low_income_workshop-newman

*Prepared for the Panel on Data and Methods for Measuring the Effects of Changes in Social Welfare Programs Conference, Committee on National Statistics, CBASSE, National Research Council, December 16-17, Washington D.C. This research was supported by generous grants from the Foundation for Child Development, the Ford Foundation, the National Science Foundation, the Russell Sage Foundation and the MacArthur Network on Socio-Economic Status and Health.

Statistical trends are necessary but not sufficient. To me, statistical trends alone are like a canary in a coal mine - they yield life or death information on the "health" of an environment, but don't always lead to improvement, causes and corrective actions.

In the years to come, researchers and policy makers concerned with the consequences of welfare reform will dwell on studies drawn from administrative records that track the movement of TANF recipients from public assistance into the labor market and, perhaps, back again. Survey researchers with panel studies will be equally in demand as federal, state, and local officials charged with the responsibility of administering what is left of the welfare system come to grips with the dynamics of their case loads. This is exactly as it should be, for the "poor support" of the future--whatever its shape may be--can only be fashioned if we can capture the big picture that emerges from the quantitative study of post-AFDC dynamics that has seen many of the nation’s poor women move from welfare to work.

Yet as the early returns tell us, the story that emerges from these large scale studies contain many puzzles. The rolls are dropping precipitously nation-wide, but not everywhere (Katz and Carnavale 1998). TANF recipients are often able to find jobs, but many have trouble keeping them and find themselves back on the rolls in a pattern not unfamiliar to students of the old welfare system. Millions of poor Americans have disappeared from the system altogether: they are not on TANF, but they are not employed. Where in the world are these people? Welfare reform has pushed many women into the low wage labor market, but we are not entirely sure how this trend has impacted their standard of living or the well-being of their children. Are they better off in terms of material hardship than they were before? Are the benefits of immersion in the world of work for parents--ranging from the psychological satisfaction of joining the American mainstream to the mobility consequences of getting a foot in the door-- translating into positive trajectories for their children? Or are kids paying the price for the lift their mothers have experienced because they have been left behind in substandard childcare? And can their mothers stick with the work world if they are worried about what is happening to their kids?

There is reason to believe that these kinds of questions will not be resolved through reliance on administrative records. States and localities rarely collect data on mothers’ psychological well-being. They will not be able to determine what has become of those poor people who have not been able to enroll in the system. They have little sense of how households, as opposed to individuals, reach collective decisions which deputize some members to head into the labor market, others to stay home to watch the kids, and yet others to remain in school. Problems like domestic abuse, low levels of enrollment in children’s health insurance programs, and the cannot easily be understood via panel studies that ask respondents to rate their lives on a scale of 1 to 10. And though one might argue that welfare reform was oriented toward "work first" and was not an anti-poverty program per se, understanding the nature of material hardship is an important goal for any public official who wants to get to the bottom of the poverty problem. Trawling along the bottom of the wage structure, we are likely to learn a thing or two about recidivism as the burdens of raising children collide with the limitations of the low wage labor market for addressing the needs of poor families.

If administrative records and panel studies cannot tell us everything we might want to know about the impact of welfare reform, what are the supplementary alternatives? I argue in this chapter that qualitative research is an essential part of the tool kit and that, particularly when embedded in a rigorous survey design, it can illuminate some of the unintended consequences and paradoxes of this historic about-face in American social policy. From this vantage point, I argue that the "right soft stuff" can go a long way toward helping us to

Understand subjective responses, belief systems, expectations and the relationship between these aspects of world view and labor market behavior;
Uncover underlying factors that drive response patterns that are overlooked or cannot easily be measured fixed choice questionnaires
Explore in greater detail the unintended consequences of policy change and
Focus special attention on the dynamics shaping the behavior of households or communities that can only be approximated in most survey or administrative record studies that draw their data from individuals. This will be particularly significant in those domains where the interests of some individuals may come into conflict with others and hard choices have to be made.

The intrinsic value of qualitative research lies in its capacity to dig deeper than any survey can go, to excavate the human terrain that lurks behind the numbers. Used properly, qualitative research can pry open that black box and tell us what lies within. And at the end of the day, when the public and the politicians want to know whether this regime change has been successful, the capacity to illuminate its real consequences -- good and bad -- with stories that are more than anecdotes, but stand as representatives of patterns we know to be statistically significant, is a powerful means of communicating what statistics can only suggest.

A wide variety of methodologies come under the broad heading of qualitative methods, each with its own virtues and liabilities. In this section, I discuss some of the best known approaches and sketch out both what can be learned from each and where the limitations typically lie. Where possible, I draw on ongoing research to illustrate the possibilities of these methods.

Obviously the great value of survey research lies in its large sample size, its representativeness, and the capacity it provides for statistical analysis and causal inference. Typically, the items on survey research instruments are close-ended questions based on fixed-choice response categories or questions that require respondents to rate their reactions on set scales. However, it is not uncommon for survey studies to include a limited number of items that are "open ended," where respondents either write short responses in their own words with no guidance from the researcher or speak their minds into tape recorders that generate brief transcripts. Open ended questions embedded in survey instruments typically follow more cut-and-dried queries (Were you "very happy, moderately happy, moderately unhappy or very unhappy with the quality of your child’s care last week?) with "why?" questions designed to learn a bit more about the reasoning behind a respondent’s answer. (What kinds of problems did you encounter with your childcare last week?). The value of the follow up question lies in the difficulty researchers may have in anticipating all the relevant fixed-choice categories. Where this is particularly vexing, open ended questions can help to illuminate complex patterns while preserving the strength in numbers that survey research provides.

At least two purposes can be served here. First, open ended responses are typically coded and then analyzed in much the same way that fixed choice questions are, but now with categories that have essentially been generated by the survey respondents rather than those forced upon them by the researcher. This is generally more reflective of the experience of interviewees as they see it and if the subjective understandings of respondents are the issue, this is an appropriate method for capturing them on a large scale. Secondly, open-ended responses (particularly in pilot studies) can be used to generate more nuanced fixed choice questions for future surveys. Clearly, the advantage to embedding qualitative research inside a survey design is that one benefits from the representativeness and sample size, while preserving the insights afforded by qualitative data.

