Statistics Canada's Low-Income Cutoffs
By Andrew Mitchell and Richard Shillington
The recent release of estimates of the Low Income Cut-offs and the Low Income Measures has raised a number of crucial issues about the measurement of poverty or "low income"; in Canada.
Ever since the Low income Cut-Offs (LICOs) were established there have been two sources of change in them. The first is the routine indexation of the lines with changes in the Consumer Price Index to maintain their real value in relation to changing prices. The second was their periodic re-basing in accordance with changing consumption patterns. To understand this we need to understand more about how the LICOs are set in the first place. To establish the LICOs Statistics Canada calculates the percentage that an average family spends on food clothing and shelter. If that was, for example, 43%, as it was for a family of four in a medium sized city in 1992, Statistics Canada added another 20% to that figure to arrive at a figure of 63%. The LICO for a family of four in a medium sized city was then set at the income level where, on average, families spent more than 63% of their income on food, clothing and shelter.
Over time as real incomes and general living standards have risen the percentage families spend on those three necessities has fallen. When that happens the percentage that families spend on necessities falls, the LICO increases and a greater number of people will be counted as poor, or "low income"; than if this adjustment had not been made. That is the essence of a relative poverty measure – as general living standards rise the income below which one is considered poor also rises. This reflects the idea that to be poor is a notion that only has meaning in relation to the society in which one lives. "Re-basing"; of the LICOs refers to this re-calculating of the LICOs to take this into account, which is why we had 1969-based lines, 1978-based lines, 1986-based lines, then 1992-based lines. Most researchers have referred to LICOs as "relative" rather than absolute measures of poverty because the re-basing meant that LICOs took account of the improving real economic circumstances of Canadians as a whole.
But since 1992, or nearly 16 years, the lines have not been re-based. It appears that Statistics Canada has made a decision to stop re-basing the LICOs in accordance with changes in family consumption patterns, so that now the only source of change to the LICOs are inflation adjustments.
The figure below shows the consequence of this by comparing the various LICOs (1968-based, 1972-based, 1986-based and 1992-based) for a single person in a large city, in relation to average family income over time. If the LICO had never been re-based, and remained based on 1968 consumption patterns, but indexed to the CPI it would been just under $12,600 in 2000. Re-basing it to reflect rising affluence resulted in a 1992-based LICO of slightly over $18,000 instead. If the LICOs had never been re-based all of the people with incomes between $12,600 and $18,000 would not have been counted as poor.
The second consequence of this is that the LICOs begin to fall relative to average incomes. In 1992 the single person LICO was approximately 30% of the average family income. By 2000 it had declined to 27%. That is, what is considered "poverty" becomes more and more distant from the mainstream. Essentially this means that over time the LICOs, and rates of low income are accordingly lower.
The LICOs themselves become more and more an absolute measure and less of a relative one.
Since Statistics Canada no long re-bases the LICOs then we should refer to them as an absolute measure of poverty which does not change, at least not automatically, and not recently with general living standards.
Past experience would suggest that re-basing the LICO would increase the income thresholds and increase the number of families below that threshold. We have no way of knowing how much of the trend of falling poverty rates over the last 15 years is due simply to the fact that the LICOs have not been rebased.
The second major issue is the abundance of poverty lines increasingly used in policy debates. The whole issue of poverty assessment has become quite muddied. We now have in Ottawa; Low-Income Cut-Offs – LICOs - (before tax and after tax), Low-Income Measures - LIMs (before tax and after tax) and now also the Market Basket Measures – MBM (created by HRSDC at the behest of the Provincial Ministers of Social Services as a clearly absolute measure of poverty).
To illustrate the variety of measures available consider the following example; let";s talk about single senior women in Quebec – median income of about $17,000.
Would we all agree that they are living in "straightened circumstances";? Are they poor? Well that depends on who you ask. We use the last year 2000 because then we can compare all our poverty measures. What proportion fell below our various poverty measures?
Fifteen years ago, if we were had had this discussion, this is the only measure we would have had. The poverty rate would have been clear. But now the poverty line, the income below which one is "poor"; varies dramatically: before tax LICO about $18,300; after-tax LICO about $15,000; after-tax LIM – about $13,000 and MBM for Quebec - $11,220. So for the year 2000, the poverty line for a single person in Montreal was somewhere between $11,220 and $18,300.
And now we have progress; we have lower poverty rates depending on the measure. We can reduce the poverty rate down from 64% to 5% without spending a penny. These changes in poverty measure had no impact of the well-being of these seniors.
Statistics Canada plays an important role on collecting data "which are used to assess poverty"; and ideally should be neutral and above the political fray. Yet Statistics Canada no longer publishes low-income rates using before-tax LICO";s so those wishing poverty rates measured this way must be able to pay for special tables.
Statistics Canada has made no secret of its discomfort with the LICOs and the uses to which social policy analysts have put them, a discomfort that seems to have mostly taken hold as a consequence of the critiques of the LICO issued by groups such as the Fraser Institute. One wonders if this is an attempt to kill the LICOs quietly by making them irrelevant.
Despite the Statistics Canada";s cautions and the war on LICOs, the Government of Canada recently informed the United Nations Committee on Economic, Social and Cultural Rights of the following: “Canada has no official measure of poverty, but typically uses Statistic Canada's LICO as a proxy.” (Source: Responses to the List of Issues Concerning Canada";s Fourth Periodic Report under the International Covenant on Economic, Social and Cultural Rights)
Experience over the last few decades has been that when poverty rates increase Canadians are cautioned that we have no official measures of poverty and those that exist have problems; over the last few years as poverty rates have declined, governments have been willing to claim credit and refer to the trend of declining poverty rates.
These are but a few of the issues one must keep in mind when using any of the variety of poverty measures currently on offer.