Skimming the Cream

In September 2011, the tents went up in New York City and by year’s end the Occupy Movement had spread to over 950 cities in over 80 countries. Millions of people across the world were on the street to protest rising income inequality and the power of money in democracy. U.S. president Barack Obama declared income inequality the defining issue of our time.

But the movement has not yet produced the fundamental social change demanded by the activists. In The End of Protest: A New Playbook for Revolution, Micah White, one of the movement’s co-creators, argues that it should continue in multiple countries sharing a common agenda, but should shift from street protests to electoral politics.

Canadians were part of this international movement. There were Occupy protests in dozens of cities. But each country’s experience is different: the nature of inequality and how it has changed is different; the politics of protest and of electoral platforms are different. These two books—Income Inequality: The Canadian Story, edited by David A. Green, W. Craig Riddell and France St-Hilaire, and The Inequality Trap: Fighting Capitalism Instead of Poverty, by William Watson—help tell the Canadian story.

After we leave the street and stop relying on protestors’ signs for our understanding, we must look at the evidence, at the data on income and its distribution. It turns out that describing and understanding income inequality in a country is not a simple task, and the evidence allows different interpretations. Depending on your value system, your assessment of the fairness of the distribution will differ. The analysis takes us into the arcane world of definitions, data and statistics. But let us follow the authors and go there.

We start with how much income there is to distribute. To a close approximation, total income is measured by gross domestic product. Canada is a high-income country. We have lots of income to go around.

And over the post-war period, real GDP has been growing—the pie has been getting bigger. Moreover, national income has been growing faster than population, so real GDP per capita has risen steadily.

Sometimes, though, this growth is interrupted by recessions—the pie shrinks, on average, and people’s incomes fall. Since the early 1950s, Canada has had three severe recessions: in the early 1980s, the early 1990s, and from 2008 to 2009, during the financial crisis. The last two recessions were especially deep and long. In the early 1990s, real GDP per capita fell 5.1 percent and took five years to return to its pre-recession high. During the most recent recession, real GDP per capita fell 4.6 percent and it has taken six years to reach the previous high.

So how has this growing income (and its periodic shrinkages) been distributed among families and individuals in Canada?

The Green, Riddell and St-Hilaire volume is the fifth volume in a series aptly titled “The State of the Art.” The volume, a joint initiative of the Institute for Research on Public Policy and the Canadian Labour Market and Skills Research Network, is a collection of commissioned papers by leading Canadian economists. The papers are first rate, rich in data and analytically sophisticated. Together they are the best source of data and analysis of income inequality in Canada: they summarize trends in inequality, examine the causes and discuss the possible role of public policy to reduce inequality. The papers seek to analyze and understand, rather than to persuade. Although many papers are not likely accessible to a general reader, the overview essay by the editors and the concluding essay by Keith Banting and John Myles (a political scientist and a sociologist) are. These two essays are must-reads for anyone who seeks a nuanced understanding of the evidence and the politics.

Banting and Myles write: “Although there is a broad academic consensus on the facts of the matter, the inequality surge has generated vigorous political debates about how the ‘facts’ should be read and about their import. It is possible to tell different stories about the new inequality by adopting different data sources or time periods, or by focusing on trends in different parts of the income distribution.” They go on: “The country has become engaged in a vigorous struggle to define or ‘frame’ the new inequality and the social stresses it brings in its wake. What is happening? Why is it happening? Is it a policy problem? There are multiple answers to these questions. And the conflicting interpretations pervade debates in the media and legislatures, as well as the proposals that parties present to the electorate.”

They suggest that disagreement can be stylized as a debate between the “inequality Cassandra” and “the inequality denier.” Neither term is accurate or fair.

Rather, it is a debate between essayists (or polemicists) in the best sense of these terms, between those of different political ideologies who seek to persuade us, using the ancient art of rhetoric, selectively introducing fact in verbal flourish in service of their cause. It is a debate epitomized by columnists of the Toronto Star and the National Post.

After reading Income Inequality, Star columnist Carol Goar wrote: “It took 35 years to turn Canada into one of the most inequitable countries in the world. It will take almost as long—plus money and discipline—to reverse the trend.”

William Watson, a National Post columnist and professor of economics at McGill, reads the data and analysis very differently. As the inside front flap of his book says, he argues “that focusing on inequality is both an error and a trap. It is an error because much inequality is ‘good,’ the reward for thrift, industry, and invention. It is a trap because it leads us to fixate on the top end of the income distribution, rather than on those at the bottom who need help most. In fact, if we respond to growing inequality by fighting capitalism rather than poverty, we may end up both poorer and less equal.” His is not inequality denial. He asserts—correctly—that our response to income inequality depends on our value system.

