11. Women's Access to EI Benefits

11.1 Executive Summary

Under both the new Employment Insurance (EI) system and the old Unemployment Insurance (UI) system, women face the same rules as men. However, although governed by the same rules, a larger proportion of working women may be unable to qualify for benefits because, on average, women are more likely to work part-time.

This report examines the job and unemployment experiences of women, with a focus on their use of EI Part I regular benefits. The report then investigates the extent to which the 1996 reform of EI led to changes for women in terms of EI eligibility, the receipt of EI and the extent of EI benefit entitlements.

Data and Methodology

This study uses information from the Canadian Out-of-Employment Panel (COEP) survey, which provides data on the job termination experiences of individuals. Important information on socio-economic conditions and other personal and employment-related information are available from the survey that enable the development of statistics that describe the experiences of women. It is also possible to link the results of this survey to EI administrative data to allow for estimates of EI receipt.

Main Findings

• Of those who lost a job during the four quarters preceding the 1996 EI reform or during the four quarters following the reform, and who had at least two consecutive weeks of unemployment, 45 percent were women. This excludes those who separated from a job due to retirement, a return to school, pregnancy or parental, or injury or illness.
• Women were more likely to experience longer spells of unemployment than men. For women with children, the average duration of unemployment was even longer.
• There was no significant difference in the rates at which men and women received EI in the eight quarters examined.
• EI eligibility rates for women were affected by the 1996 EI reform, as women were eight percentage points less likely to have had enough hours of work to qualify for EI following the reform than they had been before the reform. Men were unaffected.
• Statistical estimation results indicated that EI reform had no significant impact on the likelihood that a woman or man would collect regular EI benefits.
• EI reform did not have an impact on weeks of entitlement to EI benefits.

11.2 Introduction

Employment Insurance (EI) benefits are becoming increasingly important to women and their families, as women make up a growing percentage of the labour force. Access to the EI system is of critical importance for these workers. Working women require financial assistance that will help bridge the income gap when they are between jobs.

This paper examines the role of Employment Insurance (EI) Part I benefits for women. Under both the new Employment Insurance (EI) system and the old Unemployment Insurance (UI) system, women face the same rules as men. However, although governed by the same rules, a larger proportion of working women may be unable to qualify for benefits because, on average, women are more likely to work part-time.

This report first examines the job and unemployment experiences of women. In particular, it studies their use of EI Part I regular benefits. The report then investigates the extent to which the 1996 reform of EI led to changes in EI eligibility, EI receipt and the entitlements of women.

11.3 Data and Methodology

The Canadian Out-of-Employment Panel (COEP) survey, which is a survey of persons who had a job separation (e.g. people who were laid off or left work for any other reason), formed the basis of the analysis for this study. A one-year period before EI reform (95Q3 — 96Q2) and a one year period after EI reform (97Q1 — 97Q4) were used. The period during the phase-in of EI reform (96Q3 — 96Q4) was omitted, as it represented a period of transition.

The COEP survey is administered on behalf of Human Resources Development Canada (HRDC) by Statistics Canada. The survey collects information on sampled individuals who experienced a job separation as recorded on their Record of Employment (ROE). Collected information includes:

• Personal and household characteristics;
• Reasons for job separation;
• Detailed employment history;
• Job search activities;
• Receipt of EI/UI benefits and/or social assistance; and
• Information on household finances, assets and liabilities.

Each survey participant was interviewed twice, with the first interview occurring within one year of the job separation and the second interview taking place about nine months after the first interview. Almost 33,000 Canadians who had a change or an interruption in their employment activity were surveyed during the eight quarters that are examined.

The group of Canadians who are counted in each of these quarters is referred to as a "cohort". The eight cohorts are grouped into two periods:

• Pre-EI Reform — Participants had a job separation prior to EI reform (95Q3 — 96Q2).
• Post-EI Reform — Participants had a job separation after EI reform (97Q1 — 97Q4).

No analysis is conducted on the two-quarter period during EI reform, as the implementation of the reform was not complete and the analysis would be complex.

