what goes on in Parliament.12 Between
them, the original volunteers had already developed and
run a number of civic engagement and other related
websites, including PublicWhip.org.uk (a Parliamentary
vote tracking website) and AccessibleTrainTimes.com (an
accessible version of the National Rail Enquiries website).
As with previous sites created by this group,
TheyWorkForYou.com repurposed data already being
published online (in this case, via the Hansard website),
presenting it in a format that was accessible and searchable,
and allowed for user feedback. The site launched despite
the fact that the data it was reusing constituted a copyright
infringement in the UK: the volunteers did not have the
right to reproduce Hansard, which was covered by Crown
Copyright. Later on, and in cooperation with some of the
TheyWorkForYou.com volunteers, click-use licences were

developed at the Office for Public Sector Information

(OPSI) which among other things legitimised the sites
activities. In 2006, by mutual agreement, the running of
TheyWorkForYou.com was passed to mySociety, because
the overlap of people and goals with mySociety was so
substantial that it was best to pass the running over to
mySociety, who are an actual proper organisation with
staff and time.13
GovTrack.us was launched in the US in September 2004
by Josh Tauberer, a linguistics postgraduate and software
developer. Like TheyWorkForYou.com, it repurposes official
data (this time produced by the Library of Congress) and
publishes it in an accessible, searchable form. Because of
the way the Library of Congress makes that data available,
GovTrack.us needs to use automated techniques (for
example, pdf-scraping and subsequent information parsing
techniques). The datastream this activity creates means
that GovTrack.us also acts as a data hub for other civic
engagement websites. GovTrack.us lists at least 22 such
projects on its website.14
In 2009, Josh Tauberer incorporated Civic Impulse LLC, and
GovTrack.us became a project of this entity. Advertisements
that run on the website provide the company with a small
profit, but according to Tauberer the point has never been
to make this a business.15
Both the volunteers around TheyWorkForYou.com and
Tauberer identify themselves as civic hackers. Although the
word hacking has, in the context of computers, evolved
in meaning so as to attract broadly negative connotations
among the general public, among software engineers
hacking simply means the experimental development
of software and systems to solve particular problems.
Civic hacking is therefore understood to mean deploying
information technology tools to enrich civic life, or to solve
particular problems of a civic nature, such as democratic
engagement. Civic hackers might be described as tending
away from characteristics associated with other types
of more traditional civil society: for example, they might
privilege approaching only those civic problems to which
they can see an elegant (technical) solution. For more
advice on how to spot civic hackers, see the section below
on Advice for funders.

12 mySociety.

14 Civic Impulse LLC 2.

13 Ibid.

15 Civic Impulse LLC 1.

TAI New Technology /Open data study 11

It was the early interventions of civic hackers which arguably

served to kick off the public debate around open data, and
led to the emergence in 2009 and 2010 of the open data
catalogues data.gov and data.gov.uk. As one observer put it:

I would say that theyre the source of this debate.

Things like data.gov are being created to meet the
new demand and the new role that civic hackers
present. I just think of Josh Tauberer, whos the
University of Pennsylvania graduate student that
built GovTrack.us, which is one of the first big sites
that was taking on a huge scraping problem, taking
Congressional Bills data and scraping it and parsing
it and releasing it in XML and RDF, even in 2005
or 2006, much earlier than anyone else was even
thinking about RDF. That site really empowered an
ecosystem of other sites that use congressional data.
John Wonderlich, Policy Director, Sunlight Foundation

One question that arises from remarks such as these is

whether policy makers got interested in civic hacking
because of the obvious utility of sites such as GovTrack.
us given the meagre resources needed to produce them,
or whether instead it was the widespread adoption of
these sites among members of the general public. Several
observations indicate it was the former, not the latter, which
swung policy makers; for example:

(Josh Tauberer) operated almost completely

independently and was able to provide a more
valuable service than what was coming out of the
entire Library of Congress and he was one person
who was just a linguistics student
John Wonderlich

This theme is picked up later in the report.16

The incorporation of both these individual projects points
to a gradual consolidation of the grassroots movement
around open data issues. The two most visible grassroots
players in the UK and US open data space are mySociety
and the Sunlight Foundation respectively, but this does
not mean that open data issues were not being discussed
in other forums. The licensing issues around data reuse
in the UK (these issues appear to be absent from the US
discussion, mainly because of historic laws that proscribed
public data and databases in general from attracting
copyright protection) led in 2006 to the launch of the
Free Our Data campaign,17 run by Guardian newspaper
technology editor Charles Arthur. The Open Knowledge
Foundation (OKF) was also a key player in convening a
range of grassroots and institutional players from across
national borders around open data issues:

16 See section entitled Utility not users: absent drivers?.

17 http://www.freeourdata.org.uk/
18 The

interviewee is referring to the data catalog Kundra

pioneered as CTO of the District of Columbia, and not data.gov,

which would emerge later.

The Open Knowledge Foundation has also been

involved, from quite an early phase, in organising
events which have brought developers wanting to
do civic web services using open Government data,
together with Government representatives. John
Sheridan from OPSI has come to OKF events and
been very supportive since at least 2005 ... Some
of the discussions that happened at those kinds of
meetings have proved to have been very valuable in
the longer term. So for example in autumn 2008 we
discussed the idea of a registry of open Government
data for the UK, and at that time, data.gov didnt
exist yet people were looking at, for example,
things like Vivek Kundras ... data catalogue18
[for inspiration].
Jonathan Gray, Community Coordinator, OKF

data.gov.uk would eventually be built using the CKAN

infrastructure,19 a project developed by the OKF after it
became aware of the value of open data registries which
aims to be a lightweight open source piece of software for
helping people to find and reuse open data.20
In interviews, representatives from both mySociety and
the OKF stressed how they felt that the culture of their
communities contributed positively to getting buy-in from
government at an early stage.

Ithink it helped that the data agenda wasnt just and

solely associated with anti-politics, basically.
Tom Steinberg

Partly whats made the OKF work is an attempt to

keep things quite simple, because I guess in both
of those communities grassroots and institutional,
there might be a temptation to go off into quite
technical language.
Jonathan Gray

Theres a sense in which the OKF has been able

to mediate between those two groups, but also I
would suggest that the OKF has always been quite
neutral, the OKF has never been a campaigning
organisation.
Jonathan Gray

One interviewee was keen to stress, however, that the

cooperative approach might not translate to all political
settings: in some countries its entirely the logical thing
to do, to represent yourself as a threat, because theres no
other way of doing it.21

19 Where

CKAN stands for Comprehensive Knowledge Archive

Network. CKAN is a registry of open data and content

packages; for more details, visit http://www.ckan.net/
20 Interview with Jonathan Gray.
21 Interview with Tom Steinberg.

12 TAI New Technology /Open data study

One striking and apparent difference between the UK and

US experiences is that in the UK both mySociety and OKF,
were in some sense co-opted by the institutional middle
layer of government as the trajectory towards data.gov.
uk progressed: the OKF, to provide the software that would
eventually run the data portal; and mySociety in the form of
Tom Steinbergs appointment in 2007 by the Cabinet Office
to undertake the Power of Information Review jointly with
Ed Mayo, Director of the National Consumer Council (NCC,
now Consumer Focus).22 It is to developments in this layer
that this report now turns.

The push from the middle

I think actually you can put a lot of the success
down to very serious competence of UK civil
servants - both their understanding of the
workings of government and their understanding
of the technology.
Tim Berners-Lee

It is in the contribution of the middle layer of Government

where the trajectories of data.gov and data.gov.uk appear
to diverge. This could well be a function of this reports
methodology: the researcher was unable to conduct
interviews with any public administrator directly involved
in the launch of either platform, and literature that would
give insight into the middle layer experience is necessarily
sparse. Whereas civil society respondents from both sides of
the Atlantic painted both positive and negative pictures of
the attitudes of public servants towards open data issues, the
highest praise was reserved for the UK middle layer (see the
quote above, which was echoed by other participants). This
leads the researcher to speculate whether on top of the
early outreach undertaken by organisations such as the OKF
it was in fact the contentious issue of data licensing that
gradually brought together allies from the grassroots and
public administrator communities, building a stronger base
of expertise and shared goals than perhaps was exhibited in
the US. The different characteristics of the top-level pressures
might also be a factor, and this is explored in greater
detailbelow(see Top level drive).
In 2007, together with the director of the NCC, Ed Mayo,
Tom Steinberg was commissioned by the Cabinet Office
to conduct the Power of Information Review.23 The review
made 15 recommendations, many of which had to do
with the future role of open public data. Tom Steinberg
identifies two major opportunities offered by the Power of
Information Review. The first was to act as a cheerleader for
opening up government activities to wider civil society:

The Power of Information Review was a combination

of cheerleading and one or two really meaningful
policies. The cheerleading is because you need to set
a tone around agendas in Whitehall, and probably
governments generally, that this is of the moment,
that this is worth doing, that this is not frightening
and this is in any field, whether its technology or
agriculture or just whatever. Well, a way in which
you get things done is to talk about it as if this is the
thing that were doing right now, and the reason you
have to do that is because some of the people who
hold the decision-making power in any such public
institution will never understand what youre pitching.
So as well as coming up with good arguments, you
have to come up with a more general non-specific
emotional vibe that helps people who dont really
understand the arguments, but who for some reason
have been given powerful positions. It helps them
get over the risk-aversion. So the emotional tone is
really important.
Tom Steinberg

The second was to instigate an evidence-based review

of the Ordnance Surveys Trading Fund model:

Getting the Treasury to do an official economic

review of the charging model for Trading Funds is
the other most important thing to come out of the
Power of Information Review.
Tom Steinberg

The Government response to the review was broadly

positive, accepting each recommendation outright, in part, or
in theory.24 The Office of Fair Trading had already produced
a review of the Trading Fund model in December 2006,25 but
Steinberg remarks that this report was important as a source
of reference, but not as a source of authority or power.
More useful in getting the Government to accept the
recommendation and commission a review were personal
contacts within the Treasury.26 Steinberg admits that this
scenario is not necessarily replicable, agreeing with the
interviewer that it was more or less a function of networks
of young people who know about technology, who then
go on into different spheres of life, but stay in touch.27
However, Steinberg goes on to remark that interventions
made by Tim Berners-Lee subsequent to the Power of
Information Review may render the need for such personal
contacts obsolete (hes not young, he doesnt know anyone
in government, but hes famous enough that the Prime
Minister would pay him attention.28). Indeed, although the
report found strongly against the Trading Fund model in
terms of the overall economic value it delivered, it was not
until Tim Berners-Lee came on board at the Cabinet Office
that the reports findings were seriously acted upon.29

22 Mayo and Steinberg, 2007.

28 Interview with Tom Steinberg.

23 Ibid.

29 Crabtree,

2010. This sentiment was reflected by the reports

24 Cabinet Office, 2007.

25 Office of Fair Trading, 2006.
26 Newbury, Bently and Pollock, 2008.
27 Interviewers own remarks in conversation with Tom Steinberg.

co-author Rufus Pollock in a recent presentation on open

government data at the British Council event Copyrights 300th
birthday: What next?, attended by the researcher.

