Transportation planners take note: the National Highway Construction Cost Index is replacing the Bid Price Index as the national barometer for analyzing road-building price movements.

Planning for future highway construction requires estimating project costs and funding resources that can be reasonably expected to be available in the timeframe for the proposed improvements. Estimates are part of including projects in programs such as the statewide and metropolitan transportation improvement programs. After projects are in the plans, State highway agencies typically prepare a set of engineering plans, specifications, and estimates for the work to be accomplished. The procurement and installation of these materials become the State-defined bid items for the construction contract.

Construction companies interested in the work submit bids in which the price associated with each item includes the cost of the material itself; all other associated costs for moving, placing, and installing the material; and a component of profit and overhead associated with each item. Clearly, in this usage, bid items include all associated costs for materials, labor, equipment, and a proportional share of project overhead and profit.

For decades, the Federal Highway Administration's (FHWA) tool for determining the expected purchasing power of the financial resources available to highway agencies was the Bid Price Index (BPI), which was a compilation of nationwide data on highway construction prices. Transportation planners employed the BPI for estimating the costs of road-building projects, and FHWA used it in various analyses. But the BPI imposed excessive paperwork requirements on State departments of transportation (DOTs) and needed more data sources in order to calculate trends and perform statistical analysis.

Between 2006 and 2010, FHWA developed and deployed a new approach to generate cost compilations. The new National Highway Construction Cost Index (NHCCI) is replacing the BPI as the national database for determining the purchasing power of road-building resources. The new tool provides an understanding of historical and recent price trends in highway construction.

"For various reasons, State transportation agencies didn't use the BPI, so we knew improvements were needed," says King W. Gee, associate administrator of the Office of Infrastructure at FHWA. "The new NHCCI addresses a number of issues. This index is important to contracting agencies that are trying to develop an accurate estimate of the cost of new projects."

Origin of the Bid Price Index

In 1933, the Bureau of Public Roads (now known as FHWA) began compiling and issuing the quarterly BPI as a continuing record of nationwide construction costs, published as graphs and tables. Using recorded price trends, FHWA then extended the index back to 1922, when highway construction in the United States had reached sufficient nationwide volume and become standardized enough to provide a fairly reliable basis for comparing construction costs and price movements.

The BPI was calculated using data obtained from the award of contracts by State transportation agencies on Federal-aid highway projects, excluding contracts on the Federal-aid secondary system. Beginning in 1977, FHWA sought to reduce the paperwork burden on States by collecting data only on contracts greater than $500,000 on the National Highway System. Until 2007, FHWA collected the data on form FHWA-45, entitled Bid Price Data, and form FHWA-47, entitled Statement of Materials and Labor used by contractors on highway construction involving Federal funds. Form FHWA-45 recorded information on bid prices for major work items on Federal-aid highway construction contracts. Form FHWA-47 recorded quantity data on labor and materials used on those contracts.

How the BPI Was Used

FHWA used the data provided on the two forms as input to the BPI, which was published quarterly in the Price Trends for Federal-Aid Highway Construction. In addition to States applying the pricing information, the BPI was used in the following publications:

• Status of the Nation's Highways, Bridges, and Transit: Conditions and Performance (biennial report to Congress)
• Highway Statistics (an annual FHWA report)
• Our Nation's Highways (an FHWA report)
• Special studies of various material prices by FHWA and other organizations

FHWA also used the BPI in developing responses to policy issues related to employment and materials usage. In addition, the Bureau of Economic Analysis used the BPI as a price deflator in the production of the National Income and Product Accounts, which include many of the key aggregate variables, such as wages and market value of goods, used to describe the U.S. economy.

Concerns About the BPI

However, the BPI underwent criticism from within FHWA and from the U.S. Government Accountability Office (GAO) in its report GAO-04-113R Comparison of States' Highway Construction Costs. GAO recommended "that the Secretary direct the Federal Highway Administrator to determine whether it would be useful and feasible to collect and disseminate other [S]tate construction cost data that could supplement or supplant FHWA's bid price data."

This recommendation accelerated efforts underway at FHWA to discontinue and replace the index. The agency had found several reasons to question the tool. First, forms FHWA-45 and FHWA-47 created an excessive paperwork burden on the States. Form FHWA-47 required States to convert inputs into standard units that often did not match the units used in construction bids. For example, bituminous material on Form FHWA-47 was reported in gallons, but many States' bid items were based on area measurements rather than material quantities, thus requiring a conversion to meet the form's requirements.

