November 2002     Vol. 3, no. 11

Cumulative earnings among young workers

René Morissette

The labour market has changed markedly over the last three decades. Young people now stay in school longer than in the mid-1970s. Young men not attending school are less likely to be employed full time, and for much of the 1990s, those who worked full year, full time, received lower earnings than their counterparts in the mid-1970s. But, young women not attending school are generally more likely to have a full-time job than their counterparts three decades ago, and many of those who work full year, full time receive higher earnings.

While changes in annual earnings of various groups of workers over the last three decades have been well documented (Morissette, Myles and Picot 1994; Beach and Slotsve 1996; Heisz, Jackson and Picot 2002), the combined effect of changes in school attendance, full-time employment rates, and annual earnings on cumulative earnings-the sum of earnings that individuals receive over several years-has not been investigated.

Cumulative earnings are important for several reasons. Decreases in cumulative earnings and the resulting effect on wealth holdings may lessen the ability of individuals to buffer income losses caused by permanent layoffs. With a smaller financial cushion, they may postpone the decision to leave home, get married, have children, buy a first house, or start a business. They may also reduce consumption during unemployment spells, search more intensely for a new job if unemployed, and be less likely to quit a job.

Using the Survey of Consumer Finances and the Survey of Labour and Income Dynamics, this article first reviews changes in school attendance, employment rates, and earnings in Canada since the mid 1970s. It then provides estimates of cumulative earnings over two 12-year periods-1973 to 1984 and 1988 to 1999-and examines the extent to which changes in cumulative earnings can explain changes in wealth holdings of families.

School, employment and earnings: 1976-2001

In 1976, only about one-third of young men (aged 16 to 24) attended school full time (Table 1). Twenty-five years later, almost half of them did so. Of those who were not in school full time, 80% had a full-time job in 1976, compared with only 69% in 2001.  1  Furthermore, even though their annual pay (in constant dollars) rose substantially between 1996 and 2000, young men employed full year, full time received 9% lower wages and salaries in 2000 than their counterparts in 1975 (Chart), and even less in the mid-1980s and mid-1990s.

Longer school attendance and lower chances of having a full-time job when not in school were also observed for men aged 25 to 29. Between the mid-1980s and the mid-1990s, the annual wages and salaries of those working full year, full time were much lower than in 1975. Between 1996 and 2000, they rose markedly, returning almost to their 1975 level in 2000.  2 

Men aged 30 to 54 were less likely to hold a full-time job in 2001 than in 1976. Among those employed full year full time, only those 45 to 54 saw an increase in wages and salaries over the 25 years.

The story is different for women. While young women increased their school attendance markedly, even more so than young men, women 25 and over saw their chances of holding a full-time job once out of school increase by at least 20 percentage points over the same period.  3  Furthermore, their wages and salaries evolved much more favourably, staying fairly constant for women aged 25 to 29 and increasing at least 20% for those 30 to 54.

Cumulative earnings: 1973-1984, 1988-1999

Changes in school attendance and in the wage structure have been documented for Canada and several other OECD countries. While considerable effort has been spent assessing the causes of the changes in the distribution of earnings (Katz and Murphy 1992; DiNardo, Fortin and Lemieux 1996; Beaudry and Green 2002; Card and DiNardo 2002), little has been done to examine the consequences of such changes on the cumulative earnings of individuals over 10 or more years. To do so, the wages and salaries received by individuals over the 1973 to 1984 and 1988 to 1999 periods were summed for synthetic cohorts of individuals (see Data sources and definitions). Cumulative earnings were calculated in each 12-year period for both Canadian-born individuals and immigrants. Immigrants counted in the 1973 to 1984 period arrived in 1972 or earlier, and those counted in the 1988 to 1999 period arrived in 1987 or earlier.

