by Katherine Scott and Clarence Lochhead

This article is an excerpt from the forthcoming Social Research Report, Are women catching up in the earnings race? which will be published by the CCSD in July 1997.

A media story early this year caught our eye. According to the story, a researcher at Statistics Canada had determined that the wage gap between men and women had narrowed dramatically in recent years, and he predicted that the end of gender discrimination in the labour market was rapidly approaching. A few months later, the Canadian Labour Congress issued a report on women's work: it concluded that the wage gap between men and women was continuing and indeed for some, it was growing.

In an effort to shed more light on the real story behind women's earnings, we have conducted our own analysis using data from Statistics Canada's 1984 and 1994 Survey of Consumer Finances. In our research, we wanted to explore whether or not women's and men's earnings are converging to a point of gender parity within the labour market. This is an important question because equitable earnings afford a greater degree of autonomy for women, providing them with the means to challenge gender inequality in the home, in the workplace, in communities and in governments.

Women's earnings inch upward

When we examine average earnings, we see that over the two past decades, women's earnings have increased, while men's have stagnated. Among all female earners, the average wage climbed throughout the 1980s, from $17,169 in 1981 – the peak year before the recession – to $19,153 in 1989 (measured in 1995 dollars). Since the 1990-91 recession, women's wages have been relatively flat, inching up to $20,219 by 1995. Among full-time full-year (FTFY) female workers – those employed 30 hours or more per week for at least 49 weeks – wages also rose between 1981 and 1995. In 1981, average earnings were $25,820 (in 1995 $); by 1995, average wages were $29,700.¹

As a result, the ratio of female to male earnings has climbed from 53.6 in 1981 to 65.1 in 1995. The earnings ratio between full-time full-year workers is predictably higher, but increases in the ratio have been smaller over this period. Looking at another comparative measure, women's share of earnings is also higher. In 1981, female workers received 29 per cent of all earnings; this share increased to 35 per cent by 1995. Among FTFY workers, women now take home 33 per cent of earnings, up from 25 per cent in 1981.²

The increased presence of women in the paid labour market, combined with a declining gap in average earnings, suggests that the employment experiences and earnings of men and women are indeed converging – that the historical incorporation of women into the formal economy is leading to the eventual end of gender-based earnings inequality in the labour market. Women today are much more likely to engage in paid work than at any other time in Canadian history, and their share of labour market earnings has risen accordingly.

Yet the story told by these data is limited. A declining wage gap, for instance, can result from two very different trends: small gains being made by all women, or large increases among relatively few women. A more complete test of the "convergence" hypothesis requires us to analyse how all earners are distributed. Thus, we are interested in exploring whether the location of women within the earnings distribution is converging with the location of men. We posed the following key questions:

• What share of women are found in the bottom, middle and top ranges of the overall earnings distribution? Have these shares changed over time?
• Historically, the majority of the highest earners have been men. Can we see from the 1984 and 1994 data that women are increasingly present among top earners? Is the gender composition of low earners changing as well?

A note on the methodology

To begin our analysis, we used 1984 data to divide earners – men and women aged 18 to 64 years – into 10 equal groups, called deciles. Each decile contains 10 per cent of earners, with the bottom decile representing the 10 per cent of earners with the lowest earnings, and the highest or top decile representing the 10 per cent of earners with the highest earnings.³ The dollar values that define these deciles are shown in Table 1. So, for example, in 1984, the lowest 10 per cent of full-time full-year earners had annual earnings of $14,121 or less, while the top 10 per cent of earners had annual earnings of $57,395 or more (expressed 1994 dollars). It is important to keep in mind that these figures are annual earnings of individuals, not families.

Table 1: Full-time Full-year Earners, by Deciles, Annual Earnings in 1984

1984 Earners (Deciles)	Annual Earnings (in 1994 $)
10 (highest)	$57,395 or more
9	$48,224-$57,394
8	$42,168-$48,223
7	$36,778-$42,167
6	$32,510-$36,777
5	$28,290-$32,509
4	$24,020-$28,289
3	$19,577-$24,019
2	$14,122-$19,576
1 (lowest)	$14,121 or less
Total	All levels
Source: Calculations by the Centre for International Statistics at the CCSD using microdata from Statistics Canada's 1985 Survey of Consumer Finances, 1984 income.

Using the earnings categories that define the 10 deciles presented in Table 1, we now examine where women and men are "located" within this overall earnings structure. Our intent is to determine the respective proportions of women and men found in the bottom, middle and highest income categories as well as any changes in these proportions over the past decade. If the labour market is completely neutral with respect to gender, we would find the same distribution for women and men; if the market is moving toward a convergence in the earnings of women and men, we would find greater symmetry in 1994 than we do in 1984. Using this methodology, we are able to present a detailed picture of earnings over this period, highlighting the progress of women and men against the broader earnings trends in the Canadian economy.

Women still concentrated at the bottom

Figure 2 shows the distribution of male and female FTFY earners. The bars in Figure 2 represent the distribution of men and women in 1984, and the triangles and asterisks represent the respective distributions a decade later. In short, women are much more likely than men to be found in the lowest end of the earnings range. This was true in 1984, and it still held true a decade later. For example, in 1994, 13.4 per cent of female earners had annual earnings of about $14,000 or less (the bottom decile), while only 5.9 per cent of men were in this range. At the other end, only 3.8 per cent of female earners were found in the top decile compared to 18.1 per cent of men.

But there has been a distributional shift of FTFY women towards higher earnings. For example, in 1984, 48.0 per cent of female FTFY workers earned less than $24,000 – the bottom three deciles – compared with 18.8 per cent of men.⁴ By 1994, 39.9 per cent of female workers fell into this category, a decline of 8.1 percentage points. Among men, there has been little change in the proportion with annual earnings under $24,000. The largest share of male FTFY earners continue to be found in the upper ranges of the earnings distribution. In fact, the largest concentration of male earners is in the ninth and tenth deciles (30.9 per cent), with earnings of more than $48,000 per year.

Is wage parity for women right around the corner?

Our analysis reveals an upward shift in the earnings distribution of Canadian workers between 1984 and 1994, due in large part to gains made by women, while men's position in the overall earnings distribution has remained relatively unchanged. Even though women's gains have been important in reducing the gap between the earnings of women and men, most notably among full-time full-year workers, the majority of women continue to be concentrated at the lowest end of the earnings ladder. In brief, persistent differences in the levels of male and female earnings, combined with the relatively small gains made by some women over the past 10 years, suggest that there is still some way to go before we achieve earnings parity in the labour market.

In our larger study, we examine the impact of the population's changing age profile on the distribution of earnings. This helps us to understand which groups of women have made wage gains, and where they rank in comparison to their male counterparts.

Notes

1. Wages increased among all female workers by 18 per cent between 1981 and 1995; the rate of increase was 14 per cent among FTFY workers, and 3 per cent among all male workers. Statistics Canada, Earnings of Men and Women in 1994,Catalogue 13-217, 1995, Text Table 1.

2. Female earnings shares calculated from Statistics Canada, Earnings of Men and Women in 1994, Text Table I, III.

3. In Table 1, some deciles contain slightly more than 10 per cent of earners, and some slightly less. This is because there are no earnings cut-offs that divide the survey data into precise 10 per cent groupings.

4. We use the bottom three deciles as our summary measure to examine distributional shifts among FTFY earners.

Katherine Scott is the senior policy and research associate at the CCSD.
Clarence Lochhead is the director of the Centre for International Statistics at the CCSD.

Canadian Council on Social Development, 309 Cooper Street, 5th Floor, Ottawa, Ontario, K2P 0G5