June 9, 2000
Ottawa -- On June 12th, Statistics Canada is expected to release Income in Canada, containing family income data for 1998. This publication will include Statistics Canada's widely publicized calculation of the proportion of Canadians who fall beneath the "LICO" or low-income cut-off lines. While Canada does not have an official poverty line, the so-called "pre-tax LICOs" have been widely used as such over several years by many organizations, including the Canadian Council on Social Development.
This year Statistics Canada is shifting the way it reports on the income data by highlighting the "after-tax LICOs". This shift may lead some to the wrong conclusion that poverty is much lower than previously thought. While we hope and expect that the rate of poverty will indeed have fallen from 1997 to 1998 because of improvements in the job market, this can best be judged by changes in the proportion of Canadians falling below the pre-tax LICO lines in the two years.
The purpose of this note is to briefly comment on the change in Statistics Canada reporting, and to explain why the CCSD continues to prefer the "pre-tax LICO" line.
What is the difference between before- and after-tax LICOs?
Statistics Canada has been publishing data on the incidence of low income since the 1960s. The LICO line has been calculated in relation to the expenditure of an average family on the essentials of food, clothing and shelter. If a family spends 20 percentage points more of its total income on these essentials than the average family, then it falls beneath the LICO line. LICOs are calculated and reported for families of different sizes, living in communities of different sizes ranging from rural to large urban areas.
No single measure of poverty is ideal, and all involve some arbitrary choices. A major virtue of the LICO is that it implicitly incorporates both an absolute view and a relative view of poverty. Families fall below the LICO if they are spending a very high proportion of their income on essentials and thus live in straitened circumstances (the absolute dimension) but LICO lines also tend to rise in line with increases in average income (the relative dimension). The proportion of families falling beneath the LICOs tends to closely rise and fall in line with changes in the job market, as one would expect.
Contrary to the belief of some, the LICO measure does allow for progress to be made in reducing poverty. For example, the proportion of seniors falling beneath the LICO line has fallen considerably over the years. Most importantly, the continuity of the LICO measures over time allows Canadians to judge if poverty is rising or falling, whether because of changes in the job market or because of changes in social programs.
LICOs are calculated on the basis of both pre-tax and post-tax income. The former have in the past been released earlier each year than the latter and have been widely regarded as "the" poverty line. The CCSD appreciates that Statistics Canada will continue to make available both pre- and post-tax LICOs, and we hope that it will caution users about making comparisons between data based on the two measures.
It must be emphasized that it is possible to look at changes in the pre-tax LICO over time, or changes in the post-tax LICO over time, but one cannot compare a pre-tax LICO in 1997 to a post-tax LICO in 1998.
Post-tax LICOs are calculated in the same way as pre-tax LICOs, but expenditure on essentials is calculated as a proportion of income after income tax has been paid. This appears to lower the incidence of low income since relatively affluent families pay more in income tax than do low income families, reducing the income gap between them. For example, in 1997, the proportion of all Canadians falling below the post-tax LICO was 13.3% compared to 17.5% based on pre-tax LICOs, and the rate for children was 15.8% compared to 19.8%.
Post-tax LICO doesn't measure amount left in people's pockets
The CCSD believes that the pre-tax LICO line, expressed in dollars, is more meaningful for Canadians because most of us think of our income and what it can buy in pre-tax rather than post-tax terms. While it may be reasonable to argue that low income should be considered in relation to the actual amount of money people have in their pockets to spend, the post-tax LICO does not do a very good job of measuring this for the following reasons.
The definition of after-tax income used by Statistics Canada does not reflect the impact of payroll taxes deducted from employment income, such as EI and CPP premiums, even though these are deducted in exactly the same way as income tax. Since payroll taxes are levied as a percentage of wages only up to a threshold level, they take more of the income of low-income workers than of very high-income workers, and thus counter the progressive impact of the income tax.
The after-tax measure does not adjust for all sales and consumption taxes. However, many studies have shown that these taxes tend to bear most heavily upon the less affluent.
To summarize, the calculation from pre- to post-tax LICOs takes account only of the progressive aspect of the overall tax system, and makes no adjustment for the flat or regressive elements. Expert studies have shown that Canadians at all income levels pay about the same total proportion of their incomes in taxes to governments, so it is questionable to adjust for income taxes alone.
Finally, a true calculation of consumable income would adjust for employment expenses, such as child care and travel to work. Low-income working Canadians spend a much higher proportion of their wages on such costs than do higher income Canadians, but this is not reflected in a calculation of their after-tax income.
There is a case to be made for calculating low income estimates based on both total income and consumable income. However, the post-tax LICOs do not provide a complete picture of consumable income. As a result, they tend to understate the incidence of low income.
The major purpose of this backgrounder is to caution against any conclusion that poverty is falling or rising based on a comparison of the pre-tax LICOs from previous years and the post-tax LICOs for 1998 that will be released on June 12. The real story will be changes from 1997 to 1998 as registered by either. Hopefully, the improving job market will have brought about a reduction which we can all cheer.