An exploratory study of neighborhood choices among moving to opportunity participants in Baltimore, Maryland: the influence of housing search assistance.

This study examined the neighborhood choices of 150 families who participated in the Moving To Opportunity Program (MTO) in Baltimore, Maryland. The MTO program, utilizing an experimental design, provided intensive housing search and counseling services to the experimental subjects. This study found that the counseling services were instrumental in altering the subject's cognitive maps, and they were more likely to move to neighborhoods that were more racially integrated, safer, and, also, had higher levels of satisfaction with their new neighborhood. The authors conclude that the MTO program in Baltimore represents a clear case of public policy that, at least in the short term, worked.

Keywords: Moving To Opportunity, housing policy, public housing, low income housing, mobility

**********

Pendall (2000) has noted that since the 1970's the dominant model for U.S. federal housing policy has shifted from unit-based programs to tenant-based vouchers and certificates. The theory behind this shift is that vouchers and certificates should allow those who receive this assistance to live in better neighborhoods. Theoretically, these neighborhoods would provide access to better schools and employment opportunities, and less exposure to crime and violence, as well as other social benefits. By the early 1990's these mixed-income and dispersal strategies predominated federal housing policy (Popkin, et al., 2000).

This dispersal strategy was influenced, in large part, by the Gatreaux Program. In the late 1960's, a group of fair housing advocates filed a class action lawsuit on behalf of Chicago public housing residents against the Chicago Housing authority and HUD, charging that these agencies had employed racially discriminatory policies in the administration of the Chicago low-rent housing program. Ten years later, a Supreme Court decree ordered the formulation of a racial dispersal strategy, including the placement of 7,100 Black public housing residents or applicants in racially desegregated neighborhoods throughout the Chicago metropolitan area. The Gautreaux Program was intended primarily as a desegregation remedy. However, research by James Rosenbaum (1996) and others at Northwestern University on the families that moved through the Gautreaux Program has suggested that a move out of the central city can have positive employment, earning, and education effects.

The Gautreaux studies show that moving to the suburbs had significant positive effects on the educational attainment of the children. Not only were they less likely to drop out of school, but they were also more likely to take college-track courses, compared to those who moved within the city. After graduating from high-school, children of suburban movers were also likely to attend a four-year college or become employed full-time at a job with fringe benefits. Popkin, et al., (2000) notes that "thirty years after the initial decision, the philosophy behind Gatreaux, that public and assisted housing should be scattered throughout a range of communities or deconcentrated, has become the driving force behind the current transformation" (p. 912).

Pendall (2000) observes that these voucher and certificate programs, however, do not always "live up to their promise as mechanisms that foster mobility" (p. 882). He notes that tenants, in particular blacks and Hispanics still, often, resettle into poor, segregated neighborhoods.

South and Crowder (1997) came to a similar conclusion when they examined the mobility experiences of poor blacks and whites. They reported that blacks who moved out of poor neighborhoods were more likely than whites to move into another poor neighborhood (13.6% of black had this experience compared to 5.2% of whites). In fact, 11% of blacks moved from nonpoor neighborhoods into poor neighborhoods as compared to only 1.4% of whites. More recently, Rosenbaum and Harris (2001) cite a number of studies which conclude that among assisted households, blacks are more likely than whites to relocate to areas with higher concentrations of poverty and black residents.

A considerable body of scholarship exists concerning population mobility and residential choice, and a variety of theories have been used to explain these behaviors. In an historical overview of why people move, Shumaker and Stokols (1982) traced theories of mobility. One of the earliest theories, referred to as the "Gravity Law of Mobility", argued that people moved because they were drawn to other people. Later theorists developed models that explained that people were not necessarily drawn to other people, but instead were drawn to opportunities available within a new locale and were influenced by perceived obstacles to moving and perceived benefits within the current locale. Additionally, recent theorists assume that a rational, cost-benefit analysis underlies the relocation decision making process (Shumaker and Stokols, 1982).

