Table 2: Commute Mode

(3) Person Trips

Overall, as income rises, people make more trips and travel longer distances (Rosenbloom 1994). However, variations in travel patterns by sex, income, and family structure help to explain the travel behavior of low-income, female-headed households, of which welfare recipients are a subgroup.

Table 3 shows daily person trips and miles for seven different subgroups of the population. These data are not directly comparable since the individual studies rely on three separate surveys, were conducted in different years, and focus on slightly different population groups. However, these data can be used to support the overall following conclusion--urban women tend to make more trips than men but, on average, travel fewer miles (Rosenbloom 1994).

Hu and Young (1999) show that trip making among single-parent households with young children is higher than for all persons and for all adults without children. Rosenbloom (1994) also shows that low-income, working urban women make more trips and travel more miles than do all working women. Rosenbloom's (1994) data for female-headed households with young children do not control for employment. While she finds that this group of women makes fewer trips and travels fewer miles than working, low-income women, in all instances, they make more trips and travel fewer miles than comparable men (Rosenbloom 1994). Finally, regardless of income, single mothers travel fewer miles and make more trips than comparable men (Rosenbloom 1994).

Table 3: Daily Person Trips and Miles

Women tend to make more trips than do men because they are disproportionately responsible for household-sustaining activities including trips to the daycare center, the grocery store, and other similar destinations (McGuckin and Murakami 1999). Some of these trips are made as part of trip chains, a series of trips in a tour anchored by home or work. Sixty-one percent of women make at least one stop on the trip home from work as compared to approximately 46 percent of men (McGuckin and Murakami 1999). Twenty-eight percent of women make two or more stops on the way home from work as opposed to 18 percent of men (McGuckin and Murakami 1999). Additionally while only one-fifth of men stop on the way to work, nearly one-third of women stop on the way to work in the morning (McGuckin and Murakami 1999). This figure may be related to the fact that women drop-off or pick-up additional passengers with a greater frequency than their male counterparts (Federal Highway Administration 1995; Taylor and Mauch 1996.) Interestingly, sex differentiation in this pattern remains constant even when controlling for single fathers and single mothers. Single mothers tend to make more child-serving trips than do single fathers. For example, 65 percent of single women with children less than five years of age stop on the way to work compared to only 33 percent of single fathers (McGuckin and Murakami 1999).

(4) Commute Distance and Time

Typically, commute distance is positively correlated with earnings; therefore, on average higher-income commuters travel longer distances than lower-income commuters (Taylor and Ong 1995). A number of factors explain this relationship. First, higher-wage jobs tend to be more dispersed throughout the metropolitan area (Simpson 1992). Second, higher income leads to the desire for more housing and land, the relative costs of which are significantly lower in the suburban fringe of metropolitan areas than in the central city (Muth, 1969; Simpson 1992). Third, higher-income households often seek residential amenities such as high quality schools, low crime rates, and recreational facilities, all of which are more typical of newer suburbs than older, inner-city neighborhoods. Fourth, higher-wage workers have greater access to cars, lowering the opportunity costs of traveling to work by reducing commute time for any given distance (Taylor and Ong 1995). Finally, the positive correlation between earnings and commute distance occurs because competitive labor markets generate compensating variation in wages to offset non-pecuniary costs to workers, such as those related to long-distance commutes (Viscusi 1992).

Therefore, on average, welfare recipients and low-income commuters have significantly shorter commute distances than all commuters. Rosenbloom (1994) finds some variation in the relationship between income and commute distance. However, she finds that those in the lowest income category (earning less than $5,000) commute the shortest distances, just under six miles (Rosenbloom 1994). Murakami and Young (1999) find that approximately 66 percent of the trips made by single-parents fall within a three-mile radius compared to 49 percent of the trips made by non-low-income individuals.

(5) Time of Travel (Peak Versus Off-Peak)

Many of the work trips made by low-income women occur during off-peak hours when transit service may be limited or, in some cases, non-existent. On average, less educated women are more likely to work non-standard hours than are other women and, therefore, are more likely to travel during evenings, nights, and weekends. Figure 4 shows that women ages 18 to 34 with a high school education or less are the least likely to work a fixed daytime schedule (Presser and Cox 1997). Furthermore, women with pre-school age children are one and a half times more likely to work non-standard hours compared to women without children (Presser 1995). The results are similar for welfare recipients in Los Angeles; only two-thirds of employed welfare recipients leave for work during the peak morning hours (UCLA Lewis Center for Regional Policy Studies 2000).