For the purposes of this study, the pre-EI reform period is compared to the post-EI reform period. The changes noted in this report may be due to the move to the EI system. However, it is important to keep in mind that the observed changes may also be due to changes in the economy, patterns of work behaviour, and a variety of other factors that would make it difficult, if not impossible to control for, in a comparative analysis like this one. Further studies would be needed to investigate and assess possible alternative causes of some of the observed changes that have taken place since moving to the EI system.

The analysis focuses on individuals who lost a job during the pre- or post-EI reform period due to reasons other than retirement, return to school, injury or illness, or pregnancy or parental responsibilities. Only respondents with at least two consecutive weeks of unemployment were kept for the analyses. The main reason for this restriction is the required two-week waiting period for qualifying for benefits that was in effect over all periods spanned by the data. After applying these restrictions, the experiences of more than 18,000 people form the basis for this study.

11.4 Profile of Women Leaving Employment

This section provides a contextual background describing the major demographic characteristics of women who leave employment. The section begins with a statistical profile of the women themselves. Following that, the job and unemployment experiences of women are investigated, along with a comparison of EI receipt rates between men and women.

11.4.1 Demographic Characteristics

Averages for selected characteristics of the COEP survey respondents are given in Table 1. The characteristics examined include age, marital status, highest attained education level, area type (urban/rural), and region.

Table 1 Selected Demographics of Individuals With a Job Separation (percent)

	Men	Women
		All	Married With Children	Not Married With Children	No Children
All	55.0	45.0	13.3	4.7	27.1
Age
Youth (15-24)	19.1	15.5	4.7	24.0	19.3
Prime (25-54)	71.7	76.1	94.4	74.3	67.5
Older (55+)	9.2	8.4	1.0	1.8	13.2
Marital Status
Single	46.1	39.1	0.0	100.0	47.7
Married	53.9	60.9	100.0	0.0	52.3
Education
Less than High School	29.6	18.2	15.2	20.4	19.2
High School	28.3	25.8	32.1	26.1	22.6
More than High School	40.0	54.3	51.4	52.7	56.0
Other	2.1	1.8	1.2	0.9	2.2
Area Type
Rural	27.3	24.8	29.1	20.1	23.6
Urban	72.7	75.2	70.9	79.9	76.4
Region
Atlantic	11.1	10.3	11.4	11.2	9.5
Quebec	31.4	27.6	20.6	29.4	30.7
Ontario	27.8	33.8	40.8	26.2	31.8
Prairies	15.8	15.8	16.5	16.1	15.3
British Columbia	13.8	12.6	10.8	17.1	12.7
Sample Size	18,043
Source: COEP Survey of Job Terminations 95Q3 - 96Q2 and 97Q1 - 97Q4.

The second column of Table 1 indicates that 45 percent of COEP respondents included in the sample that was used for this study were female.⁸⁶ In comparison to men, females with a job separation were less likely to have been a youth or single and more likely to have resided in Ontario and completed a higher level of education. Variations in the remaining categories were minimal between the two genders.

Within the three subcategories for women (married with children, not married with children, no children) there were some significant variations. Women with no children or women who were not married and had children were far more likely to have been in the youth age category than women who were married with children. Women who were not married and had children were also more likely to have resided in an urban area. Finally, relative to all women, women who were married with children were more likely to have resided in Ontario while women who were not married with children were more likely to have resided in British Columbia.⁸⁷

11.4.2 Employment History Characteristics

Table 2 analyzes the employment history of individuals with a job separation, providing information related to industry as well as certain job characteristics.

Table 2 Employment History of Individuals With a Job Separation (percent)

	Men	Women
		All	Married With Children	Not Married With Children	No Children
Industry
Agriculture	2.1	2.2	2.3	4.8	1.8
Primary	6.8	1.4	1.6	0.8	1.5
Manufacturing	20.4	11.8	12.6	12.7	11.3
Construction	19.3	1.4	1.0	2.5	1.5
Transportation & Storage	6.0	3.4	4.2	2.7	3.2
Communications	1.0	1.3	1.1	1.0	1.4
Other Utilities	0.7	0.4	0.3	0.4	0.4
Trade	14.3	18.4	17.6	17.1	19.0
Finance	1.7	4.7	6.0	4.7	4.0
Education	7.9	26.0	26.0	20.5	26.9
Business Services	11.1	19.4	17.0	23.8	19.8
Public Services	2.8	3.3	3.0	4.0	3.3
Public Administration	5.9	6.3	7.3	5.1	6.0
Job Characteristics
Seasonal	22.0	14.5	14.6	12.3	14.8
Part-Time	7.7	25.5	29.3	23.8	23.9
Union	18.8	14.6	12.7	8.9	16.5
Weekly Hours of Work1	44.7	35.7	34.7	36.1	36.1
Weeks on Job1	173.2	205.0	183.1	117.9	230.4
Sample Size	18,043
Notes: 1. Figures are reported in level terms. Source: COEP Survey of Job Terminations 95Q3 - 96Q2 and 97Q1 - 97Q4.

