Urban Labor Markets

Urban labor markets are characterized by the spatial proximity of households and businesses, which offers firms and workers advantages that lead to more efficient markets, enhanced productivity, and greater economic success. Nevertheless, the nation's city, while generating a large proportion of the nation's wealth, houses much of the nation's economic disadvantaged workers. This paper describes the current conditions of urban labor markets and outlines a national urban policy agenda that addresses these concerns by taking into account cities' spatial dimension. The paper argues that a national urban labor policy should emphasize the effects of physical and informational proximity on growth, the benefits of efficient urban markets, and the importance of the access of workers to urban labor markers. These characteristics distinguish a national urban policy from simply a national policy targeted at people who happen to live in cities. Urban Labor Markets Randall W. Eberts In every developed economy, cities are the center of economic activity and opportunity. They bring together businesses, workers, and customers into close physical proximity, which offers firms and workers advantages that lead to more efficient markets, enhanced productivity, and greater economic success. These benefits come about in several ways. Large metropolitan areas allow firms and workers to specialize in activities in which they hold a comparative advantage. By being close to suppliers and customers, firms can build efficient supply networks that lower costs and enhance productivity. Also, dense, urban environments enhance the flow of ideas and the transfer of technology through informational networks, which create growthenhancing spillovers to firms. Workers also rely on informational networks to add to their general knowledge, establish support groups, enhance job-related skills, and gain access to job openings. Consequently, the physical and informational proximity of businesses and workers within metropolitan areas yields the nation's most productive activities, spawns technological and organizational innovations, and launches new business ventures. For most urban residents, economic success is determined by the jobs and income generated by labor markets. A labor market rewards those people with the appropriate skills, energy, and ideas. Acquisition of these qualifications results from long-run decisions to invest in education and training, to choose work over leisure, and to be networked with people who been successful. Labor markets also provide the signals and incentives for individuals to make these decisions and to be in a position to take advantage of opportunities when they arise. By offering the greatest opportunity for economic success, cities attract both the nation's most talented and successful individuals and the most disadvantaged. While people of all skills are important to generate the economic complementarities that enhance productivity, many of the least skilled people do not find the success they had hoped for. Therefore, cities stand as a stark dichotomy of those who have succeeded and those who have not. They generate a large proportion of the nation's wealth but also house much of the nation's poverty and homelessness. Cities have always exhibited this stark contrast between those who have achieved economic success and those who have not. The ancient Greek philosopher Plato gave commentary to his times in observing that "there are two cities: One for the rich and one for the poor" (The Republic). The ability of urban labor markets to improve the quality of life for all groups of workers varies over time. In recent years, opportunities for economic advancement have diminished. Real earnings growth has slowed to a paltry 0.9 percent during the last decade. Worse yet, only those individuals with college degrees experienced improvements in real income; people with high school degrees or less saw their real earnings decline. The slowdown in earnings and the rise in income inequality has placed considerable stress on cities. Many cities have been unable to make sufficient investments in physical and human capital required for efficient markets and the easy flow of information. As the core structure of cities deteriorates and the advantages of close proximity diminish, businesses and households have 2 left inner cities for outlying areas. This shift in location has left inner cities with fewer resources to devote to mounting social problems. This shift has also helped to diminish the nation's capacity to spawn high-productivity, high-growth activities. The purpose of this chapter is to assess the current condition of urban labor markets and to suggest a national urban policy agenda that addresses these concerns. For individuals, these issues include slow wage growth, increased earnings inequality, persistent earnings gap between whites and minorities, disparity in job opportunities and earnings between inner cities and suburbs, the spatial mismatch of jobs, and the increased isolation of many inner city residents. For the nation, the issues are the social fallout from growing income disparities and reduced income mobility and the economy's diminished growth capacity as the physical and social infrastructure supporting urban area markets deteriorates. It is appropriate for the federal government to create and carry out a national urban policy for two reasons. First, many of the issues listed above are related to the redistribution of income, which is