There are obvious limitations to this approach. Because of the expense involved in coding the material, open-ended questions cannot be attached to every fixed choice series in a survey, at least not for studies that involve thousands of respondents. When one is dealing with respondents who have literacy problems or who are not comfortable writing, the barriers to data collection can be prohibitive. Of course, among respondents with literacy issues, using mail questionnaires is problematic to begin with. Questionnaires administered face-to-face or over the telephone can still utilize open ended items by having the interviewer record the responses or by using tape recorders. Problems of accuracy can be avoided through careful training.

When one wants to collect more open-ended data from each subject, it may be appropriate to draw a smaller random sub-sample of a survey population for longer interviews designed to elicit information on a wide range of topics. Assuming that the appropriate demographic categories can be identified (including race, age, family status, children of particular ages, etc.), a representative sub-sample can be interviewed in situ or in a central location. Studies of this kind can explore in some detail the experience "informants" are having job seeking, adjusting to employment, managing children’s needs, coping with new expenses, finding transportation to work, relying on neighbors, and a host of other exploratory questions that may shed light on the post-TANF experience. As long as the sub-sample is representative, the researcher can extrapolate from it to the experience of the universe in the same way one would generalize from any representative group.

The advantage of this approach is that it permits greater depth on a larger number of subjects, yielding a more rounded perspective than is possible with only one or two open-ended questions. Such a methodology is appropriate when the study aims to understand the intricacies of subjective perspectives or the inter-twined nature of family behavior when policy change impacts directly on one household member, but indirectly on the rest. Problems of this complexity can only be understood with a great deal of qualitative information which is expensive to collect for an entire survey population, but which can be generated for a sub-sample.

The value of an in-depth interview over a fixed choice survey or even a survey with a few embedded open-ended questions is considerable. It makes possible the emergence of patterns and practices that may otherwise be obscured by a survey instrument. The longitudinal study of the Milwaukee New Hope experiment is a good case in point. New Hope provided generous child care and earnings supplements in order to guarantee families in the "treatment" group who worked at least 30 hours a week, sufficient income to bring them above the poverty line and child care and insurance supports that would make it easier to remain in the labor force. Under the direction of Greg Duncan at Northwestern University, the New Hope research team developed both a longitudinal panel survey and an embedded ethnographic study 1 which drew mainly on repeated interviews with a representative sample of participants and controls. From Duncan’s perspective, the blending of "hard" and "soft" data has been critical in understanding program impacts:

New Hope's qualitative data proved indispensable for understanding the nature and meaning of program impacts. As simple as an experimental design may seem, analyses of experimental impacts are complicated by needs to quantify the proper outcomes and isolate program impacts within important sample subgroups. Qualitative data are very helpful in both of these tasks. One of the most important - and initially puzzling - impacts of the New Hope experiment was on teacher-reported achievement and behavior of pre-adolescent children. Boys but not girls in the experimental group were .3 to .5 standard deviations better behaved and higher achieving than their control-group counterparts. Based on the survey data alone, however, we were unable to account for this gender difference. Qualitative interviews suggested that mothers felt that gangs and other neighborhood pressures were much more threatening to their boys than girls, and that experimental group mothers channeled more of the program's resources (e.g. childcare subsidies for extended-day programs) to their boys than girls. Further quantitative analyses of both New Hope and national-sample survey data support this interpretation (Romich, 1999). Further inquiry is needed to compare the strength of this hypothesis to competing explanations, but it is unlikely that this important finding about family strategies in dangerous neighborhoods would have been discovered from the quantitative data alone. (Greg Duncan, personal communication 11/29/99).

The New Hope project has also been able to provide useful analyses that separate out the experience of subgroups of participants who have responded differently to the same program opportunities. Since New Hope mirrors what some of the more generous states have tried to accomplish in their welfare-to-work programs, its experience is useful in parsing the differential impact of these supports for working families. As Duncan suggests below in his comments on labor supply and earnings, without the qualitative component, it would have been harder to identify and then "unpack" the behavior of these subgroups.

It was clear from the beginning of our quantitative work that program effects on work and earnings were heterogeneous. Roughly one-third of the families attracted to New Hope were already working more than 30 hours and viewed the program's benefits as a way of making work and family demands more manageable. If anything, experimental/control differences in the labor supply of these families were negative. In contrast, families not working full-time at baseline viewed New Hope as a way of facilitating a transition to full-time work. On balance, experimental/control impacts on labor supply were positive for these families, although stronger in the first than second year of the program.

Qualitative interviews pointed to important heterogeneity among this latter set of families. Some, perhaps one-fifth, had multiple problems (e.g., drug dependence, children with severe behavior problems, relatives in ill health) that New Hope's package of benefits could not be expected to overcome. Others had no such apparent problems, and in these cases both experimental and control families could be expected to do well in Milwaukee's job-rich environment.

But a third group, who were only one or two barriers away from making it, might well profit the most from the New Hope package of benefits. Extensive quantitative work on barrier-defined subgroups showed this to be the case (Magnuson, 1999). Program impacts on the labor supply of families with a small number of barriers were large and, if anything, larger in the second than the first year. This key set of findings would simply not have been discovered were it not for the qualitative work. (ibid)

A popular technique for exploratory research involves the use of focus groups, small gatherings of individuals selected for their demographic characteristics who engage in collective discussion following questions or prompts issued by a researcher acting as a facilitator. Focus groups operate in the native language of the participants and can last as long as two hours, providing an in depth discussion of a topic. The appeal of focus groups usually lies in the modest expense involved: this is a "quick and dirty" method of gathering data on the subjective responses of program participants. A wide range of interested parties--from politicians to business firms--utilize focus groups as a means of "testing the market," particularly where public opinion is at issue.

Of course, there are limitations inherent in the focus group approach. The contamination of opinion that occurs when individuals are exposed to the views of others can render the data hard to interpret. When particularly forceful individuals dominate the discussion, the views of more passive participants can easily be squelched or brought into conformity in ways that distort their true reactions. People are understandably hesitant to air their opinions on sensitive subjects (e.g. domestic violence, employer misbehavior, criminal behavior) in settings such as these. Moreover, it is hard to make focus groups representative of a population in any meaningful sense. They must therefore be used with caution. Focus groups are not a good tool for producing data that will withstand strict scrutiny for representativeness. What they do provide is an inexpensive and rapid means of learning about underlying attitudes and reactions, an approach that may be informative for officials or scholars looking to design more nuanced research instruments.

