Practices for Evaluating the Economic Impacts and Benefits of Transit (2017)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

18 chapter three LocaL appLication of anaLysis Methods overview Although the preceding chapter focused on national studies, this chapter reviews the range of local analyses studies and the methods they employ to assess the effects of specific transit investments, projects, and programs. It is organized to distinguish the analysis methods applied to each of the following four types of studies: 1. Existing Transit System: Current Contribution to the Economy. These studies are notable because they have no specific time dimension and no alternative scenarios. They merely describe the current economic role of transit, based on transit agency information and the application of a regional input-output (I-O) model for a specified study region. 2. Future Scenarios: Economic Impact of Alternatives (Ex-Ante/Predictive Analysis). These studies require a comparison of alternative futures (e.g., either build versus no build, or alter- native solutions to a given problem) and their associated travel characteristics (e.g., changes in travel time, cost, safety, and emissions) as calculated through a travel model or engineer- ing analysis. They also require a time frame for predicted impacts to unfold and evolve in a specified study region. The studies typically require a dynamic regional economic forecasting model to predict effects of alternative scenarios, which can then be compared. 3. Previously Completed Investments: (Ex-Post/Case Study Analysis). These studies require a comparison of pre-project and post-project economic conditions within a specified study area, based on either direct observation or past records. These studies typically also require statis- tical analysis and/or interviews to control for underlying economic trends and other factors affecting observed changes over the pre/post period. 4. Analysis of Societal Benefits. These studies require comparison of scenarios and their associated travel characteristics (as report #2). They then require spreadsheet-type tools to calculate effects on various incidence factors (e.g., fuel use, accidents, emissions, and time delays) and then apply unit valuations to those factors to generate their social welfare valuation. Some studies itemize societal benefits, whereas others place them in the context of a benefit–cost computation. For each type of study, there is a discussion of the associated terminology, overall approach or perspective adopted, analysis method applied, and the interpretation and use of results. Individual local studies are then compared in terms of their use of various data inputs, coverage of various eco- nomic impact output measures, and models employed. existing transit systeM: current contribution to the econoMy terminology These reports describe how local jobs and income are generated by the operation of a transit agency. Different terminology is used to describe these effects, including: • Impact of transit spending on the economy, • Role or significance of transit in the economy, • Contribution of transit to activity in the economy, and • Position of transit as a business enterprise.

19 approach These studies calculate how an existing transit agency currently supports jobs and income in the economy of a specified region. They show how the transit agency operating and capital budgets gen- erate (1) jobs and wages for the agency’s own workers (direct effects); (2) purchases from vendors who supply equipment, materials, and services, thus supporting jobs and wages at vendor companies (indirect effects); and (3) the re-spending of worker wages on consumer purchases, which support additional jobs and wages in retail and service sectors of the economy (induced effects). Some studies also note the extent to which commuters depend on transit to get to work, as that indicates the extent of transit-dependent jobs and wages in the region (dependency effects). A distinguishing aspect of these studies is that they are intended to describe existing conditions. Methodology These studies are intended to portray current conditions, using the most recent available data. They involve two steps. First, transit agency payroll and budget information (and sometimes data on vendors) is assembled as inputs to drive a region-specific I-O model. Second, the I-O model is then applied to calculate broader “economic multiplier” effects on jobs and income. The I-O model works by tracing how the direct inputs lead to broader supplier and income re-spending effects and calculating the extent to which supported businesses are likely to be located in the study region. There are two kinds of I-O models. The simple ones (e.g., RIMS-II) merely calculate bottom line numbers of jobs and wages, which are nearly always much larger than the agency input numbers. The more sophisticated systems (e.g., IMPLAN) can also show the mix of local area industries in which jobs and income are generated, the average wages of those jobs, and broader social accounting impacts including effects on public tax revenues. interpretation and use This type of study has two related uses for public information dissemination. One use is to show how the transit agency serves as a generator of jobs and income in the community. The second is to show how public subsidy funds come back to recirculate in the local community and end up providing jobs and income across a wide range of business activities. Because transit operations are labor-intensive, they can provide more local jobs and income than other options (such as reducing taxes so that residents can spend more money on foreign-made appliances). Some reports refer to the economic impacts of transit spending as an economic stimulative effect. Spending impact studies are subject to criticism when they are misused by either over-zealous public relations staff or under-informed analysts, who erroneously portray spending impacts as indi- cators of the “benefit” of having transit service or the “payback” from investing in public transit. Critics note that public spending on nearly any subject will also have some indirect and induced effects (leading to total business output or value-added numbers larger than the initial spending), and that these spending effects are not indicators of the actual value of resulting transit services. Authors and supporters of these economic impact studies have noted that their intended purpose was never to do BCA, but rather to simply show how transit agency spending percolates through the economy. Some reports straddle both worlds—addressing both (1) analysis of the spending effect, typically labelled as an economic or expenditure impact; and (2) analysis of mobility or transportation effi- ciency benefits. Although those reports do appropriately distinguish between the two, the inclusion of the two types of analysis can lead to some confusion as to whether the spending impacts are also to be interpreted as a benefit for local residents. The broader issue of transportation benefits is covered later in this synthesis report. reviewed studies Six studies are reviewed that assessed the current economic impact of transit agency spending. The first three focus primarily on spending effects. The remaining three cover broader societal benefits;

20 however, each also includes a section that analyzes spending impacts. These studies are compared in two tables and then individually discussed in terms of notable aspects of the study analysis methods and terminology used. Table 3 summarizes the types of study areas, transit modes, and economic models used. Four rep- resent urban transit and two rural transit. Each relies on static I-O models to calculate broader impacts on the regional economy. Three also use models to calculate tax effects. Table 4 summarizes the analysis inputs and outputs. It shows that all of these studies utilize information on current agency spending (both capital investment spending and operations and main- tenance spending), and then apply the I-O model to portray impacts in terms of regional jobs and regional income (which may be wages or value added). Some of the studies also calculate effects on output (gross business revenues) and/or tax revenues. The Dallas Area Rapid Transit (DART) study was conducted by the University of North Texas Center for Economic Development and Research (Clower et al. 2014). It focused only on (capital and operations) spending impacts. It utilized an IMPLAN I-O model to separately assess the economic impacts of current capital spending and operations spending. The model results were used to show that transit-related jobs, income, and tax revenue effects are significant in magnitude. The study also reviewed transit budget data over 5 years to show that these impacts are maintained through both recession and recovery periods of the economy. The Philadelphia Southeastern Pennsylvania Transportation Authority (SEPTA) study (Economy League and Econsult 2013) was conducted by the Economy League of Greater Philadelphia and Econsult Solutions. It focused on current spending impacts, utilizing a RIMS-II I-O model, but then expanding the financial analysis by applying a more detailed fiscal impact model. The latter was *Wages also represent worker income. GDP is Gross Domestic Product; also referred to as Gross Regional Product and is essentially the same as value added (net business income). Business output represents gross business revenues. Note that these measures are not additive. After all, wage income and tax revenues are paid out of GDP (value added), which is itself a portion of business output. Source: Review of local analysis studies. Input Data Analysis Outcome (results)* Study Capital spending O & M spending Rider survey Jobs Wages* GDP (V.A.)* Out- put* Tax Rev.* Dallas: DART Spending Impact X X X X X X Philadelphia: SEPTA’s Value X X X X X X Atlanta: MARTA Econ. Impact X X X X X X X South Dakota: Public Transit X X X X X X X Tennessee: Rural Transit X X X X X Anchorage, Alaska X X X X TABLE 4 ECONOMIC IMPACT OF TRANSIT SPENDING: INPUT DATA AND OUTPUT RESULTS Study Mode and Study Area Type Economic Model Dallas: DART Spending Impact Urban bus and rail system IMPLAN I-O Philadelphia: SEPTA’s Value Urban bus and rail system RIMS-II I-O, ESI fiscal model Atlanta: MARTA Economic Impact Urban bus and rail system IMPLAN I-O/TREDIS South Dakota: Public Transit Rural bus systems IMPLAN I-O model Tennessee: Rural Transit Rural bus systems IMPLAN I-O model Anchorage, Alaska Urban bus system University of Alaska I-O Source: Review of local analysis studies. MARTA = Metropolitan Atlanta Rapid Transit Authority. DART = Dallas Area Rapid Transit. SEPTA = Southeastern Pennsylvania Transportation Authority. TABLE 3 ECONOMIC IMPACT OF TRANSIT SPENDING: SETTING MODES AND MODELS