Other researchers have argued that mobility is a response to stress. These "stress theorists" suggest that an environment which fails to provide the resources essential to meeting one's needs produces a lack of fit between needs and environment. When the stress level reaches a critical threshold, the person seeks to relocate. Speare (1974) emphasizes a "threshold of dissatisfaction" rather than a stress response. Once individuals reach their threshold, they employ a cost-benefit analysis that includes an evaluation of their current locale compared to the cost and benefits of changing their locale.

Shumaker and Stokols (1982) note that there are some significant weaknesses in all of these theories. The first major weakness is that they assume that mobility is an alternative for all Americans. Yet, as noted earlier, the data show that mobility is not readily available to certain major subgroups in American society, particularly African-Americans. Shumaker and Stokols argue that across income groups African-Americans are restricted in their residential options.

Researchers (Clark, 1992; South and Deane, 1993; South and Crowder, 1997) have offered various explanations for the limited mobility choices of African-Americans. Prominent among them at the political/structural level are inequities such as segregation and discrimination experienced by blacks in our society. These political/structural inequities include: local zoning practices and land use regulations (Rossi and Shlay, 1982) which were designed to regulate socioeconomic spatial arrangements and indirectly influenced racially segregated spatial arrangements; the gerrymandering of school boundaries which also helped to establish segregated spatial environments; and decisions regarding highway and freeway construction that kept neighborhoods racially separate (Fairchild and Tucker, 1982).

At the individual level possible explanations may include the socio-cultural influences among poor blacks which include: having more extensive social ties in poorer neighborhoods; a greater familiarity with them; and a preference for racial homogeneity. Another way to understand these socio-cultural influences is through the concept of "cognitive maps."

Golledge (1999) defines cognitive maps as "the internal representation of perceived environmental features or objects and the spatial relations among them" (p. 5). Downs and Stea (1973) expand on this definition by stating:

   Cognitive mapping is a process composed of a series of
   psychological transformations by which and individual acquires,
   codes, stores, recalls, and decodes information about the relative
   locations and attributes of phenomena in his everyday spatial
   environment (p. 9).

In a later work Downs and Stea (1977) provide a definition that fits most closely the purposes of this paper. They state that cognitive mapping is an activity that we engage in rather than an object that we possess. Our cognitive maps represent a cross section of the world, a community, or a neighborhood at one instant in time. It reflects the world as some person believes it to be. It need not be correct. In fact, distortions are highly likely. Whether distorted or not cognitive mapping is a basic component of human adaptation necessary for human survival and everyday environmental behavior.

Our information about the world comes from both direct and vicarious sources (Downs and Stea, 1973). Direct sources involve face-to-face contact between the individual and, for example, a neighborhood; and information literally floods the person from all of this sensory modes. Vicarious information is by definition secondhand. It is literally and metaphorically seen through someone else's eyes. In either case, the information is selected and transmitted through a set of filters that necessarily distort the information, generally in a way useful to the individual in his present context.

We cannot absorb and retain the virtually infinite amount of information that impinges upon us on a daily basis. Instead, we develop perceptual filters that screen out most information in a highly selective fashion. Our views of the world, and about people and places in it, are formed from a highly filtered set of impressions, and our images are strongly affected by the information we receive through our filters (Gould and White, 1986). These filters, which are at the core of our cognitive maps, are the basis which help us decide upon and implement any strategy of spatial behavior such as neighborhood choice.

The poor often have little or no direct experience with non-poor neighborhoods and the private housing market, and have little contact with people who can give them accurate information about them. Therefore, their mobility decisions are made through distorted filters, which limits their choices, and renders their search process ineffective. Thus, through a combination of political/structural factors and socio-cultural influences, the poor, when they do make mobility decisions, often find themselves in impoverished neighborhoods much like the ones they left.

Hartung and Henig (1997) and Turner, Popkin, and Feins, (2003) among others have suggested that in addition to vouchers, residents of public housing may need "considerable support to find and keep housing in the private market" (p. 29). This support would address both the political/structural factors and socio-cultural influences affecting these residents. The MTO program attempted to provide this support through its counseling and housing search assistance.

