Getting to work in spite of the odds: commuting patterns of African Americans in Rochester and Buffalo, New York.

Differences in the residential, employment, and household characteristics of African Americans and European Americans are well documented, but racial differences in the journey to work are still not well known. Compared to the past, there are now more studies about the commuting behavior of African Americans, but the specific impact of the exodus of jobs to suburban locations on African American men and women who live in inner cities is still understudied. One study of Buffalo, NY examined racial differences in commuting, but it focused only on women (Johnston-Anumonwo 1995). In a follow-up study, men were included in the analysis (Johnston-Anumonwo, 1997), but both sets of inquiries concentrated on employment characteristics and ignored household characteristics which are likely to feature significantly for female workers.

The purpose of the present study is to examine the question of racial differences in locational access to jobs in Rochester, NY and draw parallels with the previous findings for Buffalo, NY. While presenting new data for Rochester, the study retains the critical inquiry on whether suburban employment imposes longer commute times on African Americans than on European Americans, but extends the inquiry to see if presence of children in the home affects workers' commute lengths. The results for Rochester strongly complement those for Buffalo and are consistent with the spatial mismatch hypothesis, which posited that African Americans suffer from distant suburban employment.

Following a review of the background literature of the journey to work for African Americans, a brief description of the study areas and data is provided, and then the findings are presented. The study's findings highlight the fact that African American men, and especially women, endure relatively long commutes to get to work in spite of transportation, locational and socioeconomic hindrances.

THE JOURNEY TO WORK FOR AFRICAN AMERICANS

Inquiries about racial disparities in employment accessibility are central to the spatial mismatch hypothesis (Kain, 1968). When it was first proposed in the 1960s, the hypothesis emphasized that employment opportunities are expanding in suburban locations, but because of continuing segregation of Blacks in inner cities, there exists a spatial mismatch, such that Black inner city residents face difficulties in reaching the growing job opportunities in suburbs (see Holzer, 1991 for a review). There has been little change in the residential segregation of Blacks between 1960s and now (Darden, 1990; Denton, 1994; Massey and Hajnal, 1995). This is true for the two Upstate New York counties selected for this study: Monroe County and Erie County (with Rochester and Buffalo as their respective central cities). Monroe County was 4.1 percent black in 1960 and 10.1 percent black in 1980, but 97.5 percent of the county's Black population in 1960 lived in the Rochester central city and in 1980, 87.7 percent did. A similar and even sharper pattern of disproportionate representation of Blacks in the central city prevails in Erie County. Erie County was 7 percent black in 1960 and 10.1 percent black in 1980 with 94.7 percent and 92.4 percent of the Black population respectively living in the central city in both time periods (U.S. Dept. of Commerce, 1963; 1983a; 1883b).

Early research on the effect of the exodus of jobs to suburban locations on the workplace accessibility of inner city African Americans rarely included female workers even though African American women have historically had high levels of labor force participation. Studies on the journey to work of female workers highlight two key trends. First, Black women have longer travel times than White women. Second, unlike White women who typically have shorter commutes than White men, the journey-to-work time of Black women are generally as long as those of Black men. For example, McLafferty and Preston (1991) report that in 1980, African American women in metropolitan New York spend 10 minutes longer on the average for their home-to-work trip than European American women. Other studies report longer commuting times for African American men than European American men as well (e.g., Greytak, 1974; Ellwood, 1986; Leonard, 1987). The research finding that being Black is associated with longer work-trip times can readily be attributed to racial differences in the use of a private automobile, but racial differences in other locational and socioeconomic variables that are known to affect journey-to-work time are also relevant.

The journey-to-work literature is replete with findings that workers who use public transportation usually spend a longer time getting to work than those who use a private car. Similarly, those earning high incomes have longer work-trip times than those who have low incomes (e.g., U.S. Department of Commerce, 1979, 1982; Hanson and Johnston, 1985; Pisarski, 1987; Gordon et al., 1989; Dubin, 1991; Cooke and Shumway, 1991; Ihlanfeldt, 1992; McLafferty and Preston, 1992). The evidence about the role of child status on commuting time is mixed. But in a study on women's work-trip time across racial groups, Preston et al. (1993) found that Black mothers had longer work trips than White mothers.