a traditional function of the federal government. Second, the close proximity of activities within metropolitan areas generates externalities that contribute more than proportionately to growth and that compound the many social problems facing cities. Therefore, to reflect the urban environment that generates these economic and social trends, a national urban policy must incorporate these externalities. The policy statement should emphasize the effect of physical and informational proximity on growth, the benefits of efficient urban markets, and the importance of the access of workers to urban labor markets. These characteristics distinguish a national urban policy from simply a national policy targeted at people who happen to live in cities. The chapter is divided into sections that describe several major issues facing urban workers. Each discussion presents data on these trends and identifies current research that offers insights into the underlying causes. In addition, the discussion attempts to show these issues within the context of the urban environment and to identify how these externalities should be incorporated into a national urban policy. I. Metropolitan Growth The fiscal and social distress experienced by cities in recent years has been well documented. Yet, despite increased concern about urban crime, congestion, air quality, and high housing prices, metropolitan areas continue to grow and to offer their expanding populations greater job opportunities and higher income than the declining nonmetropolitan areas. Between 1980 and 1990, the number of metropolitan residents swelled by 13.7 percent, while the U.S. population grew 9.8 percent. Nonmetropolitan areas actually lost population. As a result, the share of the U.S. population residing in the 268 designated metropolitan areas rose from 74.8 percent in 1980 to 77.5 percent in 1990. In 1990, a greater proportion of working age metropolitan residents were employed than nonmetropolitan residents, and the median income of 3 a metropolitan worker was $4,000 higher than that of their nonmetropolitan counterparts (1990 Census). This ability to create relatively high paying jobs attests to the advantages cities offer to firms through higher labor productivity and proximity to suppliers. The economic performance of central cities and suburbs is strongly linked. Workers in both central cities and suburbs benefit to a large extent from the growth of the entire region. Several articles have shown the intrametropolitan linkage between earnings and employment. For example, Stanback and Knight (1976) found that 67 percent of suburban income was generated by central city earnings in the mid 1970s. Ledebur and Barnes (1992) showed that suburban employment suffered when central city income growth lagged behind suburban income growth. In addition, job growth was markedly better for metropolitan areas with smaller disparities between central city and suburban income. The linkages between the two entities continue to strengthen as a greater percentage of holders of central city jobs reside in the suburbs. Savitch et al. (1993) show for a selected sample of 10 metropolitan areas that suburbanites claimed a rising share of an expanding central city economy. Even with these linkages, considerable disparity exists between central cities and outlying suburbs. The median income of central city residents is 40 percent lower than that of suburban residents (Figure 1). Also, central city residents are less educated and have a significantly higher unemployment rate than suburbanites. Movements between central cities and surrounding areas have only worsened these disparities in recent years. From 1989 to 1990, the median income and educational attainment of households moving from central cities to suburbs were higher than those they left behind. The median income and education levels of households leaving the suburbs for the central city were not only lower than their former suburban neighbors but also lower than central city residents. Twice as many people moved from the central city to the suburbs than from the suburbs to the central city. This exodus from central cities and the persistent economic disparities heighten the physical and intellectual barriers between central cities and suburbs. Despite the flow of people out of central cities, central city population grew during the last decade, and at a slightly higher rate than urban fringe areas. The proportion of the U.S. population living in central cities rose from 30.0 percent in 1980 to 31.3 percent in 1990. Hispanics accounted for almost half the central city population growth. Their share of central city population increased from 10.8 percent in 1980 to 15.3 percent in 1990, and their share of total metropolitan growth increased from 7.5 percent to 10.5 percent. While this increase is most evident in Southwest cities, Hispanics' presence has increased in many cities throughout the country, as the total Hispanic population has increased 53 percent during the 10-year period. The influx of Hispanics into central cities has shifted the composition of metropolitan populations toward more Hispanics and fewer whites. The proportion of blacks have remained about the same. White residents' share of central city population has decreased from 69.7 percent in 1980 to 66.1 percent in 1990, while its share of total metropolitan population has declined from 81.9 percent to 78.3. Black residents' share of central city and metropolitan population has remained constant at 22 percent and 