Critical Analysis of Conventional Transport Economic Evaluation

Transportation economic evaluation quantifies and monetizes transport project’s benefits and costs. It can significantly influence planning decisions. This report critically examines conventional transport economic evaluation. It integrates two different but overlapping perspectives: planners interested in comprehensive and multi-modal transport system analysis and economists interested in economic efficiency and economic development impacts. This analysis indicates that conventional transport economic evaluation fails to reflect basic economic principles including comprehensive and neutral analysis, economic efficiency, consumer sovereignty and integrated decision-making. It evaluates transport system performance based primarily on vehicle travel speeds and operating costs, and overlooks other accessibility factors such as the quality of other modes, transport network connectivity and geographic accessibility. It overlooks many significant impacts including parking costs, vehicle ownership costs, mobility for non-drivers, public fitness and health, and the incremental costs of induced vehicle travel. These omissions and biases tend to favor mobility over accessibility and automobile travel over other modes. Theoretical and empirical evidence indicate that these distortions often reduce economic productivity. More comprehensive and multi-modal evaluation can provide better guidance for transport planning and economic development. Critical Analysis of Conventional Transport Economic Evaluation Victoria Transport Policy Institute 2 Introduction Transportation economic evaluation refers to various methods and computer programs used in transport planning to quantify and monetize (measure in monetary units) the impacts (benefits and costs) of a transport policy or project. The assumptions and methods used in such evaluations can significantly affect planning outcomes: a policy or project may seem beneficial and desirable evaluated one way but harmful and undesirable evaluated another way. Economic evaluation often includes economic impacts analysis which evaluates economic development impacts such as changes in business activity, productivity, employment, income, property development and tax revenues (Ellis, Glover and Norboge 2012). Economic development can provide dispersed benefits and so is often considered worth of public support and subsidy. Economic development results from producer (business) savings and efficiencies. Consumer savings, such as personal travel time savings, are economic benefits, but generally provide no direct economic development benefits. Conventional transport economic evaluation methods originally developed to answer relatively simple questions, such as whether a highway project can repay its construction costs through travel time and vehicle operating cost savings. They consider a relatively narrow range of modes, objectives, impacts and transport improvement options, leading to biased results. To their credit, many transport modelers, planners and economists are working to improve evaluation methods by developing more integrated models that incorporate more impacts. However, these improvements are incremental, there are still significant gaps and biases even in the best transport evaluation models. It is important that people who use evaluation results understand these omissions and biases. This is a timely issue. Current demographic and economic trends are changing future travel demands, and the scope of modes, objectives, impacts and options considered in transport planning is expanding. These trends require more comprehensive and multi-modal evaluation to identify the policies and projects that best meet future needs. This report investigates these issues. It critically examines transport economic evaluation methods, identifies their omissions and biases, and discusses how these are likely to affect transport policy and planning decisions. It bridges the two different but overlapping disciplines: planners interested in comprehensive evaluation that responds to changing community demands, and economists interested in applying economic principles such as consumer sovereignty and economic efficiency. This critique should be of interest to anybody involved in transport planning, economic evaluation, or who uses the results. Critical Analysis of Conventional Transport Economic Evaluation Victoria Transport Policy Institute 3 Conventional Economic Evaluation Conventional transportation economic evaluation uses various computer programs to quantify and monetize (measure in monetary units) the economic impacts (changes in travel time, vehicle operating costs, accidents, and emissions, etc.) caused by transport system changes (FHWA 2012; Markow 2012; SHRP 2012; TEC 2012). Figure 1 Typical Benefits and Costs Distribution by Stakeholder Group (FHWA 2012) Conventional transport evaluation considers a relatively limited set of impacts. Vehicle ownership and parking costs, mobility for non-drivers and public fitness impacts are overlooked. Table 1 summarizes the analysis scope of various transportation evaluation tools. Although none is truly comprehensive, some newer programs (TREDIS, PECAS and RUBMRIO), consider a broader range of impacts. For example, car ownership and parking costs are now recognized and can be modeled with TREDIS and REMI (Weisbrod and Reno 2009). Table 