Category: Male/female earnings gap

Estimating non-discriminatory lifetime earnings for young females

Posted on By economica

by Christopher Bruce and Kelly Rathje

This article first appeared in the spring 2006 issue of the Expert Witness.

A number of judicial decisions have suggested that estimates of the lifetime earnings of young females should be based on data for males. Two arguments have been made in support of this view.

The first argument is that, as the differential between male and female earnings has been falling, incorporating the historical differential will understate the future earnings that young females will achieve once they become established in their careers. The second argument is that it is inappropriate for the courts to institutionalise current wage differentials that are based on discrimination.

The first of these arguments was alluded to in two BC judgments: ([B.I.Z.] v. Sams, [1997] B.C.J. No. 793; and Terracciano v. Etheridge and Fujii, [1997] B.C.S.C. B943125). The latter judgement involved a woman who was aged 16 when injuries from an automobile incident left her a paraplegic. In this judgement, Madam Justice Saunders indicated her preference for using earnings statistics of males to calculate Ms. Terracciano’s without-incident income:

[81] Indeed, it may be as inappropriately discriminatory to discount an award solely on statistics framed on gender as it would be to discount an award on considerations of race or ethnic origin. I am doubtful of the propriety, today, of this Court basing an award of damages on a class characteristic such as gender, instead of individual characteristics or considerations related to behaviour: Toneguzzo-Norvell (Guardian ad litem of) v. Burnaby Hospital, [1994] 1 S.C.R. 114.

The second argument was considered explicitly in the Alberta decision, MacCabe v. Westlock (RCSSD #110 et al [action: 9303 05787]). There, for example, the court accepted the use of income statistics for males to estimate the without-incident potential earnings of a young woman who had been paralysed while still a high school student. The court found that Ms. MacCabe had a without-accident income potential that was well above average. Importantly, the court argued:

[para468] Clearly the evidence establishes that the exceptional individual characteristics of the Plaintiff are such that her abilities would have commanded the equivalent salary of her male counterparts. She would have established a strong attachment to her career. The use of male wage tables is justified. In any event, I am of the view that any award which I grant to the Plaintiff should not and cannot be solely determined by her gender.

[para469] It is entirely inappropriate that any assessment I make continues to reflect historic wage inequities. I cannot agree more with Chief Justice McEachern . . . in Tucker, supra, that the courts must ensure as much as possible that the appropriate weight be given to societal trends in the labour market in order that the future loss of income properly reflects future circumstances. Where I differ is that I will not sanction the “reality” of pay inequity. The societal trend is and must embrace pay equity given our fundamental right to equality which is entrenched in the constitution. . . .

However, these decisions were silent about the possibility that women might earn less than men even after discriminatory practices had been removed. Importantly, for example, the Court of Appeal in MacCabe (9803-0617AC) rejected the trial court’s finding that male contingencies (such as for non-participation, unemployment, and so forth) should be used when estimating Ms. MacCabe’s losses.

[105] In general, tort law and in particular, the quantification of damages necessitates an individual approach. This is where I find the learned trial judge erred. In attempting to rectify potential inequities in the methods for quantifying damages, the learned trial judge neglected to focus on the evidence and the individual actually before her. While principles of equality should inform tort law, the learned trial judge’s application of equitable principles resulted in her ignoring some of the relevant material facts.

[106] In this case, based on the evidence, it was not reasonable to calculate MacCabe’s damages based on male contingencies. MacCabe stated she wanted to have children and would have preferred to stay at home with them for some period. This meant she would not have worked a pattern typical for male physiotherapists. There was no evidence to indicate it was more likely that MacCabe would not have had children and chosen not to take time off from full time paid employment as a physiotherapist. Thus, it would be inappropriate to apply male contingencies to her when there was no evidentiary basis that she would have worked a typical male pattern.

The appellate court in MacCabe recognised that, even in a world in which men and women experienced equality of opportunity, incomes might differ between the sexes if they made different choices. Most importantly, women might choose to take more time off to care for children; might choose to work fewer hours per week or weeks per year; and might retire earlier than men.

