Investment Prioritization Methods for Low-Volume Roads (2018)

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.

10 Investment Prioritization Methods for Low-Volume Roads disproportionally dependent on them and therefore may be, in some cases, subject to a degree of socioeconomic distress. This raises the importance of basic access considerations when planning and managing LVR. 2.3 Prioritization Practice 2.3.1 Low-Volume Roads Within the Planning Process In the United States, the place held by LVR investment decision making within the overall transportation planning process is a function of (a) who owns and/or has direct influence over LVRs and (b) the nature of the investment decisions most frequently considered for LVRs. Because of the high prevalence of nonstate ownership of low-volume facilities, state DOTs can have both direct and indirect influence over LVR prioritization processes. A review of transportation and governance in all 50 states found that 77% of the road miles in the United States are owned locally and that states distribute transportation funds to local governments through a combination of statutory formulas, legislative appropriations, and discretionary allocations by DOTs or other transportation commissions (Rall et al. 2011). This means that states may set guidelines and serve in a technical support role to local engineers and decision makers, whose role it is to prioritize LVR investments, or states may impose specific requirements for that prioritization process. For example, the Pennsylvania Dirt and Gravel Road Maintenance Program administers funds through county conservation districts and requires that these districts establish ranking sheets that result in numerical scores for use in prioritization of funding to grant applicants. The state provides sample criteria and suggestions of considerations to be incorporated into the review process but leaves the details up to the indi- vidual counties (Pennsylvania State Conservation Commission 2014). Interestingly, because this program focuses on reducing environmental impacts of roads, the State Conservation Commis- sion (under the Department of Agriculture and the Department of Environmental Protection) oversees it rather than the state DOT, but Pennsylvania DOT is an eligible funding applicant. In Minnesota, the legislatively established Local Road Research Board sponsors research of interest to counties and cities. One recent project developed a “system preservation guide” for local governments that provides guidance on how to best evaluate and select strategies for road system preservation (Local Road Research Board 2016). In addition, states own and manage additional low-volume facilities, incorporating them into their overall prioritization processes. The degree to which states are responsible for LVRs varies from state to state. For example, the State of Missouri manages 11,080 centerline miles of LVRs (<400 AADT), composing 33% of its overall system (B. Reeser, Missouri DOT e-mail correspon- dence, 2017). Arizona DOT, on the other hand, has 457 centerline miles of LVRs (<400 AADT) on the Arizona state system, or less than 10% of the total system mileage (Arizona DOT 2017). Finally, given the nature and location of LVRs, some LVRs fall under the jurisdiction of federal land management agencies or tribal governments (Federal Lands Highway 2015). Regarding decision types, LVR research and policy published in the United States and in other developed countries tend to be concerned either with asset management for an already built system (e.g., Falls et al. 2003, Hossain et al. 2003, and Yu 2015) or with ensuring or enhancing safety conditions (e.g., Bowers 2017, Chandler and Anderson 2010). In developing countries, attention is paid both to the planning and construction of new facilities (e.g., Marcet et al. 2007, Muzira et al. 2015) and to the challenge of asset management for an already existing network (e.g., Agarwal et al. 2016, Chamorro and Tighe 2015). Distinguishing LVRs from other roads is a particular focus within the literature on the environmental implications of poor construction or maintenance management practices (Dodson et al. 2007, Keller and Sherar 2003); within the

Review of the Literature 11 literature on the trade-offs associated with selecting or maintaining different types of paved and unpaved surfaces (Zimmerman and Wolters 2004, Kentucky Transportation Center 2003, and Landers 2015); and within the literature on the implications of bridge postings and closures due to inadequate maintenance (Miller 2015a, Miller et al. 2015b). 2.3.2 Evaluation Methods for Low-Volume Roads There are two broad categories of evaluation methods found in the literature for evaluating LVR investments. The first category employs engineering and asset management criteria to guide invest- ment decisions. Methods in the first category are implemented within asset management systems and work to minimize life cycle agency and user costs. Methods in this group focus on traditional engineering performance metrics such as travel time, operating cost, and safety. Methods in the second category build on the first category, incorporating additional factors to arrive at a holistic approach to investment decision making. This group of methods broadens the focus to incorporate critical strategic issues not sufficiently addressed by traditional asset management practices. U.S. Prioritization Based on Engineering and Asset Management Criteria For many agencies in North America, investment decisions and evaluation methods for LVRs begin with the use of bridge and pavement management systems. NCHRP Synthesis 510: Resource Allocation of Available Funding to Programs of Work documents the use of bridge and pavement models to assess investment needs and support resource allocation strategies. These models include databases of existing road assets, information about use patterns (e.g., vehicle volumes and truck volumes), as well as data on asset deterioration rates, project costs, user costs of deteriorated conditions, and investment triggers or minimally tolerable conditions (Duncan and Schroeckenthaler 2017). One common way of representing and tracking potential needs on roads and bridges is by using composite indicators such as sufficiency ratings or other custom indices that aggregate a weighted set of parameters that together describe how sufficient an asset is relative to proper design and maintenance. Selection and weighting of parameters can be achieved either by being based on expert input or via more quantitative methods such as principal component analysis (Pant et al. 2005). NCHRP Synthesis 501: Pavement Management Systems: Putting Data to Work reports there has been an evolution in how pavement management systems are used, progressing from initial use in the 1960s for documenting conditions and estimating investment needs to more evolved decision support, including evaluating treatment strategies from a cost-effectiveness perspective and recommending candidate projects for a preservation program (Zimmerman 2017). Needs models use benefit–cost analysis “to measure the impact of different funding levels against the economic cost to users and agencies” (Duncan and Schroeckenthaler 2017). For example, the Highway Economic Requirements System calculates operating cost savings so ideally the system can be used to analyze vehicle-operating cost savings associated with better pavement conditions (Margiotta 2016). The system also looks at other benefit categories such as safety and travel time costs as a function of road volumes, speeds, and facility design and can be used to calculate exter- nal costs from vehicle emissions. However, the system does not address other broader societal or economic impacts (FHWA 2015). Bridge management systems address the same types of user costs (travel time, vehicle operating, and safety costs) and have additional embedded informa- tion about the length of, and therefore user costs of, detours because of posted or closed bridges (Markow and Hyman 2009, Robert and Gurenich 2008). Examples of these systems include the National Bridge Investment Analysis System used by FHWA and the AASHTOWare Bridge Management software used by many agencies across the United States. While these systems are clearly important inputs to the resource allocation and prioritization process for roads, they are often just the starting point for an agency’s process. A recent survey