Figure 4: Fixed Daytime Employment Schedule, Women, 1991

There are two principal reasons why women tend to work non-standard schedules. Women are disproportionately concentrated in service sector jobs in which non-standard hours are prevalent (Presser and Cox 1997). For example, more than 45 percent of janitors and cleaners and more than 37 percent of waitresses work non-standard hours (Presser and Cox 1997). While many young women work non-standard schedules because it is a requirement of their jobs (40%), a high percentage (27%) also report that they prefer these hours so that they can arrange better care for their children (Presser and Cox 1997). The percentage of women who prefer non-standard schedules rises to 31 percent among young women whose youngest children are under 5 years of age (Presser and Cox 1997). Preferences by some women for nontraditional work hours coupled with the expected growth in jobs requiring nonstandard work schedules suggest an increase in off-peak commuters, particularly among women with young children.

(6) Expenditures

Figure 5 shows that households receiving public assistance spend 15 percent of their annual expenditures on transportation compared to 19 percent among households who do not receive public assistance (Passero 1996). The difference in expenditure patterns can be explained, in part, by employment status. Among households receiving public assistance and without working members, spending on transportation drops to 9.5 percent of annual expenditures compared to 19.1 percent among households that receive public assistance but have one or more working members (Passero 1996). These differences in expenditures may be narrowing as increasing numbers of welfare recipients are required to work. In a more recent study, Tan (2000) shows only a .01 percent difference in the percentage of annual transportation expenditures between assisted and non-assisted families.

Figure 5: Transportation as a Percentage of Annual Expenditures,
1st Quarter 1992 to 1st Quarter 1994

Figure 6: Annual Expenditures of Publicly-Assisted Families, 1998

Nevertheless, as of 1998, publicly-assisted households were spending more on transportation than any other expenditure except housing and food (Tan 2000). However, when welfare recipients are surveyed about their transportation policy preferences, they typically prefer improvements in transit service, such as more frequent service, rather than lower fares (UCLA Lewis Center for Regional Policy Studies 1999).

Job Access and Welfare Recipients

Current policies aimed at increasing welfare recipients' spatial access to employment are largely predicated on the notion of a spatial mismatch between the residential location of welfare recipients in central cities and increasing job opportunities in suburban neighborhoods. To assess whether welfare recipients face a "spatial mismatch," a number of studies have examined the geographic location of welfare recipients in relation to low-wage jobs, social and employment services, and public transit. These studies find that many welfare recipients face a spatial mismatch; most cities have at least some inner-city neighborhoods where unemployment rates are high, jobs are few, and welfare recipients live distant from employment opportunities. Moreover, some studies show that improved access to jobs results in better employment outcomes for welfare recipients.

However, the spatial mismatch is likely more relevant in metropolitan areas with high levels of residential segregation and poor transportation options for reverse commuters, and less relevant in smaller, more centralized metropolitan areas (Ihlanfeldt & Sjoquist, 1998).

First proposed by John Kain (1968) in the 1960s, the spatial mismatch hypothesis was an effort by Kain and other scholars to explain the deplorable economic position of low-skilled African-Americans living in central-city areas. Proponents of the spatial mismatch hypothesis argue that the shift in the demand for labor toward suburban areas, racial discrimination in housing markets, and poor transportation linkages between city and suburb, among other factors, isolate African-Americans in poor, central-city neighborhoods. Therefore, the argument follows that joblessness and low wages among African-Americans are, in part, the result of their systematic spatial separation from low-wage job opportunities increasingly located in suburban areas. Similar to African-Americans, welfare recipients are disproportionately concentrated in central areas distant from suburban job opportunities. The merits of the spatial mismatch hypothesis have been examined in more than 75 studies and in at least 8 comprehensive literature reviews.5 With some exceptions, the evidence supports the negative effects of the spatial mismatch hypothesis on many African-Americans.

