Nội dung

Why Some Cities Are Growing and Others Shrinking Dean Stansel Over the last three decades, large cities like Pittsburgh, Detroit, Cleveland, Buffalo, and Toledo have seen their populations shrink, while areas like Houston, Atlanta, Dallas, Tampa, and Phoenix have seen their populations grow rapidly. Examining the policy differences between high-growth and low-growth areas can provide evidence that may help declining cities reverse their fortunes. In 1980, Austin, Texas, and Syracuse, New York, were roughly the same size. The Austin metro area had a population of about 590,000, and the Syracuse metro area had about 643,000 residents. By 2007, Austin’s population had increased by more than 1 million while Syracuse’s population had been stagnant. That same disparity exists when one examines the growth of employment and real personal income. Another disparity between the two areas is the tax burden. State and local taxes accounted for nearly 13 percent of personal income in Syracuse but only about 9 percent in Austin. Although there are numerous factors that can influence the growth of individual economies, one finds a consistent relationship between low taxes and high economic growth in metropolitan areas, in states, and in nations. This article details that relationship between taxes and growth for the 100 largest U.S. metropolitan areas. In the 10 highest-tax metro areas, the state and local tax burden accounted for about 12.4 percent of personal income. In those same areas, population grew by Cato Journal, Vol. 31, No. 2 (Spring/Summer 2011). Copyright © Cato Institute. All rights reserved. Dean Stansel is an Associate Professor of Economics in the Lutgert College of Business at Florida Gulf Coast University in Fort Myers. He thanks Edward J. Lopez, Jeff Noble, and an anonymous referee for useful suggestions. 285 Cato Journal 21.3 percent from 1980 to 2007, employment grew by 40.1 percent, and real personal income grew by 75.5 percent. In contrast, taxes were only 8.3 percent of personal income in the 10 lowest-tax areas. The economic growth in those areas was much faster. Population grew by 64.4 percent, employment by 107.6 percent, and real personal income by 157.3 percent. The contrasting experiences of Austin and Syracuse occurred in countless other areas as well. This article provides 14 additional examples of pairs of metro areas that had similar tax and growth patterns.1 The experiences of all 15 pairs of metropolitan areas provide valuable lessons for distressed areas everywhere. Keeping tax burdens low appears to be an important ingredient in the recipe for economic prosperity. If high-tax, low-growth metro areas like Detroit, Milwaukee, Buffalo, and Syracuse want to be more like high-growth areas such as Dallas, Tampa, San Antonio, and Austin, they should lower their onerous burden of taxation and bring spending under control. Taxes, Economic Growth, and Prosperity In 1776, Adam Smith wrote An Inquiry into the Nature and Causes of the Wealth of Nations. Economists have been busily examining the issue ever since. It is one of the most widely studied topics in the field of economics. One of the most common findings relates to how economic activity is organized. For example, capitalist countries (those in which economic activity occurs on the basis of voluntary exchange within private markets) tend to grow faster than socialist countries (those in which economic activity is organized by government). The existence of private property rights in capitalist countries helps create stronger incentives for individuals to be productive. As a result, factors of production (including labor and capital) tend to flow out of socialist countries and into capitalist countries. The economic collapse of the Soviet Union and other bastions of socialism provide ample evidence of that. 1 These correlations do not prove that low taxes have caused the economic growth. There are many other factors that have an important influence on economic growth. For example, Walters (2010) provided evidence of the negative relationship between unionization of the local labor market and city growth. Incorporating those factors is beyond the scope of this article. See Stansel and Swaleheen (2010) and the additional articles by myself and others cited in footnotes 3 and 4 for articles that do take account of other factors. 286 Why Some Cities Are Growing Starting in the 1980s, Nobel economist Milton Friedman played an important role, along with many other economists and public policy experts, in the development of an index of economic freedom that would allow researchers to be able to measure the degree to which a country had a free market economy. Those efforts culminated in the Fraser Institute’s publication in 1996 of the first edition of Economic Freedom of the World. There have been 14 more editions published since then in what is now an annual series. Large volumes of research have illustrated a positive relationship between economic freedom at the national level and economic growth. One of the problems for economists examining that relationship is that there are many other factors that can influence growth, and those factors can vary widely across a broad selection of nations. For example, there are large differences in religion, cultural, and other institutional factors. Those types of factors are very difficult