TAI New Technology /Open data study 13

The other major outcome of the review was the founding

of the Power of Information Taskforce, convened by then
Cabinet Office minister Tom Watson MP in March 2008, and
chaired by early TheyWorkForYou contributor and former
MP Richard Allan. Grassroots and industry representatives
30 were joined on the Taskforce by various members of the
public sector, from Tom Loosemore (then at OfCom, now
head of Channel 4s own response to the web 2.0 agenda,
4IP) to Andrew Stott, Alex Allan (both at the Cabinet Office,
and the former mentioned again and again in interviews as
a changemaker within the civil service Andrew wears a tie
but actually hacks as well, says Tim Berners-Lee) and Natalie
Ceeney (National Archives). The secretariat was drawn from
the Cabinet Office (William Perrin and Alex Butler) and John
Sheridan (OPSI).31 It is perhaps important to note that the
early cross-cultural base of the OKF open data events was
repeated in the makeup of the Taskforce.
One of the most visible and striking outcomes of the work
of the Power of Information Taskforce was its Show Us A
Better Way competition,32 a prize fund of 20,000 for the
best proposals for reuse of public sector information. The
competition attracted around 500 entries, and was won
jointly by 14 projects.33 The Taskforces final report notes that:

The Taskforce has been impressed by the extent to

which access to geospatial data has been a recurrent
theme during its activities. For example, the Show
us a Better Way competition had around 500 entries
and of these over one third were for ideas around
maps and location.
Power of Information Taskforce Report34

This observation, together with mounting evidence for

example, from the Treasury study35 led the Taskforce to
give detailed recommendations around the freeing up
of geospatial data and reform of the Ordnance Survey. In
total, the Taskforce made 25 recommendations, including
the need for a well-resourced public service information
regulator, in the form of a radically better-funded OPSI
CITATION Gra09 \l 205.36
It was after the publication of the Taskforces final report
that Gordon Brown first met with Tim Berners-Lee on the
subject of open data,37 though it is unclear who made
the introduction. In June 2009, as part of a statement
on constitutional reform made in response to the MPs
expenses crisis, Gordon Brown announced the appointment
of Berners-Lee as a Government advisor.38 A Cabinet
Office press release39 confirmed that Berners-Lee would
be working with Nigel Shadbolt, Professor of Artificial
Intelligence at Southampton University both men had
met with Brown at 10 Downing Street earlier that year.40

30 The latter in the form of Richard Sargeant, formerly an official at

HM Treasury but by that time working as a public policy advisor

to Google.
31 Power of Information Taskforce 3.
32 Power of Information Taskforce 2.
33 For

the full list, see http://www.showusabetterway.co.uk/
call/2008/11/and-the-winners-are.html
34 Power of Information Taskforce 1, 2009.

Tim Berners-Lee describes his time as a Government advisor

as being characterised by a series of meetings, including
meetings with civil servants producing data. During those
meetings, he encountered a high level of enthusiasm at the
middle layer:

There were people whod actually got some datasets

that they were very happy to give out, they just
hadnt been given the okay to do it.
Tim Berners-Lee

Where people Berners-Lee encountered were unsure of

the open data strategy, he introduced them to mash-up
projects that were already transforming public sector
information:

I think when they saw just a few examples of how

information is reused by a third party who has
never been involved with the process before, thats
when they realised the value of the information that
theyve been producing is really untapped - its
not been exploited. Then they get very keen to be
a part of that infrastructure, to be part of putting
the data out there. And those two spheres, the
data sphere, as I call it sometimes, and the mashup
sphere are separate.
Tim Berners-Lee

This distinction made by Berners-Lee between the data

sphere populated by government departments and
other institutions producing data and the mash-up
sphere populated by civil society and private actors is a
crucial one, as it points to a fundamental shift in strategies
around transparency. The distinction is also highlighted
by Robinson et al, in their paper Government Data and
the Invisible Hand, which makes a case for third parties
repurposing government data along the following lines:

The biggest advantage of third party data processing

is to encourage the emergence of more advanced
features, beyond the delivery of data... Exactly which
of these features to use in which case, and how to
combine advanced features with data presentation,
is an open question. Private parties might not get it
right the first time, but we believe they will explore
more approaches and will recover more rapidly than
government will from the inevitable missteps. .. For
those desiring to build interactive sites, the barriers
to entry are remarkably low once government data is
conveniently available.
Government Data and the Invisible Hand 41

35 Newbury, Bently and Pollock, 2008.

36 Power of Information Taskforce 1, 2009.
37 Crabtree, 2010.
38 Brown, 2009.
39 Cabinet Office, 2009.
40 Crabtree, 2010.
41 Robinson, Harlan, Zeller and Felten, 2009.

14 TAI New Technology /Open data study

As observed above, it is the ease with which applications

can be developed by the mashup sphere, and not the mass
adoption of such applications, which provides the decisive
argument for opening up government data by the data
sphere. This theme is picked up in subsequent sections of
this report.42
Berners-Lee particularly applauds the decision of Cabinet
Office civil servants to release data.gov.uk in beta in
September 2009 to a self-selected community of data
mashers and other civic hackers before the official release in
January 2010, a move he saw as crucial to the success of the
official launch:

I think that was rather clever, the way they created

this community, and that was done largely by the
Cabinet Office putting together some open source
software and skinning it with CSS files
pretty promptly.
Tim Berners-Lee

It is tempting to draw parallels with this lightweight

and inclusive technical strategy which stands in stark
contrast to the vast majority of Government-commissioned
technology projects and the strategies introduced to the
Cabinet Office by the Power of Information Taskforce the
previous year.
Many of the people interviewed in the UK praised the
dedication of civil servants. This comment is typical:

Although it is questionable how far upwards the concept

of a middle layer of government extends, it is worth noting
that Wonderlich identifies actors at the most senior level
of the Secretary or Minister as being choke points in the
release of open data on occasion:

Overcoming that challenge - part of it is buy-in at

the secretary level. There are really exciting things
happening, for example, at HHS, our Health and
Human Services Agency. Thats got a variety of new
responsibilities with the new healthcare legislation
that passed, so I think its great that theyre on the
ball because they really have to be. If you look at
the Department of Defense, however, the Open
Government Plan they released yesterday was really,
for the most part, boiler plate language and plans to
make plans about other plans and to set up a new
commission that will report to another commission.
I dont know if thats characteristic of the Defense
Department or the people who are involved or if
its the secretarys fault. Im not sure how to assign
the blame.
John Wonderlich

Wonderlich goes on to describe the attitudes he encounters

from public servants unwilling to open up data:

Really, the civil servants have been what has made

this happen. Theyve understood what the benefits
might be, theyre really on it technically speaking,
in terms of understanding the web, understanding
some of these new internet technologies,
understanding the community-driven nature
of some of these projects. One cant emphasise
enough how on it theyve been.

Excuses you hear are if we release the data, youd

find out how bad it was, so its not that useful
anyway, and if we release too much data, then all
the data becomes less useful I think theyre really
just cover for insecurity about relinquishing control
because when you have a dataset that only you can
access, then you have a certain power that comes
along with that. And when you release it, you have to
take responsibility for the privacy problems
or the errors and everything else.

Jonathan Gray

John Wonderlich

And indeed, in the US, where the role of the middle

layer in the run-up to the launch of data.gov is less clear,
stakeholders still praise the attitude of at least some public
administrators:

Its a frequent occurrence for us to write about some

database or piece of information and be ignored by
the people who can make decisions about it, yet be
silently and secretly appreciated - and contacted - by
the people who actually administer the database.
Theyre our biggest allies but they are hamstrung
as far as their ability to advocate for improvement.
Its not uncommon to have meetings near Congress
or in an agency with someone who doesnt want
their boss to know because they dont know how to
appropriately advocate for an improvement or for a
data release... So sometimes well put up a blogpost
and six hours later have a really appreciative email
from someone whos happy to see that theyre being
noticed... And they usually understand the value of
their data better than anyone else.
John Wonderlich
42 See section entitled Utility or users: absent drivers?.

TAI New Technology /Open data study 15

One interviewee who had had direct experience of working

inside public administration was Nathaniel Heller, now
Managing Director of Global Integrity, who spent a short
time from 2002 as part of the US State Department. He
characterises his experiences there as follows:

Top-level drive

I think the challenge is not so much in a way a

cultural friction or the government being offended or
annoyed by civil society, its that all of the incentives
within a bureaucracy at least the government
bureaucracy tend to be aligned and oriented
around the rest of the bureaucracy and not the
public. So, for instance, I mean this was a weird era
because the interagency system in the US was so
broken during the Bush administration, but we spent
98% of our time basically fighting with the Pentagon
over various issues and trying to limit the damage.
That was our whole universe, it was centred around
dealing with interagency process, trying to promote
the department vision as opposed to somebody
elses. The engagement, talking with the public on
these things, just never happened, partly because
I was in a policy portfolio that was very superiororiented and fairly classified - its not like you would
go out and talk about these things on a daily basis.
But the point was that everything I did, my success
as a bureaucrat, my promotion, my pay raises, were
tied almost entirely to how successful I was working
within the bureaucracy. It was not at all tied to the
public or to public policy outcomes and I think thats
the challenge.

John Wonderlich

Nathaniel Heller, Managing Director, Global Integrity

You probably want to try and ask, specifically,

how did Andrew McLaughlin get his job, and how
did [Beth Noveck] get her job? Within what was
essentially a hyper-political campaign, who decided
this was worth having? Given that its very, very
different from the technology stuff that got Obama
elected that was very high-tech, but it wasnt very
open-data-y, or even open source. So thats an
interesting question, and I dont know the answer.

This frank and intriguing insight into the inner workings of

government may offer clues to both the successes and the
challenges of both data.gov and data.gov.uk.

The idea of open data as a thematic issue has been

enormously raised by the Obama administration and
the creation of data.gov

Whereas the trajectory towards data.gov.uk is a story of

policy reform especially with regards to the licensing
of public data the data.gov data portal faced no such
obstacles. There was, however, a historic lack of political will
to realise the value of public sector information:

Before Obama was elected, there were a bunch

of structures that were supposed to do this, but
didnt. So there was a clear lack of political will
to take on this kind of issue. Perennially, we have
laws that try to fix eGovernment or create more
transparency or electronic access to information
and theyve often failed in the past, or theyve had
one or two successful components and the rest of it
ends up being ignored. So I think at its heart, whats
happening now is a cultural movement and its a
change in public expectations. The structures in
Government are largely secondary to that.
John Wonderlich

Obamas campaign was characterised by the employment

of new technology in radical new ways. But as Tom
Steinberg points out, the technologies employed by
the Obama campaign were of a different hue to those
advocated by the open data community:

Tom Steinberg

Absent from Steinbergs remarks, but surely key to the

direction of data.gov, was Obamas hiring of Vivek Kundra
as the United States Chief Information Officer in March
2009, just before the data.gov portal was announced. As
CTO for the District of Columbia, Kundra was a figurehead
in the open data community. As mentioned above, Kundras
data catalogue43 had been the inspiration for early OKFsponsored discussions of a data registry for the UKs public
sector information.

43 http://data.octo.dc.gov/

16 TAI New Technology /Open data study

John Wonderlich goes some way towards answering

Steinbergs question:

I dont know when Obama first started up a

relationship with a lot of the Google officials, but its
a pretty deep one: he gave speeches at the Google
headquarters pretty early on in the campaign. And
Andrew McLaughlin, for example, was one of the top
Google people and hes now in the Office of Science
and Technology Policy in the White House. But
beyond that, I think theres a couple of factors. One is
a generational sense, that he wanted to be viewed as
a change agent, to be associated with technological
savvy. And beyond that, throughout his Senate
career, he also had a particular appreciation for the
power of what it is to post information when the
broader public has access to that information. So
one of the Bills he was responsible for in the Senate
created access to the first Federal database of Grants
and Contracts, which is USAspending.gov, so there is
that link pretty early.
John Wonderlich

However, Wonderlich notes that the long-term impact of

Obamas clear vision for data.gov is yet to be established:

I think right now what were waiting to see is to what

degree the Presidents clear political will translates to
actual change through civil servants and throughout
the bureaucracy, and its interesting to see, you
know, some agencies, that political will can transfer
rather quickly down through the ranks. In others, it
gets caught right below the secretary and doesnt
turn into very much, at least not yet.
John Wonderlich

If Obamas decision to back data.gov came as he rode

a wave of political optimism, the contrast with Browns
position when he announced the appointment of Tim
Berners-Lee could not be starker. An unelected Prime
Minister44 speaking in the shadow of one of the worst
scandals to hit British public life in decades over MPs
bogus expenses claims Browns was a very different sort
of political moment. Tim Berners-Lee gives his impression of
Browns thinking:

I think hes excited about making the country more

efficient. Transparency was a big thing, in the sense
of allowing the British public to hold the British
government to a high level of effectiveness. The
way Gordon Brown talks about transparency, its
not about trying to find people cheating. Its about
looking at the state of the nation and seeing how it
changes and thinking about how it can be better and
then providing feedback at every level.