The Moving To Opportunity (MTO) Program

MTO is a demonstration program and research study, utilizing an experimental design, authorized by the Housing and Community Development Act of 1992, which combines Section 8 rental assistance with intensive housing search and counseling services. The demonstration is testing whether, after finding private housing in low-poverty communities, MTO treatment group families will become increasingly self-sufficient, compared to those who did not make such moves and to others who made similar moves without counseling and support.

MTO provides Section 8 rental assistance to roughly 1,600 families, to learn whether the differences in neighborhood conditions affect the social and economic future of parents and children. Congress restricted the demonstration to very large cities with populations of at least 400,000, in metropolitan areas of at least 1.5 million people. From sixteen cities submitting applications, five cities were selected in March 1994 for MTO: Baltimore, Boston, Chicago, Los Angeles, and New York City.

The five local MTO programs were created via grants from the Secretary of HUD to nonprofit organizations (NPOs) to provide counseling and services in connection with the demonstration. HUD also entered into contracts with the public housing agencies to administer the Section 8 rental assistance to members of the MTO experimental group. The NPOs received funding to help pay for the costs associated with counseling the experimental group families, assisting them in finding appropriate units, and working with landlords to encourage their cooperation with Section 8 and the MTO program.

In Baltimore, the counseling and housing search assistance consisted of many components, and was designed to a) address the political/structural factors and socio-cultural influences that could potentially impair the mobility decision-making process of the MTO participants. These components included:

* the recruitment of owners and managers of property in low-poverty census tracts;

* work with landlords to obtain family tenancy history and letters of reference;

* visit MTO experimental group families in their homes and assess their strengths and weaknesses in terms of their preparedness to move and conduct credit checks;

* discuss the goals the family wants to achieve;

* provide budget and employment counseling to the families;

* provide referrals for the families to appropriate resources regarding issues such as: substance abuse problems, day care options, and parenting skills;

* provide transportation for the families to low poverty areas and inspection of potential rental units;

* coordinate discussions between landlords/property owners and participants;

* followup with individual participants and groups of participants located in the same area;

* and conduct semi-annual inspections of the rental units.

The research component of MTO utilized an experimental design that randomly assigned MTO families into three groups. From 1994 to 1998 4,608 families in the five sites volunteered for MTO and were randomly assigned. The experimental group received Section 8 certificates or vouchers that they could use for housing in low-poverty census tracts (under 10 percent poverty in 1990). They also received counseling and housing search assistance. The comparison group received Section 8 certificates or vouchers that could be used to move anywhere. They did not receive counseling or housing search assistance. A control group received no Section 8 assistance, but continued to receive assistance in the public housing or assisted housing development where they lived. This group provided a benchmark against which outcomes for the other two groups would be measured (Turner, Popkin, & Feins, 2003).

Methodology

The questions that we asked in this study concerned how successful the experimental and comparison group participants were in obtaining housing, and whether the counseling affected their locational decisions. Two data sources were utilized for this study. The first source was the Department of Housing and Urban Development (HUD) participant baseline surveys which were administered at intake to the MTO program. We used the baseline data only to examine how families felt about the neighborhoods they were moving from. The second data source consisted of structured interviews with the participants conducted by the authors of this paper after the families had moved. Interviews were not conducted with members of the control group as we were only interested in comparing participants who had moved.

There were 339 MTO participants in the city of Baltimore. The experimental group included 139 (41%), the comparison group 93 (27%), and the control group had 107 (32%) participants. The authors and trained interviewers conducted interviews with 150 of 232 participants (87 from the experimental group and 63 from the comparison group), for a completion rate of 65%. The primary reasons the other participants were not interviewed were failure to show up for the designated interview times (two appointments were scheduled), and moved and left no address.

