The More We Get Together

Regional concentration of economic activity has been one of the most important sociodemographic transformations in the United States since the Second World War. An urban research program at the Real Estate Center studied the concentration of economic activities, population and housing units as well as commercial and industrial real estate properties in Texas since 1950.

The study revealed that the relationship between productivity growth, population density and educational attainment is both a cause and a consequence of regional economic concentration in the Texas economy.

Economic Concentration in Metro Areas

The United States has 366 metropolitan areas. The top 100 cover 12 percent of the nation’s land and produced three-quarters of the U.S. gross domestic product (GDP) in 2008. The top six U.S. metros, which include Houston–Sugar Land–Baytown and Dallas–Fort Worth–Arlington, generated more than a quarter of U.S. GDP in the same year (Table 1). New York–Northern New Jersey–Long Island produced 8.8 percent of the U.S. GDP, ranking first among U.S. metro areas in terms of share of GDP produced by metro areas, followed by Los Angeles–Long Beach–Santa Ana (5.0 percent), Chicago-Naperville-Joliet (3.6 percent), Houston–Sugar Land–Baytown (2.8 percent), Washington, D.C.–Arlington–Alexandria (2.7), and Dallas–Fort Worth–Arlington (2.6 percent).

Texas currently has 25 metro areas, but the state’s top four metro areas produced about 77 percent of Texas GDP in 2008. Two metros, Houston and Dallas, accounted for 64 percent of the state’s GDP in the same year (Table 2).

On a county level, Harris County alone generated 22 percent of Texas personal income in 2007 while eight counties accounted for more than 61 percent of the state’s personal income (Table 3).

Regional concentrations of output and income in Texas have been closely associated with regional population concentration. Dallas–Fort Worth–Arlington, the fourth-largest metro area in the United States, and Houston–Sugar Land–Baytown, the sixth-largest metropolitan area, together accounted for about half of the Texas population in 2008 (Table 4). More than 70 percent of the state’s population is located in seven metro areas (Table 4). On a county level, seven Texas counties contained half of Texas’ population in 2008 (Table 5).

Regional concentration of population leads to regional concentration of housing units followed by regional concentration of commercial real estate properties. Seven Texas metro areas accounted for about 70 percent of total housing units in Texas (Table 6). About 50 percent of housing units in Texas are located in eight counties (Table 7).

Agglomeration Economics

Concentration of people and housing units in a region results from the concentration of economic activities in the area.

In urban economics, the term agglomeration is used to describe the benefits firms obtain when locating in densely populated areas or in highly concentrated markets. Study of the economics of agglomeration began in the late 19th century with Alfred Marshall, who argued that when similar firms locate near each other, the proximity encourages informational and technological spillovers. This generates higher productivity for all firms in the market because of economies of scale and network effects.

The top four metro areas in Texas provide examples of agglomeration effects. For example, 65.5 percent of business establishments with paid employees located in Texas in 2007 were in the Houston, Dallas, San Antonio or Austin metro areas (Table 8).

The Dallas and Houston metro areas were home to more than 50 percent of Texas business establishments in 2007. The Dallas metro area had 27 percent of Texas establishments with paid employees in 2007, with a larger concentration in communications equipment manufacture, computer system design and services, and air transportation services.

Houston’s share was 23.1 percent with a concentration in oil and gas production, petroleum refining, chemicals, petrochemicals, pipelines, water transportation, and computer and peripheral equipment manufacturing. The area is home to more than 4,000 energy-related firms, more than 10,000 manufacturing establishments, more than 2,000 metal manufacturing companies, and 450 chemical plants.

San Antonio’s share of total number of business establishments in Texas in 2007 was 7.8 percent, concentrated in electric and gas production and distribution firms, tourism and insurance firms. Austin’s share was 7.6 percent, mainly concentrated in electronic components manufacturing, and computer and peripheral equipment manufacturing.

The top four Texas metro areas also held similar shares of nonemployer establishments — that is, businesses without paid employees (Table 9). This group consists mostly of self-employed individuals operating small unincorporated businesses that may or may not be the owner’s principal source of income.

But agglomeration benefits cannot continually increase. At some point, they become subject to the law of diminishing marginal returns. Competition among firms in concentrated industries drives down both profit rates and profit margins. And densely populated areas have to deal with the problems of congestion and crowding. Houston was the fourth most congested metro area in the nation based on annual delay hours per traveler while Dallas ranked sixth according to a 2007 study by the Texas Transportation Institute (Table 10). Houston ranked third in terms of wasted fuel per traveler while Dallas ranked eighth according to the same study.

Productivity and Population Density

The net benefits of industrial agglomeration are divided between firms and their employees. Firms receive higher returns and employees earn higher wages. Firms pay efficiency wages — that is, wages above market clearing wages — to increase employee productivity.

One approach to forecasting the relationship between regional productivity and agglomeration is to study the relationships between average wage per job and population density across regions. This study looked at average wage per job, percent of population 25 years and older with college degrees, and population density per square mile for the top 15 most populated Texas counties (Table 11). These 15 counties were home to two-thirds of the Texas population in 2008.

County datasets are used because the Census Bureau’s most recent population density data are on a county basis. Percent of population with college degrees is an important determinant of average wage per job because wages are incomes derived from human capital and investment in education has been shown to be the most important investment to be made in human capital. Compared with the average wage per job for Texas, five counties — Harris, Dallas, Travis, Collin and Fort Bend — have above average wage rates.

Given that regional wage rates reflect regional productivity, Harris and Dallas with the highest wage rates, highest population densities, and lower than average educational attainment, have benefitted most from the positive impact of agglomeration. For Travis, Collin, and Fort Bend Counties, above average educational attainment have offset the impact of smaller agglomeration effects caused by lower population densities. Hidalgo and Cameron Counties, with wage rates about half the rates of Harris and Dallas Counties, have the lowest percentages of population with college degrees.

The fitted line on a scatter diagram shows that, in general, higher wage rates are associated with higher population density rates, but the relationship is subject to the law of diminishing marginal rates of returns (Figure 1).

Having obtained most of the benefits of agglomeration, the Houston and Dallas areas are now expected to experience diminishing marginal returns. Travis, Collin, Fort Bend, Montgomery, Williamson, Brazoria and Nueces are expecting higher income growth rates associated with higher population density.

A second scatter diagram shows the estimated regression line fitted to the wage variable and the educational attainment variable (Figure 2). The fitted line shows that, in general, higher wage rates are associated with higher educational attainment with little sign of diminishing marginal returns. The positions of Dallas and Harris Counties suggest that in addition to educational attainment there should be another factor behind the higher wage rates for these counties. That factor is the agglomeration effect (Figure 1).

Educational attainment and population density together explain about 77 percent of the variation in average wages. Calculations for population density show that if the population per square mile increases by one, say from 97 to 98 people per square mile, the average annual wage per job is forecast to increase $6.90. Estimates for college graduates show that if the percent of population with college degrees increases by 1 percent, say from 19 to 20 percent, the average annual wage per job is expected to increase by $539.

Analysis of the relationship between population density and productivity reveals that in the future the marginal economic benefits of higher population density are expected to diminish for counties included in Houston-Sugar Land-Baytown and Dallas-Fort Worth-Arlington metro areas. The Austin area is expected to benefit more from increasing population density.

Dr. Anari ([email protected]) is a research economist with the Real Estate Center at Texas A&M University.