Transport Policy 69 (2018) 35–48

Contents lists available at ScienceDirect

Transport Policy
journal homepage: www.elsevier.com/locate/tranpol

Accessibility analysis for transportation projects and plans T

a,∗ b b
Louis A. Merlin , Jonathan Levine , Joe Grengs
a
School of Urban and Regional Planning, 777 Glades Road, College of Design and Social Inquiry, Florida Atlantic University, Boca Raton, FL 33431-0991, USA
b
Urban and Regional Planning Program, College of Architecture and Urban Planning, 2000 Bonisteel Boulevard, The University of Michigan, Ann Arbor, MI 48109-2069,
USA

1. Introduction transportation is largely derived from people's demand to reach their

destinations, rather than for the sake of movement per se. Though this
The concept of highway level-of-service (LOS) and associated mo- view enjoys near-consensus status within the transportation profes-
bility metrics, including measures of congestion-related delays and sions, transportation planning practice has rarely confronted its full
prospective value of travel time saved (as incorporated in cost-benefit implications: metrics of the quality or quantity of movement alone do
analyses), have for decades shaped transportation-planning practice not indicate how well transportation is serving the purpose of accessi-
(Cohen, 1995; Laird et al., 2014; Schrank et al., 2015; Weiner, 2012). bility. Moreover, since transportation infrastructure expansion can in-
As recently as 2016, the Federal Highway Administration proposed a duce distance-increasing land-use change (Cohen, 1995), mobility im-
set of performance measures for the national highway system pursuant provements can sometimes translate, in the longer term, into degraded
to MAP-21 legislation; seven out of the proposed eight performance accessibility when increased distances are not compensated for by im-
measures are mobility-oriented measures, linked to vehicular speeds, proved travel speeds (Levine et al., 2012).
whereas none are accessibility measures (U.S. Federal Highway While this outcome would be undesirable based on the accessibility
Administration, 2017). Mobility thinking has become so ingrained perspective, it would be judged a mobility success so long as roadway
within the transportation planning profession that many evaluations of delays were reduced in the process. Standard state-of-the-practice mo-
transportation projects focus on mobility-related level of service bility metrics may be insensitive to the accessibility outcomes of plans.
without discussing the rationale or justification behind the widely ac- The conventional mobility framework, rather than presuming positive
cepted level-of-service measure. accessibility outcomes from mobility improvements, is quite indifferent
The mobility-centered approach is operationalized in transportation to accessibility implications.
planning's “predict-and-provide” formulation (Banister and Button, Aligning transportation planning with transportation's fundamental
1993) in which planners model travel flows over transportation net- purpose thus demands a shift from mobility to accessibility as the core
works, simulate the impact of future land-use and population change on goal. In some ways, this shift is underway. Regional planning agencies
those flows, and identify opportunities to improve flows through tar- have begun to evaluate alternative development scenarios based on the
geted expansions of capacity. In practice in the US, this has often meant accessibility each provides. Less common is the accessibility-based
expanding highway links whose forecast volume-to-capacity ratios analysis of individual projects or groups of associated projects; neither
presage unacceptable degradation in LOS. proposed land-use changes nor proposed transportation infrastructure
This planning approach has been criticized for overestimating the or service changes are typically evaluated from an accessibility per-
certainty of modeling results (Zhao and Kockelman, 2002), imposing an spective. Yet the shift from the regional-scenario level to the project
overly narrow technical rationality on the transportation planning level is vital, since it is the aggregation of multiple incremental deci-
process (Willson, 2001), and ignoring the futility of a strictly supply- sions that ultimately forms altered regional transportation and land use
based approach to transportation problems (Downs, 2005; Meyer, futures. In other words, operationalizing accessibility in practice de-
1997). Less common is the critique of the standard transportation pends in part on bringing accessibility-based analysis to applied trans-
paradigm, with its focus on the goal of preserving and expanding mo- portation and land-use decisions, roles currently filled by mobility-
bility, rather than accessibility (Levine et al., 2012). The accessibility- based level-of-service or traffic-impact analyses.
oriented critique argues that planning and evaluation frameworks that This shift, in the context of land-development decisions, is the
are not grounded in the fundamental purpose of transportation threaten subject of a companion article (Levine et al., 2017). For transportation
to generate perverse transportation outcomes. decisions, the shift from the regional-scenario to the project level would
An accessibility-based view starts from the idea first articulated allow planners to distinguish between transportation investments that
early in the 20th century (Bonavia, 1936) that the demand for are likely to increase accessibility over the longer term from those that

∗
Corresponding author. 777 Glades Road, SO 44 Room 284I, College of Design and Social Inquiry, Florida Atlantic University, Boca Raton, FL 33431-0991, USA.
E-mail addresses: [email protected] (L.A. Merlin), [email protected] (J. Levine), [email protected] (J. Grengs).

https://doi.org/10.1016/j.tranpol.2018.05.014
Received 18 January 2017; Received in revised form 26 January 2018; Accepted 28 May 2018
0967-070X/ © 2018 Elsevier Ltd. All rights reserved.
L.A. Merlin et al. Transport Policy 69 (2018) 35–48

