Improving Mid-Term, Intermediate, and Long-Range Cost Forecasting for State Transportation Agencies (2020)

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.

4 1 Introduction and Background 1.1 Introduction Cost forecasting is only one step in the overall construction cost estimating process. However, the importance of its role in state transportation agencies’ (STAs) planning and programming processes cannot be overstated. Accurate cost forecasting during the early stages of the development of transportation programs is critical in making sound financial decisions and optimizing the use of limited available resources (Anderson et al. 2007). An effective projection of construction costs into the future is challenging due to many factors affecting the construction market and the high volatility in the price of some construction commodities. It becomes even more challenging when the estimating process requires cost forecasting efforts over long time periods, such as those commonly involved in transportation planning. STAs are often required to forecast construction costs over long periods of more than 10 or 20 years (Molenaar et al. 2013). Some of those forecasted cost estimates are performed at a program level, based on broad infrastructure performance and capacity goals, and calculated under several assumptions with minimal to no project-specific information. The longer the forecasting time period, the greater the estimating uncertainty (Anderson et al. 2007). Numerous changes to anticipated scopes of work, schedule, right-of-way cost/alignment, and environmental requirements occur during long cost forecasting periods (and these are just a few of the many uncertainty sources), challenging, and often refuting, estimating assumptions (Shane et al. 2009). Although extensive research has been conducted on various aspects of uncertainty in construction cost estimating, most of the existing literature is focused on short-term project estimates forecasted over 1 or 2 years. Few formal studies have been aimed towards improving the effectiveness of long-range cost forecasting. Addressing this knowledge gap could significantly boost STAs’ planning and programming capabilities, as well as their ability to ensure the effective use of public capital (Janacek 2006). This has motivated STAs to seek further guidance on the appropriate implementation of mid-term (3 to 5 years), intermediate-range (up to 15 years), and long-range (more than 15 years) cost forecasting techniques. This report aims to summarize all research efforts and findings derived from the National Cooperative Highway Research Program (NCHRP) Project 10-101. This project is intended to provide guidance for STAs to improve forecasting practices for mid-term, intermediate, and long- range time horizons to better account for cost variability, economic volatility, and risk. It should be noted this study is not addressing or providing guidance on the preparation of unforecast cost estimates (current-dollar estimates). Thus, the guidance provided in the report is based on the assumption that the agency is effectively estimating unforecast construction costs with acceptable accuracy levels. All research efforts were framed around the following three objectives: 1. Identifying and analyzing current practices used by STAs to forecast costs for mid-term, intermediate, and long-range time horizons, as well as STAs’ needs and opportunities for improvement in current cost forecasting procedures.

5 2. Conducting a qualitative and quantitative assessment of the current practices by comparing forecast cost estimates with actual cost outcomes. 3. Developing a framework to facilitate the selection of effective cost forecasting approaches at the project and program level. Chapter 2 presents research outcomes associated with the first objective listed above, and to some extent, with the qualitative part of the second objective. That chapter summarizes the state-of-the- practice in construction cost forecasting over long time horizons, defined via a thorough analysis of information collected through a comprehensive literature review, an online survey administered to STAs, and feedback provided by the AASHTO Technical Committee on Cost Estimating (TCCE), which agreed to serve as an Expert Advisory Panel (EAP) for this study. The qualitative and quantitative assessment of the effectiveness of various cost forecasting practices along different forecasting time horizons was conducted through three case study agencies: Minnesota Department of Transportation (MnDOT), Colorado Department of Transportation (CDOT), Delaware Department of Transportation (DelDOT). To effectively perform a long-range “forecasts vs. actual outcomes” comparison analysis, the research team considered it necessary to involve in the analysis at least 20-years of historical pricing data from each case study agency, resulting in significant data collection, cleaning, and processing efforts. That pricing data was collected from their bid tabulations. In order to determine the implications for an STA with a smaller amount of available historical bid data, calculations for MnDOT’s case study were repeated, assuming that only the most 10 years of data were available. A considerable portion of the research efforts was focused on addressing two critical factors in effective cost forecasting: 1) the selection of a suitable construction cost index (CCI) and 2) the appropriate analysis of the selected CCI to produce a reliable inflation rate. The CCI is intended to illustrate the past behavior of the construction market, while the inflation rate is a simplified mathematical representation of that behavior, which in a typical cost forecasting process, is expected to continue along the intended forecasting time horizon. These two factors are deemed critical since both should be properly addressed in order to effectively produce cost forecasts. There is no point in implementing a mechanism to identify the most suitable CCI if the agency does not know how to analyze it to generate a reliable inflation rate. Likewise, the skills to produce reliable inflation rates from the analysis of any CCI would not be sufficient if the composition of the selected CCI does not fairly match the scope of work under consideration. Research efforts and findings associated with the two critical factors mentioned above are discussed in Chapters 4 and 5. The former presents a methodology to analyze the suitability of CCIs as potential inputs in the cost forecasting process for a given scope of work. The implementation of this methodology is illustrated in the three case study agencies as it is applied to assess and compare the suitability of several cost indexing alternatives. Those alternatives include external and in-house CCIs, as well as an alternative cost indexing system called a Multilevel Construction Cost Index (MCCI). MCCIs were developed for each agency using the collected historical bid data. The quantitative analysis demonstrated the superior accuracy of the MCCI in tracking price fluctuations over time, as well as its ability to better adapt to different