Empirical support for the spatial mismatch hypothesis has formed the intellectual foundation for the scholarship on welfare recipients and their access to jobs. Extending from the spatial mismatch hypothesis literature, researchers have produced a series of ecological studies depicting the spatial separation between welfare recipients and low-wage jobs. While not directly testing the spatial mismatch hypothesis and its application to welfare recipients, these studies rely on maps to graphically depict the location of welfare recipients, low-wage jobs, and, frequently, the public-transit service linking the two. These analyses highlight the diversity in the degree and type of spatial mismatch within and among U.S. metropolitan areas.

Table 4 summarizes some of these studies and is followed by short descriptions of the findings for select U.S. metropolitan areas. As the descriptions suggest, in some metropolitan areas, such as Detroit, welfare recipients experience a high degree of separation between residential locations in the central city and outlying suburban job opportunities. In other metropolitan areas such as Los Angeles or Boston, welfare recipients experience more localized or neighborhood level mismatches. However, in almost all cases, the studies find that many welfare recipients live in neighborhoods that are distant from low-wage jobs and poorly served by existing fixed-route, public transit.

Table 4: Degree and Type of Spatial Mismatch

(1) Atlanta, Georgia

Studies suggest that welfare recipients living in Atlanta face a substantial spatial mismatch. This mismatch includes a traditional central city-suburban mismatch as well as a mismatch across suburban areas (Rich 1999). Over the past few decades, Atlanta has become a large, polycentric metropolitan region. While the suburban job market and population have grown significantly, this growth has not been uniform throughout the outer suburban rings of the metropolitan area. The majority of low-income families and welfare recipients live within the official city limits; however, a number of suburban areas are home to concentrations of low-income families. Decentralized and multi-nodal suburban growth has resulted in a metropolitan landscape characterized by wealthy suburbs abutting low-income suburbs as well as the more traditional concentration of low-income families in the central city (Rich 1999).

In 1970, the City of Atlanta had 55 percent of the region's employment, but by 1990 the City's share of employment dropped to 29 percent (Pugh 1999). Approximately 12 percent of the population but 71 percent of all TANF recipients live within the Atlanta city limits (Rich 1999). Although the region has some of the highest employment and population growth in the United States, due to limited transit service and long commutes, welfare recipients living within the city limits have difficulty traveling to the expansive exurban employment nodes. Additionally, welfare recipients who live in poor, outer-ring suburbs face many difficulties when trying to travel to jobs in dispersed suburban areas (Rich 1999; Ihlanfeldt 1993). Therefore, welfare recipients living in multi-nodal Atlanta face numerous transit and access barriers whether they live within the central city or in one of its job-poor suburbs.

(2) Chicago, Illinois

Different than most other cities studied, Chicago, which has significant exurban and suburban growth, has also experienced notable central city growth and boasts an extensive multi-nodal transit system. The extensive transportation system in Chicago coupled with a strong regional economy significantly eases the impact of the spatial mismatch. As a whole, the Chicago area population is extremely transit dependent and a large number, 25 percent of the population, already reverse commute. The employment problem in Chicago's inner city may stem more from inadequate access to information and education than from transportation barriers (Pugh 1999).

(3) Cleveland, Ohio

Cleveland has experienced a declining central city while its surrounding suburban areas have dispersed, creating a large and distant blanket of suburban development. However, unlike some other cities, the net population of the Cleveland metropolitan area has declined since 1970 and new immigrant populations have not replaced the vacancies in the inner city (Bania, et al. 1999). Exacerbating the low economic regional growth is the lack of coordination between the transportation carriers that serve the eight counties that comprise the Cleveland metropolitan area, each of which operates independently (Bania, et al. 1999).

(4) Detroit, Michigan

Detroit offers a paradigmatic example of spatial mismatch in Midwestern, rustbelt cities at the end of the twentieth century. Similar to Cleveland, it has experienced significant central-city decline--a paucity of employment opportunities, decreasing population, concentrated poverty, and a large concentration of welfare recipients and families receiving other forms of public assistance--as well as extensive suburban growth (Allard and Danziger 2000; Laube, et al. 1997).

Over the past few decades, as the central city declined at an alarming rate, suburban Detroit experienced nearly all of the regional job growth (Laube, et al. 1997). The decline in central-city employment and the rise in suburban employment and population have resulted in a difficult commute for Detroit's poor, inner-city population (Laube, et al. 1997). Allard and Danziger (2000) found that areas with high concentrations of poverty and welfare receipt were farthest from employment opportunities.