12 Investment Prioritization Methods for Low-Volume Roads of state and provincial transportation agencies found that 40% of responding agencies reported a match between pavement management systems improvement recommendations and funded projects at least 70% of the time; about the same percentage reported that system recommenda- tions matched funded projects between 40% and 70% of the time (Zimmerman 2017). The case examples presented in Chapter 4 include additional detail on how these management systems can be input into a broader prioritization process. Moving forward, these systems are likely to become even more important as federal regulations require states to develop asset management plans and operate bridge and pavement management systems. The minimum requirements of this rule, however, apply only to management systems on the National Highway System (NHS) pave- ment and bridge assets (many LVRs are not in the NHS) and do not include analysis of road user costs and benefits in the requirement for life cycle planning (23 CFR, Parts 515 and 667 2016), although agencies are encouraged to exceed minimum requirements. Also part of the evolving asset management discussion is the process of considering cross-asset trade-offs and initiatives, including the use of multicriteria analysis for this purpose (Proctor and Zimmerman 2016). Non-U.S. Methodologies for Low-Volume Roads Within an international context, there are specific analysis systems that also fit into this category of primarily engineering-based models for decision making. Most prominent is the Highway Development and Management model (HDM-4). HDM-4 is a tool used for economic assessment of proposed road construction and maintenance investments. The World Bank originally developed the tool, and the implementation is now coordinated by the World Road Association (PIARC). While the tool can be applied to all types of roads in all types of settings, HDM-4 is commonly used across the developing world. A major element of its use is for rural roads; hence, the tool has been designed to recognize changes in road roughness, frequency of “impassability,” and effective speed as factors affecting user benefit. At its core, HDM-4 incor- porates a benefit–cost framework that considers user and environmental benefits relative to costs. The framework also allows for consideration of secondary effects of road investments by assessing effects on local agricultural or industrial output enabled by the road. (In the jargon of benefit–cost analysis, this is “producer surplus.”) HDM-4 has an additional module for multicriteria analysis to allow consideration of more qualitative social factors, including ratings of road investment impact on poverty, access to social services, and local economic development (Archondo-Callao 2008, The World Bank 2008). The Roads Economic Decision (RED) model was developed by The World Bank for the Sub-Saharan Africa Transport Policy Program (SSATP) and designed to focus specifically on applications for LVR appraisal (under 200 vehicles per day) and applications for unpaved roads. While the HDM-4 model can also be used for low-volume and unpaved roads, RED provides more detail on road roughness and impassability, considering effects of rainfall, truck volume, and dirt/gravel/cobblestone/pavement surface differences. RED also accounts for uncertainty in the assessment of future traffic, road conditions and maintenance levels, seasonality of disrupted road pass ability, and user benefits for nonmotorized vehicles. While RED also provides a multicriteria assessment framework for considering issues such as access to social services, it still requires that these factors be assessed outside the model (The World Bank 2005, Archondo-Callao 1999). The World Bank has published specific guidance in the form of Transport Note No. TRN-21 on economic evaluation for low-volume rural roads that incorporates the tools and methods described above within a more general framework (The World Bank 2005). The guidance is based on a technical paper that investigated design and appraisal methods for rural transport, address- ing the goal of providing basic access for rural communities (Lebo and Schelling 2001). TRN-21 presents a series of options available for extending more traditional consumer surplus approaches (see Table 2 for definitions) in a way that captures the objectives of LVR improvements. These