Females with a job separation were concentrated in education (26 percent), business services (19.4 percent) and trade (18.4 percent) industries. Men were more likely to have been employed in manufacturing (20.4 percent) and construction (19.3 percent) industries. There was also variation by industry among women. Women who were not married and had children were far more likely to have been employed in business services or agriculture and less likely to have been working in education than women in the other two subcategories.

In terms of job characteristics, men were more likely to have been unionized or employed in seasonal jobs. On the other hand, women were far more likely to have been employed part-time. Women worked more weeks while they were employed, but men worked nine more hours per week. There was significant variation in terms of the number of weeks of employment among women. Women who were not married and had children were, on average, employed for 117.9 weeks. This differed greatly from women in the other two subcategories (183.1 weeks and 230.4 weeks). The presence of children in the household clearly had a link to a woman's job tenure. This result may also be due to the fact that women with children who are not married are far more likely to be in the youth age category. Because they are younger, they are less likely to have worked for any great length of time at a specific, individual job.

11.4.3 Unemployment History Characteristics

Table 3 examines the unemployment history of individuals with a job separation, namely the number of weeks of unemployment and the reliance upon regular EI benefits and social assistance.

The first section of Table 3, the weeks of unemployment, indicates women were slightly more likely to have had short unemployment spells of between two and ten weeks. However, women were noticeably more likely to have experienced long-term unemployment (i.e. being unemployed for more than 52 weeks). Almost 25 percent of women were long-term unemployed whereas 17.7 percent of men were long-term unemployed. Women who were not married and had children were the most likely to have been long-term unemployed (27.5 percent).

There was almost no difference between the EI receipt rates of men and women who had a job separation lasting at least two weeks and reported being unemployed for that period (45.1 percent vs. 45.7 percent). Among women, the highest rate of EI receipt was for those who had no children (46.9 percent).

There was also little difference between men and women in the number of weeks on an EI claim and in the number of weeks of EI collected.⁸⁸ Women were more likely to have exhausted an EI claim by using up all of the weeks of entitlement (35.4 percent vs. 30.3 percent). On the other hand, men were more likely to have exhausted an EI claim by simply letting the benefit period expire, even though not all of the entitlement weeks were used (18.7 percent vs. 13.4 percent). These differences are a result of women having longer unemployment durations and, hence, being more likely to remain on EI for the duration of the claim. Gender differences in entitlement was not an explanation, as there was little difference in the number of entitlement weeks, both for all unemployment durations and for those unemployed for more than 52 weeks.⁸⁹

Finally, there was little difference between the two genders with respect to the receipt of social assistance, although women who were not married and had children were far more likely to have received social assistance than the other two groups of women.⁹⁰

Table 3 Unemployment History of Individuals With a Job Separation (percent)

	Men	Women
		All	Married With Children	Not Married With Children	No Children
Weeks of Unemployment
2 to 10 Weeks	33.8	35.5	35.1	29.0	36.8
11 to 20 Weeks	19.3	14.6	13.7	15.4	14.9
21 to 30 Weeks	14.4	11.8	9.6	15.2	12.3
31 to 40 Weeks	10.1	9.6	9.4	8.9	9.7
41 to 52 Weeks	4.7	4.0	5.0	4.0	3.5
More Than 52 Weeks	17.7	24.6	27.2	27.5	22.8
Collected Regular EI Benefits	45.1	45.7	43.6	44.2	46.9
Weeks on Claim1	34.2	33.4	32.8	33.0	33.8
Weeks Collected1	22.1	22.8	22.8	23.8	22.7
Exhausted Claim (1)2	30.3	35.4	40.0	32.5	33.7
Exhausted Claim (2)3	18.7	13.4	10.0	12.3	15.1
Received Social Assistance	8.8	8.5	3.9	32.7	6.5
Sample Size	18,043
Notes: 1. Figures are reported in level terms. 2. Claims where all entitlement weeks were used. 3. Claims where benefit period ends before entitlement is exhausted. Source: COEP Survey of Job Terminations 95Q3 - 96Q2 and 97Q1 - 97Q4.

