The 5 Es

David Levinson

Framework

Orientation of CriteriaOurselves Others

Within Efficiency EquityExpediency

Relation to classic Land use-Transportation analysis

Outside Experience Environment

Motivation

• Welfare comprises efficiency and equity. We are also concerned with the environment and the experience of users.

• An allocation is Pareto Efficient if there is no other allocation in which some other individual is better off and no individual is worse off.

• Benefit/Cost analysis concerned with net benefits, not distribution.

• Transportation projects and policies create both winners and losers from mobility, accessibility, environmental, and economic standpoints.

Externality Theory

• An externality is “a commodity bundle that is supplied by an economic agent to another economic agent in the absence of any related economic transaction between the two agents. (Spulber).”

• Arise from lack of property rights.• May be positive or negative. Positive include

consumption externalities and network externalities. Negative include congestion, pollution, accidents, etc.

• First Best Solution: Internalize Externality: Polluter Pays Principle.

• Transactions costs may make internalization difficult.

Measures of Effectiveness (MOE)

• Second-Best Solution: Regulate Developer• Infrastructure Has Multiple Attributes - A

Doctor Looks at More Than Temperature, Planners Should Consider More Than Capacity

• The Right MOE’s Vary Based on Facility Being Analyzed,

• MOE’s Should Consider Not Only Current Status of System, but also Its Direction of Change.

Second Best Regulation: Minimizes the Cost of:

Prevention (Build) The cost of infrastructure required to

maintain the performance indicators (Engineering or Statistical), or

(Manage) The cost of demand and supply management to maintain the performance measures.

Damage (Accept) The cost to the community of

worsening the performance indicators in the absence of the infrastructure.

Criteria for Choosing MOE

1. Different measures (e.g. transit and auto level of service) should be collectively complete in that one could combine them to attain an overall measure.

2. Each measure should scale or aggregate well (e.g. it should be possible to combine measures of auto level of service measured on separate links or for separate trips).

3. The measure should align with user experience and be understood by those users.

4. The performance indicator must be measurable, or calculable from available (observable) data.

5. The measure should be predictable, or able to be forecast 6. It must be useful in a regulatory or control context (so that

the measure can be used to allow or restrict new development to maintain standards, or to help guide operational traffic engineering decision).

Criteria for Selecting Measure of Effectiveness

It aids in identifying opportunities to increase the systemwide net benefits through public investment in improvements or changes in management,

It minimizes the cost to achieve necessary measurement accuracy, and

It produces the right incentives.

Normative and Positive

• To say that the speed on a link is 50 kilometers per hour tells us nothing about whether that is good or bad, it simply is.

• By comparing the measure to a normative standard (for instance, a speed limit), we can then determine whether we have a speeding problem (the speed limit is 30 km/h), a congestion problem (the speed limit is 110 km/h), or no problem.

Desired Outcomes of the Transportation System

(Caltrans)Outcome Candidate MeasuresEFFECTIVENESSAND EFFICIENCYMobility/Accessibility

Travel Time,Delay,Access to Desired Locations,Access to System

Reliability Variability of Travel TimeCost-Effectiveness Benefit/Cost Ratio,

Outcome Benefit per CostRESPONSIBILITYSustainability Household Transportation CostsEnvironmental Quality National and State StandardsSafety and Security Accident and Crime RatesEquity Benefits per Income GroupCustomer Satisfaction Customer SurveyEconomic Well Being Final Demand

(Value of Transportation to the Economy)

Data Sources and Collection

• Supply Data– Measured - Engineering Cost Study– Predicted - Statistical Cost Study

(Many Projects)

• Demand Data – Measured - Operating Agency

Utilization Data– Predicted - Statistical Forecasts

Four Perspectives on Efficiency

Perspective Profession

Mobility and Safety Engineers

Utility (Consumer’s Surplus) Economists

Productivity Managers

Accessibility Planners

Reason for Multiple Measures

• Planning, investment, regulation, design, operations, management, and assessment.

• Each profession claims to represent traveler.• Professions take the "objective" viewpoint of

the omniscient central planner (who may in fact be an engineer, manager, or economist) rather than the "subjective" perspective of the travel consumer.

