Environmental Justice AnalysisTheories, Methods, and Practice - Chapter 13 pps

13 Equity Analysis of Transportation Systems, Projects, Plans, and Policies Urban transit systems in most American cities… have become a genuine civil rights issue — and a valid one — b

13 Equity Analysis

of Transportation Systems, Projects, Plans, and Policies

Urban transit systems in most American cities… have become a genuine civil rights issue — and a valid one — because the layout of rapid-transit systems determines the accessibility of jobs to the black community If transportation systems in American cities could be laid out so as to provide an opportunity for poor people to get meaningful employment, then they could begin to move into the mainstream of American life

Martin Luther King, Jr.

The modern civil rights movement has its root in fighting the inequity in tation systems A landmark event happened in the early evening of December 1,

transpor-1955 (Oedel 1997) After work, Rosa Parks boarded a crowded bus in Montgomery,Alabama She sat down in the rows marked for whites, and later refused to relinquishher seat when sought by a white rider Her arrest led to a bus boycott organized byher pastor, Rev Martin Luther King, Jr In 1954, the Supreme Court declaredunconstitutional the infamous doctrine of “separate but equal,” which the same courtupheld when examining the Jim Crow segregated seating law of Louisiana in 1896(Bullard and Johnson 1997) Later, the Supreme Court declared discrimination ininterstate travel unconstitutional To test this decision, John Lewis and a group ofyoung people began a historic journey by bus from Washington through the DeepSouth in early 1960s These Freedom Riders challenged racial segregation anddiscrimination along their campaign roads and highways These events preceded theenactment of the Civil Rights Act of 1964 On June 15, 1999, Rosa Parks receivedthe prestigious Congressional Gold Medal of Honor, the highest award bestowed bythe U.S government

Although early struggles have focused on inequity in the operations and services

of transit systems, little attention was given to the distributional impacts of portation planning and policies until recently On April 15, 1997, the U.S Depart-ment of Transportation (U.S DOT 1997a) issued the final order to address environ-mental justice in minority populations and low-income populations The order statestwo principles to be observed:

trans-1 Planning and programming activities that have the potential to have a proportionately high and adverse effect on human health or the environment

dis-shall include explicit consideration of the effects on minority populationsand low-income populations

2 Steps shall be taken to provide public access to public information cerning the human health or environmental impacts of programs, policies,and activities

con-Specifically, U.S DOT will take actions to prevent disproportionately high andadverse effects by

1 Identifying and evaluating environmental, public health, and interrelatedsocial and economic effects,

2 Proposing measures to avoid, minimize, and/or mitigate ately high and adverse environmental and public health effects and inter-related social and economic effects,

disproportion-3 Considering alternatives, where such alternatives would result in avoidingand/or minimizing disproportionately high and adverse human health orenvironmental impacts, and

4 Eliciting public involvement opportunities and considering the resultsthereof, including soliciting input from affected minority and low-incomepopulations in considering alternatives

Any programs, policies, or activities that will have a disproportionately high andadverse effect on populations protected by Title VI (“protected populations”) will

be carried out only if

1 A substantial need for the program, policy, or activity exists, based on theoverall public interest; and

2 Alternatives that would have less adverse effects would either (a) have otheradverse social, economic, environmental, or human health impacts that aremore severe, or (b) involve increased costs of extraordinary magnitude

On December 2, 1998, the Federal Highway Administration (FHWA) issuedorder 6640.23 “FHWA Actions to Address Environmental Justice in Minority Pop-ulations and Low-Income Populations,” establishing FHWA’s policies and proce-dures for compliance with EO12898 In a memorandum dated October 7, 1999, FTAand FHWA administrators further clarified their implementation of Title VI require-ments in metropolitan and statewide planning (Linton and Wykle 1999) Specificstrategies were identified in a national conference entitled “Environmental Justiceand Transportation: Building Model Partnerships,” held in Atlanta on May 11–13,

1995 Some of the recommendations are specifically targeted at transportation andland-use systems The recommendations also include research and analysis strategiesthat emphasize the importance of GIS, qualitative and quantitative data, computermodels, and multiple and cumulative impacts

