La stratégie de réduction de la pauvreté du gouvernement fédéral devrait adopter une approche pancanadienne ; elle favoriserait la collaboration entre les différents acteurs.

Efforts to find one simple indicator of poverty are unlikely to succeed and run the risk of misdirecting policies intended to reduce poverty. Better approaches exist, but they involve hard work and collaboration. They are intense and messy, not magical.

The many actors, both governments and NGOs, who play a role in fighting poverty in Canada would benefit from indicators that track trends in those dimensions of poverty they are mandated to address, and that help determine their policy goals and priorities, measure their progress and show how their contributions fit with those of others. For most of these actors, an evolving “dashboard” of carefully selected indicators would be more useful than a single measure.

For actors with a mandate to alleviate poverty using Canada’s income security system, the chosen indicators would likely focus on lack of current income. For actors with a mandate to prevent poverty from arising in the first place, the indicators would reflect resource lacks in areas such as skills, housing, employment and health.

Indicators do not grow on trees waiting to be plucked. They require a carefully defined data feeder system and a strong analytic capacity that serves multiple purposes. The government’s Canadian Poverty Reduction Strategy provides an opportunity to accelerate the development of this capacity.

However, much current attention on the measurement of poverty is focused on the search for the holy grail of a one-size-fits-all indicator. Alain Noël’s recent Policy Options article, « How Do We Measure Poverty?, » provides a succinct and useful review of current proposals. Nearly all of the proposed indicators involve some measure of the number of people whose current family income falls below one or another cut-off line. Noël’s preferred solution is a simple indicator that tracks the number of people in families that fall below 50 percent of current provincial median incomes.

Searching for a single indicator is understandable. It would be useful to have an unambiguous indicator that tells us whether poverty is growing or falling. However, relying on a single income measure would be risky even within the income security system. As well, it would not be helpful for approaches to alleviating current poverty that do not centre on income, such as the provision of housing or workplace skills. It would be particularly unhelpful in the context of efforts to prevent poverty from arising in the first place, such as early childhood education or tackling addiction.

The diverse needs of those we label as poor

The effort to find a single income measure is taking place despite the almost universal acknowledgement that poverty has many dimensions, and that there is no consensus on a definition. It is therefore not surprising that the search for a catch-all indicator has been going on worldwide for many decades, with no common solution found. The obvious conclusion seems to have been ignored: namely, that since we cannot agree on what poverty is, attempts to measure it will always be controversial and we run the risk of devising policy solutions that won’t work.

Research has long shown that the population that is poor, or that is at risk of becoming poor, is highly diverse. While there are many ways of demonstrating this heterogeneity, a look at the duration of low incomes tells a powerful story. Brian Murphy and his co-authors show that low income is often transitory. During the early 2000s, one-third of Canadians who fell into low income left low income the following year. Few (between 1.4 and 3.5 percent, depending on the measurement chosen) were poor for six years or more.

Many of those who are persistently poor face problems in addition to lack of income, including lack of skills, health issues, addiction, stress, inadequate housing, sickness or death of a family member, lack of supports to overcome work-related activity limitations, cultural and language barriers, discrimination, weaknesses in social networks — and, among the most disadvantaged, a mix of the above. The extent of these barriers must be measured if they are to be effectively addressed by policy. The oft-stated claim that income is a good proxy for the lack of all other resources makes no sense when we look at the characteristics of many of those who are poorest.

Poverty alleviation and poverty prevention require different indicators of success

Poverty is often conceptualized as the lack of resources needed to allow people to develop their capacities. This perspective leads to thinking about poverty over the course of people’s lives. The underlying goal is sometimes cast in terms of reducing intergenerational poverty — with each generation leaving behind a world that is better and more equal than the one they inherited. Attention is accordingly placed on longer-term, preventive human development measures that expand the scope of a “poverty agenda” to include education, health, housing and employment policies and, especially important in the literature, high-quality early learning and child care in low-income families. Many preventive measures will have their main effects far in the future. Their success cannot be captured by looking only at current resources, let alone just current income.

The “law of the instrument” holds that if our only tool is a hammer, we see nails at every turn. If the main poverty indicator relates to low income, policy responses will tend to focus on income security programming, with less attention paid to instruments that address the causes of poverty. This is an especially large risk at the federal level. Because the federal government’s role in the income security system is so large, there is a real danger of downplaying its role in poverty prevention. Many poverty prevention policies in Canada have a federal component, including those related to the labour market and skills, health, housing, literacy, immigrant settlement, disability and poverty reduction in First Nations. These preventive roles are often carried out in partnership with provinces and territories, and fighting poverty is generally not their prime goal. However, while they may be inconvenient and messy, they are important.