Review of the Literature 13 options include simplifications employed by the RED model; discussions of cost-effectiveness analysis to address benefits that cannot be easily monetized and, by extension, the weighted cost-effectiveness analysis or multicriteria analysis; and a series of specific enhancements to consumer surplus approaches that are presented as being particularly relevant to LVRs. Specific enhancements include (1) better assessment of the costs or economic losses incurred from seasonal interruptions of access on impassable roads, which varies by type of activity, for example, agriculture versus job access; (2) estimation of the operating cost savings of non- motorized transport as well as savings due to shifts from nonmotorized transport to motorized travel; (3) improved valuation of time savings, including disaggregation by trip purpose and consideration of specific costs of freight transit time; and (4) valuation of social benefits from improved access to schools and health centers usually achieved through the use of sample case studies as guidance for benefit estimation. Some of these extensions address the broader strategic issues that are discussed in the section that follows. For ease of reference, Table 2 summarizes the transportation economic evaluation terminology employed by The World Bank. This terminology is consistent with that used in the United States. Prioritization with Broader Strategic Criteria While the previously mentioned asset management factors are key inputs to prioritization of LVRs, they do not capture the full range of strategic objectives that LVR investments seek to achieve. This section provides examples from the literature of how additional criteria have been brought into, or can be brought into, the decision-making process. Systematic process for accounting for and then comparing costs and benefits of a project. For transportation, typically includes effects on agencies, users, and nonusers. Used to assess the economic efficiency of an investment. All benefits and costs are expressed in comparable monetary terms that represent welfare gains/losses. Term Description Benefit–Cost Analysis* (BCA) Consumer Surplus The difference between what transportation system users are willing to pay in terms of travel costs (travel time, operating costs, etc.) and what they actually pay; traditional BCA analyzes the change in consumer surplus attributable to a transportation improvement. Producer Surplus The difference between what a producer of a good receives in a market and the minimum amount the producer is willing to accept for it (i.e., costs of production). BCA can be extended to include producer surplus effects—that is, additional industrial output enabled by a transportation improvement. Cost Effectiveness Analysis (CEA) An evaluation approach that compares the cost of interventions with their intended impacts (expressed in nonmonetary terms). CEA can be of help in situations in which benefits are not easily monetized. However, it requires clearly stated objectives as well as a process for defining a threshold of CEA acceptance below which the opportunity cost of the investment is greater than the benefits that will be received. CEA can be of particular use in screening projects/alternatives. Multicriteria Analysis This is analysis in which different criteria receive weights based on perceived importance. Transportation projects can be evaluated or ranked by using the composite score produced from multicriteria analysis. Multicriteria analysis can be useful for incorporating multiple objectives as well as incorporating difficult- to-monetize benefits. *Also called cost–benefit analysis. Source: EDR Group, based on Transport Note No. TRN-21 (The World Bank 2005) and Technical Paper 496 (Lebo and Schelling 2001). Table 2. Transportation evaluation terminology.

14 Investment Prioritization Methods for Low-Volume Roads In a recent publication, Agarwal et al. (2016) defined a maintenance priority index of a road section as composed of weighted factors for traffic safety; traffic operation (a function of surface distress); structural condition (fatigue, cracking, and rut depth); and finally a scoring for section importance. This latter score is a user rating between 0.0 and 1.0 added to allow the evaluation to capture factors such as importance to community, connectivity to market centers, connectivity to health centers, or connectivity to other district roads. Similarly, a method proposed from management of LVRs in Spain (Moya et al. 2011) combines criteria for design (15%), road con- dition (30%), traffic (15%), and function performed (40%) to determine a section’s priority. The function score for a road section comprises economic and social criteria. Different economic functions related to agriculture and forestry are assigned specific points, while the social criteria are assigned points on a scale from no importance to high importance. Agarwal et al. (2016) offer guidance about how to make a determination of social importance. For example, they propose that a road be considered of high importance when it provides health care or emergency access. A study for the Tanzanian National Roads Agency suggests using a tiered decision process with cost-effectiveness analyses to assess road management interventions for roads with very low volumes (<50 AADT) (Benmaamar 2003). While regular consumer surplus approaches would not lead to a positive result with these very low volumes, the cost-effectiveness approach was designed to take into account the strategic social or economic importance of these very LVRs. A first screening is carried out based on each jurisdiction’s economic and social indicators. In the case of Tanzania, the Human Development Index is used for this purpose. This results in a narrowing of potential investment alternatives to focus on disadvantaged communities. Then, for the more depressed areas, the road links are prioritized through a cost-effectiveness analysis. The number of served people in each road link’s catchment area is used to express road user benefits and compared with the costs for an upgrade to a minimal standard. A threshold is set to determine which maximal cost per served person is acceptable. Training developed by SSATP for rural road economic appraisal methodologies has been similarly structured to incorporate tiered decision processes that help manage levels of analy- sis effort while still sufficiently addressing the strategic objectives of providing basic access for motorized traffic (SSATP 2013). The SSATP framework represents an operationalization of The World Bank’s TRN-21 guidance and incorporates initial screening as well as subsequent ranking with guidance on when and how to use multicriteria analysis, CEA, and BCA. Of note in the training is an example of how to use poverty criteria to first screen potential investment alternatives (i.e., targeting assistance to disadvantaged areas). In addition, the training presents an example from Bhutan, in which BCA is extended for very low-volume roads (<50 vehicles per day) to include quantification of education and health benefits based on field data collected from specific case examples. U.S. states have also sought to integrate broader strategic criteria into the assessment of maintenance priorities. For example, the South Dakota DOT sponsored research to develop a methodology that would help local road agencies in the state determine the best surface type (gravel or bituminous) (Zimmerman and Wolters 2004). The approach is a guided multicriteria analysis with scoring of importance-weighted factors. Apart from agency cost and user costs (vehicle operating and crash costs), the approach incorporates optional use of additional factors, including political issues, area growth rates, housing concentration (i.e., need for dust control), mail routes, and industry/truck traffic. While these additional factors are scored on a qualitative basis (e.g., 1 to 4), the framework attempts to formalize consideration of factors that might be relevant to surface type selection based on the experiences of the technical panel supporting the research. Pennsylvania has similarly developed guidance for local decision makers charged with directing resources to dirt and gravel LVRs (Pennsylvania State Conservation Commission 2014). As part of the Dirt and

