1

Freight Behavior Research:

Rationale, Principles, and Basic Results

José Holguín-Veras,
William H. Hart Professor,
Director of the Center for Infrastructure,
Transportation, and the Environment
Rensselaer Polytechnic Institute
[email protected]
Outline
2

Freight behavior research

Lessons learned about when to intervene, pilot test,
and use freight behavior research
The JHV principle: In complex systems, obvious
solutions are (almost) always wrong…
Corollary: Congestion problems are not new…
Julius Caesar’s Rome: Lex Iulia Municipalis BC 45

QUAE VIAE IN URBAN ROMAN SUNT ERUNT INTRA EA LOCA, UNI CONTINETI
HABITABITUR, NE QUIS IN IEIS VIEIS POST K. INANUR. PIRMAS PLOSTRUM
INTERDIU POST SOLEM ORTUM, NEVE ANTE HORAM X DIEI DUCITO AGITO,
NISI QUOD AUDIUM SACRARUM DEORUM INMORTALIUM CAUSSA
AEDIFICANDARUM, OPERISVE PUBLICE FACIUMDEI CAUSA, ADVEHEI
PORTARI OPORTEBIT, AUT QUOD EX URBE EX VE IEIS LOCIS EARUM
RERUM, QUAE PUBLICE DEMOLIENDAE LOCATAE ERUNT, PUBLICE EX
PORTAREI OPORTEBIT, ET QUARUM RERUM CAUSSA PLOSTRA H.L.
CERTEIS HOMINIBUS CERTEIS DE CAUSEIS AGERE DUCERE LICIBIT.

“On the roads which are in the city of Rome or will be within the area where will be
lived joined tightly, no one is allowed after next January 1st to drive or lead a carriage
during the day after sunrise and before the tenth hour of the day, except if something
will have to be supplied or transported for building temples of the immortal gods or for
the implementation of a work for the authorities, or as from the city or from those areas
something of those things of which the demolition will be put out to tender by the
authorities, will have to be removed on behalf of the authorities, and except for those
cases in which it will be according to this law permitted to certain persons for certain
reasons to drive or lead a carriage”.
New York City in the 1900s
5
Potential Solutions… 6
The reality in the 1940s…
7

37th Street and 7th Ave, New York City, 1945

The history of the world clearly shows that…
8

The public sector has not been able to completely

solve the congestion problem…
The private sector has not been able to completely
solve the congestion problem…
Communities have not been able to completely solve
the congestion problem…
Then, why do keep using the same approaches that
failed in the past?
Complex problems cannot be solved with simple solutions (if
they were, somebody would have solved them…)
We need comprehensive approaches to the urban freight
issues…
Collaboration is key to our approach…
9

No single player could solve all freight issues by itself

Public sector  Regulates, manages infrastructure
Private sector  Operates the system
Academia  Conducts research to find solutions
Communities  Enjoy freight benefits, suffer the impacts
All players control a different piece, no one benefits
from the status quo:
The Challenge
Global Drivers
11

Economic Globalization
Urbanization:
World’s population: 7+ billion people, 9 billion by 2045
In 2010, for the first time, 50% of world population is urban,
by 2050, 70% of the world population will be urban
In US/Canada/Europe, the future is here: +80% urban
Impacts of the Internet on Supply Chains:
Millions of citizens expect fast and inexpensive deliveries
The diminished importance of proximity to customers as a
competitive advantage, together with anti-freight attitudes
and policies, leads to logistical sprawl
 Increased Citizen Expectations
This is what we all want… 12
Change is needed… Who needs to change behavior?
Answer: Supply Chains…
14

The shippers The carriers The receivers

The Economy
 15

Macy’s is a large retailer in the US:

- It is located on 34th street in Manhattan
- 100 years-old building with 9 floors
- Takes almost an entire block in Herald Square
- 20 million visitors a year
- 1 billion dollars in annual sales
- 5 loading docks for freight and 3 loading bays for construction
- 15 million pieces of merchandise sold in Christmas
- Women /shoe department is larger than a football field
Macy’s in 34th Street
16
 17