The city of Baltimore offers an excellent locale to explore the latter question concerning socio-cultural influences. Baltimore is known as a "City of Neighborhoods." Neighborhoods are a major tradition, foundation and resource for the city's civic culture (Henderson, 1993). More than 700 neighborhood associations are registered with the Baltimore City Department of Planning. These neighborhoods are mostly divided between East and West Baltimore, which are the primary geographic demarcations for the city. There are many similarities between them. For instance, the population for both is 91% African-American. The percentage of families in poverty is 49% on the East side and 47% on the West, and the percentage of dwellings occupied by renters is 81% on the East and 82% on the West. While, the overall vacancy rate for the city of Baltimore is 9% the rates for the East and West are 16.6 and 17.9, respectively. However, while similar in many ways East and West is not only geographic divide, in many ways they are also a social and psychological divides. It is very common for residents of both sides to remain "on their side" for their entire lives. This pattern is even more pronounced among poor African-American residents.

Results

In the baseline survey approximately 95% of all participants indicated that they wanted to live in a different neighborhood in Baltimore city, the suburbs, or a different city. One indicator of the effectiveness of the counseling in expanding the cognitive maps of the experimental families would be the extent of movement among experimental and comparison families from one side of town to the other. As Table I indicates approximately 50% of experimental families moved to the other side of town, as compared to approximately 30% of the comparison families (p = <.01, and a Cramer's V of .37.

Another indicator of the counseling's impact on socio-cultural influences would be how important it was for both groups to remain close to their old neighborhoods, friends and family members. As Table 2 indicates, there was a statistically significant difference in the responses between the experimental and comparison groups on the question of moving too far from their old neighborhoods (p = <.04, and a Cramer's V of .18). There was also a statistically significant difference, (see Table 3), between the groups in regards to the importance of not being too far away from family members (p = <.03, and a Cramer's V of .19).

The next question examined how successful the experimental and comparison group families were in attaining housing with which they were satisfied. Table 4 indicates that experimental group families were significantly more satisfied with their new residences (p = <.03, Cramer's V of 22).

They also were able to move to more racially integrated neighborhoods than the comparison group participants, as well as to neighborhoods they considered safer. Nearly three and one-half times as many comparison group members (47.6%) moved to mostly African-American neighborhoods than experimental group families (13.8%). About half as many more experimental group members moved to neighborhoods that were a mix of African-American and white than comparison group members. Similarly, more members of the experimental group (32.2%) moved to neighborhoods that were a mix of either Hispanic and white or African-American, Hispanic and white than had members of the comparison group (15.9%). Very few of the respondents moved to mostly white neighborhoods (4.6% of the experimental group and 1.6% of the comparison group). See Table 5.

Experimental group members were also better able to move to neighborhoods where they felt a higher degree of personal safety. Over twice as many members of the comparison group (60.8%) than members of the experimental (27.8%) indicated that they felt their new neighborhoods had problems with drugs and violence. This difference was statistically significant (p = <.00, with a Cramer's V of .33). For those respondents who indicated that their new neighborhoods had a problem with drugs and violence, we asked how serious they considered this problem. Here again the difference was statistically significant (p = <.05, and a Cramer's V of .27). Predictably, more members of the comparison group (86.7%) than of the experimental group (63.2%) felt that the problem was serious.

Discussion

Housing policy, in many respects, is much more than simply housing policy. It is also education policy, health policy, work force policy, criminal justice policy, even environmental policy. Where we live determines almost everything about how we live (Hill, 2004; Marriott, 2004). All of the participants in the MTO program, by their participation in a voluntary program, demonstrated a desire to move to better neighborhoods. Their responses to survey questions was further evidence of this desire. The main reasons they gave for wanting to move were to escape drugs and violence, find a better house or apartment, and to find better schools for their children.