Although the spatial mismatch hypothesis specifies the growing trend of workplaces in suburban locations, there are still very few empirical inquiries of the impact of suburban workplace destination on the commuting behavior of African American men and women. One study of race and commutes in Detroit found that in 1980, Blacks who worked in the Central Business District worked closer to home than Whites, but for suburban destinations Blacks commuted longer than Whites (Zax, 1990). This different commuting pattern, according to Zax, suggests that racial residential segregation restricts Black suburbanization and contributes to longer commutes for Blacks who have to work in suburban Detroit.

But disagreement persists about whether or not African American workers in U.S. metropolitan areas are more distant from centers of employment than European Americans are. This paper reports on an empirical inquiry about the question of the spatial mismatch of African Americans resident in Monroe County (Rochester) and Erie County (Buffalo) New York. Using data from the 1980 census, the study investigates whether African American working men and women in Rochester and Buffalo experience greater commuting difficulties than European Americans while taking into consideration racial differences in key factors that affect work-trip length--namely, automobile use, location, income and child status.

In Rochester and especially in Buffalo, the majority of African Americans still live in segregated neighborhoods. In fact, Buffalo remains a city with one of the highest levels of racial residential segregation in the United States (USA Today, 1991). Both Rochester and Buffalo are representative examples of cities that experienced massive de-industrialization starting in the 1970s when manufacturing jobs slumped in the urban economy, and moved out of the central city to the suburbs. The prevalent trend was one of job decentralization. Thus, one can expect to find evidence of journey-to-work constraints for inner city African American residents of Rochester and Buffalo.

A specified set of research questions is examined. Do African American men and women in Rochester and Buffalo spend a longer time commuting than European American men and women? Does unequal access to private automobiles lead to differences in the journey-to-work time of African Americans and European Americans? If differential access to private automobiles is responsible for longer commute times among African Americans, then there should be no racial difference in the work-trip times of private auto users. Is location of the workplace responsible for any racial difference in commuting? Lastly, is there any difference in the commute times of African American and European American workers with similar income and child status?

DESCRIPTION OF THE DATA

The range of factors necessary for a full understanding of racial disparities in commuting means that detailed data about working individuals are preferred. The Public Use Microdata Samples (PUMS), which is a database with information on individuals' socioeconomic characteristics and their locational and work-trip attributes, meets this requirement. The 1980 five percent Public Use Microdata Samples (PUMS) for Monroe County and Erie County, New York is used for this study. In 1980, for the first time, the census included information on journey-to-work time, i.e., the time spent traveling from home to work. Travel time, the actual number of minutes spent from home to work as reported by the respondent, is thus the measure available for work-trip length. The travel mode is the means of transportation that the worker uses to get to work e.g., public transportation or private automobile. Because of budget cut backs, the Census Bureau processed only half of the responses about the journey to work for the 1980 census, so the sample sizes are small.

While socioeconomic and demographic information provided in the PUMS is quite detailed, detail information about location is not provided in these public use data sets in order to protect the confidentiality of respondents. Only two very broad locational categories are available for this study--central city location and non central city location. In other words, locations outside the census-designated central city limits are classified simply as non central city locations (i.e., suburb). Income, the worker's annual salary in 1979, is divided into two categories: (a) below $10,000 or (b) $10,000 and above. Child status is also divided into two categories by focusing on the presence of children in the home: (a) households with no children and (b) households with school-aged children. The Census Bureau specifies race as White, Black, etc. based on respondents' self-classification (U.S. Department of Commerce, 1983c). Only White and Black employed males and females who are sixteen years old and older are selected for the study.

The five percent PUMS is used so that the study would include a large number of Black workers in these two predominantly White Upstate New York counties. In the Monroe County sample, there are 3,755 European American (White) men, 261 African American (Black) men, 2,769 European American (White) women and 282 African American (Black) women. In the Erie County sample, there are 4,876 European American (White) men, 304 African American (Black) men, 3,551 European American (White) women and 321 African American (Black) women. These then are the four race-sex groups that constitute the basis of the comparisons in the study. Racial differences in journey-to-work time are assessed using simple t-test statistics. Only racial differences that are significant at 95 percent level of confidence and above are reported. The results are presented next starting with racial differences in means of transportation used for the work trip.