21 used to also assess SEPTA impacts on property values and tax revenues (compared with a scenario in which SEPTA service is not adequately funded). Results were calculated for both the southeast Pennsylvania region and statewide. The Metropolitan Atlanta Rapid Transit Agency (MARTA) study was conducted by the Carol Vinson Institute at the University of Georgia (Clark et al. 2012). It also focused on current spending impacts (which are referred to as MARTA’s “economic contribution”). Notably, it drew on informa- tion regarding home locations of transit workers and transit riders to provide more spatial detail con- cerning impacts for the Atlanta region and for the rest of the state of Georgia. It was notable for detail concerning the mix of industries affected in the local economy. This study also included a separate analysis of mobility impacts, based on rider surveys, to assess the extent of transit-reliant jobs and costs that would be incurred if transit service was not available. The study used the IMPLAN I-O for spending effects, using TREDIS for assessing broader travel cost impacts. The South Dakota DOT study was conducted by HDR Decision Economics. It explicitly covered current spending impacts (which are referred to as “the contribution of public transit to the state economy”), as well as “benefits to society” (which include user time and cost savings and emissions benefits). The study used an IMPLAN model to calculate effects of capital and operations and main- tenance (O&M) spending, as well as household effects of transit fares (HDR Decision Economic 2011). The latter effect was an element of the calculation of user cost savings that transit riders enjoy over car reliance alternatives; the matter of user-benefit calculation is covered later in this synthesis. The Tennessee DOT study of rural transit was conducted by Oak Ridge National Laboratory (Southworth et al. 2005). It covered both current spending impacts (which it labelled as “expenditure- based value added”) and transportation “efficiency-based benefits.” For spending impacts, it applied the IMPLAN model to show the broader economic effects (value added and job) of rural transit spending, at both local and statewide levels. For efficiency benefits, it calculated the value and con- sequences of transit cost savings to rural families—a form of analysis that is discussed later on in the context of user-benefit studies. The Anchorage transit study was conducted by the Institute of Social and Economic Research at the University of Alaska (Goldsmith et al. 2006). It covered spending impacts in one chapter (“Transit as an Economic Enterprise”) of a broader report on user, social, and community benefits. It presented information on capital and O&M budgets, which were input to the statewide economic model of the University of Alaska’s Institute for Social and Economic Research Input-Output model for the state. The job and wage impact results were referred to as the “economic significance” of transit spending. The text noted that the actual impact of eliminating transit spending would be less than the economic significance since, “if the transit system disappeared . . . most of the [local] money currently spent on transit would be redistributed to other purchases within the economy.” future transit scenarios: predicted econoMic iMpact of aLternatives terminology The term “economic impact” is most widely used for studies that forecast the expected effect of a new or expanded transit facility or service on the growth of the economy of its service area in some future year. These studies are referred to as “ex ante” evaluations because the analysis is done before a proj- ect is built (in contrast to an “ex post” evaluation, which is conducted after the project is finished.) approach These studies estimate the expected level of economic activity (jobs and income) in a defined region, under alternative future scenarios. At least one scenario must portray results in which a proposed action is taken to build or invest in a transportation improvement, and it is then compared with an alternative scenario in which that proposed action is not taken. The latter scenario is commonly viewed as a base case, no build, status quo, or default funding scenario.

22 A distinguishing aspect of these studies is that they are scenario-based analyses—intended to portray the difference that a proposed project or investment would make in expected future economic conditions. The future scenario is typically portrayed for a designated year after the project is com- pleted or a target year selected for long-range planning purposes. As these studies require some form of regional economic model, they tend to forecast scenario impacts at the community, metro, or state level of geography, although they can be and sometimes are done for smaller study areas. Methodology These studies calculate the effects of a proposed transportation scenario on changes in household and business income flows, spending patterns, and productivity levels, and consequently their effects on the future economic growth in terms of jobs and income generated in the region. To accomplish this, three steps are required: a. Estimation of scenario effects on changes in transportation conditions in terms of trips, vehi- cles, travel times, operating costs, safety, reliability, zonal accessibility, etc. This is normally done by means of a travel forecasting model; however, in some cases it can be approximated through expert engineering estimates. b. Translation of step “a” outcomes into changes in direct costs and revenues for various classes of users and non-users. This requires the allocation of direct impacts to various classes of trip purpose and vehicle type, and their assignment to various elements of the economy. Those allocations provide a basis for calculating the direct value to users that is associated with travel time, cost, safety, reliability, and accessibility changes. c. Translation of step “b” outcomes into changes in regional productivity and economic growth for various sectors of the economy in designated areas, over time. This requires some form of economic impact model to calculate effects of business cost savings on demand for products (output growth) and downstream effects on supply, demand, and prices for workers and prod- ucts of other industries. There are two approaches for addressing steps “b” and “c.” The most common approach is to address both steps by using a dynamic regional economic model system that is made for transporta- tion impact analysis (e.g., REMI-TranSight and TREDIS). Both of these models respond to changes in transportation costs and accessibility factors. An equivalent approach can be constructed by apply- ing a spreadsheet to calculate direct cost savings effects for step “b” and using it to drive a regional CGE (computable general equilibrium) model for step “c.” An alternative approach is to calculate direct effects from step “b” and express them in terms of generalized cost savings or accessibility change, and then apply statistically derived coefficients or elasticities to convert them into productivity and associated GDP growth effects. The “economic rent” concept studies works in this way by applying coefficients to translate generalized cost savings into effects on jobs, income, and property values. The “wider economic impact” concept used in the U.K. studies also works in this way, using elasticities to translate changes in accessibility (effective density) into GDP growth effects. interpretation and use Because of the significant effort involved in conducting this type of study, it is most commonly used for major investment projects. There are four possible motivations for doing this type of study. First, it is often used to assess the potential effectiveness of a proposed major investment project when a major motivation for that project is to enhance the economic development of a depressed commu- nity or region. Second, it is used when required for an environmental impact study. Third, it is also used by some state DOTs (e.g., North Carolina and Ohio) to generate economic impact measures for the prioritization of transit and highway projects. Fourth, it is sometimes used to inform public discussion when there are perceived tradeoffs between project cost, environmental and community impacts, and economic impact.