6 scopes of work and to handle other program-/project-specific considerations. Further information about the MCCI and its development process can be found in Chapter 3. Chapter 5 discusses various approaches for the generation of annual inflation rates from CCIs as they are applied to the case study agencies. Those approaches include the use of simple and compounded inflation rates, as well as regression analysis and an alternative method proposed by the research team called Moving Forecasting Error (MFE). Cost forecasting approaches were evaluated on their forecasting accuracy and reliability over different forecasting time horizons; their ability to factor geographic considerations and program-/project-specific requirements; and their associated staffing, data, and information technology requirements. Finally, findings and observations from the case studies were synthesized into a Cost Forecasting Approach Selection Framework and a spreadsheet-based Cost Forecasting Toolkit. The framework and the toolkit are briefly described in Chapter 6, along with a discussion about other research efforts and resources aimed ensure the effective implementation of the cost forecasting approaches presented throughout this report. The Cost Forecasting Approach Selection Framework and Cost Forecasting Toolkit, as well as some technical information on the proposed forecasting approaches (e.g., mathematical and statistical procedures), are explained in more detail in a separate document entitled NCHRP Research Report 953: Improving Mid-Term, Intermediate, and Long-Range Cost Forecasting: Guidance for State Transportation Agencies. 1.2 Transportation Planning Programs and Cost Forecasting The maintenance, improvement, and rehabilitation of the national transportation infrastructure is a never-ending effort led by STAs. These efforts are carried out with financial resources administered on a fiscal year basis. However, the number of infrastructure needs at any given time exceeds the funding, staffing, and management capabilities of STAs within a single fiscal year. This often forces these agencies to commit funds from future fiscal years in order to meet these needs, which results in postponing lower priority projects. Approved projects could also be postponed to future fiscal years for non-monetary reasons such as environmental mitigation, permitting, right-of-way, utility relocation issues, or simply because they are part of a strategic schedule for a multi-year infrastructure maintenance, rehabilitation, or expansion program. In summary, even though financial resources are managed on a fiscal year basis, financial planning efforts must consider longer periods of time, sometimes covering periods of over 20 years. Construction and maintenance activities associated with these multi-year planning efforts are broken down into multiple plans/programs (hereinafter referred to as programs) varying by scope, purpose, and number of years. Different STAs could implement different sets of planning programs. An STA may be required to deal with some overlapping in scope and content between programs as a result of efforts to comply with regulations at different government levels. The following are brief descriptions of programs commonly implemented by STAs. It should be noted that descriptions for some of these programs may vary between agencies.