An essay by one member of the Experts Panel on Income Security of the Council on Aging of Ottawa (whose members are collectively responsible for the contents of this article) has, for example, argued that federal labour market agreements with the provinces and territories provide a rare opportunity for Canada to take a world leadership role in using the newly available tools of big data and predictive analytics to increase, perhaps dramatically, the success of skills training (including literacy training). This would enable many people with low income to find decent jobs — including many who would otherwise be persistently poor. An article by Employment and Social Development Canada evaluators describes the basic techniques.

While there is room for improvement, the role of the government of Canada in providing income security transfers is already quite effective, especially given improvements to the Canada Child Benefit and proposed changes in the Working Income Tax Benefit. The presumption should therefore be that returns on investment on other, more underutilized poverty prevention fronts (such as skills training) may be even greater. The choice of poverty indicators should encourage, not discourage, consideration of preventive options.

An opportunity to address the real measurement challenges

In places, and especially in its discussion of measurement, the background paper to the Canadian Poverty Reduction Strategy focuses on income security programming. However, the paper also opens up the possibility of a broader focus that could, for example, involve establishing a collaborative pan-Canadian umbrella framework that could provide an ongoing empirical basis for harmonizing the strategies of all of the many actors who fight poverty.

A strategy with a federal income security focus could make sense, provided there are counterpart strategies that cover federal programming in support of poverty prevention, perhaps including separate strategies for housing (such as the recently announced National Housing Strategy), or for literacy skills, or for First Nations. Separate federal strategies would make the most sense if there were also an umbrella, pan-Canadian strategy that takes account of the interactions among the different poverty-fighting programs at all levels of government.

A separate federal strategy with a focus on income security should be supported by a dashboard of income-related indicators, not a single indicator. The dashboard might well include a moving average of the Low Income Measure (LIM), perhaps supported by a regularly updated Market Basket Measure, some measures of the depth and persistence of poverty, and perhaps indicators related to work precariousness, food insecurity and material deprivation.

The search for a single measure not only ignores the multiple dimensions of poverty, it is also likely to result in continued controversy about the way poverty is measured, as opposed to dialogue on how to best fight poverty. For example, in his Policy Options article, Noël argues for a provincially based LIM, since most of the responsibility for addressing poverty is in provincial jurisdiction. While that might work for assessing provincial and territorial income security programs, a nationally based measure makes more sense in assessing the large federal programs that support those with low income, many of which are financed from general taxation, where common national standards are either explicit or implied by equalization transfers.

Our panel’s main interest, however, is to advocate that consideration be given in the Canadian Poverty Reduction Strategy to setting up a pan-Canadian strategy that would help the many actors throughout Canada work together in fighting poverty.

Toward an empirically based pan-Canadian strategy

From a measurement perspective, such a strategy would have two components: the development of the underlying database and associated analytic capacity; and the establishment of a consultative process that would result in ever-improving poverty indicators that reflect the evolving database.

These measures would allow us to provide consistent data to support the (explicit or often implicit) dashboards of indicators that underlie all poverty reduction programming and to untangle the underlying webs of causality. Policy initiatives that are informed by an understanding of their key determinants would be more effective. The data would describe the characteristics of people over the course of their lives, including their attachment to family, to the labour market and to the programs and services that provide resources such as skills, health, caring and income. The analytic capacity would facilitate tracking how people acquire and lose these resources, allow analysts to simulate the effects of proposed policy changes and provide a conceptually consistent hierarchy of indicators and supporting information to thousands of actors throughout the system.

Many of the needed data collection and analytic tools are well known in research circles under headings like big data, predictive analytics and dynamic microsimulation modelling. However, these tools are not yet widely used in policy applications. More details can be found in a separate paper prepared by the panel.

The proposed consultative process would likely occur annually, be headed by an independent body, be federally funded and be heavily reliant on Statistics Canada for data. The goals would be to identify priorities for developing the new data and analytic capacity, to identify the common data needed by the various actors to build their own indicators and to gradually develop a dashboard of pan-Canadian indicators. For example, the consultations would seek consensus on what should be included in the next year’s dashboard, based on current policy priorities and on progress in building the new data infrastructure.

Such a process would respect existing jurisdictions and would use discussions about the evidence to support more transparent and more open public discussion of policy priorities. It seems to hold much potential as a nonterritorial, collaborative, sustainable and ultimately more effective approach to fighting poverty.

This article reflects the view of the members of the Experts Panel on Income Security of the Ottawa Council of Aging: Bob Baldwin, Bernard Dussault, Peter Hicks, Andrew Jackson, Russell Robinson (chair), Richard Shillington, John Stapleton and Michael Wolfson.

Shutterstock/ By Vladimir Prusakov


Do you have something to say about the article you just read? Be part of the Policy Options discussion, and send in your own submission. Here is a link on how to do it. | Souhaitez-vous réagir à cet article ? Joignez-vous aux débats d’Options politiques et soumettez-nous votre texte en suivant ces directives.