13 percent, respectively. Since the median income of 4 Hispanic households is $9,000 less than that of white households, the shift in the composition of central city population from white to Hispanic has contributed to lower median income of central cities. Central cities have also experienced a shift in industrial composition from manufacturing to service jobs, which has contributed significantly to the income differential between central cities and suburbs. The decline in the percentage of manufacturing jobs has had a large effect on low-skilled workers. Factory jobs traditionally offered low-skilled workers higher wages than they could find in other jobs. Bound and Holzer (1991) and Bluestone et al. (1991) find that metropolitan manufacturing employment substantially impacts the employment and earnings of young men, particularly young black men, with a high school education or less. II. Slow Earnings Growth and Rising Income Dispersion The traditional role of cities as a place to seek job opportunities and achieve income advancement has been eroded during the last two decades by historically low earnings growth for most groups of workers. The less skilled workers have experienced a loss in real income, and consequently a decline in their living standards. While city dwellers on average have fared better than the national average, their earnings have followed the national trend of slower earnings growth. These trends stem from a complex assortment of economic, demographic, and social forces. Many of the factors are symptomatic of urban problems. Focusing on the more important factors offers perspective on how a national urban policy might prioritize the host of problems facing urban labor markets. During the two decades following World War II, Americans became accustomed to income growth of nearly 3 percent per year, after adjusting for inflation. All income groups enjoyed such real income gains. In fact, the lowest quintile registered the largest annual increase and the higher quintile the lowest. Only half a percentage point rate of growth separated the fastest from the slowest growing income groups (Economic Report of the President, 1994). However, starting around 1970, real income growth slowed dramatically. From 1973 to 1992, median family real income increased by only 0.1 percent per year. All income groups experienced a slowdown, but those at the lower end of the distribution were the most severely affected. The highest quintile of families enjoyed a 0.93 annual increase in median real income between 1973 and 1992, whereas the lowest quintile suffered a loss in real income of 0.69 percent annually. Slower real income growth is directly related to the slowdown in productivity growth. About the time earnings growth slowed dramatically, productivity growth slowed from an average annual rate 2.8 percent from 1958 to 1973 to an annual rate of 0.9 percent thereafter. Many economic factors are responsible for slower productivity growth, including a decline in worker skills, lack of adequate investment in private and public capital, and decline in innovative activities. With cities traditionally at the forefront of the nation's productivity gains, the slowdown draws attention to urban concerns that may have contributed to this problem. Current 5 urban problems of poor educational opportunities in inner cities, the movement of jobs away from the urban core, and the decline in investment in urban infrastructure may have contributed to the productivity slowdown. Urban social issues, such as low skills, lack of an adequate support structure, and lack of participation in the workforce, have also prevented many urban workers from enjoying the benefits of productivity growth. Changes in social characteristics also contributed to the change in household income. According to one study, social characteristics had an even greater effect than economic factors. Ryscavage, Green and Welniak (1991) show that the decline in married couple households as a percentage of all households and the increase in educational attainment had large offsetting effects on household income. Of seven factors considered--age, race, type of household, education, work experience and head of household, work experience of spouse, and industry of household head--type of household and education registered the largest effects. They estimate that as the percent of married couple households declined from 70 percent in 1969 to 55 percent in 1989, median household income fell by $3,226, or 10 percent. They also estimate that the increase in educational attainment boosted median household income by $3,010. Between 1969 and 1989, the percentage of persons age 25 and older with four or more years of college rose from 10.7 percent to 21.1 percent. The higher earnings from educational attainment resulted from two factors. The first was higher educational attainment. The second was the increase in the educational wage premium during the 1980s. During the 1970s, the earnings ratio between college and high school graduates remained constant at 1.4 for men and 1.8 for women. However, after around 1979, the education premium rose significantly, climbing to 1.7 for men and 2.1 for women by 1989, as demand for skilled workers outstripped supply. In fact, for young men, only those with four years of college or more enjoyed a real wage increase during the 1980s. Real income of high school graduates and dropouts declined during this period (Economic Report of the President, 1993, p. 92). These same factors that affected