23 reviewed studies Seventeen studies that assess proposed programs and projects are reviewed. These studies are com- pared in two tables and then individually discussed in terms of notable aspects of the study analysis methods and terminology used. Table 5 summarizes the types of study areas, transit modes, and economic models used. All of these studies focused on expected future impacts of a proposed but not yet built project (or multi-project plan), which represents expansion of an urban transit system. The first three studies focus only on the short-term effect of adding an infusion of spending on construction and operations. Accordingly, these three studies use static I-O models that are most appropriate when there are only spending effects and no change in cost factors. The remaining studies distinguish between the short-term construction effects on the growth of jobs and income in the region and longer-term effects on economic growth that are enabled by reducing travel costs and increasing accessibility. Accordingly, these latter projects rely on a dynamic economic simulation model or a statistical model—both methods that are sensitive to changes over time in cost or access metrics. Table 6 summarizes the analysis inputs and outputs. In all cases, short-term construction effects of construction spending are modeled. In most cases, the studies also account for longer-term direct effects on traveler time and expense, which affects business operating costs and household spending patterns. Most of the studies also account for spatial effects by directly considering: (1) localized effects on encouraging the growth of specific business centers and/or (2) regional effects on market scale and agglomeration economies that increase productivity. Some of these studies also captured the economic impact of enhancing access to airports, rail stations, and evolving technology business clusters. The Los Angeles study was conducted by the Economic Policy Institute (Cooper 2012). It assessed economic impacts associated with one component of Los Angeles’s 30/10 Initiative, a long-term plan that will accelerate the construction of several Metro expansion projects. The spending considered *HR = heavy rail transit, LRT = light rail transit, CR = commuter rail, BRT = bus rapid transit. Source: Review of local analysis studies. Study Mode and Study Area Type* Economic Model Analysis of Spending Effects Urban LRT RIMS II (I-O) Regional bus, LRT “proprietary” I-O Urban BRT IMPLAN I-O Analysis of Spending and Mobility Effects Multimodal downtown station TREDIS Urban HR line TREDIS Suburban LRT line Economic rent model Regional LRT system TREDIS Regional BRT, CR, bus REMI Policy Insight Regional CR line TREDIS Regional CR line TREDIS Regional bus, BRT, LRT REMI TranSight Regional LRT line TREDIS Urban BRT/LRT Line Ontario Model Regional CR TREDIS Urban HR line Statistical (density/growth) Urban BRT line TREDIS Los Angeles Light Rail Vancouver Transit Plan (Canada) Richmond BRT Atlanta Green Line Chicago Transit Benefits Maryland Purple Line Minneapolis–St. Paul: Itasca Study Hartford Capitol Region NH Capital Corridor Rail South Coast Commuter Rail West Florida Transit Needs Durham Transit (Canada) Toronto Hamilton Line (Canada) Toronto GO Electrification (Canada) London Crossrail (UK) Sydney BRT (Australia) APTA High Tech Clusters Regional bus, BRT, LRT Statistical (density/growth) TABLE 5 ECONOMIC IMPACT OF PROPOSED TRANSIT SCENARIOS: SETTING, MODES, AND MODELS

24 in this brief is related to Metro’s commitment to purchase 235 railcars to expand transit capacity throughout the region. The Vancouver (British Columbia) study was conducted for TransLink, the regional transporta- tion agency (InterVISTAS 2015). It analyzed the economic impact of a proposed long-range trans- portation and transit plan, by comparing a base case funding scenario with an expanded scenario for greater investment in new and expanded BRT and rail transit lines. The scenarios were defined in terms of spending patterns over ten years and analyzed in terms of jobs, worker income, and GDP over 30 years. Impacts were analyzed for Metro Vancouver, the rest of the province of British Colum- bia, and the rest of Canada. The Richmond study was conducted for the city of Richmond and Henrico County (Chmura Economics 2012). It assessed job impacts associated with construction and operation of a proposed BRT line. Total spending and jobs were summed over a 2-year construction period; other economic indicators were not reported. (The study also covered user benefits, property value appreciation, and related tax revenues.) The Atlanta study was conducted by a consortium of three firms for Central Atlanta Progress, a civic organization supporting downtown Atlanta (Bleakly Advisory Group et al. 2012). The study estimated the effects of developing a proposed Multi-Modal Passenger Terminal (MMPT) on spur- ring investment in land development and expanding economic activity in the central Atlanta Gulch district and broader Green Line area. The report includes both a study of MMPT impacts on local development in the surrounding area, and broader economic impacts of the MMPT, broader Green Line Plan, and other regional transit investments serving the MMPT. The Chicago transit study was conducted by EDR Group for Chicago Metropolis 2020, a civic organization (“Chicago Metropolis 2020” 2007). It evaluated effects of new transit spending pro- *Note on inputs: “Land devel.” refers to investment in the development of business clusters; “Mkt. access” refers to effects on inter-zonal accessibility and affecting the scale or density of labor markets and business product markets. **Note on outputs: Wages also represent personal income. GDP is Gross Domestic Product, also referred to as Gross Regional Product and is essentially the same as value added (net business income). Business Output represents gross business revenues. Note that these measures are not additive. After all, wage income and tax revenues are paid out of GRP (value added), which is itself a portion of business output. (a) Tax revenue impacts were calculated, but they were associated with a separate analysis of property value growth. Source: Review of local analysis studies. Input Data* Output Results** Study Constr./ spend. Travel time + cost Land devel. Mkt. access Jobs Wages GDP (V.A.) Out- put Tax rev. Analysis of Spending Effects X — — — X X — X — X — — — X X X — — X — — — X — — — (a) Analysis of Spending and Mobility Effects X X X X X — X X X X X X X X X X X — X X — X X X — — X X X — — X — — X — X X — — X X X X X X X — X X X — X — X X X X X X — X — X X X X X X — — X X — X X X X X — X X — — X X X — — X X — X X X X — X X X X X X X X — X X X — X X X X — — Los Angeles Railcar Purchase Vancouver Mayor’s Transit Plan Richmond BRT Atlanta Green Line Chicago Transit Benefits Maryland Purple Line Mpls.–St. Paul: Itasca Study Hartford Capitol Region NH Capital Corridor Rail South Coast Commuter Rail W. Florida Transit Needs Durham Transit (Canada) Toronto Hamilton (Canada) Toronto GO Electr. (Canada) London Crossrail (UK) Sydney BRT (Australia) APTA High Tech Clusters — X — X X X X X — Electr. = Electrification. TABLE 6 ECONOMIC IMPACT OF PROPOSED TRANSIT PROjECTS: INPUT DATA AND OUTPUT RESULTS