Although families in both the experimental and comparison groups were equally motivated to move, families in the experimental group were more successful than those in the comparison group. Our findings strongly suggests that without counseling and other assistance families employed a housing search strategy we have labeled, "Go where you know," which is influenced, to a large extent by an individual's cognitive map. The comparison group families were more likely to employ this strategy, while the experimental families moved to better (e.g. low poverty) neighborhoods that were further away from friends and family, and prior residences, and to neighborhoods that were outside of their cognitive maps. This is not to imply that there is anything inherently wrong with living in a poor or all-black neighborhoods. However, decades of overt and covert support for segregation in housing policy has left many of these neighborhoods devastated, and it is these neighborhoods which public housing residents are most familiar.

The lack of economic resources and opportunity are largely responsible for poor families being unable to attain a satisfactory level of residential mobility. The MTO program provided economic resources and opportunity to both the experimental and comparison group families to move into better neighborhoods. Without counseling support and the requirement to move to low poverty neighborhoods, however, comparison group families moved into primarily high poverty neighborhoods, where they felt drugs and violence were serious problems, and expressed more dissatisfaction

with these new neighborhoods when compared to the experimental group families.

As Varady and Walker (2003) have noted, providing housing search and counseling assistance is expensive, and may not be needed by all users of Section 8. However, the Baltimore experience demonstrates that market forces and opportunity were not enough to substantially change the residential circumstances of poor families. This study shows that these alone only tend to steer poor residents into areas where other poor residents already live. In conclusion, this study tells the story of very poor families, living in conditions that most Americans would agree are intolerable, who were given the opportunity under an innovative federal program to move to better neighborhoods. One group of movers was, on the face of things, more successful in their moves than the other. As measured by such criteria as the extent to which their new neighborhoods were free from crime and drugs, the members of the experimental group were more successful. It must be noted that because MTO participants volunteered for the program self-selection bias cannot be ruled out, and these participants may differ in unknown ways from the larger population (Popkin, et al., 2000). However, MTO in Baltimore represents a clear case of public policy that, at least over the short term, worked. Further analysis of the MTO participants over the next several years will be needed to provide answers about the long-term impact of the program.

References

Clark, W. A. V. (1992). Residential preferences and residential choice in multiethnic context. Demography, 29:451-466.

Downs, R. M. & Stea, D. (1973). Image and environment: Cognitive mapping and spatial behavior. Chicago: Aldine Publishing Company.

Fairchild, H. H. & Tucker, M. B. (1982). "Black residential mobility: trends and characteristics." Journal of Social Issues, 38:51-74.

Golledge, R. G. (1999). Wayfinding behavior. Baltimore: Johns Hopkins Press.

Gould, P. & White, R. (1986). Mental maps. Boston: Allen & Unwin.

Hartung, J. M. & Henig, J. R. (1997). "Housing vouchers and certificates as a vehicle for deconcentrating the poor: evidence from the Washington, D.C. metropolitan area." Urban Affairs Review, 32:403-419

Henderson, L. J. (1993). "Baltimore: managing the civics of a "turnaround" community." National Civic Review, 82:329-339.

Hill, M. (2004, January 18). Still separate. The Baltimore Sun, p. 1, 6C.

Marriott, S. S. (2004). Fundamentally flawed. The Baltimore Sun, p. 1, 13A.

Pendall, R. (2000). Why voucher and certificates users live in distressed neighborhoods. Housing Policy Debate, 11:881-910.

Popkin, S. J. Buron, L. F., Levy, D. K. & Cunningham, M. K. (2000). The Gautreaux legacy: What mixed-income and dispersal strategies mean for the poorest public housing tenants? Housing Policy Debate, 11:911-942.

Rosenbaum, E. & Harris, L. E. (2001). Residential mobility and opportunities: Early impacts of the Moving To Opportunity demonstration program in Chicago. Housing Policy Debate, 12:321-346.

Rosenbaum, J. E. Stroth, L. K., & Flynn, C. (1996). Lake Parc Place: A study of the first four years of a mixed-income housing program. Working Paper WP-96-7. Evanston, IL: Northwestern University, Center for Urban Affairs and Policy Research.

