Low Income Trends in the 1990s
by Andrew Jackson
Director of Research
Canadian Council on Social Development
This paper was presented by Andrew Jackson at the Regina City Hall (January 16, 2001) and Saskatoon Public Library (January 18, 2001). At the same events, CCSD's Executive Director, Marcel Lauzière, made a presentation entitled Social Policy at Risk: Can Canada make the right choices?
Defining Low Income
Unlike the US and some other countries, Canada has no official 'poverty line', not least because popular and expert opinion alike clashes on whether and to what extent low income should be defined in absolute terms (inability to meet basic needs) or in relative terms (distance from the community norm.) However, there is very active public interest in the changing fortunes of those at the bottom end of the income distribution. This reflects ethical concerns about social equity, as well as growing recognition of the high social and economic costs of low income in such terms as ill health and poor developmental and educational outcomes for children and youth.1
Statistics Canada regularly publishes data on the proportion of Canadians falling below three low income lines, soon to become four. The most familiar of these are the pre tax and post tax Low Income Cut-Offs (LICOs), which conceptually define a low income household as one which spends much more than an average equivalent household on the necessities of life - food, shelter and clothing - and thus has much lower absolute and relative 'discretionary' income than the norm. Housing costs in particular tend to swallow a very high proportion of the income of low income households. In 1998, the post tax LICO line in a large urban centre was $14,510 for a single person, and $27,890 for a family of four and the overall low income rate was 12.2%, or about 1 in 8 Canadians. 2
The Low Income Measure or LIM, which is used a great deal in international comparisons, explicitly defines low income as being much worse off than average, and is drawn at one half the median income of an equivalent household. In the mid 1990s, the overall incidence of low income after tax by this measure was 11.2% in Canada, much below the 17.7% rate in the US, but well above the 6.5% rate in Sweden.3 Canada fares badly in international comparisons when the focus is on low income among children, but well in terms of low income among the elderly. The UN Agency UNICEF recently reported that the child poverty rate in Canada (falling below half of median household income) for the mid 1990s is 15.5%, compared to 22.4% in the US and just 2.6% in Sweden, but the low income rate among the elderly, defined on the same basis, is just 4.8% in Canada, compared to 23.7% in the US and 6.0% in Sweden.4
Osberg has found that the wide gap between Canadian and US low income rates fell in the 1990s, to the extent that some US states now have low income rates below some Canadian provinces.5
Starting in 2001, Statistics Canada will provide data based on a new line called for by federal and provincial governments, the Market Basket Measure or MBM. This is an attempt to calculate the income needed by a given household to meet its needs, defined not just in subsistence terms, but also in terms of what is needed to approach 'creditable' community norms. While controversial in that the MBM involves subjective judgements on what to include in the basket, and would lower the low income rate compared to the traditional pre tax LICO measure, many anti poverty groups have followed the same approach of attempting to calculate basic budgets. In summary, the different lines provided in Canada reflect different approaches and provide different rates, but all define low income as more than inability to meet very basic needs, and all tend to show the same broad trends over time.6
The Causes of Low Income
Canadian households draw virtually all of their income from two major sources: market income gained from employment and investments, and transfer income from government programs such as public pensions, Employment Insurance (EI), and provincial social assistance. Investment income provides only a modest share of personal income except for very high income households and the elderly in receipt of private pensions.
Some individuals and groups - most of the elderly, those with serious long-term disabilities, many single parents with young families - receive only limited market income, so their level of income and vulnerability to low income is primarily determined by the 'generosity' of government transfers.
The Canada/Quebec Pension Plan, Old Age Security and the Guaranteed Income Supplement provide virtually all seniors with an income above or very close to the low income lines, and have greatly reduced both low income and income inequality among retirees since the 1960s.7 By contrast, social assistance incomes for all family types in all provinces, including child and other benefits paid on top of basic benefits, are well below the LICO lines.8 Table 1 provides illustrative data for Ontario. Note that data provided is for an "employable" adult; persons with disabilities receive more.
The great majority of the working-age population and their families normally rely very heavily on the labour market for their livelihood. Annual labour income is a function of wage rates multiplied by hours and weeks of work in the year, and rates and trends of low income are very strongly driven by the business cycle and by changing levels of employment and unemployment. Table 1 shows the total number of weeks of full-time work (40 hours) needed at the median hourly wage (and two-thirds of median) needed to reach the LICO line. A household with a single earner working 40 hours per week at the median hourly wage ($13.86 per hour) is unlikely to experience low income (defined by the pre tax LICO) unless the earner experiences long term unemployment over the year since 32 weeks suffice. However, a household with a single earner in a low wage job (defined as paying two-thirds the median wage) is highly vulnerable to low income even if fully unemployed over the year. Indeed, a single parent family with one child cannot climb above the low income line on earnings from a low wage job alone. A two adult/two child household with two low earners can only experience a total of about 3 months of combined unemployment to maintain earnings above the low income line.