If this is true, one approach to predicting the earnings of young women might be to take current earnings data for men and adjust them downward for these differences in contingencies. In this article, we consider the impacts that each of four such contingencies might have on the male/female earnings differential.

Participation rates

The percentage of any group that is either working or available for work at any time is known as the “participation rate.” Primarily because of the impact of child-rearing, women have lower participation rates than do men at virtually every age. Hence, if everything else is equal, women’s earnings in an average year will be lower than men’s in proportion to the differences in participation rates.

Statistics indicate that the best predictor of whether a particular woman will be in the labour force in the future is whether she was in the labour force in the past. In particular, married women who had been working prior to the birth of their first child have a strong tendency to return to work within one to two years of the birth of that child. Nonetheless, there is a large number of women who delay re-entry until their youngest child is of school age. [1]

Also, many authors have detected a strong positive correlation between education level and female participation. That is, the most highly educated women tend to have the strongest attachment to the labour force. Table 1, for example, indicates that 86.4 percent of 25-54 year-old, university-educated women in Alberta participated in the labour force in 2002, compared with 80.7 percent of 25-54 year-olds who had high school education. (Participation rates drop significantly after age 54 due to retirement.)

Table 1

Note that the figures reported in Table 1 reflect the current labour force participation of women. It seems likely, however, that women who are currently in their 20s and 30s will maintain their high participation rates as they age. Therefore, the figures for the older age categories, in Table 1, might be adjusted upwards when projecting the future behaviour of plaintiffs who have not yet reached the age of majority.

Given the figures in Table 1, and recognising that male participation rates are approximately 95 percent for all age and education groups (between 24 and 54), the earnings for females might be estimated by reducing male earnings by the difference between 95 percent and the relevant female labour force participation rate. For example, given the information in Table 1, the earnings for females with post-secondary education could be estimated by reducing the earnings of comparable males by approximately 10 percent. Similarly, earnings for those with high school education might be obtained by reducing male earnings by approximately 15 percent.

Part-time employment

Even when they are in the labour force, women report lower annual incomes than men because they are more likely to work part-time. As Table 2 indicates, in the age group 25-54, approximately 20 to 25 percent of women work part-time, whereas only 2 to 5 percent of men do so. That is, on average, the number of women working part-time is approximately 20 percentage points higher than the number of men. If part-time is interpreted to mean “half time,” this implies that women in the labour force work 10 percent fewer hours than men. Everything else being equal, therefore, this difference suggests that male earnings should be reduced by a further 10 percent, in order to obtain an estimate of female earnings.

Table 2

Hours Worked

There is also evidence to suggest that, even among individuals who work “full-time,” women work fewer hours per week than do men. Table 3 indicates, for example, that women’s “usual hours per week” are about 80 percent of those of men. Furthermore, Sweetman [2] reports that this ratio varies only slightly by education level: in the age group 40-44, for example, the ratio of female to male hours worked is 81 percent for those with high school and 85 percent for those with a bachelor’s degree.

Table 3

It would be double-counting, however, to reduce male earnings by both 10 percent for part-time work and 15 to 20 percent for hours worked, as the latter differential includes the effect of the former. Rather, it appears that, among those individuals who work full time, women work 5 to 10 percent fewer hours than do men. It is this contingency that should be applied to male earnings.

Retirement

It is seen in Table 4 that women retire approximately two years earlier than men, on average. Hence, any estimate of female lifetime earnings will have to take this difference into account.

Table 4

Summary

Even if there was no discrimination in the labour market – that is, even if women received the same hourly wages as men – on average, women’s annual earnings would still be lower than men’s. Among those with high school education, the differential would be as much as 35 percentage points, due to differences in labour force participation rates (15%), propensity to work part-time (10%), and hours worked per week (10%). Among university graduates, the differential would be approximately 25 percentage points.

Accordingly, if it is concluded that women’s attachment to the labour market will not change in the future, women’s average annual earnings cannot be expected to rise above 65 percent (high school graduates) to 75 percent (bachelor’s degrees) of men’s earnings.

Only if women increase the percentage of their time that they devote to the labour market will that 25 to 35 percent differential begin to fall.