There is no doubt that the combination of a world-famous

inventor, bestowed with the backing of the highest Minister
in the land, was a catalyst to change across all Government
departments, and particularly in the case of the stillcontentious Ordnance Survey data.

So the sequence was to go and talk to people at a

ministerial level and get them to urge the different
departments to do this, and simultaneously to talk
to developers, the people who had been working
for a while on this sort of stuff in a relatively quiet
way...if you look at the ministers, each minister had
to basically command each department to say,
Yeah, were going to do this.
Tim Berners-Lee

The success of Berners-Lee in bringing Ministers on board,

in contrast to reports from the US experience that data.
gov aspirations often got stuck at or just below Secretary
level, leads the researcher to speculate on the strategic
importance of the Prime Ministers chief advocate for the
project being at one remove from traditional political
power. Did the presence of Berners-Lee famous and
celebrated in a sphere completely removed from politics
encourage Ministers to tune down their political
sensitivities, put aside personality politics and point-scoring,
and just get on with it? It is tempting to believe so.
Finally, those who have followed the debate sound a note
of caution to anyone wishing to suppose that it was Tim
Berners-Lee alone who created the success of data.gov.uk:

The received picture in the press, certainly in the

last year, has been one of Tim Berners-Lee coming
in and revolutionising everything in Government.
Which of course is, to a large extent, true in terms
of the excitement and the catalytic potential of
him coming in and making all of this work more
high-profile. And also in terms of communicating
it to the public, saying that the inventor of the
World Wide Web is involved adds credence to
something that could otherwise be viewed as quite
obscure. Bringing Tim Berners-Lee on board at that
time, I think, was crucial, in terms of the political
value thats been obtained. That said, recent
developments in the last year or so with data.gov.
uk have cast a much longer historical shadow, going
back to the Show Us A Better Way competition, but
before that too. The fantastic work that OPSI have
been doing for a long time, which have their roots
in, for example, recommendations made in studies
such as the Cost Cutting Review, back in the late
nineties or early 2000s.
Jonathan Gray

Tim Berners-Lee

44 Brown

took over the position when he became Labour Party

leader after Tony Blair resigned in 2007. His election as party

leader was unopposed and automatically bestowed on him the

role of Prime Minister, and he declined to call a General Election

afterwards, as might have been expected.

TAI New Technology /Open data study 17

Utility not users

the absent driver?
Was end-user takeup a key driver in the journey towards
data.gov and data.gov.uk? Certainly, the potential for
broadening civic engagement of applications developed
by the small communities of civic hackers that congregated
around, say, GovTrack.us in the US, or MySociety in the UK,
came up frequently in interview. But more often, it was
the utility of applications (in contrast with the resources
expended to produce them) and not their broad user bases,
which seem to have inspired officials further up the line to
engage with the open data agenda.
In the Power of Information Review which covered many
aspects of the interface between Government and Web 2.0
beyond open, linked data Steinberg and Mayo highlight
the existence of large communities of ordinary British
citizens congregating around user-generated information
websites, such as the parenting website Netmums
(described in the report as having 275,000 users) and
MoneySavingExpert (2.5 million users).

The proportion of people using such sites to

help themselves and others is now on a par
with the friendly societies and mutuals of the
nineteenth century

And yet where user-generated information websites were

promoted on the basis of existing large user communities,
websites which repurposed government information were
promoted by advocates on the basis of the utility they
provided when compared with their low cost; on the basis
of their potential for wide and beneficial usage, rather than
on existing usage patterns. This is despite the fact that one
case study of TheyWorkForYou.com shows that it attracts
between 250,000 and 500,000 users each month46 - a not
inconsiderable number while a separate study shows that
another mySociety civic engagement tool, WriteToThem.
com, had around 100,000 users in 2008.47
A discussion of barriers to broad use of civic engagement
websites in developing and middle-income country
settings takes place in the following sections of this report.
Among other topics, this discussion highlights the roles of
intermediaries, such as traditional civil society groups, in
taking information from open data-driven web applications
and disseminating it across populations less wired than
those found in the US and UK. Clearly, to ensure real impact,
a potential user base should be a prerequisite when it
comes to deploying such initiatives elsewhere. But from a
strategic standpoint, it is worth noting that the existence of
this user base was not a key part of the rhetoric used to talk
up the open data idea. Rather, it was utility, and not users,
which won the arguments in the UK and US.

Power of Information Review45

45 Mayo and Steinberg, 2007.

46 Tactical Technology Collective, 2009.

47 Escher, 2009.

18 TAI New Technology /Open data study

International
perspectives

TAI New Technology /Open data study 19

The UK structure and the way this happened in the UK is totally unlikely to
work in the developing world. And I think we should just let go of it.
Ethan Zuckerman

This may be one of those things where putting

a certain amount of effort in now will have a
disproportionate effect.
Tim Berners-Lee

The researcher spoke to a number of regional and domain

experts (see list of interviewees) in order to understand how
the strategies employed to realise data.gov and data.gov.uk
might translate to a middle-income or developing country
setting, and to uncover additional factors that might come
into play should such initiatives be attempted. Interviewees
ranged from experts on the introduction of freedom of
information (FOI) laws and computerised administrative
systems around the world, to experts in fiscal transparency
and budgetary monitoring, to civil society practitioners
in India, Kenya, Tanzania, South Africa and the central and
eastern European (CEE) region.
A high level of scepticism was encountered on a number of
issues. Questions were raised around whether privileging
open data initiatives over information access rights was a
desirable strategy in countries where such rights were not
yet enshrined in legislation. Interviewees also questioned
whether datasets equivalent to those being released on
data.gov and data.gov.uk existed in some of the countries
discussed. Interviewees also questioned the impact of data.
gov and data.gov.uk, which remained, for them, unproven.

The relationship between

FOI and open data
One interviewee, who had spent considerable chunks of his
career advocating globally for FOI laws, felt that any move
away from this reform programme towards open data-type
initiatives would represent an unfortunate change
of priorities.

I think that for those countries which dont have a

law yet and these are not mutually exclusive by
any means, but certainly getting a law is a major
priority and that if there are limited resources
available for the country, I think for me that would
be a greater priority.
Toby Mendel, Executive Director
Center for Law and Democracy

Such scepticism was not limited to the old guard. One

interviewee at the centre of the open data movement wished
to make the following comment on an unattributed basis:

There is a gap between initiatives that are based

on governments giving out things that they want to
give out, and governments creating rights that mean
that they give things out all the time that they maybe
dont want to give out. The Freedom of Information
Act is a right, it gives people a right, whereas these
data initiatives in America and Britain tend to be not
rights, but more like gifts. And that doesnt make
these data websites a bad thing, not at all it just
makes them not as good as rights. Because there
are two problems: firstly, theyre not necessarily as
strong; and secondly, they assume to some degree
that the government knows what the public wants.
And in some cases thats really obvious, because
[campaigners] are shouting all the time that the
public wants maps. But when it turns out that actually
the really valuable data exists in something that only
one person knows about, and if they could get hold
of it, then they could make a billion pounds, if they
dont have a right to get that data, and they dont
have any leverage, then were not likely to see that
value added.
Anonymous

20 TAI New Technology /Open data study

Nonetheless, a countering viewpoint was expressed by an

activist working in Kenya, who felt the need to call time
on expensive advocacy efforts in favour of FOI, given the
limited success of such efforts:

For a long time the focus was on getting the

Freedom of Information Act passed. Lets do that
first and then once that law has passed, then well
have the data. And what you see is that I think
theres, even [after all the] money and advocacy,
theres still only two countries in Africa that have
Freedom of Information Acts that have gone
through. In Uganda they passed the Bill but they
didnt operationalise it.
Ory Okolloh, co-founder Mzalendo
co-founder, Ushahidi (Kenya)

Notwithstanding this contrast in viewpoints, there were

common threads to discussions on the relationship
between FOI efforts and open data initiatives. For example,
the idea that open data catalogues might capacitate
bureaucracies48 in those countries which did have FOI laws,
but whose public servants struggled to meet both the
proactive and reactive responsibilities enshrined in those
laws, proved to have some traction:

If you look at the right to information legally,

there are two major sides to it: one is proactive
disclosure, so all of these laws or most of these laws
place an obligation on governments to push out
information... And then there is a request-driven
side of it. And what you find is that governments in
many, many countries - I would say most countries
outside the highly-developed country zone - have a
great struggle meeting their proactive obligations.
I think we need to recognise the very important role
of capacitating the public sector in this area and
not working with civil society initiatives: it almost
sometimes may even undermine the government...
[But] I think that trying to link these kinds of
initiatives to promoting efficiencies and perhaps
even building back capacity in government would
be a good antidote to some of the problems that
Im highlighting.
Toby Mendel

48 Robinson, Harlan, Zeller and Felten, 2009.

A story of bureaucracies struggling to meet disclosure

responsibilities shifting to a proactive publication agenda
was told in the context of India:

So after this law was enacted, thousands of requests

have reached Government officers and Government
officers are snowed under by these transparency
requests. So some of them are trying to deal with
this deluge of requests by proactively publishing
datasets on their websites.
Sunil Abraham, Executive Director
Centre for Internet and Society (India)

A broader context of an emerging democracy struggling

under the weight of including all stakeholders in public life
was related in Tanzania:

Tanzanias been going through reforms for many

years now, and... there were all these committees
created and a very elaborate and rather cumbersome
machinery set up to have everybody involved in the
process. And I think now theres a disillusionment
with that and the emphasis has partly shifted to look
at [whether], if everything was open then different
people in their own time could get on with analysing
and using the data as and when needed.
Rakesh Rajani, Founder
Twaweza (Tanzania)

TAI New Technology /Open data study 21

Questions over impact

Data characteristics

Open data catalogues as a valid initiative to pursue in a

setting outside of established, higher-income democracies
was not the only concept to attract scepticism. Interviewees
questioned the impact of the US and UKs own data portals:

Data availability in developing and middle income

countries was also flagged as an issue by global experts.

We all think the techniques are kind of sexy and

exciting, but does it matter whats the impact?...
The corruption drivers in the US, particularly around
money and politics, are completely unsolved by
greater amounts of money and politics information
being put out there... I think the political economy
context in which one is working matters hugely.
Theres going to be a limit at which point it doesnt
matter anymore - this information is getting out
there, everybody knows the problem, everybody
sees it, we dont need more databases, more
mashups, more maps - there is something else
missing, which is some nasty combination of a lack
of leadership and the very real political economy
choke point that simply prevents these kinds of
reforms from being possible.
Nathaniel Heller

I think the whole open data thing is a bit Im not

saying its emperors new clothes, but its verging
on that... I believe in the idea that stuff should be
open and I believe that in principle if you make it
open then... people will use it for socially positive
ends as well as possibly [the complete opposite].
But... the thing that annoys me about it, isnt so much
the people who are doing it, but [that] people think
thats it, thats the job done. You know... the idea
that if we make the data open there is this layer of
civically savvy hackers who will then do the necessary
job. And I dont think thats the case basically.
Because the necessary job isnt TheyWorkForYou
or even WhatDoTheyKnow. Thats not reaching out
into communities or engaging with... social agendas.
And it could do, open data really could do [that]. I
think it could be part of a picture of civic activation,
but it isnt at the moment and I think its not because
whats happening is a bad thing, but because its
only 50% of the picture. And the worry I have is that
that 50% of the picture is being portrayed as 100% of
the picture.
Dan McQuillan, Social Innovation Camps
(Central and Eastern Europe)

Although the researcher has a lot of sympathy with

such viewpoints, she notes that it might be too early to
measure the impact of open data portals on wider social
and political issues.