1 Transport Economic Evaluation Tools (Ellis, Glover and Norboge 2012) This table summarizes various tools used to evaluate transport policy and project economic impacts. Note that many (CDSS, HEEM-III, LEAP, MicroBenCost, REIMHS, REIMS and SPASM) are outdated and should not be used, others (SMITE, SCRITS and Highway 1) are very specialized, HEAT is an example of a state-specific REMI shell program (other states have similar versions) and EMME3 is an example of a travel demand forecasting model, not an economic model. The box on the following page describes the general capabilities and limitations of the various types of models. Critical Analysis of Conventional Transport Economic Evaluation Victoria Transport Policy Institute 4 Choosing Economic Analysis Software – Overview of Model Tools Glen Weisbrod (2003), Economic Development Research Group (www.edrgroup.com) Input-output (I-O) models calculate the economic development impacts (jobs, income and GDP) resulting from changes in regional business activity. For instance, they can be used to calculate the impacts of a new or expanded airport, power plant or construction activity. However, they have no longterm forecasting dimension, and no internal ability to forecast impacts of changes in transport costs or market access. For the US, the most commonly used RIMS-II and IMPLAN. For Canada, Statistics Canada offers provincial level models. More specialized models are also available. Economic impact forecasting models are more comprehensive evaluation tools that calculate changes in business attraction as well as growth. They incorporate I-O models and add capabilities to calculate the economic growth consequences of changes in household and business costs (due to travel time and travel cost changes). Newer ones also calculate impacts of changes in market access and trade over time. For the US and Canada, the most commonly used are REMI and TREDIS. Both are multi-regional, spatial economic models that can also tax revenue impacts and social benefit-cost measures. Other models with more limited or specialized uses are also available. Land use models forecast change in the location patterns of population, employment, housing and business activities. The newer versions are sometimes referred to as spatial input-output models because they base their allocation of business growth on I-O models, with greater spatial detail and less industry detail. They account for market access but not business attraction, because they assume fixed regional growth. The most widely used are PECAS and UrbanSim. User benefit/cost models are designed to help engineers and planners identify, rate and select optimal highway projects. They assess highway improvement benefits travel speed and delay, safety, and sometimes emissions rates. They do not consider economic development impacts, although some of their impacts are incorporated into forecasting models. BCA.net provides project level analysis. HERS-ST assesses statewide highway investment needs and project priorities. LCCA evaluates facility lifecycle costs. Other widely recognized benefit-cost tools used by state DOTs are CalBC and NetBC. A variety of other transportation planning tools are also available. Economic development tools are models and datasets designed to assist in business attraction and site location decisions. BizCosts, LocationSelector and FacilityLocations, Site Selector Pro and LEAP all compare alternative locations in terms of business operating costs, market conditions, labor force, land, transportation access, etc. They can assist businesses in making site location decisions for new facilities, and economic developers can use them to identify relative their area’s strengths, weaknesses and best targets for business attraction. They generally consider transportation access in limited terms, such as distance to nearest interstate highway and airport. Critical Analysis of Conventional Transport Economic Evaluation Victoria Transport Policy Institute 5 MobilityVersus Accessibility-Based Planning Conventional transport evaluation often assumes that transportation refers simply to mobility (physical movement) and so evaluates transport system performance using indicators such as average travel speed, congestion delay, and costs per-mile or -kilometer. More comprehensive and multi-modal planning recognizes that the ultimate goal of most travel activity is access to services and activities (CTS 2010; Litman 2003). Several factors can affect accessibility:  Motor vehicle travel conditions. Automobile travel speeds, affordability and safety.  Quality of other modes. Walking, cycling, public transit, telework, delivery services speeds, convenience, comfort, affordability and safety.  Transport network connectivity. Density of paths and roadway connections, and therefore the directness of travel between destinations, plus the quality if connections with public transport.  