to quantify, thus our ability to account for their influence on the process is quite limited. One way to avoid that problem is to look at sub-national jurisdictions. For example, the 50 U.S. states have much less variation in religion and culture than do two nations such as the United States and China. In 2002, the Fraser Institute produced its first edition of the Economic Freedom of North America, which provided an index of economic freedom in U.S. states and Canadian provinces (see Karabegoviç and McMahon 2008).2 Because smaller jurisdictions share a more common set of cultural institutions, it is easier for researchers to accurately examine the relationship between economic freedom and economic growth. There is growing evidence that states with higher economic freedom and lower taxes are more prosperous, even when the effects of many other growth-related factors are incorporated.3 Examining states addresses some of the challenges inherent in using national data, but not all of them. The boundaries of states and provinces are relatively arbitrary and some local economies cross them. For example, the Washington, D.C., metropolitan 2 In addition, the Pacific Research Institute has produced a state-level index, U. S. Economic Freedom Index (McQuillan, Huang, and McCormick 2004), and in 2009 the Mercatus Institute produced a broader state index, Freedom in the 50 States: An Index of Personal and Economic Freedom (Sorens and Roger 2009). 3 See, for example, Vedder (1990), Bartik (1991), Becsi (1996), Wasylenko (1997), Crain and Lee (1999), Kreft and Sobel (2005), Ashby (2007), Campbell and Rogers (2007), Ashby and Sobel (2008), Hall and Sobel (2008), Reed (2008). 287 Cato Journal area includes counties in Maryland, Virginia, and West Virginia. There are more than 30 other metro areas that cross state boundaries and a few cross national boundaries. San Diego’s metro area is on the U.S.-Mexican border, while Buffalo’s is on the U.S.Canadian border. Furthermore, economic conditions can vary widely within those boundaries. Conditions in Dallas’s metro area are quite different from those in the Lubbock and Amarillo metro areas a few hours to the west. Using the local economy as the unit of analysis helps to address the problems related to using nations or states. In the United States, the metropolitan area is a countybased concept designed to reflect the boundaries of local labor markets or local economies. Although there is no economic freedom index for metropolitan areas, there are data available on taxes and spending. One of the most important components of the various economic freedom indices is the tax burden. Taxes remove resources from private decisionmakers and put them in the hands of elected officials and bureaucrats. The latter face much weaker incentives to use those resources efficiently and lack the information to be able to do so. As a result, jurisdictions with higher tax burdens will tend to have less prosperous economies. Furthermore, high-tax areas will tend to be less attractive to residents and businesses. Because people and employers are mobile, high taxes will discourage in-migration and encourage out-migration. The literature examining local jurisdictions is limited. However, there is growing evidence that localities with higher taxes—and larger government in general—have less prosperous economies, even when the effects of many other growth-related factors are incorporated.4 Taxes and Economic Growth in the 100 Largest U.S. Metropolitan Areas To test the hypothesis that high-tax areas have less prosperous economies, one can observe data on taxes and economic growth for the 100 largest metro areas in the United States—that is, those with 2007 populations over 575,000—during the last three decades. The 4 See, for example, Bradbury, Downs, and Small (1982); Dalenberg and Partridge (1995); Crihfield and Panggabean (1995); Holcombe and Lacombe (2004); Higgins, Levy, and Young (2006); Stansel, Gohmann, and Hobbs (2008); Stansel (2009); and Stansel and Swaleheen (2010). 288 Why Some Cities Are Growing tax data measure total state and local taxes as a percentage of personal income. The local tax data are collected by the U.S. Census Bureau’s Census of Governments every five years. The state average is then added to the local figure to provide for more valid comparisons across states.5 To track changes in the tax burden over time, we can take the average of the tax burden for 1977, 1982, 1987, 1992, 1997, and 2002. Economic growth is measured by the change from 1980 to 2007 in population, employment, and real personal income. For consistency each metro area is defined as it was for 2009 (see U.S. Office of Management and Budget 2008). The data for the largest 100 metro areas show that areas with low taxes do indeed tend to grow faster than those with high taxes. As Figure 1 shows, this is true no matter how growth is measured. Population growth from 1980 to 2007 was three times higher in the 10 lowest-tax metro areas than in the 10 highest-tax areas. In those same areas, employment growth was more than two and a half times higher and real personal income growth was twice as high. Table 1 provides the data for each of those 20 metro areas. Five of the 10 lowest-tax areas are in Florida or Texas, states that do not tax personal income. Three others are in Tennessee, which only taxes dividend and