Review of the Literature 15 Gravel Road Maintenance Program, the Commission requires that county conservation dis- tricts develop ranking sheets and numerical scoring for project prioritization. The state-offered sample criteria are based primarily on environmental conditions and effectiveness of proposed solutions from an environmental perspective but also suggest other considerations, including “types of road use (residential, school bus route, timber, agriculture, etc.)” and “future road use plans (developments, drilling, etc.).” The ROADEX research program in Northern Europe (Johansson 2006) proposes a particu- larly interesting way of prioritizing maintenance decisions for LVRs, with a focus on the social benefits that accrue from having roads in rural areas in good condition. The method was devel- oped in recognition of benefits for increased social welfare like improved possibilities for attend- ing schools, dissemination of knowledge and technology, health and other services, increased market competition, and increased possibilities for starting businesses and creating new jobs. Johansson proposed a transportation need index (TNI) to be used as a multiplier for the road user benefits. The TNI summarizes social, economic, and road user needs, incorporating scores for the following subcriteria: fragility (F), lifeline urgency (L), and the road user needs for people (P) and business (B). These subcriteria are defined in Table 3. In addition to research on types of criteria to address broader strategic objectives, the literature also investigates the issue of resolving different perspectives, preferences, or objectives across stakeholders involved in or affected by a potential road investment decision. For example, Otun documented the application of a multicriteria analysis to prioritize road provision in Nigeria. The process was designed to explicitly account for multiple and potentially conflicting objectives by assigning weights to different criteria based on the judgment of interviewed stakeholders. This framework enables arrival at an acceptable decision despite divergent views. It does not presuppose a singular optimal solution (Otun 2016). Additional examples of applied multicriteria analysis processes are found in the compilations by Zopounidis and Doumpos (2017) and in Dalal et al. (2010). Overall, the approaches to evaluating broader strategic objectives of LVRs found in the literature vary in their complexity but generally employ some mix of (a) qualitative judgment-based scoring of a list of criteria and (b) the use of quantitative indicators of a road’s strategic importance. In addition, a significant portion of the literature cited in this section is still within the realm of research or guidance, rather than documentation of currently applied practice. Chapters 3 and 4 delve in more depth into current practices as reported by state DOTs. Critical Strategic Issues in Overall Highway and Road Funding Prioritization In addition to approaches that are specific to LVRs as discussed above, there are also a variety of approaches adopted by states in the United States that bring strategic social and economic policy issues into highway and road funding prioritization in general and thus in practice also into LVR decisions. In response to the federally required transportation improvement program, many state DOTs have put forth criteria for project capital investment and project ranking and, in many cases, there is explicit recognition of the importance of maintaining roads that support locally important social and economic functions even when they have low traffic volumes. This includes enabling (a) access for farm and resource extraction products coming from local sites; (b) access to health, education, and jobs from isolated local communities; and/or (c) secondary feeder roads that connect to major economic corridors. These criteria show up in both formal scoring systems and informal qualitative rating systems, in addition to the traditional traffic flow impact criteria (vehicle miles traveled, vehicle hours traveled, and safety). While the traditional criteria all favor roads with higher traffic volumes,

