Small Cities USA: Growth, Diversity, and Inequality

APPENDIX: TECHNICAL INFORMATIONON DATA SOURCES AND STATISTICAL ANALYSES

155 In this book, I draw upon a wide variety of data sources for the analysis in chapters 2 through 5. I explain in this appendix where the data were collected from and the types of manipulation undertaken to make it suitable for analysis. All tables for the appendix are found on the website www .smallcitiesusa.com. Metropolitan Statistical Areas Through the book, I use “metropolitan statistical area” (MSA), “central city” and “suburbs” in reference to the places I am studying. I first outline how MSAs were used and then move onto central cities and suburbs. I drew a sample of eighty MSAs using U.S. Census FIPS (Federal Information Processing Standards) codes to designate places. I used a standard 1990 metropolitan statistical area (MSA) definition for data over time when possible. The State of the Cities Data Systems (SOCDS) generated by the Department of Housing and Urban Development (U.S. Department of Housing and Urban Development 2006) contains data from 1970, 1980, 1990, 2000 and for certain years from 2001 to 2011 for MSAs using the constant 1990 definition so that data are comparable across years. Unfortunately , the SOCDS data are not comprehensive for all of the variables that I use in my analysis; therefore, some data were collected based on the geography of the MSA definition of the year they were collected (in the case of data collected by the U.S. government). Other data, that were timeinvariant , were collected to match the 1990 MSA definitions (e.g., data on the number of educational institutions). APPENDIX: TECHNICAL INFORMATION ON DATA SOURCES AND STATISTICAL ANALYSES 156 APPENDIX To facilitate analysis over time, I did not use the Core Based Statistical Area (CBSA) geographic definition that the U.S. Census Bureau moved to in 2000. The advantage of the CBSA definition is that it brings together metropolitan and micropolitan geographic areas that formerly had existed under different coding schemes. The main disadvantage of using it is that it is impossible to use for 1970 data. I coded data for four MSAs using either a PMSA (Primary Metropolitan Statistical Area) or NECMA (New England Central Metropolitan Area) codes, although the data from these places were available in the SOCDS data. These four cities are Trenton, New Jersey; Salem, Oregon; Flint, Michigan; and Ann Arbor, Michigan. While PMSA data may include data for places that are subordinate to a larger central city, the four cities for which I used PMSA data were more than thirty miles from the nearest city with a population of 50 percent of the central city I was examining. Thus, each of these four cities fit my criterion for inclusion in the sample. To create a better understand the different geographic entities, I include the U.S. Census Bureau’s definitions (U.S. Census Bureau 1994): Metropolitan Statistical Area [MSA]. An MSA consists of one or more counties that contain a city of 50,000 or more inhabitants, or contain an urbanized area (UA) as defined by the Census Bureau and have a total population of at least 100,000 (75,000 in New England). Counties containing the principal concentration of population—the largest city and surrounding densely settled area—are components of the MSA. Additional counties qualify to be included by meeting a specified number of commuters to the counties containing the population concentration and by meeting certain other requirements of metropolitan character, such as a specified minimum population density or percentage of the population that is urban. MSAs in New England are defined in terms of cities and towns, following rules concerning commuting and population density. Primary Metropolitan Statistical Area [PMSA]. A PMSA is a sub-area of an MSA that has a population of one million or more. PMSAs consist of a large urbanized county or a cluster of counties (or, in New England, cities and towns) that demonstrate strong internal DATA SOURCES AND STATISTICAL ANALYSES 157 economic and social links as well as close ties with the central core of the MSA. When a PMSA is recognized, the entire area of which it is a component becomes a Consolidated Metropolitan Statistical Area (CMSA). All territory within a CMSA is also part of the associated PMSA. New England County Metropolitan Area [NECMA]. NECMAs are county-based alternatives to the city- and town-based MSAs and CMSAs for the six New England states...