Down for young men born here

The results are striking. Over the period from 1973 to 1984, Canadian-born men aged 26 to 35 in 1984 received $289,500 (in 1999 dollars) in wages and salaries (Table 2). However, between 1988 and 1999, their counterparts received only $246,500.  4  In contrast, Canadian-born women of the same age amassed $171,600 between 1988 and 1999, about $20,000 more than between 1973 and 1984.  5  For this age group as a whole, cumulative wages and salaries dropped by roughly $10,000 between the two periods.

Why the decline for 26 to 35 year-old men? The answer is simple. First, young men now stay in school longer than their counterparts did during the mid-1970s, thus decreasing the number of years in which they receive significant wages. Second, once out of school, they are less likely to have a full-time-and therefore relatively well-paid-job than in the past. Third, those who did work full year, full time earned less annually during much of the 1980s and 1990s than their counterparts did previously.

Why the rise for women? Changes in school attendance cannot be an explanation since young women also stay in school longer. The greater presence of women in full-time jobs may have increased their cumulative earnings by giving them greater access to better-paid jobs and also by increasing their annual workhours. Second, for some of those who worked full year, full time-for example, those 30 and over-annual wages and salaries may have increased partly because of the growing tendency of women to be employed in high-paying occupations.

For older age groups, the increasing participation of women in the labour market and their move towards better-paying occupations has resulted in massive increases in cumulative wages and salaries. For instance, Canadian-born women 46 to 55 increased their cumulative wages and salaries by more than $100,000 between the two periods. The increase for Canadian-born women 36 to 45 was about $70,000. In contrast, Canadian-born men saw drops of about $10,000 and $50,000 respectively. As a result, while cumulative wages and salaries of Canadian-born men aged 36 to 55 were three times as high as those of their female counterparts during the 1973-1984 period, the ratio dropped to about 1.8 during the subsequent period.

All the above patterns held for cumulative earnings-that is, wages and salaries plus net self-employment income.

Down even more for young immigrant men

During the 1973 to 1984 period, immigrant men aged 26 to 35 cumulated about $319,000 in wages and salaries, roughly $30,000 more than their Canadian-born counterparts. However, while cumulative wages and salaries of Canadian-born men aged 26 to 35 fell by fully $40,000 between the two periods, they dropped by more than $75,000 for immigrant men. As a result, during the 1988-1999 period, the cumulative wages and salaries of immigrant men no longer exceeded those of their Canadian-born counterparts. Immigrant men aged 36 to 45 also experienced a sharper decline.  6 

The poorer performance of immigrants was also observed among women. While Canadian-born women 26 to 35 enjoyed an increase in cumulative wages and salaries between the two periods, their immigrant counterparts experienced virtually no change. Similarly, cumulative wages and salaries of older Canadian-born women rose by at least $15,000 more than those of their immigrant counterparts.

Once again, all aforementioned patterns held when cumulative earnings rather than cumulative wages and salaries were examined.  7 

Why have cumulative earnings of immigrants evolved less favourably? At least part of the answer lies in the average weeks worked and weekly earnings in the two 12-year periods.

Among men aged 26 to 35, weekly earnings fell roughly 15% for both the Canadian-born and immigrants (Table 3). However, weeks worked by immigrant men fell 7% while those of Canadian-born men remained virtually constant. Therefore, cumulative earnings of immigrant men fell more. Among men aged 36 to 45, the sharper decline in cumulative earnings of immigrants resulted from greater drops in both their weekly earnings and weeks worked. In contrast, Canadian-born women of all age groups enjoyed a greater increase in cumulative earnings than immigrant women, mainly because they increased their number of weeks worked at a much faster pace. Hence, for most age-sex combinations, cumulative earnings of immigrants evolved less favourably, mainly because of differences in time worked rather than differences in weekly earnings.

Wealth implications

Unless declines in cumulative wages and salaries are completely offset by increases in other types of income (for example, Employment Insurance, social assistance benefits, or interests and dividends) or increases in savings rates, individuals are likely to feel an effect on their wealth holdings. For instance, the drop of about $10,000 in average cumulative earnings from one 12-year period to the next was likely a major factor underlying the decline in median wealth of family units whose major income recipient was Canadian-born and aged 26 to 35.  8  Between 1984 and 1999, median wealth of such families fell by about $8,000 (Table 4).  9  Likewise, the drop of more than $30,000 in the average cumulative earnings of young immigrants probably explains a large fraction of the $23,000 decline in median wealth observed among family units headed by young immigrants.

