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

This section uses Canadian Out of Employment Panel (COEP) and administrative data from the last quarter of 1995 to identify new entrants. One of the main objectives is to estimate the number of regular beneficiaries who have 20 to 25 insured weeks and who, under Bill C-12, were negatively affected from the increase of the entrance requirement from 20 to 26 weeks.

2.1 Methodology

The starting point of the analysis was to identify who, among those who terminated a job in the last quarter of 1995 and claimed Employment Insurance (EI), were new entrants. Also, identify who would not have been able to qualify for benefits, had the higher entrance requirement under Bill C-12 already been in effect. The analysis focuses on those who:

1. terminated a job in the last quarter of 1995; and
2. received a Record of Employment (ROE) after they terminated their job (regardless if they consequently claimed EI or not).²

The main methodological challenge faced by the study was reconstructing the employment history of individuals over a two-year period and determining who was a new entrant. ROE and Status Vector records were used to estimate the following:

1. the qualifying period (52 weeks prior to the benefit commencement period or the period since the start of the previous claim, which ever was shorter);
2. the number of weeks of insured employment during the qualifying period;
3. the pre-qualifying period (the 52 week period prior to the qualifying period); and
4. the number of weeks of insured employment and benefit weeks during the pre-qualifying period (individuals are defined as new entrants if the weeks of insured employment and EI benefits during the pre-qualifying period were less than 14).

The methodology used to estimate insured weeks from the ROE and Status Vector records is complex, because often job spells overlap and/or contain weeks of non insured employment. The basic principle used in the logic of the algorithm used by this study was to allocate weeks in such a way as to give the benefit of doubt to the claimant. This is the same principle applied by field staff in determining EI eligibility.³

Since in the case of EI claimants the actual number of insured weeks of employment is known from their Status Vector, it was possible in these cases to test the methodology of estimating insured weeks, by comparing the study estimates to the number of insured weeks in the Status Vector. Chart 1 shows the frequency distribution of beneficiaries by the difference between estimated and actual number of insured weeks: in 93 percent of the cases the difference was more or less 2 weeks. This suggests that the methodology employed here approximates closely to the process followed by EI field staff. Of course, no such direct test is possible in the case of non-beneficiaries or for estimates of insured weeks and benefit weeks during the pre-qualifying period.

2.2 Number of New Entrants

The above methodology led to the following estimates (Table 1):

1. 645,941 individuals terminated a job in the last quarter of 1995;
2. 370,248 of them (57.3 percent) received regular benefits (an additional 41,968 individuals, or 6.5 percent, received special EI benefits);⁴
3. 70,287 of all regular beneficiaries (19.0 percent) were new entrants — i.e. had less than 14 weeks of insured employment or weeks of EI benefits during the pre-qualifying period;⁵
4. 26,273 of the new entrant regular beneficiaries (7.1 percent of all regular beneficiaries) had 20 to 25 insured weeks; it is this group of claimants that would have been disqualified from benefits if Bill C-12 were already in effect (Chart 2), assuming no behavioural change.

Footnotes

2	Of those who claimed benefits after their job was terminated in the fourth quarter of 1995, 80 percent started their benefit period within the same quarter.
3	Example: suppose an individual had two jobs that covered the same 10 week period. In this case, the weeks are staggered and the individual is credited with 10 insured weeks. The logic is more complicated when ROEs overlap partially.
4	76 percent of the regular beneficiaries started their benefit period within the quarter, while the rest started in 1996.
5	Prequalifying period refers to the 52-week period preceding the qualifying period.