Application

How should these statistics be applied? We believe that two broad cases can be distinguished. In the first, the plaintiff was old enough at the time of her injury that it is possible to determine both the occupation she would have entered and the strength of her attachment to the labour market. (This was the situation in MacCabe, for example.) In those cases, information specific to the plaintiff should be used to predict her (non-injury) earning capacity.

In the second case, the plaintiff was young enough that neither her career nor her labour market attachment can be predicted. In such cases, we believe that information about the plaintiff’s family background is sufficient to allow the court to identify approximately what her educational attainment would have been. Census data concerning incomes by education can then be used to predict the plaintiff’s earning capacity.

But, is it census data for females, or for males, that should be used for this purpose? We have argued that, if the court believes either that labour force discrimination will largely disappear over the next few decades, or that the effects of discrimination should not be institutionalised in damage awards, it is male data that should form the basis of the award. However, the information we have presented in this paper suggest that, even in the absence of discrimination, women will earn less than men because of differences in attachment to the labour force. For that reason, we would propose that, for young females, the forecast of earnings capacity should be based on male data; but that those data should be adjusted downwards as we discussed above.

Footnotes

1. C.J. Bruce, Assessment of Personal Injury Damages (Butterworths: Toronto, Vancouver), fourth edition, 2004, page 167. [back to text of article]

2. Arthur Sweetman, 2002. “Working Smarter: Education and Productivity,” The Review of Economic Performance and Social Progress, in: Andrew Sharpe, Executive Director & France St-Hilaire, Vice-President, Research & Keith Banting, Di (ed.), The Review of Economic Performance and Social Progress 2002: Towards a Social Understanding of Productivity, volume 2 Centre for the Study of Living Standards. [back to text of article]

leaf

Christopher Bruce is the President of Economica and a Professor of Economics at the University of Calgary. He is also the author of Assessment of Personal Injury Damages (Butterworths, 2004).

Kelly Rathje is a consultant with Economica and has a Master of Arts degree (in economics) from the University of Calgary.

Male Versus Female Earnings – Is the Gender Wage Gap Converging?

Posted on By economica

by Kelly Rathje

This article first appeared in the spring 2002 issue of the Expert Witness.

When we estimate the potential income of a young female (without- or with-accident) who does not have a well-established career path, we rely on census data and usually present earnings for both males and females. As is well-known, men have, on average, earned more than women. A number of reasons have been offered for this, including: labour force discrimination, different occupational choices, differences in labour force participation trends, and so forth. However, it is also well-known that the average income earned by women has been increasing relative to that earned by men. In 1967, women’s earnings were approximately 58 percent of men’s earnings. By 1997, women’s earnings were approximately 73 percent of men’s earnings. But will this trend continue, and will the gender wage gap continue to close in the future?

A recent paper by Michael Shannon and Michael Kidd addresses the question of the size of the gender wage gap in the future.* Using recent Canadian data, they project future trends, based on current trends in educational attainment and labour force participation. They then use these predicted trends to estimate the wage gap from 2001-2031 using a statistical model. They find that the wage gap will continue to close, however a wage gap of approximately 22 percent will still exist in 2031.

In this article, I first examine current and projected trends in educational attainment and labour force participation – two factors influencing earnings. Then, I present Shannon and Kidd’s results regarding the projected gender wage gap. Finally, I consider the implications of their results for the estimation of the potential incomes of young females.

Educational attainment

One factor that influences earnings is educational attainment. In recent years, female educational attainment has increased relative to that of males. To incorporate recent trends in educational attainment, Shannon and Kidd create an age-education pattern for both males and females. In 2000, it is found that individuals in the 25-29 year age category are better educated than individuals in the 55-59 year age group, and that this trend will continue into the future. For example, in 2000 approximately 2 percent of individuals (either males and females) in the 25-29 year age category have less than a high school education, compared to approximately 23 percent in the 60-64 year age category. As the individuals in the 25-29 year category age, the pattern of educational attainment is carried through into future years. The number of individuals in the 55-59 year age category in 2030 (individuals who were in the 25-29 year age category in 2000) that have less than a high school education will decline to approximately 2 percent, and we see higher education levels for all age groups in the future.**

In addition, female university enrollment has increased. In fact, women now account for the majority of university students, and females are entering fields that were typically male-dominated (such as engineering, applied sciences, and mathematics).

