Much has been written about income inequality. You can hardly find a current affairs magazine, newspaper or television program that has not dealt with income inequality recently. In December 2013, President Obama called America's growing income gap "the defining challenge of our time." Whether the income gap is a problem or a "challenge" is something for another article. However, the question of whether the income gap, at least in Canada, is indeed growing is something I wish to address here.

In a new study for the Fraser Institute, I and my colleagues examine the measurement of income inequality and how changes in the definition of income or the definition of the income recipient can have a major impact on the results. Indeed, and this is a principal finding of the study, we find that after-tax income inequality for individuals has actually declined over the past three decades for which we have publically available data.

It is important to explain why differences in income definition and in income recipient matters. A great many studies of income inequality focus on "earnings"—the wages and salaries that employed people receive and any net income from small business (non-corporate) enterprises. The problem with earnings is that today we have proportionately more people with zero earnings than was the case three decades ago. Despite the fact that there are more two-earner families, we also have more seniors with no earnings; more students living on their own (with no earnings); and more government transfer (i.e. welfare) recipients with little or no earnings. The more people with zero earnings, the greater the measured level of inequality. As well, the more million dollar earners (business, sports and entertainment superstars), the greater is measured inequality.

All of this has happened, so we are not surprised to see greater earnings inequality. From 1982 to 2010, earnings inequality rose by 24 per cent using the widely known Gini coefficient as our indicator of inequality.

Further, families are smaller than they were three decades ago and this means that, on average, there are fewer people with which to share family income. So, if we define income as after-tax income and then adjust for family size, we get a different picture.   Today the income inequality of economic families has only increased by 10 per cent over the past three decades. And, if we look at individuals (any adult with any income at all) and their after-tax (or disposable) income, we show that inequality has actually declined by 4 per cent over the past three decades, again using the Gini indicator.

These findings have several implications: First, they show that inequality measurements are quite sensitive to the choice of income recipient and the choice of income definition.  Second, they show that looking at individuals rather than families, we find that income inequality has, in fact, declined a bit since the early 1980s—a result that is confirmed using American data as well.

Finally, these results suggest that caution is advisable when we look at studies of income inequality. Dramatic differences can be obtained if we change our initial definitions and indicators.

Income inequality is a very complex matter. Substantial social and economic changes have happened over the past few decades and it would be surprising if measured inequality was not impacted as well. The fact, however, is that it has not increased very much—if we look at economic families—or at all, if we look at individual incomes.

Of course, income inequality is not the best measure of how the living standards of Canadians changes. For that, we use the goods and services that people actually consume, which is the subject of a future study on inequality to be published by the Fraser Institute.