A Cautionary Tale: The Efficiency Paradox

The Efficiency Paradox
William S. Jevons’ “The Coal
Question” (1865) concluded:
If demand is (long-term), lower
prices increase consumption

We need a holistic approach

to sustainability…
Technology is part of the
solution, not THE solution
http://en.wikipedia.org/wiki/File:PSM_V11_D660_William_Stanle
y_Jevons.jpg#file
Before “Solving the Problem” make sure you
have the facts right…

Which industry sector produces the bulk

of congestion attributed to freight?
 Establishment based freight trip generation …
20

NY-North NJ- Palm Bay- Fargo, ND-

NAICS Description Lebanon, PA
Long Island Melb...FL MN
44 Retail trade 39.06% 44.19% 34.85% 37.50%
42 Wholesale Trade 19.41% 11.04% 17.89% 13.57%
72 Accommodation / Food Services 15.72% 16.87% 13.97% 14.35%
23 Construction 11.47% 14.35% 16.14% 12.18%
31 Manufacturing 8.17% 8.80% 8.11% 15.35%
48 Transportation / Warehousing 6.16% 4.74% 9.03% 7.05%
Total freight trip generation (FTG) for FIS 1,024,477 25,682 15,515 10,285
Population 19,949,502 550,823 223,490 135,486
Number of establishmentsRetail/Accommodation/Food
(Total) 545,197 > 50% of 6,709
13,597 FTG 4,272
Number of establishments (FIS) 235,325 5,893 3,317 2,185
Employment (Total) 7,568,043 172,925 119,626 79,543
Employment (FIS) 3,061,899 84,821 63,186 47,164
Establishments (FIS)/1000 persons 11.796 10.699 14.842 16.127
Employment (FIS)/1000 persons 153.482 153.990 282.724 348.110
FTG/1000 employees (all sectors) 135.369 148.517 129.698 129.302
FTG/1000 employees (FIS) 334.589 302.783 245.549 218.071
FTG/1000 persons 51.354 46.625 69.423 75.913
Average FTG per establishment 4.353 4.358 4.677 4.707
Average employment per establishment 13.011 14.394 19.049 21.585
FTG vs. Establishment Size
21

50%
45% NY-Northern NJ-Long Island
40% Palm Bay-Melbourne-Titusville, FL
Percent of total FTG

35% Fargo, ND-MN

30% Lebanon, PA
25%
20%
15%
10%
5%
0%

Establishment employment
NYC Metro area
22

Freight trip Freight trip

Area
Population Establishments Employment attraction % FTA production % FTP
(miles2)
County (FTA) (FTP)
BRONX 42.15 1,332,650 7,754 91,787 19,900 4% 14,048 4%
BROOKLYN 70.88 2,465,326 23,262 232,199 58,114 13% 40,883 12%
NASSAU 287.96 1,334,544 24,142 314,287 62,828 14% 46,956 14%
MANHATTAN 23.09 1,537,195 40,415 692,260 113,069 26% 76,874 23%
PUTNAM 245.91 95,745 1,731 14,937 4,040 1% 3,298 1%
QUEENS 109.71 2,229,379 23,276 290,156 55,737 13% 46,390 14%
RICHMOND 58.74 443,728 4,268 49,668 10,136 2% 8,182 2%
ROCKLAND 192.39 286,753 4,547 60,963 11,600 3% 8,895 3%
SUFFOLK 926.81 1,419,369 26,787 357,405 69,234 16% 52,788 16%
WESTCHESTER 465.79 923,459 15,127 204,525 38,498 9% 30,477 9%
Grand Total 2,423.43 12,068,148.00 171,309.00 2,308,184.50 443,155.77 100% 328,790.82 100%

Based on the models estimated by National Cooperative Freight

Research Program Project #25 “Freight Trip Generation and Land
Use” available at: http://www.trb.org/Main/Blurbs/168396.aspx
Where do these deliveries go?
23

In Manhattan:
- 80 buildings and large traffic generators produce
4-8% of the total freight traffic…
- Restaurants and drinking places (10,000) produce
four times the freight traffic produced by the port…
Freight Behavior Research
For a policy to be effective:
25