11.4.4 EI Receipt

The percentage of men and women with a job separation receiving regular EI benefits is provided in Table 4. Column three in Table 4 provides the t statistic, which determines whether the EI receipt rates between men and women are significantly different.⁹¹ When examining Tables 4 through 6, it is important to note that the "N" column indicates the number of sampled individuals in the first two columns. This information provides an informal measure of sample reliability.

During the eight quarters analyzed, there was no significant difference between men and women in the EI receipt rates of regular benefits.⁹² However, there were some significant differences across some of the categories. Female youths were far less likely to have collected EI than their male counterparts, as were married women with children and women from the Prairies. Married women without children and women from British Columbia were more likely to have collected EI than married men with children and men from British Columbia.

Table 4 Percentage of Individuals With a Job Separation Receiving Regular EI Benefits by Selected Demographics (percent)

	Men	Women	t statistic	N
All	45.1	45.7	0.38	10,393
Age
Youth (15-24)	32.4	22.0	-3.45	1,822
Prime (25-54)	47.9	49.6	1.00	7,575
Older (55+)	49.8	53.6	0.78	996
Family Type
Single with Children	37.3	44.2	1.43	563
Single without Children	40.9	40.3	-0.27	3,937
Married with Children	50.2	43.6	-2.52	3,030
Married without Children	47.8	53.1	1.99	2,858
Education
Less than High School	50.5	48.6	-0.65	3,544
High School	44.4	46.4	0.73	2,957
More than High School	41.8	44.4	1.22	3,635
Other	42.2	42.3	0.01	257
Area Type
Rural	50.7	48.1	-1.03	3,816
Urban	43.0	44.9	1.08	6,577
Region
Atlantic	51.9	53.2	0.78	3,477
Quebec	51.3	50.7	-0.20	1,325
Ontario	39.1	42.6	1.11	1,064
Prairies	39.6	35.9	-1.97	3,219
British Columbia	44.1	48.9	1.71	1,308
Source: COEP Survey of Job Terminations 95Q3 - 96Q2 and 97Q1 - 97Q4.

11.5 The Impact of the 1996 EI Reform

The degree of support provided to women by EI Part I will be examined in terms of the percentage of those experiencing a job termination who qualified for EI, the percentage who received EI, and the maximum number of weeks that they were entitled to collect EI if they did qualify.

11.5.1 Impact of the 1996 EI Reform on EI Eligibility

The percentage of women with a job separation with enough weeks/hours to qualify for EI is given in Table 5, both before and after the 1996 EI reform.

Table 5 Percentage of Women with a Job Separation with Enough Weeks/Hours to Qualify for Benefits (percent)

	Pre-EI Reform (95Q3-96Q2)	Post-EI Reform (97Q1-97Q4)	t statistic	N
All	83.0	74.9	-4.39	3,542
Age
Youth (15-24)	74.7	62.5	-2.29	491
Prime (25-54)	84.0	76.2	-3.82	2,739
Older (55+)	88.2	86.4	-0.35	312
Family Type
Single with Children	75.2	75.8	0.10	344
Single without Children	80.6	75.4	-1.51	920
Married with Children	82.2	68.0	-3.90	1,125
Married without Children	88.2	80.6	-2.75	1,150
Education
Less than High School	82.2	74.2	-1.91	706
High School	80.7	73.6	-1.94	1,009
More than High School	84.5	75.6	-3.61	1,755
Other	79.1	80.1	0.08	72
Area Type
Rural	81.5	73.4	-2.21	1,199
Urban	83.4	75.5	-3.77	2,343
Region
Atlantic	82.9	79.0	-1.80	1,131
Quebec	81.9	76.2	-1.47	419
Ontario	83.0	71.3	-2.96	483
Prairies	82.6	72.6	-3.99	1,082
British Columbia	85.3	81.0	-1.32	427
Source: COEP Survey of Job Terminations 95Q3 - 96Q2 and 97Q1 - 97Q4.