FINDINGS

African Americans Rely More on Public Transportation

African American workers in Rochester and Buffalo used public transportation more than European Americans did in 1980. Indeed, a much higher proportion of African American women than European American women used public transportation for their work trip. In Monroe County, 24.8 percent of African American women used public transportation compared to 7.8 percent of European American women, while 12.3 percent of African American men used public transportation compared to only 3.9 percent of European American men. In Erie county 33.6 percent of African American women used public transportation compared to 9.9 percent of European American women, while 11.8 percent of African American men used public transportation compared to only 4.1 percent of European American men (see Figure 1 for Rochester and Figure 2 for Buffalo). This greater reliance of African Americans on public transportation conforms with well documented patterns, and it is expected to increase their average work-trip time since public transportation is typically a slower and more time consuming mode of travel.

African Americans Spend More Time for the Work Trip

When average work-trip times are examined, the work trips of African American men and women in Monroe County and Erie County are longer than for European American men and women (see Tables 1 and 2 for racial differences in commuting time in Rochester and Buffalo respectively). Significant differences are charted in the graphs displayed in Figures 3 to 8 and discussed in the detailed presentation of findings under the specified headings. African American men in Rochester spend 21.4 minutes versus 19.4 minutes for European American men, while the African American women spend 24.4 minutes versus 17.4 minutes for European American women (t-values are significant at p=<0.01). This means that African American women in Rochester spend 7 minutes longer on the average for their work trip than European American women counterparts. There is also a significant racial gap among women in Buffalo where African American women spend 22.8 minutes versus 18.1 minutes for European American women (Figure 3).

Since more African Americans use public transportation, and since research consistently confirms the pronounced lengthening effect of public transportation on workers' commute times (e.g., U.S. Department of Commerce, 1982; Taylor and Ong, 1995; McLafferty and Preston, 1992), one should make allowance for this, and examine only workers with the same mode of travel. For the remainder of the study, all comparisons about racial differences in work-trip time are restricted to respondents who use a car.

When the work-trip times of private automobile users are compared, there are no remarkable racial differences among men in Rochester and Buffalo, or among women in Buffalo; while the racial gap among women in Rochester reduces to four minutes: i.e., 20.9 minutes for Black women and 16.8 minutes for White women (note however, that this reduced difference is still statistically significant p=<.01--Table 1) (Results are displayed in Figure 4).

The disproportionate use of public transportation by African Americans therefore accounts to a large measure for their relatively longer travel times. Greater dependence on public transportation does not however offer complete explanation for continuing longer work-trip times of many African Americans because when the samples are examined further by workplace location or by income and child status, significant racial differences persist in the commuting time of some auto users. In spite of the use of a private automobile, some African Americans spend a longer time than European American counterparts. The continuing racial differences in travel time among auto users with similar locational and socioeconomic profiles are reported below. First, we look at workers with the same workplace location.

Some African Americans Spend a Longer Time than European Americans with the Same Workplace Location

If a work trip begins and ends in the same area, it is likely to take a shorter time than if the trip starts in one area and ends in another area. For instance, trips that start in the central city and end in the suburbs may take longer than those that begin in the central city and end in the central city. That is, intra-area trips typically take a longer time than opposite direction trips (Hanson and Johnston, 1985; Johnston-Anumonwo, 1995). Racial differences in geographic location of the home and workplace are likely to contribute to racial gaps in commute times, so it is more accurate to compare workers with the same home and work location.

Primarily because of insufficient sample sizes for African Americans who reside outside the central cities of both Rochester and Buffalo, and precisely because the spatial mismatch hypothesis is concerned with the situation of inner city workers, I restrict the remaining comparisons on location effects to respondents with central city homes.

African American male and female auto users who live in the central city and whose workplaces are in Rochester central city, spend longer times than European American counterparts. In Buffalo, African American male auto users whose workplaces are in the central city also spend a longer time than European American men with central city workplaces (Figure 5). According to Figure 5 then, African American respondents who reside and work in central city locations are seen to spend a longer time than the European Americans.