Rossi, P. H. & Shlay, A. B. (1982). "Residential mobility and public policy issues: why families move revisited." Journal of Social Issues, 38:21-34.

Shumaker, S. A. & Stokols, D. (1982). Residential mobility as a social issue and research topic. Journal of Social Issues, 38:1-19.

South, S. J. & Crowder, K. D. (1997). Escaping distressed neighborhoods: individual, community, and metropolitan influences. American Journal of Sociology, 102:1040-1084.

South, S. J. & Deane, G. D. (1993). Race and residential mobility: individual determinants and structural constraints. Social Forces, 72:147-167.

Speare, A. (1974). Residential satisfaction as an intervening variable in residential mobility. Demography, 11:173-188.

Turner, M. A., Popkin, S., Feins, R. (2003). Environmental impact. Journal of Housing and Community Development, 60:25-29

Varady, D. P. & Walker, C. W. (2003). Using housing vouchers to move to the suburbs: The Alameda County, California experience. Urban Affairs Review, 39:143-180.

JAMES X. BEMBRY

University of Maryland Baltimore County

Department of Social Work

DONALD F. NORRIS

University of Maryland Baltimore County

Department of Public Policy

Table 1
Area of Origin/Destination

                    East         West         Total

Area of Origin    #     %      #     %      #      %

Experimental
East             22    51.2   21    48.8   43     100
West             18    46.2   21    53.8   39     100

Comparison
East             25    69.4   11    30.6   36     100
West              7    31.8   15    68.2   22     100

p = <.01; Cramer's V = .37.

Table 2
How important was it that the neighborhood or area the apartment (or
house was located in not be too far away from your old apartment (or
house) and neighborhood?

                Experimental Comparison   Total

                No.   %      No.   %      No.   %

Important       23    26.4   27   42.9    50    33.3
Not Important   64    73.6   36   57.1   100    66.7

Total           87    58.0   63   42.0   150   100.0

p = <.04; Cramer's V = .18.

Table 3
How important was it that the apartment (or house) not be too far away
from your family?

                Experimental Comparison   Total

                No.    %     No.    %     No.     %

Important       48    55.2   45    73.8   93    62.8
Not Important   39    44.8   16    26.2   55    37.2

Total           87    58.8   61    41.2   148   100.0

p = <.03; Cramer's V =19.

Table 4
How satisfied are you with your new apartment or house?

                Experimental Comparison   Total

                No.    %     No.    %     No.     %

Satisfied       78    89.7   48    76.2   126    84.0
In the Middle    6     6.9    5     7.9    11     7.3
Dissatisfied     3     3.4   10    15.9    13     8.7
Total           87    58.0   63    42.0   150   100.0

p = <.03; Cramer's V = .22.

Table 5
Describe the racial makeup of your neighborhood

                              Experimental    Comparison   Total

                              No.    %        No.    %     No.     %

Mostly African-American       12    13.8      30    47.6   42     28.0
Mix of Black/White            43    49.4      22    34.9   65     43.3
Mix of Black/White/Hispanic   28    32.2      10    15.9   38     25.3
Mostly White                   4     4.6       1     1.6     5      3.3

Total                         87    58.0      63    42.0   150   100.0

p = <.00; Cramer's V =38.

Table 6
In your opinion, does the neighborhood you moved into (the one you
live in now) have a problem with drugs and violence?

          Experimental  Comparison    Total

          No.    %      No.    %      No.     %

Yes       20    27.8    31    60.8    51     41.5
No        52    72.2    20    39.2    72     58.5

Total     72    58.5    51    41.5    123   100.0

p = <.00; Cramer's V =33.

Table 7
In your opinion, how serious is the problem with drugs and violence in
your new neighborhood?

                  Experimental Comparison    Total

                  No.    %     No.     %     No.     %

Serious Problem   12    63.2    26    86.7   38     77.6
Not a Problem      7    36.8    4     13.3   11     22.4
Total             19    38.8    30    61.2   49    100.0

p = <.05; Cramer's V =.27.