There has clearly been a significant change between the pre-EI reform period and the post-EI reform period. In the four quarters preceding EI reform, 83 percent of women had enough weeks to qualify for EI. All other factors being equal, the move from a weeks-based system to an hours-based system appears to have decreased female eligibility rates to 74.9 percent in the four quarters following EI reform.⁹³ Taking a closer look, this change primarily reflects a change for married women. This result is not unexpected, as Table 2 indicated that women, and especially married women with children, were far more likely to have worked part-time and for fewer hours per week than men.

The largest impact of the EI reform experienced by women was for those living in the Prairie Provinces. Eligibility rates declined from 82.6 percent in the pre-EI reform period to 72.6 percent in the post-EI reform period. Other groups that were more significantly affected than others included women who were married with children, women of prime age, women from urban areas, and women with more than a high school education.

11.5.2 Impact of the 1996 EI Reform on EI Receipt of Regular Benefits

Table 6 provides a breakdown of the percentage of women with a job separation receiving regular EI benefits in the periods preceding and following the 1996 EI reform. There was no significant change in the receipt of EI between the pre-EI reform period and the post-EI reform period in any of the categories, although overall EI receipt rates decreased from 46.8 percent to 44.6 percent.⁹⁴ This would suggest that, in relation to the probability of receiving EI benefits, women were not affected by the 1996 EI reform in the four quarters following the reform.⁹⁵ The only female group significantly affected by the EI reform was women in the Prairie Provinces, whose rate of EI receipt decreased in the post-EI reform period.

A number of factors may have contributed to the small size of the drop in the EI receipt rate for women. The most obvious reason is that the female take-up rate⁹⁶ increased slightly, albeit not significantly, from 56.4 percent in the pre-EI reform period to 59 percent in the post-EI reform period.⁹⁷ Because of this slight increase in the take-up rate, EI receipt rates did not decrease as much as they would have, had the take-up rate not changed at all. In fact, it was estimated that EI receipt rates would have decreased significantly in the absence of an increased change in the take-up rate.

Table 6 Percentage of Women With a Job Separation Receiving Regular EI Benefits (percent)

	Pre-EI Reform (95Q3-96Q2)	Post-EI Reform (97Q1-97Q4)	t statistic	N
All	46.8	44.6	-1.03	3,542
Age
Youth (15-24)	23.2	20.9	-0.53	491
Prime (25-54)	50.9	48.3	-1.04	2,739
Older (55+)	51.8	55.1	0.41	312
Family Type
Single with Children	46.7	42.2	-0.69	344
Single without Children	39.9	40.6	0.15	920
Married with Children	45.7	41.5	-1.07	1,125
Married without Children	54.2	52.0	-0.57	1,150
Education
Less than High School	50.1	47.3	-0.57	706
High School	48.3	44.4	-0.93	1,009
More than High School	44.9	44.1	-0.26	1,755
Other	49.7	32.9	-1.13	72
Area Type
Rural	50.2	46.3	-0.97	1,199
Urban	45.8	44.0	-0.70	2,343
Region
Atlantic	53.3	53.2	-0.02	1,131
Quebec	50.8	50.6	-0.04	419
Ontario	44.4	40.7	-0.79	483
Prairies	38.8	33.5	-1.84	1,082
British Columbia	49.6	48.3	-0.30	427
Source: COEP Survey of Job Terminations 95Q3 - 96Q2 and 97Q1 - 97Q4.

Table 7 gives the results of a statistical estimation of the probability that a woman with a job separation will collect regular EI benefits based on some key demographic and work characteristics.⁹⁸ The first column shows the likely change in the probability of receiving EI benefits when compared to a control group. Only women are included in the regressions in Tables 7 and 8.