(5) Los Angeles, California

Geographic data for Los Angeles show that there is a spatial mismatch between the residential locations of welfare recipients and the locations of low-wage jobs. However, the mismatch is not necessarily between central city and suburbs. Welfare recipients' access to employment varies depending on their residential location and their commute mode (Blumenberg and Ong 2001). Many welfare recipients in Los Angeles live in central-city neighborhoods adjacent to the central business district (CBD) and are able to reach many jobs within a reasonable commute by either car or public transit. In these neighborhoods, spatial access to jobs is good and high unemployment rates are likely the result of a myriad of other employment barriers (Danziger et al. forthcoming). However, some welfare recipients live in job-poor, central-city neighborhoods where, if transit-dependent, they face long and difficult commutes that limit their likelihood of finding and sustaining employment even if they are traveling to destinations within the central city (Blumenberg and Ong 2001).

(6) Milwaukee, Wisconsin

Like Atlanta and Detroit, Milwaukee has a very high incident of spatial mismatch and racial segregation. Unlike Atlanta and similar to Detroit, the spatial mismatch is limited to the classic central city-suburb dichotomy. Furthermore, the root of the extreme spatial mismatch in Milwaukee seems to have less to do with expansive suburban development, although its suburbs are expansive, than with entrenched racial tensions between the inner-city, African-American population -- 97 percent of the African-American population lives in the central-city -- and the outlying white suburban community. Among one of the most significant barriers to employment in Milwaukee is the non-coordination of central city and suburban transit providers (Pawasarat 1997).

(7) Philadelphia, Pennsylvania

While not subject to a severe spatial mismatch, welfare recipients living in Philadelphia continue to face significant spatial barriers to employment. Pugh (1999) links this to two problems -- localized or "micro" gaps in public transit and employment as well as slow job growth in the entire region. In addition to low overall growth, Philadelphia, like Detroit and Milwaukee, is extremely racially segregated. In cases where there is low regional employment growth, the studies on both Cleveland and Philadelphia suggest that adequate transportation may be extremely important (Pugh 1999).

(8) Rural Areas

Welfare recipients living in rural areas face unique challenges in making the transition into the labor market. Rural areas tend to offer fewer job opportunities; average earnings are lower in rural than urban areas; and, in some counties, available jobs are concentrated in the highly seasonal agricultural sector where the demand for labor fluctuates from month-to-month (Kaplan 1998; Weber and Duncan nd). Without local employment, rural welfare recipients may find themselves living distant from urban employment centers and in areas where public infrastructure--public transportation, social service programs, and other services--is minimal. On average, rural residents--including welfare recipients--must travel longer distances to work, to reach services, and to make household-sustaining trips (Dewees 2000).

Studies suggest that close to 40 percent of all U.S. rural residents live in areas without public transportation and another 28 percent of rural residents live in areas with low levels of transit service (Rucker 1994). Many studies assert that the opportunities of rural welfare recipients vary by community and are based on the relative access that welfare recipients have to jobs and support services (Fletcher et al. 2000). However, very little is known about the role of transportation, in particular, in limiting the employment opportunities of welfare recipients. Most of the existing evidence is anecdotal. For example in one study of welfare recipients in Iowa, a welfare participant is quoted as stating:

I could have had a job on the 15th [of the month] but I didn't have a vehicle. It takes about half an hour to 45 minutes just to get downtown on the bus. Then another 20 minutes after transferring to the appropriate bus. The buses don't even start out here until 6:15 in the morning. So how the heck can I get to work by 6:30? (Fletcher et al. 2000:15)

The unique conditions of rural areas have motivated the adoption of particular types of policies aimed at improving rural transit service. These include demand-responsive service where customers call at least 24 hours in advance to arrange an appointment to be transported to a particular site, vanpooling, the coordination of travel to a single site, and captivated-rate bus service where a bus company receives funds based on the number of recipients in the region and is required to provide services to all recipients in that region.

Conclusion

These spatial analyses of welfare recipients highlight the diversity in the degree and type of spatial mismatch within and among various U.S. metropolitan areas. In some areas, such as Detroit, welfare recipients experience a distinct central city-suburban mismatch. In other metropolitan areas such as Los Angeles and Philadelphia, welfare recipients experience a more localized or neighborhood level mismatch. While these variations are significant, the studies find that most cities have at least some inner-city neighborhoods where unemployment rates are high, jobs are few, and welfare recipients live distant from employment opportunities.