In this chapter, we begin with a look at environmental impacts of transportationsystems Next, we discuss the regulatory environment for incorporating equity analysis

in the transportation planning process and outline major methods and tools that can

be used for conducting equity analysis in transportation planning The first of thesemethods is based on refinement of transportation system performance measures, whichhave different strengths and weaknesses for equity analysis We devote a lot of attention

to equity analysis of mobility and accessibility, including basic concepts, measurementmethods, their uses and limitations, empirical evidence about mobility and accessibilitydisparity, and spatial mismatch In Section 13.5, we discuss use of a hedonic pricemethod for evaluating distributional impacts on property values In Section 13.6, weoutline an integrated GIS and modeling approach to assess differential environmentalimpacts and examine a few analytical issues In Section 13.7, we review the equityimplications of some transportation policies such as congestion pricing Finally, wediscuss the Los Angeles MTA case and some analytical issues involved

13.1 ENVIRONMENTAL IMPACTS OF TRANSPORTATION SYSTEMS

Highways and transit are a mixed blessing Accessibility is important in the realestate market, where there is a well-known motto of “Location, Location, Location.”Good accessibility is valued, but this does not mean living right next to an interstatehighway Proximity to highways or transit is “a double-edged sword” (Landis andCervero 1999) On one hand, people value good accessibility to jobs and servicesbecause of highways and transit In order to obtain good accessibility, householdsand employers look for their residential or non-residential locations a certain distancefrom highways or transit On the other hand, households typically do not like to liveright near highways because of negative externalities generated such as noise andpollution Here is a tradeoff between accessibility and environmental quality Thosehomebuyers who are aware of and sensitive to the negative externality of highwaysand transit and who believe that the benefits associated with locating near highwaysand transit are less than the associated cost will locate away from highways andtransit if they can afford it Those homebuyers who are unaware of or insensitive tothe negative externality of highways and transit will include houses along highwaysand transit in their house-hunting domain They will locate right near highways andtransit if they perceive benefits such as good accessibility and good home valueoutweigh costs such as noise and air pollution

The U.S DOT’s order defines adverse effects from transportation “Adverseeffects mean the totality of significant individual or cumulative human health orenvironmental effects, including interrelated social and economic effects, which mayinclude, but are not limited to: bodily impairment, infirmity, illness or death; air,noise, and water pollution and soil contamination; destruction or disruption of man-made or natural resources; destruction or diminution of aesthetic values; destruction

or disruption of community cohesion or a community’s economic vitality; tion or disruption of the availability of public and private facilities and services;vibration; adverse employment effects; displacement of persons, businesses, farms,

destruc-or nonprofit destruc-organizations; increased traffic congestion, isolation, exclusion destruc-or aration of minority or low-income individuals within a given community or fromthe broader community; and the denial of, reduction in, or significant delay in thereceipt of, benefits of DOT programs, policies, or activities” (U.S DOT 1997a)

sep-Transportation emissions continue to be a significant cause of air pollution,although today’s cars are 70 to 90% cleaner than their 1970 counterparts Thishappens partly because of the rapid increase in travel activity since 1970 Vehiclemiles traveled have almost doubled in the U.S from 1970 to 1990, tripled from

1960, and increased even faster in many metropolitan areas (U.S EPA 1998d) In

1970, vehicle miles traveled totaled 1,114 billion, compared with 2,144 billion in

1990 and 2,405 billion in 1995 As shown in Table 10.1, in 1997, the transportationsector accounts for 76.6% (67,014,000 short tons) of all CO; 49.2% (11,595,000short tons) of NOx; 39.9% (7,660,000 short tons) of VOCs; and 13% (522 tons) oflead (U.S EPA 1998c) In 1993, mobile sources contributed 21% (1.7 million tons)

of 166 air toxics, and on-road gasoline vehicles emitted 76% of all mobile sourceHazardous Air Pollutants (HAPs) nationwide (U.S EPA 1998d)

Transportation impacts decay with distance from roadways related air pollution is the most serious near roadways, and tends to reduce to thebackground level between 500 and 1000 meters (1640 to 3280 feet) from the roadway(FHWA 1978) Traffic noise depends on the volume, speed, and composition oftraffic, and often lessens to the background level at about 1000 feet (300 m) from

