Preface

Social protection systems are intended to support households in financial difficulties, a role that has been underlined during the recent “Great Recession” in many countries around the world. Over the period 2007–2011, social transfers paid in cash increased by more than 2 percentage points of GDP in a majority of OECD countries. ¹ Patterns of transitions into and out of benefit receipt are key measures of the effectiveness of social protection systems. Increases in economic need should be reflected in an increase in inflows to receipt, but benefit receipt should not hinder and should ideally support a swift return to self-sufficiency. Means-tested social assistance (SA) or “welfare” and other minimum-income benefits are of particular interest as the stakes of finding a suitable policy configuration are especially high. On the one hand, those requiring support typically have little or no other resources to fall back on and may suffer acute poverty if minimum-income support is difficult to access. On the other hand, benefit levels tend to be well below commonly used poverty thresholds, and long spells of benefit receipt are therefore a social concern as well as being costly for governments.

Almost all OECD and EU countries operate comprehensive means-tested benefit programs for working-age individuals and their families. These benefits have a major role as “social protection floors,” either providing last-resort income support alongside primary income-replacement benefits or acting as a principal instrument for delivering support. As anti-poverty measures, they reduce income disparities at the bottom of the income range and, as such, are important redistribution policies. Equally important, they provide safety nets for individuals experiencing low-income spells and so help smooth income levels over time. Government spending on such benefits is a significant item in the public finances of OECD and EU countries. As with other components of social spending, the aftermath of the Great Recession has created not only greater demand for social support but also increased pressures to reduce or control spending.

The nature of last-resort benefits has been changing over time, notably as a consequence of the “welfare to work” reforms introduced in countries such as the United States, the United Kingdom, Australia, New Zealand, and Germany. These reforms may affect the extent of income support that these benefits can provide, the transitions into and out of benefit receipt, and the levels, trends, and composition of social spending. Social assistance benefit receipt is therefore at the core of social policy concerns.

This volume presents new results on the dynamics of social assistance, minimum-income and related out-of-work benefits in a range of different country contexts. It contains eight articles, seven of which were presented at the IZA/OECD/World Bank Conference on “Safety Nets and Benefit Dependence” in Paris in May 2013. ² The articles shed light on benefit spell durations, the movements into and out of receipt of safety net benefits, the individual or family characteristics associated with these movements, the extent of state dependence or “scarring,” and the interaction of various welfare programs. The results establish an evidence base for an informed policy debate in a range of OECD countries. They also provide methodological background for future work on benefit receipt patterns.

One article is methodological, evaluating alternative approaches to differentiate genuine state dependence from spurious correlations in binary outcomes using dynamic panel data models; three articles focus on the dynamics of social assistance receipt in various countries using the dynamic random effects probit version of these models; two articles deal with transitions into and out of social assistance emphasizing the role of public employment services and activation measures; and finally two deal with welfare program interactions. All articles utilize longitudinal data and several use register data linked to surveys.

We are pleased to publish as our lead article a previously unpublished “classic”. The 2000 working paper by Kenneth Y. Chay and Dean R. Hyslop on “Identification and estimation of dynamic binary panel response models” was a pioneering and influential piece of research that has been much cited ever since, including by several of the articles in this volume. Chay and Hyslop examine the roles of sample initial conditions and unobserved individual effects in consistent estimation of the dynamic binary response panel data model. Many different specifications are considered. These include alternative random effects models in which the conditional distributions of both the unobserved heterogeneity and the initial conditions are specified, and fixed effects conditional logit models that make no assumptions on either distribution. In their revision of their paper for this volume, Chay and Hyslop now include models with the “Wooldridge” treatment of initial conditions (also used by other papers in the volume). The models are fitted to longitudinal data on welfare and labor force participation for female respondents to the US Survey of Income and Program Participation (SIPP). There are several findings. First, the hypothesis that the sample initial conditions are exogenous is rejected by both samples. Misspecification of the initial conditions drastically overstates estimates of the state dependence and understates estimates of the short- and long-run effects of children on labor force participation. The fixed effects conditional logit estimates are similar to the estimates from the random effects model that is flexible with respect to both the initial conditions and the correlation between the unobserved heterogeneity and the covariates. For female labor force participation, there is evidence that fertility choices are correlated with both unobserved heterogeneity and pre-sample participation histories.

Between 1991 and 2005, the United Kingdom recorded a falling SA receipt rate, driven by a decline in the SA entry rate, while the SA exit rate changed much less. In the article “The Dynamics of Social Assistance Benefit Receipt in Britain,” Lorenzo Cappellari and Stephen P. Jenkins analyze the dynamics of social assistance benefit receipt in Britain during this period using a multivariate dynamic random effects probit model applied to British Household Panel Survey data. The model is used to derive year-by-year predictions of aggregate SA entry, exit, and receipt rates. The analysis highlights the importance of the decline in the unemployment rate over the period and other changes in the socioeconomic environment including two reforms of the income maintenance system in the 1990s. It also illustrates the effects of self-selection (“creaming”) on observed and unobserved characteristics.