25 posed by the Chicago Regional Transit Authority. The analysis considers four alternative scenarios: a base case that includes no new funding for transit; a maintain case that includes funding needed to provide existing service and maintain existing ridership levels; an expansion in service quality, frequency, and geographic coverage; and an “expand and plan” scenario that pairs expansion with land-use practices that encourage transit-oriented development. The Maryland Purple Line study was conducted for the Maryland Transit Administration (TEMS 2015). It assessed the construction impacts of a proposed light rail line that would connect several suburban communities using RIMS-II. It then utilized the Economic Rent Model to calculate gener- alized cost savings associated with affected travel zones, and applied statistically derived coefficients to portray productivity value and its capitalization in terms of job, income, and residential property values in those zones. The Minneapolis–St. Paul study was conducted for the Itasca Project, a business-led civic group (Cambridge Systematics 2012). The study was designed to explore the regional economic impacts of accelerating or expanding transit line development in the region. The study included analysis of both wider economic impacts associated with expanded labor market access and societal BCA (the latter analysis is discussed later in this chapter). The Hartford study was conducted by the Connecticut Center for Economic Analysis at the Uni- versity of Connecticut, for the Capital Region Council of Governments (Carstensen et al. 2001). It assessed the economic growth effects of a proposed expansion of regional transit, including new BRT and commuter rail service, as well as expansion of existing bus services. The study estimated effects of travel time and cost savings, reduced congestion, and the “amenity value” (regional attrac- tiveness) effect of pollution, safety, and household cost changes. Impacts were assessed at a local, regional, and state level. The New Hampshire Commuter Rail study was conducted for the New Hampshire Rail Transit Authority (Economic Development Research Group and TranSystems 2010). It assessed the eco- nomic growth effects of initiating the proposed Capital Corridor commuter rail service on exist- ing tracks between Boston, Massachusetts, and Concord, New Hampshire. In this study, economic impacts are estimated for a build scenario and no-build scenario. The South Coast Commuter Rail study was conducted by a consortium of planning and consult- ing firms for Massachusetts DOT (Goody Clancy et al. 2009). It assessed the economic development growth that would be associated with the restoration of commuter rail service to an economically depressed region of Massachusetts. The plan also involved initiatives to encourage station area development and the development of satellite clusters of high tech industries outside of the Boston– Cambridge center. The study assessed effects on labor markets for high tech industries and forecast economic growth impacts for various portions of the rail corridor. The West Florida study was conducted for the West Central Florida Chairs Coordinating Com- mittee, representing transportation planning organizations in eight counties (Cambridge Systemat- ics 2005). The study provided both economic impact and BCA for the Tampa region’s long-range transportation plan, compared with a smaller set of projects that could be implemented at current funding levels. It included bus, BRT, and light rail transit (LRT) elements, and their effects on users and broader economic productivity. Besides the standard economic growth impacts predicted by the economic model, it also estimated impacts on land values in station areas. The Durham (Ontario) study was undertaken by a consortium of planning and consulting firms for the Municipality of Durham, Ontario (HDR Decision Economics et al. 2010). The study included both economic impact and BCA, comparing alternative scenarios involving BRT and LRT investments along with highway high-occupancy vehicle lanes. The EIA covered the impacts of reducing conges- tion reduction (affecting travel times, costs, and reliability) and business market access. The Toronto Hamilton B-Line Rapid Transit project was conducted by a consortium of three con- sulting firms for MetroLinx (Steer Davies Gleave et al. 2010). The study evaluated a proposed new

26 rapid transit line for the MetroLinx transit system, to connect McMaster University with downtown Hamilton, Ontario. Three options have been developed for the B-Line corridor: BRT, LRT, and LRT developed in phases. All three options are evaluated for economic impacts, costs, and benefits. The Toronto GO Electrification study was conducted by a consortium of consulting firms for MetroLinx (Delcan Arup et al. 2010). The study evaluated the economic impacts of electrifying GO Transit Commuter Rail Trains under six different scenarios. It examined impacts on the provin- cial economy associated with changes in travel times, reliability, highway congestion, safety, and emissions. The London study (England) was conducted for Crossrail, a subsidiary of London’s transport agency Crossrail study (Colin Buchanan and Volterra Partners Limited 2007). This economic ben- efits study forecast the impacts of the proposed Crossrail transit line on the economy of the London region. A key factor affecting employment and GDP growth was that Crossrail would enable con- tinued employment growth in Central London, whereas failure to build it would lead to congestion growth that would represent a capacity constraint. The study also calculated changes in value added and earnings per worker, based on changes in worker productivity and economic agglomeration effects. The expected impacts on national tax revenue were also calculated. There was also a separate BCA study of Crossrail, which is discussed later in this chapter. The Sydney BRT study was sponsored by Transport for New South Wales, the state transport agency (Institute of Transport and Logistics Studies, University of Sydney, and Economic Develop- ment Research Group 2012). The study assessed impacts of a proposed BRT line connecting down- town with a new high tech cluster and growth center north of the city. Impacts were examined in terms of economic development for the entire metropolitan area, based on business cost savings and agglomeration benefits (of enhanced accessibility). There was also a related BCA study (discussed later). Although the BCA results were borderline, the economic impact study (EIS) strengthened the strategic case for the project, which is now moving forward. The APTA Cluster studies assessed the impact of investing to expand future transit service for 11 high-growth employment clusters that are focused on nationally important technology industries (Economic Development Research Group 2014, 2016). As with the Crossrail study, the economic impact was calculated by showing the growth constraint imposed by existing transportation infra- structure and labor market loss threatened by future road congestion growth. These studies also showed the agglomeration benefits for technology clusters and the additional effect of future transit expansion to enable continued development of these clusters. previousLy coMpLeted investMents (ex-post analysis) These studies observe the actual results of projects. They require a comparison of pre-project and post-project economic conditions within a specified study area, based on either direct observation or past records. These studies typically also require statistical analysis and/or interviews to control for underlying economic trends and other factors affecting observed changes over the pre-/post-period. They are referred to as ex-post evaluations because the analysis is done after a project is completed (in contrast to ex-ante evaluations, which are conducted before a project is started.) terminology The term economic impact is also sometimes used to describe changes that have been observed in the level of economic activity occurring in an area over time, which can be attributed to the opening of a new or expanded transit facility or service. overview These studies measure the level of economic activity occurring in a defined area at times before and after the completion of a transit project or the opening of a transit service. Pre-/post-data are used to

27 calculate a change over time, and steps are then taken to determine attribution of credit; that is, the extent to which the transit facility or service can be credited as causing the change. The most direct economic measures of change are jobs, building investment, and business sales; however, sometimes measures of property value, property sales, and/or property tax revenue are used as indicators of economic change. Because these studies require direct observation or the assembly of local data, they tend to be most commonly defined to assess neighborhood and community scale impacts of specific transit lines and stations. The approach can and sometimes is also applied to assess broader scale (regional) impacts, but the problems of isolating transit effects from other urban and regional development factors becomes more problematic as the study area becomes larger. Methodology These studies require access to public or proprietary databases that provide measures of jobs, are jobs, building construction permits, building square footage, business sales, property values, prop- erty sales, and/or property tax revenues. The data must exist on an annual basis for a period of time that includes a year or two before the project was started, and at least 5 to 10 years after the project was completed. The information must also exist at a level of spatial detail that is sufficiently fine to capture changes resulting from the transit project. The observed changes are referred to as “gross change.” A second step is to also assemble a “control” measurement, which means the same data measures but for a broader surrounding area (e.g., rest of a city or a state), or a similar study area where no such change in transit service occurred. A comparison of the study area change with the control area change yields a net change. If the net change is greater than zero, then a third step is usually done; conduct interviews with public planners, analysts, and private-sector organizations to determine “attribution of credit”; that is, the extent to which the transit project can be credited with the observed change in economic activity. interpretation and use This type of historical EIS has several different uses, as noted in the TCRP Report 186 (EDR Group 2016): • To confirm the justification for current and past investments by determining the extent to which a plan or policy is achieving intended effects and hence is worthy of continued funding and operation; • To provide insight for planners (lessons learned) regarding factors and processes that affect project outcomes and need to be considered in future project development and implementation; • To validate analysis methods by determining the accuracy of current analysis methods used to predict costs and/or benefits, and to enable improvement in future prediction methods; and • To assemble data for subsequent statistical analysis and market research on the relationships of transportation investment, land development, and economic development. reviewed studies Seven studies that analyzed historical data to assess the impact of past transit investments were reviewed. These studies are compared in two tables and then individually discussed in terms of notable aspects of the study analysis methods and terminology used. Table 7 summarizes the types of study areas, transit modes, and economic methods used. All seven studies observed changes in economic conditions covering time periods before and after completion of new light rail, heavy rail, or BRT stations. Five of the seven studies apply statistical (regression) analysis to compare changes occurring over time, relative to a control area where no new transit investments were made. The other two are collections of case studies where information was provided on pre-project and post-completion points in time, for both the project area and broader surrounding regions. Table 8 summarizes the analysis inputs and outputs. It shows that all of the studies include before and after economic data for the study areas. In addition, the two case study collections and one of