The idea that this stuff exists, is in digital format of

some kind and is anywhere close to standardised is
probably just an impossible leap I just dont think
its there.
Nathaniel Heller

My simplest example for this would be years ago,

talking with the government in Senegal and trying
to plan an intervention based on electronic property
records and... the Senegalese government was at
first very enthusiastic. And then we started talking
about the physical challenge of it and what we
ended up discovering is that before we built an
electronic property records system we actually had
to build a property records system. It wasnt clear
that that data existed in paper form and that to build
the sort of government data transparency system
we needed, in many cases we would have to do the
basic data collection.
Ethan Zuckerman, , Senior Researcher
Berkman Center for Internet and Society

Kenya is probably one of the most open societies

in sub-Saharan Africa, but it is still a very serious
challenge to get that sort of data, and I think... its
a challenge because the data doesnt exist and,
in some cases, it is a challenge because people
understand just how important that data actually
is and what it could be used for.
Ethan Zuckerman

However, regional civil society actors pointed to various

sources of data that they used in their everyday practice.
As might be expected, availability, timeliness and digital
accessibility varied widely:

Broadly speaking there are three kinds of datasets

that come to mind. One is governance or
administrative data that is collected usually year on
year from schools, health centres and so forth. Then
there is survey data..., and there the key actor would
be the National Bureau of Statistics, whereas for
the administrative data the key actors would be the
Ministries. And then the third one is the data related
to ad hoc studies as well as audits that are... often
donor-driven reports such as expenditure tracking
surveys. Now theres push to try to see if all three can
be available.
Rakesh Rajani (Tanzania)

22 TAI New Technology /Open data study

He made the following remarks about trends towards

online availability across the world:

Apart from these types of data which is the law and

GIS and census and infrastructural data, [for] all
the other Department-related data, availability in
digital format varies from state to state. Theres no
standardisation, very little standardisation.

[For] almost every government that we assessed in

the previous round of our survey that was published
in 2008, the Ministries of Finance have functional
websites. So its increasingly catching on. I think
there are a couple of countries where they didnt
have that, so its the exception where information is
[not] available online... [but] in many countries what
weve noticed they do is that they make PDF versions
of these documents available. In some cases they are
scanning the documents and making them available
in a PDF version after its been scanned. And so that
obviously is indicative that theyre not really working
on these documents necessarily online or at least
theyre making a choice here to make it available by
scanning it rather than making the original electronic
document available.

Sunil Abraham (India)

The Auditor General whose reports we do rely quite

heavily upon for the purposes of reflecting on service
delivery in prior years, that information is available in
PDF format
Jay Kruuse, Head of Monitoring and Research
Programme, Centre for Social Accountability,
South Africa

All the work we do [with Mzalendo] is manual, so we

have to literally cut and paste information if we can
find it. Its gotten a lot better from when we started.
Now things like the Hansard are on the website
pretty much in soft copy and up-to-date. When we
started you couldnt get the Hansard at all online. So
its improved but its still either in a PDF or in a Word
document that you cant crawl
or extract information from.

Vivek Ramkumar, Manager, Open Budget Initiative

Ory Okolloh (Kenya)

Vivek Ramkumar of the International Budget Partnership

provides the following data49 with regards to the availability
of government documents in 85 countries that the
organisation regularly scrutinises as part of its monitoring
efforts. (See table below).

Key budget
documents

No. of Countries
Making Available
Online

No. of Countries
Making Available
On Request

No. of Countries
Producing the
Information but
not Publishing It

No. of Countries
not Producing
the Information

Pre-budget statement

27

29

26

Executives budget
proposal

49

13

23

Citizens budget

13

67

Enacted budget

68

13

In-year reports

63

13

Mid-year review

18

21

42

Year-end report

50

14

14

Audit report

50

21

49 International Budget Partnership.

TAI New Technology /Open data study 23

Several interviewees agreed that centralised information

on the current publishing practices of governments
either in the context of proactive requirements enshrined
in FOI law, or in a looser context was lacking. A call was
made to map the availability of datasets (see Avenues for
further research section):

I think it would be great to start mapping what

datasets exist within governments, but Im going to
stand by my scepticism; I think a lot of the data that
you want data is not an easy thing to generate
and, you know, Ive spent a whole lot of time in a
whole lot of African ministries and a lot of the critical
records are still on paper its not clear to me that
those records are getting digitised or are getting
digitised in a meaningful way I dont thinks its as
simple as just taking a database and publishing it in
many cases.
Ethan Zuckerman

It is worth noting that many interviewees spoke of

personal data, such as census data, in the same context as
non-personal data. Indeed, in interview, Tim Berners-Lee
appeared to exclude the idea that government data portals
might encompass personally identifiable information only
in the first six months of the project:

the rule of that first six months was to get the lowhanging fruit, to show that online data was valuable
but to do it without attempting anything which
was questionable, like not going anywhere near
personally identifiable information.
Tim Berners-Lee

The researcher notes that funders who wish to foster public

goods through the better provision of digital information
should consider the dramatic implications for the right
to privacy of datasets containing personally identifiable
information being released to the public.

50 Gray, Hatcher, Hogge, Parrish and Pollock, 2009.

Other issues around the character of governmentmaintained data were raised in interviews. The specification
of standard formats for data publication was one such
issue. The format in which a government releases data can
have a positive or negative impact on that datas reuse by
third parties, particularly if the government chooses to
release data in proprietary, as opposed to open format. The
following guidance is taken from a techno-legal manual on
open data in the context of international aid, but applies
universally to open data projects:

Proprietary, non-proprietary and open file

formats: On the one hand some file formats are
non-proprietary and open, which means they can
be used or implemented by anyone with little or
no restriction. Prominent examples include HTML/
XHTML, OpenDocument, PDF, TXT, XML7.
On the other hand some file formats are
proprietary, which means that there may be
restrictions on how the format may be used, and
certain software packages may be required to
read the files. Prominent examples include MPEG
Audio Layer 3 (MP3), Windows Media Video (WMV),
Microsoft Word (DOC/DOCX) and Microsoft Excel
(XLS/XLSX)8.
...If the file format specification is publicly available,
then there is less risk that prospective re-users will be
required to use a particular piece of software, or that,
in the worst case scenario, the format will become
obsolete and unreadable without software that is no
longer supported.
Unlocking the potential of aid information 50

24 TAI New Technology /Open data study

Whereas proprietary standards lead to monopolies and

lock-in, open standards allow for open competition, and
so are vital in encouraging vibrant competition in the
mash-up sphere of open data both in the context of
private sector application development, and civil society
application development. However, the dominance of
commercial information technology vendors around
discussions of format standards, each one keen to have
their format approved as the national standard and to reap
the benefits of the monopoly conditions that would flow
from such a decision, risks the development of national and
international open data standards in some regions:

The Open Standards Policy document is in [its third]

draft and there have been two rounds of consultation
over the last two years. There have been two rounds
of consultation with the corporate sector and civil
society, but there are several areas of contention.
Some corporates representing proprietary interests
want the Open Standards Policy to have space for
patented technologies which might have royalty
implications. So thats the first point of contention
and the second point of contention is that some
corporate sector organisations believe that in a
single domain, multiple standards should be allowed
to operate, there shouldnt be a mandate on a single
standard. Over these two particular disagreements
the policy has been delayed and even today its
very difficult to say whether that policy would ever
be fully published.
Sunil Abraham (India)

If the donor and the ministry are not specifying data
export standards, the idea that this data is open,
what youre going to get instead is someone coming
in and [proposing] a proprietary system, heavy focus
on security, a system where it is almost impossible to
squeeze the data out of it in the end. I mean, I think
where there might be an opportunity in all of this is
thinking about how you construct a data standard
and policy for donors... What Im not sure about
is how long it might take for international or local
contractors to figure out how to respond to that and
figure out how to build good bids around that.
Ethan Zuckerman

Recently in Kenya there was a seminar on connected

government. There was huge vendor representation
there and I was following the conference on Twitter
and all of them were pushing a solution integrated,
customised solution, but solutions that are internal to
Government, nothing really about the public.
Ory Okolloh (Kenya)

51 http://www.pmg.org.za/

Commercial models of cost recovery around public data

of the sort that caused so many problems for the open
data movement in the UK - were also described in at least
two of the regions covered by the expert interviews:

Yeah, so GIS geographic information is available...

But until very recently the Government always saw
this as a revenue stream because certain private
organisations were willing to pay for these datasets.
Sunil Abraham (India)

Definitely with geographical information, I think

maps, they sell maps or they sell that kind of data,
so that might be an issue.
Ory Okolloh (Kenya)

A new type of commercial/cost-recovery model was also

encountered, involving civil society practitioners in South
Africa. The Parliamentary Monitoring Group (PMG)51 is a
consortium of NGOs in South Africa working to publish
data about the activities of South African Parliamentary
committees. The data it publishes is based on reports and
recordings made by volunteer monitors it enlists from the
general public and who attend the committee meetings in
this capacity. Although PMG is funded by various grantgiving organisations (it lists current funders on its website
as the Open Society Foundation, the Royal Danish Embassy
and USAID), it cross-subsidises free access to its materials
for South African civil society by charging commercial
organisations a subscription fee:

PMG will have minutes of meetings of committees.

Itll have audio recordings. Itll have submissions by
interested parties and business and political parties
and itll have draft Bills and Gazettes which are being
considered by Committees. But not everything is
accessible....At this stage because were a non-profit
we are not being charged, but I believe that there
are certain charges that are applicable. But thankfully
because we meet the requirements for exemption
we dont need to.
Jay Kruuse (South Africa)

TAI New Technology /Open data study 25

The implications of this for open data portals are unclear, but
the existence of PMG points to a different path being taken
by civil society actors around government data than that
seen in the UK and US. Those planning interventions around
open data in South Africa should consider the impact of such
interventions on existing civil society initiatives.
The issue of data quality was also raised:

There is a series of public interest disclosures that

are expected from every single electoral candidate.
However, when the candidates make this disclosure
they usually send it in paper form, not in electronic
form and what they do is they write their names
differently. So its very difficult for you to run [any]
analysis without spending a lot of time... manually
cleaning up the data.
Sunil Abraham (India)

Finally, a set of three questions to use when evaluating

different countries suitability for a data.gov-type
intervention was proposed:

Does the data exist? Is it released and does it

threaten someone? and then, Is it usable by
journalists or somebody else?
Ethan Zuckerman

The researcher agrees that the existence of digitised

data seems like a necessary prerequisite to the success
of a project like data.gov, and recommends that further
research be conducted to establish rates of administrative
digitisation across the world. In subsequent sections of
this report, the researcher goes on to examine who open
data might threaten (in particular see the final paragraph
of this section on high stakes faced by some government
bureaucrats), and who might be the user-end beneficiaries
of open data projects.