16 Investment Prioritization Methods for Low-Volume Roads TNI Subcriteria Subcriteria Classification Fragility (F) Developed in the Scottish Highlands, the concept of “fragile areas” identifies areas “where there is a risk that the local communities may not be strong enough to survive.” Fragile areas are defined as communities “being in decline or in danger of decline” as determined through a composite of fragility indicators: • Social fragility—population. Total % population decline; % decline in population aged 0–15 years; population density; % of people retired because of health or age. • Economic fragility—unemployment. Long-term unemployment rate and income support claimant rate. • Accessibility—to key services. Population residing outside of a 20-minute one-way drive to 5 key services: Post Office, Primary School, Grocery Store, GP Surgery, and Petrol Filling Station. • Remoteness indicator—from the main service centre. Population residing outside a 1.5-hour one-way drive from city. 1. No fragility 2. Little fragility 3. Medium fragility 4. High fragility Lifeline Urgency (L) Lifeline or lifeline urgency rural roads are defined “as a transport link which has no substitute, or where the substitute entails a considerable increase in time or money expenditures, where any diminution in the quality, reliability or availability of the former is likely to have a significant impact on the social or economic viability of an affected community.” 1. The road is not a lifeline road. 2. The road has a substitute that presents a minor increase in time and cost. 3. The road has a substitute that presents a major increase in time and cost. 4. The road has no substitute. Transportation Need for People (P) Apart from considering the number of people traveling on a road, their trip purposes and time-specific access needs are also regarded as essential. Travel to access education and work is considered most essential, compared with other trip purposes. 1. Few road users, only temporary use 2. There are only a few permanent residents with no time-scheduled access need. 3. Schoolchildren and commuters 4. High priority use (schoolchildren, commuters, daily bus routes) Transportation Need for Business (B) The importance of business traffic depends on its regularity and the dependence of the businesses on high accessibility. 1. No business traffic 2. Only a few businesses with no need for regular daily transportation 3. Few businesses with needs for regular daily transportation 4. Several businesses requiring daily transportation service with high accessibility needs Source: Based on Johansson 2006. Table 3. Transportation need index subcriteria for LVRs. the social and economic criteria allow for roads with lower volumes but particularly important roles to also receive due recognition in prioritization. Various states also structure their highway funding and ranking systems in such a manner as to recognize the importance of specific classes of roads. For instance: • Kansas considers local community input as a major factor in scoring road projects for state funding (Kansas DOT 2012); • North Carolina divides its program into three tiers. The Division Needs program includes LVRs and is scored 50% based on quantitative measures (congestion, benefit/cost, safety,

Review of the Literature 17 freight and military, and accessibility and connectivity) and 50% based on qualitative input from local DOT divisions and planning partners (North Carolina DOT 2017); • Texas DOT designates and provides funding support for “farm-to-market roads” and “ranch-to-market roads” (Texas DOT 2017); • Ohio DOT’s investment prioritization considers projects that address local economic distress (Ohio DOT 2008); • Wisconsin DOT has a forest road aid program; it also had a highway prioritization scoring process that recognized projects that bring money into the state (by serving tourism or farm/ industry sites) (Wisconsin DOT 2010); and • The Appalachian Regional Commission, which covers 13 states, has a public works program that funds local industrial access roads to work with the states to complete a system of Appalachian Development Highways to connect isolated communities to broader markets (Appalachian Regional Commission n.d.). None of the examples focuses specifically on LVRs, but each one represents a broader recogni- tion of factors that can justify investment in road projects entirely aside from the considerations of traffic volume. Documented Challenges in Low-Volume Road Management and Prioritization While the above sections do point to available approaches for prioritizing LVR investments based on a broad awareness of their strategic importance, there nevertheless remain barriers and challenges to implementing holistic approaches in practice. The challenges, documented in a large portion of LVR literature, include contextual and management factors affecting the overall ability of agencies to manage LVRs, and relatedly, the sufficiency of existing fund allocation and prioritization processes in addressing LVRs. Contextual factors and management challenges. Inadequate maintenance and funding of LVR is a problem observed worldwide including in both developing and OECD countries (Agarwal et al. 2016, Kemp et al. 2016, Chicoine et al. 1989, and Pinard et al. 2003). Funding challenges and preservation burdens have been well documented within the United States across the transportation network including in NCHRP Synthesis 480: Economic and Development Implications of Transportation Disinvestment (Duncan and Weisbrod 2015). Contributing to this challenge in the United States (as well as in many other places) are waning revenues from fuel taxes, an aging highway network with significant reinvestment needs, and construction cost escalation. These factors can be particularly acute for LVRs because as Zimmermann and Wolters (2004) and Yu (2015) explained, decision kings tend to prioritize higher volume or higher classification facilities such as those of the NHS over LVRs. Kemp et al. (2016) highlighted how decision makers often end up deferring maintenance to save in the short term even if defer- ring maintenance results in higher life cycle costs overall. Higher life cycle costs stem from higher costs of reconstruction once a road has deteriorated beyond the point at which less intensive maintenance is no longer an option (Agarwal et al. 2016). Locally owned facilities face unique challenges. Local governments in rural areas, in par- ticular, struggle to manage the costs of LVR upkeep due to dwindling tax bases and high ratios between road mileage and taxable economic activity (Chicoine et al. 1989, Clemmons and Saa- ger 2011). Kemp et al. also described how expected increases in heavy-truck volumes in Australia would further exacerbate the issue of maintaining an adequate state of repair (2016). The U.S. Federal Highway Administration similarly forecasts truck vehicle miles traveled to grow at a faster rate than light-duty vehicles (2017). In some areas, challenging terrain and environmental considerations further contribute to the imbalance between funding availability and LVR needs (Coghlin n.d., Keller and Sherar 2003).