median income--household type, educational attainment, and industry mix--also shifted the income distribution. Had these factors not changed during the last 20 years, the income distribution would have been more equal. These two social factors affected the various income groups differently. Household type significantly affected the lower quintiles of the income distribution but not the higher quintiles. Educational attainment affected the high end of the distribution but not the low end. Industry mix affected the lowest quintiles, suggesting the heavy dependence of low-skilled workers on the rents they received from manufacturing jobs (Bartik, 1993). III. Minority Earnings Gap One of the most troubling and persistent concerns in the performance of urban labor markets, and one that a national urban policy must address, is the earnings gap between blacks and Hispanics and whites. Blacks and Hispanics earn considerably less than whites, and neither 6 group has made any significant strides in narrowing this gap during the last decade. In 1980, black household median income was 58 percent of white household income. By 1991, the ratio stood at 60 percent. The ratio of the median income of Hispanic households to white households remained steady at 72 percent during this period. Even considering married-couple households, the gap, although not as large, has not narrowed appreciably since the mid-1970s. Before then, black married households advanced from 72 percent of white income to 80 percent. When all households are considered, household type--married versus single head, usually female--plays a large role in explaining differences among the three major racial and ethnic groups. Between 1969 and 1989, the proportion of all black households that consisted of married couples dropped from 56.7 percent to 35.7 percent. By 1991, the percentage had fallen even further to 32.8 percent. In contrast, the proportion of all white households that were married couples shrunk from 73 percent to 59 percent between 1969 and 1989, and stood at 57 percent in 1991. The depressing effect of the rise in female-headed households can be seen by simply comparing the median income of female-headed households versus married-couple households. In 1991, the median income of a black female-headed household was $11,414 versus $33,307 for married-couple households. The income of single female parents was even lower--$9,417. IV. Lack of Economic Progress of Black Men With household type accounting for much of the black-white household income differential, it is tempting to place most of the blame on the deterioration of the married-couple family for the lack of progress in narrowing the gap between blacks and whites. However, while important, it detracts from the underlying reasons for family breakups and ignores the problems facing household heads, in particular black males. Since joblessness among black males leads 1 to fewer marriages, the decline in black married couples is self perpetuating (Welch, 1990). In addressing the lack of economic progress among urban minority workers, a national urban policy must focus on the disadvantages minorities face in seeking employment and the increasing disenfranchisement of low-income black males. The widening black-white earnings and employment gaps among young men cut across all education levels. Bound and Freeman (1992) show that the gap is most severe for college graduates and high school dropouts and that blacks in the Midwest experienced the largest decline. Bound and Freeman conclude that the black economic experience is too diverse to support a single-factor explanation. For college graduates, changed occupational composition, shifts in industry demand, and growth in relative supply are major contributors. For high school dropouts, the reduction in the real minimum wage lowers earnings, and the increase in the proportion of young blacks with criminal records are major causes of reduced employment. For blacks in the Midwest, decline in manufacturing, changes in occupational composition of jobs, 7 decline in unionism and the real minimum wage, and the growth of relative supply were major factors. A more important factor in explaining the decline in real income of low-skilled men is a significant decline in their labor market participation rate. For example, the employment to population ratio of white male high school dropouts fell fourteen percentage points from 0.89 in 1967 to 0.75 in 1987. The decline was even greater for black male high school dropouts (Juhn 1992). Central city black men are less likely to find work than their suburban counterparts. In 1990 when the national unemployment was around 5.5 percent, the difference in unemployment rates between central city and suburban residents was two percentage points. The fall in employment is largely accounted for by a rise in the number of men who experienced long spells of labor market inactivity, and not in more frequent spells (Juhn 1992). The increasing number of working age males who are detached from the labor force has long-run economic and social consequences. Without more frequent work experience, a growing portion of the work force loses valuable job skills, which is detrimental to the national economy in wasting valuable labor resources as well as to individuals who lose income. They also lose the job contacts necessary to retain employment or find better job prospects. Nonparticipants in the labor market contribute to social problems. Those who are unemployed and no longer actively look