Second, the reporting forms were not timely because they were due after completion of highway projects so all price changes could be integrated. Unfortunately, this requirement reduced responses to the survey because most State DOTs tended to move onto the next project rather than finishing Federal paperwork on completed projects.

Third, GAO contacted 12 States and found that they did not use the BPI because the DOTs collected more detailed information than what was included in the Federal index. The GAO report reinforced what many at FHWA believed: that few States actually used the BPI data.

Furthermore, the quality of the BPI data was below U.S. Department of Transportation standards as expressed in the 2003 Guide to Good Statistical Practice in the Transportation Field. Reviews of data reported on forms FHWA-45 and FHWA-47 showed uneven data quality. The process used to develop the BPI did not allow statistical examination, and only limited data review was possible.

In April 2007, FHWA discontinued the collection of forms FHWA-45 and FHWA-47 because of the data quality issues and concerns about the work burden on States. The final BPI report was for the fourth quarter of calendar year 2006.

Development of the Replacement

Coordinated research into a replacement for the BPI began in 2000. FHWA evaluated 14 databases but found none that could be used to replace the BPI or serve as the basis for creating a new tool. An outside review by an FHWA contractor proposed a new detailed data collection process with broader scope than forms FHWA-45 and FHWA-47 but with fewer data providers. FHWA rejected this proposal as being both too costly to implement and presenting a small likelihood of success given the regulatory hurdles to issuing new surveys.

Ultimately, FHWA decided to purchase a database from Oman Systems, Inc. (OSI) to provide construction cost data for the NHCCI. This database captures Web postings by States on bids submitted on highway construction contracts and contains all States except Alaska and Hawaii. Some States have data in the database back to the mid-1990s, but others start in 2000 or later.

Steve Hemphill of the New Mexico Department of Transportation says that he uses the database to "sort through historical prices for bid items by date, district, geographic region, size of project, quantity window, etc., for mean and standard deviation information." The detail available makes it a robust database for creation of a price index.

National Highway Construction Cost Index

The new NHCCI quarterly construction cost index is replacing the BPI with data collected starting March 2003. At that point, that database's owner was gathering information from 45 States, including the largest ones. The NHCCI overlaps with the BPI for 2003 through 2006, but the two datasets and index structures do not match closely enough to allow for a continuous index. However, the NHCCI does allow analysis of annual and quarterly price changes.

For FHWA's purposes, the relevant information included with each pay-item is State, bid price, unit of measure, general expenditure category, and date the contract was awarded. The NHCCI combines prices of individual goods and quantity weights to track the percentage change in prices over time for a particular selection of goods and services. Implicitly, the quality of goods and services represented in a given timeframe is assumed to be constant. In the NHCCI, individual goods in the data are represented by "pay-items" for successfully bid contracts. Pay-item definitions differ from one State to another, but FHWA overcomes this problem by holding definitions constant within each State and then creating the national index, which does not compare State prices but does arrive at a national average. The differences in State-defined pay-items mean that a user could not compute comparable State indexes.

The NHCCI procedures are different from those used by the U.S. Department of Labor's Bureau of Labor Statistics (BLS) to create the former Producer Price Index (PPI) for Highway and Street Construction. For that index (discontinued in July 2010), BLS contacted input sellers such as steel and cement manufacturers to obtain their current sale prices. BLS then employed the Bureau of Economic Analysis benchmark input-output relationships to construct the PPI for highway and street construction. Therefore, the NHCCI could be thought of as a construction (or output) cost index, whereas the PPI was a sales (input) price index.

The NHCCI differs from both the PPI and BPI since it uses a Fisher index methodology, which embodies the idea that, for a fixed market basket, changing relative prices will lead to changes in the relative quantities being purchased in the basket as entities make substitutions within an item category. This concept is easily illustrated by an analogy to grocery shopping. When a shopper notices that blueberries have doubled in price, the consumer will shift to another choice, say bananas. Over time, then, the market basket is changing, and the Fisher index recognizes this explicitly by using the average market basket rather than a fixed base-year market basket. The BPI attempted to account for the changes in the market basket of construction goods by updating the market basket of highway construction materials every 10 years. This process was so time-consuming that the BPI's market basket had not been updated since 1987.