Transportation-a highwTransportation-ay (HokTransportation-anson et Transportation-al 1981) Acting Transportation-as bTransportation-arriers, the first row of houses Transportation-along

a transportation corridor absorbs most noise impacts (Stutz 1986) In old city centers,

a street network is often in grids Houses are usually attached and face a street More

or less, most houses in city centers are impacted by noise associated with vehicles.However, a street network consists of links to various functions with various trafficvolumes and speed Some streets serve a local purpose, others serve as minor arteries,and still others serve as major arteries This means that environmental impacts fromthe transportation network are different for different neighborhoods This has beencommonly recognized in the real estate market

Besides environmental impacts, transportation projects and plans also have a widerange of social and economic impacts on the communities The U.S Department ofTransportation’s Community Impact Assessment reference guide (FHWA 1996) iden-tifies a variety of potential community impacts that need to be addressed (Table 13.1).Clearly, transportation projects, programs, and plans have much broader impacts

13.2 INCORPORATING EQUITY ANALYSIS IN THE TRANSPORTATION PLANNING PROCESS

The Clean Air Act Amendments (CAAA) of 1990 mandate integration betweentransportation and air quality processes at all levels of governments The planningprocess is the focal point for consideration of existing transportation system perfor-mance, transportation management strategies, CAAA requirements, transportationimprovement program and implementation, funding, and other factors such as social,economic, and environmental factors TEA-21 consolidates the previous sixteenplanning factors into seven broad areas to be considered in the planning process(same as for statewide planning)

As mentioned earlier, FTA and FHWA administrators jointly issued a dum to further clarify their implementation of Title VI requirements in metropolitan

memoran-and statewide planning (Linton memoran-and Wykle 1999) The memo emphasizes that thelaw applies equally to the processes and products of planning as well as to projectdevelopment They request actions to ensure compliance with Title VI during theplanning certification reviews conducted for Transportation Management Areas(TMAs) and through the statewide planning finding rendered at approval of theStatewide Transportation Improvement Program (STIP) The reviews assess Title VIcapability in terms of overall strategies and goals, service equity, and public involve-ment Review questions address specific procedural and analytical capabilities forcomplying with Title VI “Does the planning process have an analytical process inplace for assessing the regional benefits and burdens of transportation system invest-ments for different socio-economic groups? Does this analytical process seek toassess the benefit and impact distributions of the investments included in the planand TIP (or STIP)?” In 2000, FHWA and FTA revised planning regulations (23 CFR

450 and 49 CFR 619) and environmental regulations (23 CFR 771 and 49 CFR 62)

to propose appropriate procedural and analytical approaches for Title VI compliance.These laws and regulations prompt a paradigm shift in transportation planning

It moves away from the old paradigm, which emphasizes how fast vehicles move.The new paradigm emphasizes how well people’s travel needs are met economicallyefficiently, environmentally friendly, justly, and socially

TABLE 13.1 Community Impacts of Transportation

Social and psychological impacts

Redistribution of population, community cohesion and interaction, social relationships and patterns, isolation, social values, and quality

of life Physical and visual

impacts

Barrier effect, sounds (noise and vibration), other physical intrusions (dust and odor), aesthetic character, compatibility with community characteristics

Land-use impacts Farmland, induced development, consistency and compatibility with

land-use plans and zoning Economic impacts Location decision of firms (move-in, move-out, close, stay-put), direct

impacts of construction on local economy, tax base, property value, business visibility

Mobility and access impacts

Pedestrian and bicycle access, traffic-shifting, public transportation, vehicular access, parking availability

Public services impacts Relocation or displacement of public facilities or community centers,

inducing or reducing use of public facilities Safety impacts Pedestrian and bicycle safety, accidents, crime, emergency response Displacement Effect on neighborhood, residential displacements (characteristic of

displaced population, types and number of dwellings displaced, residents with special needs), business and farm displacements (types and number of businesses and farms displaced), and relocation sites

Source: Federal Highway Administration 1996 Community Impact Assessment: A Quick ence for Transportation Publication No FHWA-PD-96-036 Washington, D.C.