for employment are much less likely to marry and live with a spouse than labor market participants (Welch 1990). As shown earlier, the increase in female-headed households has led to a decline in household income and contributes to a persistent cycle of welfare dependency. V. Barriers to Economic Success Urban workers benefit from metropolitan economic growth only if they have access to the urban labor market. For minorities and low-skill workers, this access is impeded by physical barriers of spatial isolation and intellectual barriers of poor education and inadequate support groups and referral networks. Therefore, in designing a national urban policy, efforts should also be made to increase access to the urban labor market. The shifting industrial mix of central cities and suburbs has left some groups of workers isolated from decent-paying jobs. The movement of manufacturing jobs to the suburbs has made it increasingly difficult for many inner city residents to finds jobs close by that offer the same wages as manufacturing jobs. Various types of service jobs have concentrated in central cities, but either the pay is lower or the skill requirements are higher than the qualifications of many residents. Beginning with Kain (1968), several studies have proposed that physical distance between jobs and workers could be a major factor in the low employment rate and income levels of blacks, particularly black males. Kasarda (1989) found that black males without high school degrees who held full-time jobs spent 10 minutes longer traveling to work than their white counterparts. Ihlanfeldt and Sjoquist (1990, 1991) presented convincing evidence that spatial mismatch has 8 lowered the earnings of central city non-college graduates significantly. Using more aggregate data, Price and Mills (1985) estimated that the concentration of blacks in central cities could explain as much as one third of the black/white earnings gap. Focusing on minority youths, Ihlanfeldt (1992) estimated that a five-minute reduction in commuting time would increase the probability of employment by 0.075, which would cause the mean employment rate of this group to rise by 26 percent. Physical distance is not the only barrier that prevents low-income inner city residents, particularly minorities, from gaining access to better-paying jobs. Intellectual barriers also play a major role. Intellectual barriers prevent people from participating in the formal and informal information networks through which people learn about job opportunities, are referred to employers, and gain appropriate role models and support groups. These barriers stem in part from physical isolation, such as the increasing concentration of poverty. But they also result from a growing gap in the ability of inner city residents to meet the skill requirements of the workplace and to communicate and move comfortably within an increasingly sophisticated work environment. While distances may become effectively shorter because of better transportation, intellectual barriers may continue to grow as businesses demand higher skills than many inner city workers can match. The growing gap between skill requirements of the workplace and skills of inner city residents deters many workers from participating in urban labor markets. Without appropriate skills and workplace know-how, workers are shut out of many jobs, even entry-level ones. The lack of quality education for residents of inner city areas has been blamed for the detachment from the workplace. However, focusing on education to explain the difference in economic success between black men and white men is somewhat controversial. One would expect a greater demand for high skill workers to worsen the absolute and relative labor market position of black men, since they have on average less educational attainment and less work experience. However, Bound and Freeman (1992) find strong evidence that interracial differences in measured educational attainment explain little or none of the deterioration in black men's labor market position. Evidence of poor quality inner city schools compared to suburban schools raises the possibility that this difference is a significant factor in explaining black/white earnings differentials. Card and Krueger (1992) use state-level data on school attributes to show that improvement in school quality can narrow the black/white wage differential. Bound and Freeman (1992) are not as convinced that educational quality is a major factor and cite statistics that show standardized test scores of blacks rose modestly relative to whites during the 1980s. A major barrier to low-skilled persons, particularly blacks, is the lack of basic workplace competencies. Detailed studies conducted in Cleveland (Hill, Rittenhouse, and Allison 1994) and in Michigan (Bartik and Erickcek, 1992) found that many disadvantaged workers do not meet the basic requirements of entry-level jobs. Through interviews and focus groups, it became evident that many disadvantaged individuals have difficulty finding or holding jobs because they do not 9 understand what employers expect of them, such as proper dress, getting to work on time, coming to work every day, working well with fellow employees and management. They also lacked other basic skills necessary to perform any job, such as problem solving, resource and time management, and organization. Many employers interviewed by the studies believed that schools were not preparing students with these competencies. If schools do not properly prepare