African American Women with Suburban Work Destinations Spend a Longer Time Getting to Work

When we focus on those workers with non central city destinations (i.e., reverse commuters), we find that for both study areas, it is only among women that a very large and significant racial gap in commuting time is observed. African American women residing in Rochester and who commute to work locations outside the central city spend almost 8 minutes longer than European American counterparts (26.6 minutes compared with 18.9 minutes--Figure 6). In Buffalo, African American female reverse commuters spend almost 6 minutes longer than European American counterparts (26.1 minutes compared with 20.3 minutes--Figure 6). It appears that unlike Black men, Black women bear a bigger cost for work trips to non central city destinations when compared to their White counterparts. Generally therefore, the African American women auto users with work destinations in the suburbs of Monroe County and Erie County spend considerable time commuting to work. In addition, some other African American women spend considerably longer time getting to work as discussed next.

Low Income African American Women and Working Mothers Spend a Longer Time Getting to Work

The two socioeconomic factors examined are income and the presence of children in the household. The common expectations are that low wage earners will be less able to afford long commutes, and workers with high levels of domestic duties will try to meet time pressures by avoiding time-consuming commutes. African American and European American workers usually differ in income level and household composition. African American workers should have shorter commutes if their lower incomes disallow long commutes, or if they have greater domestic obligations, such as household responsibilities associated with children, that places more demand on their time.

First, I examine racial disparities in work trip length of workers in two income groups. As shown in Figure 7, African American low income female workers in Monroe County spend a longer time for their work trip than European American low income female workers (22.2 minutes versus 15.3 minutes--significant at p=<.01 in Table 1). There are no other significant racial disparities among workers with similar income levels. The substantial racial difference in commute time among Monroe County's low-income women (seven minutes) is note worthy. It indicates that these Black women are enduring relatively long commutes to low-waged jobs.

For child status, the commute times of workers in households without children are compared with those in households with children (school-aged children). Significant racial differences are observed only among women (see Figure 8). African American mothers in Monroe County spend 5 minutes longer time than European American mothers (20.6 versus 15.4 minutes (significant at p=<.01)); and African American mothers in Erie County spend almost 4 minutes longer time than European American mothers (19.7 versus 16.0 minutes (significant at p=<.01 in Table 1).

This set of findings run contrary to common expectations that those workers earning low incomes or with household responsibilities for children will have shorter commutes. In spite of the constraints, low-income African American women in Monroe County and those with children in both counties put up with longer commutes than their European American counterparts.

Keeping in mind that since these comparisons are conducted only for auto users, the findings show that even when access to an automobile is not a hindrance, many African American workers (especially female workers) in Rochester and Buffalo still bear a bigger time cost than European Americans. Also, with the widest racial gap being observed among women, the results counter the welfare queen stereotype of Black women since the women in the analysis are all employed. Lastly, although the time differences may appear small, the time cost for African American workers is not trivial. For instance the cumulative time of the two-way work trip is quite considerable, and it amounts to time lost from other tasks. In the final section of the paper, I elaborate on the implications of these findings.

CONCLUSIONS AND DISCUSSION

According to the findings of this study, African Americans in Monroe and Erie counties as a group do have longer commutes than do European American workers. As expected, the reason lies largely in the greater reliance of African Americans on public transportation. Once travel mode is taken into account, the racial difference in commuting time among most auto users is negligible. Taylor and Ong (1995) also note that among workers who have access to automobiles, there is reduced racial gap in travel time.

By failing to conduct race-specific and sex-specific comparisons, many previous studies masked continuing significant differences among sub groups of workers, and by ignoring the impact of workplace location on the duration of the work trip, racial disparities in locational access to jobs are understated. Hence, one main contribution of this study is to demonstrate that taking into account the workplace location provides a clear assessment of the existence and nature of commuting difficulties that African Americans face when they live in central cities and/or work in suburbs.

Most of the African Americans in both counties do work in the central cities. And while the observed racial disparities are less pronounced among men, Black men who live and work in the central cities still spend about three minutes more for their work trip than do White counterparts, but note that the average trip length is under 20 minutes. Among women in Rochester, Black women with intra-city trips also spend a longer time than the White counterparts. But the more striking finding for African American women is the substantially longer commuting time of reverse commuters (approximately 26 minutes). Indeed, a central line of inquiry in the study is the specific impact of suburban employment on the commutes of African Americans. The need to work in suburban destinations (i.e., outside the central city) of Rochester and Buffalo in 1980 imposes a disproportionate commuting time burden on inner city Black women than on White women. This study thus provides support of a spatial mismatch in Black women's geographical access to employment in Monroe and Erie counties in 1980. It validates the early conclusions of McLafferty and Preston (1991) that many African American women experience a very insidious form of spatial mismatch and face significant transportation and locational barriers in traveling to work.

