Directorates
·	Director General's Office
·	Data Development
·	Strategic Evaluations
·	Partnership and Continuous Evaluations
Publications
·	Evaluation Reports
·	Evaluation Briefs
·	Labour Market Development Agreement Series
·	Lessons Learned Series
·	Evaluation Tool Kit Series
·	Evaluation Forum Newsletter
·	From Theory To Practice

• Introduction and Background
• Purpose
• Key Findings
• Biographical Notes

Introduction and Background

Unemployment Insurance (UI) and social assistance (SA) are the two most important income security programs for Canada's working age population. In many respects, UI and SA (or welfare) are fundamentally different programs. The UI program's primary objective is the provision of insurance to labour market participants for the temporary loss of income due to unemployment. UI is not a universal income support program; it specifically excludes all individuals who are not in the labour force or do not have the requisite labour force attachment, who are self employment or who have exhausted their benefit entitlement. In addition, benefits paid under the program are based on previous earnings and benefits have a limited duration.

The objective of the Canadian welfare programs is to provide financial assistance to all individuals and families in need, irrespective of the causes of the hardship. Unlike UI, welfare is a universal program. Eligibility for the receipt of welfare is based on a ""needs"" test and is not tied to previous employment; benefits are based on the household's assessed budgetary deficit and not related to prior earnings; nor is there a limit on the length of time a person may receive benefits.

There is an established literature examining the use and behavioural effects of UI in Canada. Recently, researchers have begun describing and analyzing how individuals and families interact with the welfare programs. However, there is very little work which directly examines the interaction between the UI and SA programs.

There are several avenues through which the UI and SA programs may interact. Firstly, SA is the income security program of last resort. Individuals in need and who are ineligible for benefits under other targeted programs my turn to SA for assistance. Furthermore, individuals who do not obtain employment by the time their UI claim expires may turn to SA for income support. As a consequence, changes to UI that restrict eligibility or decrease the generosity of the program may shift some people from UI onto SA.

Secondly, changes over time in the economy and the labour market may contribute to SA becoming a form of UI for segments of the workforce. For instance, the 1990's have seen the collapse of the youth job market and an increase in the incidence of long-term unemployment among young workers. These developments may contribute to new labour market entrants relying relatively more on SA, rather than UI, than previous generations.

Administrative practices are a further source of interaction between the programs. In particular, in the late 1980's significant backlogs developed in the processing of UI claims. As a result, the processing time for a UI claim was substantially longer than the 2 week waiting period. This was a cause of hardship to many individuals who turned to SA while their UI claim was processed. Consequently, there was an increase in the "UI pending" SA caseload which was purely a result of the administrative features of the UI program.

Purpose

Give the potential interactions between the programs, the purpose of this study is to examine the extent to which these interactions occur in practice. One important policy question is whether UI and SA serve essentially separate clienteles. To examine this question, we compare the characteristics and experiences of the group of people who rely on both programs to the characteristics of the people who participate in either program.

After examining descriptive information on program participants, the next phase of the research is to analyze specific forms of interaction between the programs. In this study, the hypothesis that the duration of UI receipt is influenced by an individual's previous use of welfare is tested. Future work will examine the impact of the UI reforms introduced in November 1990 on shifting potential UI users onto SA. Additionally, interprovincial variation in SA program parameters will be used to identify the impact of SA program generosity on individuals' relative use of SA and UI.

Data and Methodology

The empirical work uses Human Resources Development Canada's (HRDC) UI Status Vector File, derived from the administration of the UI program, matched with a corresponding Welfare File derived from the administration of the SA programs in British Columbia over the period September 1986 - December 1992. Both data sets are 10 percent random samples of the case records of the participants of both programs, drawn identically (based on the masked Social Insurance Number) so that individuals may be linked across the two files. A key advantage of these data is that they provide a detailed and accurate history of an individual's use of UI and SA over an extended period of time. Additionally, the data include a fine level of time aggregation, corresponding to the time scale by which the programs are administered.

The study proceed by matching the UI and SA files and:

• generating summary statistics for the population of individuals and families in B.C. who began a spell of welfare recipiency in 1986, 1989 and 1992. Recipients are then grouped according to whether they had a UI claim open during the same year, and descriptive statistics are generated for each of the subgroups.