Some of the limitations of focus groups can be addressed to a modest degree through the careful selection of focus group members. Sensitive subjects may best be addressed by drawing together people who are as similar as possible, who have experienced a common dilemma in the hopes that the similarities between them will lessen the discomfort. Hence investigators often construct focus groups along the lines of racial or ethnic groups, gender or age groups, or residents of a particular neighborhood. The "contamination" of forceful individuals can be limited by the guiding hand of an equally forceful facilitator who makes sure that others have a chance to participate. However, none of these approaches fully eliminate the difficulties inherent in public discussions of this kind.

Focus groups are therefore probably best used to gather data on community experience with and opinions toward public assistance programs rather than systematic data on individual perspectives. For example, the problems associated with enrollment in children’s health insurance systems could probably be well understood by convening focus groups. Indeed, one of the strengths of the method is that it prompts individuals who may not be able to express themselves easily in a one-on-one setting to recall and describe difficulties they have encountered. Information of this kind is far more textured and complete than fixed choice questionnaires and can help public officials address the deficiencies in outreach programs, for example.

Welfare reform is a process unfolding over a number of years, where the before and the after may be widely separated and the "in between" states of at least as much interest as the ultimate outcomes. We have good reason to believe that families pass through stages of adaptation as their children age, new members arrive, people marry, jobs are won and lost, and the hold of new requirements (work hours, mandated job searches), exert their influence. For this reason, it will be critical that at least some of the nation’s implementation research follow individuals and families over a period of years, rather than rest easy with cross-sectional studies. Indeed, one need only look at how the Panel Study of Income Dynamics has altered and enhanced our understanding of income over the lifespan or movements in and out of poverty over time to recognize the value of panel studies of this kind.

Because of the expense involved in stemming attrition from the panel, researchers are generally inclined to rely on survey-style research methods and at best to include a small number of open ended questions. However, anthropologists and sociologists have developed longitudinal interview studies where the same participants are interviewed in an open-ended fashion at intervals over a long course of time. I have two studies in the field at the moment--one on the long range careers of workers who entered the labor market in minimum wage jobs in poor neighborhoods and the other on a sample of working poor families, intended to assess the impact of welfare reform on those who were not the targets of policy change--that utilize this approach. In both cases, a representative sample of approximately 100 subjects (for each project) has been interviewed at three to four year intervals for a total of six to eight years worth of data collection. Here it has proven possible to capture changes in perceptions of opportunity, detailed accounts of changing household composition, the interaction between children’s lives and parent’s lives, and the impact of neighborhood change on the fate of individual families. While the sample is very small by the standards of survey research, the depth and nuance of the data that emerges from such an approach is of great value in opening the "black box" that may resist interpretation in studies based solely on administrative records or fixed choice instruments.

Qualitative panel studies are, however, labor intensive and expensive. They ask a great deal from participants who typically have to give up several hours of their time for each wave. In order to generate adequate response rates, given these high demands, it is generally important to provide honorariums that exceed the incentive payments of typical survey projects by quite a bit. Longitudinal interview studies can be conducted by having interviewers take notes recording responses, but are typically done via the use of tape-recorded interviews which must then be transcribed and possibly translated. Given the nature of the data that studies of this kind are aiming for, it is particularly important to use interviewers who are matched for race and gender, and imperative to have staff fluent in the languages of the subjects. All of these requirements add to the costs involved in research and, unless budgets are overflowing, limit to small numbers the size of the sample that can be tracked over time.

Nonetheless, the quality of the data obtained in this fashion can make qualitative, longitudinal studies well worth the effort. This may be particularly true when one wants to go beyond a scholarly or policy audience to engage either the general public or political figures in the exploration of welfare reform. Illustrating statistical trends with "real life" examples of the dilemmas and success stories of former welfare recipients is of great value in this regard. Researchers should not cede to journalists the entire responsibility for telling the story of welfare reform "with a human face" since reporters do not select their "informants" systematically and there is no guarantee that their accounts will be anything more than anecdoctal.

Qualitative panel studies can be developed with an original sampling strategy that picks up a representative population based on neighborhood residence or participants and matched controls who participate in a social service program. However, they are probably most valuable when they are embedded in panel studies using a survey design and are therefore subsets of the much larger population of survey respondents that can serve as a better basis for statistical analysis.

There is one disadvantage to this embedding strategy: if the underlying survey is part of a longitudinal panel study, the selection of a sub-sample that will be accorded more attention may bias the responses of this group to succeeding waves of the survey. Researchers need to evaluate this possibility, though it need not be a serious flaw. Most surveys seeking to track the consequences of welfare reform are going to focus on "objective" and measurable outcomes: hours worked, income earned, job acquired, jobs lost, health insurance enrollment, and so forth. Qualitative sub-studies may yield additional information on how these states of being were arrived at (job search strategies, barriers to insurance enrollment), but in most instances will not compromise the underlying information in a negative (concealing) direction. The experience of providing more information through open-ended interviews may, in fact, encourage greater revelation among the participants in the qualitative study. Researchers will want to check for any systematic biases that may be emerging and, for some purposes, exclude the sub-sample from statistical analyses of the survey population.

However, I would argue that the advantages of selecting the qualitative sample from an original panel population far outweigh the disadvantages. "Soft" studies of this kind are often suspect on grounds of representativeness and the value of their contribution dismissed as a result. While one could, in theory, recruit participants in a qualitative study who are similar to those in the survey population, it is always possible that these "add ons" differ enough from the participants to raise doubt. Hence, in my view, is it a safer bet to draw the qualitative sample from the original research universe and risk the chances that their involvement may alter their responses to a longitudinal study. (Obviously this is not a problem if the underlying survey is cross-sectional.)

Whether based on close ended questions or intensive open-ended formats, interviews can never do more than provide the research community with a window on the self-reported state of mind or experience of those undergoing the transition from welfare to work. Administrative records can provide an objective means of measuring enrollment patterns or recidivism rates, but often yield little insight into the reasons for the differential behavior of particular subgroups. Interviews can supplement administrative records to answer these "why" questions, but here too they are limited by the self-report nature of the data. When subjects are unaware of the reasons for their conduct, inclined to conceal some aspect of their behavior, or unable to recall critical details, puzzles may remain.

