Dion’s misuse of poverty stats
In the poverty debate, those on the right believe the focus should be first on alleviating the plight of people at or below the subsistence level. Instead, those on the left insist that government should work at assisting people in so-called relative poverty, defined, in various ways, as people who have incomes or spending at certain levels below that of average Canadians. This broader poverty focus of the left is motivated by the socialist philosophical desire to flatten the income distribution, or, in other words, to make incomes more equal.
So, I suppose it is not surprising that the federal Liberals have announced that, if elected, they would fight a relative poverty war. In a Nov. 9, 2007 speech to the Learning Enrichment Foundation, Stéphane Dion pledged to reduce, over five years, the number of Canadian persons living in households with incomes below the so-called after-tax LICO (or low income cut-off ) by 30% (from 10.8% of the population to 7.6%), and reduce the number of children living in households with incomes below the LICO by 50%. Mr. Dion choose this LICO measure of poverty because, in Mr. Dion’s words, “Most experts use Statistics Canada’s low-income cut-offs to measure poverty.” However, I’m very sure the “experts” Mr. Dion was referring to did not include those people actually living in real poverty. Nor do these “experts” include those who invented the LICO concept in the first place, and measure it year in and year out.
“At the heart of the debate [about poverty] is the use of the low-income cutoffs as poverty measures, even though Statistics Canada has clearly stated, since their publication began over 25 years ago, that they are not,” Statistics Canada says.
In fact, of course, the “experts” Mr. Dion refers to are those on the left whose interest it is to define poverty as broadly as possible.
In his speech, Mr. Dion said that he was “embarrassed” by the poverty situation in Canada. No doubt this embarrassment is due to the fact that the LICO measure shows only a small decrease in the percentage incidence of poverty, or, when factoring in population growth, a 34% increase in the number of people in poverty since Pierre Trudeau’s 1968 “Just Society” speech. This increase in LICO poverty has occurred in spite of the fact the Liberals were in power for all but nine of the years between 1968 and 2005 (the latest year for which LICO figures are available).
But consider the LICO definition for one moment, and the resulting LICO poverty measure that the Liberal “experts” are employing. The LICO measures the income level at which people spend 20 percentage points more of their income than the average person does on food, clothing and shelter. For instance, if the average person spends 43% of their income on these three items, and if people with after-tax incomes of $25,000 spend 63% (43%+20%) of their incomes on these three items, the after-tax low-income cut-off would be $25,000. Using the LICO to measure poverty, one would then argue that all people with after-tax incomes below $25,000 are living in poverty. Though the definition is relatively straightforward, LICO measurement is anything but — requiring surveys of people’s expenditure, income estimates, econometric curve drawing and extrapolation. Statistics Canada also estimates LICOs for various-sized households and for various-sized population centers.
But here’s the kicker with LICO. Over the decades since the late 1960s, in spite of Canadians eating more, having bigger wardrobes and living in better accommodation, the improvement in Canadian incomes has meant that the fraction of aftertax income that the average Canadian spends on food, clothing and shelter has fallen. Now, before you read the next sentence, what would you think this fact would mean for the level of poverty in Canada? If you answered “nothing” or “lower it,” you would be quite wrong — at least according to the LICO measure. In actuality, the way in which LICO poverty is defined, as the average Canadian spends smaller and smaller fractions of his or her income on food, clothing and shelter, the LICO measured rate of poverty increases. As the result of changing average Canadian spending patterns between 1969 and 1992, the extent of LICO-measured poverty increased by about six percentage points. Since Statistics Canada currently still uses the 1992 expenditure patterns to measure LICO, this means that if spending habits had not changed over the years, LICO poverty would have been measured in 2005 at about 4.8%, rather than the announced 10.8%. That’s almost a 60% reduction, or twice the Liberal’s promised target. And, unfortunately for Mr. Dion, should Statistics Canada choose to update LICO poverty measures for more recent expenditure data, this would likely increase the measured incidence of LICO poverty once more and screw up the Liberal targets. So I guess the advice to Mr. Dion would be to pray Statistics Canada doesn’t do that.
Concerning his party’s “30-50 plan,” Mr. Dion states, “This is a bold goal” and “Specific targets allow the electorate to know when something has been a success. They also give a very clear idea of when their politicians have failed them.”
Well, I’m not sure we need any clearer picture about the failures of politicians. Instead, what we do need are policies that are at least honest in their goals. Especially when dealing with poverty, we need policies that utilize scarce resources (taxes) to target those most in need. This is the most compassionate approach, not policies that have good sound bites. Casting a relative poverty net (one full of holes, at that) over as wide as possible a population for socialist income redistribution purposes is cynical at best and, at worst, heartless toward the most needy among us.