Both FHWA's BPI and NHCCI reflect the changing mix of inputs over time. Since the final product measured by the NHCCI is highway construction, it is appropriate that the index reflect a mix of inputs. FHWA intends to monitor how the mix of goods in the NHCCI changes over time. One of the advantages of the Fisher index is that the market basket is updated throughout the index, removing the inherent bias that would otherwise occur if the mix of goods changes over time. A description of the mathematics of a Fisher index is available at www.fhwa.dot.gov/ohim/nhcci/math.cfm.

Businesses use the NHCCI to gauge the market. "An output price measure such as the NHCCI is valuable for demonstrating that highway costs may be escalating much faster than consumer prices or even a broad producer price index," says Ken Simonson, chief economist for The Associated General Contractors of America. "In addition, some contractors may use the index to compare the change in their bid prices [relative] to the overall market."

Construction Cost Index Data

Price indexes are only as good as the data employed to construct them. The proprietary database used to build the NHCCI contains a large volume of data on State roadway projects. The universe of data is larger than the BPI's, which covered only Federal-aid contracts over $500,000 on the National Highway System. The NHCCI is intended to cover all highway projects and therefore arrive at an average cost index for all road construction.

FHWA's objective in constructing a highway cost index is to provide a measure that is useful not only as a national cost index for constructing highways, but also as an index that the States can use to gauge differences in their relative cost experiences for highway construction. The NHCCI is not intended to replace State-specific cost indexes, which offer two key advantages. First, they help States monitor how quickly unit costs are rising to better match budget appropriations with actual expenditures. Second, they provide States with better information for planning and budgeting a statewide program of projects.

State DOTs and metropolitan planning organizations (MPOs) are responsible for developing the Statewide Transportation Improvement Programs (STIP), the Transportation Improvement Programs (TIP), and metropolitan transportation plans (MTP) that identify projects that they propose for Federal aid funding. The STIP/TIP and MTP contain project cost data and revenue estimates. Since 2007, the project cost data and revenue estimates in these documents are required to be reported in year of expenditure (YOE) dollars. The NHCCI provides information that planners and project staff can utilize to track cost trends and develop reasonable project cost estimates for use in the programming documents and the long-range transportation plans. The NHCCI provides the States and MPOs a source of data where project construction cost trends can be evaluated and used to estimate future project costs in YOE dollars.

Given the relevant indexing and economic theory, the requirements for the data are the following:

• Consistent and specific product characteristics so that price comparisons over time compare apples to apples rather than excavation in corn fields to excavation in mountains. Consistency is fundamental to creating a well-defined cost index and involves taking care in specifying what constitutes an individual product. FHWA used the guidelines developed by BLS to further inform its data analysis (www.bls.gov/ppi/ppifaq.htm). To summarize, BLS treats products that are characterized by different price determinants as distinct products. BLS has a large staff of industry experts and statisticians to design survey questionnaires to isolate different products and evaluate the responses.
• Continuity in the time series for the individual pay-items so that there are few breaks in the series. Mathematically, the index methodology requires prices and quantities to be available in consecutive periods. The requirement arises because the index computes the average change in price between consecutive periods. Products that are purchased infrequently or are inherently unique, such as traffic control devices, cannot be included in the index since prices are not comparable quarter to quarter.
• A sufficiently large or representative set of goods so that the resultant index is relevant for those who make the real budget decisions. The proprietary dataset has an advantage because it represents a virtual universe of FHWA's items of interest — the components of State highway costs. The wealth of data in the database enables FHWA to use alternative methods to arrive at the same objective: a reliable indicator of highway construction costs that can be used for both general cost comparisons and for States to gauge changes in their costs against national averages. FHWA's approach is to eliminate pay-items that may be defined too broadly or have statistical properties that imply variable price-determining characteristics.

Edits Used to Improve Data Quality

FHWA's approach to creating the NHCCI attempts to reliably reflect changes in the prices of the underlying goods. To achieve this goal, FHWA procedures first eliminate three categories of data.

The first is nonstandard pay-items that have the same pay-item number but different pay-item descriptions (or units of measure) from project to project. For example, on one project an item is described as Control Survey, and on another project in the same State, the description is Cleaning Existing Paved Ditch. These types of pay-items, even if they fit all the statistical criteria, cannot be included due to the differing types of work from project to project.