Anthropologists and qualitative sociologists often combine interviews with a large "N" with direct observation of behavior in order to fill in these gaps. For some years now, for example, I have been conducting a study of the impact of welfare reform on the working poor in New York City. This is a longitudinal interview study involving 100 families in three ethnic groups spread around Manhattan, Brooklyn and the Bronx. Three waves of interviews over a period of six years provide a detailed sense of the difficulties these families have encountered securing childcare or finding work that dovetails with family responsibilities, even though they were not the targets of reform per se (since they were already in the labor market). The two waves of interviews we have completed thus far indicate that although employment opportunities have improved in recent years, problems remain for the working poor precisely because their wages do not push them above the poverty line. Family budgets are strained; relatives are doubling up; children are moving back and forth from New York to Puerto Rico because, as parents are preoccupied with work all day, some of them are having trouble supervising their children. These observations are clear enough in the interviews.

However, this data provides only a sketchy sense of how these dilemmas surface at the neighborhood level and how, in turn, that ecological context impacts back on the families in our study. Hence we developed a community study component of the project, a year’s worth of intensive fieldwork in three New York neighborhoods--one primarily African-American, one largely Dominican, and one with a large number of Puerto Ricans as well as other Latino immigrants from Mexico and central America. For the past seven months we have been tagging along beside police officers, sitting in classrooms, visiting with congressmen and church leaders, talking with local employers, and devoting a lot of attention to twelve families drawn from our interview sample who live in these three communities. Participant-observation has proven to be an invaluable addition to what we know from the interview data. For example, we have been able to see for ourselves what the teen culture of the communities is like and the ways in which is it influencing the behavior of particular members of the households we study as the parents are occupied at work. We have witnessed the dilemmas of poor, working mothers who cannot easily control their sons when they reach adolescence and we know how they adjust their work lives to try to provide more opportunity for surveillance. Having worked with these families over a long period of time--before and after their re-entry into the labor force--we have learned that their capacity to steer their adolescents has declined sharply.

For example, one family we have come to know quite well has a teenage son who is faring poorly in his middle school. When his mother was receiving public assistance, she was able to visit the school during the day to confer with his guidance counselors at length and to learn directly from them (as opposed to the filtered news from the son) that he was in danger of being held back a full year. Now that this mother is working full time, she is unable to exercise this level of involvement. Her family clearly benefits from her increased earnings: there is less tension over finances in the household and the departure of a paying boarder was less of a cataclysm than it might have otherwise been. However, the mother is worried about what will become of her teenage son in school and is now dependent on him for information on his progress. He is clearly at risk for dropping out altogether which may have an impact on his mothers employment stability and will surely influence his own trajectory in adulthood.

Our home based and day-care based fieldwork has also helped us understand the dilemmas of working poor mothers who have been unable to afford or locate quality childcare. The youngest children in some of these households are showing the effects of poor quality care, with some displaying seriously worrying behaviors that their mothers believe are the result of untrained or unconcerned child care providers (including relatives pressed into service). When we compare these children to their older siblings, most of whom had more attention from their mothers when they were little, the differences are striking. This tells us that there is a problem to be understood here, for the good fortune of mothers (most of whom report being happier because they are working) may be paralleled by the declining fortunes of their youngest children, an outcome many of Edin et al’s (1999) interviewees worried about in advance.

The knowledge we have gained about the work lives of our main informants is complemented by the fieldwork we have done in the neighborhoods. We know, for example, that while opportunities for factory work are very limited in the city itself, that a whole system of private, off the books, van pools carries Dominican workers out to New Jersey factories where they earn just above the minimum wage. Our observational data has taught us that the van pools themselves have become a major source of information on job openings for low skilled workers. The cost of this reverse commute are fairly onerous for low earners, however, amounting to more than 10% of take home earnings for most users.

The perspectives of service providers, teachers, police officers, local politicians and employers are equally valuable, for they are in a position to look beyond the immediate concerns of particular families to assess the consequences of welfare reform for neighborhoods and the institutions within them that must absorb the demands that policy change visits upon them. Service providers, particularly those in the child care and medical care fields, are concerned that they cannot respond adequately to the additional needs that have surfaced since time limits became the law of the land. Medical care personnel in poor neighborhoods continue to report that they have not been able to enroll enough children in Child Health Plus and that they are seeing a steady, and often overwhelming, demand in emergency rooms for treatment of conditions that should have been seen long before they reach this critical point.

Teachers and guidance counselors have noticed that they have a harder time getting parents to pay attention to children’s school behavior or academic problems because they are not as available as they once were. The coincidence of welfare reform and the imposition of new state testing standards for children at all levels of the school system has ratcheted up the stakes in classrooms throughout New York City, leaving teachers and school principals even more concerned about bringing those with educational deficits up to speed. Without easy access to parents, this is proving a complex task. Ironically, however, this very demand, has spurred the city to provide summer school classes, which has been an answer to many a working mother’s prayers for childcare.

Police officers report steep declines in crime and much safer streets in the three neighborhoods we are studying. There is no evidence that this trend is related in any direct way to welfare reform, but is instead part of a nation-wide pattern that experts have yet to fully understand. In New York City, however, the move toward more aggressive policing in minority enclaves has met with mixed responses as a number of notorious cases involving police violence have shown. On a day to day level, however, these pressures have surfaced in a higher level of street surveillance and some resentment of "police harassment" by youth in the families we study who report being told to "move on" when they are talking with friends on the corner. Young men, in particular, feel somewhat less welcome in their own neighborhoods than they once did. For adults, particularly women and elderly men, however, these changes have been a blessing. They can pass without as much fear, walk to and from the subway without worrying about being harrassed by drug dealers. Some report that the drug trade has moved indoors and off the streets, which makes them feel more vulnerable than they did before, but on the whole they approve of the changes or at least are willing to tolerate the increase in police aggression because it means fewer worries accompanying their ordinary movements.

It remains to be seen whether neighborhood safety will improve to the point where one of the chief worries of women moving off of welfare and into work--that their children will not be safe if left unsupervised or will get into trouble in the absence of their mothers--will be assuaged (Edin et al 1999, Edin and Lein 1997, Newman 1999, Anderson 1999). This reservation has played a key role in the past in keeping mothers out of the labor market. Until now, crime rates seemed responsive mainly to levels of community social capital (Sampson 1998) which could, in turn, be boosted through the deliberate efforts of stay-at-home mothers and elders. The absence of mothers from neighborhood streets as they head into the workplace renders this strategy less effective. In any case, if crime continues to decline, we may see that a key purpose of welfare reform (to get mothers into jobs) will be furthered by policy changes that had nothing to do with it (investment in community policing [Winship 1999] or the drive to lower crime rates).