may degrade it and is the subject of this article. Transportation projects understand how transportation and land-use systems facilitate or im-
quite obviously influence the speed of travel, but less obvious are the pede access to opportunities spread across metropolitan areas for spe-
influence transportation projects have on the land-use system: if a cified population segments.
transportation investment induces greater proximity of origins and Regional scenario-based accessibility analysis is made easier by the
destinations (compared with a no-build alternative), the land-use im- fact that these future scenarios are typically “what if” possibilities ra-
pact can amplify the effect of the transportation investment in enabling ther than predicted futures. Planners can characterize how the future
accessibility. By contrast, when the transportation investment induces might look in terms of both land-use patterns and transportation in-
greater spread of origins and destinations, the land-use effect can di- frastructure at a particular point in time according to desired patterns,
minish, negate, or even reverse a transportation project's mobility- and from these data, the relative accessibility performance of various
based accessibility gains. An accessibility-based evaluation of a trans- scenarios can be readily compared.
portation project requires appropriate tools to distinguish between ac- This contrasts with the analysis of transportation projects, an ana-
cessibility-enhancing and accessibility-degrading transportation in- lysis that demands forecasting of induced land-use changes before ac-
vestments. cessibility impacts can be meaningfully gauged. Accessibility-based
This article defines and demonstrates such tools, testing them out analysis has rarely been used for project-level analysis of this type.
for two sets of affiliated roadway projects in metropolitan San Antonio, Transportation projects can be thought to have two effects on accessi-
Texas, a rapidly growing sunbelt region in the United States. The first is bility, one short term and direct and the other longer term and indirect.
the highway elements of the Long-Range Transportation Plan for the Traditionally only the short-term, mobility-based impacts have been
San Antonio metropolitan area for 2015, incorporating multiple pro- considered in transportation project evaluations. The short-term impact
jects throughout the region. The second is a series of expansions and is the reduction in travel times which increases accessibility by redu-
improvements to a suburban ring road, referred to as Loop 1604. The cing generalized travel costs. The long term impact is from the induced
analyses forecasts land-use impacts of each of these set of transporta- land use effects, which may counteract or further enhance the short-
tion projects using TELUM, a freely available transportation/land-use term mobility benefits (Boarnet and Haughwout, 2000).
model funded by the Federal Highway Administration and housed at In metropolitan area travel-demand modeling, analysts commonly
the New Jersey Institute of Technology. Impacts on each suite of pro- take the future spatial pattern of employment and households as fixed
jects on future auto accessibility to employment are then gauged based and exogenous, and then predict trip patterns based upon that spatial
on the projected changes to land use patterns. The results show in both pattern. This was the case for the study area of this article. The Alamo
cases that taking an accessibility-based approach reveals outcomes that Area Metropolitan Planning Organization has a single land-use forecast
differ substantially from the standard mobility-based approach. each 5-year period between 2020 and 2040. This future land-use
This article seeks to develop an approach to accessibility analysis of forecast is not contingent on the adopted transportation plan (Alamo
transportation projects that is relevant to planners in applied practice; Area Metropolitan Planning Organization, 2015). The transportation
for this reason, availability and usability of the land-use model is cen- infrastructure is then planned in response to the predicted trip patterns
tral to our goal. This is the principal benefit of TELUM, an im- that result from the exogenously provided land use pattern.
plementation of a transportation/land-use interaction model that has However, this common practice of assuming fixed future land uses is
been in broad planning use since the 1970s. Our approach is available problematic, since the pattern of future land uses is itself influenced by
to researchers employing more advanced models including those based the provision of transportation infrastructure. If new development sig-
on microsimulation, yet for the reason described above we chose to nificantly decentralizes due to new transportation infrastructure, the
demonstrate it with a model that is already embedded in planning speed benefits of such infrastructure may be partially or completely
practice. counteracted in accessibility terms by increasing travel distances
(Grengs et al., 2010; Levine et al., 2012). An accessibility analysis of a
2. Accessibility evaluation of transportation projects transportation plan that does not allow for induced land-use change
implicitly assumes none; such analysis amounts to a mobility analysis in
Accessibility is broadly defined as the ease of access to destinations another form, since all mobility improvements will translate into ac-
of interest. A growing number of researchers and analysts have been cessibility improvements if land use patterns remain unaltered.
arguing for a shift from mobility-based evaluation to accessibility-based To ascertain the full accessibility impacts of a transportation project,
evaluation for years (Cervero, 1996). Accessibility measures can take a land-use forecast sensitive to the impacts of the proposed transpor-
into account up to four components: A transportation component, a tation projects is required. That is, an accessibility analysis of the
land-use component, a temporal component and an individual com- proposed projects must account for both the short-term mobility im-
ponent, though often the temporal and individual components are pacts and the longer-term land use changes.
omitted due to methodological complexities (Geurs and Van Wee, Fig. 1 illustrates the method used here to address these require-
2004). Accessibility can be measured through various formulae, in- ments. On the bottom the current state of the practice is diagrammed,
cluding the cumulative opportunity, gravity, person-based, and utility where a future land use forecast serves only as an input into describing
measures. In this paper, we focus on gravity-based measures of acces- future travel patterns. This assumes that proposed infrastructure has no
sibility to all employment as a succinct summary measure of the per- role in shaping future land use patterns. The new proposed method is
formance of the transportation-land use system (Ahlfeldt, 2011). In a on the top. It assumes that the proposed transportation infrastructure
more detailed analysis, accessibility to other, more specific types of will reshape land-use patterns. It builds on existing procedures of mo-
destinations, or the distribution of accessibility benefits across the po- bility-based analysis but incorporates both changes to travel patterns
pulation may be of interest as well. and changes to land use patterns within the broader framework of an
Accessibility-based analysis has made some inroads in the practice accessibility analysis.
of regional scenario planning. Seattle, Chicago, and San Francisco
metropolitan regions are among those that have employed accessibility 3. Project-level analysis: the differences between transportation
performance measures to evaluate potential regional futures (Chicago and land use project evaluation
Metropolitan Agency for Planning, 2010; Metropolitan Planning
Commission, 2009; Puget Sound Regional Council, 2008). Accessibility A companion article (Levine et al., 2017) argues that project-by-
metrics have been employed to understand the economic, equity, and project decision-making is an essential component of most transporta-
multimodal performance of future regional transportation/land-use tion and land-use planning decisions. It argues that project-level ana-
scenarios. In other words, accessibility measures have been used to lysis is conceptually different from regional-scenario analysis and

36
L.A. Merlin et al. Transport Policy 69 (2018) 35–48

Fig. 1. Accessibility-based transportation Project evaluation.

requires distinct tools and methods. Therefore, the companion piece comparison. For example, the accessibility impacts of transportation
proposes a method for carrying out accessibility-based evaluation of project “X” may be judged in comparison with a No-Build scenario, or
individual land-use projects. possibly in comparison with transportation project “Y.” By contrast,
To summarize the key argument of this companion piece briefly, the land-development, while regulated through public planning, is typically
key difference between regional-scenario analysis and project-level a private affair. This impedes the delineation of clear basis for com-
analysis is that the impact of proposed, marginal projects must be parison for land-use projects, as discussed in more detail in the com-
projected into the future. That is an individual land-use project, such as panion piece.
a proposed development in a specific location, must have its marginal The second difference between accessibility analysis of transporta-
transportation impacts forecast to assess its accessibility impact. tion projects and land-development projects is one of scale. The com-
Likewise, an individual transportation project, such as a proposed new panion paper demonstrated the accessibility analysis of land-develop-
rail line or highway, must have its marginal land use impacts forecasts. ment projects as small as around 200 residential units or 100,000
Such forecasts are not usually part of regional scenario analyses; sce- square feet of retail. Any land-use change whose mobility impacts can
narios are often constructed via a “what if” method for the purpose of be analyzed through traffic-impact analysis can be evaluated in acces-
maximizing policy contrasts. Also, in scenarios land-use and transpor- sibility terms. By contrast, small transportation projects (such as an
tation changes are envisioned as occurring in concert, rather than in- intersection improvement) are not as readily evaluated for their land-
crementally and marginally. The companion article also emphasizes the use impacts. For this reason, this paper analyzes large groupings of
need for practical methods that can be put into widespread use by transportation projects: the 2015 Long-Range Transportation Highway
planners with a minimum of technical sophistication required. Simpler, Projects for the San Antonio region, and a series of corridor projects
easy-to-adopt tools will abet wider adoption of these new methods. associated with Loop 1604, a suburban ring road.
These key differences between regional-scenario analysis and project- Such large bundles of transportation projects have a discernible
level analysis are detailed in Table 1 below.Where the companion ar- impact on regional measures of average household accessibility. This
ticle proposed a new approach to accessibility-based analysis for land- contrasts with land-development projects, whose accessibility impacts
use projects, this article focuses on the prospective accessibility-based are likely to be a “drop in the ocean” relative to total regional acces-
analysis of transportation projects. sibility. The land-development case addressed the problem of small size
Accessibility analysis for transportation projects differs from that of via an accessibility-elasticity. However, this accessibility-elasticity
land-use projects in three fundamental ways: 1) Basis of comparison 2) metric is not necessary in the analysis of sizable transportation projects,
Scale of effect and 3) Technical requirements. The basis of comparison and it is not used in the current study.
refers to what the transportation projects are compared to in the eva- The third difference that makes the evaluation of transportation
luation process. Transportation projects are generally the product of projects more challenging than land-use projects is that projecting fu-
public planning and investment, an attribute that facilitates analysis ture land-use impacts is technically more difficult than projecting future
because alternatives are readily identified that would form a basis for transportation impacts. For the case of land-use projects, planners and

Table 1
Different types of accessibility analysis.
Scenario Plan Project

Land Use Future transportation and land use changes are envisioned as part of a Individual development proposal is incorporated as a marginal change to land-use
holistic set of changes; Future state is evaluated for accessibility patterns. A model is used to forecast accompanying changes to travel patterns;
benefits Marginal difference to transportation and land use systems is input to accessibility
analysis
Transportation One or more transportation projects (i.e. new or expanded roadways) is assessed for
how it will change travel patterns; A model is used to forecast accompanying changes
to land use patterns; Marginal difference to each is input to an accessibility analysis.

