Poverty Measures Need Measuring

Commentary, David Seymour, Poverty, Uncategorized

As German philosopher Immanuel Kant pointed out, a man whose fishing net has two-inch holes will likely conclude all fish are more than two inches wide. The discipline of measuring poverty is one where models and approximations have been passed off as reality for too long. Quotes like this recently published one: “it is inexcusable that 546,000 British Columbians, 13 per cent of the total population, live in poverty…” raise more questions than they answer. Such figures often reflect a measurement system with the measurer’s prejudices more than anything else. (The author of the above quote used a Statistics Canada income measure which Statistics Canada warn is not a poverty measure).

The art of generating poverty statistics is an attempt to generate a total picture of everyone’s living standards with limited information. Short of an Orwellian state, nobody can observe every single person’s real activities and judge their quality of life. Just as with dictionaries, poverty statistics are short hand summaries of complex activity. Just like atomic models, they simulate things that we cannot observe. Language scholars and physicists each realize that they are using limited information to create approximations of reality, not reality itself. Problematically, poverty measurements are often presented without explanations of their limits.

For the most part, poverty measures are based on income statistics, which are easily available. Unfortunately the assumption that recorded income reflects standards of living is tenuous. For example Canada is full of restaurant servers who wear designer clothes, eat and drink out, and attend endless “industry” parties. These people are below the (income based) poverty line because money from tips doesn’t always show up when government inspectors ask.

Even when monies do officially show up, single-year snapshots of income are highly misleading. From 1993 to 1998, 24 per cent of the Canadian population was deemed low-income at some point, but only 2.9 per cent were in this category for all six years. One British poverty report found that three out of every hundred people classed in the bottom 10 per cent for income were in the top 10 per cent for expenditure. Poverty statistics based on a single year’s income don’t reflect true lifestyles because savings can cross financial years and sometimes not all income shows up officially.

An even more difficult problem in accurately measuring poverty is the mistaken notion of some that when others become richer, that makes you poorer. Much of the poverty debate is based upon relative measures of income. Absolute measures of poverty compare the money that people (theoretically) have to spend with the cost of food, shelter, clothing and transport. Relative measures, on the other hand, compare income to the income of others. If an entrepreneur like Bill gates moved from the US to Canada, he would bring massive wealth and skill with him. However relative poverty would increase here and decline there. It’s a bizarre way to try and calculate “poverty”.

There is a cynical species of crabs that, if placed in a dry bucket to drown in air, would rather drag down any crab down to share their grizzly fate than help it escape. Canada wasn’t built on that spirit, but relative poverty measures are. Regardless of how the poor fare, relative poverty increases if wealthy people become wealthier. Indeed, Statistics Canada points out that higher taxes on the rich appear to reduce relative poverty, not because they give the poor more money, but they pull higher income earners down and close relative gaps.

Worldwide, top income tax rates have tumbled from heights of up to 68 per cent (Canada) since the 1970s and so people have become more prepared to declare high incomes. So higher incomes started to show up in official statistics. Does this mean that poverty has increased or that we have saner tax levels? Relative poverty measurements would mistakenly conclude the former.

That is the nub of our problems with taking poverty statistics too literally. Misleading statistics can introduce solutions like punitive taxes that don’t actually help. Nothing here is intended to deny poverty exists or to belittle attempts to reduce it. All of the above is designed to sharpen focus on poverty.

Next time you see a proclamation that poverty is X%, or Y% compared to year Z, just remember it is partly a reflection of poverty, and partly of a measurement system. Understanding measurement systems is essential for interpreting the poverty and the current debate takes poverty statistics far too uncritically at the cost of more targeted solutions.