A number of researchers have suggested that the local human capital level has external effects on others nearby, with much of this literature focusing on the external effects of average education levels on productivity and wages.1Rauch, 1993, Moretti, 2004b, Dalmazzo and de Blasio, 2007a and Dalmazzo and de Blasio, 2007b and others offer support for the existence of human capital externalities by showing that the local level of human capital is positively correlated with wages even after controlling for individual worker characteristics.2 However, because workers are mobile, the external effects of human capital on wages will not be the end of the story. Higher productivity in an area will attract new workers and cause the area to grow in population (Glaeser et al., 1995, Simon, 1998, Simon, 2004, Simon and Nardinelli, 2002 and Glaeser and Saiz, 2004).3Berry and Glaeser, 2005 and Waldorf, 2009 also suggest that the existing stock of human capital is especially important in attracting educated in-migrants. Furthermore, as new workers move in, they compete with existing residents for housing and bid up housing prices (Rauch, 1993, Shapiro, 2006 and Dalmazzo and de Blasio, 2007a). Shapiro (2006) suggests that the local human capital level increases the implicit value of an area's consumption amenities; i.e., the stock of human capital makes an area a more desirable place to live and increases the quality of life. Shapiro (2006) estimates that 40% of the growth effects of human capital are due to increased quality of life.
Because the local human capital stock appears to make an area better in so many ways, it is important to understand why areas differ in aggregate human capital levels. One of the most important determinants of the local human capital level is the presence of colleges and universities in the area, and many of the most highly educated areas are home to major state universities (Winters, 2011). Higher education institutions increase the local human capital stock in at least two ways: 1) they increase access to higher education for local residents and make it more likely that local high school graduates will pursue post-secondary education (Card, 1995 and Alm and Winters, 2009); and 2) they bring in students from outside the area seeking an education and some of these student in-migrants end up staying in the area after their education is complete (Blackwell et al., 2002, Huffman and Quigley, 2002, Groen, 2004, Groen and White, 2004 and Hickman, 2009).4Winters (2011) suggests that most of the differential in-migration to high human capital cities is due to students moving to pursue higher education and that most of the growth of so-called “smart cities” is due to recent student in-migrants staying in an area after finishing their education. These students stay in the area because it gives them greater utility than other areas. Importantly, some students might stay in the area where they moved for higher education because it offers a high quality of life.5 Colleges and universities, therefore, affect their surrounding areas in several important ways.
This paper considers the separate effects of the local human capital level and the relative importance of higher education institutions on the quality of life in U.S. metropolitan areas. Building on Rosen, 1979 and Roback, 1982 this paper measures quality of life by differences in “real wages” across areas, i.e., wages adjusted for differences in cost of living and worker characteristics (Winters, 2009 and Winters, 2010). A brief discussion of the theoretical model is presented in Appendix A. Following previous literature, we measure the local human capital level by the share of adults (age 25 and older) with at least a four-year college degree. Notable examples of high human capital stock cities include San Francisco, CA, Portland, OR, Boston, MA and Washington, DC. We measure the relative importance of higher education institutions by the share of an area's population enrolled in higher education. While there are other criteria by which we could measure the relative importance of higher education (see for example, Gumprecht, 2003), this measure is quite intuitive. Areas for which colleges and universities are relatively important will have a large share of their population enrolled in higher education. It is worth emphasizing that this is a relative measure. Most large metropolitan areas have at least one college or university, but a single higher education institution in a highly populated area may not have much of a relative effect. Quite often the areas in which higher education has the greatest relative importance are small to mid-size metropolitan areas home to large flagship state universities, i.e., what we sometimes think of as “college towns.” Examples include State College, PA, College Station, TX, Madison, WI and Athens, GA.6
The local human capital level and the relative importance of higher education institutions are expected to improve the quality of life in an area for a number of reasons. First, highly educated residents may be more likely to support local public goods such as museums, parks, symphonies, and theaters. Similarly, educated residents might facilitate the density and diversity of consumer services such as restaurants, coffee shops, and bars that consumers find desirable (Glaeser et al., 2001 and Waldfogel, 2008). Educated persons are also more likely to be politically active (Milligan et al., 2004) and may elect better government officials and help build clean cities with low pollution and low crime. They are also less likely to commit crimes (Lochner and Moretti, 2004), and more likely to be tolerant of others different from themselves (Florida, 2002).
A strong relative importance of one or more higher education institutions in an area may also be an important amenity. First, the presence of higher education institutions increases the local human capital level, which increases quality of life as discussed above. However, the effects of colleges and universities on quality of life go beyond the effect of increasing the local human capital level. Many college towns provide consumption opportunities not readily available elsewhere. One of the most obvious examples is live collegiate sports. Many diehard fans want to watch their teams in person and living near their schools makes this more convenient. Other examples of college town amenities include college bars and local music scenes, which are often considerably more abundant in areas with a strong college or university presence. A number of other consumer services are also likely to be more abundant in college towns including bookstores, record stores, bicycle shops, health food stores, pizzerias, and ethnic restaurants (Gumprecht, 2003). The local human capital level and the relative importance of higher education institutions, therefore, are both likely to affect quality of life in important ways.
In this paper we find that the quality of life in an area is positively affected by both the local human capital level and the relative importance of higher education institutions. Furthermore, these effects persist when these two measures are considered simultaneously, even though these two variables are highly correlated. That is, the human capital stock and the relative importance of higher education institutions have a shared effect and also separate effects on quality of life. Controlling for the share of the population enrolled in college, a 0.10 increase in the share of adults with a college degree increases the quality of life in an area and causes workers to accept roughly 0.9% lower real wages. Controlling for the share of adults with a college degree, a 0.10 increase in the share of the population enrolled in college increases quality of life sufficiently so that workers are willing to accept 2.9% lower real wages.