37
L.A. Merlin et al. Transport Policy 69 (2018) 35–48

engineers can readily project transportation impacts by turning to the article. The goals of the present article are similar but include the de-
widely available and accepted method of traffic impact analysis monstration of the general applicability of accessibility-based analysis
(Institute of Transportation Engineers, 2006). By contrast, for the case to transportation projects and the illustration of the consequences of
of transportation projects, practitioners have no widely adopted this approach by directly contrasting how accessibility and mobility
method for projecting land-use impacts. Projecting land-use impacts analyses can lead to divergent results.
typically requires complex software tools and a range of data – often
including parcel-level land categories and small-scale employment – 4. Data and methods
that are not readily within reach of most local and even many regional
transportation or land-use planning agencies (Brown and Lee, 2013). We selected the San Antonio, Texas metropolitan area, also known
as the Alamo Area, for analysis because it is a fast-growing metropolitan
3.1. Examples of accessibility-based transportation project analysis region with significant investment planned in new highways. According
the Alamo Area's long range transportation plan, the Alamo
We evaluated the peer-reviewed literature for similar project-level Metropolitan Area expects growth from a population of 2.0 million in
analysis by examining it along three dimensions. First, does it analyze a 2010 to 3.4 million by 2040 (Alamo Area Metropolitan Planning
specified project or bundle of proposed transportation projects? This is Organization, 2015). Employment is likewise expected to grow rapidly
the criterion by which we distinguish project-level analysis from sce- from 0.9 million jobs in 2010 to 1.7 million by 2040. Fig. 2 illustrates
nario analysis. Second, is it a prospective analysis? That is, is the ana- the current employment density in the 5-county metropolitan region,
lysis of accessibility impacts a forward-looking evaluation of the im- with employment centers located in historic downtown San Antonio as
pacts of the proposed set of projects. Retrospective analysis is of interest well as along a band across the northern suburbs. The metropolitan San
for researchers but is not useful for the evaluation of proposed trans- Antonio region embodies a strong decentralization trend, with the
portation projects in practice. And third, are induced land-use impacts highest population and employment growth in recent years occurring in
forecast? For a method to be useful for the prospective accessibility- suburban locations and along major highway corridors.
based evaluation of transportation projects, it must meet all three of the Analyzing highway projects was of particular interest because these
above criteria. offer the potential to improve mobility without necessarily producing
We found a small body of work that examines the impacts of large concomitant accessibility benefits. Highway projects usually can im-
transportation infrastructure projects on accessibility (Fan et al., 2010; prove vehicular speeds, but they are also prone to inducing decen-
Geurs et al., 2012; Gjestland et al., 2012; Gulhan et al., 2014). However tralized land use patterns. Therefore, analyzing sets of highway projects
almost all these papers examine only the mobility effects of the trans- potentially allows for the distinction between transportation projects
portation projects they analyze (see Table 2 below). In other words, that enhance mobility from those that enhance accessibility.
these analyses assume there is no induced land-use change, only im- We analyze two sets of transportation projects for their accessibility
provements or changes to travel times. This is a very limiting analysis, impacts. The first set of projects includes the highway components of
because if only mobility effects are accounted for any transportation the region's long-term transportation plan adopted in 2014, known as
project that improves speeds will be accessibility enhancing. “Mobility 2040” (Alamo Area Metropolitan Planning Organization,
Interestingly, most of the prospective studies of the impact of pro- 2015). Note that we only consider effects of these projects for the period
posed transportation projects on accessibility focus on transit and from 2010 to 2020. The goals of the plan are typical for a metropolitan
railways projects. Presumably if such projects induced additional de- long-range transportation plan, and include decreasing traffic conges-
velopment near their transit stations, their accessibility impacts might tion, improving public transit, mitigating environmental impacts, sup-
be greater than stated in these analyses, because each of these assumed porting economic growth, improving safety, and coordinating with
no land use impact. The possible under-scoring of the accessibility local land use plans. The 25-year LRTP (Long Range Transportation
benefits of transit projects highlights the significance of considering Plan) includes a total of $17.2 billion in transportation funding for
induced land use change as part of an integrated accessibility analysis. operations, maintenance, safety improvements, and roadway expan-
Geurs et al., 2012 is a rare example of a prospective analysis for the sions. Approximately $2.1 of this total is applied to roadway expan-
accessibility impacts of a transportation project that includes antici- sions. A map of the proposed roadway expansions is included below in
pated impacts to land use change. This study employs a land-use model Fig. 3.
to forecast variation in employment locations as a result of the proposed The second set of transportation projects is a series of transportation
transportation projects and thus exemplifies an approach to accessi- improvements along San Antonio's outer loop known as “Loop 1604.”
bility-based transportation project analysis proposed in the present For this analysis, we assume that the rest of the region's planned

Table 2
Literature incorporating the prospective accessibility analysis of transportation projects.
Author Year Project Level Prospective Land Use Forecast Transportation Project Type

(Jiang, 2016) X Transit

(Gulhan et al., 2014) X X Transit
(Bocarejo et al., 2014) X Transit
(Hensher et al., 2014) X X High Speed Rail
(Stepniak and Rosik, 2013) X Highway
(Bocarejo and Oviedo, 2012) X Transit
(Kilby and Smith, 2012)2 X NA
(Geurs et al., 2012) X X X Transit
(Gjestland et al., 2012) X Bridge
(Tiwari and Jain, 2012) X Transit
(Curtis, 2011) X Transit
(Fan et al., 2010) X Transit
(Bertolini et al., 2005) X NA
(Department for Transport (UK), 2005) X NA
(Gutiérrez, 2001) X High Speed Rail
(Lomax, 1997) X X NA

38
L.A. Merlin et al. Transport Policy 69 (2018) 35–48

Fig. 2. Employment density for Alamo region, 2010

transportation projects go ahead as planned but subtract out the asso- 4.1. The TELUM land use model
ciated Loop 1604 projects to examine their marginal impacts. The
analysis is framed this way because we take the existing long-range An objective of this study is to develop a methodology that could be
transportation plan as the status quo. Loop 1604 is in most locations a widely used with minimum demand for technical capacity or data-in-
4-lane expressway with frontage roads running parallel. The Loop 1604 tensive input requirements, and this goal guided the selection of a land-
projects primary consist of expansions from 2-lane to 4-lane expressway use model for this project.
in select locations, with occasional expansion to 6 lanes. The Loop 1604 A variety of land use models have been used in planning research
projects are displayed below in Fig. 4. and practice. These include the California Urban Futures Model, the

Fig. 3. Mobility 2040 roadway projects.

39
L.A. Merlin et al. Transport Policy 69 (2018) 35–48

Fig. 4. Loop 1604 projects.