Refer-Equity analysis of transportation systems is more complicated than most otherenvironmental justice issues While most environmental justice controversies origi-nate from specific site-based projects in a particular time, transportation systemsinvolve much broader spatial and temporal dimensions In the spatial dimension,transportation systems (either highway or transit) are linear and penetrate everycommunity, compared with scattered points with limited impact areas of someLULUs in a few communities In the temporal dimension, transportation systemshave definite planning horizons well into 20 years in the future, as well as theirlegacy from the past While transportation projects are carried out one by one indifferent communities, they generally have to go through transportation planningprocesses at the city, county, regional, or state levels Those federally funded projectswith regional significance require evaluation in the metropolitan planning process,which involves both project and policy levels Clearly, equity analysis has to dealwith these two levels.

Table 13.2 lists major methods and tools for conducting environmental justiceanalysis in transportation planning They are built upon existing analytical capabil-ities of metropolitan planning organizations Technical staff for MPOs need to adaptthese methods and tools to deal with environmental justice issues in their regions.For example, many MPOs use a traditional four-step transportation modeling chain(see Figure 13.1) to forecast travel demand and evaluate air quality conformity.These models can be refined and adapted to provide measures for equity analysis,

TABLE 13.2 Methods for Conducting Environmental Justice Analysis in Transportation Planning

User benefits Consumer welfare measures by population groups

• Travel demand models/urban models Direct valuation of travel time saving, operating costs, and safety improvement

• Travel demand models

• Benefit-cost analysis Fiscal impacts of transportation

funding and pricing

Welfare economics and incidence of tax analysis

such as accessibility In particular, mode choice and trip generation submodels arevery useful for equity analysis A mode choice submodel can provide accessibilitymeasures and consumer welfare measures (see Chapter 9 and Section 13.4 below).These measures can be stratified by income and race for equity assessment Tripgeneration and vehicle availability submodels can provide measures for evaluatingthe impacts of transportation projects and plans on mobility by various populationgroups Most data needed to carry out these analyses are available In the following,

we will discuss some of these methods in detail

13.3 TRANSPORTATION SYSTEM PERFORMANCE MEASURES

With a shift in the planning paradigm, new performance measures are adopted toevaluate transportation systems Vehicle Miles of Travel (VMT), vehicle trips, andaverage travel speed are three performance measures that are very important inlinking transportation planning and air quality planning

Planners and engineers have used a lot of variables to measure transportationsystem performance For example, people often feel and complain about congestionduring commuting, but it is not easy to define and measure congestion Congestion

is vaguely defined as “the level at which transportation performance is no longeracceptable due to traffic interference The level of acceptable system performancemay vary by type of transportation facility, geographic location and/or time of day”(Interim Final Rule on Management and Monitoring Systems) With this definition,the federal government gives state and local governments substantial flexibility tomeasure congestion

FIGURE 13.1 Traditional four-step transportation modeling.

Each performance measure shows different strengths and weaknesses whichhave various equity implications (Table 13.3) These measures can be segmented bypopulation group for evaluating distributional impacts of transportation projects Forexample, a VMT measure was used to assess the disproportionate impacts of theBarney Circle Freeway Modification Project in Washington, D.C (Novak and Joseph1996) The goals of the project were to complete a vital freeway link as a majorcommuter route in the District and to divert through traffic from residential streets

to freeways The project was controversial and opposed by a coalition of civic,neighborhood, and environmental groups (Sphepard and Sonn 1997) To measurethe distributional impacts of the changes in freeways and residential streets, VMT-persons were estimated as the changes in VMT on each residential street link andeach freeway segment multiplied by the population that is affected (Novak andJoseph 1996) Reduction in VMT in residential streets is a benefit that accrues tolocal residents, while an increase in VMT on freeways is a burden to local residents.VMT-persons measure the magnitude of benefits and burdens for local residents.Holding population constant, a census tract with a large VMT reduction in residentialstreets benefits more than a census tract with a small VMT reduction VMT changes

in each census tract were derived from traffic modeling To assess distributionalimpacts, VMT-persons were estimated separately for minority and low-income pop-ulations and compared to the proportion of the total population that each groupcomprises in the area Any disproportionate impacts were then spatially identified