7  Long-Range Transportation Plan (LRTP): “A document resulting from regional or statewide collaboration and consensus on a region or state's transportation system, and serving as the defining vision for the region's or state's transportation systems and services” (FHWA 2017). LRTPs are required to cover a period of no less than 20 years (23 USC §135 2018). For metropolitan areas, LRTPs are usually referred to as Long-Range Metropolitan Transportation Plans (MTPs).  Intermediate-Range Plan (IRP): Some STAs have implemented “an intermediate-range plan forming a bridge between short-range programming […] and long-range planning” (ITD 2009). “The baseline project definition, cost, and schedule should be set prior to programming a project into the IRP or no later than before a project is included in the STIP” (Molenaar el at. 2013).  State Transportation Improvement Programs (STIP): The STIP is a mid-term transportation and capital improvements program. It “lists Federally-funded transportation projects that are located outside Metropolitan Planning Organization (MPO) boundaries” (Georgia DOT 2017).  Transportation Improvement Program (TIP): “The metropolitan transportation planning process shall include development of a transportation improvement program (TIP) for the metropolitan planning area by the MPO in cooperation with the State and public transit operators” (23 USC §134 2018). A TIP is included into the STIP directly or by reference, without making any modifications to the plan approved by the MPO (23 USC §135 2018). The TIP updating cycle must match the cycle of the STIP (23 USC §134 2018).  Transportation Asset Management Program (TAMP): “Transportation Asset Management Plans (TAMPs) act as a focal point for information about the assets, their management strategies, long-term expenditure forecasts, and business management processes” (FHWA 2017).  Bridge Management Program (BMP): The BMP is usually part of the TAMP. It is intended to manage the bridge inventory, as well as to “evaluate bridge condition, predict deterioration, and guide decision-making” (Indiana DOT 2016).  Pavement Management Program (PMP): The PMP is also part of the TAMP. The PMP: “1) assesses the current pavement condition, 2) predicts future pavement condition, 3) determines maintenance and rehabilitation needs, and 4) prioritizes these needs to make the best use of anticipated funding levels (i.e., maximizing benefit while minimizing costs)” (Mississippi DOT 2017). Some programs may be of mandatory implementation due to federal, state, or local regulations. For example, according to the US Code of Federal Regulations, when federal funds are involved, STAs are required to develop LRTPs with a minimum 20-year forecast period as well as mid-term STIPs (23 USC §135 2018). Likewise, MPOs must coordinate with their respective STAs to prepare MTPs and TIPs consistent with statewide LRTPs and STIPs (23 USC §134 2018). The Moving Ahead for Progress in the 21st Century Act (MAP-21), enacted in 2012, “requires all State DOTs to develop a risk-based TAMP that, at a minimum, addresses pavements and bridges on the National Highway System” (NYSDOT 2014). The TAMP must be reviewed and approved

8 “not less frequently than once every four years” (23 USC §119 2018). To comply with this act, some STAs have broken down their TAMPs into multiple smaller programs, such as bridge and pavement management programs. TAMPs are not exclusively intended for budgeting purposes. In fact, they are mainly aimed to monitor the physical condition of existing infrastructure assets, predict deterioration, and coordinate maintenance and construction activities across the state. However, TAMPs are required to include a financial plan and lifecycle cost analyses (23 USC §119 2018), whose effectiveness relies on the agency’s cost forecasting practices. This is where this study will contribute to the effective development of TAMPs. It must be noted that some states are still in the process of developing their TAMPs. IRPs are not mandated by federal regulation, but they may be implemented by some agencies to comply with state/local statutes. This study found that at least eight STAs are currently using IRPs. As stated by the Idaho Transportation Department (ITD) (2009), an IRP works as a “bridge between short-range programming […] and long-range planning.” This means that when STAs do not use an IRP, projects are moved from the LRTP directly into the STIP (Molenaar et al. 2013). Less formal version of IRPs can also be found in virtually all STAs in the form of major investment studies, corridor plans, concept studies, and needs analysis, among others. LRTPs, IRPs, and STIPs also define three important milestones in the cost estimating process of transportation construction projects before entering into a short-term planning phase. The estimation of transportation construction costs is an iterative process that occurs at multiple points during the project life cycle (Schwaber 2003). As shown in Figure 1.1, the longer the time horizon, the greater the estimating uncertainty due to the fact that many factors influencing construction costs are undefined at early project development phases (Touran and Lopez 2016). In fact, specific projects are usually undefined and unknown in an LRTP, and when identified, they are defined at a conceptual level. Cost estimates in LRTPs are usually presented on a lump sum basis or broken down into broadly defined goals, such as “maintain state of good repair for existing state-owned bridges” (TxDOT 2015). As a program evolves into downstream planning and programming activities, more details are available to facilitate and define specific projects, and consequently, allowing for more accurate cost estimations (Gransberg et al. 2015).

9 Figure 1.1 Cost Forecasting Uncertainty over Time Figure 1.1 illustrates the increasing cost certainty experienced by transportation programs as they move from a long-range planning stage to their actual execution through the award of construction contracts. Additionally, this figure links the different forecasting time horizons with their respective estimate range based on the AASHTO Practical Guide for Cost Estimating (Molenaar et al. 2013). The “Estimate Range” in this figure refers to estimating accuracy. Thus, early in the planning phase, when developing LRTPs for time horizons over 20 years, construction cost estimates are expected to be either as low as half of the actual construction cost at program completion, or as high as twice that amount (see Figure 1.1). Although the AASHTO Practical Guide for Cost Estimating is not completely clear about the sources of uncertainty considered in the proposed estimate ranges, they seem to represent the total uncertainty accumulated throughout the entire cost estimating process. The quantitative analysis presented in this report includes the calculation of similar estimate ranges for the case study agency, but mainly for the portion of the total uncertainty contributed by the cost forecasting process.