The three-tiered approach:

top-level drivers
Given the clear emergence of a three-tiered approach in both
the data.gov and data.gov.uk trajectories, interviews with
domain and regional experts focussed on their experiences
with each of these layers of society. Several issues around
the top level were discussed, including how a character like
Tim Berners-Lee might wield influence at this level, and how
different political cultures might characterise the top level
response to open data strategies in different regions.
A push from the top level was seen as a good strategy
for overcoming wider inertia around opening up to citizens
in Tanzania:

I think the larger constraint is Government is

simply not in the habit of, it hasnt quite got to
understanding that this is public data. As far as
theyre concerned this is their stuff and Why the
hell should we give it to anybody?. Now the way to
get around that, if one got a sufficiently high level
commitment from the President and I wonder
whether theres also a role here for Parliament and
so forth and possibly legislation that was on the
one hand broad enough to not get stuck in some
detailed problem that somebody might have a
beef with, but sufficiently detailed so it doesnt just
become a wishy-washy commitment in principle,
but has some teeth to it. So if one got that balance
right then the technical piece can kick in, meaning
the middle level can then come in and start doing
their part provided there was this high-level
commitment that was publicly made.
Rakesh Rajani (Tanzania)

The role of superstar power, such as that clearly fulfilled

by Tim Berners-Lee in the UK experience, was generally
greeted positively, with some caveats. Other potential toplevel personalities were also mentioned:

[India] is a country where film stars get voted into

power, so there is a bit of personality cult for sure.
And thats why theyre using Nandan Nilekani
whos the former boss of Infosys to run the national
identity project.
Sunil Abraham (India)

26 TAI New Technology / Open data study

[Tim Berners-Lee] would not be viewed by the

Government as someone like me or other lobbyists
who have been working on this theme. Theyre
like Obviously we know where shes coming from.
But that leverage and the fact that he could point
to other examples I think it would make a huge
difference. Ive also been encouraging Google
which engages with a lot of Governments to
leverage their access and their conversations, even
as a vendor. They could be pushing the Cloud or
whatever it is that they sell to Governments. But, as
part of that process, they could certainly influence
the standards or the format of information thats
going into the Cloud.
Ory Okolloh (Kenya)

Tim BernersLee, I doubt whether there are more

than 50people in the whole country who even
know who he is.
Rakesh Rajani (Tanzania)

One needs to be careful that it doesnt sound like

the bad old imperialist US. There are ways of getting
it right and there are ways of botching it. But, done
well, I think theres nothing that can match Obamas
star power particularly in East Africa to get this done.
Rakesh Rajani (Tanzania)

Different trends in political structures and cultures were

also highlighted. The impact of nepotism, and the big man
culture of African politics, were mentioned negatively in
the context of capacitating the middle layer of government
administrations:

Its highly centralised. There are high levels of control

over what can be released and in many instances
its completely unnecessary but its part and parcel
of where weve come from. And I think the fact that
we have a lot of people within Government who
are appointed to a position in the Public Service on
the basis of connections. We do have high levels of
nepotism and patronage in the public sector and
so people are very much aligned to the wishes of
those who hold power and that results in a culture of
secrecy or I cannot do this without approval from on
high. And the result is that information which really
isnt sensitive and which should be distributed quite
freely is caught up in the red tape. So we would need
to change that climate.
Jay Kruuse (South Africa)

Yes, absolutely, there is a culture of the big man, and

the big man culture isnt just a governmental culture,
its also a commercial culture. The biggest problem
we had doing software development work in subSaharan Africa from 1999 to 2004 was management,
and it was basically how do you persuade an
entrepreneur who runs his own company that he
needs to treat his software developer as a precious
resource because they could go elsewhere and, at a
certain point, you actually ended up with companies
essentially saying, Well, I dont want my employees
to be any smarter, that would be bad for me.
Ethan Zuckerman

Its still a big man culture in most African countries.

So they would not take well to feeling like theyve
been left out of the loop. Its a protocol-type issue.
So maybe you do [the two streams of advocacy]
in parallel, but Im not sure it would work as well
...if the Minister finds out people are working, the
junior guys, without him even being aware. Its more
of an ego management issue than anything else
and I think that will be a bigger problem in African
countries maybe than it was in the UK.
Ory Okolloh (Kenya)

TAI New Technology / Open data study 27

The three-tiered approach:

the middle layer
In many ways, the importance of the middle layer of
government bureaucrats in the success or failure of an open
data-type initiative was as present in the experiences of
interviewees as in the US and UK experience:

But other interviewees saw capacity-building as a simple

challenge to overcome:

I think the primary constraint is not going to be

technical. So even where there is a technical gap,
it could be easily filled.
Rakesh Rajani (Tanzania)

I think there would have had to be some baseline of

support at the mid-levels of management in order to
get this done

The journey towards greater collaboration between civil

society and the middle layer of government seems to have
already begun in some of the regions discussed:

Nathaniel Heller

I think the non-profit sector is also beginning

to realise that it is not very useful to be in an
antagonistic position with the Government. It would
be best if the sector also took some responsibility for
building the necessary data ecosystem.

The idea that independent groups out there may

want to analyse the data themselves and come
up with perhaps different conclusions from what
Government comes up with is not quite accepted.
Rakesh Rajani (Tanzania)

Messaging to change attitudes at this layer might also play

out in a similar fashion:

Something else that I find useful is saying...

Hey you guys do so much good work or The
perception out there in the public is that you guys sit
around and do nothing. By putting this data out, its
in your interests.
Ory Okolloh (Kenya)

Question marks were raised over the capacity of this layer of

government to meet the open data challenge:

Government bureaucrats, unlike British bureaucrats,

they move around a lot. They dont have the domain
expertise or they dont know computer science or
information sciences
Sunil Abraham (India)

One of the things to think about is that in the

computerisation of African government ministries,
the computers usually go first to the minister and the
deputy secretaries, and those arent necessarily the
people who are going to open things up.
Ethan Zuckerman

I think sometimes they just dont know any better or

theres just not... someone making that compelling
argument about why a PDF is not enough.
Ory Okolloh (Kenya)

Sunil Abraham (India)

Even those NGOs who have been working on

electoral transparency have realised that they cannot
do all the data cleaning themselves, that they have to
work together. So I think there will be more and more
working with the Government in the near future and
various top-down and bottom-up projects that try
and build consensus around data standards.
Sunil Abraham (India)

You need to change mindsets. I belong to a civil

society network of access to information practitioners
who are requesting information, but there is also the
odd information officers forum so people who are in
Government who are appointed by law to consider
requests for information, there are regular meetings
of those forums to try and break down the history of
secretiveness, defensiveness and unwillingness to
provide information.
Jay Kruuse (South Africa)

28 TAI New Technology / Open data study

In others, there was still much ground to cover:

I think, what I havent seen enough of is people

engaging not even necessarily with the top officials,
just with the guys who have to deal with this on a
day-to-day basis. So a lot of Departments now have
CIOs, CTO type guys. Those are the people, or the
e-Government Directorate and not necessarily the
Minister himself because he can just say Yeah okay,
we agree. Thanks, bye.
Ory Okolloh (Kenya)

And one region-expert interviewee saw the separation

of civil society and government, when it came to new
movements for social change that included technological
innovation, in a fairly positive light:

I think the incumbents havent woken up to [social

innovation camps]. I dont know how much youve
been to these places, but the incumbent situation
is pretty hardcore. Power is a very live issue. And
networks of influence are pervasive, and corruption
is pandemic. I dont think theres any interest at
all [at that layer of government]. Which is nice in
a way because it means theres a strong contrast
with this stuff. Yeah, theres none of this tokenistic
doo-dab, so theres still a possibility to believe in
its radical potential.
Dan McQuillan, (Central and Eastern Europe)

Finally, the high stakes faced by bureaucrats in the least

developed parts of the world were highlighted:

In 2003, I was asked by the new USAID administrator

in Kigali to go sit down with the head of the Rwandan
Customs service and my task was to try to figure
out why the Rwandan Customs service hadnt
computerised. Theyd gotten all the money from the
US government, the system had been on the ground,
installed for 18 months; no one was using it. And I
spent four hours with the director of the Customs
service and, for three hours, he basically told me it
was hard, it was difficult, people didnt want to use
it, there were problems with it. Finally, he got up, he
closed the door to his office, and he said, Look, I
dont want to get killed and I dont want you to get
killed and then proceeded to explain to me that
corruption in the Customs service was so rife within
Rwanda that the computerisation of it would reveal
whos stealing what and that there was so much
money at play that he literally feared for his life if he
used the system. So, its important to realise that in
some of these cases, its not just something simple.
Ethan Zuckerman

TAI New Technology /Open data study 29

The three-tiered approach:

civil society
Where the existence of civil society actors pushing for open
data was remarked upon by interviewees, such groups were
characterised as small, a situation consistent with that in the
UK and US:

I think the specific push for data being transparent

is driven again by a relatively small group of people.
Theres a broader push for Government to become
transparent, but that isnt focused on data.
Rakesh Rajani (Tanzania)

The main constraint is that it all comes down to

people and there are very few people who have
the wherewithal to really do this within Tanzanian
civil society

Although the ability to stimulate civil society was

highlighted, it was in the context of general transparency
advocates, and not the focussed, tech-minded actors
witnessed in the US and UK contexts:

Getting movements going to push for openness is

almost the easiest advocacy thing you can do I
mean you go into a country, bring together ranges
of potentially interested stakeholders, and youre
almost certain to launch a campaign.
Toby Mendel

However, the attraction of new technology was

described as something universal, leading the same
interviewee to conclude that finding such people in any
relatively technologically-developed country would not
be a great barrier:

So there is engagement, there are people who get

it and there is certainly a group of developers who, if
the data was open, perhaps could be incentivised in
the way theyve done in the US and UK with Apps for
Democracy type competitions. But not necessarily
on the advocacy.

People just love to play around with this stuff,

and that is a potent force to harness and its not
susceptible of those problems with parachuting in.
Of course, local people, if its uninteresting, wont do
it. But I think what drives whoever puts that stuff on
the UK and the US sites, I think that the similar kind of
person will exist in most of the more technologically
developed countries where this thing might be a
goer

Ory Okolloh (Kenya)

Toby Mendel

Ushahidi has gotten so much attention because it is

one of the very few players in the space; it is one of
the few really innovative ideas to come up.

Certainly, practices of civil society actors in US and UK, in

particular the hijacking of data outside of the permitted
licensing terms but in the public interest, were already
being mirrored in Kenya:

Rakesh Rajani (Tanzania)

Ethan Zuckerman

Even in the US context, the need to stimulate civil society to

participate in open data projects was highlighted:

I think the bar has been set in many ways almost

too high and theres an implicit expectation or
assumption that if we just get it out there people
will figure out how to use it and automatically push
a demand side agenda thats for the good of the
public and that leap, to me, is really a big one. There
has been very little evidence that thats ever been
done in a meaningful way.
Nathaniel Heller

In my work Ive always appropriated stuff that even

we shouldnt really have access to, but our argument
has been that its public information and we have the
right to it.
Ory Okolloh (Kenya)

30 TAI New Technology / Open data study

Donors: a fourth tier?

The centrality of international aid donors in much of the
developing world was highlighted on a number of occasions
during this round of interviews, leading the researcher to
conclude that these actors might represent a fourth tier in
the drive towards opening up government data.