Human-Induced Land Transformations Project (HILT), the Land Use As an aggregate, zonal land use model, TELUM can mostly be cali-
Evolution and Impact Assessment Model (LEAM), MEPLAN, TRANUS, brated with commonly available US Census data. Land use data must be
and UrbanSim (Clarke et al., 1997; Deal and Pallathucheril, 2008; Hunt analyzed via a GIS system to associate different household and em-
and Simmonds, 1993; Johnston and de la Barra, 2000; Landis and ployment activity types with their land consumption requirements. The
Zhang, 1998; Waddell, 2002). Nevertheless, land use modeling is still underlying causal framework of the TELUM model assumes that newly
unusual among MPOs. A recent survey of 100 MPOs found that just located households are attracted by accessibility to employment, other
39% of MPOs that examined future scenarios had a functional land use similar type households, available land, and zonal-specific factors. New
model available to help construct those scenarios (Brown and Lee, employment is attracted to the availability of local workforce, proxi-
2013). UrbanSim is the land use modeling system that has most likely mity to other employment locations, land availability, and zonal-spe-
had the longest effort to become a readily adaptable tool for MPO cific factors.
analysis (“UrbanSim: Our Story, 2017”). As of 2017, there were 20 out For this case study, the launch year (the year from which the fore-
of more than 340 MPOs that had experience using UrbanSim (personal casts begin (Smith, 2001)) for the land use models was 2010 while the
communication with Paul Waddell). Hopefully, land use models such as base year (the prior year from which data trends are extrapolated) was
UrbanSim will continue to improve in usability and adoption; but as our 2005. For each of these two years, we gathered information on
goal was to demonstrate the possibility of accessibility analysis with the households by income category and employment by industrial grouping
most accessible land use modeling tools available, we opted for TELUM for the 436 Census Tracts in the region. Household data by Census Tract
instead. The concepts illustrated here do not depend upon using TELUM for 2005 and 2010 are from the American Community Survey. Em-
and are completely transferable across land use modeling platforms. ployment data by industry and by Census Tract are from the LEHD
TELUM was developed under the guidance of the Federal Highway Origin-Destination Employment Statistics (LODES) Dataset. These data
Administration to allow middle-range MPOs the capability to conduct are used to calibrate the attractiveness of different features for future
their own land-use modeling in house (New Jersey Institute of growth, including land supply and accessibility to employment and
Technology, 2005). TELUM is a software program that incorporates workforce.
data from Microsoft Excel and Microsoft Access, and produces land use Land consumption is determined by a different model within
forecast data that can be visualized in ArcGIS. A disadvantage of TELUM. This model is calibrated based upon launch-year (2010) land-
TELUM is that it is not integrated with any travel demand model, so use patterns in relation to launch-year employment and population
land use forecasting and travel demand forecasts must be fed iteratively patterns. Land use parcel data were provided by the Alamo Area
back and forth, limiting the level of synchronization between the Metropolitan Planning Organization (AAMPO) with the following land
transportation forecast and the land use forecast. use categories: Residential, Commercial, Industrial, Undevelopable,
TELUM is an aggregate, zonal-based model rather than a micro- Right of Way, and Vacant Developable.
simulation model. Provided zonal data on households and employment The TELUM model uses a single zone-to-zone impedance matrix to
from two time periods and land use consumption data for one time define how employment accessibility influences future residential
period, the TELUM land use model calibrates parameters explaining growth and how accessibility to workers influences future employment
land use change based upon recent trends. Then based upon these ca- growth. We used peak-hour auto-based travel times provided from the
librated parameters, future land use can be forecast via the same zonal AAMPO travel demand model as the source of these impedance ma-
structure up to 6 time increments into the future. TELUM allows for a trices. The travel demand model was run with different transportation
distinction of up to 6 household types and 8 employment types in its plan configurations for future years by the AAMPO at our request. The
zonal forecasts. AAMPO's travel demand model is a traditional four-step travel demand

40
L.A. Merlin et al. Transport Policy 69 (2018) 35–48

Fig. 5. Forecast household shift due to Alamo transportation plan to 2020.

model calibrated to year 2005 data (Alamo Area Metropolitan Planning For equations (1.1) and (1.2)), Ai is accessibility for zone i, Hi is the
Organization, 2011). number of households residing in zone i, Dj is the total number of jobs
The TELUM land use model was used to forecast 2020 land-use located in zone j, tij is the auto-based, peak-hour travel time between
patterns for both the Build and No-Build scenarios for each set of zone i and j, and −0.14 is the impedance coefficient with units in in-
transportation projects evaluated. The model was calibrated based verse minutes.
upon population and employment data from 2005 to 2010, land-use We also examined the changes to mobility for each of the two sce-
consumption data from 2010, and travel impedance data from 2010. narios by examining aggregate travel time for 2020 vehicle flows across
Then two different 2020 travel impedances were fed into the model to the metropolitan region. This was done by summing up all projected
produce two distinct land use forecasts, one for the Build Scenario and zone-to-zone vehicle flows by the peak-hour vehicular travel time be-
one for the No-Build Scenario for each project set. tween all zones for each scenario.
We calculated accessibility to employment by automobile for each
of the two scenarios with a gravity potential measure (See equations T= ∑ Fij ∗tij
i, j (2)
(1.1) and (1.2)). Accessibility to employment was analyzed because it
was the most readily available measure from the land use forecasts, and In equation (2), T is the aggregate travel time in minutes, Fij is the
because accessibility to employment is of concern for promoting eco- flow of vehicles from zone i to zone j over the peak hour and tij is the
nomic opportunity and reducing the length and cost of commutes (Hu, peak-hour travel time from zone i to zone j.
2016; Levinson, 1998). Because TELUM produces population forecasts
by household type and employment forecasts by industry type, a variety 5. Results
of accessibility measures are possible; for example, it is possible to
calculate the accessibility impacts on low-income households sepa- The results section details the implications of the Mobility 2040
rately from high-income households. However, we report only ag- LRTP Highway Projects and the 1604 Loop Projects. For each set of
gregate accessibility impacts here for purposes of brevity. The travel projects, the spatial implications in terms of household locations, em-
cost for the accessibility calculations in this case only accounts for peak- ployment locations, and accessibility shifts is mapped. Then accessi-
hour auto travel times. If the data were available, other types of ac- bility and mobility impacts of these projects is also summarized in
cessibility such as transit accessibility could be calculated. The coeffi- tabular format. The accessibility impacts take into account both
cient of impedance (−0.14) is derived from AAMPO's travel demand changes to travel speeds and land-use patterns, whereas the mobility
model impedance curves (San Antonio-Bexar County Metropolitan impacts only take into account changes to travel speeds. “Speed-Only”
Planning Organization, 2011). Fig. 8.4 of the model documentation accessibility impacts display the impacts to accessibility if no land use
provides the Friction Factors for Home-Based Work travel times. With change had occurred.
the friction factor declining by 50% between 7 and 12 min, resulting in
an impedance coefficient = -ln(0.5)/5 = 0.14.
5.1. Analysis 1: Mobility 2040 transportation plan vs. No-Build
Equations (1.1) and (1.2):

Ai = ∑ Dj f (cij) = ∑ Dj e−0.14∗tij Our first analysis examined the accessibility implications of the
j j (1.1) Mobility 2040 Highway Projects versus a No-Build Scenario. We ex-
amined how mobility and accessibility change between the Build and
ARegion = ∑ Ai Hi/ ∑ Hi No-Build scenarios for the year 2020, which involved a 10-year land-
i i (1.2)
use forecast from launch-year 2010 data. Although 2040 data were

41
L.A. Merlin et al. Transport Policy 69 (2018) 35–48

Fig. 6. Forecast employment shift due to Alamo transportation plan to 2020.