students for viable jobs or prepare them to enter and succeed in college, their physical isolation in inner city ghettos will be reinforced by intellectual isolation from networks that allow them to take advantage of the opportunities of the urban labor market. Business hiring practices and informational and referral networks may also place blacks at a disadvantage. The move to more informal referral systems with the decline of unionism and the emerging importance of small businesses in job creation may have made this more of a factor in recent years. The majority of workers learn about job openings and obtain referrals through informal networks of friends and relatives (see Holzer (1991) for a literature review). Holzer (1988) found that the most frequently used methods of job search by youths are checking with friends and relatives and making direct applications without referrals. O'Regan and Quigley (1993) found that family networks of fathers, mothers, and siblings were most important in helping at-home youths gain access to jobs. Moreover, jobs obtained informally are generally associated with higher earnings and more prestigious occupations (Granovetter, 1974). Blacks also appear to have a disadvantage throughout much of the job process from interview to promotion. Black men fare worse than whites in interviews (Kirschenman and Neckerman, 1991) and are at a disadvantage with respect to specific employer practices and attitudes in recruiting, hiring, and promotion (Braddock and McPartland, 1987). Wilson (1987) suggests that geographic concentration of low-income households may lead to social isolation, further limiting the ability of residents to improve their economic situations. The spatial separation of households with different income and educational levels is more pervasive than simply between central cities and surrounding suburbs. Concentration of poverty, and thus the loss of networks of families and friends with positive work experience, has occurred within central cities. Studies have shown that low-income households have become increasing concentrated in central city neighborhoods during the last several decades (Reischauer (1988). For example, a recent analysis of poverty in the Cleveland area by Coulton, Chow and Pandey (1990) shows that it has become more geographically concentrated during the 1980s. In the City of Cleveland, between 28 and 47 percent of the population reside in high poverty neighborhoods. The analysis uses 40 percent as the high poverty threshold. The effects of concentrated poverty are more evident for the nonwhite population than the white population. Migration of the non-poor out of neighborhoods contributes significantly to the rising spatial concentration of poverty. Coulton, et al.(1990) estimated that migration of non-poor from emerging poverty areas accounts for 60 percent of the increased concentration of poverty in Cleveland. Between 1980 and 1988, 23 percent of the non-poor moved, while only 1 percent of the poor moved out of these areas. 10 The effects of neighbors or members of a support group are critically important in the acquisition of skills, incentive and value structures, and tendencies toward various types of behavior. Crane (1991) finds that high school dropout and teenage pregnancy rates are significantly lower in neighborhoods with a higher proportion of adults holding management and professional jobs. Case and Katz (1991) find contagion effects between neighboring youths for several different outcomes: criminal activities, drug use, unemployment, gang membership, and church attendance. Therefore, the breakdown of informal networks as friends and relatives with jobs or job experience leave a neighborhood reduces the prospect of finding employment for those who remain. Crime and the participation in illegal business activities are also detrimental to a young black man's employment prospects and earnings potential. Living in a crime-ridden area also indirectly affects a person's ability to function properly in the labor market. Bound and Freeman (1992) report that greater criminal activity may induce some young men to forego employment. Conversely, the deteriorated job market for the less skilled may make crime more attractive. Freeman (1992) underscores the long-term adverse consequences of involvement in crime on youth employment and argues that improvement of skills and increase in legitimate opportunities should be part of a crime reduction effort. Unfortunately, the tremendous increase in crime in inner city neighborhoods may discourage the location of businesses, which reduces the number of jobs available to these residents. VI. Policy Issues and Suggestions An effective national urban policy must embody the unique characteristics of cities while addressing the trends that pervade the national economy. The distinguishing features of cities include the spatial proximity of businesses and households, the flow of information between these parties, and the linkages of markets within urban areas. These characteristics generate both positive and negative externalities, which give cities a prominent position in the economy. Consequently, government's role and the primary purpose of a national urban policy should be to design programs that promote growth, internalize externalities, improve market efficiency, and strive for a more acceptable income distribution. Current trends facing urban workers and the findings of a large body of research described in this chapter suggest the following policy directives.