Fact #1: Policy makers must have a good idea of how

the target will respond to the policy, lack of such
insight leads to:
(Negative) unintended consequences, ineffective policies
Lack of credibility and public trust
Fact #2: Policy makers have to:
Select the most appropriate initiative
Select the best policy lever (regulation, pricing, incentives)
Apply the policy stimuli on the best target (not obvious)
Select the strength of the stimuli correctly: if the stimuli is
too weak, nothing will be accomplished; if too strong, it may
lead to misallocation of resources and political problems
Behavior research provides the answers
Behavioral Influencers
Key aspects to consider…
27

Agent rationality
Markets conditions, pricing, and industry structure
Agent interactions
Behavioral patterns at the agent level
Example: The Off-Hour Delivery Project
 Rationality

Holguín-Veras, J. (2008). "Necessary Conditions for Off-Hour Deliveries and the

Effectiveness of Urban Freight Road Pricing and Alternative Financial Policies in
Competitive Markets." Transportation Research Part A: Policy and Practice 42(2): 392-413
For off-hour deliveries to be feasible:
30

Carrier and receivers must be better of because of

policies targeting carriers (pc) and receivers (pR)
Mathematically: Marginal Revenues > Costs

G j (p C )  C j (p C ) Carrier is better off

Receivers
Gi (p R )  C i (p R ) i   O
j are better off
Technical condition to
 O
i  O
min i   O
j ensure minimum duration
of deliveries
NOTE: Cost savings are negative,
cost increases are positive
For supply chains to change behavior…
31

All participants must benefit, or

The participant(s) with most power must force the
others to change…
The shippers The carriers The receivers

The Economy
Market Conditions: the PANYNJ TOD Initiative

Holguín-Veras, J., Q. Wang, et al. (2006). "Impacts of Time of Day Pricing on the
Behavior of Freight Carriers in a Congested Urban Area: Implications to Road Pricing."
Transportation Research Part A: Policy and Practice 40(9): 744-766.
We all know that…
33

If prices go up, transportation demand goes down

In freight road pricing:
Tolls are imposed on truck traffic
Carriers pass the toll to the receivers / shippers
Receivers / shippers will react by moving their operations to
the off peak hours
Right?
Not quite….. Reality is more complex than we think
The experience of the Port Authority of New York and
New Jersey’s Time of Day Pricing Initiative (2001)
provides great insight
Observed Behavioral Changes
34

Facility usage

Legend:
Increasing impact
on receivers

Productivity Increases
Changes in Facility
and Changes in
Usage and Cost
Facility Usage
Transfers

Productivity Increases
and Changes in Facility
Usage and Cost
Transfers

Productivity Cost transfer

Productivity Increases
increases
and Cost Transfers

Only the carrier is Carrier and receiver Only the receiver

impacted are impacted is impacted
Observed Behavioral Changes
35

Facility usage
(0%)
Legend:
Increasing impact
on receivers

Productivity Increases
and Changes in Changes in Facility
Facility Usage (4.8%) Usage and Cost
Transfers (27.6%)

Productivity Increases
and Changes in Facility
Usage and Cost
Transfers and (19.3%)

Productivity Cost transfer

Productivity Increases
increases (42.8%) (5.1%)
and Cost Transfers (0.4%)

Only the carrier is Carrier and receiver Only the receiver

impacted (42.8%) are impacted is impacted
(24.6%) (32.7%)
 Agent Interactions

Holguín-Veras, J., M. A. Silas, et al. (2007). "An Investigation on the Effectiveness of Joint
Receiver-Carrier Policies to Increase Truck Traffic in the Off-Peak Hours: Part I: The
Behaviors of Receivers." Networks and Spatial Economics 7(3): 277-295.
The decision about delivery time
37

Is made jointly between receivers and carriers

40% by receivers, 38% between receivers and carriers and
22% by carriers
Let’s take a look at the payoff matrix
The first sign represents the impact on carrier and the
second the impact on receiver (This is the solution
(These are non-feasible solutions) preferred by most receivers)
Receiver
Strategy Regular hours Off-hours
(I) (II)
Regular hours (-,+) (-,-)
Carrier (III) (IV)
Off-hours (-,-) (+,-)
The key is to use policy tools to make (This is the solution
OHD the best option to receivers preferred by most carriers)
 Behavioral Patterns: Staffed OHD

Holguín-Veras, J., M. A. Silas, et al. (2007). "An Investigation on the Effectiveness of Joint
Receiver-Carrier Policies to Increase Truck Traffic in the Off-Peak Hours: Part I: The
Behaviors of Receivers." Networks and Spatial Economics 7(3): 277-295.