available for land use forecasting, we found that extrapolating the land- accessibility. Although accessibility changes are illustrated via Fig. 7,
use forecast over 30 years produced exaggerated results; on the other this map can be misleading because accessibility gains can be con-
hand, a land-use forecast over less than 10 years might not show en- centrated in areas which are spatially large but low in population. To
ough land-use change for a meaningful accessibility analysis. summarize the region-wide accessibility performance, the number of
people residing in each Census Tract is taken as the weight for that
Census Tract, and the job accessibility score is averaged over all Census
5.1.1. Land use forecast and accessibility shifts
Tracts; this produces the average job accessibility experienced across all
Forecast land-use shifts due to Mobility 2040 are illustrated in Fig. 5
residents of the metropolitan region. Percent differences are reported
and Fig. 6. These shifts illustrate how the Build Scenario differs from
for the Build Scenario in comparison with the No-Build Scenario and
the No-Build Scenario by taking the difference between the two for each
the Year 2010 baseline.
zone. Household and employment shifts are shown with total popula-
The mobility impacts of the Alamo's Mobility 2040 Plan through
tion change rather than percent change. Household gains for the Mo-
2020 are significant, however despite this, its accessibility impacts are
bility 2040 projects are concentrated along the northern metropolitan
negligible. The Build Scenario reduces aggregate travel time by 5.5%,
fringe, in particular the far northeast. Household losses are con-
so travel times are quite improved. In addition, if the associated
centrated on the northern, inner ring suburbs, but also encompass much
transportation projects had no land-use effect–if they only changed
of the central city. Employment gains due to the transportation plan
interzonal speeds–they would increase accessibility by 3.1%. The
cluster in the northwest and northeast, while employment losses are
reason the accessibility improvement is less than the mobility im-
scattered throughout the central city and the northern inner ring.
provement is that accessibility creates greater weights where current
The resultant geography of accessibility changes is illustrated in
households reside, and presumably the speed improvements are pre-
Fig. 7. Accessibility losses are greatest for the central city, but also there
dominantly weighted towards less-populated areas of the region.
are some accessibility losses along the metropolitan fringe. The acces-
However, after also considering the land-use effects of the Mobility
sibility gains occur along the outer-edge suburbs. These accessibility
2040 Plan through the year 2020, the net accessibility benefits are
shifts are largely explained due to shifts in employment location away
−0.2%. There is no discernible benefit to the average households’ ac-
from central areas and towards peripheral areas illustrated in Fig. 6.
cessibility as a result of the Mobility 2040 Plan.
However, this map shows the geography of accessibility change without
accounting for population weights. Aggregate accessibility impacts
weighted based on resident population in each tract are presented in 5.2. Analysis 2: Loop 1604 vs. No loop
Table 3, discussed below.
The second analysis focuses on the accessibility effects of a series of
related transportation projects, a bundle of expansions and improve-
5.1.2. Accessibility and mobility effects
ments to a suburban ring road known as Loop 1604. The existing
Table 3 illustrates mobility and accessibility performance across
transportation plan is examined with all Loop 1604 projects included
four scenarios: Build, No-Build, Speed-Only Effects, and Year 2010.
versus the same transportation plan but with all Loop 1604 projects
Speed-Only Effects show what the impact would be on accessibility if
excluded. The first scenario we will refer to in shorthand as “Loop
the proposed transportation projects had no land-use impacts but only
1604” or “Build” and the second as “Plan Minus Loop 1604.”
speed impacts. Year 2010 mobility and accessibility are included as a
baseline for comparison. The primary mobility effects are reported via
aggregate peak-hour travel time and average peak-hour trip time. The 5.2.1. Land-use forecast and accessibility shifts
primary accessibility effects are reported via population-weighted The Loop 1604 projects as a whole result in more centralized

42
L.A. Merlin et al. Transport Policy 69 (2018) 35–48

Fig. 7. Accessibility shift due to Alamo transportation plan to 2020.

locations for households as illustrated in Fig. 8. Household increases are mobility effects.
found throughout the central city and along the northern inner ring Table 4, like Table 3, examines mobility and accessibility impacts
suburbs. Household losses are found throughout the metropolitan across four scenarios: Build, Plan Minus Loop 1604, Speed-Only Effects,
fringe, both to the north and to the east. Employment changes are more and Year 2010, and compares the Build Scenario's performance as a
centralized (see Fig. 9), with employment gains occurring in a crescent percent change relative to Plan Minus Loop 1604 and Year 2010. The
from the central city to the northern edge. The largest employment mobility effect, as measured via average trip times, is an improvement
losses are in the far northwest, the far north, and the far east of the of 2.5% in comparison with the Plan Minus Loop 1604 Scenario.
metro area. Shifting to the accessibility impacts of Loop 1604 with respect to the
The geographic area that benefits from these shifts in terms of ac- Speed-Only Effects, the improvement to the average household's ac-
cessibility is the central city, as shown in Fig. 10. Although the geo- cessibility is 5.0%. Accessibility benefits exceed mobility benefits be-
graphic area with accessibility losses is much larger, from the per- cause these accessibility calculations weigh such benefits in locations
spective of residential population the area with accessibility gains is where households and jobs are clustered. (see Table 4).
larger in magnitude. Total accessibility impacts, including the effects of land-use changes
in addition to speed effects, raise the accessibility impact of Loop 1604
projects to +9.5% for Build in comparison to Plan Minus Loop 1604.
5.2.2. Comparison of accessibility and mobility effects
Loop 1604 is the opposite kind of case from the Mobility 2040; it has
fairly strong mobility benefits, but its accessibility benefits are much 6. Discussion
higher than these mobility benefits alone. The centralizing (and
proximity-increasing) effect that building the Loop 1604 Projects have The Mobility 2040 Projects demonstrate mobility benefits through
relative to the Plan Minus Loop 1604 Scenario compounds its positive decreased average travel times, however despite this they offer no net

Table 3
Mobility 2040 plan mobility and accessibility impacts through 2020.
Build Scenario No-Build Scenario Speed-Only Effects Year 2010

Mobility Impacts Aggregate Travel Time

Vehicular Travel Time, Minutes 82,098,736 86,908,769 NA 73,189,146
Average Trip Time, Minutes 24.59 26.03 NA 21.92
Percent Diff. No-Build −5.5% NA NA NA
Percent Diff. with 2010 12.2% 18.7% NA NA

Accessibility Impacts Gravity-Weighted Accessibility to All Employment

Minimum Zone 22 21 21 66
Maximum Zone 96,340 96,069 97,071 107,978
Mean Zone 28,105 27,790 28,298 29,330
Standard Deviation 23,335 23,635 23,553 25,382
Population Weighted 24,406 24,446 25,194 27,988
Percent Diff. No-Build −0.2% NA 3.1% NA
Percent Diff. with 2010 −12.8% −12.7% −10.0% NA

43
L.A. Merlin et al. Transport Policy 69 (2018) 35–48

Fig. 8. Household changes due to loop 1604 to 2020.

Fig. 9. Employment changes due to loop 1604 to 2020.

accessibility benefits. A major reason for this is that the Mobility 2040 The Loop 1604 Projects, on the other hand, illustrate accessibility
Highway Projects decentralize both households and employment; after benefits greater than their purported mobility benefits. Nearly half of its
factoring in these land-use shifts, the average household has no greater accessibility benefits are due to induced land-use changes – i.e. a 4.3%
accessibility in 2020 than if no transportation projects were completed increase in accessibility is attributable to land use changes, while a
at all (i.e. the No-Build Scenario). The advantage of faster speeds, which 5.0% increase is due to mobility changes.
are documented in the mobility analysis above, are completely coun- The reason that the Loop 1604 projects are particularly effective for
terbalanced by the land use decentralization induced by the transpor- improving auto-based accessibility is because they offer a triply com-
tation plan. pounded benefit. First, average travel speeds are increased. Second,

44
L.A. Merlin et al. Transport Policy 69 (2018) 35–48

Fig. 10. Accessibility changes due to loop 1604 to 2020.

Table 4
Loop 1604 mobility and accessibility impacts through 2020.
Build Plan Minus Loop 1604 Speed-Only Effects Year 2010

Mobility Impacts
Aggregate Trip Time, Minutes 81,890,310 83,949,731 73,189,146
Average Trip Time, Minutes 24.5 25.1 – 21.9
Percent Diff Baseline −2.5% NA
Percent Diff with 2010 11.9% 14.7% NA

