Voir en Français

New data provides full story on income gap in Canada

A version of this commentary appeared in the Globe and Mail, iPolitics and Hill Times

Canadian Dollar CloseupStatistics Canada released data recently on the incomes of the top 1 per cent of tax filers, and compared these to the incomes of the remaining 99 per cent. Not surprisingly, this small segment of the population receives a disproportionate share of the pie  about one-tenth of all individual income, with a median income at $283,400, about 10 times the median of the bottom 99 per cent.

Statscan has further provided comparable data going back to 1982, and not only by province and for the five largest cities, but also for men and women separately. Statscan has also used its CANSIM data dissemination tool (now free of charge) to provide a tremendous range of much more detailed breakdowns  enough to keep data junkies busy for days and weeks.

For example, the threshold to be in the top 0.1 per cent in terms of after-tax income, at $2.2-million, is almost 14 times as high as the threshold for the top 1 per cent. Virtually all the news coverage so far has discussed only the numbers in the Statscan text for the data release, which referred only to before-tax income.

What you may not have noticed is that this is the first time Statscan has ever produced such data as part of its standard suite of statistics. One of the challenges for a national statistical agency is to stay relevant to the issues of the day. The Occupy movement has been news for more than a year, and we even have the elite of the business community in Davos recently putting income inequality at the top of their agenda. So these new data are most welcome.

Issues related to income inequality have been bubbling in the background amongst economists for decades. Interest amongst official statisticians recently reached a high point when former French prime minister Nicholas Sarkozy appointed a blue ribbon (many Nobel laureates) commission to examine serious gaps in national statistics.

The three major areas addressed by the Stiglitz/Sen/Fitoussi commission report in 2009 were the environment, well-being, and incomes  including especially income inequality.  The commission’s reasoning on the latter was that most people cannot relate personally to GDP statistics, not least because economic growth has not been spread evenly. Middle income individuals have experienced stagnant incomes while GDP has grown over past decades. The commission’s advice, therefore, was for official statistics to provide more detail on the distribution of income for individuals and families.

Statistics Canada has a long and exemplary history of producing data on incomes and income inequality. For example, in parallel with leading work by Mollie Orshansky in the U.S., Jenny Podoluk in a 1967 census monograph introduced the low income lines that are widely used as poverty indicators. Data on the numbers of individuals and types of families with low incomes have been published annually and in detail ever since. The same household surveys used for these data are also used to provide data on those with middle and upper-middle incomes. But these surveys were never sufficiently reliable to provide data on the top 1 per cent, so such data were not published until this past Monday.

Columnist Terence Corcoran wrote a blistering and unwarranted attack on Statscan for pandering to the Occupy movement by publishing data on the rich, as if these were the only income distribution data published. But if anything, the story is the opposite. Statscan is to be commended for balancing its long-standing statistical series on those with low  and middle incomes with these newly available data on those with high incomes.

Michael Wolfson is an adviser with EvidenceNetwork.ca, and Canada Research Chair in population health modelling/populomics at the University of Ottawa. He is a former assistant chief statistician at Statistics Canada, and has a PhD in economics from Cambridge.

January 2013