There is a growing literature that suggests that spatial isolation from jobs leads to adverse economic outcomes for welfare recipients. Ong and Blumenberg (1998) show that long distance commutes in the low-income community are related to lower earnings.

Blumenberg and Ong (1998) show that access to employment leads to lower welfare usage rates. Allard and Danziger (2000) show that access to employment opportunities is related to higher employment among Detroit welfare recipients. Finally, the research conducted by Cervero et al. (forthcoming) in Alameda County, California shows that automobile access leads to higher employment among welfare recipients; so, too, does highly accessible transit when it is within a walkable distance from residences.

Cars, Public Transit, and Welfare Recipients

In 1997 the U.S. Department of Health and Human Services reported that 6.7 percent of all welfare recipients owned vehicles (U.S. Department of Health and Human Services 1997). This figure has been widely publicized and used by federal agencies to promote the use of federal funds to enhance public transit services for welfare recipients and other low-income commuters. However, other research indicates that a higher percentage of welfare recipients have access to automobiles than initially reported and that access to cars, rather than public transit, is strongly related to positive employment outcomes. While far from a problem-free solution, cars offer flexibility in trip making which is particularly important in searching for jobs and in making child-serving trips. They also offer women personal security which is particularly important when traveling during off-peak hours when service may be limited, riders few, and waiting at dark, isolated stops is dangerous.

(1) Automobiles

Auto ownership among all households has been increasing over time. Figure 7 shows the steady decline in zero-vehicle households from 21 percent in 1969 to 8 percent in 1995 (Hu and Young 1999). Still, access to automobiles varies dramatically by income, race, ethnicity, and family structure. One-fifth of all low-income households, those with incomes below $25,000, do not own vehicles compared to only one percent of higher income households, those with incomes over $55,000 (Hu and Young 1999). African-Americans have the highest incidence of zero-vehicles, comprising approximately 12 percent of all households and 35 percent of households without vehicles (U.S. Bureau of Transportation Statistics 1995). Single parents are also overrepresented among zero-vehicle households, comprising 5 percent of the population and 12 percent of households without cars (U.S. Bureau of Transportation Statistics 1995).

Figure 7: Zero-Vehicle Households, 1969-1995

As Table 5 shows, car ownership rates among welfare recipients vary greatly across studies. Regardless, these figures show higher rates of auto ownership than initially reported by the U.S. Department of Health and Human Services (1997) but much lower rates of auto ownership compared to the vehicle ownership rates of the general population.

Table 5: Welfare Recipients and Auto Ownership

Differences in vehicle ownership rates may be due to a number of factors. Welfare administrative data may underestimate vehicle ownership rates since they are derived from vehicle asset reports. Welfare recipients may be inclined to underreport their vehicle ownership since, in many states, welfare recipients are not eligible for benefits if the value of their vehicles exceeds a certain dollar amount. In California, this figure has been set at $4,650. Vehicle ownership data from other sources may also underestimate vehicle use since it does not incorporate access to vehicles that are owned by other household members or access to cars owned by friends, family, or neighbors. For example, in Los Angeles, while 55 percent of all welfare recipients own vehicles, 68 percent of recipients commute by car (UCLA Lewis Center for Regional Policy Studies 2000).

In addition to owning fewer automobiles, low-income households typically own older, less reliable vehicles. Murakami and Young (1997) find that low-income, single parent households own vehicles that are, on average, 11 years old compared to the vehicles owned by non-low-income households that are, on average, 8 years old. In Los Angeles, 69 percent of the cars owned by welfare recipients were 10 years old or older (UCLA Lewis Center for Regional Policy Studies 2000). Over one-half of all survey respondents in Los Angeles (55%) had at least one mechanical problem, and 23 percent had three or more mechanical problems over the last three months that prevented them from reaching their destinations (UCLA Lewis Center for Regional Policy Studies 2000). Fifty-nine percent of welfare recipients in Los Angeles stated that mechanical problems were one of the two major problems with owning cars (UCLA Lewis Center for Regional Policy Studies 2000).