For the purposes of their calculations, Shannon and Kidd make the conservative assumption that educational enrollments will remain constant into the future. In 2000, approximately 22 percent of women (aged 25-64) had a high school diploma, 32 percent had a post-secondary diploma, 14 percent had a bachelor’s degree, and 5 percent had a graduate degree. By 2031 it is predicted that approximately 17 percent of women will have a high school diploma, 35 percent will have a post-secondary diploma, 18 percent will have a bachelor’s degree, and 8 percent will have a graduate degree.

Male educational attainment, as a comparison, is predicted to remain relatively unchanged over the 30 year period considered. In 2000, approximately 19 percent of males had a high school diploma, 33 percent had a post-secondary diploma, 13 percent had a bachelor’s degree, and 8 percent had a graduate degree. By 2031 it is predicted that approximately 20 percent of men will have a high school diploma, 36 percent will have a post-secondary diploma, 14 percent will have a bachelor’s degree, and 7 percent will have a graduate degree.

These results indicate that women are “catching up” to males in the percentage that obtain higher levels of education. Since higher education tends to lead to higher wages, the increased educational attainment of women, and the constant attainment of males, contributes to a closing of the gender wage gap.

Labour force participation

Another factor that influences women’s earnings is that they tend to take time away from the labour force (either to withdraw entirely or to reduce hours to part-time status) for a period of time – as is common for women who choose to have families. Thus, women, on average, bring less experience to their jobs, which means they tend to have lower incomes at any given age.

Labour force participation rates have shown steady growth over the last three decades, and many experts anticipate that they will continue to rise. Moreover, relying upon historical participation rates by age cohort may be misleading as many women are delaying the onset of pregnancy. In 1986, on average, women were approximately 25 years old when their first child was born. By 1996, women were approximately 27 years old when their first child was born. During the early years between finishing school and starting a family, women are tending to work full-time in their careers. It is in the early years of one’s career that substantial wage growth usually occurs. By delaying starting a family, women can be more flexible in career decisions such as traveling, relocation, overtime, etc. Thus, women may benefit from the higher wage increases earlier on in their careers. Also, they may be able to exit the labour force at a time that will have less impact on their careers, and their earning potential.

Shannon and Kidd predict that women will have increased their number of years of work experience by 2031. A summary of the actual (1994 and 2001) and estimated (2001 and 2031) years of work experience is outlined in Table 1.

Figure 1

Two important predictions are made in Table 1. First the number of years of experience obtained by males at each age group will not change significantly over the next 30 years. Whereas males 45-49 had worked an average of 25.6 years in 1994, for example, they are predicted to have worked 25.5 years in 2031. Similarly, the work experience of 55-59 year-old males is predicted to change by only 0.2 years – from 36.6 to 36.8 years over the same time period.

Second, whereas the work experience of young females is predicted to remain relatively unchanged, older women are predicted to obtain more years of lifetime employment. For example, while the work experience of 35-39 year-old females is predicted to change by 0.1 years between 1994 and 2031 – from 12.5 to 12.4 – the work experience of 55-59 year-olds is predicted to increase by 2.3 years – from 23.9 to 26.2. And the experience of 60-64 year-olds is predicted to increase by 5.1 years.

Shannon and Kidd concluded that these changes will produce only a slight narrowing of the wage gap between men and women – and then only in older age groups. But their results did not allow for changes in number of hours worked in a lifetime. It is also possible that some wage gains could be obtained by women if they were to work more full-time hours, and less part-time, and if they were to increase their full-time hours. In 1997, for example, women working full-time, worked 39 hours per week on average, whereas men worked 43 hours.

Shannon and Kidd incorporate the trends summarized above to determine the future wage gap. Their results are shown in Table 2 below.

