The Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) famously ended welfare “as we knew it”—Aid to Families with Dependent Children—and replaced the 60-year-old AFDC with a new program pointedly called Temporary Assistance for Needy Families, or TANF. TANF substantially increased state discretion in choosing how to aid families with children, but it also established various process requirements that were intended to increase efforts to move unemployed adult recipients into jobs. The new regulations generally require close interaction between benefit recipients (often disingenuously termed “customers”) and local agency personnel. Presumably, it is in these interactions that recipient understanding of the nature and terms of public assistance is to some extent formed. If it is the objective of welfare reform to change these perceptions, reform must address what happens when those seeking help encounter the system.

How Management Matters addresses the questions concerning how what happens in the assistance encounter might be influenced by management, whether management indeed responded to the objectives of the 1996 reforms, and whether these responses worked. The book is drawn from the results of a field research study on PRWORA's consequences fielded by the Rockefeller Institute of Government in Albany, New York, and funded by the U.S. Department of Health and Human Services. Data for the research were collected in 1998 and 1999, soon after PRWORA's implementation, in 11 locations in four states: Georgia, Michigan, New York, and Texas. The analysis is based on the study of agency documents; interviews with administrators, managers, supervisors, and front-line workers; a written survey of front-line staff; and direct observation of interaction between agency staff and TANF applicants and participants in TANF.

Norma Riccucci finds that in practice, the goals of welfare as these states were coming to know it conflicted. While high-level officials commonly cited movement to self-support as a central intent of the TANF system, most interaction between workers and clients involved eligibility assessment and benefits determination. Front-line staff, she found, “rely heavily on professional experience, work norms, and familiar routines in performing their jobs” (p. 77). The result is that welfare as we were coming to know it at the end of the last decade continued to have many features similar to what the author believes to have been true of welfare in the past. One strategy adopted by state officials to counteract the intractability of street-level bureaucrats was to place responsibility for delivering work services in a separate bureaucracy—the state employment services. This, she reports, “sent a clear signal to street-level bureaucrats” (p. 117). “The decision to separate income maintenance services from employment or work-first services is a very compelling explanation for how the welfare system was reformed. And this factor may help to explain the decline in caseloads” (pp. 117–18).

Field studies of this type pose many problems for readers. One is the issue of selection. While the targeted four states may have been “purposely selected to maximize variation in policy, administrative structure, and political culture” (pp. 10–11), the book includes no data on the distribution of these features across all states to help in positioning those under scrutiny. A second is absence of control, real or hypothetical, against which the observed outcome might be measured. What was the welfare we were supposed to have known really like? Given that social assistance always involves some sequence of application, eligibility determination, and service response, where in the assistance process should we look for change, and how will we know when we see it? A third concerns just what is to be expected. PRWORA changed federal financing for welfare from a matching grant to a block grant. As a result, states pay all marginal assistance costs wholly from state funds. Are we surprised that eligibility determination continued to be a major focus of operations, despite the lip service paid to the promotion of self-sufficiency? Does this really reflect the “intractability” of the culture of welfare offices or what we should have expected, given that the federal government no longer pays more than half the cost of each new addition to the rolls?

Beyond a lack of framework for directing attention, readers will suffer at various points from the lack of vital methodological detail. Regressions are reported, for example, with no explanation of the metric or distribution of the dependent variables (p. 72). While the literature on managing street-level bureaucrats is richly covered, available research pertinent to some assertions—for example, the role of state employment agencies in delivering services to AFDC and TANF recipients—is ignored. The analysis of allocation of caseworker time suffers from a well-known issue in welfare studies: Welfare workers' attention is never distributed equally across cases. Most effort is devoted to a small, problem-plagued subset. As a result, while time-allocation data provide insight into the agonies of social assistance work, without careful weighting they are unreliable for drawing inferences about the average experience of applicants or established clients. Each caseworker may actually talk about time limits with a client only once a day and spend far more time assessing child support needs, ensuring food stamp receipt, or requesting birth certificates. But if every new applicant is told about time limits at some point during welfare intake, that would be a profound change. The methodology pursued by the Rockefeller Institute tells us a lot about caseworker jobs, but not, in the end, much about the effect of reform on client experience and understanding.

On the other hand, only fieldwork gives us stories. How Management Matters includes many fascinating excerpts from the project's recordings of interactions between those seeking assistance and agency personnel. Despite the various selection biases, when studied with care, the near-verbatim presentation of these episodes has much to offer those who would formulate and reform welfare policy but care to think in advance of the kinds of encounters such reforms are intended to influence.