Changes in cumulative earnings over 12 years will likely be less correlated with changes in wealth holdings for older families than they are for younger ones for at least two reasons. First, wealth holdings of individuals aged 46 to 55, for example, depend among other factors on earnings received over at least 30 years (since these individuals were 16 to 25). Second, for a given level of cumulative earnings, consumption patterns are likely to change more over two distinct 30-year periods than over two 12-year periods, thereby weakening the relationship between cumulative earnings and wealth holdings.

Changes in family composition and in the correlation between spouses' earnings may also affect family wealth. The growth in the proportion of unattached individuals over the last 20 years implies that an increasing proportion of individuals do not benefit from the economies of scale associated with cohabitation. This tends to depress savings and, thus, wealth holdings of family units. Furthermore, the growing tendency of highly paid men to be married to highly paid women increases wealth holdings of many dual-earner couples and likely increases wealth inequality.

The role of student debt

The greater debt load of students may also have contributed to a decrease in wealth holdings of young families. However, its effect is limited for two reasons. First, student debt is likely to be carried mainly by postsecondary graduates, who represent only a fraction of young individuals. Second, between 1982 and 1995, average amounts owed at graduation by bachelor's graduates increased only by about $3,700 (in 1999 dollars) for men and $4,000 for women (Finnie 2001). Furthermore, average amounts owed by other postsecondary graduates increased even less. In contrast, cumulative earnings of young Canadian-born men fell by at least $40,000 between the periods 1973 to 1984 and 1988 to 1999. Clearly, the decline in cumulative earnings of young men contributed much more to the decline in wealth holdings of young families than did the growth in student debt.  10 

Other factors must also have played a role. Young individuals now get married later, thereby delaying benefits from the economies of scale associated with cohabitation. This may be offset by some individuals staying longer with parents or cohabiting in other ways.  11 

Homeownership implications

If individuals are waiting until they have a certain level of savings before buying a first house, one would expect the decline in cumulative earnings of young workers to result in a postponement of homeownership. Indeed, this was the case for families with a young immigrant as the major income recipient. In 1984, 55% of these families owned a principal residence (Table 5). Fifteen years later, only 43% of their counterparts did so.  12  In contrast, family units headed by young Canadian-born individuals experienced virtually no decline in homeownership. About half owned a house both in 1984 and 1999.

The fall in homeownership among young immigrants but not among young Canadian-born individuals is consistent with the greater decline in cumulative earnings among young immigrants. The easier access to mortgage loans observed during the 1990s may be one reason why homeownership did not fall among Canadian-born individuals, despite a fall in their cumulative earnings.

Summary

Relative to their counterparts in the mid-1970s, young men today stay in school longer, are less likely to be employed full time and, until recently, received lower earnings (in constant dollars) when working full year, full time. Taken together, these three factors explain why they had much lower cumulative earnings over the 12-year period from 1988 to 1999 than their counterparts did from 1973 to 1984.

Among Canadian-born individuals, this decline was partly offset by the growth of cumulative earnings of young women. Overall, young Canadian-born individuals cumulated roughly $10,000 less between 1988 and 1999 than between 1973 and 1984. Young immigrants fared worse. Their cumulative earnings dropped by more than $30,000. The decline in cumulative earnings was likely a major factor underlying the decline in median wealth of young families. While the growth in student debt also played a role, its contribution was much more limited.

Consequently, as measured by median wealth, the typical young family in the late 1990s had less assets than its counterpart in the mid-1980s.  13  Having less assets reduces the ability of a family to absorb financial shocks in the event of a permanent layoff, unforeseen expenses or health problems, or the decision of one of its members to quit a job. It may also influence the effort expended by an individual to find a new job or the decision to leave a job with unsatisfactory working conditions.