Figure 2

There, it is projected that the gender wage gap will decline in the future. On average, it is projected that the difference between incomes for males and females in the 25-44 year age category will fall to approximately 17 percent by 2031. That is, full-time, full-year wages for females within the 25-44 year age category are projected to be approximately 83 percent of their male counterparts. By comparison, women in the 45-64 year age group will earn approximately 71 percent of their male counterparts’ incomes.

Conclusions & implications

Shannon and Kidd’s results imply that the gender wage gap will continue to close, but a gap of approximately 22 percent will still exist in 2031. Increasing female labour force participation and educational attainment, coupled with the relative stability of the male labour force participation and education attainment contribute to the wage gap closure.

In comparison to the wage gap closure from 1967 to 1997 (42 percent to 27 percent, or 15 percentage points), the results for the next three decades suggest that convergence of the gender wage gap will slow from 2001 to 2031 (29 percent to 22 percent, or 7 percentage points). The authors’ findings also suggest that changes in the wage gap for older individuals (within the 45-64 year age group) will produce the greatest convergence (43 percent to 29 percent, or 14 percentage points).

Part of the projected wage gap in 2031 is due to the differences in the labour market characteristics addressed by Shannon and Kidd. Since women tend, on average, to work fewer years over their work-life; work fewer hours per week; and are more likely to withdraw from the labour force or reduce their hours to part-time for the purposes of raising a family, their wages will, on average, be less than those of their male counterparts. However, these characteristics historically have accounted for only half of the wage gap. The portion of the wage gap that cannot be explained by labour market characteristics is generally attributed to discrimination and to differences in preferences between men and women. For example, women tend to be the primary caregivers. Thus, they may choose to work in lower-paying jobs that have more flexibility regarding sick days and hours worked, or within positions that are easily entered and exited. These are factors which also contribute to the wage gap, but are not easily captured using traditional statistical methods, such as those used by Shannon and Kidd.

What do these findings imply for using male earnings when predicting the potential income for young females? It seems reasonable to conclude that the findings suggest that historical average income figures for women underestimate the future potential income of an average young woman today. This is because historical income figures reflect women who (on average) had a much different labour force experience than today’s average young woman will experience, and that young women in the future will experience. It seems that the “reality” for today’s average woman lies somewhere between historical figures for males and females. It appears that even young women who will follow a “traditional” average female career path will earn more than the average women represented by historical data since today’s females are acquiring higher education levels and displaying a greater labour force attachment by participating full-time in the labour force longer.

Thus, it may be appropriate to use average earnings for males to predict the future potential income of an average young female, and then to apply contingencies to reflect the possibility of labour force absences and part-time employment. I emphasize, however, that this approach still carries difficulties. For example, women tend to enter different careers than men, even when they are working full-time. That is, there is still a tendency for occupations to be “male-dominated” or “female-dominated”, and the female-dominated occupations tend to pay less, even considering the same level of educational attainment between men and women. Thus, using male earnings data for any given level of education (considering all occupations) may overstate the potential life-time earnings of a young female.

Footnotes

* Shannon, Michael and Michael Kidd, 2001, “Projecting the Trend in Canadian Gender Wage Gap 2001-2031”, Canadian Public Policy. Vol. XXVII, No. 4, 447-467. [back to text of article]

** Shannon and Kidd also consider a scenario in which enrolment increases in the future at the same rate it had increased in the prior 12 years. For the purposes of this article, I focus on the more conservative scenario in which they assume that there is a one-time jump in enrolment from 1994-2000, and then enrolment remains constant over the 2001-2031 period. [back to text of article]

leaf

Kelly Rathje is a consultant with Economica and has a Master of Arts degree (in economics) from the University of Calgary.

The MacCabe Judgment: Allowing the Use of Earnings Statistics for Males When Estimating the Future Income of a Female

Posted on By economica

by Derek Aldridge

This article was originally published in the autumn 1998 issue of the Expert Witness.