As in the United Kingdom, Canada also saw falling SA participation rates since the mid-1990s and modestly increasing benefit receipt rates in the years following the recent Great Recession. In the article “State Dependence in Social Assistance Receipt in Canada,” Jorgen Hansen, Magnus Lofstrom, Xingfei Liu, and Xuelin Zhang use longitudinal data from the Canadian Survey of Labour and Income Dynamics for the years 1993–2010 to document that there are substantial provincial differences in social assistance participation in Canada with higher participation rates in the Eastern part of the country. Using dynamic random effects probit models, the authors find that there is significant state dependence in social assistance, even after controlling for endogenous initial conditions and unobserved heterogeneity. The level of state dependence varies across provinces and might be associated with the generosity of provincial benefit systems.

Unlike the United Kingdom and Canada, Germany experienced relatively stable SA participation patterns in the 1990s, but thereafter there was a steep upward trend which peaked in 2006. As in the United Kingdom during the 1990s, Germany also implemented a major reform in the provision of social assistance receipt benefits in the mid 2000s (the “Hartz” reforms). In the article “State Dependence in Social Assistance Benefit Receipt in Germany before and after the Hartz Reforms,” Sebastian Königs studies state dependence in social assistance receipt in Germany using annual survey data from the German Socio-Economic Panel for the years 1995–2011. As in the previous two articles, he estimates a series of dynamic random effects probit models controlling for endogenous initial conditions and unobserved heterogeneity. The estimates suggest substantial structural state dependence in benefit receipt, which is higher in Eastern Germany than in Western Germany. The results indicate that state dependence was similar before and after the Hartz reforms, and the article discusses possible reasons for this finding.

Participation rates vary not only over time but also within countries. In the article “How Do Exit Rates from Social Assistance Benefit in Belgium Vary with Individual and Local Agency Characteristics?,” Sarah Carpentier, Karel Neels, and Karel Van den Bosch exploit the decentralization of the administration of social assistance benefits in Belgium to examine how much of the variation in spell lengths of benefit receipt is associated with differences in policies across agencies. Using administrative record data they analyze the monthly hazard of benefit exit for an inflow sample of benefit recipients across 574 welfare agencies. Controlling for differences in both the observed and unobserved characteristics of beneficiaries and local agencies, the authors find higher exit rates from social assistance receipt in bigger municipalities and in agencies which provide more generous cash-based supplementary assistance. They also find strong evidence of shorter benefit episodes in agencies where active labor market program participation rates are higher.

Activation of welfare recipients through job preparation and job search assistance programs, or through “workfare” in publicly provided jobs have been introduced in several countries in the past decades as a way to enhance job search incentives, work opportunities, and employment outcomes. In the article “The Effects of Mandatory Activation on Welfare Entry and Exit Rates,” Anna Persson and Ulrika Vikman use register data on the entire population of Stockholm municipality in Sweden to examine the effect of mandatory participation in active programs on benefit receipt as well as on the transitions into and out of SA. Exploiting the sequential implementation of mandatory activation across the districts of Stockholm, the authors find a negative effect of activation on the SA entry rate for young and single individuals, and a positive effect on the SA exit for singles, suggesting a higher response among those with fewer family responsibilities.

Temporary assistance programs may create job search disincentives and may also influence application to other programs, which are designed for those with longer term needs such as disability insurance. In the article “The Impact of Temporary Assistance Programs on Disability Rolls and Re-Employment,” Stephan Lindner and Austin Nichols examine whether participation in four temporary assistance programs in the United States influences application for disability insurance or Supplementary Security Income and re-employment. Instrumenting temporary assistance participation using variation in policies across states and over time, the authors find evidence suggesting that increased access to unemployment insurance benefits reduces disability insurance applications, increases the probability of return to work, and reduces the probability of claiming supplementary benefits.

In the last article “What Impact Does Old-Age Pension Receipt Have on the Use of Public Assistance Programs among the Elderly?,” Norma B. Coe and April Yanyuan Wu examine the reasons behind the low take-up rate in means-tested programs in the United States among the elderly. By exploiting the interaction between old-age pension benefits and public assistance programs to instrument for the expected public assistance benefit level, they estimate the causal impact of benefit levels on the take-up of means-tested programs. The authors find that the low take-up among the elderly is not driven by changes in the composition of the eligible pool, that old-age pensions decrease the take-up of public assistance programs by decreasing the gains of participation, and that eligible individuals who begin receiving old-age pensions continue to participate in programs providing cash more often than in programs offering in-kind transfers. This different program response suggests that the elderly may have preference for cash over in-kind transfers.

As with past volumes, we aim to focus on important issues and to maintain the highest levels of scholarship. We encourage readers who have prepared manuscripts that meet these stringent standards to submit them to Research in Labor Economics (RLE) via the IZA website (http://rle.iza.org) for possible inclusion in future volumes.

Stéphane Carcillo

Herwig Immervoll

Stephen P. Jenkins

Sebastian Königs

Konstantinos Tatsiramos

Editors