Clearly, household composition is a key factor. Single households and single families with children are more vulnerable to low income because of dependence on one potential earner, and because single households cannot realize economies of scale.
Vulnerability to low income because of short or long term unemployment or family composition is, of course, strongly affected by the availability and generosity of EI and social assistance benefits, as well as by child benefits and GST credits which supplement earnings to some degree.
|Table 1: Vulnerability to Low Income|
| ||2 Adults/2 Children||1 Adult/1 Child||1 Adult|
|LICO before tax (city 500,000+)||$33,063||$21,962||$17,571|
|Weeks of work needed at median wage||60||40||32|
|Weeks of work needed at low wage||90||60||48|
|Total welfare income (Ontario)||$18,130||$13,704||$6,822|
An Overview of Low Income Incidence and Trends in the 1990s
Table 2 provides some basic data on persons on low income (post tax LICO) for 1989, the peak year before the recession of the early 1990s; for 1993, when the overall economic recovery began to take hold; and for 1998, the last year for which data are available. As shown, the overall low income rate for persons in 1998 was 12.2%, down from 13.1% in 1993, but still well above the 10.2% level of 1989. Low income rates for seniors - already very low for seniors in families - fell over the 1990s. By contrast, low income rates among the single non elderly rose, to 33.9%, from already high levels. The low income rate of children fell slightly for single parent families, but remained at a very high level in 1998 (45.3%). The rate rose from 7.0% to 8.3% for children in two parent families. Data from the 1995 Census show that low income rates among visible minority and aboriginal Canadians have recently been more than double the national rate.9
|Table 2: Persons in Low Income After Tax (% of Group) and % Composition of Low Income Population|
| ||1989||1993||1998||% of Low-Income Population in 1998|
|Children <18|| ||24.8|
|(2 parent families)||7.0||9.9||8.3||13.0|
|Age 18-64 in families||6.4||8.9||8.1||35.9|
|Source: Statistics Canada Cat. No. 75-202-RPE. Income in Canada. 1998. Table 8.1|
The final column of Table 2 provides a different picture by looking at the overall composition of persons in low income in 1998. While low income rates for working age adults in families and children in two parent families are relatively low, these two groups combined still make up almost half of the total population living in low income, and children from two parent families make up more than half of the total number of children in low income.
Given that social assistance incomes were far below the low income line in both 1989 and 1998, cuts to benefits in some provinces in the 1990s are unlikely to have had much impact on the low income rate as opposed to the depth and severity of low income. The key driver of the trend in the rate among the working age population in the 1990s has been the failure of the labour market to provide as many weeks and hours of work to vulnerable working age households as in the pre recession year of 1989. Certainly this is consistent with the steep 32% fall in the real market income of the bottom quintile of all households between 1989 and 1998, and the 16% fall in the second quintile, as compared to the 2% average decrease in the market income of all households over the same period.10 For two parent families with children with one earner, the incidence of low income rose from 14.3% in 1989 to 17.9% in 1998; and for non elderly single individuals who worked in the year, the incidence of low income rose from 17.3% to 20.5% between 1989 and 1998 among men, and from 23.5% to 25.9% among women.11
Reductions in wage income at the bottom of the income distribution in the 1990s would have been strongly influenced by still high unemployment in 1998 compared to 1989, and by the pattern of net job growth in the 1989 to 1998 period. This was heavily tilted towards precarious and often low paid own account-self employment, temporary and involuntary part-time work. A contributing factor appears to have been reduced entitlements to Employment Insurance benefits after 1995. Myles and Picot report that average UI benefits fell sharply for low income families with children, particularly two parent families, between 1993 and 1996.12
Depth of Low Income
The Nobel Prize winning economist Amartya Sen once pointed out that the proportion of the population living in poverty could be perversely reduced by redistributing income from the very poor to those just under a poverty line. This illustrates the fact that falling below a line tells us nothing about the depth of low income, or variations of income among the low income population.
In 1998, the average income gap relative to the after tax LICO was considerable: $6,638 for all families; $7,342 for two parent families with children; $6,194 for single parent families; and $4,919 for unattached individuals.13 In 1997, the total income gap compared to the pre-tax LICO came to $20 Billion or some 2% of total national income, of which just 10% represented a shortfall in the incomes of elderly persons.14 Income gaps and the proportion of low income persons living in very low income households have increased significantly among the non elderly in the 1990s.