On October 5, the Alberta court released its decision in the case of MacCabe v. Westlock RCSSD #110 et al (action: 9303 05787). The judgment is important for many reasons, though the most important aspect from an economist’s point-of-view is that it recommended the use of male earnings statistics to estimate the future earnings potential of a female. In particular, it was found that Ms. MacCabe would have followed a career path similar to that of the average male. That is, the court concluded that she would not have taken significant amounts of time out of the workforce for child rearing, and she would not have worked part-time. Therefore, it found that earnings statistics for males should be used to predict what her income would have been.

Some of the most important sections from the decision (related to the male/female income statistics issue) are reproduced here:

[para468] Clearly the evidence establishes that the exceptional individual characteristics of the Plaintiff are such that her abilities would have commanded the equivalent salary of her male counterparts. She would have established a strong attachment to her career. The use of male wage tables is justified. In any event, I am of the view that any award which I grant to the Plaintiff should not and cannot be solely determined by her gender.

[para469] It is entirely inappropriate that any assessment I make continues to reflect historic wage inequities. I cannot agree more with Chief Justice McEachern . . . in Tucker, supra, that the
courts must ensure as much as possible that the appropriate weight be given to societal trends in the labour market in order that the future loss of income properly reflects future circumstances. Where we differ is that I will not sanction the “reality” of pay inequity. The societal trend is and must embrace pay equity given our fundamental right to equality which is entrenched in the constitution. . . .

[para470]  . . . The Court cannot
sanction future forecasting if it perpetuates the historic wage disparity between men and women. Accordingly, if there is a disparity between the male and female statistics in the employment category I have determined for the Plaintiff the male statistics shall be used, subject to the relevant contingencies. . . .

[para481]  I agree with Dr. Bruce that absent
the accident, the Plaintiff would have been committed to her career and there would not have been a significant withdrawal from the labour force. . . .

So what does this imply about future cases involving injured or deceased females? It seems clear to us that if it is accepted that a young female would have followed a career path similar to that of the average male (in which she works full-time and does not take significant amounts of time out of the workforce for child rearing), then it follows that income statistics for males should be used to estimate her pre-accident income. (We discussed this issue in the Autumn 1997 issue of the Expert Witness.)

But what if it is found that a young woman would have followed a traditional female career path? In this case we suggest that using income statistics for females will still probably underestimate the true income path, but using those statistics for males will probably overestimate the true income. The reality likely lies somewhere in between the two alternatives. However, the MacCabe judgment appears to leave open the possibility that earnings statistics for males could be used even for female plaintiffs who would have followed “traditional” female career paths. It may be the case that the courts will choose to apply earnings statistics for males, regardless of the evidence about the woman’s likely career path – as a sort of “social justice” choice (Recall paragraph 469 of MacCabe: “I will not sanction the ‘reality’ of pay inequity.”)

However, the same argument could possibly apply to other situations in which a certain group of people earn less, on average, than the average male. For example, it is well-known that, on average, Natives earn less than non-Natives. From the MacCabe decision it may follow that one should use average income statistics for males to estimate the potential income of a young Native male (or female), with no adjustment to account for the reality that the average Native earns less than average non-Native. Conversely, perhaps the defense could argue that a person who has been disfigured in an accident should not be compensated for the “appearance-discrimination” component of his loss of income because that would be an endorsement of the “reality” of discrimination. If the court chooses to correct for the reality of pay inequity, then this could raise some difficult issues for those of us involved in loss of income cases.

leaf

Derek Aldridge is a consultant with Economica and has a Master of Arts degree (in economics) from the University of Victoria.

Using Male Earnings Data to Forecast the Future Income of Females

Posted on By economica

by Derek Aldridge

This article first appeared in the autumn 1997 issue of the Expert Witness.

In two recent British Columbia judgments ([B.I.Z.] v. Sams, [1997] B.C.J. No. 793; and Terracciano v. Etheridge and Fujii, [1997] B.C.S.C. B943125), the court accepted use of average earnings statistics for males to estimate the future earnings of female plaintiffs. In this article, I investigate the reasoning behind the court’s decision, and the circumstances in which it might be appropriate for us to use male data when estimating female earnings.

As an introduction, I first consider some statistical evidence regarding the “wage gap” between men and women. How much of this gap is due to discrimination, and how much can be explained by other factors? Second, I examine how the wage gap has diminished somewhat over time. Finally, I consider the implication of the economic evidence, together with the recent court decisions.