Holguín-Veras, J., M. A. Silas, et al. (2008). "An Investigation on the Effectiveness of Joint
Receiver-Carrier Policies to Increase Truck Traffic in the Off-Peak Hours: Part II: The
Behaviors of Carriers." Networks and Spatial Economics 8(4): 327-354
Behavioral research conducted
39

Behavioral surveys conducted to assess the responses

of carriers and receivers to alternative policies
400 carriers and 400 receivers (Manhattan and Brooklyn)
Policies studied:
Carriers: Customer request from receivers, customer request
& toll savings, customer request & financial rewards,
customer request & parking, customer request & paid permit
Receivers
 Tax deduction (financial incentive), Shipping discounts

The stated preference data were used to estimate

behavioral choice models
Econometric results: Receivers
40

Variable Name Coefficient T-value

Utility of off-peak deliveries: C1CHOICE
A tax deduction in any employee assigned to OPD TDEDUCT 8.392E-05 1.410
Reasons for not receiving OPD
No access to building/freight entrance after hours REASON1 -1.234 -1.571
Interferes with normal business REASON2 -0.591 -1.208
All receivers are moderately sensitive
Additional costs to the business if accepting more OPD COST
to tax-0.888
deductions
-3.232
Policy interaction terms
Tax deduction for Wood/lumber TDCOM8 6.968E-04 2.219
Tax deduction for Alcohol TDCOM4 4.356E-04 2.209
Tax deduction for Paper TDCOM9 2.627E-04 2.988
Tax deduction for Medical supplies TDCOM22 2.598E-04 3.188
Tax deduction for Food TDCOM2 1.875E-04 3.973
Tax deduction for Printed Material TDCOM21 1.652E-04 1.802
Tax deduction for Metal TDCOM13 1.415E-04 1.410
Other interaction terms
Number of employees in a branch facility BRANEMP 9.867E-03 1.612
Utility of no off-peak deliveries:
These industry segments are more
Alternative specific constant
sensitive
CONSTANT
than the
1.599
rest
4.151
R2 0.172
Adjusted R2 0.140
Econometric results: Carriers
41

Variable Name Coefficient t-value

Utility of off-peak deliveries: C4CHOICE
Percentage of customers requesting OPD PCUST 0.017 2.912

(Variables deleted) ALL truckers are sensitive to

customers requesting OPD
Parking infractions in Manhattan per driver per month
Nothing FINE0 ONLY -1.083
some segments
-2.600 of the
From $1-$100 FINE100
industry-0.521 -1.665
are sensitive to tolls
Policy interaction terms
Toll savings for Petroleum/coal TOLCOM10 0.440 1.606
Toll savings for Wood/lumber TOLCOM8 0.340 1.912
Toll savings for Food TOLCOM2 0.209 2.733
Toll savings for Textiles/clothing TOLCOM6 0.217 2.022

(Variables deleted)
R2 0.194
Adjusted R2 0.146
Behavioral Modeling: Unassisted OHD