Despite the problems associated with auto ownership, an increasing number of studies suggest that access to cars is associated with positive employment outcomes. Raphael and Stoll (2000) examine employment gaps across racial and ethnic groups. They find that raising minority car-ownership rates to white car-ownership rates would narrow the employment differential for Blacks and Latinos relative to Whites (Raphael and Stoll 2000). With respect to welfare recipients, Ong (1996) finds that those with automobiles have a statistically higher likelihood of employment and higher mean hours and monthly earnings compared to welfare recipients without automobiles. In an analysis of Alameda County, Cervero et al. (forthcoming) also find that car ownership significantly increases the probability that welfare recipients move from welfare into the labor market. While cars are not required for most jobs (Holzer and Danziger 1998), they may enable welfare recipients to search more widely for employment than can welfare recipients without cars (Danziger and Holzer reported in O'Regan and Quigley 2000).6 Moreover, they typically increase the number of available jobs located within a reasonable commute distance (Blumenberg and Ong 2001).

(2) Public Transit

In the face of insufficient private vehicle ownership and a general reluctance on the part of policymakers to advance automobile-centered transportation policy, public transit provides many low-income commuters with an essential service. Public transit usage among all commuters is quite low; approximately 2 percent of all trips and 3.5 percent of work trips are made on public transit (Hu and Young 1999). In contrast, the percentage of welfare recipients reliant on public transportation for their daily commute is markedly higher, close to 50 percent. (See Table 5.) For transit-dependent welfare recipients, the success of welfare reform may well rest on the level and quality of transit service in their neighborhoods.

The high percentage of transit-dependent welfare recipients suggests that there is a role for public transit in assisting welfare recipients in their work-related travel. Fixed route public transit is best suited for travel in job-rich areas with high concentrations of welfare recipients (Blumenberg and Ong 2001). If welfare recipients live in job-rich neighborhoods, public transit may allow them access to a fair number of jobs (Blumenberg and Ong 2001). Also, once transit-dependent welfare recipients move from the job search into employment and are traveling to a single employment destination, they typically travel with greater ease (UCLA Lewis Center for Regional Policy Studies 2000).

While public transit has its problems, some welfare recipients find that public transit can meet their travel needs provided that service levels are high--in other words, that buses run frequently and cover a large and well-coordinated area (UCLA Lewis Center for Regional Policy Studies 2000). Therefore, many of the policy discussions related to public transit center on service improvements such as increasing levels of service in low-income neighborhoods, better integrating of services between and among transit agencies, and reducing fares (American Public Transit Association 1999; Community Transportation Association of America, 1999).7 Also, recent legal actions by the NAACP-Legal Defense Fund and the Bus Riders Union against the Metropolitan Transit Authority (MTA) in Los Angeles highlight efforts to improve transit service for low-income riders (Brown, 1998). Bus proponents argued successfully that the MTA's distribution of transit funds shifted resources to rail and away from buses and, in doing so, negatively affected the levels of transit service available to low-income riders (Brown, 1998). As a result of a consent decree, the MTA was mandated to improve bus service to low-income areas (Brown, 1998).

However, studies also reveal the many limitations of existing fixed-route public transit. During the job-search process, when low-income individuals must travel to many unknown destinations, transit may not be a viable option. In Los Angeles, transit users were twice as likely to state that their job-search trips were somewhat or very difficult compared to those traveling by car (UCLA Lewis Center for Regional Policy Studies 2000). Further, evidence on the relationship between access to public transit service and employment is weak. Sanchez (1999) finds that access to public transit is positively related to the labor participation rates (average annual weeks worked) of residents in Portland and Atlanta. However, with respect to welfare recipients, access to public transit does not appear to be significantly related to employment. Cervero et al. (forthcoming) find that measures of transit service quality are not significant predictors of employment among welfare recipients in Alameda County, California. They also find that increasing the percentage of welfare recipients with cars has stronger effects on employment than improving transit mobility (Cervero et al. forthcoming). In a working paper, Ong (2001) also finds no relationship between public transit access and employment for welfare recipients in Los Angeles.

