Effects of Bill C-113 Disentitlement on UI and Welfare Take-up Rates - August 1995

Effects of Bill C-113 Disentitlement on UI and Welfare Take-up Rates

View Whole Report

• Introduction
• Methodology and Data
• Key Findings
• Conclusions and Issues for Policy
• Biographical Notes

Effective at the beginning of April 1993, Bill C-113 made two changes to Canada's regular unemployment insurance system. For most individuals, benefits were cut from 60 to 57 percent of insurable earnings. However, individuals who either voluntarily quit their jobs without just cause or were dismissed from their jobs had their benefits cut to zero. This study of Bill C-113's effects attempts to answer the following questions:

1. What were the effects of Bill C-113 on the UI take-up rates of job quitters? Did it drastically reduce UI take-up rates and thus disentitle a lot of quitters from benefit receipt, or (because of the "just cause" exemption) did it have only a small effect on the take-up?



2. By what mechanisms did Bill C-113 reduce the claim rates of job quitters? Did fewer quitters apply for UI, were more of the applications by quitters rejected by HRDC and/or did the disentitled workers become re-employed more rapidly than before?



3. What were the implications of Bill C-113 for quitters on welfare take-up rates? Did welfare receipt rise, thus effectively transferring the social welfare burden from the federal to the provincial governments?

Partly in order to examine the effects of Bill C-113 on labour markets, Human Resources Development Canada (HRDC) commissioned a panel survey of individuals separating from their jobs about one month before (January 31 through March 13, 1993) and one month after (April 25 through June 5) the effective date of the Bill. The resulting data set, including merged administrative records pertaining to the concurrent and past use of the UI system, has come to be called the Canadian Out of Employment Panel (COEP) survey.

The sampling frame for the COEP is a 10 percent random sample of the population of individuals receiving a Record of Employment (ROE) form in one of the two sample periods. Canadian employers are required to issue an ROE form whenever a job separation occurs. Aside from ROE's issued for participation in a Work Sharing program, apprenticeship, and retirement at age 65, ROE's for all separation reasons were sampled. Of the approximately six million annual job separations for which ROE's are issued, and of the approximately 12,000 separations sampled in the COEP, about half are for reasons of "short work" as reported by the employer, about 15 percent are "voluntary departures" (VQ's), or quits, while the remainder consist of a wide variety of codes, including 18 percent labelled "other" (the second largest single category of separations). In this paper we focus largely on the VQ group, though we shall occasionally present comparative results for the "short work" and "other" groups. Indeed, since we are interested primarily in the effects of the changes on individuals who quit into unemployment, workers who reported that they "quit to take another job" are excluded from the analysis as well.

One concern in using a comparison of the February and May cohorts of the COEP survey to infer the effects of Bill C-113 on job quitters is the problem of seasonality: perhaps some or all of the differences between these two cohorts are just the normal differences one would observe between individuals leaving jobs in February and May of any year. In this study we adjust for seasonal influences in three alternative ways. First, using administrative data on earlier years, we can compare the 1993 February-May changes in several key variables to those of earlier years, to see whether they are indeed typical or not. Second, we use individuals who separated for reasons other than quitting as a rough control group for quitters. Third, we use regression analysis to control for any observable differences in the characteristics of individuals quitting their jobs in February versus May.

Claim Rates of VQ's

Did quitters' VQ claim rate fall due to Bill C-113? Examining differences in separations occurring before and after the policy change, the VQ claim rate did fall, by about 9.6 percentage points.

Typically, the UI claim rate falls between February and May, and 1993 was an absolutely typical year in this regard. The claim rate of VQ's also falls during this time period, by an average of about 3 percent. 1993 differs from previous years (from 1985) in that the VQ claim rate fell by 6.4 percent, more than double the average decrease in the VQ claim rate.Thus, although part of the February-May 1993 decline in the take-up by quitters was likely due to seasonal factors, some was also attributable to Bill C-113. Overall, however, this decline in UI take-up was not large, falling from about 29 to 19 percent. Thus, although Bill C-113 did disentitle some job quitters from UI receipt, even after its imposition, almost 20 percent of individuals who quit their job claimed Ul.

Source of Decline in Takeup

By what mechanism did C-113 reduce the VQ take-up rates of job quitters? The analysis indicates that by far the most important factor contributing to the decline in UI takeup by job quitters was a "discouragement effect" whereby fewer quitters applied for UI.

In the paper we develop an analytical framework for decomposing the decline in UI takeup into a set of components, each associated with a different mechanism of adjustment to the disentitlement. These mechanisms are a "discouragement effect" (fewer quitters applying for UI), a "rejection effect" (more UI applications by quitters rejected by HRDC), and an "incentive effect" (disentitled individuals induced to become re-employed more quickly as a result of the benefit cut).

Our findings indicate that by far the most important factor contributing to the decline in UI takeup by job quitters was a "discouragement effect", accounting for about 78 percent of the total drop in UI takeup. This drop in takeup was mainly observed among workers who had no re-employment success following their separation: the percentage of this group applying for UI fell from 73 to 49 percent. Further evidence of a discouragement effect is that the percentage of non-applying quitters who said they did not apply because they "believed they were ineligible for benefits" rose from 40 to 54 percent. A much smaller portion of the decrease in takeup was due to a "rejection effect". In our analysis, we find no evidence of any positive incentive effect of the UI benefit cut on reemployment rates of affected workers.

Welfare Takeup and Program Substitution

Did workers who were disentitled from claiming UI under the C-113 provisions turn to welfare as a substitute source of income support? An increase in welfare takeup was observed among workers who had their UI benefits reduced to zero by Bill C-113.

Welfare takeup rates were compared for VQ's before and after the policy change. A small increase in welfare takeup was observed for all groups between the two time periods, regardless of UI receipt or re-employment success. However, for the households of job quitters who had no re-employment success after separation we observe a much larger increase in welfare takeup. Before the policy change about 18 percent of non re-employed VQ's received welfare compared to 28.6 percent after, a difference of 10.6 percent. Thus, at least for the subgroup of workers most affected by the quit provisions of Bill C-113, the loss of eligibility for UI appears to increase their use of welfare.

This study has two main implications for policy formation. First, policymakers should be aware of the relatively small effect that the voluntary quit provisions of Bill C-113 had on UI takeup rates. Probably because of the long list of reasons for justified quits, only a fairly small decline in the UI takeup rate of job quitters ensued. Thus, the policy change did not lead to a major cost saving to the UI system.

Second, the paper shows that there are measurable interactions between Canada's UI and welfare systems: restricting UI access for certain workers led to an increase in welfare takeup among those workers after the policy change. It seems clear that estimates of the effects of further cutting the UI program on provincial welfare expenditures need to take this into account.

Tom Crossley is a Ph.D student at McMaster University, specializing in labour economics.

Peter Kuhn received his Ph.D. in economics at Harvard University in 1983. Since then he has taught economics at the University of Western Ontario, Princeton University, and McMaster University, where he is currently professor of economics. Professor Kuhn's area of specialization is labour economics, including the issues of trade unionism, discrimination, immigration, displaced workers, and unemployment insurance. He is currently the program director of the Canadian Employment Research Forum, sits on the editorial board of the international journal Labour Economics, and is a former member of the editorial board of the Canadian Journal of Economics.