The EI reform variable did not have a statistically significant impact on the probability that an unemployed female worker would collect EI benefits,⁹⁹ in line with the data given in Table 6, even after controlling for various demographics and job characteristics.¹⁰⁰

Also included in the estimation was a variable for the impact of EI reform on women with children. This variable was also statistically insignificant, indicating no change in the probability of collecting EI benefits due to EI reform for that group of women.

Table 7 Probit Regression for the Probability that a Woman With a Job Separation will Claim Regular EI Benefits (percent)

	% diff.	P value	Confidence Interval (90%)
EI Reform	-1.3	0.667	-6.2	3.6
With Children * EI Reform	-3.5	0.439	-11.0	3.9
Age
Youth (15-24)	-24.6	0.000	-29.3	-20.0
Prime (25-54)	Control	Control	Control	Control
Older (55+)	-2.8	0.528	-10.0	4.4
Family Type
Single with Children	4.4	0.271	-2.2	10.9
Single without Children	-0.2	0.966	-7.0	6.7
Married with Children	Control	Control	Control	Control
Married without Children	6.3	0.116	-0.3	12.8
Education
Less than High School	-1.4	0.664	-6.9	4.0
High School	Control	Control	Control	Control
More than High School	-2.1	0.449	-6.7	2.5
Other	-7.0	0.392	-20.1	6.2
Area Type
Rural	0.3	0.915	-4.2	4.8
Urban	Control	Control	Control	Control
Region
Atlantic	7.5	0.033	1.7	13.4
Quebec	5.9	0.102	0.0	11.9
Ontario	Control	Control	Control	Control
Prairies	-5.7	0.050	-10.5	-0.9
British Columbia	7.4	0.026	1.9	12.9
Industry
Agriculture	-0.7	0.932	-13.9	12.5
Primary	5.6	0.435	-6.2	17.4
Manufacturing	Control	Control	Control	Control
Construction	19.0	0.011	7.3	30.7
Transportation & Storage	-3.1	0.666	-14.7	8.6
Communications	7.5	0.466	-9.4	24.3
Other Utilities	-9.9	0.449	-30.5	10.8
Trade	-2.4	0.590	-9.5	4.8
Finance	-2.0	0.768	-13.2	9.1
Education	4.9	0.236	-1.9	11.7
Business Services	-0.1	0.975	-7.1	6.8
Public Services	-5.9	0.381	-16.7	5.0
Public Administration	-13.6	0.015	-22.3	-4.9
Job Characteristics
Seasonal	-2.3	0.442	-7.3	2.6
Part-Time	-12.6	0.000	-16.8	-8.4
Union	-2.4	0.484	-8.0	3.2
Weeks on Job	0.0	0.005	0.0	0.0
Unemployment Rate	0.7	0.075	0.1	1.3
Sample Size	7,532
Source: COEP Survey of Job Terminations 95Q3 - 96Q2 and 97Q1 - 97Q4.

The statistical estimation results in Table 7 also show some other factors that were significant in determining whether or not a woman would collect EI. In particular, youths were far less likely to have collected EI benefits than workers of prime age. This is not surprising, given the fact that youths were more likely to have been employed in positions with fewer hours of work per week, thereby reducing their EI eligibility.

Other significant estimation results indicated that:

• Workers in Atlantic Canada and British Columbia were more likely to have collected EI than workers in Ontario, while those in the Prairies were less likely to have done so;
• Those employed in construction were far more likely to have collected EI than workers employed in the manufacturing industry, while those employed in public administration were less likely to have done so;
• Part-time workers were far less likely to have collected EI than full-time workers;
• Workers who had spent more weeks at their previous job were more likely to have collected EI than workers with fewer weeks employed at their previous job; and
• Women in areas with higher unemployment were more likely to collect EI.

11.5.3 Impact of the 1996 EI Reform on Weeks of Entitlement

Although entitlement is an adequacy issue and not an access issue, it does require analysis, as it relates to EI receipt and the length of time that claimants are entitled to receive EI benefits.

Statistical estimates of the impact of EI reform on the number of weeks that a woman with a job separation is entitled to collect EI are given in Table 8.¹⁰¹ The EI reform variable did not have a statistically significant impact on the weeks of entitlement for women.¹⁰² For women with children, EI reform also had no impact.