10 1.3 Current Practice vs. Ideal Practice The current state-of-the-practice of construction cost forecasting is further discussed in Chapter 2, but Figure 1.2 could be considered a fair representation of the typical cost forecasting process currently implemented by STAs. In general, once the scope of a given program has been defined, a cost estimate in current dollars is performed, which is projected into the future using a given inflation rate. Ideally, the inflation rate should be determined as a function of the intended scope of work, but that does not seem to be the case among STAs. In fact, a number of agencies use standard one-size-fits-all inflation rates to forecast costs for all transportation programs regardless of their anticipated scopes. Figure 1.2 Current Typical Cost Forecasting Process A few agencies estimate inflation rates using CCIs to identify market trends; however, the CCIs being used are also usually one-size-fits-all indexes or CCIs with calculation inputs that do not align or are completely unrelated to the intended scope of work. The mathematical procedure to apply the inflation rate across the desired forecasting time horizon depends on the type of rate (i.e., fixed simple, fixed compounded, or variable), but regardless of the calculation approach, it usually yields a single-value estimate that ignores the unavoidable uncertainty inherent in the cost forecasting process. The research team found that the longer the forecasting time horizon, the less likely formal risk analysis methods are used by STAs to account for estimating uncertainty. TCCE’s input and other effective practices found in the literature were used to design the ideal cost forecasting process shown in Figure 1.3. It was concluded from a discussion with TCCE members that an ideal cost forecasting system should be able to handle different scopes of work at various levels of detail and for different forecasting time horizon. STAs are dealing with a certain degree of variability in the level of detail in the scope of construction activities forecasted across long time periods. For example, LRTPs usually involve broad scopes of work, but sometimes, they could include specific capital projects defined at a higher level of detail, and whose associated costs are forecasted over 20-25 years. TCCE members also indicated that forecasting a capital project expected cost over a 20-year period does not necessarily mean that the agency is planning to execute the project in approximately 20 years. Rather, it means that decision-makers are considering to execute the project within the next 20 years. It could potentially be approved and awarded in 15 years or less. Therefore, it would be more appropriate to provide decision-makers Define  Program Scope Estimate Construction  Costs in Current Dollars Define Inflation  Rate  One‐size‐fits‐all inflation  rate or derived from a  non‐scope‐based CCI Forecasted Cost Estimate  (single value)

11 with a forecasting timeline showing the progression of the cost forecast as it moves across the desired forecasting time period. Finally, the TCCE highlighted the importance of producing risk- based outputs to account for estimating uncertainty and to facilitate the communication of such uncertainty to different types of stakeholders and decision-makers. This led the research team to propose the risk-based forecasting timeline shown in Figure 1.3 as the ideal cost forecasting output. Figure 1.3 Ideal Cost Forecasting Process From the Federal Highway Administration’s (FHWA) perspective, an ideal cost forecasting system would use in-house historical cost data as the main reference for the determination of applicable inflation rates. “Local historic cost data and experience with cost inflation are valuable data sources for use in projecting future rates” (FHWA 2017). This logic explains the use of in- house historical cost data suggested in Figure 1.3. There has also been a concurrent study sponsored by the FHWA aimed towards developing a methodology to improve cost estimating accuracy and reliability through the appropriate use of CCIs. This study has recognized the need for flexible cost indexing methodologies that allow the customization of CCIs to the specifics of each project, such as the scope-based CCIs shown in Figure 1.3 that facilitate the generation scope-based inflation rates. Cost indexing systems with that level of flexibility, and built through a similar methodology, have been developed by two separate studies conducted for MnDOT (Gransberg and Rueda 2014) and the Alabama Department of Transportation (ALDOT) (Pakalapati and Rueda 2018). Those studies have demonstrated the ability of an innovative cost indexing system to overcome the limitations of traditional CCIs. This innovative system is called Multilevel Construction Cost Index (MCCI). An MCCI consists of a group of indexes organized in a multilevel arrangement, allowing to forecast each individual cost element in a program or project with the MCCI index that best matches its scope. Costs for different programs/projects are forecasted with different sets of indexes, offering great flexibility to customize the forecasting process to the specifics of each scope of work. Although the MCCIs developed and evaluated in this study follow a similar arrangement of multiple indexes, the research team has taken advantage of this opportunity to improve the methodology used in the previous studies by proposing a more effective method to calculate and update index value. This approach is considerably different from the one used in the previous two studies, and this is the first time that the MCCI methodology is tested for a forecasting application.

