Welfare Reform and Its Effects on the Social and Emotional Development of Young Children (Birth to Five Years): Comments on Morris and Kamerman

Introduction

The welfare reform legislation of 1996, as noted by Morris and Kamerman, overhauled Aid to Families with Dependent Children (AFDC) from a federal entitlement to a system that has become increasingly focused on encouraging and requiring very low-income parents to find jobs and stay employed. Mandatory work requirements and time limits remain the most hotly debated components of current federal and state TANF (Temporary Assistance for Needy Families) programs. Faced with meeting the dual goals of decreasing welfare roles and reducing public expenditure, and with helping families attain self-sufficiency, federal and state policy-makers have traditionally been focused on the implications of welfare reform policies on parental economic well-being and sometimes on their marital stability.  How such policies might affect children is often of secondary concern during policy debates, and rightly so, since, with the exception of teen parents, welfare policies are not designed to directly target children.  Fortunately, due to the foresight of and financial support from federal and state agencies and several private foundations, efforts were made just over a decade ago to collect information about the outcomes of children in welfare and low-income families in the context of welfare reform.  Morris and Kamerman summarize several of the main findings in this emerging body of important research.

Research and Conclusions

Morris summarizes findings on the effects of welfare policies on children’s outcomes from a set of random assignment studies of welfare and employment programs that were originally conducted by MDRC and re-examined and synthesized as part of MDRC’s Next Generation project. As a group, these studies, which include the Milwaukee New Hope project and the Canadian Self-Sufficiency Project (SSP), tested three distinct policy components (earnings supplements, mandatory employment services and time limits). Taking place during the early 1990s, largely under federal welfare waivers, these studies were not designed to test 1996 welfare reform, but nevertheless did test policies that are key components in several current TANF programs. Because of the random assignment design of the studies, Morris rightly argues that these experiments provide some of the best evidence to date about the causal effects of welfare policies on children.  Morris should also acknowledge that in fact, because New Hope and SSP are included in this body of work, this summary of research goes above and beyond lessons for U.S. welfare policy and more generally applies to employment and income-support policies.

The key finding in this research is that welfare and employment programs that increase parental employment and income had small and consistently positive effects on the developmental outcomes of children who were preschool- and elementary school-aged at study entry. Programs that increased employment but not income had few and inconsistent effects on young children’s development. These benefits of earnings supplement programs appear to be concentrated in school achievement and cognitive test scores and are sustained in the long term. In contrast, these same programs have unfavourable effects on the schooling outcomes of children who were in pre-adolescence and early adolescence at study entry. Further work shows that the small positive effects on achievement primarily occur for those children making the transition into middle childhood, whereas the small negative effects occur among those children making the transition out of middle childhood. These main conclusions are similar to complemented research conducted by the Project on State Level Child Outcomes (led by Child Trends).

That these particular programs, tested via well-implemented rigorous random assignment design, affected children in the ways summarized above is largely undisputable. Because of this, these findings have been widely disseminated and have had an enormous impact on the conceptual thinking behind welfare reform. However, experimental research has its limitations. Since the studies all took place during a particular time and place, the findings might not be applicable to a different context or environment.  In addition, this particular set of studies was not designed to appropriately measure the well-being of infants and toddlers, nor of adolescents (with the one exception of SSP as noted by Kamerman). Yet it is these two groups of children that policy-makers should be increasingly concerned about, given the potentially low quality and instability of child care available for very young children and the unexpected hints of potentially negative effects among adolescents. Finally, the child outcomes collected in these studies are primarily based on maternal reports; though the random assignment design ensures that the effects are not biased by maternal perceptions, “what” these outcomes are actually capturing is a question, particularly with respect to children’s social behaviour. 

Indeed, Kamerman touches upon many of these weaknesses of experimental research in her summary of findings to date, which includes the Morris et al. work as well as research conducted under the Three-Cities Study, a longitudinal study of low-income families in Boston, Chicago and San Antonio that collected detailed observational assessment as well as survey data on families and their preschool- and adolescent-aged children.  Kamerman’s conclusions have a slightly broader focus than those of Morris. In her view, the first major finding from the experimental research is that “there were no or weak impacts [of welfare programs] on children, but what impacts were found included both positive and negative impacts.” Across age groups and across a variety of outcomes, this main conclusion is an important one in that, as Kamerman notes, the detrimental, or “horrible,” effects on children that were expected as a consequence of welfare reform did not occur.  The substantial increases in maternal employment did not translate into equally substantial negative effects on their children.  This is true even for adolescents where the noted negative effects on achievement and grade repetition, for example, are quite small in magnitude. 

Many of Kamerman’s other conclusions are similar to those of Morris with the exception of two.  It is not clear from the experimental studies that a decrease or a neutral change in employment and income can lead to negative effects on children’s outcomes.  Time limits have not resulted in dramatic decreases in family income, possibly because of extensions and because families rely on informal support. The current set of experimental studies has also not clearly shown that increases in mother’s education can benefit children. How very low-income children are faring in general (poorly vis-à-vis their middle-income counterparts), and how transitions into and out of employment might affect very low-income children are two additional and important points that Kamerman makes, drawing on findings from the Three-Cities Study and related projects. 

Implications for Child Development and Policy

These findings pose difficult trade-offs for policy-makers as they grapple with balancing budgets and providing a cohesive and comprehensive social safety net for low-income families and children.  While one potential policy response is to develop strategies to increase both family income and employment, several open questions remain about whether and how increased income should be targeted (e.g. pure transfers or in-kind), about the role of child care for very young children and the role of institutions and social context for adolescents (as suggested in recent ethnographic research) and how these policy responses can be sustained under tight budget constraints.  The next wave of non-experimental research is positioned to tackle these questions, with more experimental studies potentially becoming available. This next wave of research should also focus appropriately on collecting more and better measures of social development, understand how children are affected across their various stages of development from infancy to young adulthood and strive to improve upon the application of developmental theory to inform hypotheses about the effects of welfare and employment policy across domains of children’s development.