Accessibility Impacts
Minimum 19 26 26 66
Maximum 88,860 85,630 87,130 107,978
Mean 28,380 26,770 27,980 29,330
Standard Deviation 22,949 21,982 22,478 25,382
Population Weighted 25,108 22,939 24,094 27,988
Percent Diff Plan Minus Loop 1604 9.5% NA 5.0% NA
Percent Diff with 2010 −10.3% −18.0% −13.9% NA

these travel speed improvements are focused on areas of concentrated analysis to accessibility analysis will result in a new prioritization
population and employment. And third, the Loop 1604 projects spur across proposed transportation projects. Projects that improve speeds
centralization of population and employment relative to the Plan Minus but encourage decentralization will score more poorly under an ac-
Loop 1604 Scenario. Accessibility benefits of this suite of projects are cessibility-based evaluation system. Projects that capitalize on existing
therefore created by changes to speeds and induced land use shifts as concentrations of population and employment will perform better and
well. be more likely to be selected for implementation. In other words, the
As evidenced here, the mobility-based benefits of major transpor- goal of a shift to accessibility-based project evaluation is to change
tation projects do not necessarily correspond with their accessibility which transportation projects are selected for implementation based
benefits. While we verified that a major set of highway investments upon their anticipated accessibility benefits.
does produce mobility benefits as expected, after considering induced In the US, policy is starting to shift away from mobility-based
land-use change the net effect on average household accessibility to analyses and towards accessibility-based analyses, at least in some se-
employment was virtually zero. On the other hand, we found that a lected cases. The State of California has decided to omit traditional
mobility-based analysis of the Loop 1604 suite of projects would sig- traffic impact analysis from its environmental analysis process, de-
nificantly underestimate its accessibility benefits in comparison with a emphasizing traditional mobility concerns (Governor's Office of
Plan Minus Loop 1604 Scenario. In sum, this paper demonstrates that Planning and Research (CA), 2016). Meanwhile, the State of Virginia
mobility-based analysis alone cannot provide a meaningful indicator of has adopted a new SmartScale project prioritization process that ex-
whether a plan provides transportation benefits to households (as plicitly includes an accessibility component (Virginia Department of
measured by accessibility changes), nor can it indicate the relative size Transportation, 2016). The methods presented here are hoped to ac-
of those benefits. celerate and improve this shift towards the accessibility-based evalua-
The implications of this result are that shifting from mobility tion of transportation projects.

45
L.A. Merlin et al. Transport Policy 69 (2018) 35–48

In this paper we focus on a comparison of mobility and accessibility- consider the trade-offs between model complexity and data hungriness
based performance measures for transport project evaluation. In fact, versus demonstrated model accuracy and the strength of its theoretical
the world of transport project evaluation is a complex one extending basis. Our goal in this case was to focus on minimum barriers to entry
beyond either of these types of measures. Cost-benefit measures are for the adoption of land use modeling as a required part of an integrated
widely used, especially in developing countries and for larger-scale accessibility analysis.
projects. Cost-benefit measures, however, are often criticized because of The aggregate nature of TELUM forecasts supports place-based ac-
the questionability of many of their assumptions, such as their valua- cessibility measures, as opposed to disaggregate person-based measures
tion of time savings, and their inability to account for equity impacts such as utility-based approaches. The utility-based approaches show, on
(Jones et al., 2014; Shi and Zhou, 2012). But more fundamentally for a disaggregate basis, the varying accessibility that even neighboring
the purposes of this study, cost-benefit analysis is most typically a form individuals experience based on their daily opportunities and con-
of mobility analysis, because travel-time savings, rather than accessi- straints and are, for this reason, a more accurate depiction of accessi-
bility gains, form the basis of transport-project benefit. By contrast, bility than the highly aggregate accessibility measures used here. While
Geurs et al. (2010) demonstrate the possibility of evaluating accessi- these metrics can readily be estimated for an existing population, their
bility gains in monetary terms in a fashion that would be analyzable use in prospective planning is limited by the unknown preferences,
within the cost-benefit analysis framework. opportunities, and constraints of a projected future population. Basic
Furthermore, there is a growing trend towards the consideration of population projections themselves are quite technically challenging and
multiple criteria in transportation project evaluation using Multi- disaggregating future populations across multiple demographic and
Criteria Decision Analysis (Avineri et al., 2000; Frohwein et al., 1999; social characteristics as a high level of geographic detail is likely not
Macharis and Bernardini, 2015; Sinha and Labi, 2011). This includes feasible with any reasonable degree of accuracy. For this reason, we
the capability of incorporating both quantitative as well as qualitative believe that the aggregate forecasting and accessibility measurement
criteria (Avineri et al., 2000; Macharis and Bernardini, 2015). This used here are appropriate tools for the transportation and land-use
literature focuses on the integration of numerous economic, environ- planning tasks at hand.
mental, and societal considerations in transportation evaluation, but A second limitation is due to the baseline of comparison. We took
generally employs mobility-based approaches to transportation bene- the existing plan – Mobility 2040 – as the status quo for comparison
fits. As with cost-benefit analysis, multicriteria evaluation can be re- purposes, even though our analysis shows no accessibility benefits due
vised with an accessibility framework for transportation benefits as in to the Mobility 2040 Plan. Our analysis shows that the Loop 1604
the Virginia SmartScale process (Virginia Department of projects provide an accessibility improvement when the rest of the
Transportation, 2017). Mobility 2040 Plan projects are taken as the baseline of comparison.
The more widespread adoption of accessibility-based performance However, since we find no benefit to the Mobility 2040 Plan, a better
analysis faces several obstacles. Some obstacles may be conceptual, comparison might be the Loop 1604 projects versus a true no-build
such a misunderstanding of the accessibility concept or a confusion of scenario, i.e. no new roadway projects for the region. Only then would
the concepts of accessibility and mobility (Boisjoly and El-Geneidy, we be able to discern if the Loop 1604 projects are truly accessibility
2017a; Proffitt et al., 2017). Other obstacles may be political, that ac- enhancing.
cessibility as a planning goal has no natural constituency, whereas there Although we only present average household accessibility impacts
are obvious constituencies for the concerns of traffic congestion and in this paper, the method presented here readily supports different
environmental impacts. Yet another barrier is in relation to oper- types of disaggregate accessibility impacts, including a differentiation
ationalization. In many cases accessibility is understood at the con- of accessibility impacts by household type, accessibility to work as well
ceptual level but there is a lack of means to implement it properly as non-work destinations (i.e. shopping, health care), and accessibility
within decision making (Boisjoly and El-Geneidy, 2017b). This paper by transit as well as by auto modes.
aims to offer a concrete methodology for analyzing the accessibility Future research could improve upon the methods presented here in
impacts of transportation projects to help overcome this obstacle. As several ways. More sophisticated integrated transportation-land use
such, it is one of a growing family of accessibility-based analysis tools or models would improve the accuracy of forecast accessibility impacts.
approaches that are increasing available for practitioners (Papa et al., Sensitivity analysis of forecast future land use is warranted given the
2015). inherent uncertainty of the land use forecasting exercise. More diverse
transportation projects could be analyzed, for example comparing a
7. Limitations proposed highway project with a proposed transit project, while taking
into account the expected land use impacts of each. Equity analyses are
The analyses here are based on a single round of iterative trans- an additional layer that could be added to the evaluation of the ac-
portation and land-use forecasts. That is, a new zone-to-zone travel time cessibility impacts of proposed transportation projects. While the a
matrix is estimated for the Build and No-Build scenarios, and a different priori evaluation of proposed transportation projects is perhaps as old
regional land-use outcome is forecast for each scenario. More iterations as the field of transportation planning itself, the application of an ac-
are possible: the new land-use forecast would alter the zone-to-zone cessibility lens to such evaluations is a relatively recent phenomenon
travel time matrix, and so on. The decision to iterate travel time and and many methodological improvements are possible.
land use just once each was largely a function of the core purpose of this
project: demonstrating manageable tools by which transportation 8. Conclusion
planners in local practice could evaluate the accessibility impacts of
proposed transportation projects. Given lowered barriers to the use of When transportation planners evaluate the transportation-related
integrated transportation-land use models, they could be used to fore- benefits of a proposed project based upon mobility alone, they miss an
cast accessibility impacts, and presumably they would capture more essential part of the dynamic relationship between transportation in-
comprehensively the dynamic interplay between the transportation and frastructure and land-use patterns. Improved transportation infra-
land use systems. In addition, any improvement to the land use fore- structure influences the pattern of new land uses, yet these impacts are
casting enterprise itself would clearly improve the accuracy of the re- rarely accounted for in transportation project analysis. Even though the
sults. The TELUM land use model deployed here is an aggregate model need for accessibility-based transportation analysis has been well es-
and does not involve the detailed simulation of land development tablished, consideration of the land-use impacts of major transportation
markets; therefore, its land use forecasts may be less accurate than projects is routinely ignored and accessibility is either not accounted for
other, more nuanced land use models. Analysts must continually or inadequately estimated as a consequence.