Participant observation in TANF offices and in welfare-to-work programs is an important part of the picture as well. Mark Rank’s (1994) study is one of the few that attempted to get inside the culture of the old AFDC system and it was very valuable for understanding the perspective of welfare clients as they were processed by caseworkers. We shall have to await a new generation of organizational studies based upon similar fieldwork methods in order to understand how the new goals of TANF offices-- especially job placement--are being absorbed into a bureaucracy that was designed for entirely different purposes (Ellwood 1988). Qualitative research on welfare-to-work programs can tell us a great deal about the job retention problem as well. Watkins (1999) offers a compelling account of the disjuncture that plagues some programs that try to build self-esteem as a means of retaining participants, only to discover that graduates expect much more from the labor market as a result than they actually find. High job turnover rates follow as the frustration of discovering that an "I am somebody" campaigns run headlong into the low wage labor market where the message may be something closer to "You are not important."

What these examples are intended to illustrate is the value of contextual information generated through the use of long term fieldwork. Among other things, this approach provides something close to a continuous monitoring of a small sample of families or participants in organizations. Rather than let weeks or years go by between short-term contacts, fieldwork permits ongoing contact and the capacity to check what informants say about their state of mind, their survival strategies, their relations with others, and their neighborhood or institutional conditions against what fieldworkers can observe for themselves and/or learn from "experts" situated in the community.

The data derived from interview and participant observation projects can be used in at least two different ways: to generate hypotheses that might be turned into survey research questions for use with a much larger representative sample or as an end in and of itself. These two aims are not mutually exclusive. The difficulty, of course, with the "end in itself" approach is that questions of representativeness are always vexing with very small samples and for most research in this genre, small samples are the only affordable possibility.

My own approach has involved embedding the selection of informants within a larger survey design in order to respond to this concern. In 1995-96, we undertook a survey of 900 middle aged African Americans, Dominicans and Puerto Ricans in New York City, chosen to be representative of ethnically diverse and ethnically segregated neighborhoods, with both high and low levels of household income. From this population, a representative subsample of 100 respondents were chosen for in-depth interviews at three year intervals (1998 and again in 2001). Finally, twelve individuals - four from each of the ethnic groups of central concern - living in the three neighborhoods described in the previous section were selected from this qualitative subsample. The choice of these particular twelve people was guided mainly by their employment status and family type, with a mix of single parents and intact couples. This nested design has made it possible to generalize with a reasonable degree of confidence from the families we have come to know best to the population as a whole that we began with.

A similar approach has been pursued by the Manpower Demonstration Research Corporation’s "Urban Change" project, a study of the impact of devolution and the time limits of the TANF system on poor families in four cities: Philadelphia, Cleveland, Miami and Los Angeles. A multi-disciplinary team of social scientists are drawing upon "administrative records; cross-sectional surveys of food stamp recipients; census tract-level neighborhood indicators; repeated interviews with Executive Directors of community-based social service organizations; repeated ethnographic interviews with welfare-reliant women in selected neighborhoods; and repeated interviews with and observations of welfare officials and line staff" (Edin et al 1999:6). 2

The qualitative interview part of the Urban Change project has been following 80 families from high and medium poverty neighborhoods in Cleveland and Philadelphia. Under the direction of Kathryn Edin at the University of Pennsylvannia, this project has thus far collected a large amount of baseline information on a series of topics including:

Apirations for [women’s lives] and their children; experiences with case workers and the welfare system; knowledge about and attitudes toward welfare reform; income and expenditure patterns; educational and work experiences; family life; attitudes toward marriage and future childbearing; health and caregiving; social support; material hardship; use of social service agencies; and perceptions of the quality of their neighborhoods. (Edin et al 1999: 6).

Families were chosen for this part of the study by selecting three neighborhoods 3 in each city with moderate to high concentrations of poverty (over 30% living below the poverty line) and welfare receipt (20% or more of families receiving welfare). 10-15 families were recruited in each neighborhood in a fashion that avoided the liabilities of drawing from lists provided by TANF offices. Since the Urban Change project investigates behaviors that are technically disallowed (esp. unreported income, in-kind contributions from fathers, etc.), it was important for the researchers to be able to present a truly independent face to their informants, untainted by connection to enforcement agencies that could affect their benefits. Instead, the researchers posted notices in the target neighborhoods, knocked on doors, requested referrals from community leaders and local institutions. They attempted to guard against the over-representation of any given social network by utilizing no more than two recruits through any of these sources.

A strategy of this kind probably over-represents people who are higher on social capital than some of their more isolated counterparts. They have connections. A strict sampling design from an established list may pick up people who are less "hooked in" to institutional resources or private safety nets and will therefore tell us something about people who confront welfare reform from a socially isolated vantage point rather than from a more integrated position. However, the liabilities of this approach are considerable as well, for it is much harder to disassociate from official agencies when pursuing a sample generated randomly from, for example, a TANF office case load.

The neighborhood strategy employed by the Urban Change project insures that the qualitative study is representative of white, black and latino families who are particularly disadvantaged. As Edin et al (1999:7) have explained, the design will not pick up welfare recipients who live in mixed income or more affluent neighborhoods. It is possible that this strategy will yield a slightly darker perspective on the consequences of welfare reform if we take as our universe of concern the entire range of long term recipients, many of whom moved off of the rolls with apparent ease as unemployment declined. These are the people whose human capital, including prior work experience, has made them relatively easy to place. What the Urban Change project will tell us is how this transition has effected those with less going for them, because their neighborhoods (and the contacts they derive from them) are less likely to provide useful information for job hunting. The same environments undoubtedly present safety concerns that mothers will have to factor in as they scramble to figure out how to care for their children. In the end, these are the more pressing questions in need of answers, hence the wisdom of the Urban Change project’s approach.

Urban Change is not an ethnographic project in the strict sense of the term. Contact is maintained intermittently with the target families, often utilizing telephone interviews in place of face to face contact. Intervals of contact are approximately six weeks, though this varies by the informants’ situation. Nonetheless, it will provide a very rich data base, spanning the before and after of the imposition of time limits, that will tell us an enormous amount about the challenges women and their families have faced in transitioning to from public assistance to the world of work. The size and ethnic diversity of the sample (including poor whites, often overlooked in studies of the poor), the multi-city approach, and the fusion of administrative records, expert perspectives, and the inclusion of welfare-reliant families in communities with varying levels of poverty, will help to address many of the more important theoretical questions before us, most especially the consequences of race and ethnic differences, neighborhood effects and human capital differences in the unfolding of welfare reform.