28 the statistical studies also include local interviews to better ascertain the attribution of credit for observed impacts. The observed outcome (change) data vary—five studies track land value change, five employment change, three land-use change, three population change, and two public revenue or cost change. An academic paper authored by an Arizona State University faculty member examined the impact of light rail on property values (Golub et al. 2011). It employed a hedonic price (regression) model that incorporates variables related to individual properties, neighborhoods, and transit accessibility. The Dallas study is an academic paper authored by staff of the Center for Economic Development and Research at the University of North Texas (Clower et al. 2014). It focused on the DART light rail system, and examined changes over time in land use, land development, and office space rents around stations. The Atlanta study is an academic journal article written by University of Kentucky faculty (Bollinger and Ihlanfeldt 1997). It applied econometric analysis to estimate the impact that MARTA rail rapid transit stations have on surrounding population and employment in the Atlanta metro area. The Small Cities study is an academic journal article by Ball State University faculty (Faulk and Hicks 2010). The study used regression analysis to estimate the impact bus service in small cities and rural areas has on the following demographic measures: growth in transfer payments, income growth, employment growth, and population growth. The National BRT study is a report prepared by University of Utah faculty for the National Insti- tute of Transportation and Communities (NITC) (Ganning and Nelson 2015). The study used the Census Longitudinal Employment Housing Dynamics database and a CoStar real estate database to examine changes in jobs, population, households, and office rents in the vicinity of BRT stations in 10 cities, over a 5-year period. Regression analysis controlled for changes in surrounding areas over the period was used. Source: Review of local analysis studies. Study Input Data Source Outcome Measures Pre-/post- statistics Inter- views Real estate value Land devel. or use Employ- ment Popu- lation Public revenue or cost Phoenix Light Rail X — X — — — — Dallas Light Rail X X X X — — X Atlanta Study X — — — X X — Small Cities Study X — — — X X X National BRT Study X — X — X X — SHRP 2 Case Studies X X X X X — — TCRP H-50 Case Studies X X X X X — — TABLE 8 CASE STUDIES OF PREVIOUSLy BUILT PROjECTS: INPUT DATA AND OUTPUT RESULTS *HR = heavy rail transit, LRT = light rail transit, BRT = bus rapid transit. Source: Review of local analysis studies. Study Mode and Study Area Type* Economic Analysis Phoenix Light Rail Urban LRT station areas Time series regression analysis Dallas Light Rail Urban LRT station areas Time series regression analysis Atlanta Study Urban HR station areas Time series regression analysis Small Cities Study Rural and urban bus service areas Time series regression analysis National BRT Study Urban LRT station areas Time series regression analysis SHRP 2 Case Studies Urban HR and bus station areas Pre-post case study TCRP H-50 Case Studies Urban HR, LRT, BRT station areas Pre-post case study TABLE 7 CASE STUDIES OF PREVIOUSLy BUILT PROjECTS: SETTING, MODES, AND MODELS

29 The SHRP 2/EconWorks study is a database of transportation project case studies that feature economic impact measurements. The database grew out of the federal SHRP 2 project, which origi- nally funded a series of highway economic impact case studies, and later expanded the cases to include nine passenger intermodal projects—rail transit stations located along highways to support intermodal passenger rail stations. The case studies covered impacts on surrounding areas including economic development, land values, and land development. The web access to this database is now housed in the AASHTO-operated EconWorks website https://planningtools.transportation.org/13/ econworks.html. The TCRP H-50 study (Economic Development Research Group 2016) was conducted by the authoring consulting firm for TCRP. It included seven case studies of transit-oriented development and their economic impacts. The cases were developed as part of a pilot effort to demonstrate the feasibility of a transit economic impact case study database. The case studies were intentionally pat- terned to parallel the SHRP 2 case studies noted previously, and they were added to the EconWorks database (described earlier). anaLysis of societaL benefits (including bca studies) These studies focus on the valuation of societal benefits instead of impacts on the economy. They are also referred to as social welfare analyses because the economic value of benefits is based on social welfare (willingness to pay) principles. The results may be expressed in terms of a single year— representing either the current or expected future benefits of actions. These same measures may be used to drive a BCA, in which case they are expressed in terms of net present value. The BCA studies are discussed later in this chapter. In all cases, the calculation of benefits is made by comparing sce- narios with some difference in their associated travel characteristics. terminology There is a range of terminology used in various reports to describe the subject matter covered by this type of analysis. It includes: • alternative aggregate terms: social welfare benefit, societal benefit, or social benefit; • components: user benefit, environmental benefit, economic development benefit, social/ community benefit, low income mobility benefit, or avoided public costs; and • methods: benefit–cost (or cost–benefit) analysis, and multiple account evaluation. approach These studies calculate the economic value of benefits to society (various people and businesses). The benefits may represent the value provided by an existing facility or service, or the added value that would be provided if a proposed (new or improved) facility or service is completed in the future. Either way, the methodology to be used and interpretation of results is the same. In the context of economic analysis, societal benefits have a dollar value that can be determined either from either direct observation (e.g., cost savings, medical costs of injuries, and vehicle repair costs), from revealed preferences (e.g., survey research on willingness to pay for cleaner air or better access to transit lines, or faster time getting to shopping destinations). Thus, societal benefits capture many elements of benefit that have value to people, but do not directly affect business growth. In this respect, societal benefit studies may be seen as capturing broader effects than an EIS. Another distinguishing aspect of societal benefit studies is that they are typically expansive in terms of covering non-user benefits. Some focus on aggregate net benefits, without considering a specific study area. This broad, aggregate view does not typically capture distributional effects among neighborhoods, population groups, or elements of the economy. Therefore, a neighborhood revitalization effort that might be captured by an ex-post EIS may be seen from this perspective as a zero sum redistribution of income.