Mobility

• Highway Capacity Manual (segments)• Texas Transportation Institute (metro

areas)• Quantitative and Qualitative• Auto and Non-auto• Scale: Intersection, Link, Subnetwork,

Trip, Network• Basis: Time or Flow

Roadway Mobility Measures

Measurement Scale Volume and Capacity Time

Intersection approach Volume to Capacity Ratio: Stopped Delay:

Queue Length

Total Intersection Critical Lane Volume: Average Delay:

Road Segment Density: Average Delay:

Volume to Capacity Ratio: Average Travel Time:

Road Network Area Cordon: Average Travel Time (Distance):

Area Screenline: Average Percent Delay:

Average Congestion Index: Average Trip Time (Distance) Ratio:

Average of Area Intersection:

Shoulder Hour Index:

Distribution Measure: Distribution Measure:

Qualitative Mobility Measures

Volume &Capacity• Parking Availability

and Cost• Connectivity• Conflict with Non-

auto System• Hazard• Auto Service

Stations• Comfort

Time• Coverage• Aesthetics• Destination

Distribution• Information

System

Non-Auto Mobility Measures

Measurement Stage Volume and Capacity Time

Walk (Bike) and Walk Access and Egress to Transit

Sidewalk (Bikeway) Ratio Coverage

Connectivity Circuity

Hazard Delay

Bicycle Parking Aesthetic

Travel Time

Auto Access and Egress Parking Availability and Cost

Park and Ride Access Time

Waiting Waiting comfort Frequency

In-Vehicle Usage Opportunity

Service Comfort Reliability

Absolute Time

Relative Time

Directness

Productivity Measures

Where: T = Travel on the system in question (person-km or ton-km), H = Hours of labor by employees ofthe highway agency (including professional drivers), D = Hours of time by the driver and passengers spenton the network in question (excluding professional drivers), K = Dollars of public capital spent (buildingand maintaining the infrastructure), V = Dollars of private capital spent (the share of the cost of owningand operating a vehicle, exclusive of taxes to pay for public capital for its use on the network in question), ldenotes links in the set of links L under question.

Description Formula

Productivity of Public Labor (PGL)

Productivity of Private Labor (PPL)

Productivity of Public Capital (PGK)

Productivity of Private Capital (PPK)

€

PGL =

Tl

l

∑

H l

l

∑

€

PPL =

Tll

∑

Dll

∑

€

PGK =

Tll

∑

K ll

∑

€

PPK =

Tl

l

∑

Vl

l

∑

Accessibility MeasuresDescription Formula

Accessibility (A) in zone i depends on the opportunities (e.g. jobs P) in zone j and the transportation cost cij between them

Job - Worker Ratio (R) in zone i at radius r (in transportation cost) is the Jobs (P) within radius r divided by Workers (Q) within radius r

Density (D) in zone i is the sum of jobs and workers within radius r, divided by the area contained within

Difference () in zone i is the difference between the number of jobs and workers in radius r

Force (F) between zones i and j is the product of the jobs (P) in zone j and the workers (Q) in zone i and a function of the transportation cost cij between them

€

Ai = Pj f cij( )j

∑

€

Ri =

Pjj =1

r

∑

Qjj =1

r

∑

€

Di =

Pi +Qii =1

r

∑

πr 2

€

i = Pi − Qii =1

r

∑

€

Fij=QiPjfcij ()

Accessibility

Accessibility

0.00001

0.0001

0.001

0.01

0.1

1

10

100

1000

0 20 40 60 80 100 120 140

Travel Time

Propensity, Normalized Trips

0

50000

100000

150000

200000

250000

300000

Job Supply

Propensity NormalPJS JobSupply

)(*)(*)2/1(onoffoffon

PPQQCS −+=

Travel Time(Price)

Total Flow(Quantity)

Supply (Metering-off)

Supply (Metering-on)

Demand

Poff Pon

Qoff Qon

Change in Consumer Surplus

Consumers’ Surplus

Consumers’ Surplus Criticisms

• Transportation rather than activities as the base for consumer's surplus

• Aggregation error involved. • No consideration of choice and the existence

of non-user benefits in the consumers’ surplus metric.

• The costs and benefits associated with spillovers and externalities are often improperly captured

Equity: Some Terms

• Horizontal equity: allocation of benefits and costs among individuals and groups who are similar.

• Vertical equity: distribution of benefits and costs across different groups.

• Process equity: equal access to the planning and decision making process.

• Result equity: examines the outcome.