Policy and Process Evaluations

Numerous studies show that transportation is one of the major program ingredients to helping welfare recipients make successful transitions from welfare into the labor market. The results of this research formed the basis for the passage of the Job Access and Reverse Commute Grant Program, a federal program to assist states and localities in developing new or expanded transportation services that connect welfare recipients and other low-income persons to jobs and employment-related services. This program combined with other federal resources has helped to establish and/or expand numerous transportation programs and services for welfare recipients. However, the overall effectiveness of many of these programs has not been established. While there are numerous reports describing what has become known as "best practices" in the field (American Public Transit Association 1998; Community Transportation Association of America 1999; Mathematica Policy Research Inc. 1998; Reichert, 1998), formal program evaluations of these services are few. Since many of these programs have only recently been established, it may be too early to evaluate their success.

There have been a few studies that extend beyond simply describing the services provided and use data to highlight the strengths and weaknesses of the program under review. One of the earliest programs, a program that preceded welfare reform, was the Bridges to Work demonstration project, a joint project of Public/Private Ventures (P/PV), a Philadelphia-based nonprofit research and program development organization, and the U.S. Department of Housing and Urban Development (HUD). The purpose of the project was to connect inner-city residents with suburban employment opportunities by providing job placement and transportation services to mitigate the problems created by travel to distant and unfamiliar employment locations. The demonstration project was located in nine sites and began the first of four years of operations in late 1996. As of March 1999, there were 1,960 participants, 982 of whom were eligible to receive services as part of the program. By this time, 599, or just over 61 percent, had been placed in jobs (Elliott et. al. 1999). However, it is difficult to assess the effectiveness of this particular intervention. While the research design included the random assignment of participants to experimental and control groups, at least thus far the published report does not compare results between the two groups.

A second high-profile transportation program is JOBLINKS, sixteen demonstration projects designed to test a variety of transportation strategies that help unemployed and underemployed people reach economic self-sufficiency. The project began in 1993 with funding from the Federal Transit Administration and has been administered by the Community Transportation Association of America. Goldenberg et al. (1998) evaluated the programs and concluded that the availability of transportation services enabled participants, many of them welfare recipients, to travel to jobs, interviews, and job training sites. Evaluations of these demonstration projects highlight four key factors in their success: improved and expanded service; community support; coordination among public agencies; and flexibility, responsiveness, and the serious consideration of non-traditional transportation services (Goldenberg et al. 1998; Applied Management and Planning Group 1999). Once again, the overall effectiveness of the demonstration projects is difficult to determine since a rigorous research design was not followed.

The U.S. General Accounting Office (GAO) is responsible for evaluating the success and implementation of the Department of Transportation's Access to Jobs program. The initial report (U.S. General Accounting Office 1999) analyzes the implementation of the Job Access and Reverse Commute Program after its first year and, therefore, does not use performance measures to evaluate the "success" of the transit services initiated under this federal funding program. The GAO found that coalition building among federal agencies, local organizations, and federal grant recipients is essential to the provision of useful transportation services (U.S. General Accounting Office 1999; U.S. Department of Transportation 1998). While the majority of applicants for federal assistance grants were traditional transportation service providers, the program encouraged the application of "consolidated" organizations--coalitions of local organizations and agencies, who utilized existing transportation services to round out their own services (U.S. General Accounting Office 1999). The U.S. Department of Transportation (DOT) further encouraged the development of locally-appropriate solutions, facilitated by the coming together of numerous area stakeholders (U.S. Department of Transportation 1998).

In nearly all of the programs, transportation service providers relied upon the existing infrastructure and directed the DOT grants towards expanding and improving services, such as increasing the hours and frequency of services provided (U.S. General Accounting Office 1999). Among the other services provided by grantees were additional lines, demand-responsive service, improved connections, van and shuttle service, limited carpooling and ridesharing, increased marketing information, and access to childcare and other social service agencies. In response to the GAO report, the DOT moved away from welcoming consolidated grant applications to actively encouraging transit providers and local organizations and agencies to work together to coordinate services (U.S. General Accounting Office 1999).