12 Additionally, the research team has designed and applied a more reliable methodology to assess the forecasting performance of cost indexing systems. As explained in Chapter 4, it was used to find the best MCCI configuration and to evaluate the suitability of external and/or in-house CCIs. Thus, the STA that decides not to implement the propose MCCI system could still use this methodology to identify the most suitable non-MCCI alternative. 1.4 Cost Indexes and Inflation Rates The study found that, regardless of the length of the forecasting period, there are two essential elements involved in the cost forecasting process: CCIs and annual inflation rates. CCIs are time series aimed to quantify average price fluctuations in the construction market over time. Inflation, on the other hand, refers to the overall increase in the price of goods and services at a microeconomic (for an individual, group, or industry) or macroeconomic (national economy) level (Munday 1996). Thus, an inflation rate is the average measure for that increase during a given period of time (e.g., annual inflation rate). A negative inflation rate is called deflation, and it corresponds to an overall decrease in the price of goods and services under consideration. Calculating an inflation rate for a single item or commodity is a fairly easy task. It becomes more difficult when the needed inflation rate is intended to represent a group of items or commodities. The following is a hypothetical example used to explain the difficulty in the calculation of the latter. An STA estimates that the price for Commodity A has increased 200% over the last 5 years, while the price for Commodity B decreased by 10% during the same period of time. What is the combined inflation rate for these two items (single rate)? The considerable increase in the price of Commodity A could suggest a positive combined inflation rate. However, what if Commodity B is asphalt and Commodity A is steel. As occurs with most STAs, asphalt is the most relevant material for the agency under consideration in terms of cost, while the cost of steel has a considerably lower impact on the STA’s budget. In order to calculate the combined inflation rate for this example, it is first necessary to determine how much more relevant Commodity B is compared to Commodity A. After determining the relative relevance of each commodity, the agency would need to find a mechanism to facilitate an “apples-to-apples” integration of these two commodities. That mechanism is a CCI. Cost indexes can track prices for a single item or can be designed to integrate multiple goods and services into a single economic indicator taking into consideration the level of relevance (relative weight) of each item. That is called a composite CCI. Thus, a composite CCI between Commodities A and B could be developed and analyzed to define current market trends that would provide an overall combined inflation rate for both commodities. Macroeconomic inflation rates in the US are commonly estimated using the Consumer Price Index (CPI) published by the Bureau of Labor Statistics (BLS). The CPI is calculated from monthly price fluctuations of about 80,000 items in a market basket of goods and services purchased by urban consumers (BLS 2018a). To have an idea of the wide range of items used in the calculation of the CPI, it includes items such as milk, shampoo, rent, household keeping supplies, apparel, gasoline, medical care, recreation services, college tuition and fees, and funeral services (BLS 2018b). The

13 Bureau of Economic Analysis (BEA) maintains a similar broad index called the Personal Consumption Expenditures (PCE) price index. The PCE price index is calculated for a slightly different market basket using other quantitative methods and under different assumptions than those applied to the CPI, but it is still based on a broad set of goods and services regularly consumed by the general public. Despite the fact that the CPI and PCE are not calculated with construction-related inputs, they seem to be a popular option among STAs to support cost estimating processes. A further discussion on the use of the CPI or PCE for cost forecasting purposes is included later in Section 2.4. Two separate indexes developed for the same purposes, such as the CPI and PCE, are likely to yield different inflation rates (Rueda 2016). Such differences would depend on the data source, index composition, and index calculation approach, and might pose a dilemma for STAs when a CCI needs to be selected. Rueda and Gransberg (2015) approached this dilemma by suggesting that the most suitable index is the one that best satisfies two principles: the matching and proportionality principles. 1.4.1 Matching and Proportionality Principles Even though the use of in-house CCIs is the preferred approach recommended by the FHWA, the research team decided to consider the MCCI methodology as an alternative cost indexing technique since it is designed to overcome some limitations associated with the implementation of traditional CCIs. Some of those limitations are discussed by Rueda and Gransberg (2015), who introduced the matching and proportionality principle. According to their study, these two principles are frequently violated when using traditional CCIs for cost estimating purposes at the program or project level. The matching principle refers to the degree of similarity between the components used in the calculation of a CCI and the scope of the program or project to be forecasted. Once the matching principle has been fairly met, the proportionality principle appears. It refers to the degree of consistency between the relative weights of index components and the actual relevance of the same components in the intended program/project. Therefore, an ideal, but unlikely scenario, would be one in which each cost element in the program is represented by an input element in the CCI, and the relative weight of each element is the same in the CCI as in the program. It should be noted that a violation of the matching principle implies a violation of the proportionality principle since not sharing the same components would make it impossible to match the weights. Moreover, while there is great variability in the scope and configuration of programs and projects within an STA, the set of input components in a typical CCI usually remains unchanged over time. This means that the matching principle cannot always be met. The MCCI methodology is designed to better meet the matching and proportionality principles by providing STAs with the ability to adjust the configuration of the MCCI to align with the scope of the intended program.