[Government] will grudgingly accept [that] donors

[who] give a lot of money [can] raise questions, but
theyll be less open to Tanzanians, particularly if its
done in an open space.
Rakesh Rajani (Tanzania)

I think in many cases NGOs hold just as much data

as the Government. And so Ive always argued
that similar efforts [should be directed at NGOs],
especially in Africa where NGOs are so involved in
health or water, or disease. Youve seen the World
Bank starting to release its datasets now, but theyre
sometimes just as bad as Government in terms of
holding onto information.
Ory Okolloh (Kenya)

Ive argued for a long time that opening up data

is really the way to go, both with Government and
especially in Africa with NGOs as well or multilateral
institutions like the World Bank. Youll probably find
the World Bank has better stats on health than our
Government does. And so making sure that theyre
not left out of the conversation
[is important] as well.
Ory Okolloh (Kenya)

One interviewee pointed out that getting donors to open

their books was only one half of the problem it would be
necessary to identify where money had eventually ended
up in order to create a useful system. Rather than look to
national governments to provide this data, crowd-sourcing
it from individual beneficiaries on the ground might be just
as, if not more effective. Nonetheless, issues persisted:

You cant just look at how the donor hands

the money out, you have to have some sort of
confirmation from the other side as far as where
the money ended up going. Because the real
problem in many of these cases is you hand out a
large sum of money and then its possible that big
percentages of it disappear before it hits any of the
desired recipients.
Ethan Zuckerman

So, one way you could do this is you could have

the donor essentially disclose what their funding
matrix looks like and then who their beneficiaries
are and then, ideally, you go to the beneficiaries
and say, Well look, how much of this did you get?
and then try to figure out how you resolve this back
and forth. Then, what youre doing is building a very
complex map of who the donors are funding and
questions about how that money is actually getting
to recipients. To the extent that that money is going
into governments as far as governmental support of
the situation, like Tanzania, Im curious [whether] you
really would be able to get the budgetary data that
youre looking for. I think what youre going to find
is that a lot of that money is fairly general support
you know, it is money that goes to support and
strengthen a particular agency.
Ethan Zuckerman

On several occasions, interviewees observed that donors

could tie their aid to the opening up of government data.
Good arguments were made about how what looked like
technocratic (and therefore fairly unthreatening) reforms
on the surface, could lead to transformative outcomes:

The technocratic [reforms] are the only ones that the

donors - at least the multilaterals and the bilaterals
can usually engage on. I mean, the World Bank
being a very good example: by mandate, it cant
touch politics or it cant touch the justice sector, for
instance. Those are huge areas that are in desperate
need, but they literally cannot fund anything. So,
instead, they can do budget reform, they can do
public financial management, they can do capacity
building of the various civil servants, all that kind
of stuff, which... starts to smuggle in a culture of
accountability and transparency.
Nathaniel Heller

TAI New Technology / Open data study 31

Depending on a particular countys situation, donors could

take one of two approaches:

What we have found is that, in a large number of

countries, the lack of public information is due to
two reasons. One is because the governments
dont produce the information even for internal
purposes. Now this is where there is lack of technical
capacity or theres a lack of an institutional system
that requires government to produce that particular
kind of information on a routine basis. The role
of donors there is to be able to... support public
finance management systems, the development
of infrastructure, providing technical assistance to
governments in order to develop their capacities
to start producing this information. So thats one
set of countries. But in another set of countries, an
almost equal number of countries, we find that the
governments dont publish information, but this
information is being produced for internal purposes
and for their donors. So donors do have access to
these reports, budget reports. In those countries the
lack of public availability of these documents is due
to political will, its a political decision taken that we
are not going to be transparent or were not going
to publish this document or weve not done it in
the past and its not a priority for us at this moment.
There donors have a different role, because we know
they have access to these documents or could very
easily get access to these documents if they wanted
to. There we are encouraging donors to actually
use their own influence over government to start
encouraging them.
Vivek Ramkumar

Some countries, however, might resist such approaches

on account of perceived hypocrisy:

The hypocrisy issue is not trivial even if its blown

out of proportion, I think ... the perception of it is very
real in low-income countries. So for the World Bank
to walk in and talk about the broader transparency
agenda rings hollow, I think, for a number of
governments and they say, We dont even know
whats going on at the executive board level
Nathaniel Heller

And it was suggested that the strategic value of some aid

recipients as allies (around, for example, security issues),
might mean some donors hesitate to enforce transparency
conditions attached to aid:

The chances that the US is going to put too

much pressure on some of its key allies in this
environment is highly unlikely. Theyre going to
continue to give aid to certain countries irrespective
of whether those countries are transparent and
meeting certain benchmarks and thats the way
with other countries too.
Vivek Ramkumar

Having donors tie aid could also lead to unintended

consequences, either through poor execution of open
data projects (which could be expected in the absence of
a strong local movement able to guide and monitor open
data initiatives made in response to donors) or through
countries rejecting aid tied to opening up data, in favour
of aid which is tied in a less burdensome fashion:

Yes, donor pressure has worked, but it has also

backfired, I think its something you have to
be careful with. Probably the leading example
of a backfire is Pakistan. Basically the Asian
Development Bank imposed adoption of an FOI
law as a loan conditionality, and Pakistan adopted
an FOI law. But the bank didnt know what it was
doing and was unable to assess the quality of that
law, so it accepted it at face value. It was actually
an ordinance passed by a military government, so
it had no credibility among the people anyway, but
also it was not a progressive rule and its been
I mean, not completely its being very ineffective
and, to some extent, I think it is I mean, this is
arguable, but I think, to some extent, it has blocked
further progress.
Toby Mendel

I can imagine some governments or some

departments refusing that as a condition and looking
for essentially less demanding aid, which is to say
that if the US and the UK come in and say, This
is how were tying our aid and the Japanese say,
Well, weve got a very different tie were actually
not going to touch transparency, were going to tie it
to the use of Japanese contractors. That would be
my opinion.
Ethan Zuckerman

I think we are still in such early days that its hard to

tell just how systematic or systemic the shift may be
and, frankly, what the harm is or isnt. I think there are
some negative externalities that would come out of
that shift, but I think we may not know for a decade
or two just how serious they are or arent.
Nathaniel Heller

32 TAI New Technology / Open data study

Issues for developing and middle

income country users

The importance of SMS and other mobile platforms in the

ecosystem of developing world communications was also
remarked upon:

In terms of the widespread existence of end-users for

applications derived by civil society or other actors from
Government data, the researcher received mixed reports:

There are only 200 postings on Vote Report India

of some kind of noncompliance or some kind of
irregularity with particular voting booths. Thats a
very small number if you look at India as a country
with one billion people.

There are only 20million broadband connections in

India and only 71million people claim to have used
the internet last year. So I think using the internet
to exercise your citizenship, thats quite a mature
usage and in India the internet is still very, very
recent, at least the PC and desktop version of the
internet. Maybe the next generation of open data
e-governance tools will have mobile interfaces.

Sunil Abraham (India)

Sunil Abraham (India)

Things are changing. In the last municipal elections

that I went to vote in, outside where we had the
polling booths there were young people with
laptops and they were holding onto spreadsheets
of the voter lists and they were verifying or helping
voters find their election ID before they went into the
polling booths. So things are slowly changing, but as
far as the internet goes we are still a very long way
off from the West.

The researcher was interested in how the cost structure

of SMS, as opposed to internet communication, would
impact on user takeup of open data applications designed
for mobile. One interviewee had already been working in
partnership with mobile providers to reduce costs at the
supply end:

Sunil Abraham (India)

A lot of us in this country were completely

disempowered due to an apartheid state where you
didnt question, you werent entitled to ask, you were
completely disenfranchised. And now weve gone
towards a country which has these advanced laws but
a lot of people remain illiterate, impoverished and
disempowered to ask those questions that theyre
entitled to ask and obtain explanations
from Government.
Jay Kruuse (South Africa)

The importance of institutional users of derived

government data was highlighted:

Weve got a number of organisations that we work

with who are beginning to realise the benefit of
evidence-based monitoring and evidence-based
advocacy, where you engage with assurances and
promises or plans by Government to improve their
services to communities.
Jay Kruuse (South Africa)

Its certainly a long road, speaking from my

experience with Mzalendo.... Mzalendos site will
never reach more than a couple of hundred hits
a day unless we do a huge media campaign. But
then [we can produce] information in a format that
if youre an advocacy group you can come to our
website, print out really easily and use to do your
advocacy work.
Ory Okolloh (Kenya)

The commitment we have is really at a pilot level

now. What were trying to do with the mobile phone
companies is to say to them Look, if you allow us
to send SMSs to citizens at your off-peak hours
... so we would send them just to subscribers
within their own network so they dont have to pay
interconnection charges, in other words it doesnt
cost them anything additional
So if you let us do that you do two things. One
is that you reach citizens and you provide a service
at virtually minimum cost to yourself, so youre just
using your unused capacity. But secondly when
citizens either forward those SMSs to others or
respond with monitoring information, in fact what
theyre doing is they are paying normal rates,
normal SMS rates and in this way you are generating
revenue for yourself.
Rakesh Rajani (Tanzania)

Another interviewee, when challenged on cost, pointed out

that engagement through SMS still represented value for
money, when compared with other advocacy interventions:

If we wanted to send out SMSs also, theres

potentially a cost implication, as you say, with that
as well. But I always argue, I say Compared to a lot
of the other transparency projects I see [its cheap,
if you] really look at how much has been spent on
advocacy around the Freedom of Information Bill.
[And it] has not really gone anywhere.
Ory Okolloh (Kenya)

The pace of change on the network was also highlighted:

In a matter of years, in a matter of three/four/five

years we will get to a situation whereby we are
not dealing with SMS anymore, we are dealing
with internet via mobile phone in East Africa
which I think will open up the possibility for
citizen engagement remarkably.
Rakesh Rajani (Tanzania)

TAI New Technology /Open data study 33

Complementary strategies
for open data
During the second round of interviews, a number of
complementary strategies were suggested for pushing
open data initiatives in developing and middle income
countries. As observed in the UK and US contexts, Obama
and Brown, in their endorsement of data.gov and data.
gov.uk respectively, were operating in distinct political
moments. Toby Mendel drew on his experience of
promoting FOI laws around the world to describe a number
of other political moments worth looking out for.
The first (which is arguably similar to Obamas moment)
was the context of a fresh administration brought in on
a popular mandate to replace a corrupt or otherwise
politically dicey old regime:

Mexico when the Fox administration came

in very much embraced transparency as a very
important value and set up a fantastic system: great
law and a great oversight institution. But one of the
reasons they were interested in that was because
all of the skeletons in the closet, as it were, didnt
belong to them.
Toby Mendel

In 1974 there were very, very progressive

amendments to the US FOIA, falling on the
heels of the impeachment of Nixon
Toby Mendel

The second was against a background of the fear or reality

of corruption, in countries focussed on rapid economic
development:

China has a... sensible worry that their economic

growth will be stunted by massive corruption
and they see openness as an antidote to that.
In Indonesia, too - the anti-corruption angle in
Indonesia had a stronger practical precedent
than in China because corruption was seen
as having brought about the economic collapse
that eventually led to Suhartos downfall.
Toby Mendel

Regional peer pressure could also play a role:

Of course, there is the whole process of developing

international standards and that includes regional
pressure. So the Philippines is about to adopt a law
and Indonesia has just adopted one, and Thailand
has had one for a while and... once the Philippines
does, then Malaysia will start to say, Well, you know
were falling behind in the Asian region so theres
a larger political and peer pressure.
Toby Mendel

Finally, sub-national laws were also identified

as important drivers:

Malaysia is a country which is very, very secretive.