30 However, there is a second type of benefit study that focuses specifically on segments of the population or specific neighborhoods, and then counts benefits associated with a redistribution of investment to those target groups or areas. For instance, a project that enhances mobility for carless, elderly households may be seen as beneficial in terms of reducing social service costs. Similarly, a project that brings business activity to a depressed neighborhood may be viewed as beneficial in terms of reducing blight and crime. Methodology These studies calculate the effects of a proposed transportation scenario on changes in the safety, air pollution, development, land values, and the associated income and spending patterns of households, businesses, and government agencies. To accomplish this, four steps are required: a. Estimation of scenario effects on changes in transportation conditions, in terms of trips, vehi- cles, travel times, operating costs, safety, reliability, accessibility, etc. This can be done by means of a travel forecasting model or expert engineering and planning estimates. b. Translation of step “a” outcomes into direct impacts for various classes of users, affected non- users, and public agencies, as well as direct consequences for emissions of air pollution. This requires the allocation of travel impacts to various classes of trip purposes and vehicle types, and the application of various coefficients representing emissions rates, vehicular accident, congestion related delays, or accessibility metrics. Some studies also use statistical analysis to relate land use and density increases to accessibility improvements. c. Conversion of step “b” impacts into money values typically requires the application of unit cost factors, or in some cases, accessibility impact elasticities that represent impacts on land values or business productivity. d. Packaging of step “c” results by: (1) adding them together to portray annual benefits, or (2) calculating the net value of benefit and cost streams through BCA, or (3) describing benefits accruing to different parties in the form of a multiple account evaluation. Steps “b,” “c,” and “d” can be accomplished using spreadsheets or various models that are freely available on the web. A discussion of available BCA tools is provided in the “Transportation Benefit Cost Analysis” web guide, operated by TRB’s Standing Committee on Transportation Economics (available at: http://bca.transportationeconomics.org). Further guidance on the application of BCA, including issues such as the choice of perspective and selection of discount rates, can be found at that source. interpretation and use There are two types of uses of these studies. Some are done in the form of BCAs, intended to show the efficiency or return on investment justification for selecting preferred selections and funding alternatives. Others are done to show the value of effects that are otherwise missed by economic impact or classical BCA studies. reviewed studies Fourteen studies that cover social benefits were reviewed. These studies are compared in two tables, and then individually discussed in terms of notable aspects of the study analysis methods and termi- nology used. Table 9 summarizes the types of study areas, transit modes, and economic analysis methods used. The 14 studies cover individual bus and rail transit lines, as well as entire systems. The first ten studies examine a wide range of societal benefits associated with the continuation of existing transit systems or the addition of new transit lines. All include analysis of public data concerning land values/use, tax revenues, and/or public service costs. Half also utilize transit rider survey information to help establish transit benefits. Three of the ten societal benefit studies make use of regional travel models

31 to calculate travel time and cost savings benefits, whereas all four of the BCA studies rely on regional travel model results. Table 10 summarizes forms of benefits that were measured and valued in the studies. It shows that all of the studies cover the basic user benefit analysis factors of travel time and cost, nearly all include safety, and a majority also includes environmental benefits. In addition, each of the societal benefit studies includes some additional benefit classes that affect specific parties. They vary from study to study, but include avoided infrastructure costs for government, infrastructure spending effects that support local jobs, increases in property values for specific areas, or costs associated with social services and access to health care and education. Each of the formal BCA studies focused more narrowly on user and environmental benefits, but most also included some form of productivity impact associated with access (agglomeration) ben- efits. Each of these studies was done as an accompaniment to a separate EIA. Several of them were also part of, or accompanied by, a larger EIA. Descriptions of these studies are provided here. Readers are warned that not all of these studies represent best practices to be replicated. They do represent a wide variety of different perspectives; some involve strong assumptions that may not be universally embraced and some apply economic methods in non-conventional ways. Aspects of the study that may be seen as non-conventional— both innovations and applications counter to standard practice—are noted in the following descrip- tions. These issues are then discussed more fully in chapter three, “Analysis of Societal Benefits.” The New york City study was prepared for Vision42, local a civic group (Urbanomics 2005). This report, despite having “economic impact” in the title, is actually about the value of benefits of a pro- posed LRT line. It covers all of the traditional travel user benefits for transit riders and car travelers, plus expected increases in land values that generate added rental income for land owners. The annual benefits are computed and compared with annual O&M costs for an average year. Study Mode and Study Area Type* Economic Analysis Methods Travel model (a) Survey analysis (b) Public records (c) Societal Benefit Studies NYC: Proposed LRT Urban LRT line — — X Tennessee Rural Transit Rural bus systems — X X SF MTA Benefits Urban HR, LRT and bus transit system — — X Wisconsin Transit Benefits Rural and urban bus services — X X Michigan Transit Benefits Rural and urban bus services — X X Anchorage Econ Benefits Regional bus transit system X X X Austin Urban Rail Regional LRT and bus transit system X — X Washington WMATA Benefits Regional HR and bus transit system X X X Seattle: Sound Transit Plan Regional LRT and bus transit system — X Milwaukee County Transit Regional bus transit system — X X Standard BCA Studies (done in addition to Economic Impact Studies) Mpls.–St. Paul: Itasca Study Regional LRT and bus lines X — — Durham Transit (Canada) Regional LRT and BRT lines X — — Toronto Go Electr. (Canada) Regional CR electrification X — — London Crossrail (UK) Regional HR line X — — *HR = heavy rail transit, LRT = light rail transit, BRT = bus rapid transit, CR = commuter rail. (a) refers to network simulation and/or mode choice models. (b) refers to travel surveys covering mode choice and trip purpose. (c) refers to land value/use, tax revenues, and government spending data. Electr. = Electrification Source: Review of local analysis studies. TABLE 9 TRANSIT BENEFIT VALUE STUDIES: SETTING, MODES, AND MODELS

32 The Tennessee study was conducted by Oak Ridge National Laboratory (Southworth 2005). This study assessed the benefits of rural transit and urban transit services, compared with what would occur today if transit was not funded. Besides travel-related time, cost, and safety, the annual mobility benefits included avoided costs (from what would have occurred if there was no transit service) associated with users foregoing trips to access medical, educational, employment, and personal business destinations. Lost income from transit spending (and multiplier effects) was also included. The San Francisco study was prepared for the San Francisco Municipal Transportation Authority (EPS and CHSS 2015). This economic benefit study developed an analytical framework to demon- strate how maintaining and expanding Muni services and infrastructure provide a positive ROI to the city and is an essential part of ongoing economic sustainability. The study considered a hypothetical scenario in which Muni does not exist. Besides the standard user benefit of vehicle operating cost sav- ings, the BCA also included the value of savings in parking costs. The Wisconsin study was prepared for Wisconsin DOT (HDR-HLB Decision Economics 2006). This report calculated annual statewide benefits associated with the availability of transit services in the state. Besides the standard user benefits, the study included the benefit of avoided cost savings that would be incurred without transit—including public assistance and home health care costs, as well as higher costs of travel to health care, education, employment, and social and recreation desti- nations. It also compared average year benefits with average year transit operating costs. The Michigan study was prepared for Michigan DOT (HDR Decision Economics 2010). This report calculated annual statewide benefits associated with the availability of transit services in the state. The income generated by transit spending as well as household cost savings were recognized as Forms of Benefit Study Time & cost (a) Safety (b) Envir. (c) Prod. (d) Infra (e) Land value (f) Trans spend (g) Gov. serv. + health (h) Govt. rev. (i) Societal Benefit Studies NYC: Proposed LRT X X — — X X — — — Tennessee Rural Transit X X X — — — X X — SF MTA Benefits X X X — — — — — — Wisconsin Transit Benefits X — — — — — — X — Michigan Transit Benefits X X X — — — — X — Anchorage Econ. Benefits X X X — — — x* X — Austin Urban Rail — — — X X X — — X Washington WMATA Benefits X X X — X X x* — — Seattle: Sound Transit Plan X X X — — — x* — — Milwaukee County Transit X X X — X — — X — Standard BCA Studies (done in addition to Economic Impact Studies) Mpls.–St. Paul: Itasca Study X X X — — — — — — Durham Transit (Canada) X X X X — — — X — Toronto Go Electr. (Canada) X X X X — — — — — London Crossrail (UK) X X — X — — — — X (a) including travel time and vehicle operating cost reduction, and improved reliability. (b) collision costs reduction. (c) including pollution and carbon emissions reduction. (d) includes economic productivity gains related to enhanced access and agglomeration. (e) refers to avoided costs of additional infrastructure and parking investment. (f) refers to residential and/or commercial property valuation gains. (g) refers to household income generated by increased spending on transit operations and maintenance due to new project (note that this is usually accounted for outside of the total benefit calculation). (h) refers to avoided government costs associated with unemployment payments, health care, and social services that would be incurred without transit investment. (i) refers to added government revenue associated with household income or property taxes. x* refers to an impact that was calculated separately and not added to total societal benefit. Source: Review of local analysis studies. TABLE 10 TRANSIT BENEFIT VALUE STUDIES: INPUT DATA AND OUTPUT RESULTS