While some program evaluations focus on successful attempts to increase public transportation services, Waller and Hughes (1999) critique the institutional reliance on public transit systems as the only viable solution to low-income job access. They cite the institutional reliance on public transportation as a lazy willingness to provide service that is "good enough" and thereby perpetuates the disadvantaged position of the urban underclass (Waller and Hughes 1999). While Waller and Hughes (1999) fail to find truly successful paratransit programs, they have found successful attempts to facilitate private automobile ownership. They note that 43 states lifted automobile asset limitations and a number of local programs offer insurance assistance to ease the purchase or lease of vehicles and allow the use of TANF funds for automobile repair (Waller and Hughes 1999). Policy recommendations include the following: welfare eligibility based on income, not based on current or partial receipt of assistance; revised asset limitations to allow for each household worker to own a car; public-private partnerships; and the use of non-welfare-to-work funds for transit funding, so as to eliminate overlap between TANF and welfare-to-work funds.

Nearly all of the program evaluations stressed similar elements necessary to the provision of effective transportation services for welfare recipients. These elements include expanded and improved transit service; coordination among agencies; increased access to information; the inclusion of some paratransit alternatives, such as vanpools and ridesharing; and, to a lesser extent, vehicle purchase assistance (U.S. Department of Transportation 1998; U.S. General Accounting Office 1999; Waller and Hughes 1999).

Travel Behavior and Needs of Welfare Recipients: Focus Group Findings

In late summer of 2000, we organized a series of focus groups in Fresno County to initiate our examination of the transportation barriers facing welfare recipients. The interviews were arranged with the assistance of the Fresno County Department of Social Services, although no agency personnel were present during the interviews themselves. Welfare participants were invited to participate in focus groups following a job-search workshop or a welfare intake session. Welfare participants volunteered their participation and were offered Target gift certificates upon completion of the focus group.

A total of 81 individuals participated in the focus groups, which were held in offices of the Fresno County Department of Employment and Temporary Assistance (DETA) or in career centers located in the cities of Fresno, Reedley and Kerman. (See Figure 1 for the location of these cities.) We asked focus group participants to fill out a short written questionnaire and then to engage in informal discussions around a number of transportation-related topics, including mode of travel, commute time, and commute ease or difficulty. The focus groups were conducted in a discussion format, allowing participants to talk freely without the constraints imposed by a formal survey.

The focus group transcripts are included in Appendix Four. Overall the focus groups highlight the following issues:

The majority of focus group participants relied on public transportation as their means for routine travel including their search for employment.

The majority of focus group participants reported difficulties traveling to and from work-related destinations.

Focus group participants were concerned about the quality and quantity of public transit service particularly in rural areas.

Respondents from rural areas reported more transportation-related difficulties than did urban respondents.

The comments made by focus group participants were essential in developing the formal survey instrument used for the analysis presented in Chapter Four of this report.

Profile of Focus Group Participants

To establish the basic profile of focus group participants, we asked participants to fill out a short written survey. The survey included the date and location of the focus group, basic demographic characteristics (such as race/ethnicity, sex, and age) and a few transportation questions, including whether they owned a car, their relative access to cars, and their perceptions of their ease or difficulty of travel. Table 1 summarizes the information from this survey. The major distinction among the six focus groups lies in the nature of participants' involvement in the welfare program. The participants in two of the focus groups were new enrollees who had just completed an orientation to the requirements of the welfare program. The participants in the other four focus groups were involved in job-search programs intended to help them find employment.8

As indicated by the data presented in Table 6, most of the participants were involved in job-search workshops (65); the remaining 16 participants were contacted at welfare intake sessions. The welfare intake interviewees were female and slightly younger on average than those interviewed at the job search workshops. Overall, 71 percent of focus group participants was female with an average age of approximately 29.

Table 6: Focus Group Participant Characteristics

Table 7 presents data on the ethnicity of the focus group participants. Approximately 50 percent were white, and the remainder were Latino, African-American, of mixed racial or ethnic backgrounds, or from other racial backgrounds. One notable omission from the focus groups was Asians--particularly Hmong, ethnic Cambodians who comprise a sizeable percentage of the Fresno County caseload. The underrepresentation of Hmong welfare recipients in the focus groups may be due to a number of factors. It may be due to a self-selection bias, an underlying reason why Hmong welfare recipients might be less likely to volunteer their participation. The bias could occur if Hmong welfare recipients are less likely than other racial or ethnic groups to be involved in the initial phases of the welfare-to-work program. Or it may be due to the particular geographic location of the focus groups, perhaps outside of the neighborhoods in which Hmong welfare recipients travel.

Table 7: Ethnicity of Focus Group Participants