September 2012, Vol. 135, No. 9
Organizational philosophy and welfare-to-work policies
Download the PDF
Book reviews from past issues
What Works in Work-First Welfare. By Andrew R. Feldman, Kalamazoo, MI, W.E. Upjohn Institute for Employment Research, 2011, 183 pp., $40/cloth; $18/paper.
The passage of the Personal Responsibility and Work Opportunity Reconciliation Act in 1996 marked the realization of the goal, as expressed in the words of President Bill Clinton, to "put an end to welfare as we have come to know it." The relatively unpopular and frequently criticized Aid to Families with Dependent Children (AFDC) was consequently dismantled and replaced by the current federal welfare program, Temporary Assistance for Needy Families (TANF). Although welfare reform encompassed a large number of substantive changes, TANF can now be fairly characterized as resting upon a central premise: the function of welfare is to provide temporary financial assistance in order to facilitate the transition into employment and self-sufficiency. This proposition serves not only as an encapsulation of the philosophy behind welfare reform, but also as a statement of the fundamental challenge faced by state welfare administrators.
Andrew Feldman's What Works in Work-First Welfare is a study that takes a different approach. Unlike books such as Grogger and Karoly's Welfare Reform: Effects of a Decade of Change, which focus primarily on the effects of policies upon the incentives of welfare recipients, What Works in Work-First Welfare addresses policies that affect the incentives faced by frontline staff and managers who administer employment services. The analysis proceeds from the perspective of evaluating how organizations implement their welfare employment programs, with an emphasis placed upon managerial practices. Organizational practices are separated into two broad categories: "strategic policies," designed to directly influence the behavior of welfare recipients, and "managerial practices," aimed at influencing the behavior of middle management and frontline staff. These organizational practices are examined within the context of New York City's welfare system, a system that is markedly different from those in the vast majority of states and cities.
New York City's unique welfare system presents several analytical advantages to the study of employment program implementation. As Feldman describes, the system is partially privatized, with all of the employment services provided by 19 privately contracted organizations that operate 26 Employment Services and Placement (ESP) programs. The ESPs are given a certain measure of freedom to implement their programs on the basis of what they believe is most effective. After state-run Job Centers determine eligibility, welfare recipients are randomly assigned to the 26 ESPs. The analytical benefit of this procedure is that it essentially creates a natural experiment that should theoretically function to reduce selection bias when estimating the impacts of policies, strategies, and practices as they vary among the different ESPs. The presence of for-profit, nonprofit, faith-based, and secular organizations provides further opportunities to investigate differences between practices within these types of establishments.
To identify those policies and practices which influence employment outcomes, Feldman used a multilevel regression model to analyze approximately 14,000 individuals assigned to the 26 ESPs. What was found to work? Three practices were revealed to have statistically strong results. First, full pay-for-performance significantly improved the job placement rate compared with partial pay-for-performance. ESPs are paid in a three-step process based on job placements, retention after 3 months, and retention after 6 months. The partial pay-for-performance ESPs are operated by a single organization that pools the total compensation, blurring the perception of a direct link between the bottom line and performance for the individual ESPs. The results indicated that the partial pay-for performance ESPs exhibited significantly lower placement rates. Second, an emphasis on quick job placement, in contrast to a longer process relying upon more extensive case management, yielded better results. Feldman explains that this finding may indicate that welfare recipients are more resilient and ready to assume the responsibility of working than is sometimes assumed and that the greater urgency on the part of staff to place individuals may translate into greater motivation for the recipients. Finally, promoting an immediate job search, rather than waiting for job training to be completed, is more productive: the placement rate and the caseload employment rate, defined as being placed in a job and working 6 months later, were lower for those ESPs emphasizing job training. Feldman suggests that the attributable to welfare recipients not being well suited to classroom education, with 50 percent never completing training and only 10 percent being placed in jobs related to their field of training. It is also interesting to note that the de-assignment rate (the rate at which the ESP sends those considered unemployable back to the Job Centers for reevaluation), the sanction rate (the rate at which the ESP sends those who break program rules back to the Job Centers), and the size of the program showed no statistically significant relationships.
While the peculiarities of New York City's welfare system facilitate the analysis of interesting organizational relationships, they also serve to limit the extent to which other states can emulate that system. The privatization of employment services is a major undertaking that goes far beyond the implementation of strategic policies or managerial practices. For many states and cities, it may simply not be an option. Furthermore, the adoption of pay-forperformance for middle managers and frontline staff may be restricted by the structure of employee compensation arrangements in some states. More importantly, New York City's very lenient sanction policy creates an environment that may alter the behavior of welfare recipients with respect to the effort they expend in finding and maintaining employment. As a result, the effectiveness of strategic and managerial practices may vary from state to state, depending on the severity of the sanction policies in place.
One of Feldman's major findings is that job training worsens employment outcomes; however, it is necessary to offer a few words of caution regarding this finding. Strictly speaking, a general analysis of the effectiveness of job training cannot be performed with Feldman's data, because of the restrictions placed on training by New York City and the limitations of the data. In New York City, job training cannot exceed 3 months and the types of training available are limited to those performed by approved job training providers. These two constraints effectively eliminate most forms of classroom-based human capital development from consideration. Moreover, Feldman's data measure only employment outcomes for up to 6 months after placement. Recent research has shown that classroombased human capital development typically does not exhibit positive employment outcomes in the short run, but does produce strong positive effects two or more years after training. Therefore, although Feldman's results regarding job training may be valid under the conditions present in New York City, it is important not to draw general conclusions regarding the effectiveness of job training based on the New York City experience alone.
If What Works in Work-First Welfare suffers a weakness, it is that several managerial practices are analyzed only informally. Feldman qualitatively evaluates the impacts of defining a clear mission statement, setting organizational goals, motivating staff around goals, and measuring and monitoring performance, but, as he acknowledges, the results are not particularly solid. Nevertheless, Feldman does proceed to draw the conclusion that frontline managerial practices are not as important as higher level strategic policies, an inference that is surely tenuous given the qualitative nature of the analysis. In fairness, though, quantifying and statistically analyzing some of these practices is particularly difficult. Accordingly, the criticism just raised should not be given undue weight.
The audience that will likely benefit most from a reading of What Works in Work-First Welfare consists primarily of those involved in the administration of federal and state welfare services. Case managers and job developers should take away a better understanding of how welfare programs operate and their roles within it; statistical results should be of interest to frontline staff. However, policymakers and program directors responsible for determining strategic policies and managerial practices stand to gain the most from the book. Although not all of the policies and practices described therein will be realistically applicable beyond New York's unique welfare system, some are certainly general enough to merit consideration in most states. And for those practices treated only informally within the book, the discussion should still serve to stimulate creative thinking about how to evaluate these difficult-to-measure aspects of welfare-to-work programs.
–John Kenneth Krantz
Utah Department of Workforce Services
Salt Lake City, UT
Within Monthly Labor Review Online:
Welcome | Current Issue | Index | Subscribe | Archives
Exit Monthly Labor Review Online:
BLS Home | Publications & Research Papers