33 benefits to residents, and an I-O multiplier model was applied to calculate broader economic impacts. Total benefits for an average year were compared with average annual transit O&M costs. The study also included a spreadsheet model designed for use by individual transit agencies. The Anchorage study was conducted by the Institute of Social and Economic Research at the University of Alaska, Anchorage (Goldsmith et al. 2006). This report calculated the benefit of transit relative to a scenario in which the transit system does not exist. The study calculated transit benefits as the sum of user, social, and community benefits. User benefits are expanded to include savings in parking and car ownership costs. Social benefits include savings in costs of access to jobs, medi- cal services, social services, educational opportunities, and recreation. Community benefits include safety, environment, and reduced costs for building parking facilities. These benefits are compared with annual local costs of property taxes to subsidize transit operations. The Austin study was conducted by the Center for Sustainable Development at the Univer- sity of Texas, Austin, with staff of the Capitol Area Council of Governments and city of Austin (Austin Economic Development Department 2013). This study identified a series of benefits accruing from proposed urban rail expansion in Austin. It used a geographic information system and land value model to estimate property value increases, a fiscal impact model to estimate tax revenue growth, and an I-O model to translate household savings in car ownership, parking, and vehicle operating costs into jobs and wage growth, and applied agglomeration elasticities to estimate impacts on GDP. The Washington, D.C. study was prepared for the Washington Metropolitan Area Transit Author- ity (AECOM and Smart Growth America 2011). This study identified benefits associated with transit services within the Washington, D.C., metro area, compared with a scenario in which transit services do not exist and another scenario in which highway capacity is added instead of transit investment to achieve base case levels of service. Besides traditional user benefits and environmental benefits, the analysis covered transit benefits associated with added property values and avoided costs of highway and parking infrastructure. In the report, values are shown for each benefit or impact category, but are not added together. The Seattle study was prepared for Sound Transit (PB Consult 2008). This report described BCA associated with the rail transit portion of the long-range transit plan. As with other rail transit BCA studies, benefits to highway users (from reduced congestion) comprise the largest category of travel- related benefits. A separate volume, prepared for the environmental impact report, also discussed broader societal impacts; however, they are not presented as monetized benefit values. The Milwaukee study was prepared for Milwaukee County (Huron Consulting Group et al. 2015). This study identified benefits associated with transit services within the Milwaukee area compared with what would occur if transit did not exist. Besides traditional user and environmental benefits, the study included avoided household costs associated with car ownership and operation, parking, taxis, and chauffeurs, as well as avoided public costs associated with pavement damage, unemploy- ment compensation, food stamps, and medical care for those left unable to get to work or access medical care. These annual benefits are totaled and then compared with the transit agency’s annual operating cost. The Minneapolis–St. Paul study was prepared for the Itasca Group, a business-led civic organiza- tion (Cambridge Systematics 2012). The report examined the net benefits of scenarios that accelerate and/or expand fixed guideway transit system development in the region. Both EIA and BCA were done. The latter accounted for benefits of savings in travel time, vehicle operating costs, air pollution, safety, and pavement maintenance. Rather than a formal BCA with a discounted net present value cal- culation, benefits in this study were calculated for a number of years and compared with average O&M costs to provide an internal rate of return. The Durham study was prepared for the Regional Municipality of Durham, Ontario (HDR Inc. and iTRANS 2010). This study calculated net benefits and benefit–cost ratios for alternative transit investment scenarios (in addition to providing EIA, which was described earlier). It covered standard

34 travel, safety, and environmental benefit categories, but also added the benefit of saving on social service costs owing to better mobility for ride-dependent households. It is notable for showing how benefit–cost ratios can differ depending on whether benefits for road users are also counted. The Toronto study was prepared for Metrolinx (Delcan Arup et al. 2010). This study exam- ined the net benefits and benefit–cost ratios for electrification of commuter rail lines. The report calculated benefit–cost ratios for alternative investment scenarios (in addition to a separate EIA). Besides the standard user travel time and vehicle operating cost benefits, the BCA included the value of enhanced reliability. Additional social and community benefits were noted as qualitative benefits. For the London study, the city and national governments funded a series of benefit–cost studies for the proposed Crossrail transit line across London (in addition to a separate EIA, discussed earlier) (Crossrail 2010). Both benefit–cost ratios and discounted net benefits were calculated, once in terms of traditional user benefits (time, vehicle operating cost, safety, etc.) and again with the addition of wider economic benefits (including added worker productivity and GDP growth from economic agglomeration). A separate analysis effort also estimated impacts on land value uplift in the vicinity of new stations as a result of enhanced accessibility. Revenue gains to government were also counted as a benefit, as stated in the U.K. guidelines for BCA. LocaL studies—Key observations similarities Overall, transit EIS widely utilize comparisons between the actual multi-modal system performance in contrast to a comparable situation without transit. This generally follows a “base case” versus “build case” structure consistent with economic analysis typically applied to other modes. However, within this context it is also notable that local studies are often highly tailored around specific policy contexts. Unlike impact studies for other modes of transportation, transit impact studies often rely on measures that extend well beyond traditional transportation categories such as overall managerial efficiency for the transit operator, public health, and land values as shown earlier in Table 8. differences The studies that were reviewed in this chapter illustrate opportunities to capture important benefits and impacts that can be relevant for decision making. They also illustrate limitations associated with various measurement perspectives and the pitfalls of being unclear about the viewpoint being taken. These opportunities, limitations, and pitfalls can be shown by considering how each of the measure- ment methods treats the various forms of economic impact and benefit. The studies that were reviewed in this chapter generally apply one of three basic methodologies: 1. EIA applies rigorous models of the economy (either static I-O models, dynamic simulation models, or statistical models) to calculate the net effects of transportation scenarios on jobs and income in a designated area. The study area may be a small area or a large area. Income concepts include worker wages, total household income, value added (GDP), and business output (revenue). 2. BCA applies rigorous net present value calculations to generate measures of the efficiency of spending in terms of the present value of a time stream of total societal benefits, net of total costs. A benefit–cost ratio may also be shown. 3. Societal benefit studies (also referred to as social benefit analysis or multiple account evaluation studies) portray the economic value of various forms of benefits that accrue to various parties in various areas at various times. This form of analysis is powerful in capturing socially important distributional effects—particularly benefits to specific area and population groups that are vul- nerable or in need of support. Many of these effects disappear in BCA studies because current BCA practices aggregate over space, time, and elements of the economy and population.