William Julius Wilson and his colleagues are in the midst of a similar three-city study of welfare reform and its consequences. This project involves a survey of 2,800 low income households from poor and moderate income communities which began earlier this year. The sample is divided between TANF recipients and those who do not receive these benefits. It is restricted to households with young children (younger than four) and those with children between 4 and 14. A developmental study of 800 of these families who have children ages 2-4 will be embedded in this larger design, with interviews from caretakers and the fathers of these children to be included.

[This component] will track how welfare policies affect the daily lives and neighborhood resources of poor families. In-depth interviews will be conducted over the course of two years and will cover such topics as the respondent’s life history and daily routines of life. This component also includes diary studies and observations of the participant when she goes to social service offices for assistance." (Moffitt and Ver Ploeg 1999: 67).

The great advantage of the three city study is the way in which the ethnographic sample is nested inside a huge research machine that can analyze neighborhood variables, state and local level employment data, the statistical power of a 2800 person survey, and the repeated interviews and family assessments in the child development portion of the project.

This project has an enormous budget and is therefore the "Cadillac" model that few other studies of welfare reform will be able to match. Nonetheless, it is theoretically possible to utilize a very rich fieldwork approach as long as the resources for this labor intensive form of data gathering are on hand. Few social scientists would disagree that moving from macro-level findings based on surveys to the most micro-level data drawn from fieldwork, with mid-range interviews and focus groups in between is the best possible approach for preserving representativeness but building in the richness of qualitative research.

Very few research projects will be able to match the scale of the MDRC study or the Three City project. Indeed, even my own more modest study of l00 families in one city required a substantial research budget and a rotating team of fieldworkers willing to commit a total of more than six years to the enterprise. Of course, not all studies of welfare reform need to be as long in duration as the ones described here. For state and local officials whose aim is less to explore the theoretical questions that motivated these studies and more to learn in depth about the family management problems of their caseloads, it may be possible to arrange with local universities to organize neighborhood-based research projects that will provide "snap shot" versions of the same kinds of questions.

Another sampling strategy involves the use of "snow ball" samples which attempt to capture respondents who share particular characteristics (e.g. low wage workers or welfare reliant household heads) by asking those who meet the eligibility criteria to suggest friends or neighbors who do as well. Some classic studies in the annals of poverty research have used snowball samples (e.g. Lillian Rubin’s Worlds of Pain, Elliot Liebow’s Tally’s Corner) to great effect. More recently, Edin and Lein’s Making Ends Meet relies upon referrals from a variety of sources, including the personal contacts of individuals already in their study population, to build a sample in four cities. The defining feature of a snowball sample is that it gathers individuals into a sample who have some acquaintance with those who are already involved. They are, in this respect, not a random sample, though they may come to approximate the demographics of a sample this is genuinely random.

Snowballs can be tightly bound to a particular network, as was the case in Tally’s Corner, or can guard against the possibility that membership will not represent truly independent cases. When object of study is densely connected webs of friends and relatives, it will be important to capture naturally occurring social networks. Hence the initial selection of the key informant needs to pay attention to representativeness. Thereafter, however, there will be nothing random about the study participants: they will be selected members of the original informant’s trusted associates.

For example, in my recent study of the working poor in central Harlem (Newman 1999), a representative sample of workers in fast food restaurants formed the core of the research, but a selected subsample were central to a final phase of intensive participant-observation that focussed on the survival strategies of ten households and the social networks attached to them. The ten key informants were selected to represent the racial and gender diversity of the universe of workers. Branching out from there, in concentric circles around the 10 key informants, we took in the friends, neighbors, schoolmates, teachers, preachers, distant relatives, and street contacts of these individuals. Hence, while the original subsample was representative, the snowballs grew around them because the purpose of the study was to learn about how these households managed the many challenges of low wage work in naturally occuring contexts (school, home, church, extended family, etc.) Ultimately, perhaps as many as 500 additional people were included in this phase of the research, though they were hardly a random sample.

Others have used snowballs to generate the "master sample," however and here it is important to guard against the possibility that network membership is biasing the independence of each case. Some snowball samples are assembled by using no more than one or two referrals from any given source, for example. Edin and Lein’s (1997) important study, Making Ends Meet, is a good example of a partial snowball strategy that has made independence of cases a high priority. Initially, they turned to neighborhood block groups, housing authority residents’ councils, churches, community organizations and local charities in order to find mothers who were welfare reliant or working in the low wage labor market in Boston, Chicago, Charleston, and San Antonio. Concerned that they might miss people who were disconnected from organizations like those who served as their initial sources, Edin and Lein turned to their informants and tried to diversify:

To guard against interviewing only those mothers who were well connected to community leaders, organizations and charities, we asked the mothers we interviewed to refer us to one or two friends whom they thought we would not be able to contact through other channels. In this way, we were able to get less-connected mothers. All in all we were able to tap into over fifty independent networks in each of the four cities. (ibid: 12).

Using this approach, Edin and Lein put together a heterogeneous set of prospective respondents who were highly cooperative. Given how difficult it can be to persuade poor people who are often suspicious of the motives of researchers (all the more so if they are perceived as working for enforcement agencies), working through social networks can often be the only way to gain access to a sample at all. Edin and Lein report a 90% response rate using this particular kind of snowball technique. Since this is higher than one usually expects, it may be that there is less independence among the cases than would be ideal under random sample conditions, but it is far preferable to an approach that is more random but with very low response rates.

Qualitative research of any kind -- open-ended questions embedded in surveys, ethnographic interviews, long-term fieldwork with families or "neighborhood experts" -- generates mountains of text. Anthropologists and qualitative sociologists accustomed to working with this kind of data have developed various means for boiling it down in ways that make it amenable to analysis. At the simplest level, this can mean developing coding schemes that transform words into numeric representations that can be analyzed statistically as one would do with any kind of close-ended survey data. Hence, turning to the Urban Change project, for example, we find that initial baseline interviews show that respondents are hoping that going to work will provide them the wherewithall to provide a variety of opportunities for their children. Mothers also report that they expect their social status to rise as they depart welfare and note that their children have faced taunting on account of their participation in AFDC that they trust will cease once they are independent of state support. These findings come from tape recorded base line interviews intended to capture their prospective feelings about moving into the labor market some two years in advance of the imposition of time limits. These responses can be coded into descriptive categories that reflect the variety of expectations respondents have for the future, or the hopes they have expressed about how working will change their lives for the better.

