Mr. Dion’s Anti- Poverty Plan

I’m a big fan of setting clear and attainable targets and timetables to eliminate poverty, and applaud last week’s Liberal Party commitment to reduce the number of those living in poverty by 30% and the numbers of children living in poverty by 50% within 5 years. http://www.liberal.ca/story_13293_e.aspx

Clear targets and timetables have recently been called for by the National Council of Welfare and by Campaign 2000 with the laudable aim of keeping governments’ feet to the fire, and have clearly played a significant role in sustaining tangible reductions in poverty in the UK and other countries, as detailed in a good new Campaign 2000 policy paper. http://www.campaign2000.ca/res/dispapers/summoned_to_stewardship.pdf

That said, I have a couple of major quibbles with the metric for the poverty reduction target which has been embraced by Mr. Dion and the Liberals, namely incidence of poverty as measured by the Statistics Canada after-tax LICO (LICO-AT.) My preferred metric would be the poverty gap measured in dollars relative to a consumption-based poverty line such as a reformed Market Basket Measure of poverty. (MBM)

This might seem like a boring, technical, policy-wonkish kind of point. But we need a target that will register progress in implementing good progressive policies that work to improve the lives of low income Canadians, and incidence by the LICO-AT just won’t do that. It could, in fact, discourage measures to boost the incomes of the very poor, and will register no progress at all to the extent that we address poverty through increased investments in affordable child care and affordable housing programs.

Any poverty line is arbitrary and the LICO-AT does a reasonable job of tracking changes in income poverty over time. Effectively it defines poverty in terms of a household having an income which means that much more than average has to be spent on the necessities of life – a compromise between a consumption or bare necessities of life standard, and a pure relative income standard such as having an income below half or two thirds of the median for a household of the same size. (As an aside, a consumption based standard is often miseladingly described as an absolute measure , but any consumption standard has to be set relative to the consumption norms of the society in which a person lives. Even the absurdly frugal Fraser Institute poverty line is explicitly set relative to Canadian consumption norms.)
The problem with defining a poverty reduction TARGET by exclusive reference to the numbers living below a line is that we totally fail to take into account the depth of poverty. As Amartya Sen has argued for years and as Lars Osberg has long argued here in Canada, poverty depth is at least as great a problem as poverty incidence. The quickest, cheapest and simplest way to cut poverty as measured exclusively by incidence would be to modestly increase the incomes of those just below the line, while ignoring those well below it. Sen even noted that we could make great but obviously illusory progress in the war on poverty by taking income from the very poor and giving it to the just poor, pushing the latter over the line and thus cutting the poverty rate while increasing the depth of poverty of the very poor.

While this may seem a wonkish point, as Campaign 2000 detail in their most recent report card on child poverty in Canada, the incomes of low income families with children are well below the LICO-AT poverty line – by an average of $9,400 in the case of single parent families, and by $10,400 in the case of two parent families. Increasing social assistance benefits would raise the incomes of many single parent families with young kids, not to mention those of adults with disabilities who can’t work on a continuing or regular basis, but would have virtually no impact on the poverty rate for children or persons with disabilities. By the same token, raising minimum wages (not on Mr. Dion’s list of anti poverty measures, by the way) or increasing refundable child benefits would lift the incomes of many working poor households, but this would be only incompletely captured by the impact on the numbers of people living below the poverty line.

The key point here is that we need to target a reduction in the poverty GAP – the total amount by which the incomes of poor households fall below the poverty line. Any measures which boost the incomes of poor households will lower the GAP, but not necessarily the numbers living below the line. (As an aside, the elimination of the GAP is a surprisingly affordable proposition. The annual revenue cost of cutting the GST rate by one percentage point – $6 Billion – would completely eliminate the poverty gap for children, according to Campaign 2000 based on CCSD research.)

The second point is that the even the poverty GAP measured relative to the LICO-AT would not capture progress in addressing poverty through supports and services as opposed to cash transfers. For years, anti poverty activists have been calling for more affordable housing and affordable child care. Housing and child care are obviously huge expenses for most working and non working poor families with children, and any reduction of these costs through good public policies would mean huge improvements in the lives of low income people. But these improvements would simply not be captured by changes in the incidence of poverty based on the LICO-AT if policy took the form of delivering in kind benefits as opposed to cash benefits. A rental subsidy paid in cash would count, but not access to a subsidized co-op or rental unit. A child care subsidy would count, but not access to a subsidized and affordable child care space.

At best, we would fail to capture the impact of good anti poverty policies. At worst, we would create a bias for governments to address poverty exclusively through cash transfers. Mr. Dion’s speech made almost exclusive reference to such transfers.

This is a difficult issue to resolve, but consumption based measures of poverty such as the Market Basket Measure (MBM) of poverty developed by Human Resources and Social Development Canada could be tweaked or supplemented to take into account the actual living costs of low income households so that development of the in kind services and supports we want was captured. The MBM did attempt to capture actual child care costs paid by low income families with children, and was designed to be sensitive to actual community housing costs as opposed to the nation-wide averages captured in the LICO-AT.

By way of conclusion, we are making some real progress in moving poverty reduction and elimination up the public policy agenda. The efforts of Campaign 2000, the National Council of Welfare and others are pushing politicans to adopt clear targets. But we need to do the hard work of developing measures which will capture real progress in implementing a progressive agenda.