interest income. The seven highest-tax areas are all in New York, which has one of the highest state income taxes in the nation. New York City, the highest tax area, has its own local income tax in addition to the state tax. Figure 2 shows that there is a negative correlation between state and local taxes and employment growth in the 100 largest metro areas. The correlation coefficient is ⫺0.405. Similarly, for real personal income growth the correlation with taxes is ⫺0.374 and for population growth it is ⫺0.346. While correlation does not prove causation, if taxes were not a drag on economic growth we would expect to see positive correlations. Furthermore, it should be noted that the tax data slightly lag the growth data. Using average tax burden for 1977–2002 and growth over 1980–2007 helps strengthen our results. Another way to examine the issue is to sort the metro areas by growth rather than by tax burden and then look at tax burdens in high-growth and low-growth areas. The data for the largest 100 5 In the case of metro areas that cross state boundaries, the state tax burden for the state with the largest central city in that area was the one used. 289 Cato Journal FIGURE 1 Low-Tax Metro Areas Had Higher Economic Growth 180% 10 LowestTax Metro Areas 160% 10 HighestTax Metro Areas 157% 140% Percentage Change, 1980–2007 120% 108% 100% 80% 75% 64% 60% 40% 40% 21% 20% 0% Population Employment Real Personal Income metro areas show that areas with high growth tend to have lower taxes than those with low growth. Figure 3 illustrates that this relation holds true for all three measures of growth. The 10 metro areas with the lowest population growth from 1980 to 2007 had about a 13 percent higher state and local tax burden than the highest population growth areas. Tax burdens were 19 percent higher in the areas with the lowest employment growth and about 15 percent higher in those with the lowest growth of real personal income. Table 2 details the tax and growth data for the highest and lowest population growth 290 8.5 8.6 8.3% Low-Tax Area Average 7.9% 8.0 8.2 8.2 8.3 8.3 8.4 8.4 Ten Lowest-Tax Large Metro Areas Jacksonville, FL MSA Bradenton-Sarasota-Venice, FL MSA St. Louis, MO-IL MSA Colorado Springs, CO MSA San Antonio, TX MSA Knoxville, TN MSA Tampa-St. Petersburg-Clearwater, FL MSA Nashville-Davidson—Murfreesboro— Franklin, TN MSA Memphis, TN-MS-AR MSA Fort Worth-Arlington, TX MD State and Local Taxes as a Percentage of Personal Income, 1977–2002 Average 64.4% 28.2 100.8 75.1% 93.1 19.3 89.8 70.8 34.3 66.9 66.1 Population 107.6% 59.9 125.9 118.2% 158.6 68.2 120.8 107.7 80.3 126.3 109.8 Employment continued 157.3% 107.9 180.8 192.7% 221.1 81.0 187.8 166.7 102.2 152.2 180.6 Real Personal Income 1980–2007 Growth in: TABLE 1 Low-Tax Metro Areas Grew Faster Than High-Tax Metro Areas Why Some Cities Are Growing 291 292 14.0% 13.9 12.6 12.5 12.2 12.1 11.8 11.7 11.6 11.5 12.4% 10.0 Ten Highest-Tax Large Metro Areas New York-White Plains-Wayne, NY-NJ MD Nassau-Suffolk, NY MD Syracuse, NY MSA Buffalo-Niagara Falls, NY MSA Poughkeepsie-Newburgh-Middletown, NY MSA Rochester, NY MSA Albany-Schenectady-Troy, NY MSA Milwaukee-Waukesha-West Allis, WI MSA Bakersfield, CA MSA Minneapolis-St. Paul-Bloomington, MN-WI MSA High-Tax Area Average 100 Largest Metro Area Average State and Local Taxes as a Percentage of Personal Income, 1977–2002 Average 21.3% 51.0 14.7% 9.9 0.2 ⫺9.2 32.2 6.1 10.3 10.5 93.7 44.9 Population 40.1% 80.0 25.2% 41.2 24.7 10.5 49.8 26.9 40.4 30.1 81.7 70.2 Employment 75.5% 129.4 102.3% 86.9 45.3 27.2 94.8 44.6 74.6 59.7 90.4 128.8 Real Personal Income 1980–2007 Growth in: TABLE 1 (cont.) Low-Tax Metro Areas Grew Faster Than High-Tax Metro Areas Cato Journal Why Some Cities Are Growing FIGURE 2 State and Local Taxes Are Negatively Correlated with Employment Growth State and Local Taxes as a Percentage of Income, 1977–2002 Average 14% 13% 12% 11% 10% 9% 8% 7% -50% 0% 50% 100% 150% 200% 250% 300% 350% 400% Employment Growth, 1980–2007 areas. Six of the 10 highest-growth areas are in three states with no personal income tax (Florida, Nevada, and Texas). All 10 of the lowest-growth metro areas are in the higher-tax Northeast or Midwest regions of the country. Taxes and Economic Growth in Selected Pairs of U.S. Metropolitan Areas Since residents and businesses are mobile, they have the ability to vote with their feet (Tiebout 1956) by locating in their most 293 Cato Journal FIGURE 3 High-Growth Metro Areas Had Lower Taxes 12.0% 10 HighestGrowth Metro Areas State and Local Taxes as a Percentage of Income, 1977–2002 Average 11.5% 10 LowestGrowth Metro Areas 11.0% 11.0% 10.7% 10.6% 10.5% 10.0% 9.5% 9.4% 9.3% 9.2% 9.0% 8.5% 8.0% Population Growth, Employment Growth, Real Personal Income 1980–2007 1980–2007 Growth, 1980–2007 desired jurisdiction. Metro areas with tax burdens that are much higher than others with whom they compete will tend to have less prosperous economies. To more closely examine this issue, we focus on 15 selected pairs of metro areas. Each pair contains one area with relatively low taxes and high growth and one with relatively high taxes and low growth. The metro areas within each pair have roughly similar population size either in 1980 or 2007. The first set was chosen from among the 100 largest metro areas (those with 2007 population greater than 575,000) regardless of geographic 294