Holguin-Veras, J., R. Marquis, et al. (2013). Fostering the Use of Unassisted Off-Hour
Deliveries (forthcoming in Transportation Research Part A
Receivers’ Behavioral Model: Unassisted OHD
Model Model 2
Key influencers: Independent variables Parameter t-stat
- One time incentive (OTI) Constant 0.22 (1.00)
Number of deliveries -0.08 (-11.66)
- Trusted vendor Incentives
One time incentive in $1000 (OTI) 0.17 (6.76)
- Business Support Services Carrier discount in percent (CDR*100) 3.10 (7.12)
- Public Recognition Business Support (BS) 0.51 (3.52)
Public Recognition (PR) 0.38 (2.48)
Trusted Vendor (TV) 0.94 (4.29)
The One-time-incentive NAICS
Clothing stores, binary variable -2.46 (-4.32)
is most influential in Performing arts, binary variable -4.80 (-12.38)
Interaction terms: OTI and NAICS
these sectors OTI for food and beverage stores 0.20 (4.24)
OTI for apparel manufacture stores 0.11 (1.88)
OTI for clothing stores 0.25 (3.40)
OTI for nondurable wholesalers 0.37 (7.62)
Interaction terms: CDR and NAICS
CDR for personal laundry -2.08 (-3.25)
Interaction terms: TV and NAICS
TV for food and beverage stores 2.02 (3.17)
TV for performing arts 13.49 (11.16)
TV for clothing stores 2.24 (4.06)
TV for miscellaneous stores retailers 3.17 (5.86)
43
Unassisted OHD: Behavioral Research
44

Key determinants in OHD participation:

One-Time-Incentive (financial)
Discounts from vendors (financial)
Business support is worth  US$1,000 - $3,000
Public recognition is worth  US$600 - $1,100
Trusted vendor is worth  US$24,000 - $36,000
Suggestion:
Public sector  Incentives and public recognition
Carriers/vendors  Shipping discounts
Trucking groups  Create a “Trusted vendor” program
Re-align federal/state incentive programs:
Environmental, economic, etc. to support OHD
Require recipients of incentives to accept OHD
Freight Behavior Research
45

Instrumental in the success of the OHD project

Helped us find the:
Right industry segment
Right combination of incentives
Right amount of incentives
The prevailing wisdom was that OHD would not work,
freight behavior research cleared the path for a
transformative project…
 46

Lessons learned about when to intervene, pilot

test, and use freight behavior research

Holguín Veras, J., et al., The New York City Off-Hour Delivery Project: Lessons for City
Logistics, in Innovations in City Logistics, E. Taniguchi and R.G. Thompson, Editors.
2013, Elsevier.
Lesson #1: Make sure there is a market failure
47

Too often, we act when we do not need to:

If private interest coincides with social interest
 Do not do anything, you will make things worse
There are plenty of examples
Confirm the existence of the market failure
Try to understand why the market does not reach the most
efficient solution
Once that is done, act based on science and pragmatism
Lesson #2: Listen to and engage stakeholders
48

Proper and honest private sector engagement:

Confirms/rejects the analysis of the problem to be solved
Provides a thorough idea of the constraints and expectations
of the various stakeholders
Enables the public sector to chart implementation paths that
have a better chance of succeeding
Do not surprise the private sector!
How to engage:
Get input from as many independent sources as possible
Positions expressed may be biased
Designate one person at the agencies as the freight-person
Create an Industry Advisory Group (IAG)
Complement IAG input with targeted outreach to avoid bias
Lesson #3: Follow the Science
49

Private sector behaviors are difficult for “outsiders”

like us to fully understand:
They know what they are doing…
We do not have full access to all the information they have
If they do things that look “crazy” to us, there is a reason
Behavior research is key for successful policy making
Behavioral modeling based on attitudinal surveys
Focus groups, in-depth-interviews
Design policies on the basis of the Science
Lesson #4: Pilot tests could be a good idea’s 50

best friend, or worst enemy

A well-designed pilot test:
Provides a real-life test of a new idea or program
Is an excellent way to identify problems that need to be
ironed out before full-scale implementation
Develops new, unexpected insights into overlooked aspects
Is a great mechanism to attract attention and support
A poorly-designed pilot test:
Could damage the reputation of a good idea, and its
proponents, considerably (false negative)
Could “confirm” the “validity” of a bad idea (false positive)
Do pilot tests based on the science
 51

Concluding comments
Key thoughts
52

Complex problems defy simple solutions…

If things were easy to solve, they would have been
solved already…there is no way to escape complexity
Unilateral solutions do not work in complex systems,
there two many interconnections that will push back
Freight behavior research is the key to navigate these
complex environments
It will give you credibility to negotiate solutions with the
private sector
It will ensure that you do not pursue wrong ideas…
It will prevent political embarrassments…
There is power to tap… 53
Thanks!