The decline in cumulative earnings may also have affected the decision of some young families to buy a home or have children. Indeed, homeownership fell among families headed by young immigrants, and the average age of mothers at the birth of their first child rose from 25.7 to 27.1 between 1986 and 1996 (Statistics Canada 1999). The extent to which the decline in wealth holdings may have contributed to postponing having children in some families cannot be examined with the data currently available and thus remains an open question.  14 

The increase in the labour force participation of women is well documented. This increased participation and the move towards better-paying occupations have resulted in massive increases in cumulative earnings during a 12-year period for women in their mid-30s and older.

It should be pointed out that these results are based on averages and so may not apply equally to all individuals. For instance, changes in the cumulative earnings of young men may have been different for university graduates than for those with only a high school diploma. Likewise, cumulative earnings of full-year, full-time workers may have evolved differently from those of the 'average' individual. Because many individuals increase their education level or make transitions into and out of full-year, full-time employment over a 12-year period, data based on synthetic cohorts cannot provide accurate estimates of changes in cumulative earnings by education level or by full-year, full-time status. To do so, longitudinal data are needed.

Notes

1. The unemployment rate of young men not attending school full time was 10.4% in 1976 and 11.8% in 2001.
2. To produce earnings trends for the 1975-2000 period, earnings data from the 1996 Survey of Labour and Income Dynamics (SLID) were first compared with those of the 1996 Survey of Consumer Finances (SCF). If earnings in the SCF were, say, 2% higher than those in SLID, the SLID data for the 1996-2000 period were adjusted upwards by 2% in order to produce comparable numbers. The adjustment was done separately for each age-sex combination.
3. The increase in school attendance of Canadian-born young individuals was similar to that of young immigrants. Census data indicate that the percentage of Canadian-born young men enrolled in school full time rose by 15 percentage points, from 39% to 54%, between 1981 and 1996. The percentage of immigrant young men enrolled in school full time increased by 16 points, from 45% to 61%. The corresponding rates for Canadian-born young women in 1981 and 1996 were 37% and 56%, and 39% and 59% for immigrant young women.
4. Cumulative wages and salaries of men aged 26 to 30 and those aged 31 to 35 fell by about $50,000 in both cases.
5. The growth in cumulative wages and salaries of women aged 26 to 35 was due to a solid rise of $40,000 for women aged 31 to 35. Cumulative wages and salaries for women aged 26 to 30 did not rise.
6. Although immigrant men aged 46 to 55 experienced a somewhat larger drop in cumulative wages and salaries than their Canadian-born counterparts, they experienced a much smaller decrease in cumulative earnings.
7. Since immigration status is unknown for some respondents in the Survey of Labour and Income Dynamics, it is worth investigating the extent to which the poorer performance of immigrants could be affected by missing information. Data from the Survey of Consumer Finances indicate that immigrants aged 24 to 43 in 1997 (therefore aged 26 to 45 in 1999) who came to Canada in 1987 or earlier represented between 7% and 11% of the selected population in 1997. Accordingly, the cumulative earnings of immigrants were calculated under two assumptions. First, all individuals who did not know their immigration status were considered immigrants who came to Canada in 1987 or earlier. Second, of all individuals who did not know their immigration status, only those in the top decile of the distribution of earnings were considered immigrants who came to Canada in 1987 or earlier. Results based on the first assumption strengthened the study's finding of the poorer performance of immigrants 26 to 45 or younger (in 1999) in respect to cumulative earnings. Under the second assumption, cumulative earnings of immigrant men aged 26 to 45 fell at least $14,000 more than those of their Canadian-born counterparts, and the cumulative earnings of similar-aged immigrant women rose at least $10,000 less than those of their Canadian-born counterparts. Thus, the finding of poorer performance by immigrants aged 26 to 45 does not appear to be a result of excluding observations of unknown immigration status.
8. Ideally, one would like to correlate changes in cumulative earnings with changes in wealth holdings of individuals. Since the 1984 Assets and Debts Survey and the 1999 Survey of Financial Security measure wealth only at the family level, this analysis is limited to changes in wealth holdings of family units.
9. Ideally, the present discounted value of earnings (at the beginning of each period) received over each 12-year period should be computed and the resulting changes in wealth holdings estimated-assuming no changes in consumption expenditures between the two periods. Using a discount rate of 3%, the present discounted value (in 1973) of earnings received by Canadian-born men aged 26 to 35 amounted to $252,700 for the 1973-1984 period, compared with $213,900 (in 1988) for the 1988-1999 period. The corresponding amounts were $128,700 and $147,600 for their female counterparts, and $190,800 and $180,800 overall. If one assumes that the extra earnings were saved and generated a real rate of return of 3%, then differences in the present value of discounted earnings would generate changes in average wealth holdings of -$53,600 for the men, +$26,200 for the women, and -$13,800 overall.
10. While the increase in school attendance may lead to an increase in student debt and a decrease in the number of years young individuals will be able to work full time before they reach, say, age 30, more education may increase their earnings in the longer term.
11. Card and Lemieux (1997) showed that between 1971 and 1994, the proportion of youth living with their parents rose more in Canada than in the United States. They concluded that the greater increase observed in Canada was related to poor labour market conditions in this country.
12. The decrease is statistically significant at the 10% level (two-tailed test).
13. Part of the decrease may have resulted from some young dual-earner couples' deciding to hold less in precautionary assets because the risk of income loss is spread over two earners.
14. Admittedly, changes in values and lifestyles have played an important role.