• generating summary statistics on the UI claims initiated in B.C. in 1986, 1989 and 1992. The set of UI claims are divided into three groups: claims where the claimant had an SA spell overlapping with the UI claim period; claims where the claimant began a welfare spell immediately on termination of the UI claim; and those claims not associated with a concurrent or subsequent welfare spell. This division of the joint UI-welfare population is intended to distinguish the ""UI pending" from the "UI exhaustee"" group of welfare recipients.

• after presenting general descriptive information on the set of people who use both UI and SA in a limited period of time, one specific interaction between the programs is examined in more detail. The influence of an individual's history of SA receipt on subsequent UI spell durations is tested within a hazard function framework. For this analysis, the UI sample is restricted to regular claims by males between 1987 and 1992.

Key Findings

1. There is a substantial overlap in the clientele of the UI and SA programs in B.C., and the extent of this overlap increased over the 1986-1992 period under study.

2. The majority of welfare recipients in 1986 were employable (78 percent), were single (65 percent) and were male (61 percent). Lone parent families accounted for 18 percent of the welfare spells which commenced in 1986. The average duration of welfare spells was approximately 14 months. One-third of the welfare recipients also had a UI claim open in 1986. This ""UI history"" group were disproportionately younger, make, single or part of a two-parent family, employable and UI pending. This group also tended to be more mobile, with the average welfare spell being 8.9 months.

3. Over the 1986-1992 period, the number of welfare recipients increased (by 50 percent), as did the proportion in the ""UI history group"" (to 46 percent). By 1992, 86 percent of all recipients were classified as employable and 14 percent were UI pending; of the ""UI history"" group, 92 percent were employable and 39 percent were UI pending.

4. Examining the sample of UI claims, in 1986 approximately 10 percent were in one of the ""UI-Welfare"" groups. The concurrent UI-welfare users tended to have longer UI spells and below average weeks of insured employment and insured weekly earnings. The group of UI-subsequent-welfare users were more likely to be UI exhaustees, and in cases where benefits were actually paid, this group had the longest claim period and benefit payments durations.

5. Over the 1986-1992 period, the number of UI claims increased as did the proportion in the UI-welfare groups. By 1992, 17 percent of claimants were in one of the joint UI-welfare beneficiary groups.

6. In testing for the impact of prior welfare receipt on the duration of UI claims, the individual's maximum potential duration of benefits was constructed. We find that individuals with a recent welfare spell tend to have longer UI spells (by about 2 weeks). However, after controlling for maximum potential duration of UI benefits, they exhibit very similar exit rates from UI and had very similar incidences of UI exhaustion as the non-welfare population.

Biographical Notes

Garry Barrett is a PH.D student in Economics at the University of British Columbia. Previously, he completed an honours B.A. (Economics) at the University of Sydney Australia, and an M.A. (Economics) at the University of British Columbia. His primary areas of research have included industrial relations, equilibrium business cycle theory, earnings distribution, and UI-SA interactions.

Denise Doiron completed a B.A. at l'Université de Moncton, an M.A. at the University of British Columbia, and a Ph.D at the University of British Columbia (1987). She has held appointments as Assistant Professor at the University of Western Ontario and the University of British Columbia, and is moving the Department of Economics, University of Sydney, Australia, in January 1995. Areas of teaching and research include industrial relations, labour economics and labour market policy.

David Green is an Assistant Professor of Economics at the University of British Columbia. He completed his Ph.D at Stanford University in 1990. His main areas of research interest are the effects of Unemployment Insurance in the labour market, the impact of immigration policy and the adaptation of immigrants to the Canadian economy. More recently, he has also worked in the area of earnings and income distributions.

W. Craig Riddell is Professor and Chair of the Economics Department at the University of British Columbia. He has published extensively in scholarly and professional journals, mainly research in the areas of industrial relations, labour economics and public policy. He was the research co-ordinator for Labour Markets and Labour Relations for the Royal Commission on the Economic Union and Development Prospects for Canada, 1983-1985 (Macdonald Royal Commission). Professor Riddell is currently the academic co-chair of the Canadian Employment Research Forum.

"The Interaction of Unemployment Insurance and Social Assistance" by Garry Barrett, Denise Doiron, David Green and Craig Riddell, is work in progress which will examine UI-SA relationships for five provinces. This background study is in preparation for publication by Human Resources Development Canada as an Insurance Program evaluation report, 1994.