13.4 EQUITY ANALYSIS OF MOBILITY AND ACCESSIBILITY

13.4.1 C ONCEPTS AND M ETHODS

Mobility, literally the ability to move around, “relates to the day-to-day movement

of people and materials” (U.S DOT 1997b) Is mobility a merit good or a right?The majority of 1,600 people randomly surveyed in New Mexico considered mobility

as a right (Hamburg, Blair, and Albright 1995) They were asked the question: Doyou believe that the ability to get where you want to go in a reasonable time andfor a reasonable cost is or should be a basic right in the same sense as freedom ofspeech or the pursuit of happiness? Of the approximately 1,600 people surveyed,58.9% responded yes, and 20.8% responded no The survey also shows that femalesand low-income people are more likely to consider mobility as a right

Mobility is an easy to define concept but is not easy to measure The measuresoften used include vehicle availability and number of miles traveled or trips takenduring a given time, the latter referred to as revealed mobility Factors affectingmobility trends include income growth, household size, labor force participation(particularly women), aging, changing levels of immigration, residential and joblocation, changes in the nature of work and workplaces, and advances in informationtechnologies (U.S DOT 1997b)

Accessibility is a concept that is not easy to define and measure When peopletalk about how accessible a place is, they relate that place to other places Accessi-bility involves relative locations of activities or opportunities such as working places,

TABLE 13.3

Transportation Performance Measures and Their Equity Implications

Can be aggregated to any spatial level (link, facility, subarea, region) and to any temporal level (peak-hour, peak-period, daily, weekly, monthly, annually)

Can be used by mode, including trucks Reflects persons’ real demand for travel

Does not directly address air quality impacts from vehicle trips

Favors long-distance travelers

Vehicle miles of travel

Favors users of motorized modes such as car, truck, or transit Favors particularly drivers with more mileage than average Vehicle hours of travel

Does not address non-motorized trips

Favors users of motorized modes

Average speed Speed of person trips

between origin and destination averaged for a period of time

Intuitively appealing Can be aggregated by trip purpose, mode, spatially, and temporally

Related to air quality

Overemphasis on vehicle trips Favors users of motorized

modes

Average trip length, time Distance and time of

person trips between origin and destination averaged for a period

of time

Intuitively appealing Can be aggregated by trip purpose, mode, spatially, and temporally

Difficult to define free flow speed Free flow speed may change and affect the public’s understanding

Favors travelers using congested roads Congestion

(volume/capacity ratio;

level of service)

Volume/capacity Level of Service is a set

of descriptors (such as

A to F) to measure transportation system performance and is defined based on travel time, cost, number of transfer,

volume/capacity ratio, etc

Data are readily available Easy and inexpensive to estimate

Difficult for the public to understand Requires a good estimate of capacity, which is difficult Cannot measure a volume that is greater than capacity in reality Deficient in over-saturated conditions

Difficult to aggregate to a higher level of geography such as a region

Favors travelers using congested roads

Congestion (speed or travel

time)

Defined by speed/travel time threshold range

by facility For freeways, severe congestion occurs if speeds less than 30 MPH

Directly addresses traveler perspective of congestion

Expensive to collect data Favors travelers using

Person Trips Averaged number of

trips per person

Can be aggregated by trip purpose, mode, spatially, and temporally

Does not address accessibility Can show mobility

disparity

TABLE 13.3 (CONTINUED)

Transportation Performance Measures and Their Equity Implications

shopping centers, recreational facilities, residential locations, and so on ity is the ease in reaching them from an origin Proximity does not necessarily meanaccessibility The fact that A is near B does not mean that one can get to B from Aquickly because there may be some barriers between them, such as a river Theconcept of accessibility includes not only relative spatial locations of activities oropportunities but also how fast one can travel from A to B by what mode at whattime It depends on the structure and performance of the transportation network In

Accessibil-a broAccessibil-ader sense, the determinAccessibil-ants of Accessibil-accessibility mAccessibil-ay encompAccessibil-ass reAccessibil-al Accessibil-and perceivedtravel cost, and modal, personal, and location attributes For example, modal char-acteristics that affect perception include comfort, speed, directness, consistency,degree of physical effort, and extent of waiting

Measuring accessibility is the subject of a great deal of empirical investigations,but there is no consensus on a universally accepted measure for accessibility Avariety of accessibility measures appear in the literature (Ingram 1971; Guy 1983;Song 1996) Song (1996) identifies and compares nine accessibility measures In

a concise manner, accessibility measures can be classified into three categories:distance- or time-based measures, cumulative opportunity measures, and gravity-type measures