46
L.A. Merlin et al. Transport Policy 69 (2018) 35–48

In order to understand the transportation benefits of a major objectives and indicators in metropolitan transportation plans. In: Transportation
transportation project, an accessibility-based evaluation framework is Research Board 96th Annual Meeting, vol. 55. pp. 38–50. (February). http://doi.org/
10.1016/j.tranpol.2016.12.011.
required. And to evaluate the accessibility benefits of such a project, Boisjoly, G., El-Geneidy, A.M., 2017b. The insider : a planners ’ perspective on accessi-
induced land-use changes must be accounted for, in addition to the bility. In: Transportation Research Board 96th Annual Meeting. Transportation
better-understood travel-time impacts. It is possible – as demonstrated Research Board, pp. 1–19.
Bonavia, M.R., 1936. The Economics of Transport. Nisbet and Co. Ltd, Cambridge, UK.
here – that the decentralizing effects of a transportation project might Brown, C.M., Lee, D.J.-H., 2013. How Does Modeling and Forecasting Support
counterbalance or outweigh its speed-enhancing benefits. Alternatively, Performance-based Planning? Results of an MPO Survey. Citlabs.
if the transportation investment leads to more concentrated land de- Cervero, R., 1996. Paradigm shift: from automobility to accessibilty planning. In:
Working Paper. Institute for Urban and Regional Development, Berkeley.
velopment than might have occurred otherwise, the land-use impacts Chicago Metropolitan Agency for Planning, 2010. Go to 2040 Comprehensive Regional
will magnify the accessibility benefits of the transportation investment. Plan. Chicago Metropolitan Agency for Planning, Chicago, IL.
The practical method we present here illustrates how a land-use model Clarke, K.C., Hoppen, S., Gaydos, L., 1997. A self-modifying cellular automaton model of
historical urbanization in the San Francisco Bay area. Environ. Plann. Plann. Des. 24,
can be integrated into the analysis of a transportation project to as-
247–261.
certain its net accessibility impacts. Cohen, H.S., 1995. Expanding Metropolitan Highways: Implications for Air Quality and
Land-use impacts are by no means marginal in either of the cases we Energy Use (Appendix B: Review of empirical studies of induced traffic). .
analyzed. In the Mobility 2040 case, the land-use impacts completely Curtis, C., 2011. Integrating land use with public transport: the use of a discursive ac-
cessibility tool to inform metropolitan spatial planning in Perth. Transport Rev. 31
neutralize the purported mobility benefits. In the second case, the Loop (2), 179–197. http://doi.org/10.1080/01441647.2010.525330.
1604 Projects, the land-use impacts augment the mobility benefits, ul- Deal, B., Pallathucheril, V., 2008. Simulating Regional Futures: The Land-Use Evolution
timately constituting 45% of the accessibility impacts. The land-use and Impact Assessment Model (LEAM). In: Brail, R.K. (Ed.), Planning Support
Systems for Cities and Regions. Lincoln Institute of Land Policy, Cambridge, MA.
impacts of a major transportation project are too large to be ignored, Department for Transport (UK), 2005. Guidance on Accessibility Planning in Local
and accessibility analysis cannot be replaced by a simple mobility Transport Plans. London.
analysis, which does not account for these land use shifts. Downs, A., 2005. Still Stuck in Traffic : Coping with Peak-hour Traffic Congestion.
Brookings Institution. Washington, D.C. Brookings Institution Press Retrieved from.
The tools we employ in the paper are widely accessible in planning http://www.loc.gov/catdir/toc/ecip0415/2004004190.html.
practice, and there is no technical reason that practicing transportation Fan, Y., Guthrie, A., Teng, R., 2010. Impact of Twin Cities Transitways on Regional Labor
professionals cannot evaluate transportation projects based upon their Market Accessibility: a Transportation Equity Perspective. Transitways Impacts
Research Program Report #5. .
accessibility impacts. This approach can bring accessibility evaluation Frohwein, H.I., Lambert, J.H., Haimes, Y.Y., Schiff, L.A., 1999. Multicriteria framework to
out of the somewhat abstract world of regional scenario planning by aid comparison of roadway improvement projects. J. Transport. Eng. 125 (3),
linking it to everyday decision making and the prioritization of pro- 224–230.
Geurs, K.T., de Bok, M., Zondag, B., 2012. Accessibility benefits of integrating land use
posed transportation projects.
and public transport policy plans in The Netherlands. In: Geurs, K.T., Krizek, K.J.,
Reggiani (Eds.), Accessibility Analysis and Transport Planning, pp. 135–153.
Acknowledgements Geurs, K.T., Van Wee, B., 2004. Accessibility evaluation of land-use and transport stra-
tegies: review and research directions. J. Transport Geogr. 12 (2), 127–140.
Geurs, K.T., Zondag, B., De Jong, G., de Bok, M., 2010. Accessibility Appraisal of Land-
Thanks to the Alamo Area Metropolitan Planning Organization for use/transport Policy Strategies: More than Just Adding up Travel-time Savings.
generously sharing data and even conducting transportation model runs Elsevier. http://doi.org/. https://doi.org/10.1016/j.trd.2010.04.006.
on our behalf. Special thanks to experienced travel demand modeler Gjestland, A., McArthur, D., Osland, L., Thorsen, I., 2012. A bridge over troubled waters:
valuding the accessibility effects of a new bridge. In: Geurs, K., Krizek, K.J. (Eds.),
Zachary Graham for accommodating our modeling and data requests in Accessibility Analysis and Transport Planning2. MPG Books Group, Cheltenham, UK,
the midst of his real work for AAMPO. Also thanks to David Kruse for pp. 173–194.
sharing his implementation of the TELUM model with us. We also ap- Governor’s Office of Planning and Research (CA), 2016. Revised Proposal on Updates to
the CEQA Guidelines on Evaluating Transportation Impacts in CEQA. Sacramento,
preciate the support of the TELUM team, especially Brijesh Singh of the CA. Retrieved from. https://www.opr.ca.gov/docs/Revised_VMT_CEQA_Guidelines_
New Jersey Institute of Technology. Proposal_January_20_2016.pdf.
This work was funded by a grant from the NEXTRANS regional Grengs, J., Levine, J., Shen, Q., Shen, Q., 2010. Intermetropolitan comparison of trans-
portation accessibility: Sorting out mobility and proximity in san Francisco and
university transportation center grant #DTRT07-G-0005 and also sup-
Washington, D.C. J. Plann. Educ. Res. 29 (4), 427–443. http://doi.org/10.1177/
ported by the Dow Sustainability Postdoctoral Fellowship at the 0739456X10363278.
University of Michigan. Gulhan, G., Ceylan, H.H., Baskan, O., Ceylan, H.H., 2014. Using potential accessibility
measure for urban public transportation Planning : a case study of Denizli, Turkey.
Traffic Transport. 26 (2), 129–137. http://doi.org/10.7307/ptt.v26i2.1238.
References Gutiérrez, J., 2001. Location, economic potential and daily accessibility: an analysis of
the accessibility impact of the high-speed line Madrid-Barcelona-French border. J.
Ahlfeldt, G., 2011. If Alonso was right: modeling accessibliity and explaining the re- Transport Geogr. 9 (4), 229–242. http://doi.org/10.1016/S0966-6923(01)00017-5.
sidential land gradient. J. Reg. Sci. 51, 318–338. http://doi.org/10.1111/j.1467- Hensher, D.A., Ellison, R.B., Mulley, C., 2014. Assessing the employment agglomeration
9787.2010.00694.x. and social accessibility impacts of high speed rail in Eastern Australia. Transportation
Alamo Area Metropolitan Planning Organization, 2011. San Antonio/Bexar County 5- 41 (3), 463–493. http://doi.org/10.1007/s11116-013-9480-7.
County Travel Demand Model Documentation. San Antonio, TX. Hu, L., 2016. Job accessibility and employment outcomes: which income groups benefit
Alamo Area Metropolitan Planning Organization, 2015. Mobility 2040. the most? Transportation 1–23. http://doi.org/10.1007/s11116-016-9708-4.
Avineri, E., Prashker, J., Ceder, A., 2000. Transportation projects selection process using Hunt, J.D., Simmonds, D.C., 1993. Theory and application of an integrated land-use and
fuzzy sets theory. Fuzzy Set Syst. 116 (1), 35–47. http://doi.org/10.1016/S0165- transport modeling framework. Environ. Plann. Plann. Des. 20, 221–244.
0114(99)00036-6. Institute of Transportation Engineers, 2006. Transportation Impact Analyses for Site
Banister, D., Button, K., 1993. Transport, the Environment and Sustainable Development. Development: an ITE Recommended Practice (Washington DC). .
Elsevier. Jiang, H., 2016. Accessibility and the Evaluation of Investments on the Beijing. pp.
Bertolini, L., le Clercq, F., Kapoen, L., 2005. Sustainable accessibility: a conceptual fra- 395–408. http://doi.org/10.5198/jtlu.2016.884.
mework to integrate transport and land use plan-making. Two test-applications in Johnston, R.A., de la Barra, T., 2000. Comprehensive regional modeling for long-range
The Netherlands and a reflection on the way forward. Transport Pol. 12 (3), 207–220. planning: linking integrated urban models and geographic information systems.
Boarnet, M.G., Haughwout, A.F., 2000. Do Highways Matter? Evidence and Policy Transport. Res. Pol. Pract. 34, 125–136.
Implications of Highways' Influence on Metropolitan Development. University of Jones, H., Moura, F., Domingos, T., 2014. Transport infrastructure project evaluation
California Transportation Center. using cost-benefit analysis. Procedia - Soc. Behav. Sci. 111, 400–409. http://doi.org/
Bocarejo, J.P., Portilla, I.J., Miguel Velasquez, J., Cruz, M.N., Pena, A., Oviedo, D.R., 10.1016/j.sbspro.2014.01.073.
2014. An innovative transit system and its impact on low income users: the case of Kilby, K., Smith, N., 2012. Accessibility Planning Policy : Evaluation and Future
the Metrocable in Medellin. J. Transport Geogr. 39, 49–61. http://doi.org/10.1016/j. Directions Final Report, (June). pp. 1–35. Retrieved from. https://www.gov.uk/
jtrangeo.2014.06.018. government/uploads/system/uploads/attachment_data/file/3190/accessibility-
Bocarejo, S.J.P., Oviedo, H.D.R., 2012. Transport accessibility and social inequities: a tool planning-evaluation-report.pdf.
for identification of mobility needs and evaluation of transport investments. J. Laird, J., Nash, C., Mackie, P., 2014. Transformational transport infrastructure: cost-
Transport Geogr. 24, 142–154. http://doi.org/10.1016/j.jtrangeo.2011.12.004. benefit analysis challenges. Town Plan. Rev. 85 (6), 709–730.
Boisjoly, G., El-Geneidy, A., 2017a. How to get there? A critical assessment of accessibility Landis, J., Zhang, M., 1998. The second generation of the California urban futures model.
Part 1: model logic and theory. Environ. Plann. Plann. Des. 25, 657–666.