18 Investment Prioritization Methods for Low-Volume Roads Methodological and data challenges. Looking beyond contextual factors, the reviewed literature also highlights methodological challenges that can make LVR management and investment decision making a challenge. Common themes emerge from the literature in both developing and developed countries. These include (a) challenges related to lacking necessary data or uncertainty in necessary data to support LVR analysis and management and (b) a lack of sensi- tivity in existing evaluation approaches to broader strategic issues that are of known importance. In India, Agarwal et al. reported that most existing methods for collecting data on road- way conditions are too sophisticated, costly, or require specialized equipment and operators and are thus not practical in many situations (2016). The researchers therefore endeavored to develop a management strategy that minimizes data inputs and focuses on input data that can be easily obtained. Similarly, the LVR RED model, developed to improve decision making for LVRs in Africa, targets addressing identified deficiencies in existing models. These deficiencies include the need to reduce input requirements and to address greater uncertainty of inputs— particularly traffic volumes and road conditions—through an emphasis on the sensitivity of results (Archondo-Callao 1999). Literature from the United States also points to data limitations that affect decision making for LVRs. The FHWA, for example, found that states experienced difficulty implementing the federally funded High Risk Rural Roads Program because of challenges in determining crash rates on eligible roadways. These challenges included imprecise crash location information in rural areas where reference points are far apart (e.g., intersections or properties) and where there is a lack of volume data needed to normalize crash rates for locally owned roads rather than state-owned roads (Chandler and Anderson 2010). In developing a “social return on investment framework for Iowa rural bridge transportation asset management,” Miller reported that lack of LVR data was a significant research challenge (2015a). Rural roads in Iowa receive less data collection efforts than more primary roads and therefore need models to estimate missing data based on available field data for closely compara- ble assets (Miller 2015a). In North America, inventory and pavement conditions data availability vary by road class. In a recent survey conducted as part of NCHRP Synthesis 501, all DOT respon- dents had inventory and conditions information for the NHS, which is consistent with federal reporting requirements for the Highway Performance Management System. On local roads, on the other hand, only 11 out of 41 respondents had full coverage of inventory information and even fewer (6 out of 41) had complete pavement conditions data (Zimmerman 2017). Coghlin described how in many cases LVR investment strategies are based on “informed estimates rather than actual data” because of incomplete coverage in inventory, traffic, condition, and accident data, and that this, along with limited analytical resources for LVRs, make decision making all the more challenging (Coghlin, page 2, n.d.). While data availability and quality are likely to be more severe challenges in developing countries than in developed countries, they still influence LVR prioritization within the United States. For example, South Carolina’s prioritization pro- cess for non-federal aid secondary pavement improvement projects has a provision within their scoring system to correct for discrepancies between actual pavement conditions as observed by district engineers and the most recent data collected by the South Carolina DOT on pavement conditions (South Carolina DOT 2017a, 2017b, 2017c). In addition to data adequacy challenges, there needs to be sensitivity to strategic issues. The literature also cites difficulties in applying traditional evaluation methods to LVRs due to their inadequate sensitivity to broader strategic goals that provide the justification for LVR investments. In seeking an appropriate road management system for the Tanzanian National Roads Agency, Benmaamar found existing methods unsatisfactory for very low-volume roads (<50 AADT) (2003). These roads, he argued, merit special attention to strategic issues such as their economic and social importance for rural populations (2003). The roads provide access to

Review of the Literature 19 markets, alternative income sources, schools, health facilities, and other social activities. They also play a central role in increasing opportunities for disadvantaged communities. Benefits of investments are related to basic access, which is difficult to describe in monetary terms, and therefore requires new and different evaluation approaches (Benmaamar 2003). Kemp et al. in Australia echoed this argument, stating in a study prepared for Austroads that “typical road project evaluations are inadequate for assessing and prioritizing regional and remote road project funding” (page iv, 2016). Estimates of travel time and vehicle operating costs savings primarily drive typical BCA. This is a problem according to Kemp et al. because travel time and operating costs saving are not the primary reasons for investing in remote and rural roads. Rather, analysis methods need to address investment objectives such as improving accessibility and therefore reducing logistics and freight delivery costs, or providing community health and educational benefits. These objectives are more difficult to convert into monetized units and therefore Australian practice excludes (Kemp et al. 2016). Johansson made a nearly identical point in research on LVRs in Northern Europe, namely, that the social benefits of LVRs to rural areas are difficult to monetize within a benefit–cost framework (2006). Relevant social benefits cited by Johansson include comfort; the ability to access education, health, or other services; influence on social life; and influence on industrial production and levels of investment within a region (2006). Following this line of research, Edvardsson described how very small improvements on heavily traveled roads yield large benefits within a traditional evaluation framework, against which even large improvements on LVRs cannot compete. He also suggests the idea of minimally tolerable conditions for safety and access (2013). Coghlin focused on volume as an inadequate proxy for roadway importance, enumerating rea- sons why existing road volume may not capture a true picture of an LVR’s role within society and the economy, including (a) low volume but high vehicle loads, (b) suppressed demand due to poor conditions, and (c) unconventional traffic mix (e.g., machinery or bicycles) (n.d.). In Iowa, Miller documented how bridge prioritization approaches that use volume as their primary decisions criteria “puts low-volume bridges on farm-to-market roads at a disadvantage in com- petition for scarce funding, as shown by the many structurally deficient low-volume bridges located in croplands of Iowa, a state whose economy is based on agriculture” (page 83, 2015a). While the research literature does propose solutions to these sensitivity challenges, the level of implementation of holistic approaches is not fully apparent from the literature alone. To that end, the next chapters in this report delve deeper into U.S. practices as reported by DOT staff. In addition, note that there is an intersection between resource challenges and methodological challenges of LVR prioritization. Low-volume roads are roads that by their nature require more than traditional engineering or BCA but at the same time, because of their status in the overall network hierarchy, fewer resources tend to be available to collect data and perform analysis for these roads. 2.4 Review of the Literature—Key Observations This chapter summarizes relevant literature pertaining to investment prioritization methods for LVRs. The following key observations are noted: • Definition. There is no single definition of a low-volume road. Volume thresholds vary from as low as 200 vehicles per day to as high as 5,000 vehicles per day. In the United States, the AASHTO definition of very low-volume load roads <400 ADT is commonly adopted. In addi- tion to volume, roads may also be classified by their location (rural versus urban), ownership, functional class, or some combination thereof. Because rural, local, and lower class facilities tend to be lower volume, practices developed for managing and evaluating rural and/or locally owned roads are often directly relevant to LVRs.