The second category of data is eliminated because of unit of measure problems that make it difficult to track price changes. Many of these are lump sum items where the quantity of the item is listed as "1." The prices bid on these types of items generally are not related to any specific price trend but rather are due to other factors such as project type, duration, location, size, and traffic patterns.

The third eliminated category consists of suspect pay-items in the historical databases grouped into 31 predefined work categories. Some of these categories relate to aspects of a contract such as startup costs and incentives. As with the units of measure problems, some categories relate to groups of pay-items that generally are not related to any specific price trend but rather relate to the project type, location, and size. These categories are mobilization and alternates, bonuses, and time.

Additional statistical edits eliminate pay-items that are unlikely to have consistent price-determining characteristics. These edits include removing pay-items that are not represented for eight or more quarters and those that represent outliers.

Effect of the Edits On the Data

Most of the changes in value and number of observations come from the edit that removed pay-items without eight or more quarters of data. This edit is recomputed each quarter so that, as a pay-item becomes more pervasive, it will be included in the index. This is a major advantage of using a chained-index. A chained-index utilizes expenditure data in adjacent time periods to reflect any substitution that highway departments make across item categories in response to changes in relative prices.

Data tests show that losses due to the edits leave about 40 percent of the value of all available pay-items in the final database. Another feature to note is that shares of the various categories in total value do not change much from the initial share for all retained pay-items. Only three categories change by more than 1 percent, leaving the final basket of pay-items with a similar makeup to the basket made up of all pay-items in the bid.

The number of observations falls by almost 430,000 after all the edits are incorporated. Note that the number of observations declines by a much larger percentage than the value, which indicates that the pay-items lost through the edits tend to be low-priced items.

Overall, the effect of the edits on the data, although large, still leaves a useful dataset for constructing the NHCCI. The dollar value of the retained pay-items database is large relative to the unedited database, with almost $70 billion of the initial $170 billion still available for index computation. Further, the distribution over the expenditure categories, which will determine the weights in the NHCCI, is largely unaffected by the edits. Similar statistical analysis could not be performed on the data provided by the BPI, highlighting another advantage the NHCCI has over the discontinued index.

Future Research

Although the NHCCI offers many improvements, further advances are desirable. Additional research will focus on the following areas:

1. Monitoring how the NHCCI performs with updated data, particularly related to input substitution over time.
2. Developing methodology and a software application to derive subindexes based on category groupings, such as excavation, reinforcing steel, structural steel, structural concrete, structures, asphalt, etc.
3. Developing software applications to provide indexes by project types, such as capital improvement and maintenance, if possible to identity these in the dataset, and analyzing how different project types perform over time.
4. Further investigating the randomness of the items excluded from the dataset.
5. Investigating combining the proprietary database and American Recovery and Reinvestment Act of 2009 dataset to identify and analyze selected characteristics.
6. Asking States to review and possibly better define some items within the dataset to reduce data edits.

The NHCCI is intended as a price index that can be used both to track pure price changes associated with highway construction costs and to convert current-dollar expenditures on highway construction to real- or constant-dollar expenditures.

Although many States maintain excellent cost indexes, the NHCCI provides a national view for use in models associated with the Status of the Nation's Highways, Bridges, and Transit: Conditions and Performance report and will be published in the annual Highway Statistics. FHWA will continue working with State DOT partners to enhance and improve forecasting techniques that are relevant and useful for project cost estimates.

Karen White, Ph.D., is a transportation specialist with FHWA's Office of Transportation Policy Studies. She oversees and participates in a broad variety of transportation policy issues, such as highway finance and the impacts of highways on economic productivity. White received doctorate and master's degrees in economics from the University of Houston and bachelor's degrees in economics and finance from the University of Texas at Austin.

Ralph Erickson is the leader of FHWA's Highway Funding and Motor Fuel Team in the Office of Highway Policy. He began his career with the agency as an economist in 1975 and has worked on projects providing development and implementation of new highway policy directions. Erickson served as chair of the Transportation Economics Committee of the Transportation Research Board from 1992 to 1998. He holds a B.A. in history from Otterbein University with economics graduate work at the University of Maryland-College Park.

For quarterly updates to the NHCCI and background materials with additional information, see www.fhwa.dot.gov/ohim/nhcci/index.cfm, or contact Karen White at 202-366-9474, karen.white@dot.gov, or Ralph Erickson at 202-366-9235, ralph.erickson@dot.gov.