The longer times spent by Black women may be expected to decrease if Black women have more access to private automobiles. However, the longer commutes of African American women who reverse commute suggest that it is reasonable to speculate that as employment opportunities continue to expand in suburban locations and not in central city locations, African American women (even those who use a car) are still likely to suffer the inconvenience of significantly longer commutes to suburban workplaces than European American women.

There is reason to believe that concerns about spatial mismatch remain current in U.S. cities. For example, a recent study presents evidence of spatial mismatch as well as evidence of the negative treatment of Blacks by suburban police officers and White residents as testimony of the multiple barriers facing African Americans in gaining access to employment opportunities in suburban Detroit (Turner, 1997). Similarly, in Buffalo, the current validity of both an automobile mismatch and a spatial mismatch for African American women proved tragically true in the case of a Black woman who was killed while crossing an expressway in suburban Buffalo on her way to the shopping mall where she was employed. This particular case had racist underpinnings because the management of the suburban mall seemed to have pursued explicit policy decisions preventing buses coming from inner city Buffalo from stopping at the mall.

One should note that the results for these two Upstate New York cities corroborate those of other cities. For example, Shumway and Cooke's (1991) analysis of 1980 PUMS data for three large Midwest cities--Chicago, Cleveland and Detroit--revealed that central city residents experience constrained access to employment. If more jobs were available in central cities, there would be less need to reverse commute to reach suburban jobs. Alternatively, if Blacks had unhindered access to suburban housing, the racial disparity in locational access to jobs would be lessened. It is inaccurate to minimize the importance of locational access to Black employment outcomes, and it would be premature to abandon inquiries about the possible role of location in the mismatch of workers and jobs.

Apart from difficult access to suburban work destinations, the study also finds evidence of other travel time constraints for Black women. Specifically, in Monroe County, it is the low income Black women who have relatively long commutes, while in both Monroe County and Erie County, it is African American mothers who have relatively long commutes. It is clear from the study that not all African American women face the constraints of long journeys to low-paying jobs, and not all African American mothers have very long travel times. But the emerging profile of some Black women who combine parenthood with wage earning and endure long commutes to suburban destinations suggest that policy makers need to recognize the efforts of these workers and reward their diligence with remunerative job opportunities.

It is essential to stress that like all studies that use commuting data, this study understates the general problem of access to jobs since it excludes the unemployed, many of whom are unemployed probably due to locational constraints. However, the use of journey-to-work data, and the focus on travel time in particular is appropriate. Time is a resource. In some instances, time is money, therefore lost time is lost money. Much of the extra time that African Americans in Monroe and Erie counties spend longer than European Americans range from 3 minutes to over 6 minutes. Using simple calculations, this can be extrapolated into between 6 minutes and 12 minutes per day, or between 30 minutes and 60 minutes a week, or between 25 and 50 hours a year--the equivalent of about a week. Cast in this light, the longer commute times of African Americans can be interpreted as constituting a race tax burden.

In conclusion, this analysis of work trips in Rochester and Buffalo complements as well as expands the empirical literature on racial differences in the commuting patterns of American urban residents. The study has shed additional light on analyses about racial differences in commuting by highlighting the significance of several factors--race, gender, means of transportation, residence, workplace location, income, and presence of children--in the job access constraints of groups of African Americans. These factors have to remain central in upcoming analyses of other more up-to-date data particularly for informing contemporary policy debates such as welfare reform. For now, the evidence from Rochester and Buffalo about the journey to work of African American workers in spite of transportation, location, income or household responsibility constraints strongly counters prevailing stereotypes of welfare dependent African Americans.