Most qualitative interview instruments pose open-ended questions in a predefined order. They may also permit interviewers some latitude to permit informants to move the discussion into topic areas not originally envisioned. Within limits, this is not only acceptable, it is desirable, for understanding the subjective perspectives of the respondents is the whole aim of this kind of research and the instrument may not effectively capture all the relevant points. However, to the extent that the original format is followed, the coding can proceed by returning to the responses that are contained in approximately the same "location" in each interview transcript. Hence, every participant in our study of the working poor under welfare reform was asked to talk about how their neighborhood has changed in the last five years. Their responses can be categorized according to the topics they generally raised: crime declining, gentrification reflected in rising rents, new immigrant groups arriving, and so forth. We develop codings that reflect these routine responses in order to be able to draw conclusions such as "50% believe that crime has declined precipitously in their neighborhood," or "20% object to police harrassment of their teenage children."

However, we also want to preserve the nuances of their comments in the form of text blocks that are "dumped" into subject files that might be labeled, "attitudes toward the police" or "comments on neighborhood safety." Researchers can then open these subject files and explore the patterned variety of perspectives on law enforcement or the ways in which increasing community safety have effected patterns of movement out of the home or the hours that mothers feel comfortable commuting to work. When qualitative researchers report results, we typically draw upon these blocks of text to illustrate the patterns we have discovered in the data, both to explore the nuances and to give the reader a greater feeling for the meaning of these changes for the informants. To have this material ready at hand, one need only use one of a variety of text processing programs, including Atlas.ti, Nud*ist, and Ethnograph each of which has its virtues. 4 Some proceed by using key words to search and then classify the text. Others permit the researcher to designate conceptual categories and then "block" the text with boundary markers on either side of a section so that the entire passage is preserved. It is even possible to use the indexing capacities of standard word processing programs, such as Microsoft Word 6.0 and above, which can "mark" the text and dump it into subject files for later retrieval.

Most qualitative projects require the analyst both to digest the interviews (which may be as long as 70 pages or more) into subject headings and to preserve the flow of a single informant’s interview through summaries that are preserved by person rather than by topic. I typically maintain both kinds of qualitative data bases, with person-based summaries boiling a 70 page text down to 5-6 pages that give me a thumbnail sketch of each interview. This is an approach of primary value to an academic researcher and may be of less importance for practitioners who may be less interested in life histories for their own sake and more concerned with responses to welfare reform per se.

There can be little doubt that qualitative research is an essential part of the picture if we are to understand the real consequences of welfare reform. Equally, though, we have to admit that it is a complex and expensive undertaking, one not easily suited to the resources of local TANF officials who may nonetheless rely on research as a means of understanding the dynamics of their case loads or who seek to improve service delivery. One imagines this will be even more important if rolls continue to fall, leaving only the most disadvantaged to contend with. If the pressure to find solutions for this harder-to-serve population grows, it may well become critical for administrators and policy makers to figure out new strategies for addressing their needs. This will not be easy to do if all we know about these people is that they have not found work or have problems with substance abuse or childcare. We may need to know more about how their households function, about where the gaps are in their childcare, about the difficulties of accessing drug treatment, or the concerns they have about the safety of older children left unsupervised in neighborhoods with crime problems.

Is this qualitative research challenge one that federal and state officials should move to meet? Will they be able to make use of this information, above and beyond the more normative studies they conduct or commission on caseloads in their jurisdictions? To answer this question, I turn to several interviews with officials at the federal and state levels whom I’ve asked to comment on the utility of qualitative data in their domains. Their observations suggest is that the range of methods described in this paper do indeed have a place in their world and that the investment required to have this material "at the ready" has paid off for them in the past. However, the timing of these studies has everything to do with the resources available for research and the information demands to which officials have to respond. For some, the time is right now. For others, qualitative work will have to wait until the "big picture" based on administrative records and surveys is complete.

Dennis Lieberman, Director of the Department of Labor’s Office of Welfare to Work, is responsible for demonstrating to Congress and therefore to the public at large that the programs under his jurisdiction are making a significant difference. As is true for many public officials, Lieberman’s task is one part politics and one part policy science: political in that he has to communicate the value of the work this program accomplishes in the midst of competing priorities and scientific in that the outcomes that show accountability are largely "bottom line," quantitative measures. Yet, as he explains below, this is a complex task that cannot always be addressed simply by turning to survey or administrative records data:

One of the major responsibilities I have is to demonstrate to the Congress and the American people that an investment of $3 billion (the size of the WtW grants program) is paying off. Numbers simply do not tell the story in its entirety or properly. Often times there are technical, law-driven reasons why a program may be expanding or enrolling slowly. These need to be fixed, most often through further legislative action by Congress.

From a surface perspective a program may appear as a poor investment. Looking behind the numbers can illuminate correctable reasons and present success stories and practices whose promise may lie buried in a statistical trend. As an example: one of the Welfare to Work program criteria (dictated by statute) would not allow service providers to help those individuals who had a high school diploma. We were able to get that changed using specific stories of individuals who were socially promoted, had a high school diploma (but couldn't read it) and were in very great need. Despite all this, they were walled out of a program designed specifically for them. A high school diploma simply did not lift them out of the most in need category. The numbers showed only low enrollment, appearing at first glance, like recruitment wasn't being conducted vigorously enough. (Lieberman, personal communication).

As this comments suggests, qualitative work is particularly useful for explaining anomalies in quantitative data that, left unsolved, may threaten the reputation of a program which officials have reason to believe is working well but which may not be showing itself to best advantage in the standard data bases.

These evaluations are always taking place in the context of debates over expenditures and those debates are often quite public. Whenever the press and the public are involved, Lieberman notes, qualitative data can be particularly helpful because it can be more readily understood and absorbed by non-specialists:

Dealing with the media is another occasion where numbers are not enough (although sought first). Being able to explain the depth of an issue with case histories, models, and simple, common sense descriptions is often very helpful in helping the press get the facts of a program situation correct. There is a degree of "spin distrust" from the media, but the simpler and more basic the better. This, of course, also impacts on what Congress will say and do.

Quite apart from these questions of evaluation and public debate, some federal officials have found qualitative data useful in the context of program design and "tinkering" to get the guidelines right. Focus groups and case studies help policy makers understand what has gone wrong, what might make a difference, and how to both conceptualize and then "pitch" a new idea after listening to participants explain the difficulties they have encountered. Lieberman continues:

I personally have found qualitative data (aside from numbers) as the most useful information for designing technical assistance to help grantees overcome program design problems, to fix processes and procedures that "are broken", to help them enrich something with which they have been only moderately successful, and to try something new, which they have never done before.