The Wage Gap

As we know, women, on average, tend to earn less money than men. In fact, in 1991, average income for women was only 61.5 percent of that for men. However, part of this gap is because a higher proportion of women workers are part-time. If we compare women working full-time to men working full-time, we find that (in 1991), women earned about 70 percent as much as men.

However, a wage gap of about 30 percentage points remains. Can this gap be explained by educational differences – are women earning less than men simply because they do not invest in as much education? The answer is no. Even when researchers compare men and women with the same level of education, the wage gap remains. For example, statistics indicate that Canadian women with university degrees earn only 72 percent as much as Canadian men with degrees. Some, but not all, of this wage gap is due to discrimination in the labour market.

In fact, it appears that much of the wage gap is due to factors other than labour market discrimination. One source is that women tend to work fewer years in total, over their careers, than do men – they have a greater tendency to interrupt their careers and withdraw from the labour force (especially to raise children), and they are also more likely than men to work part-time. These factors are reflected in labour force statistics which indicate that while 95 percent of 25-44 year old male workers are employed full-time, only 77 percent of women workers of the same age are full-time. These factors effectively reduce the average amount of work experience that women accumulate over their careers (part of what economists call human capital). So because women, on average, bring less experience to their jobs, they also tend to earn lower incomes at any given age.

Based on this, we might expect that if we compared men and women in the same jobs, with the same education, and the same amount of work experience, the wage gap would disappear. However, that is not the case. A Canadian study examined this question by surveying men and women who graduated from Canadian universities in 1982 and comparing their annual incomes two years and five years after graduation. The study compared men and women who had completed the same type of degree in university and who had worked continuously over the study period. The conclusion was that, even when controlling for education and experience factors, women still earned less than men – after two years the study found that the women graduates were earning 88 percent as much as men, and after five years they were earning about 82 percent. This trend was visible even for women with master’s and doctoral degrees (though women with doctorates in medical and health sciences were earning more than their male counterparts after two and five years). A particularly notable result is that, on average, the gap between men’s and women’s earnings actually increased as their careers progressed.

This leaves an unexplained wage gap of at least 10 percent – it is this portion of the gap which is generally attributed to discrimination. However, not all of this “discrimination wage gap” is due to discrimination in the labour market. Some of it seems to be due to the type of career paths that women tend to choose within occupations – perhaps they are positioning themselves somewhat for a future point at which they expect to temporarily withdraw from the labour force or drop to part-time status. It also seems that, to some extent, women tend to be socialized – within their families, in school, and culturally – to choose different sorts of career paths than men. The portion of the wage gap that remains after accounting for these factors is due to labour market discrimination – maybe 3-5 percent.

Note however, that the tendency for women to be socialized toward lower-paying careers may result from systemic discrimination – discrimination between boys and girls in the way that they are raised. It is also sometimes argued that traditional women’s occupations are lower-paid because women predominate in these professions (to the extent that this is true, the labour market is responsible). If we include these forms of discrimination, then the total wage gap due to discrimination is more like 10-15 percent.

The Wage Gap: Changes Over Time

When deciding whether to rely on historical income statistics to forecast future earnings, it is important to consider whether the historical relationships of the past can be expected to apply in the future. Examination of historical average earnings statistics for men and women working full-time reveals that the average wage gap has shown a clear decreasing trend over time. Specifically, the ratio of average female earnings to average male earnings increased from 59.7 percent in 1971 to 73.1 percent in 1995. This trend is illustrated in Figure 1 below.

Figure 1: Ratio of Average Earnings of Females to Average Earnings of Males

Figure 1

If we adjust average earnings statistics for the effects of inflation (so that, for example, earnings in 1971, 1981, and 1991 are all expressed in 1996 dollars), we discover that average male earnings only benefited from slight real increases over the last 25 years. (Specifically, average male earnings grew at approximately 0.10 percent per year over 1971-95.) The average earnings of women, on the other hand, experienced noticeable real growth – approximately 1.28 percent annually over 1971-95. This suggests that the male-female wage gap is decreasing over time because women are experiencing significant real wage gains, while men are not. We suspect that this trend is largely due to women spending more time in the labour force (increasing participation rates), pursuing higher paying occupations (including many “traditional male occupations”), and facing less discrimination than in the past. The annual real wage gains of men and women over this period are shown in Figure 2 below.