Land use proximity. Development density and mix, and therefore distances between activities. Planning decisions often involve tradeoffs between different forms of accessibility. For example, roadway design features intended to increase automobile accessibility, including wider roadway with higher traffic speeds, and hierarchical road networks that have fewer intersections, allow higher vehicle travel speeds, but create barriers to pedestrian travel, and since most public transit trips involve walking links, reduces transit access, and increase travel distances, as illustrated below. Locations convenient for automobile access, such as major highway intersections, tend to be difficult to access by other modes, while more central locations that are easier to access by walking and transit often have more intense traffic and parking congestion. Figure 2 Hierarchical Versus Well-Connected Road Networks Hierarchical road networks (left) have numerous dead-end streets that connect to higher-speed arterials. This lengthens trip distances and concentrates traffic on a few roads, which tends to increase congestion. A well-connected road network (right) offers multiple, direct routes between destinations, which reduces trip distances and distributes traffic. Well connected road networks often have lower average speeds but total travel times are generally less. Table 2 summarizes the degree that conventional transport economic evaluation considers various accessibility factors, and requirements for more comprehensive evaluation. Conventional evaluation primarily considers automobile and public transit travel speeds, but gives little consideration to non-motorized travel conditions, and if transport network connectivity and land use accessibility are considered at all, it is only at a regional scale. Critical Analysis of Conventional Transport Economic Evaluation Victoria Transport Policy Institute 6 Table 2 Consideration of Accessibility Factors In Transport Planning Evaluation Factor Consideration in Conventional Evaluation Required for Comprehensive Evaluation Motor vehicle travel conditions – traffic speed, congestion delays, vehicle operating costs and safety Usually considered using indicators such as roadway level-of-service, average traffic speeds and congestion costs and crash rates. Impacts should be considered per capita (per capita vehicle costs and crash casualties) to take into account the amount that people travel. Quality of other modes – convenience, comfort, safety and affordability of walking, cycling, ridesharing public transport Considers public transit speed but not comfort. Non-motorized modes ignored. Multi-modal transport system performance indicators that account for convenience, comfort, safety, affordability and integration Transport network connectivity – density of connections between paths, roads and modes, and therefore the directness of travel between destinations Traffic network models consider major regional road and transit networks. Local streets, nonmotorized networks (paths and sidewalks), and connections between modes are often ignored. Fine-grained analysis of sidewalk, path and road network connectivity, and consideration of the connections between modes, such as the ease of walking and biking to public transit terminals. Land use proximity – density and mix, and therefore the distances between activities Often ignored. Some integrated models consider some land use factors. Fine-grained analysis of how land use factors affect accessibility by various modes. Conventional planning evaluates transport system performance based primarily on motor vehicle travel speed and operating costs. New methods are needed for more comprehensive accessibility evaluation. Often-overlooked accessibility factors are frequently significant. Various types of research indicate that vehicle traffic speeds influence overall accessibility less than the quality of transport options, roadway connectivity and land use accessibility. For example, a major U.S. study found that development density has about ten times more influential on the number of destinations that can be reached in a given time period than the same percentage increase in vehicle traffic speeds (Levine, et al. 2012). Another major study found that in Phoenix, Arizona, residents of more central neighborhoods make substantially shorter trips, drive a third fewer daily miles and experience less congestion delays their suburban counterparts due to their improved travel options, more connected streets and increased proximity to destinations (Kuzmyak 2012). The Texas Transportation Institute (TTI) estimates that in large U.S. cities, congestion increases average annual commuting costs by 34 hours and 16.5 gallons of fuel. In contrast, residents of more automobile-dependent U.S. urban regions such as Jacksonville, Nashville and Houston average more than 30 daily vehicle miles traveled, over 50% more than the 20 vehicle miles traveled in compact, multi-modal communities such as New York, Sacramento and Portland. This additional driving adds 104 additional hours and 183 gallons of fuel per capita on average, far more than per capita congestion delays (Litman 2012). These and other studies illustrate the importance of evaluating transport system performance based on accessibility rather than just mobility. Mobility-oriented evaluation favors policies and projects that increase vehicle travel speeds without regard to other accessibility factors. It overvalues roadway expansion and undervalues alternative modes, demand management strategies and smart growth policies. Critical Analysis of Conventional Transport Economic Evaluation Victoria Transport Policy Institute 7 Economic Principles This section discusses basic economic principles that should guide economic evaluation. Comprehensive and Neutral Analysis Accurate evaluation must consider all significant impacts (benefits and costs). One