An examination of low income intensity among Canadian children by Myles and Picot highlights the under-appreciated fact that the depth of low income among children fell in the 1980s because of rising transfers, even though the widely reported rate remained quite steady.
As would be expected, low income intensity increased in the recession, 1989-1993. More surprisingly, intensity increased 1993-96 as modest gains in employment income were more than offset by reduced social assistance and EI benefits. For those outside the labour market or in very long-term unemployment, social assistance benefits are a direct determinant of the depth of low income, while the availability of EI benefits is a significant income variable for those who move frequently between jobs and short-term unemployment.
Duration of Low Income
Knowledge of the dynamics of low income has only come to light in the 1990s with the advent of surveys like the SLID which follow panels of people over time. It is now clear that a large group of Canadians regularly move in and out of low income, while a much smaller proportion are stuck in this situation over several years. In the six years between 1993 and 1998, just under 1 in 4 Canadians (24.2%) experienced low income at least once, double the rate of low income in the single year of 1998. Of this group experiencing low income at least once, about half (12.8%) spent 'only' one or two years in low income, while the other half spent a significant period of time of 3 - 6 years in low income. Just over 5%, or 1 in 20 Canadians, spent either 5 or 6 years of the entire 1993-1998 period in low income.15
The groups at highest risk of long-term low income are those most likely to be dependent on social assistance or trapped in very low wage and insecure jobs for long periods of time: single parent families headed by women, persons with serious disabilities, and those particularly vulnerable to long-term unemployment such as youth and the near elderly with very low levels of education and skills. Continuing low income is highly likely to overlap with deep low income, particularly for social assistance recipients.
The reality of deep, continuing low income for perhaps 1 in 20 Canadians has raised concerns that we may be seeing the emergence of a US inner city type 'under class.' This has particularly been the case in the context of the relative concentration of low income persons in very low income neighbourhoods in some large urban centres. In 1995, 17.9% of all low income persons in Canada lived in very low income neighbourhoods (census tracts with double the national rate), up marginally from 17.3% in 1990, but well up from 11.8% in 1990.16 However, the characteristics of low income neighbourhoods differ considerably by major urban centre, and it is unclear how much movement there is in and out of distressed neighbourhoods over time. This is likely to be a major topic for future research.
Analysis has shown that about half of all movements of families with children in and out of low income are driven by 'family formation events' such as marriage and divorce, and about half by 'labour market events' such as unemployment and changes in the number of earners in a family.17 Studies analyzing flows in and out of low income have found that a relatively large group of working age Canadians seem to regularly move a little above and below the low income line as they alternate between relatively low paid and insecure employment on the one hand, and unemployment and dependence on EI or social assistance benefits. Setting aside relatively rare 'family formation events', the odds of falling below the low income line are much higher than average for those with incomes 25% or more below the median, while the odds of those with above median incomes falling into low income are quite low.
The persons most vulnerable to cycling in and out of low income are, as would be expected, those who have the greatest difficulty finding and keeping relatively well-paid and secure jobs: younger and older workers alike with limited education and skills; recent immigrants whose education and skills are either limited or unrecognised and who may experience language difficulties; aboriginal persons; and persons with disabilities. Vulnerability is much greater for single persons and particularly single parents who are disadvantaged in the labour market.18
The rate of low income has likely declined since 1998. Continued strong gains in full-time employment will have increased the weeks of work of many vulnerable working age households, and there may have been some wage growth even in low wage jobs as recovery has really taken hold. Meanwhile, the tax/transfer system has slowly begun to deliver modest boosts rather than cuts to the incomes of lower income households. At the federal level, the Canada Child Tax Benefit, which delivers significant net benefits to most low and middle income families with children, has been increased with each Budget; the basic tax exemption has been raised, and it and tax credits such as the GST credit have been reindexed to inflation; and reduced entitlements to EI benefits have been partially reversed. The deep cuts of the mid 1990s to provincial social assistance programs may have come to an end, and some provinces have begun to increase tax credits and support services for persons with low incomes. However, higher incomes for middle and higher income households will raise the thresholds relative to which low income lines are set.