Figure 2: Real Changes in Earnings of Males and Females

Figure 2

We noted above that women tend to participate less in the labour market than men – they interrupt their careers more often, and for longer periods of time. The tendency toward work interruptions among women is changing though – recent information from Statistics Canada indicates that women’s labour force interruptions are now significantly shorter than they were in the past: over half of all Canadian women now return to work within two years of an interruption, compared with only an eighth in the 1950s. As it continues, this trend will further narrow the wage gap. We also know that women with more education tend to return earlier to the labour force.

Given these trends, past earnings averages for women will not accurately reflect what the average woman will earn in the future. Women are catching up to men, and it seems reasonable that today’s young women can expect to earn approximately the same lifetime income as today’s young men, if they follow similar career paths.

Implications

What do these findings tell us about when we should use earnings statistics for men to forecast the future earnings of a woman? It seems that this would be appropriate if we have reason to believe that the woman involved would have followed a career path more typical of men than of women (historically). For example, if it is believed that a young woman would have worked full-time (or very nearly so), and without interruption, throughout her career, then it would appear to be appropriate to use earnings data for males in her occupation.

Of the two BC judgments noted at the beginning of this article, one involved a woman who had already established her career path at the time of the accident, while the other involved a young girl who had not completed high school. In the former, it was apparently reasonably clear that she was following the sort of career path that has been typical of males, rather than females. In the latter, it was argued that the girl would have followed a typical male career path. In either of these types of situations it seems that using average male earnings statistics will better predict what a woman’s future earnings will be (or would have been, but for an injury or death).

However, what if a plaintiff has not established a career path at the time of her injury, and it is unclear whether she would have followed a typical male or a typical female career path?

In these cases, economists have typically chosen to forecast a young woman’s income based on her expected level of education – using statistics representing average earnings for women with a certain level of education. Our discussion so far may seem to suggest that average earnings for males of the given education level might be a better choice than using that for females. However, there are some difficulties with this approach. As noted, on average, women tend to enter different careers than men, even when they are working full-time (that is, we still observe a trend of “typical male occupations” and “typical female occupations”). And the typical female occupations tend to pay less. Given this, we would expect that the average income for women of a given education will continue to be less than the average for men of the same education – even if the women are working full-time without interruption. This holds even if we believe that labour market discrimination will end.

This suggests that using male earnings data to forecast the earnings of a young woman might overstate the woman’s true earning potential if we are basing our income estimates solely on a given education level, rather than on a given education and a given occupation. (Though, using earnings data for females will almost certainly underestimate the earnings potential of a young woman.)

Alternatively, if one is calculating the young woman’s potential income by assuming that she would have worked at a specific occupation (as an economist, for example), then it would probably be more accurate to rely on male earnings data, and then explicitly apply contingencies reflecting the impact of possible labour force absences and part-time employment. By using historical data for males, we can hopefully correct for the errors introduced when we use historical data for women (which reflects women who followed different career paths and faced greater discrimination than women today and in the future). By directly applying the appropriate contingencies for non-participation and part-time employment, based on our knowledge of the particular plaintiff, we will adjust for the probability that the woman may or may not have followed a “traditional” woman’s career path. These two adjustments will allow us to determine a reasonable forecast of a woman’s earnings, knowing that even if she follows a “traditional” career path, she will likely not face the same degree of discrimination as faced by past women whose earnings formed the basis for current statistical averages.

We should emphasize, however, that these generalizations can always be overridden by the facts of a particular case. If it is reasonable to assume that a young girl would have followed a career path more typical of men than of women (even if we do not know what that career would have been), then it is also reasonable to use male earnings data to forecast her income.

leaf

Derek Aldridge is a consultant with Economica and has a Master of Arts degree (in economics) from the University of Victoria.