TABLE 1 AVERAGE WORK-TRIP TIME (MINUTES) MONROE COUNTY, 1980

                               Black   White         Black   White
                               Men     Men           Women   Women

Full Sample                 N  261     3755          282     2769
  Travel Time                   21.4     19.4   **    24.4     17.4  **
Auto Users Only             N  201     3421          197     2394
  Travel Time                   19.9     19.3   ns    20.9     16.8  **
Work Location
Central City Residence to
Central City Workplace      N  114      529          107      380
  Travel Time                   17.5     14.7   **    18.0     13.9  **
Central City Residence
to Suburban Workplace       N   35      206           48      142
  Travel Time                   22.3     20.4   ns    26.6     18.9  **
Income Status
Low Income (= <$10,000)     N   41      599           82     1261
  Travel Time                   18.6     16.5   ns    22.2     15.3  **
High Income (= >$10,000)    N  154     2641          108     1047
  Travel Time                   20.5     20.0   ns    20.1     18.5  ns
Child Status
No Child in Household       N   53     1284           49      959
  Travel Time                   20.6     19.5   ns    20.2     17.9  ns
Child Present               N   63     1089           76      778
  Travel Time                   21.5     19.1   ns    20.6     15.4  **

** -- difference is significant at 95 percent level of confidence or
higher.
ns -- difference is not significant.

TABLE 2 AVERAGE WORK-TRIP TIME (MINUTES) ERIE COUNTY, 1980

                                Black   White         Black   White
                                Men     Men           Women   Women

Full Sample                 N   304     4876          321     3551
  Travel Time                    22.5     20.9   ns    22.8     18.1  **
Auto Users Only             N   241     4442          186     2924
  Travel Time                    22.0     20.9   ns    18.5     17.1  ns
Work Location
Central City Residence to
Central City Workplace      N   121      596          133      477
  Travel Time                    19.7     16.4   **    16.1     15.4  ns
Central City Residence
to Suburban Workplace       N    88      373           34      173
  Travel Time                    24.6     24.0   ns    26.1     20.3  **
Income Status
Low Income (= <$10,000)     N    66      889           99     1902
  Travel Time                    19.5     17.7   ns    17.6     16.0  ns
High Income (= >$10,000)    N   168     3335           79      906
  Travel Time                    22.3     21.9   ns    19.2     19.7  ns
Child Status
No Child in Household       N    88     1803           60     1281
  Travel Time                    22.5     20.8   ns    18.8     18.0  ns
Child Present               N    72     1438           78      936
  Travel Time                    21.4     20.9   ns    19.7     16.0  **

** -- difference is significant at 95 percent level of confidence or
higher.
ns -- difference is not significant.

Figure 1: Racial Differences in Means of Transportation (Monroe County)

                       Black Men  White Men   Black Women  White Women

Private Automobile        77%        91%          70%          86%
Public Transportation     12%         4%          25%           8%
Walk and Other Means      11%         5%           5%           6%

Note: Table made from pie chart.

Figure 3: Average Work-Trip Time (Full Sample)

Minutes

Monroe County
Black Men                21.4
White Men                19.4
Black Women              24.4

Erie County
White Women              17.4
Black Women              22.8
White Women              18.1

Note: Table made from bar graph.

Figure 4: Average Work-Trip Time for Auto Users

Minutes

Monroe County
Black Women              20.9
White Women              16.8

Note: Table made from bar graph.

Figure 5: Average Work-Trip Time for Central City Destinations

Minutes (auto users only)

Monroe County
Black Men                        17.5
White Men                        14.7
Black Women                      18

Erie County
White Women                      13.9
Black Men                        19.7
Black Women                      16.4

Note: Table made from bar graph.

Figure 6: Average Work-Trip Time for Suburban Destinations

Minutes (auto users only)

Monroe County
Black Women                      26.6
White Women                      18.9

Erie County
Black Women                      26.1
White Women                      20.3

Note: Table made from bar graph.

Figure 7: Average Work-Trip Time for Low Income Women

Minutes (auto users)

Monroe County
Black Women                   22.2
White Women                   15.3

Note: Table made from bar graph.

Figure 8: Average Work-Trip Time for Mothers

Minutes (auto users only)

Monroe County
Black Women                    20.6
White Women                    15.4

Erie County
Black Women                    19.7
White Women                    16

Note: Table made from bar graph.

WORKS CITED

Cooke T. J. and J. M. Shumway. 1991. Developing the spatial mismatch hypothesis: Problems of accessibility to employment for low wage central city labor. Urban Geography 12:310-323.