35 The EIS become particularly useful for transit analysis because they can be tailored to capture impacts on specified (small or large) areas served by transit. EIS can also be important in showing the effects of transit investment on achievement of longer-term economic development strategy out- comes. (This was notably done for several of the predictive studies.) In addition, the societal benefit studies can be particularly useful for transit analysis because they capture distributional effects that are of interest for public policy and that are missed by the more aggregate BCA accounting. However, the societal benefit studies that were reviewed here vary widely in terms of rigor, which is defined as the notation that some benefits apply only from the perspective of a specific group, place, or type of agency. Some of these studies keep these benefit accounts separate; others combine benefits and costs that involve different perspectives, ignoring that some of the benefits come through costs incurred by others. Some of the studies also invoke strong assumptions, particularly that cost savings to households can be translated into equal or greater job and wage growth by industries. There are apparent cases where transit advocates, in their eagerness to demonstrate transit ben- efits, fail to note that benefits to some parties may come at the expense of others. For instance, some studies fail to note that savings in car parking costs at workplaces may be achieved by enabling free or low-cost park & ride facilities at transit stations, which come at taxpayer expense. how analysis Methods treat different types of impacts and benefits The remainder of this section presents the basic categories of transit impacts and shows how they lead to different economic outcome measures when viewed from the different perspectives taken by these methodologies. Reducing the Cost of Travel This refers to savings in travel time, travel-related expense, and/or accident costs that (transit, car, or truck) travelers incur as a result of new or improved transit. The calculation is based on current or projected future origin–destination patterns. • Societal benefit analysis and BCA methods recognize the value of these savings as a benefit for (transit or car) travelers. However, care must be taken to ensure that expense savings are true societal gains and not merely cost savings to one party received through a subsidy provided by another party. • EIA treats the savings for business-related travel as a reduction in business operating costs, which increases the competitiveness of local businesses and hence leads to greater local eco- nomic growth in the specified area. However, it treats the other travel savings as a shift in household spending patterns, reducing spending on vehicles, fuel, and repair services, and shifting it to purchases of other consumer products and services. The spending shift will have no impact on income in the study area unless it can be shown that it enables some “import sub- stitution”; that is, a greater share of the purchase revenues going to local suppliers in place of imports from outside the region. Market Access: Enhancing Productivity This refers to market scale and agglomeration economies that accrue to businesses because transit can expand access to labor and business customer markets, and provide opportunity to access more specialized worker skills and serve more specialized customer needs. • BCA studies recognize this effect for major projects that have regional impact. It is recognized as an additional benefit since it leads to GDP (income) growth that is beyond the effect of travel time and travel cost savings. It is the largest component of what the U.K.’s BCA accounting guidance refers to as “wider economic benefits.” • Societal benefit studies could also recognize this productivity gain generated by broader acces- sibility. However, in practice these studies seldom do so. Instead, these studies tend to measure

36 accessibility gains in other ways, most often as reducing the cost for dependent households to access employment, education, and health care destinations (to be discussed later). • EIS recognize this productivity gain as a wider economic impact, which increases local busi- ness competitiveness and hence leads to greater economic growth in the specified area. Market Access: Increasing Land Values This refers to the effects of transit improvements on increasing access to, and expanding access from, areas that are within walking distance of transit stops and stations. This makes those locations poten- tially more attractive for residents and businesses; increasing demand then drives up land values. • Societal benefit studies often treat this localized effect as a benefit for the local area, since the local land value uplift represents the capitalization of its access benefit (i.e., a “revealed prefer- ence” measure of how much local households and businesses value that access improvement). • Formal BCA studies tend to take a broader area view, in which there is likely to be a spatial redistribution of location demand patterns, as some areas gain land value relative to other areas. If wider regional income is not changed, then the spatial demand shift is likely to be a zero sum effect (in which gains and losses offset each other). However, if there are agglomeration effects that increase regional productivity, then there may be net inward investment to the region and, in that case, there can be a net gain land values. • EIS at the regional level do not recognize localized land value shifts. Instead, access gains are recognized by their effect on broader GDP growth (see preceding productivity effect). However, local EIS (e.g., studies of transit-oriented development) may focus on the economy of a specific neighborhood or community. In those cases, an uplift in commercial land values or office rents may be recognized as an indicator of increased business income being generated there. Changes in the value of residential land is seen as a transfer of wealth among landowners in which there is no net income change, although this effect may be recognized as an indicator of desired societal benefit. Market Access: Providing Transportation for Dependent Populations This refers to savings in household or government costs associated with providing access for depen- dent populations (individuals who are poor, do not drive, or do not have car access) to get to jobs, education, health care, and personal business destinations. Without transit, one of three outcomes occurs: (1) these households incur higher travel costs for taxis (that could be directed instead to purchase food or housing), (2) these households depend on social service vans that are publicly sub- sidized, or (3) these households forego job, education, and/or health care services; thus, reducing net household income and/or raising public service costs. • Societal benefit studies often recognize these savings for specific population segments as a benefit for those segments. • Formal BCA studies tend to take a broader view aggregating impacts on all sectors of society, viewing these effects as income transfers between taxpayers, government, and recipients. Hence, there is only an economic benefit if this effect leads to changes in worker productivity or net public costs of social services. • EIS usually ignore this effect, insofar as they are income transfers between households and businesses, which have no impact on jobs and income unless they lead to either worker pro- ductivity changes or net inward investment flowing to an area. Avoided Cost of Road Infrastructure This refers to savings in government costs enabled because greater transit reliance can reduce con- gestion growth and the need for investment in new and expanded road infrastructure. • Societal benefit studies sometimes recognize these savings as a government benefit associated with reduced highway spending.

37 • Formal BCA studies consider the opportunity cost of transit and highway spending; therefore, cross-allocations of costs do not matter unless they also affect multi-modal performance mea- sures of service quality or net public expenditures. • EIS usually ignore this effect, which they view as transfers between households, government, and business sectors of the economy that have no impact on jobs and income unless they lead to shifts in business costs, productivity, or capacity for continued economic growth. Transit Spending Effect on Generation of Jobs and Income This refers to the effect of transit (capital or operations and maintenance) spending on the generation of local jobs and income in a community. • Societal benefit studies may recognize the job and income generated by transit spending to be a benefit for a local study area. This is particularly true when the study area is a rural region, isolated community, or urban neighborhood that is economically depressed, has high unem- ployment, and the influx of spending would not otherwise be flowing to that area. • Formal BCA studies consider transit spending to be part of the cost side of the benefit–cost equation, which cannot also be counted on the benefit side. • EIS often recognize the spending effect on jobs and income as a short-term impact, which is kept separate from reporting on the longer-term job and income impact derived from transpor- tation service improvements. Time Frame Effects This refers to the treatment of impact and benefit changes over time. • Societal benefit studies typically focus on a single year—usually the current year, but some- times a designated future year. They report on transit benefits occurring in that year (com- pared with a scenario without transit or without a proposed new project). The benefits may be summed or reported as separate benefit accounts. If they are summed, they may be compared with transit spending in that year. Most of those studies compare benefits with transit operations and maintenance costs, ignoring capital investment costs. • Formal BCA studies consider the opportunity cost of transit capital and operations spending, evaluate the full stream of benefits and costs over a period of time, and apply a discount rate to calculate a net present value of benefits—cost. • EIS are designed to represent comparative scenarios. They most often report on the net differ- ence in the size of the economy of a specific area for a specific year, relative to what would be the case without transit (or the proposed project). The result may be calculated for the current year, for a future year that is 20 or 30 years away, and/or a series of years. The net difference in economic activity is calculated in constant dollar terms, so that scenarios can be viewed in today’s terms.