To study the distribution of total income, we cannot use the national income accounts data; rather, we must turn to surveys of households, which ask the income of all families and unattached individuals. The surveys can tell us market income, total income (that is, market income plus government transfers such as the old age pension) and after-tax income.

Each income measure is useful depending upon the purpose of the analysis. If we want to understand what has been going on in the labour market, the analysis would use market income. But when the focus is on the fairness of the income distribution, the analysis would use after-tax income because the tax and transfer systems are designed specifically to affect the fairness of how much money people have to spend.

From these surveys, we know how much income each family and individual has. But how are we to characterize the distribution across all the thousands of households?

The most intuitively understandable way to characterize the distribution of income across households is to first arrange the households from lowest income to highest income and then look at the share of total income going to the lowest ten percent, the next ten percent and so on; this is the decile distribution of income. If total income were evenly distributed and every household had the same income, each decile would have ten percent of the income.

The data can be further refined to recognize that households differ in size and there are economies of scale (as the adage tells us, two can live as cheaply as one) so we should look at size-adjusted income. In Statistics Canada’s terminology, this is adult-equivalent-adjusted income. So analysts look at the decile distribution of AEA income.

There are other ways to represent the distribution of income than by looking at deciles. Some measures represent the whole distribution in a single number; one such measure is the widely used Gini coefficient. As before, households are arranged from lowest to highest, and then a calculation is made based upon the cumulative percentage of AEA income received by the cumulative percentage of the population. For complete equality the Gini is zero and for complete inequality the Gini is one. Both books summarize the data on income distribution by looking first at the Gini coefficients; Watson’s book offers a clear discussion of how a Gini coefficient is calculated.

The Canadian story of income inequality has a number of themes. And much of the story can be told through three figures: one looking at the overall distribution of income (Gini coefficients), another looking at the lower end of the distribution (poverty) and the third looking at the top end (the one percent).

The first theme is that the distribution of market income has become more unequal. From 1976 to 2011, the AEA Gini rose from 0.365 to 0.446. This long-term trend is unambiguous.

But the distribution has not continuously been getting more unequal. It rose during the recessions of the early 1980s and ’90s, but curiously not during the financial crisis recession. And market income inequality has not grown since the mid 1990s.

There are several standard explanations offered for rising inequality of market income: skill-based technological change and globalization tend to increase earnings at the top of the distribution and depress earnings at the bottom of the distribution; institutional factors such as declining levels of unionization tend to depress earnings at the bottom of the distribution. These all seem to be at work in Canada. Watson asks us to push further, to consider other reasons why the Gini might rise. For example, if female labour force participation rises and more women get university degrees, and if high-income women tend to marry high-income men, the distribution of income will become more unequal—the Gini will rise. He asks: should we worry about rising inequality due to these changes in the role of women?

The second theme is that after-tax income is more equally distributed than market income. It too has become more unequal, although less so. The Gini for after-tax income rose from 0.300 to 0.313.

Should we worry about a current Gini of 0.313? The Gini coefficient has the advantage of capturing the distribution in one number, but it is difficult to envisage what this number implies. Data on the decile distribution are more intuitively understood. In 2013, the top decile had 23.7 percent of after-tax AEA income, the fifth decile had 8.2 percent, and the bottom had 2.5 percent. The top decile has slightly less than ten times more than the bottom decile. Are we doing enough to lower it? The tax-transfer system has significantly redistributed income: the average AEA income in the top decile has fallen from $146,800 to $113,200 and the average income in the bottom decile has risen from $1,300 to $12,000. The answers depend upon your value system.

Our answers are likely affected by what has been happening to the actual incomes of each decile. Have the rich been getting richer while the poor got poorer? The answer is no. From 1976 until the mid 1990s, the AEA after-tax real income of all deciles was quite level: no one was getting better off. Then real incomes for all began to rise (although somewhat faster for the top decile, hence the rise in the Gini). And since 2001, they have continued to rise for all deciles but at comparable rates. Since the mid 1990s, the average AEA real after-tax income of the lowest decile has risen 25 percent. The poor have not been getting poorer. And measures of relative poverty (the percentage of the population with incomes below one half of the median), what Statistics Canada calls the low-income measure, have been steady.

Another way to look at the situation of the poor is to examine absolute poverty. Strikingly, and against the narrative of rising inequality, poverty has gone down. We have come to recognize the Statistics Canada low-income cut-off as the measure of absolute poverty. The LICO after-tax AEA measure of poverty spiked during the early 1980s and ’90s recessions, although not during the financial crisis recession, but overall has declined since the mid 1990s. The poverty rate has fallen from 15.2 percent to 8.8 percent. Vulnerable groups—aboriginal persons, recent immigrants, youth not in school, lone parents and unattached older individuals—have higher rates of poverty. But the poverty rates of these vulnerable groups have declined even more than in the general population.