Darden, J. T. 1990. Differential access to housing in the suburbs. Journal of Black Studies 21:15-22

Denton, N. A. 1994. Are African Americans still hypersegregated? In Residential Apartheid: The American Legacy, ed. R. D. Bullard, J. E. Grigsby III and C. Lee, 49-81. Los Angeles: Center for Afro-American Studies, University of California.

Dubin, R. 1991. Commuting patterns and firm decentralization. Land Economics 67:15-29.

Ellwood, D. T. 1986. The spatial mismatch hypothesis: are there teenage jobs missing in the ghetto? In R. B. Freeman and H. J. Holzer (eds.) The Black Youth Employment Crisis. Chicago: University of Chicago Press, pp. 147-187.

Gordon, P., A. Kumar, and H. W. Richardson. 1989. The spatial mismatch hypothesis: Some new evidence. Urban Studies 26:315-26.

Greytak, D. 1974. The journey to work: racial differences and city size. Traffic Quarterly, Vol. 28 No. 2 241-256.

Hanson, S., and I. Johnston. 1985. Gender differences in work-trip length: explanations and implications. Urban Geography 6:193-219.

Holzer, H. J. 1991. The spatial mismatch hypothesis: What has the evidence shown? Urban Studies Vol. 28 No. 1 105-22.

Ihlanfeldt, K. R. 1992. Intraurban wage gradients: Evidence by race, gender, occupational class, and sector. Journal of Urban Economics 32:70-91.

Johnston-Anumonwo, I. 1995. Racial differences in the commuting behavior of women in Buffalo, NY, 1980-1990. Urban Geography 16:23-45.

_________. 1997. Race, gender, and constrained work trips in Buffalo, NY, 1990. Professional Geographer 49:306-317.

Kain, J. 1968. Housing segregation, Negro employment, and metropolitan decentralization. Quarterly Journal of Economics Vol. 82 No. 2 175-97.

Leonard, J. S. 1987. The interaction of residential segregation and employment discrimination. Journal of Urban Economics, Vol. 21 323-346.

Massey, D. S., and Z. L. Hajnal. 1995. The changing geographic structure of black-white segregation in the United States. Social Science Quarterly Vol. 76 No. 3 527-42.

McLafferty, S., and V. Preston. 1991. Gender, race and commuting among service sector workers. Professional Geographer 43:1-15.

_________. 1992. Spatial mismatch and labor market segmentation for African American and Latina women. Economic Geography 68:406-31.

Pisarski, A. 1987. Commuting in America. Westport, CT: Eno Foundation

Preston, V., S. McLafferty, and E. Halmilton. 1993. The impact of family status on black, white and Hispanic women's commuting. Urban Geography 14:228-50.

Taylor, B. D., and P. M. Ong. 1995. Spatial mismatch or automobile mismatch? An examination of race, residence and commuting in US metropolitan areas. Urban Studies 32:1453-73.

U.S. Department of Commerce. 1963. U.S. Census of Population: 1960, Volume 1, Characteristics of the Population, Part 34, New York. Washington, DC: U.S. Bureau of the Census.

U.S. Department of Commerce. 1979. The Journey to Work in the United States: 1975. Current Population Reports P-23, Special Studies No. 99. Washington, DC: U.S. Bureau of the Census.

U.S. Department of Commerce. 1982. The Journey to Work in the United States: 1979. Current Population Reports P-23, Special Studies No. 122. Washington, DC: U.S. Bureau of the Census.

_________. 1983a. 1980 Census of Population and Housing, Buffalo, New York. PHC80-2-106. Washington, DC: U.S. Department of Commerce, Bureau of the Census.

_________. 1983b. 1980 Census of Population and Housing, Rochester, New York. PHC80-2-306. Washington, DC: U.S. Department of Commerce, Bureau of the Census.

_________. 1983c. Census of Population and Housing, 1980: Public-Use Microdata Samples, Technical Documentation. Washington, DC: U.S. Bureau of the Census.

U.S.A. Today. 1991. By the numbers, tracking segregation in 219 metro areas. November 11:3a.

Zax, J. F. 1990. Race and commutes. Journal of Urban Economics 28:336-348.

Ibipo Johnston-Anumonwo (1)

(1) Ibipo Johnston-Anumonwo (Ph.D) is an Associate Professor in the Department of Geography at the State University College at Cortland, NY.