20 Investment Prioritization Methods for Low-Volume Roads • Significance. There is broad agreement across the literature that adequately maintained LVRs directly affect travel costs, improve circumstances for the economy of the served area, and enhance the social welfare of the individuals using the road. Moreover, the literature docu- ments how factors that go beyond travel cost savings are disproportionately important when considering LVR investments. These factors include trade and labor market access, access to public services, and the role of connectivity in supporting economic development. In many cases, LVRs are different from other higher volume roadways because of the high level of dependence of users and the lack of (or high cost of) travel alternatives. • LVRs in the planning process. In the United States, LVR decision making tends to be most concerned with management of an already built system, rather than with system expansion. Additionally, states have both direct and indirect control of LVRs because of the significant level of local ownership of lower volume facilities. • Prioritization approaches. Two broad categories of evaluation methods support LVR invest- ment decisions. They are prioritization based on engineering and asset management criteria (often implemented within bridge and pavement management systems) and prioritization with broader strategic criteria. The first category is the core of traditional practice, while the latter category seeks to supplement decisions with criteria to capture the full range of strate- gic objectives of LVR investments. The approaches for incorporating broader considerations typically employ a mix of qualitative scoring and quantitative indicators of a road’s strategic importance. • Challenges. LVR management suffers from funding constraints, data availability challenges, and the challenge of needing to expand beyond traditional engineering analysis (e.g., BCA) to capture strategic issues that are of known importance in LVR decision making. Practices that respond to these challenges are addressed through an investigation of North American experience in the rest of this report.

21 3.1 Overview This chapter examines the current state of agency practice concerning the prioritization of investments in LVRs. A survey gathered information on the experiences, challenges, and approaches of agencies responsible for prioritizing LVR investments. The focus was on state DOTs, but Canadian provincial transportation agencies are also included in the set of respondents. 3.2 Survey Instrument, Distribution, and Response A survey of state DOTs and Canadian provincial transportation agencies was conducted between January and April 2017. Surveys were distributed to members of the AASHTO Stand- ing Committee on Planning (representing the 50 states and the District of Columbia) and to individuals in Canadian provincial transportation departments (through the assistance of the Transportation Association of Canada). Forty out of 50 states responded to the survey for a state DOT response rate of 80%. Three Canadian provinces also responded to the survey. Figure 1 shows the survey questions. The survey was structured in the following ways: (1) to identify the respondents, (2) to determine the agency’s broad approach to defining and under- standing the importance of LVRs, (3) to solicit information about how agencies make LVR capital investment decisions, (4) to ask how agencies make LVR preservation and maintenance decisions, (5) to identify trends in approaches to LVR and agency interest in future improve- ments, and (6) to determine the appropriate point of contact for follow-up interviews in support of case example development (see Figure 1). The use of branching logic within the survey means that respondents only answered questions that were relevant to their state. In addition, agen- cies were invited to provide additional detail and supporting documentation of their practices. This material, along with responses to survey questions, provided the basis for identifying case examples that are further investigated in Chapter 4. Recognizing that agencies may have a variety of approaches to defining and understanding LVRs, the survey questions were designed to try to address as many situations as possible. Survey respondents were informed that if any of the questions or categories did not fit their situation, they were to use the open-ended questions to address these issues. Additionally, the survey covers capi- tal investment decision making and practices related to the allocation of preservation and main- tenance funds separately, with the understanding that different sets of tools and different groups within an agency tend to be involved in these two areas of LVR decision making. For the purposes of the survey, capital investments were defined as “construction of new or upgraded low-volume roads.” Respondents were encouraged to obtain input from others in their agency, as needed. To ensure an adequate response rate, recipients of the survey were provided with periodic reminders by e-mail and phone. The full survey and a list of responding agencies are provided C H A P T E R 3 Survey of Agency Practice