We can measure the relative accessibility of one location based on distance,travel time or cost, or composite or generalized cost of travel from other places Acommon measure used in the early literature was distance from the center of thecentral business district (CBD) to a location in a city or metropolitan area Thismeasure assumes the predominant role of the CBD employment base in a metro-politan area’s population distribution It corresponds to a monocentric model ofurban structure Clearly, it fails to account for the decentralized nature of urbanforms in the U.S and other countries, particularly the polycentric structure thatpenetrates metropolitan America Not surprisingly, it is one of the poorest accessi-bility measures (Song 1996) The measure of average commuting distance or time

to work, popularly used as a transportation system performance measure, correctsthe inadequacy of the monocentric assumption However, it does not contain anyinformation about the magnitude and distribution of opportunities such as employ-ment at different destinations Use of this measure is perplexed by the “commutingparadox” (Gordon, Richardon and Jun 1991) National and local surveys have repeat-edly reported slightly increasing, or constant and stable commuting time over theyears, although commuting distance increases significantly as does congestion inmost metropolitan areas Evidently, accessibility measures based on commuting timeand distance will show different results Further, the difference in commuting timebetween central city and suburban residents in urbanized areas is small, 18.2 min

vs 20.8 min from the 1990 NPTS data Does accessibility really remain constantover time? Not necessarily Is accessibility a little different between central city andsuburban areas? Not necessarily Of course, the distance measure can be refined toincorporate the importance and attractiveness of different destinations by varyingweights However, distance-based accessibility measures do not fare well overall inrepresenting population distribution in an urban area (Song 1996)

A cumulative-opportunity measure is to estimate the number of opportunitieswithin a specified amount of time One example is the number of shopping facilities

located within 10-min travel by car or transit Another example is the number ofjobs reachable within an average commuting distance or time in a region, or within

a proportion of such averages (Song 1996) For accessibility of a residential location

to employment, we can estimate the number of employees within a 30-min or min commute by car or transit This type of measure has gained increasing popu-larity in analysis of the distributional effects of long-range regional transportationplans In its equity evaluation of the 1998 Regional Transportation Plan, the SouthernCalifornia Area Governments (SCAG) measures accessibility improvement byincome and race/ethnicity in two modes: job opportunities accessible through tripsless than 30 min by transit and job opportunities accessible through trips less than

45-30 min by auto One strength of this measure is its easy, commonsense interpretation

It has gained endorsement from some environmental groups, which facilitates itsincreasing adoption in other metropolitan areas such as Washington, D.C However,its major limitation is the assumption that each and every opportunity is equallyimportant within the specified time For example, a job reachable in 5 min is treatedthe same as a job reachable in 30 min In addition to the lack of spatial differentiation,the specified time or distance is often arbitrary and arbitrarily cuts off opportunitiesjust outside the boundary If this cut-off value is large enough, we will not easilydetect any changes in accessibility caused by transportation improvement projects

in the future As a result of using this measure, distributional impacts could bedistorted In comparing with eight other accessibility measures, Song (1996) findscumulative job opportunity within the average commuting distance to be the poorestaccessibility measure in explaining population distribution in the Reno–Sparks met-ropolitan area in Nevada

Gravity-type accessibility measures, as commonly used by transportation ners and engineers, take the form of opportunities weighted by a distance-decayfunction The common assumption underlying this type of measure is that theimportance of opportunities at the destination declines as the distance from an originincreases They can be derived from a gravity trip distribution model or a logitdestination choice model For transportation modeling, a study area is divided intorelatively homogeneous zones The accessibility for zone i is

plan-A i = ∑j W j f(C ij)where W j = the amount of activity in zone j or the total trips attracted to zone j;

f(C ij) = impedance function or friction factor that represents the nonutility of travel

or spatial separation between zones i and j We obtain a normalized measure bydividing A i by the total amount of activity in the study area