Gini coefficient and Lorenz Curve

21

1

AA

AGini

+=

Gini = 0: Perfect Equity Gini = 1: Perfect Inequity

Equity

Measuring Equity: Entropy

To analyze traffic data, we can take: yi = proportion of total delay accrued by each

individual

H statistic approaches zero as the distribution approaches complete inequality

€

H = − yj • logk yj( )j

∑

H = the entropy statistic

yj = the proportion of average net gains to the jth class

k = the log base

Measuring Equity: Redundancy

• R-value of 0% represents complete equality €

R = 1− HH max

R = the measure of redundancy

H = the calculated entropy

H max = the maximum possible entropy

Equity Impact Statement

Process Outcomes

Stratification Opportunity toEngage inDecision-Making

Mobility Economic Environmental Health Other

Population

Spatial

Temporal

Modal

Generational

Gender

Racial

Cultural

Ability

Income

Racism & Transit

• Bus transit is perceived as largely serving poor and minority passengers.

• Rail transit is perceived as largely serving middle class white passengers.

• Efforts to build rail transit absorb a large share of transit funding, and serve those who have choices, at the expense of those who don’t.

NIMBY and Equity

• NIMBYs - Not in My Back Yard ‘selfishly’ oppose new road projects

• Assumed to be on “property value” grounds.• May in fact be on “mobility” basis.• Neighbors do not gain mobility benefits in

same way as through trips. Roads often benefit non-locals at expense of locals.

• Most projects create both winners and losers.• Losers use politics to stop projects which may

have an overall net benefit to society.

Environmental Justice

• Only considers "fair treatment for people of all races, cultures, and incomes (Executive Order 12898)" regarding the development of environmental laws and policies. It thus only examines environmental outcomes and only addresses a few strata.

• What about Transportation Justice?• What about Economic Opportunity?

Environmental Costs

Cost Category Long Run Average Cost

($/vkt)

User $0.13

Infrastructure $0.0174

Freeflow Time $0.15

Congestion $0.0045

Accidents $0.031

Noise $0.006

Air Pollution $0.0056

Total $0.34

Environmental Impact Statement

Figure: Flowchart depicting the decision process options for NEPA• Node 1: Proposed Action• Node 2: Coordination and Analysis• Node 3: Significant Impact ?

o IF YES:ß Node 3-A: Notice of Intent & Scoping Processß Node 3-B: Draft EISß Node 3-C: Public Commentß Node 3-D: Final EISß Node 3-E: Record of Decision (ROD)ß Node 3-F: Agency Action

o IF NO:ß Listed Categorical Exclusion (CE) OR Documented CEß Coordination and analysis as neededß Document appropriatelyß Agency Action

o IF Unknown:ß Environmental Assessment

• IF No significant impactso Finding of No Significant Impact (FONSI)o Agency Action

• IF Significant impacto GO TO NODE 3-A AND CONTINUE

http://environment.fhwa.dot.gov/tutorials/a2.htm

Experience: Travelers and Subjectivity

• Just as Einstein noted that the point of view of the observer shaped the measurement of time, point of view also affects the perception of time as a measure of transportation level of service.

• Moving towards trip-based measures of effectiveness will more closely align with user experience

Summary

• Externalities Provides Underlying Efficiency Rationale for Development Regulation

• Unfortunately, “First Best” Solution (P=MC) is Not Always Feasible

• Regulating Supply is a “Second Best” Solution

• Multiple Measures of Effectiveness are Required to Understand Impact of Development on Capacity Utilization

• Efficiency is not the only issue.

Conclusions: Efficiency is Multi-Splendored

• Four Classes of Efficiency Measures: Mobility, Utility, Productivity, Accessiblity.

• Each is a gauge, none should be exclusive.

• None captures the subjective perspective of travelers.

• New measures must be developed which do reflect the customer.

Conclusions: Equity is Efficient

• Equity matters for 2 reasons– Equity is the “right thing” to do– As minority groups become empowered, Equity becomes the

efficient thing to do, since in the absence of its recognition, nothing could be built. In economic terms, the “side payments” are worth the cost

• Race and location are the source of among the greatest inequities in our society, though not the only ones.

• Need to think beyond single project: Develop means for compensation of losers from gains of winners. Side payments, bargains, and bundles of projects may accomplish this.

• Danger of log-rolling turning into pork barrel.• Equity must be broadly considered. • Things need not be strictly fair, but the unfairness inherent

should not be unknown.

Conclusions

• Select Measures of Effectiveness• Collect and Forecast Data• Establish Standards (Absolute or

Relative)• Open System to Peer Review and Public

Scrutiny• Continuously Evaluate System • Implement Monitoring System in

Regulation