47
L.A. Merlin et al. Transport Policy 69 (2018) 35–48

Levine, J., Merlin, L., Grengs, J., 2017. Project-level accessibility analysis for land-use Sinha, K.C., Labi, S., 2011. Transportation Decision Making: Principles of Project
planning. Transport Pol. 53, 107–119. October 2016. http://doi.org/10.1016/j. Evaluation and Programming. John Wiley & Sons.
tranpol.2016.09.005. Smith, S.K., 2001. State and local population projections : methodology and analysis. In:
Levine, J., Grengs, J., Shen, Q., Shen, Q., 2012. Does accessibility require density or Swanson, D.A., Tayman, J. (Eds.), The Plenum Series on Demographic Methods and
speed? J. Am. Plann. Assoc. 78 (2), 157–172. http://doi.org/10.1080/01944363. Population Analysis. Kluwer Academic/Plenum Publishers, New York Retrieved
2012.677119. from. http://search.lib.unc.edu?R=UNCb3905323.
Levinson, D.M., 1998. Accessibility and the journey to work. J. Transport Geogr. 6 (1), Stepniak, M., Rosik, P., 2013. Accessibility improvement, territorial cohesion and spil-
11–21. lovers: a multidimensional evaluation of two motorway sections in Poland. J.
Lomax, T., 1997. NCHRP 398 Quantifying Congestion. Washington DC. Transport Geogr. 31, 154–163. http://doi.org/10.1016/j.jtrangeo.2013.06.017.
Macharis, C., Bernardini, A., 2015. Reviewing the use of multi-criteria decision analysis Tiwari, G., Jain, D., 2012. Accessibility and safety indicators for all road users: case study
for the evaluation of transport projects: time for a multi-actor approach. Transport Delhi BRT. J. Transport Geogr. 22, 87–95. http://doi.org/10.1016/j.jtrangeo.2011.
Pol. 37, 177–186. http://doi.org/10.1016/j.tranpol.2014.11.002. 11.020.
Metropolitan Planning Commission, 2009. Equity Analysis Report, vol. 34. U.S. Federal Highway Administration, 2017. National Performance Management
Meyer, M.D., 1997. A Toolbox for Alleviating Traffic Congestion and Enhancing Mobility. Measures; Assessing Performance of the National Highway System, Freight
Atlanta, GA. Retrieved from. https://pdfs.semanticscholar.org/91b9/ Movement on the Interstate System, and Congestion Mitigation and Air Quality
88e181902de7823e3ef77956e7c0830d0f57.pdf. Improvement Program. AGENCY, Washington DC.
Papa, E., Silva, C., Te Brömmelstroet, M., Hull, A., 2015. Accessibility Instruments for UrbanSim: Our Story. (2017). Retrieved from http://www.urbansim.com/new-page/.
planning practice: a review of European experiences. J. Transport Land Use 3 (3), Virginia Department of Transportation, 2016. Smart Scale Technical Guide.
1–20. http://doi.org/10.5198/jtlu.2015.585. Virginia Department of Transportation, 2017. Smart Scale: Finding the Right
Proffitt, D., Bartholomew, K., Ewing, R., Miller, H.J., 2017. Accessibility planning in Transportation Projects for Virginia. Retrieved February 13, 2017, from. http://
American metropolitan areas: are We there yet? Urban Stud. 609, 1–26. http://doi. vasmartscale.org/.
org/10.1177/0042098017710122. Waddell, P., 2002. UrbanSim - modeling urban development for land use, transportation,
Puget Sound Regional Council, 2008. Vision 2040 Draft Environmental Impact Statement. and environmental planning. J. Am. Plann. Assoc. 68, 297–314.
Puget Sound Regional Council, Seattle, WA. Weiner, E., 2012. Urban Transportation Planning in the United States: History, Policy,
Schrank, D., Eisel, B., Lomax, T., Bak, J., 2015. Urban Mobility Report, vol. 81 Inrix and Practice. Springer Science & Business Media.
(August). http://doi.org/DTRT06-G-0044. Willson, R., 2001. Assessing communicative rationality as a transportation planning
Shi, J., Zhou, N., 2012. A quantitative transportation project investment evaluation ap- paradigm. Transportation 28 (1), 1–31.
proach with both equity and efficiency aspects. Res. Transport. Econ. 36 (1), 93–100. Zhao, Y., Kockelman, K.M., 2002. The propagation of uncertainty through travel demand
http://doi.org/10.1016/j.retrec.2012.03.002. models: an exploratory analysis. Ann. Reg. Sci. 36 (1), 145–163.

48