The data used in an assessment undergoes the same scrutiny for scientific and professional integrity, as does the methodology used in obtaining and analyzing data. Data sources should be referenced in a manner consistent with the professional or scientific discipline. Although there are no prescribed data sets to be used in an assessment, they should possess the following basic attributes: relevant, current enough to be useful, reputable or from a credible source, and in a usable format. Of these, obtaining usable data is often the most problematic. All too often, existing data sets are misrepresented and turn out to be something other than what they were represented or no one has any confidence in these data. Most data used in EISs comes from established sources--the most common of these are state and federal natural resource agencies and libraries. More recently, data clearinghouses have been a source of digital geospatial data.

In most cases, there is a preconceived idea among the EIS prepares and the regional FHWA as to what constitutes "best available data." There is, however, no official definition of best available data; this is principally a byproduct of experience. Consequently, this is rarely an issue for dispute. This does not imply, however, that there may not be a better way of acquiring or analyzing necessary information, but through years of experience, EIS preparers and lead agencies have become familiar with certain data sets and have grown accustomed to their application for various assessments.

To fulfill the requirements of NEPA, engineers have to address a number of issues related to socioeconomics and environmental justice, topics that were not fundamentally a part of their usual formal training.  There are numerous sources of socioeconomic data that can be accessed via the internet from federal and state archives. Herein, we provide background information about how to obtain accurate, reliable, and complete socioeconomic data from the public domain. The orientation of this document is towards regional projects, although most of the content can be adapted for smaller areas. Given the wide range of engineers' exposure to the social sciences, it is difficult to prepare a document that is comprehensible to the newcomer, but not too superficial for more experienced workers; decision on what to include in this document are based on discussion between engineers and social scientists in formal settings together with experience of the authors in working with multidisciplinary teams in the field. Below are the specific motivations:

1. The need to include societal and institutional issues in transportation: The traditional transportation planning process accommodates the disparate views of stakeholders by a variety of means including public hearings and information sessions. These meetings serve to address economic issues, access to activities, neighborhood vitality, safety issues, availability and convenience of transportation system, and equity issues associated with costs and benefits of the system (Unsworth 1994; Balog, Schwarz, and Simon 1994; Cunningham et al. 1996). Availability of socioeconomic data for the neighborhood or the area will help reduce the amount of time and other resources employed in the planning process.

2. Environmental Justice Issues: EO 12898 requires Federal agencies to achieve environmental justice by identifying and addressing disproportionately high and adverse human health and environmental effects, including the interrelated social and economic effects of their programs, policies, and activities on minority and low-income populations. Socioeconomic data, such as race, income, population (by age variables), and employment, are readily available from Federal databases for assessment of environmental justice. Identifying the size and location of low-income and minority population groups is an important first step toward assessing whether or not transportation system investments disproportionately burden or fail to meet the needs of any segment of the population under consideration.

3. Transportation Needs Assessment: Constructing demographic profile such as age, type of employment, and of course income level is very essential in transportation needs assessment.The identification of these population groups through and their geographical locations, helps identify the transportation needs of the target people. For example, a neighborhood or community with very old and low-income people will be obviously more dependent on public transportation than a prime-age with high income one. This information will help the transportation agency provide safer, more easily accessible and user-friendly transit facilities. Data on businesses (i.e. shopping, manufacturing, and other services) is necessary to be incorporated in transportation or public transit planning.

4. Obtaining Data from Public Domain: Data used in transportation planning and regulatory processes undergoes the same scrutiny for scientific and professional integrity, as does the methodology used in obtaining and analyzing data. Data sources should be referenced in a manner consistent with the professional and scientific discipline. The data sets should possess the following basic attributes: relevant, current enough to be useful, reputable or from credible source, and in a usable format. Of these, obtaining usable data is often the most difficult. All too often, existing data sets are misrepresented and turn out to be something other than what they were represented to be or no one has any confidence in them. Other problems in data acquisition are incompleteness, previously unavailable, incompatibility, and so on. It is very difficult to obtain historical data for specific time frame from one source. You may find breaks in years, starting later years, or covering different variables or the same variables but different dimensions. U.S. Census Bureau most often covers different variables in each census years. Furtherer more, census data such as the U.S. decennial census or metropolitan area household surveys are based on once per decade or an even more infrequent basis.

Data Descriptions

1. Age: Three age groups were analysed from 1970 to the year 2000: 0 to 19 years, 20 to 59 years, and age 60 and older. Maps showing the percentage of the total population in 5-year intervals for all three age groups are shown in Table 1.

2. Population Density: Population density is measured as the number of persons per square mile and covers the period between 1970 and 2000. Maps showing percent change in population density for every five years are included and coded in similar fashion as age. For example, the change in population density from 1970 to 1975 is P_dens_70_75.

3. Per Capita Income: This also covers 1970 to 2025, where the years 1970 to 2000 are actual data and 2001 to 2025 are predicted. It was derived from total income from county by multiplying the per capita income share by total income. Percent changes for 5-year intervals are provided including maps from 1970 to 2000, using the same intervals. These percent changes are coded as P_in_70_75.

4. Employment: Employment is categorized as follows: (a) farm employment, (b) non-farm employment, (c) wage employment, and (d) total employment for the period between 1970 and 2000. Percent change in 5-year intervals is mapped.

5. Household Income: Household income is based on the value of the U.S. dollar in 1992. Data were acquired for the period between 1970 and 2000. Percent change in 5-year intervals is mapped.

