Why measurement matters in income inequality

Oversimplifying this issue to arrive at predetermined results risks making the situation worse

By Jason Clemens
and Chris Sarlo
The Fraser Institute

Canada, like most industrial countries has its fair share of economists and politicians arguing that Canada has an inequality crisis requiring large-scale, even unprecedented government intervention to solve. More often than not the issue of inequality and how we measure it is grossly over-simplified.

The reality of income inequality is far more complicated than such analyses and hyperbole suggest. For example, both the level of and growth in income inequality is incredibly sensitive to the definition of income used.

Jason Clemens
Jason
Clemens

Many studies of income inequality use “earnings” to measure income. Earnings is a particular and rather narrow measure of income, however, since it only includes wages and salaries and any net income received from unincorporated businesses. As such, earnings as a measure of income ignores government transfers, progressive income taxes, and tax credits that benefit lower-income individuals and families. In other words, “earnings” ignore almost all of the current interventions undertaken by government to reduce income inequality.

A recent study examining the sensitivity of income inequality to how income is defined demonstrates how the results can change when different definitions are used. In 2010, the latest data included in the study, the top 10 per cent of families received 36.8 per cent of all earnings in Canada. This is a 34.2 per cent increase in the share of earnings received by the top 10 per cent of families since 1982. Such results are the hallmark of studies calling for large-scale intervention by government.

However, if we include the policies that governments in Canada have already implemented – income transfers for instance – and measure after-tax income rather than just earnings (and adjust for the size of the family to ensure it doesn’t influence the results) both the level and growth of income inequality are significantly lower.

Chris Sarlo
Chris Sarlo

The share of after-tax income received by the top 10 per cent of families in 2010 was 25.3 per cent. In other words, the level of inequality as measured by the share of income received by the top 10 per cent of families was more than 30 per cent lower than if just earnings are used to measure income.

Similarly, the growth in income inequality is significantly lower when after-tax income adjusted for family size is used. Specifically, the growth in income inequality falls from 34.2 per cent when earnings are used to just 12.9 per cent (1982 to 2010). This is critically important since after-tax income incorporates many of the mechanisms currently used by government to reduce income inequality.

The measurement of inequality is also complicated by issues such as whether individuals or families are analyzed, whether income is really the best measure of one’s standard of living, how goods and services have improved over time (which isn’t largely captured in the data), and how our circumstances naturally change over the course of our lives.

All of these factors need to be considered when we analyze inequality. This is not to say that inequality isn’t an important issue but rather that it is a complicated one. Oversimplifying this complex social and economic issue to arrive at predetermined results risks making the situation worse. Understanding how it’s measured is one crucial step in understanding the reality of inequality.

Jason Clemens and Christopher Sarlo are co-authors of the recently released Income Inequality Measurement Sensitivities, which is available at www.fraserinstitute.org.

Jason and Chris are Troy Media Thought Leaders. Why aren’t you?

© Troy Media


income inequality

The views, opinions and positions expressed by columnists and contributors are the author’s alone. They do not inherently or expressly reflect the views, opinions and/or positions of our publication.

You must be logged in to post a comment Login