As one of the originators of Canada’s market basket measure (MBM), developed with my colleague Michael Hatfield, I was interested to read the argument for complementing it with a material deprivation index (MDI) to better measure poverty, as outlined recently in Policy Options by Richard Matern, Dana Vreeswijk and Sofia Seer.
They are onto something important. With a few caveats.
Canada would benefit from incorporating a well-designed material deprivation measure into its understanding of poverty. But the strongest argument for the MDI is not that it reveals the inadequacy of the market basket measure. It is that it tells us something the MBM was never designed to tell us.
The MBM asks a relatively straightforward question: does a family have enough disposable income to afford a specified basket of goods and services representing a modest, basic standard of living in its community?
Canada is too rich to be failing this badly on economic and social rights
That made the poverty threshold concrete. Rather than defining poverty simply by a position in the income distribution, the MBM connects income to the cost of necessities. It also lets us examine what happens to families after taxes and transfers, and therefore lets us evaluate an important part of what governments actually do for their residents.
But there is another question worth asking: whatever a household’s measured income, what necessities is it actually unable to afford?
That is the territory of a material deprivation index.
Two households can have the same disposable income and experience quite different standards of living. One may have unusually high medical or disability-related expenses. Another may carry substantial debt. Some may possess savings that allow them to absorb an income shock, while others may not.
An income measure cannot capture all of these circumstances. Nor should we expect it to do so.
A material deprivation index approaches the problem from the other direction. Instead of asking what resources people have available, it asks whether they are actually going without things regarded as necessary for an acceptable standard of living.
That makes the results reported by Food Banks Canada interesting. But it also means we should be careful about comparing its 28-per-cent material deprivation figure directly with the official MBM poverty rate of 11 per cent.
Those numbers are not competing estimates of precisely the same phenomenon. They are different measures answering different questions. Indeed, the difference between them may itself contain valuable information.
Suppose income poverty falls while material deprivation does not. Perhaps housing or other essential costs are increasing in ways existing measures do not adequately capture. Perhaps household indebtedness matters. Perhaps particular populations face costs poorly represented by averages. Conversely, material deprivation might improve even when measured household income changes only a little.
The dental care example discussed by Matern, Vreeswijk and Seer is especially intriguing. They report a substantial decline in respondents unable to afford an annual dental visit, coinciding with the expansion of the Canadian Dental Care Plan.
If further evidence confirms that relationship, it illustrates something important. Governments do not improve people’s material well-being only by increasing their cash incomes. They can also reduce what people must pay to obtain necessities.
If government assumes some of the cost of dental care, a family may be materially better off even though its measured income has not increased by a single dollar. An income measure may have difficulty detecting that improvement. A well-constructed deprivation measure potentially can.
Material deprivation information may also help policymakers understand not merely how many people are experiencing hardship, but what kind.
If inadequate income is the problem, income transfers may be appropriate. If housing costs are driving deprivation, the problem may be housing supply or there may be a need for housing assistance. If people cannot obtain dental care, publicly supported dental coverage may work better than an equivalent generalized income intervention. If disability-related costs are producing hardship among households whose incomes appear adequate, a targeted response may be necessary.
Good measurement should help us distinguish among those problems rather than compress them into a single poverty number.
There are, of course, questions for any material deprivation measure. Which items belong in the index? How do we distinguish inability to afford something from a decision not to purchase it? How stable are judgments about what constitutes a necessity? At what point does deprivation warrant classification as poverty? How comparable are results across regions and over time?
These are not objections. They are normal questions accompanying any serious social indicator. The MBM itself has required — and continues to require — review, refinement and methodological judgment.
Canada should not turn this into a contest between the MBM and the MDI.
The MBM tells us whether household disposable income is sufficient to purchase a defined modest standard of living. A material deprivation index offers another perspective: whether Canadians are actually able to obtain particular necessities associated with that standard.
The objective should be to understand people’s circumstances well enough to design policies that improve them.
The question is not whether the MDI should replace the market basket measure. It should not. The better question is: What can we learn when we put the two beside each other?
For policymakers interested in understanding poverty — and in knowing whether their interventions are actually reducing it — that may prove considerably more useful than asking either measure to tell us everything.