Other estimation results suggested that the weeks of entitlement were less for youths. Residents of the Atlantic Provinces and Quebec had more weeks of entitlement than Ontario residents, while those in the rural areas had fewer. Seasonal and part-time workers also had fewer weeks of entitlement, while unionized workers had more. Finally, workers from high unemployment areas and those who were employed for long periods at their previous job had a higher number of entitlement weeks.

Table 8 OLS Regression for the Impact on the Weeks of Entitlement of Women with a Job Separation

	Coef.	P value	Confidence Interval (90%)
EI Reform	-0.3	0.628	-1.4	0.8
With Children * EI Reform	-1.1	0.319	-2.9	0.7
Age
Youth (15-24)	-2.8	0.001	-4.3	-1.4
Prime (25-54)	Control	Control	Control	Control
Older (55+)	-0.9	0.461	-2.8	1.1
Family Type
Single with Children	1.0	0.294	-0.6	2.5
Single without Children	0.4	0.661	-1.2	2.1
Married with Children	Control	Control	Control	Control
Married without Children	1.1	0.196	-0.3	2.5
Education
Less than High School	-0.5	0.505	-1.9	0.8
High School	Control	Control	Control	Control
More than High School	1.0	0.149	-0.1	2.1
Other	1.2	0.567	-2.3	4.7
Area Type
Rural	-1.0	0.086	-2.0	0.0
Urban	Control	Control	Control	Control
Region
Atlantic	1.5	0.096	0.0	2.9
Quebec	1.7	0.047	0.3	3.1
Ontario	Control	Control	Control	Control
Prairies	-1.1	0.104	-2.2	0.0
British Columbia	0.2	0.835	-1.2	1.5
Industry
Agriculture	2.4	0.112	-0.1	4.8
Primary	0.6	0.612	-1.5	2.7
Manufacturing	Control	Control	Control	Control
Construction	-0.1	0.936	-2.2	2.0
Transportation & Storage	-1.4	0.339	-3.8	1.0
Communications	0.5	0.790	-2.8	3.8
Other Utilities	0.5	0.861	-4.2	5.3
Trade	1.3	0.223	-0.5	3.1
Finance	-0.1	0.968	-3.3	3.1
Education	-0.5	0.592	-2.2	1.1
Business Services	-0.2	0.865	-1.9	1.6
Public Services	-1.3	0.402	-3.8	1.2
Public Administration	-1.2	0.436	-3.9	1.4
Job Characteristics
Seasonal	-4.1	0.000	-5.1	-3.0
Part-Time	-3.3	0.000	-4.4	-2.2
Union	1.4	0.039	0.3	2.6
Weeks on Job	0.0	0.000	0.0	0.0
Unemployment Rate	0.6	0.000	0.5	0.7
Sample Size	7,022
Source: COEP Survey of Job Terminations 95Q3 - 96Q2 and 97Q1 - 97Q4.

11.6 After EI Reform

Using the most recent data available, further research was conducted to monitor any changes that may have occurred from the second quarter of 2001 (cohort 24) to the second quarter of 2002 (cohort 28). Based on the results, there is reason to believe that eligibility for EI benefits and EI receipt rates have declined more for women than for men. However, it should be noted that if cohorts 22 and 26 are compared, eligibility and EI receipt rates increase more for women than for men.

11.7 Conclusions and Further Research

The first part of this paper examined selected demographic and job characteristics of women. COEP survey results indicated that women comprised 45 percent of all persons leaving their jobs. Women with a job separation were more likely to have been married and were more likely to have had higher levels of education. They were also far more likely to have been employed part-time and be unemployed for longer durations. For women with children, the job separations were even longer. It was also revealed that men and women collected EI at roughly the same rate.

The latter part of the paper assessed the impacts, all other things being equal, of the 1996 EI reform on female eligibility for benefits, the likelihood that women would claim EI benefits and on their weeks of entitlement to EI benefits. It was discovered that EI reform had reduced the eligibility rates of women. However, the analysis found no evidence that EI reform had any significant impact on EI receipt rates for women or on their number of entitlement weeks. Further research is required to investigate the reasons for why EI receipt rates did not go down significantly for women, while eligibility rates did.