Applied Research Bulletin - Volume 4, Number 1 (Winter-Spring 1998) - May 1998

Is It Possible to Move Up on the Income Scale?

Previous

Contents

Next

A number of recent studies have shown that over the past ten or fifteen years, Canada has experienced a significant increase in the inequality of individuals' annual earnings. Does this increase mean that there has been a freeze or a decline of opportunities in the labour market? Does it imply that earnings inequalities will become greater in the long term?

Using the new Longitudinal Administrative Databank (LAD), Ross Finnie offers some answers to these questions. In a series of studies, Finnie has attempted to measure and characterize the mobility of individual earnings from the early 1980s to the early 1990s. The results considered in this Bulletin concern all male and female workers between 20 and 64 years of age (except full-time students, self-employed persons and persons who declared less than $1,000 in annual income from employment).

By examining the movement of workers from one earnings bracket to another, Finnie shows that the Canadian economy is characterized by a high degree of earnings mobility over time. According to his results, 59 percent of the people who were in the labour market between 1982 and 1992 moved from one income level to another during that period. Nonetheless, earnings stability (or immobility) is significant since 41percent of working Canadians remained in the same earnings bracket during this ten-year period.

Earnings for the most part tend to rise

Finnie observes that, in the long term, earnings mobility tends to follow an upward curve. As a group, Canadian workers are more likely to experience an increase in their real earnings than a decrease. Between 1982 and 1992, the average probability of an increase in real earnings was 39.5percent, while the probability of a decline was only 19.2 percent. Moreover, the highest probability of upward mobility occurred at the lower end of the income from employment scale, since over the ten-year period 71.8percent of workers in the first (lowest) earnings quintile experienced an increase in their real earnings.

Text version

The overall picture that emerges from this examination of the distribution of income from employment indicates that the circumstances of individuals who initially are at the lower end of the scale tend to improve over the years, while individuals in upper brackets tend to maintain their position. The vacuum thus created at the bottom of the distribution scale is filled by new arrivals to the labour market.

Moreover, the extent and direction of real earnings mobility rather closely reflect the normal trends that can be observed in the labour market over its life cycle. For example, the probability of upward earnings mobility decreases with age, while the probabilities of downward mobility and stability increase. At the bottom of the earnings scale, upward mobility still remains very substantial and is always more common than downward mobility, regardless of the age of the individuals concerned. For example, male workers between 35 and 54 years of age who were in the second or third quintile had an average probability of upward mobility of approximately 47 percent, while their likelihood of experiencing a decline was only 30 percent.

Disturbing trends

Over the years, trends in individual earnings mobility have changed. A comparison of the periods 1982-1987 and 1987-1992 shows that for all individuals, upward mobility declined by approximately 7 percent from the first period to the second, while the probability of downward mobility increased by about 15 percent. The probability of remaining within the same earnings bracket increases at the bottom of the scale, and declines slightly in the upper brackets. These results seem to indicate that it is becoming harder to move upwards in the scale of earnings, and individuals at the top of the distribution generally seem to have more difficulty in maintaining their position. These results must, however, be interpreted with caution, since they in part reflect economic conditions that differed from one period to the other. Moreover, the results do not contradict the fact that upward mobility remains generally more probable than downward mobility.

Text version

On the other hand, the trend observed for workers between 20 and 24 years of age seems rather disquieting. We find that between the early 1980s and the early 1990s, the upward mobility of these young workers declined considerably in comparison to the mobility of their older counterparts. This is particularly true for young male workers.

An interesting fact noted by Finnie is that income mobility trends for men and women are diametrically opposed over time. This opposition translates in concrete terms into a convergence of probabilities of transition for the two groups. More precisely, it is observed that between the early 1980s and the early 1990s, male workers suffered a decrease of approximately 12 percent in the probability of increasing their real earnings, and the probability of a decline for this group rose by 21 percent. For women, by contrast, the probability of an increase in real earnings rose by approximately 9 percent, while the probability of a decline fell by about 10 percent.

The positive impact of mobility on annual earnings disparities

One of Finnie's more important observations is that mobility tends to have an equalizing effect on the distribution of individuals' earnings. Finnie noted that long-term inequality was nearly nine percent lower than the average disparity observed on an annual basis. In other words, the commonly used measurement of annual inequality of earnings overestimates long-term inequality, because of mobility. An idea of the importance of the mobility factor can be obtained by noting that the decrease in inequality attributable to mobility is greater than the disparity of the indices. The annual inequality observed between the peak and the trough of the business cycle reflects a disparity of about five percent.

However, the extent of the impact of mobility on earnings disparities remains constant from the beginning to the end of the 1980s. This means that the observed increase in the inequality of individual earnings during this decade was not offset by an increased mobility of earnings. The rise in short-term inequality was thus followed by a more or less similar increase in long-term inequality.

It remains to be seen whether the trends described here continue to develop in the 1990s.

Note:

The Gini coefficient is a measure of income inequality based on the Lorenz curve. When the Gini coefficient is zero, all the units have the same income. When the Gini coefficient is one, one unit has all the income, and all the other units have zero income. Consequently, the higher the Gini coefficient the greater the inequality in the distribution of income.

Text version

Previous

Contents

Next