VANCOUVER, BC-Efforts to accurately measure and define poverty in Canada have been hindered by inconsistent and poor quality data, resulting in a confusing picture that is often further distorted by politicians and activists, says a new study from noted poverty researcher Professor Chris Sarlo of Nipissing University.

In What is Poverty? Providing Clarity for Canada, released today by independent research organization the Fraser Institute, Sarlo finds that poverty, whether measured by income or consumption, has remained in the four to six per cent range since 1996. His numbers stand in stark contrast to media reports last week that claimed census data showed increased levels of poverty and a growing gap between rich and poor.

"Last week's report from Statistics Canada described 'relative poverty,' which is really an estimate of the proportion of Canadians who are less well off than average; it's a measure of inequality and tells us nothing about the state of deprivation in Canada," Sarlo said.

"Media commentators and politicians then take those numbers and describe poverty in absolute terms, using graphic images and over-the-top language that brings to mind hunger and misery usually associated with third-world countries. This problem of definition "switching" confuses people and impedes intelligent public discussion of this important issue."

Sarlo, who is also a senior fellow with the Fraser Institute, has long argued that the most realistic and credible measurement of poverty is one based on the necessities of life. He defines poverty as the cost of a list of basic needs required for long-term physical well-being, including nutritious food purchased at grocery stores fulfilling all Canada Food Guide requirements, rental accommodation, clothing purchased new at major department stores, household furnishings, supplies, personal hygiene items, laundry, insurance, and out-of-pocket health costs such as medications, dental, and vision care.

"We need to know how many of our fellow citizens cannot afford all of their basic needs. This measurement of genuine deprivation is based on the absence of these basic needs," Sarlo said.

"Relative poverty, which is what the census report from Statistics Canada described, is largely a function of the degree of inequality in a nation. It is not about poverty as most people understand that term. Unless a nation becomes more or less unequal over time, there will be no change in relative poverty despite any improvements in real income. A rising living standard, by itself, will do nothing to reduce relative poverty."

In What is Poverty? Providing Clarity for Canada , Sarlo tracks the latest information about the incidence of basic needs poverty in Canada, utilizing two different sets of data (one focused on family spending and the other on labour market information), and two different equivalence scales in the estimation of poverty in Canada.

Among Sarlo's findings:

• Poverty rates have declined over the past 36 years, but it is unclear what has happened to poverty over the past 31 years. One of the sets of data shows poverty declining sharply after 1996 (after being relatively stagnant over the preceding 22 years) and the other set of data has poverty rates rising slowly from 1974 to 2005.

• Using incomes drawn from the spending databases, overall poverty (that is, the poverty rate for everyone, including children, as distinct from the child poverty rate) fell sharply from about 12 per cent in 1969 to approximately three per cent in 1974 and then drifted slowly upwards to about 4.5 per cent in 2005. Using incomes drawn from the labor market databases, overall poverty fluctuated between eight and 12 per cent between 1973 and 1996, and then fell sharply after that to around five per cent.

• The estimation of child income poverty closely follows the patterns for overall income poverty.

• Consumption poverty for all people and for children fell sharply between 1969 and 1974 and has drifted up slowly after that using either equivalence scale.

Sarlo notes that his latest research again reveals concerns about data quality related to the issue of underreporting and hidden income, particularly in regards to income data, which is often used to indicate economic well-being in studies of poverty. He calls for Statistics Canada to improve the quality and reliability of its data collection.

"Underreporting of true income which includes gifts or subsidized rents understates how a household is doing. All of the income data collected by Statistics Canada, including the census and tax filer data, is "reported income" that relies on people giving accurate information about their incomes for the previous year," Sarlo said.

He suggests that more work is needed to get a more accurate picture of the extent and varying levels of poverty in Canada.

"The measurement of poverty in Canada is considerably hampered by the apparent sensitivity to the type of survey that income is drawn from, and by real concerns about data quality, specifically possible omissions and underreporting. Statistics Canada needs to examine these considerations and investigate issues related to data quality."