It has very, very draconian secrecy rules. It has
very repressive media rules, its a country where
information is highly controlled, but its not a tinpot
dictatorship like Burma or North Korea or even
Vietnam. And there, civil society groups working
with opposition politicians have taken advantage of
the loss to the governing party of, I think, five states
in the last election, and they worked at the state
levels, so the sub-national level. And actually thats
been a pattern across the larger Asian countries. So
if you look at India, Japan, China, Indonesia - all of
those countries had sub-national laws first and then
the countries started to develop some sub-national
experience. Because a lot of this is about fear and
the fear of the unknown... and I think a practical
experience at the city level, at a provincial level,
or a state level can really help with that.
Toby Mendel

In the context of forcing administrations to meet their

proactive publishing requirements where an FOI regime
is nominally in place, Jay Kruuse whose Public Service
Accountability Monitor organisation has in the past brought
proceedings against government departments for failing to
meet publishing requirements - identified the judiciary as
an important ally:

The most positive signals that we see are...

pronouncements by our judiciary on these issues.
Weve had some really landmark decisions in recent
times which have vindicated these issues and pointed
to the fact that democracy presupposes participation
and dissemination of information so that people can
make informed choices.
Jay Kruuse (South Africa)

Prizes were also highlighted as a means to encourage best

practice in the middle layer of government:

The Human Rights Commission who have

a mandate in terms of the current access to
information legislation in South Africa, they look to
hold a Golden Key Award every year to recognise
information officers who have excelled and who
provide guidance to other information officers.
That has been one way of improving access.

34 TAI New Technology /Open data study

Jay Kruuse (South Africa)

An aside: advice for funders

One thing I suggested to people who are trying to

work on this within Kenya is to have an award for the
most open Ministry or something, get them prizes
for the best Information Department, also incentivise
so theres something in it for them rather than them
feeling that theyre just being pressured to change.

During the second round of interviews, advice was offered

directly to the funders who commissioned this report:

Ory Okolloh (Kenya)

Finally, at least two interviewees recognised the importance

of individual political actors in their domain spheres, and
extrapolated to suggest strategies targeted at identifying
and engaging such transformative individuals:

the other possibility is that you might start looking

for ministers who embrace this. So, for instance, the
former Nigerian finance minister, who was a fierce
anti-corruption advocate... when she was in power
she would have been an excellent minister to try this
with. Its probably the sort of thing that if were going
to do it, we would want to make bets on a small
number of ministers who want it and who get it.
Ethan Zuckerman

There are lots of pockets of where the Government is

putting that information out there. It just needs to be
pushed to the next step or to find a champion who
gets it. The new Speaker of the Kenyan Parliament is
much better than the previous one and so hes been
okay with allowing TVs to broadcast... So again, its
not really an issue of secrecy or the Civil Service [not
wanting to] change. It was just someone who got it.
But now the next step becomes then if you find that
person, educating them and saying Oh okay, its not
just about putting it out there but you can do X, Y
and Z and this is why its important.
Ory Okolloh (Kenya)

If [funders] want to make headway on these issues,

they need to do two things. One is they need to be
very clear about the overall end goal, a big-picture,
strategic goal. They then have to be open to dealing
with people who can make things happen and those
people, they often are not your usual development
sets. They might be quite different characters, often
in the private sector. You might find that the best
way to move this is to hire some really expensive
private sector guy whos going to charge you
$50,000 for 20days of work or 25days of work, but
that $50,000 may end up getting you much farther
than some long-term project where you sink in half
a million dollars.
Rakesh Rajani (Tanzania)

Another interviewee backed up this assertion that civic

hackers might look fundamentally different to the sorts
of civil society actors with whom funders are used to
engaging:

[Theyre] possibly technocrats actually... Maybe [civic

hackers] have a slightly cybernetic view of systems:
they can work well, and[for civic hackers] value and
good is about removing the obstacles to the smooth
functioning of your cybernetic system. Whatever.
I dont know. But its very particular, its great, its
inspiring, its internationally inspiring and also real.
But thats not the same as social impact.
Dan McQuillan (Central and Eastern Europe)

Common donor practices, such as avoiding project

replication, were also highlighted as an issue:

But theres a systemic problem which is all donor

agencies, including the ones youre preparing the
report for, incentivise uniqueness and incentivise
unique propositions. They dont incentivise
replication of good ideas. They dont incentivise
joining the bandwagon, right?
Sunil Abraham (India)

Further, its worth noting that the transition towards data.

gov and data.gov.uk took place across a longer timescale
at least four or five years than the traditional two year
funding cycle followed by many of the major grant-giving
organisations. It is the researchers view that funders should
consider the long historical shadow cast by these positive
final outcomes when setting goals for similar interventions
elsewhere in the world.

TAI New Technology /Open data study 35

Avenues for further research

This report has thrown up two valuable avenues for further
research. The first is around the social impact of existing
data catalogues such as data.gov and data.gov.uk. Much
hope has been invested in the transformative potential
of these catalogues to open issues in public life to a wide
audience thanks to the interactivity and ubiquity of the
World Wide Web in the US and UK. In order to build on their
call for open data to be pursued across the world, funders
have an interest in tracking over time to what extent open
data catalogues deliver on this promise.
The second, and perhaps more pressing, research need is
for a systematic evaluation of government data collection
activities across all regions being considered for data.govtype interventions. Such an evaluation should capture not
just what data is being collected, but also:

During the interview process, the researcher encountered

several international monitoring NGOs with existing
networks of in-country researchers which could be
leveraged to complete such a study, perhaps in conjunction
with an organisation such as mySociety or the OKF, or any
other organisation that has special expertise in open
data contingencies.
Finally, its worth mentioning that several research projects
which seek to provide technical guidance to those wishing
to build open data portals are ongoing, including a nascent
project by the OKF to produce an open data manual,52 and
a project to map open government data initiatives around
the world that is being conducted jointly by access-info.org
and the OKF.53

how that data is being maintained (ie digitally

or on paper);
to what extent data collection practices are
standardised across a country, or whether
they are district- or region-specific;
what formats the data is stored in, and whether
those formats are open or proprietary; and (perhaps
most importantly)
what licensing practices are associated with
redistribution or reuse of that data.

52 See http://lists.okfn.org/pipermail/okfn-discuss/2010-

April/007244.html. The manual will be based on the OKFs

pamphlet Unlocking the Potential of Aid Information (Gray,
Hatcher, Hogge, Parrish and Pollock, 2009).

53 See http://opengovernmentdata.org/

36 TAI New Technology /Open data study

An open data
strategy checklist

TAI New Technology /Open data study 37

Based on the evidence uncovered in this report, the following checklist

is proposed for those considering initiating a campaign for data.gov
and data.gov.uk-type projects in other countries.

Status of FOI

Potential end-users

Does the country have a Freedom of Information

law? Has it been implemented? Is it delivering on
transparency? Is it delivering on social outcomes?

How free is the press? How wired?

Does the Freedom of Information regime include

proactive disclosure/ publishing requirements as well as
reactive disclosure, or is it based on request-only access?

Are there specific examples of those groups using

data in their advocacy/monitoring or other civic
engagement activities?

How is the administration coping with the proactive

disclosure and reactive disclosure elements of the
Freedom of Information regime that is in place?

Are there specific examples of take-up of data by

end users that may inform open data initiatives?

Has the judiciary been active in upholding Freedom

of Information obligations?
If the country does not have a Freedom of Information
law, is there an active movement advocating for one?
If the country does not have a Freedom of Information
law, are there sectoral freedom of information laws,
transparency provisions within the countrys constitution,
or applicable international standards (e.g. Aarhus
convention) that are relevant and implemented?

Current data collection activities

What level of data collection is undertaken
by the government?
o Is this data collected in a systematic
and timely fashion?
o Is this data stored digitally or on paper?
o In what format is digitised data collected and stored?
Are these open or proprietary formats? Are they
machine-readable?
o What is the state of e-government activities in the
country? Does the government have an e-government
strategy? Is it implemented?
A
re government data or compilation of government data
currently protected by copyright or another intellectual
property-like regime?
I s the data subject to any licences that restrict
reuse? Are fees charged for access (for instance,
to aid cost recovery)?
T o what extent are conversations around the digitisation
of government subject to vendor capture?
What sorts of privacy laws are operational in the
country and how might they impact on open access
to government data?

I s there a user base of traditional civil society groups

that may make use of targeted data?

I n what ways did that data need to be made

accessible in order for it to be used?
W
hat level of internet penetration is there across
the country?
W
hat level of mobile penetration is there and how are
people accessing mobile data services (SMS, 3G etc)?

Tier 1: Civil society

H
ow technically literate is civil society?
Are civic hackers present?
A
re there instances in-country where local civil
society is appropriating government data already?
Is there an organised, technology-led local group, such
as the Sunlight Foundation, or MySociety, in the country?

Tier 2: The middle layer

How empowered is the middle layer by the current
political environment, in which departments and
at what level?
I s openness a high-stakes issue at the middle layer?
How threatened might individuals advocating for
openness within this layer feel? To what degree and
how might middle level civil servants resist opening
government data that was not collected with the
intention of being released?
How technically competent is the middle layer?
H
ow much capacity is there within the middle layer?
How far has e-government penetrated?
D
oes an existing network / community exist that bridges
the gap between the middle layer and civil society?

38 TAI New Technology /Open data study

Tier 3: The top-level

Tier 4: Donor drivers

I s the country at a particular stage in the political

cycle that would make opening government data
advantageous in a political sense?

What levels of multilateral and bilateral donor

involvement are there in the countrys budget and
governance?

Politically, who might the winners and losers be

if government data is opened?

Are donors already releasing their own data openly?

To what degree and how is the expense of opening

significant amounts of government data likely to
be an issue?
R
egionally, could the country be described as lagging
behind neighbouring countries in terms of openness
and transparency?
C
ould regional peer pressure have an impact on top level
political will to open data (e.g. ASEAN, SADC)?
What are the countrys economic aspirations? Could
an argument for efficiency be made?
Are there exemplars of open data practice at the
sub-national level?
H
ow is the top political tier likely to react to advocacy
from any one of the following actors:
o Tim Berners-Lee?
o Barack Obama?
o Commercial operators (eg Google)?
o Entertainment industry: film stars?
How does the country rank in corruption assessments?
Do allegations of serious corruption reach the highest
levels of government, or are they restricted to middle
and lower levels of government?
I s there an anti-corruption movement in the country,
and could it be an effective ally?

H
ow thoroughly does the administration report on
aid spending?
H
ow has the country reacted to previous tied aid?
Is there scope for positive conditionality?
A
re there private donors (local or international)
active in the country who could be useful allies?