6. Farm Income: This covers 1970 to 2025, and follows the same pattern as per capita income - the years 1970 to 2000 are actual data and 2001 to 2025 are predicted. It was derived from total county income by multiplying the corresponding share by total income. Percent change for every 5 years have been computed for the entire period and corresponding maps from 1970 to 2000 are available too. The 5-year percent changes are coded as i_f.

7. Business: This only ranges from 1984 to 2000 and is obtained from Pro CD Phonebooks. Pro CD Phonebook is an easy-to use CD-ROM reference tool that gives an instant access to millions of businesses and residential listings from printed phone directories across United States and Canada. They are point files (business density by county) grouped by the year of establishment in 3-year intervals as follows: 1984 – 1986, 1987 – 1989, 1990 – 1992, 1993 – 1995, and 1996 – 1998. The data was in spreadsheet dbase4 table and a script was built from event themes. This took the longitude and latitude fields in the table and used them as X and Y coordinates for a point theme for each desired county. Efforts to find a source(s) for income from these commercial businesses have proved futile, as it is not reported individually.

8. Race: This data ranges from 1970 to 1999. The portion from 1970 to 1990 is decennial data obtained from U.S. Department of Commerce, Bureau of the Census publications from Auburn University Library, University of Georgia Library, the University of Alabama Center for Business and Economic Research-Tuscaloosa, Alabama, and the Center for Demographics, Auburn University in Montgomery, AL (AUM). Consultation with all these sources reveals the original source is Census Bureau and they were gathered through door-to-door survey. The between census data were estimated using trend analysis. The growth rate between 1970 and 1980 and that of 1980 and 1990 for each race in each county was determined and divided by 10 to obtain the annual growth rate and the trend was estimated for that interval. Data from 1990 to 1999 was obtained from the Census Bureau website as annual data. County population by race has not been released yet for the 2000 census. The major population by race considered, are White, Black, and Other. Where "Other" includes American Indian, Asian Indians, Chinese, Latinos, and so on. Maps indicating the distribution of each race are provided for 1973, 1980, 1991, and 1999.

9. Housing By Construction Type: This is also a decennial data and obtained from Bureau of Census publications from Auburn University Library, Center for Demographics, Auburn University at Montgomery, and Georgia Tech, GA. Since the housing census for 2000 was yet to be released and there is no such information on the Census Bureau website, we could only obtain years 1970 to 1990. Housing census for 2000 had been released only for the first five states in alphabetical order, which Alabama is included but it was only the total number of housing, not by type and also at the state level only. The same procedure applied to the race data was followed in filling in the years between census data years. However, instead of linear trend, exponential function was most of the time appropriate. Especially, in counties where the rate was high for particular ten-year period, exponential function was used. Due to the varying objectives of each census, the census data for 1970, 1980, and 1990 were not uniformly compatible. The years 1980 matches with both 1970 and1990, but not 1970 and 1990 are not compatible. Any effort to merge these three census years resulted in a whole lot of information loss. We therefore, grouped them into two categories: 1970 to 1980 and 1980 to 1990. The 1970-80 group has the following information: 1 unit, 2 units, 3-4 units, and 5 or more units structures, and mobile homes and trailers while the 1980-90 has 1 unit detached, 1 unit attached, 2 units, 3-9 units, 10-49 units, and 50 or more units, plus mobile homes and trailers. Maps for each of these constructional types have been provided for years 1973, 1980, and 1990.

10. Housing By Value: This 1980 to 2000 residential housing and value data was purchased from U.S. Census Bureau Manufacturing and Construction Division. This place level delimited text file was imported into excel spreadsheets and all the extraneous period characters from the place field removed. Also removed, were all counties outside of the desired area and converted all files to county level in dbase4 format. However, due to the fact that there are multi-places in a county, any county shape file one place will always superimpose over the others. However, for exhibition purposes four shape files for 1980, 1991, and 2000 have been made to be compared to the satellite pictures taken for these years.

Balog, J.N., A.N. Schwarz, and R. Simon.  Involving Individuals with Disabilities in ADA  Complementary Paratransit Planning and Implementation.  In Transport Research Record 1463, TRB, National Research Council, Washington, D.C., 1994, pp. 53-60.

Cunningham, L.F., K. Christensen, D. Dunn, E. Gonzales, and M.P. Hirsch., Recommendations for Developing Customer Focus in Statewide.  Transportation Planning.  In Transportation Research Record 1552, TRB, National Research Council, Washington D.C., 1996, pp 19-26.

Hoffeld, S., L. Cobb, and D.A. Griffin.  Determining Disproportionate Impacts in Environmental Justice Evaluations, Wilmington Bypass, Wilmington, North Carolina, Proc., 22^nd Annual Conference of the National Association of Environmental Professionals, Jacksonville, FL, 1997, pp. 350-365.

Richardson, Barbara C. and Lidia P. Kostyniuk.  Methods for Including Societal Issues in Transportation Decisions. In Transportation Research Record 1626, TRB, National Research Council, Washington D.C., 1998.

Transportation Decisions.  In Transportation Research Record 1626, TRB, National Research Council, Washington D.C., 1998.

Sabol, Scott A.  Effects of Highway Bypasses on Rural Communities and Small Urban Areas.  Research Results Digest, NO 210, National Cooperative Highway Research Program, May 1996.