way to determine which impacts should be considered is to start by defining planning objectives (specific conditions that a community wants to achieve, such as reduced congestion and improved public fitness and health). Table 3 indicates the degree that common planning objectives are considered in conventional transport economic evaluation. Table 3 Consideration of Objectives in Transport Economic Evaluation Objective Consideration In Conventional Evaluation Congestion reduction Generally considered. Often dominant Roadway cost savings Generally considered Parking cost savings Often ignored in transport planning but considered in other types of planning User cost savings & affordability Vehicle operating costs and transit fares usually considered, vehicle ownership costs often ignored. Improved mobility for nondrivers Generally ignored in major roadway planning, but recognized in urban street and transit planning Traffic safety Generally considered, but measured per vehicle-mile, Energy conservation Sometimes considered, but measured per vehicle-mile Air emission reductions sometimes considered, but measured per vehicle-mile Efficient land use Not generally considered in individual transport plans Public fitness and health Not generally considered in individual transport plans Only a portion of common transport planning objectives are considered in conventional evaluation. Some impacts are measured per unit of travel which ignores the incremental costs of induced vehicle travel. Table 4 indicates the impacts that are typically included or overlooked in conventional economic evaluation. Overlooked impacts are sometimes called intangibles, with the implication that they are difficult to monetize and insignificant in magnitude. However, many of these impacts have been monetized and are significant in magnitude (Litman 2009). For example, UK and New Zealand transport agencies have standard methods for monetizing parking costs, habitat preservation, changes in mobility options for disadvantaged people, and changes in public fitness and health caused by transport planning decisions (DfT 2006; NZTA 2010). Table 4 Conventional Transport Evaluation Scope of Monetized Impacts Usually Considered Often Overlooked Financial costs to governments Travel speed (reduced congestion delays) Vehicle operating costs (fuel, tolls, tire wear) Per-mile crash risk Project construction environmental impacts Downstream congestion Parking costs Delay to non-motorized travel (barrier effect) Vehicle ownership and mileage-based depreciation Indirect environmental impacts Strategic land use impacts Public fitness and health Conventional transportation economic evaluation tends to monetize a limited set of impacts. Critical Analysis of Conventional Transport Economic Evaluation Victoria Transport Policy Institute 8 Since congestion costs often dominate conventional transport economic evaluation it is important to consider the accuracy of these values. Congestion costs are usually measured by comparing vehicle travel speeds and operating costs between peak and freeflowing conditions (i.e., level-ofservice A), and monetizing increased travel times and fuel consumption. Many economists believe that this approach exaggerates congestion costs by using freeflow speeds as a baseline, and unit travel time values that are higher than most motorists would willingly pay for travel time savings (Litman 2013; iTrans 2006). More realistic baseline and travel time values typically reduce congestion cost estimates by 40-60%. In addition, conventional evaluation often overlooks or undervalues generated traffic and induced vehicle travel, which tends to exaggerate the benefits and underestimates the full costs of urban roadway expansions (Litman 2001). Several studies using various analysis methods indicate that per capita transport expenditures are higher in automobile-oriented communities that have more road and parking supply, and lower fuel and parking fees, than in more compact and multi-modal communities with higher vehicle operating costs, because lower costs per vehicle-mile are more than offset by increased vehicle travel (Makarewicz, et al. 2008; Litman 2008). Typical North Americans spend 350 to 550 annual hours and $4,000 to $6,000 annual dollars on vehicle travel, plus indirect costs including road and parking subsidies, accident and environmental damages. As discussed above, severe congestion adds about 35 travel hours and $60 in fuel expenses annual per capita, which is small compared with the incremental travel time and vehicle costs that can result from automobiledependent transport systems and sprawled development. This suggests that conventional evaluation exaggerates the importance of congestion relative to total transportation costs. Figure 3 illustrates estimates of various automobile costs, showing the portions that are typically considered and overlooked in conventional evaluation. Many significant costs are often overlooked. For example, conventional models ignore vehicle ownership and most parking costs, based on the assumption that transport planning decisions do not affect total vehicle ownership or parking demands. Such assumptions are inappropriate for evaluating alternative modes, demand management strategies or smart growth policies that often do affect these costs. Figure 3 Automobile Costs (Litman 2009) $0.00 $0.05 $0.10 $0.15 $0.20 $0.25 $0.30 Ve hic le Ow ne rsh ip