This project was made possible by a grant from the Center for Rural Pennsylvania,
a legislative agency of the Pennsylvania General Assembly.
Responsibility for this work and its policy implications remain with the author.
Special thanks to research assistants Christopher Collins and Matthew Dubowski,
whose able assistance and attention to detail added greatly to this research.
Published in the
International Regional Science Review
Vol. 26, #1, pp. 86-116 (January 2003)
Is the Cost of Living Less in Rural Areas?
There seems to be a general presumption that rural areas benefit from lower costs of living than urban areas. However, there is relatively little systematic data, other than anecdotal evidence, to support this. To test this presumption, this study develops spatial cost of living estimates for each of the 67 counties of Pennsylvania. In addition to the overall cost of living, it generates indexes for each of six component subindexes: groceries, housing, utilities, transportation, health care and miscellaneous goods and services. These estimates find that the average urban resident of Pennsylvania pays about 6% more than rural residents for a broad basket of goods and services. Moreover, urban residents pay more on average for all six major categories of goods, with the greatest difference (12.7%) occurring for housing costs. Interestingly, the paper also finds that the efficiency of the local public sector can affect local cost of living. The paper also identifies policy implications of these differences for economic development and other purposes.
Is the Cost of Living Less in Rural Areas?
There seems to be a general presumption that rural areas benefit from lower costs of living compared to urban areas. However, there is relatively little systematic data, other than anecdotal evidence, to support this.
For example, McHugh (1990) says: "While it is known that cost of living differences between rural and urban areas exist and that they affect the pattern of economic development, there is no consistent and comprehensive measure of these cost differences currently available." More recently, Isserman (2001, p. 45) says that rural America offers: “lower land costs, lower building costs, lower housing prices, lower labor costs, lower security costs, lower parking costs, and lower taxes.” But he cites no data to support this, perhaps because it seems self-evident.
Lower costs in rural America seem intuitive, but might there be other costs that are higher in rural areas, more than offsetting the lower costs that Isserman cites? McHugh (1990) found that rural households spent more than urban households on some items, such as transportation, health care and tobacco. And Marshall (2001) points out that we should expect rural telecommunications services to be more expensive due to greater distances, lower population densities and the inability to take advantage of economies of scale. Kurre (2000) found that rural families commuted farther in 1990 than their urban counterparts, presumably incurring higher money travel costs (if not time costs, since they might be expected to face less congestion.) Lack of scale in rural areas might also be expected to result in fewer local sellers of many goods and services, and so less competition and less competitive prices.
It seems, then, that an a priori case could be made either for a lower or a higher cost of living (COL) in rural areas.1 Given this, it is logical to turn to the data. The problem is that there is relatively little systematic data about rural costs of living. The work that has been done on spatial cost differences generally (not just urban-rural) leads to broad agreement that the cost of living varies from place to place within the nation. Virtually every study that looks at spatial cost of living finds significant differences between places. For example, Walden (1997, p. 237) says: "It is now well-established that prices vary between states." McMahon (1991, p. 426) says: "Significant differences in the cost of living exist among different parts of the country, as well as among different rural and urban counties of the same state." And studies of interregional income variations and convergence have found that cost of living differences play a major role (for example, DuMond, Hirsch and Macpherson 1998, Wojan and Maung 1998, Walden 1997, Deller, Shields and Tomberlin 1996, Bishop, Formby and Thistle 1994, and Eberts and Schweitzer 1994).
Unfortunately, data on COL variation are relatively scarce. There is currently no official government program in the United States to provide information on this important topic. Johnston, McKinney and Stark (1996, p. 568) say: "There is without doubt a need for data on regional variations in prices or costs of living." Deller, Shields and Tomberlin (1996, p. 110) say: "our findings…are… limited by the availability of good regional price data. …regional scientists need to develop a research program to address the shortcomings of our data." And Koo, Phillips and Sigalla (2000, p. 135) say: “There is a great need for information about regional COL’s.” But a panel of price research experts convened in 1999 by the National Bureau of Economic Research to “…present their individual views on what research agendas they would propose as meriting highest priority over the next 20 years” (Abraham et al. 2000, p. 31) did not even mention the topic of spatial price comparisons. This is very disappointing to regional scientists.
The private sector has responded to some extent, though, to this lack of data. Runzheimer International makes a substantial part of its revenues from estimating living and travel costs in different locations worldwide. Over 2,000 clients pay for this kind of spatial COL data from them, and according to Runzheimer over half of U.S. companies pay salary differentials based on geographic COL differences. (Runzheimer, 1998) The Federal Government also adjusts salaries for some of its employees based on COL differences, despite having no official COL statistics available to the public. (U.S. Office of Personnel Management, 1997 and 1999.)
Probably the most widely available source of data on COL differences comes from the American Chamber of Commerce Researchers Association (ACCRA), which has been publishing data on cost of living differences in American urban areas since 1968. (The ACCRA data are explained in detail in the next section.) More recently, other firms have begun to provide COL data on the web. HomeFair (http://www2.homefair.com) offers comparison of costs in pairs of cities, including some foreign cities, in their "Salary Calculator." HomeFair bases its COL comparisons on data collected by its own staff and that of the Center for Mobility Resources. Its five major categories for U.S. data are housing costs (33%), utilities (8%), consumables (16%), transportation (10%), and other services (33%). Taxes are not included.
DataMasters (http://www.datamasters.com) offers similar bilateral COL comparisons, using the ACCRA data as the basis for their own data. But DataMasters expands on the ACCRA data by incorporating information on tax differences between locations, which are conspicuously absent from the ACCRA database. DataMasters claims that this results in as many as two-thirds of the index values decreasing, some by double-digit amounts. And ReloSmart at VirtualRelocation.com provides a similar service in its Cost of Living Tool, but it requires detailed information on the user’s income, expenses, mortgage, and even expected interest rates to provide a customized COL value for each user.
All of these websites typically provide comparisons only for pairs of cities, rather than data for all places at one time. And, unlike ACCRA, they are not always willing to provide detailed information about their data sources and methodology. Most importantly for our purposes, most of these COL data are for urban areas. For example, the requirements of the ACCRA program concerning the types of products and number and types of pricing locations make it very difficult for a rural area to participate. Given the paucity of rural COL data, those who wish to study rural-urban COL differentials must typically create their own estimates. The next section reviews some of those approaches.