Conclusions and Review
This study has used income tax longitudinal data over the period 1982-96 to
look at business-cycle effects on short-run earnings mobility patterns of
workers in the Canadian labour market. The approach has involved looking at
the proportion of workers in six earnings intervals over time and at dynamic
one-year transition matrix summary measures of earnings mobility over time.
Four major results have been found.
First, there are major cyclical effects in changes in the degree of earnings
polarization and in general upward or downward shifts of the earnings
distribution over the 1982-96 period, and these effects are most markedly
concentrated in the two recessions over this period, particularly between 1990
and 1992. Weak labour markets increase earnings polarization significantly
with generally much stronger effects for men than for women. In terms of
general distributional shift, quite opposite cyclical patterns are observed.
Weak labour markets shift the earnings distribution for men down
significantly, but the higher unemployment shifts the women’s earnings
distribution up relative to the men’s as women’s earnings are relatively less
sensitive to weakened labour markets.
Second, when these cyclical patterns are observed across age groups, the
cyclical effects across all age groups for each gender generally mirror that for
the gender as a whole, but the strength of the effects differs across age groups
with the strongest effects typically occurring among entry and younger
workers. Higher unemployment increases polarization rates across all age/sex
groups. Among males, it is the cyclical sensitivity of the earnings of workers
at the lower end of the distribution which is driving the results. Among female
workers, it is generally the upper end of the earnings distribution where the
greatest cyclical sensitivity occurs. Entry and younger workers of both gender
show far the greatest sensitivity to unemployment rate changes, while older
workers show generally the weakest effects.
Third, year-to-year dynamics of workers’ earnings are summarized in
terms of measures of earnings mobility, the probability of moving up the
distribution across earnings intervals, the probability of moving down the
distribution across earnings intervals, and the net probability of moving up
across earnings intervals. It is found that higher unemployment decreases net
earnings mobility significantly for men by reducing the probability of moving
up the earnings distribution and increasing the probability of moving down
about equally, so that the average net probability of moving up is significantly
decreased. For every percentage point increase in the (adult male)
unemployment rate, the average net probability of moving up is estimated to
Cyclical Changes in Short-Run Earnings Mobility in Canada
481