TAI New Technology /Open data study 39

Conclusion

40 TAI New Technology /Open data study

This report has been produced on the premise that there are substantial social and economic gains
to be made from opening government data to the public, and that the combination of geographic,
budget, demographic, services, education and other data, publicly available in an open format on
the web promises to improve services as well as create future economic growth. Several interviewees
approached during the research that went into this report have sought to challenge that premise,
and those challenges have been noted. It is beyond the scope of this report to investigate those
challenges in any further detail. However, the researcher suggests that those who wish to take this
research further would do well to monitor the wider impact of data.gov and data.gov.uk, if only to be
able to respond to such challenges in the future.
This research has sought to explore the feasibility of
applying the approach to open data taken by the US and UK
administrations (with the open data catalogues www.data.
gov and www.data.gov.uk respectively) to relevant middle
income and developing countries. It has sought to identify
the strategies used in the US/UK contexts, and it has proposed
a set of criteria to guide the selection of pilot countries,
criteria which in turn suggest a template strategy to open
government data in a middle income or developing country.
The resulting checklist appears to give equal weight to
each factor, and indeed one of the most striking aspects
of the success of the two open data projects studied is
how, in each case, that success was brought about by a
broad range of concurrent factors and events. Nonetheless,
the researcher would like to suggest that one aspect of
the open data strategy deserves special highlighting: the
existence, in both the US and the UK, of highly established
data collection activities operated by a well-resourced,
broadly independent and highly skilled middle layer
of government administrators. This is one aspect of the
trajectory towards data.gov and data.gov.uk that has
perhaps attracted the least attention of onlookers, for
two reasons: the activities of this section of society do
not generally attract attention (at least not when they are
functioning well), nor do its members seek it; and secondly,
the activities of another group that has contributed to the
open data initiatives the so-called civic hackers are fresh
and exciting, and therefore likely to be more attentiongrabbing. All this is worth highlighting since, as several
regional and domain experts have made clear during
interviews, these middle layer activities both the data
collection and the political position the administrators who
undertake it exist within are potentially weak or absent in
most middle income and developing countries.
This report has focussed in some detail on the difficulty
experienced in the UK of opening up geospatial data
that was at the time commercially licensed in order to aid
cost recovery. Such commercial / cost recovery activities
have also been highlighted in the checklist. However, the
relationship of commercial licensing with the eventual
success of a data catalogue like data.gov.uk strikes the
researcher as complex. It has been suggested in the course
of this report that the barrier these activities imposed
in the UK may have served as a common call to action
among both civil society and the middle-layer government
administrators, which in turn served to strengthen the
crucial communication between these two groups in the
trajectory towards data.gov.uk, and ultimately enrich the
final proposition when compared with data.gov.
Of course, without the intervention of Tim Berners-Lee, it
is unlikely whether the UK government would have ever
sanctioned the opening up of geospatial data. But there is
a risk that funders, seeing the impact of this intervention,

might conclude that such an intervention is all that is

needed to establish similarly vibrant open data initiatives
in other countries. This would be a mistake. It is this
researchers opinion that without existing and meaningful
activity around open data on all three tiers of public life
identified in this report (with a potential fourth tier, that of
international donors, also coming into play in developing
and middle income countries) funders could be at risk
of seeding a kind of cargo cult. The researcher agrees
with Ethan Zuckerman, who, when asked to qualify his
scepticism with a vision of how things might eventually play
out around open data in Africa, made the following remark:

I think African governments end up opening their

data through a combination of citizen pressure in
countries where that matters and donor pressure
in countries where that matters. So, for instance...
in Ghana, you actually have an extremely engaged
citizenry, and a sceptical and technically competent
press starting to demand certain critical pieces of
information. What that might require in Ghana is
focusing on the journalists and essentially saying,
Why dont we have this data and this data and this
data?, What could we do, what questions could we
ask and answer if we have it?, Why cant we get it?
and then suing the government to get it if they dont
have it. And I could imagine that leading gradually
toward the release of data in that country. There are
other countries where the government is a lot less
[responsive] to popular pressure. Im going to go out
on a limb and say that Rwanda, which is functioning
as a one-party dictatorship, is one of those countries.
But donors are very powerful and Kagame is
extremely concerned with being perceived as a
modern, forward-looking benevolent dictator. And,
I think in that case, using the pressure of the aid
industry which represents a remarkable part of the
government budget, and essentially saying, For
our sake, for your sake, for your citizens sake, this
is going to become mandatory, and were going to
release that information and its going to be available
not just within the international community, but to
the world at large and particularly the journalists. I
can imagine that working if only because weve got
so much leverage in that situation.
Ethan Zuckerman

TAI New Technology /Open data study 41

Annexes

42 TAI New Technology /Open data study

Annex I: Methodology

This research was conducted over ten days in April 2010.

It drew on published research studies, commissioned
reports, evaluations and other grey literature (listed in the
bibliography) as well as telephone interviews.
The researcher was guided to a core group of interviewees
who acted as primary sources of information (listed
below). An additional interviewee Jonathan Gray was
added at the researchers discretion, on the basis of
concurrent work he was conducting on the topic (which
had been flagged by the researchs commissioners), and
on his status as a coordinator of an international network
of open data enthusiasts.
Interviews were conducted in two rounds. The first round
focussed on individuals who had been directly or indirectly
involved with the development of data.gov and data.gov.
uk. The researcher then used information garnered from this
process, and from the literature review, to shape questions
for the second round of interviews. The second round of
interviews focussed on individuals with wider domain and
region expertise.

TAI New Technology /Open data study 43

Annex II: List of interviewees

Round one interviewees

Round two interviewees

Round two interviewees

Tim Berners-Lee
www.w3.org/People/Berners-Lee/

Domain experts

Regional experts

Steve Bratt, CEO

World Wide Web Foundation
www.w3.org/People/Bratt/

Sunil Abraham, Executive Director,

Centre for Internet and Society (India)
www.cis-india.org/publications/cis/sunil

Jonathan Gray, Community Coordinator,

Open Knowledge Foundation Network
www.okfn.org/

Nathaniel Heller
Managing Director Global Integrity
www.globalintegrity.org/aboutus/team.
cfm#nheller

Jay Kruuse, Head of Monitoring

and Research Programme
Centre for Social Accountability
South Africa
www.psam.org.za

T
om Steinberg, Director, mySociety
www.mysociety.org/
John Wonderlich, Policy Director,
Sunlight Foundation
sunlightfoundation.com/

T
oby Mendel, Executive Director
Center for Law and Democracy
V
ivek Ramkumar
International Budget Partnership
www.internationalbudget.org/
Ethan Zuckerman, Senior Researcher
Berkman Center for Internet and Society
http://ethanzuckerman.com/

Dan McQuillan
Social Innovation Camps (CEE)
www.sicamp.org/
O
ry Okolloh, co-founder Mzalendo
co-founder, Ushahidi (Kenya)
www.kenyanpundit.com/
Rakesh Rajani, Founder
Twaweza (Tanzania)
www.twaweza.org/

44 TAI New Technology /Open data study

Bibliography

Allan, R. (2009, November). The Power of Government

Information State of the eUnion: Government 2.0
and onwards Retrieved April 16, 2010 from
http://21gov.net/wp-content/uploads/e-book.pdf
BBC News. (2010, April 1). Ordnance Survey offers
free data access. Retrieved April 16, 2010, from
http://news.bbc.co.uk/1/hi/8597779.stm
Berners-Lee, T. (2009, February). Tim Berners-Lee on the next
Web. (T. Berners-Lee, Performer) TED Conference 2009.
Brown, G. (2009, June 10). Oral Answers to Questions
Prime Minister. Retrieved April 16, 2010, from House
of Commons debates:
www.theyworkforyou.com/debate/?id=2009-06-10c.795.1
Cabinet Office. (2009, June 10).
Pioneer of the World Wide Web to advise the government on
using data. Retrieved April 16, 2010, from Cabinet Office:
www.cabinetoffice.gov.uk/newsroom/news_
releases/2009/090610_web.aspx
Cabinet Office. (2007). The Governments Response
to The Power of Information Review. Retrieved from
www.official-documents.gov.uk/document/
cm71/7157/7157.pdf
Civic Impulse LLC 1. (n.d.). About This Website.
Retrieved April 16, 2010, from GovTrack.us:
www.govtrack.us/about.xpd
Civic Impulse LLC 2. (n.d.). Sites that Use GovTrack Data.
Retrieved April 16, 2010, from GovTrack.us:
www.govtrack.us/downstream.xpd
Crabtree, J. (2010, January 27). Mash the State. Prospect .
Escher, T. (2009, March 5). Citizens contacting
representatives. Presentation to the Oxford Internet
Institute Internet and Democracy Workshop.
Executive Office of the President, Office of Management
and Budget. (2009, December 8). Open Government
Directive. Retrieved April 16, 2010, from
www.whitehouse.gov/omb/assets/memoranda
_2010/m10-06.pdf
Gray, J., Hatcher, J., Hogge, B., Parrish, S., & Pollock, R.
(2009, December). Unlocking the Potential of Aid Information
Version 0.2. Retrieved April 23, 2010, from
www.unlockingaid.info/wp-content/uploads/2010/02/
UnlockingAidInformation.pdf
International Budget Partnership. (n.d.).
Open Budget Survey 2008. Retrieved from
www.openbudgetindex.org/index.cfm?fa=fullReport
Kirkpatrick, M. (2010, January 20).
UK Launches Open Data Site; Puts Data.gov to Shame.
Retrieved April 16, 2010, from ReadWriteWeb:
www.readwriteweb.com/archives/uk_launches_open_
data_site_puts_datagov_to_shame.php

Mayo, E., & Steinberg, T. (2007, June). The Power of Information:

An independent review. Retrieved from http://195.99.1.70/
advice/poi/power-of-information-review.pdf
mySociety. (n.d.). About Us.
Retrieved April 16, 2010, from TheyWorkForYou.com:
www.theyworkforyou.com/about/#1
Newbury, D., Bently, L., & Pollock, R.
(2008, February 26). Models of Public Sector Information
Provision via Trading Funds. Retrieved from
www.berr.gov.uk/files/file45136.pdf
Office of E-government and IT; Office of Management
and Budget. (2009, December 3). Data.gov Concept of
Operations: Draft. Retrieved April 13, 2010, from
www.ideascale.com/userimages/sub-1/736312/
ConOpsFinal.pdf
Office of Fair Trading. (2006, December).
The commercial use of public information. Retrieved from
www.oft.gov.uk/shared_oft/reports/consumer_protection/
oft861.pdf
Power of Information Taskforce 1. (2009, February). Power
of Information Taskforce Report. Retrieved from http://poit.
cabinetoffice.gov.uk/poit/wp-content/uploads/2009/03/
poit-report-final-pdf.pdf
Power of Information Taskforce 2. (n.d.).
Show Us A Better Way. Retrieved April 16, 2010, from
www.showusabetterway.co.uk/
Power of Information Taskforce 3. (n.d.).
Taskforce Membership. Retrieved April 16, 2010,
from Power of Information Taskforce:
http://powerofinformation.wordpress.com/
taskforce-membership/
Robinson, D., Harlan, Y., Zeller, W., & Felten, E. (2009).
Government Data and the Invisible Hand. Yale Journal
of Law and Technology , 11, 160.
Tactical Technology Collective. (2009).
Case Study: mySociety. Retrieved April 26, 2010,
from Ten Tactics for turning information into action:
www.informationactivism.org/viewtactic1
Torkington, N (2010, February 2). Rethinking Open Data.
Retrieved April 16 2010 from OReilly Radar:
http://radar.oreilly.com/2010/02/rethinking-open-data.html

TAI New Technology /Open data study 45

Acknowledgements

About the author

The researcher would like to thank the Transparency

and Accountability Initiative (a donor collaborative)
for providing her with the opportunity to conduct this
research, and in particular Martin Tisne, Programme
Manager at the Transparency and Accountability Initiative,
as well as Janet Haven and Darius Cuplinskas of the Open
Society Institutes Information Programme, for their advice
and guidance during the research. Thanks too to Steve
Bratt of the Worldwide Web Foundation and Stephen King
of the Omidyar Network. Finally, the researcher would like
to extend her heartfelt thanks to all of those who agreed to
be interviewed for this project, for giving up their valuable
time to share their expertise and insights with others.

Becky Hogge is a freelance writer and broadcaster.

Her writing on information politics, human rights and
technology appears regularly in the UK press, and she has
produced research papers on internet governance, online
campaigning and online markets for various international
NGOs and grant-giving organisations. She has served on
several national awards panels, and has been invited by
the UK and other European governments to consult on
information policy. Her opinions on internet governance
and internet censorship have been broadcast around
the world. Becky is the former Executive Director of the
UKs Open Rights Group, and she sits on the board of the
Open Knowledge Foundation. Her first book, Barefoot into
Cyberspace, will be published in summer 2011.

This work is licensed under a Creative Commons

Attribution 3.0 Licence:
http://creativecommons.org/licenses/by/3.0/us/