My office often convenes groups of similar-focus programs for idea sharing and then simply listens as practitioners outline their successes, failures, needs, and partnerships. We convene programs serving non-custodial fathers, substance abusers, employers and others. We have gotten some of the most important information (leading to necessary changes in regulation or law) this way.

Gloria Nagle, the Director of Evaluation for the Office of Transitional Assistance in the State of Massachusetts, faces a different set of demands and therefore sees a slightly different place for qualitative work. She notes (personal communication, 11/30/99) that her organization must be careful to conduct research that is rigorous, with high response rates and large representative samples in order to be sure that their work is understood to be independent and scientific. Moreover, because collecting hard data on welfare reform is a high priority, her office has devoted itself primarily to the use of survey data and to the task of developing data bases that will link various administrative records together for ongoing tracking purposes. However, she notes that the survey work they are doing is quite expensive and that at some point in the future the funds that support it will dry up. It is at that point, she suggests, that qualitative data will become important:

Administrative data is like scattered dots. It can be very hard to tie the data together in a meaningful way. Quarterly UI earnings data and information on food stamps might not give a good picture of how people are coping. For example, what about former welfare recipients who are not working and not receiving food stamps? How are they surviving? We can't tell from these data how they are managing. When we no longer can turn to survey data to fill in the gap, it would be very useful to be able to do selective interviews and focus groups.

Nagel sees other functions for qualitative research in that it can inform the direction of larger evaluations in an efficient and cost effective fashion.

Finally, she notes that focus groups and participant observation research is a useful source of data for management and program design purposes:

A more comprehensive opinion poll of federal and state officials on the program and on the research evaluation side would no doubt generate other perspectives. Suffice to say for the moment, there is potential for qualitative data to take its place in the arsenal of research approaches we will need to bring to bear if we are to understand what welfare reform has really meant over the long haul.

Given the complexities of this style of research, it would probably be most effective for state agencies to provide RFP’s that local universities can respond to as part of their public service and training activities. Students are a good source of research labor and are often very interested in the problems of the poor. Sociologists, demographers, political scientists, and anthropologists can all be drafted to assist state officials in understanding how welfare reform is unfolding. With proper planning, long term panel studies that embed qualitative samples inside a large-scale survey design, can be conducted in ways that will yield very valuable information to policy makers and administrators. Utilizing this kind of partnership has the advantage of independence from the enforcement agencies with whom TANF participants may be reluctant to cooperate. Since most states have a network of public universities distributed throughout the territory, one can make use of their location to generate appropriately diverse research populations -- urban/suburban/rural, multiple ethnic groups, neighborhoods with different levels of poverty, and areas with higher and lower levels of unemployment, being among those probably most important to represent.

Research units of state agencies can also invest in in-house capacities for qualitative research. Even when research resources are tight, making sure that ethnographers and interviewers are part of the team is an important management decision. This may appear to be a "frill," but as I have already noted it may actually save the day when survey results cannot explain the findings on recidivism or childcare. The presence of ethnographers and interviewers in federal agencies is commonplace now. For example, the Census Bureau maintains a staff of anthropologists and linguistics who study household organization in order to frame better census questions. In years past, the Bureau has employed teams of ethnographers to conduct multi-city studies of homeless populations in order the check under-representation in the census. As devolution progresses, it will be important to replicate this expertise at the state level in the field of welfare reform as well.

Whether research partnerships or in-house teams are chosen, the greatest success will undoubtedly be achieved when qualitative research is embedded inside quantitative studies that are either cross-sectional or longitudinal panel studies. The fusion of the two approaches provides greater confidence in the representative nature of qualitative samples and the capacity to move back and forth between statistical analyses and patterns in life histories renders either approach the richer for its partner.

The design of the qualitative sample in New Hope took a random draw from all program and control cases that fell into the family and child sample (essentially, cases with at least one child 0-10 at the point of random assignment). The research team did some stratification before drawing the sample, sorting the list by program vs. control status, then by race. Thereafter, the sampling was random within these cells. (See Weisner et al 1999).
See Quint et al 1999 for more detail on the methodology of the Urban Change project.
Neighborhoods are defined in terms of census tracts. Ranging in size from one to four contiguous tracts, these neighborhoods must meet the poverty, welfare receipt and racial/ethnic composition (Edin 1999: 7)
For helpful reviews of these software packages, see Barry (1998) or "QDA Overview" on the web at http://www.quarc.de/body_overview.html.

1999 My Children Come First: Welfare-Reliant Women’s Post-TANF Views of Work-Family Tradeoffs, Neighborhoods and Marriage. Paper Presented at the Northwestern/University of Chicago Joint Center for Poverty Research Conference, “For Better or Worse: State Welfare Reform and the Well-Being of Low-Income Families and Children,” Washington D.C., September 16-17.

1997 Making Ends Meet: How Single Mothers Survive Welfare and Low-Wage Work. New York: Russell Sage Foundation.

1998 The State of Welfare Caseloads in America’s Cities. Brookings Institutions Center on Urban and Metropolitan Policy. Washington DC.

1999 Evaluating Welfare Reform: A Framework and Review of Current Work. Interim report of the Panel on Data and Methods for measuring the Effets of Changes in Social Welfare Programs. Committee on National Statistics, National Research Council. Washington, D.C.: National Academy Press.

1999 No Shame in My Game: The Working Poor in the Inner City. New York: Knopf/Russell Sage.

1999 Big Cities and Welfare Reform: Early Implementation and Ethnographic Findings for the Project on Devolution and Urban Change. New York: Manpower Demonstration Research Corporation.

1995 Living on the Edge: The Realities of Welfare in America. New York: Columbia University Press.

1999 Operationalizing the Welfare to Work Agenda: An Analysis of the Development and Execution of a Job Readiness Training Program. Unpublished manuscript, Department of Sociology, Harvard University.

1999 Getting closer to understanding the lives of economically poor families: Ethnographic and survey studies of the New Hope experiment. Poverty Research News. The Newsletter of the Northwestern University/University of Chicago Joint Center for Poverty Research. Based on the Workshop on Qualitative and Quantitative Methods sponsored by the Joint Center, June 1999, Chicago, IL.