14 1.4.2 Types of Inflation Rates Inflation rates are used to estimate future construction costs in “year of expenditure dollars.” Basically, when used in cost forecasting, an inflation rate is intended to represent an anticipated future trend in the construction market inferred from the analysis of relevant historical data. However, there are different approaches that can be used to approximate and incorporate inflation rates into the cost forecasting process. This study has identified two main types of inflation rates:  Fixed Annual Inflation Rate – Simple (Not Compounded)  Fixed Annual Inflation Rate – Compounded Annually Some STAs set fixed annual inflation rates, which represent the average expected annual growth in construction prices during the forecasting time period. This is a common practice among STAs, but not all STAs apply the fixed inflation rate in the same way. Some apply a simple inflation rate while others prefer a compounded inflation rate. When a simple inflation rate is used, the projected cost is increased by the same number of dollars every year, and the magnitude of the increase is equal to the cost estimate in current dollars multiplied by the fixed annual inflation rate. On the other hand, a compounded annual inflation rate is applied every year to the cumulative inflation in the previous year. Figure 1.4 shows the difference between a 5% simple and a 5% compounded inflation rate when applied to a $10-million project (current-dollar estimate) over 20 years. Even though the two curves start deviating from each other after the first year, the difference between them starts becoming evident after the fifth year, suggesting that there is no significant difference in applying a simple and compounded inflation rate for mid-term forecasts. The difference between the approaches increases as the forecasted time horizon is extended. This is always the case for positive fixed annual inflation rates, which is the common assumption made by STAs. Even though it is not unusual for STAs to experience deflation in their construction prices due to the drop in the price of key commodities, these are usually short-term downward trends. As shown in Figure 1.4, a fixed compounded inflation rate is more suitable when an exponential increase in construction prices is expected or assumed while a simple inflation rate assumes a linear growth trend.

15 Figure 1.4 Fixed Compounded Inflation Rate vs. Fixed Simple Inflation Rate 1.5 Organization of the Report This report has been organized into six chapters, as follows:  Chapter 1 presents some background information to facilitate a better understanding of the purpose and content of the report. The chapter includes general information on transportation planning programs, a brief comparison between current and ideal cost forecasting practices, and some relevant information on the use of CCIs and inflation rates.  Chapter 2 discusses the state-of-the-practice in long-term cost forecasting for transportation construction projects and programs. Information presented in this chapter is the result of a comprehensive literature review on current practices used by STAs and by practitioners in other industries, an online survey administrated to STAs, and feedback provided by the EAP.  Chapter 3 discusses the objectives and methodology followed by the research team to conduct case studies with MnDOT, CDOT, and DelDOT. The chapter also presents some general information on the development and use of MCCIs.  Chapter 4 details a process to conduct a comparative suitability analysis to identify the most suitable cost indexing alternative for a given scope of work. The process is explained as it is applied to identify the best cost indexing alternative for different geographic regions associated with each case study agency.  Chapter 5 continues with the case studies by applying various forecasting approaches on sample scopes of work. The purpose of this chapter is to illustrate the application of different cost forecasting approaches, as well as to assess their ability to handle mid-term, intermediate, and long-range forecasting periods.

16  Chapter 6 discusses the efforts made in this study to ensure that all cost forecasting practices and techniques resulting from this study were practical, align with the current needs of the transportation construction industry, and were of easy implementation by state transportation agencies. Those efforts include project tasks intended to gather feedback and suggestions from subject matter experts and the development of tools that have been made available to STAs to facilitate the effective implementation of all guidance and cost forecasting methodologies generated from this research project.