22 Investment Prioritization Methods for Low-Volume Roads Source: Survey questionnaire (Appendix A). PART I – RESPONDENT INFORMATION PART II – CONCEPT OF LOW VOLUME ROADS This section asks about your agency’s broad approach to understanding and defining low volume roads. Does your agency make a distinction between low volume roads and other roads for the purposes of planning, funding, or design? What percentage of your system is classified as low volume, or equivalent? Does your agency use a volume threshold to define low volume roads? If you answered no, how does your agency define low volume roads? For what purpose(s) does your agency differentiate low volume roads from other roads? Please score the purposes of low volume roads according to their relative importance in your state/province/jurisdictional area. PART III - CURRENT FUND ALLOCATION PRACTICES (Capital Investments) The following section asks about how your agency makes decisions regarding the allocation of funds to capital investment in low volume roads. How do you ensure that low volume roads are considered in capital investment decisions? If there is a separate process or application of criteria, what are its characteristics? What factors do you consider when evaluating capital investments in low volume roads? Please provide any additional detail regarding your prioritization methods for low volume road capital investments. PART III - CURRENT FUND ALLOCATION PRACTICES (Preservation and Maintenance) The following section asks about how your agency makes decisions regarding the allocation of funds to preservation and maintenance of low volume roads. How do you ensure that low volume roads are considered in preservation and maintenance decisions? If there is a separate process or application of criteria, what are its characteristics? What factors do you consider when evaluating preservation and maintenance in low volume roads? Please provide any additional detail regarding your prioritization methods for low volume road preservation and maintenance. PART IV – FUTURE PRACTICE FOR LOW VOLUME ROAD FUND ALLOCATION How would you characterize recent trends in the importance of low volume road consideration in your agency’s funding decisions? Is there interest within your agency in better addressing low volume roads within your fund allocation and prioritization processes? Which of the following would enable improved investment decision making for low volume roads? PART IV- CASE EXAMPLES/OTHER AGENCIES As part of this project, the consulting team will be developing case examples of select agency experience with investment prioritization methods for low volume roads. Who would be the appropriate contact person for any follow up questions regarding your agency’s experience? – – – – – – – – – – – – – – – – – – Figure 1. Survey questionnaire.

24 Investment Prioritization Methods for Low-Volume Roads – Class II and III – Federal Aid Local Roads – Essentially all non-NHS classified as low volume. – When we prioritize "regionally significant" projects, that significance is directly related to volumes. – Low-volume roads are not clearly defined within our department. It is recognized that there are several roads which would be considered low volume and are treated differently but we do not have set policies for these roads. – Low-volume roads typically have a national functional classification as collector or local. Our DOT does not have a specific agency defined term. AASHTO provides us with a definition of "ultra-low" as an ADT less than 400. – We don’t specifically differentiate low-volume roads but use functional classification and state/federal system designation to differentiate different types of roadways. – Major capital program is classified in three categories: Interstate, Primary, and Secondary. – Minor collectors, local on system roads, farm to market roads. – No, other than the federal functional class system. Note that our DOT has jurisdiction for the primary roadway system and local entities such as counties and cities have jurisdiction over the secondary system. – Noncommerce Routes – Not systemically; Low Volume Road Green Book is used for routes under 400 AADT. No statewide definition of "low-volume roads" currently exists. – Roads with more than 2,000 ADT are on a locally defined network called the PHN. A low-volume road is not on this network. – The agency has five customer service levels; these service levels are significantly influenced by AADT. We have not yet established clear policy to guide our actions relative to these service levels but the higher service rated roads do get and will continue to get more attention. – The agency has a classification of roadways referred to as Home Access Road Program Roads that are low-volume roads that do not receive dedicated funding. Funding is provided on an as-needed basis. Some of the county roads are also low volume; however, the funding stipulation does not apply to them. – The low-volume roads are not given a specific label but the volume information is often a criterion used when funding, planning, and design decisions are made. – The roadway network is divided into road prioritization categories based on ADT, ESALs, highway classification, and required frequency of rehabilitation based on pavement prediction models. – Tier 4 – Use functional classification – We designate five classifications of roadways on our state system. The lowest three tiers would be the equivalent of "low-volume roads," even though that term is not used per se. – We don't have a term for low-volume roads but ADT is in all of our funding formulas/applications. We provide specific funds to locals and counties that normally have most of the low-volume roads. – We use functional classification in a general sense to stratify investment decisions. – Local roads that generally have 400 vpd or less. Note: Low Volume Road Green Book refers to the AASHTO Guidelines for Geometric Design of Very Low-Volume Local Roads (AASHTO 2001); PHN = Primary Highway Network; vpd = vehicles per day. Source: Survey of state DOTs and Canadian provincial transportation departments. This survey has been edited for anonymization. Please provide the term used to classify these facilities in your agency, if different from “low volume roads.” Alternately, if your agency does not distinguish low volume roads, but recognizes a class of roads that are effectively low volume, please describe. Figure 3. LVR classification terms.