References

• Beach, Charles M. and George A. Slotsve. 1996. Are we becoming two societies?: Income polarization and the myth of the declining middle class in Canada. Toronto: C.D. Howe Institute.
• Beaudry, Paul and David A. Green. 2002. Changes in U.S. wages 1976-2000: Ongoing skill bias or major technological change? National Bureau of Economic Research (NBER). Working paper no. 8787. Cambridge, Mass: NBER.
• Card, David and John DiNardo. 2002. Skill-biased technological change and rising wage inequality: Some problems and puzzles. National Bureau of Economic Research (NBER), Working paper no. 8769. Cambridge, Mass: NBER.
• Card, David and Thomas Lemieux. 1997. Adapting to circumstances: The evolution of work, school, and living arrangements among North American youth. National Bureau of Economic Research (NBER). Working paper no. 6142. Cambridge, Mass: NBER.
• DiNardo, John, Nicole Fortin, and Thomas Lemieux. 1996. "Labor market institutions and the distribution of wages, 1973-1992: A semi-parametric approach." Econometrica 64, no. 5 (September): 1001-44.
• Finnie, Ross. 2001. "Student loans: The empirical record." Canadian Journal of Higher Education 31, no. 3: 93-142.
• Heisz, Andrew, A. Jackson, and Garnet Picot. 2002. Winners and losers in the labour market of the 1990s. Analytical Studies Branch, Research Paper Series no. 184. Ottawa: Statistics Canada.
• Katz, Lawrence F. and Kevin M. Murphy. 1992. "Changes in relative wages, 1963-1987: Supply and demand factors." Quarterly Journal of Economics 107, no. 1 (February): 35-78.
• Morissette, René, John Myles, and Garnett Picot. 1994. "Earnings inequality and the distribution of working time in Canada." Canadian Business Economics 2, no. 3: 3-16.
• Statistics Canada. 1999. Vital statistics compendium, 1996. Catalogue 84-214-XPE. Ottawa.

René Morissette is with the Business and Labour Market Analysis Division. He can be reached at (613) 951-3608 or perspectives@statcan.gc.ca.

[ Highlights ] [ Main menu | Editor's corner | More news | Contact us | Survey information | Back issues ] [ Statistics Canada home page | In depth | Français ]

© Statistics Canada - Conditions of  use Published: 2002 11 19