Economic recovery will do a great deal to reduce absolute and even relative low income for many, but it will not do everything. Even at very low unemployment rates, some working age persons will remain outside the job market, or will be unable to command even modest annual earnings. Further reductions in low income will require social investments to raise the skills of the disadvantaged, and to provide the range of supports and services such as affordable child care and disability supports which are needed to improve access to jobs. Reducing low income will also require some increases in transfers, notably social assistance benefits, and some re-thinking of the structure of our income security system. The modest success of the Canada Child Tax Benefit and of the Earned Income Tax Credit in the US has led many to advocate a more comprehensive system of income supplements for the working poor as an important way forward.19
1Canadian Institute for Health Information, Report on the Health of Canadians, 1999. David Ross and Paul Roberts, Income and Child Well-Being: A New Perspective on the Poverty Debate. Canadian Council on Social Development, 1999. The LICO line is calculated by adding 20% to the spending of an average equivalent household on food, clothing and shelter. LICOs are calculated for different sized families and communities, and rebased periodically to take changes in household spending patterns into account. Currently, the average household spends about 35% of income on the three necessities so a low income household is one spending more than about 55% on necessities. While geographical variations in household spending were found to be quite minor in the past, LICOs have been criticized for failing to take into account large housing cost differences between major urban centres. The post tax LICO which is now highlighted by Statistics Canada is based on spending on necessities relative to after rather than before tax income. While the trend of the two LICO measures over time is basically the same, the proportion of low income households is lower for the post tax measure by about 4 percentage points since the gap between pre and post tax income grows as one moves up the income scale because of the progressive nature of the personal income tax. The post tax LICO is not adjusted for payroll taxes, but this is being reconsidered by Statistics Canada.
2Statistics Canada. Cat. 75-202 - RPE Income in Canada 1998. Table 8.4.
3OECD Economics Working Paper #189. Income Distribution and Poverty in Selected OECD Countries 1998
4Innocenti Report Card #1, A League Table of Child Poverty in Rich Nations. UNICEF June, 2000; Luxemburg Income Survey data from www.csls.ca for the elderly.
5Lars Osberg. "Poverty in Canada and the USA: Measurement, Trends and Implications". Presidential Address to the Canadian Economics Association, 2000. Forthcoming in the Canadian Journal of Economics.
6Statistics Canada Cat. 98-13. "Measuring Low Income and Poverty in Canada: An Update" May, 1998. Information on the Market Basket Measure can be found in Human Resources Development Canada, Applied Research Bulletin Vol. 4 #2, Summer-Fall, 1998.
7John Myles "The Maturation of Canada's Retirement Income System: Income Levels, Income Inequality and Low-Income Among the Elderly" Statistics Canada Cat. 11F0019MPE #147 March, 2000.
8 National Council of Welfare - Welfare Incomes 1999 Autumn, 2000. Table 6
9Statistics Canada "The Daily", May 12, 1998.
10Calculated from Statistics Canada Cat. 75-202-XIE Income in Canada, 1998 Table 7.2
11Statistics Canada Cat. 75-202-RPE. Income in Canada 1998. Table 8.3
12John Myles and Garnett Picot, "Social Transfers, Earnings and Low Income Intensity Among Canadian Children, 1981-96". Statistics Canada Cat. 11F0019MPE#144 March, 2000
13Statistics Canada Cat. 75-202-RPE. Table 8.3.
14CCSD Fact Book on Poverty, Chapter 6. Canadian Council on Social Development, 2000.
15Statistics Canada Cat. 75-202-RPE Income in Canada, 1998. Table 8.2. For a detailed empirical analysis based on tax data, see Ross Finnie, "The Dynamics of Poverty in Canada: What We Know, What We Can Do" C.D. Howe Institute, 2000.
16M. Hatfield, Concentrations of Poverty and Distressed Neighbourhoods in Canada. W. - 97 - 1E. Applied Research Branch, Human Resources Development Canada, 1997. For detailed analysis see John Myles, Garnett Picot and Wendy Pyper, "Neighbourhood Inequality in Canadian Cities" and Kevin Lee, "Urban Poverty in Canada" Canadian Council on Social Development, 1999.
17Garnett Picot, Myles Zyblock and Wendy Pyper, "Why Do Children Move Into and Out of Low Income? Changing Labour Market Conditions or Marriage and Divorce?" Statistics Canada Analytical Studies Branch Research Paper #132, 1999.
18For a detailed analysis see The CCSD Fact Book on Poverty and "High Risk Factors Behind Poverty and Exclusion", Human Resources Development Canada, Applied Research Bulletin Vol. 6 #1, Winter/Spring 2000.
19 For recent policy prescriptions see the 2000 Budget Submissions made to the Minister of Finance by the Canadian Council on Social Development, www.ccsd.ca, and the Caledon Institute for Social Policy, www.caledon.ca.
Canadian Council on Social Development,
190 O'Connor Street, Suite 100,
Ottawa, Ontario, K2P 2R3
Tel: (613) 236-8977, Fax: (613) 236-2750, Web: www.ccsd.ca, Email: firstname.lastname@example.org