And the middle class, the middle deciles of the distribution, has not been suffering. Its share of total income has been about constant. Its average real income was stagnant until the mid 1990s (as was everyone else’s), but has risen strongly since.

How significant are these patterns? Some would say the overall rise in income inequality is deeply troubling; others that the rise is real but modest.

But the data do not support the claim of exploding inequality evoked by some commentators.

There is one large and enduring change in the distribution of income, but it does not get captured by the Gini coefficient or the decile data: the rise in the share of market income going to the top one percent (and within this one percent, the share going to the very, very top, the top decile of the one percent). To study this very top end of the distribution, we cannot rely on household surveys but must use personal income tax records. Thus these are data on individuals (not individuals grouped into households as above).

The share of market income going to the top one percent fell steadily over the post-war period, from about 11.5 percent to a low of about 7.5 percent in the early 1980s. Since then it has risen, topping at over 13 percent just before the financial recession, and then falling back to about 11.5 percent. (In contrast in the United States, the share of the one percent has continued to rise.)

The rising share of the one percent is unquestionably the most significant change in the distribution of income in Canada over the last 35 years.

Watson challenges those who protest the one percent by arguing that high income due to extraordinary talent, hard work or innovation is good inequality. He unabashedly titles his second chapter “The Deserving Rich,” using people such as Steve Jobs and Sidney Crosby as examples. Bad inequality “occurs when people prosper from morally dubious or even illegal actions or from some unfair advantage they have contrived for themselves or conspired with others, often through governments, to acquire.” The latter is sometimes called rent seeking.

Green et al. offer a different and more balanced approach to understanding the rise of the one percent. They note that most members of the one percent are not the Steve Jobses or Sidney Crosbys of the world, but are senior executives and those working in the financial and business services sectors. There are competing theories in economics about how compensation is determined in the market. One, based in models of perfect competition, concludes that workers and executives are paid the value of their marginal product. Executives get high salaries because their contribution is very high. The other, based in models of imperfect information, asserts that workers/executives have some -discretion/power to shift a portion of total profit toward themselves. Wages beyond their contribution are called rents. These models conclude that the rise in executive pay is not because of rising contribution, but because executives have secured rents for themselves. (And yes, these rents are what Watson would have to call bad inequality.)

There is no absolute consensus, but much evidence points to the rise in executive pay as being due to increasing extraction of rents, especially through stock-option compensation packages.

The Canadian story is filled with puzzles.

Over the last decade, the real incomes of all deciles have been rising, but there is a longer-term upward trend in the Gini coefficient. Absolute poverty had been falling and the share going to the top one percent has risen. How should we characterize these patterns of inequality?

And we should remember that economic growth in Canada usually increases the incomes of all deciles. Recessions usually cause inequality to spike upward. Recessions hurt everyone, but hurt the vulnerable more. As Tony Fang and Morley Gunderson write in Income Inequality, “a growing, full-employment economy and greater labour demand remains the most effective first line of defence against poverty, particularly among vulnerable groups.”

The Occupy Movement forced income inequality to the top of the political agenda. Protestors called for fundamental social change. Watson took this call seriously, seeing it as an attack on capitalism. If we attempt to alter income distribution at the top end, he argues, like Friedrich Hayek did, we are on the road to serfdom.

But Watson need not worry. Canadian political parties did not offer radical social change. Is this because those with high incomes are too powerful? But does money actually corrupt Canadian democracy? Perhaps it was because voters worried more about a recession and wanted to focus on income growth rather than income distribution. Or was it because the actual rise in income inequality was judged by voters to be much less than the protestors claimed?

Have we misunderstood the reality of the Canadian experience? This is certainly possible. In today’s world, movements for radical social change such as the Occupy Movement are international. The actual experience of each country can get lost in the meta narrative. Canada is especially prone to accepting the U.S. story as our own (ours is just a little less extreme).

The inequality issue in Canada did move into electoral politics, as Micah White recommended. But during the 2015 national election, none of the major parties talked much about income inequality or about poverty. Instead, they talked about getting the economy growing and helping the middle class. The victorious Liberals promised growth, modest redistribution to the middle class and a change in tone, a return to “sunny ways.”

Income inequality, the “defining issue of our time,” produced a change in tone but little fundamental. Curiously neither book makes any mention of the other defining issue of our time—climate change. This is a serious gap in our understanding. How (or if) we make the transition to a low-carbon economy will powerfully affect both income growth and income distribution in the decades ahead.