For a common, simple example of evaluating access to jobs, W jis the number

of jobs in zone j, and f(C ij) is typically a negative linear or nonlinear function ofdistance between zones i and j Another example is exponent function (d ij– β) Themost commonly used β-value is 1 β-value can also be calibrated in the trip distri-bution stage of the four-step travel demand modeling chain This measure is some-times referred to as the Hansen accessibility index, credited to Hansen (1959) Twoother functional forms are an exponential distance decay function [exp(–βd ij)] and

a Gaussian function {exp[–β(d ij)2]} For any β-value less than 1, the Gaussian

function decays more gradually at the beginning and then more quickly, with ing distance than an exponential or exponent function does Song (1996) shows thatthe exponent accessibility measure with the most commonly used β value of 1 isnot statistically inferior to any other measures in explaining population distribution.Overall, gravity-type accessibility measures perform better than other measures

increas-A measure of accessibility can also be derived for the multinomial logit model

in the following form (Ben-Akiva and Lerman 1985):

V n = 1/β ln ∑kexp(βV kn)where V kn is the utility function for individual n for option k, and β is the parameter

of the exponential function This measure is, in fact, “the systematic component ofthe maximum utility of all alternatives in a choice set” and “a measure of theindividual’s expected utility associated with a choice situation” (Ben-Akiva andLerman 1985:301) As shown in Chapter 9, it is also the average benefit perceived

by the individual It is often referred to as the log-sum index

Gravity-type accessibility measures can be estimated by modes of transportationsuch as private automobiles and public transit This is done simply by using differentimpedance measures by modes To aggregate accessibility by modes, Shen (1998)proposed a general accessibility index, which takes into account the level of auto-mobile ownership:

A i G = αi A iauto + (1 – αi) A itrans

where A i G = the general employment accessibility for residential location i;

αi = the percentage of workers in location i whose household has

at least one automobile;

A iauto, A itrans= employment accessibility for auto drivers and transit riders,

A cautionary note for using gravity-type accessibility is that the zonal systemaffects the results (Brunton and Richardson 1998) For a large zone structure, it iscertain that some short trips, particularly intrazonal trips, are omitted Thus differentzonal systems will result in different trip length distributions The analyst should becautious in comparing gravity-type accessibility measures for two regions with zonalsystems at different scales

13.4.2 U SING A CCESSIBILITY FOR E QUITY A NALYSIS

Accessibility measures described above can be used to conduct equity analysis forTransportation Improvement Programs (TIPs) and Long-Range Transportation Plans

(LRTPs) in the metropolitan planning process The spatial distribution of accessibilitychanges associated with transportation improvement projects, programs, and planscan be related to the spatial distribution of population in the metropolitan areas Thishelps answer the question of who gains and loses in a regional transportation plan.This analysis consists of three processes:

• Measure changes in the accessibility due to transportation improvementprojects or plans;

• Correlate these changes with distributions of population by income andrace/ethnicity;

• Compare protected populations with other population groups

GIS-based highway accessibility can be measured in the following steps:

1 Obtain or build a road GIS layer, which includes, at least, the attributes

of segment length and design (post) speed Additional useful attributesinclude free-flow speed, congested speed, and turn penalty

2 Calculate design travel time for each segment: time = segmentlength/design (post) speed, or calculate congested travel time = segmentlength/congested speed, or calculate free-flow travel time = segmentlength/free-flow speed

3 Create a road network in GIS such as TransCAD, ArcView/NetworkAnalyst, or ArcInfo This network should include, at least, the calculatedtravel time data

4 Identify origins and destinations in a GIS layer Origins and destinationscould both be TAZ centroids Alternatively, origins could be block-groupcentroids, and destinations could be TAZs

5 Run the shortest path procedure in a GIS to obtain the shortest travel timebetween origins and destinations

6 Obtain data on the amount of activity in each destination zone such astotal employment, retail employment, and households

7 Calculate the accessibility index based on the aforementioned equation.This index could represent employment, retail, or residential accessibility,depending on the type of activity used in the calculation

Use this procedure to calculate the accessibility measurements for a base-yearroad network Then repeat for a year with transportation improvements The dif-ference between the two represents the accessibility changes caused by transpor-tation improvements

The accessibility index based on free-flow speed does not account for tion in some segments of the road network, thereby representing an ideal situationand a high end of true index values On the other hand, the accessibility indexbased on congestion speed represents the real-world situation Congestion speeddata can be obtained through traffic monitoring, which, however, has limited geo-graphic coverage Usually, we can obtain congestion speed through the four-steptransportation modeling for a base year and any forecast year Without access to a