Poverty in Canada (2nd Edition)

THE ESTIMATE THAT JUST UNDER one million Canadians (or about 4 percent of the population) had reported incomes lower that the poverty line in 1988 must be qualified. It is clear that in a number of important cases the poverty lines, based on 1988 market costs, overestimate true costs to the household, thus resulting in an exaggerated poverty rate. In several other cases, the low reported incomes belie true living standards—again resulting in an overestimate of the incidence of poverty.

Home ownership

Mortgage-free home ownership does not eliminate shelter costs but, in most cases, substantially reduces them. Though the household is freed from payment of principle and interest on the mortgage loan, there do remain four additional categories of costs: (1) maintenance, repairs and replacements; (2) property taxes; (3) homeowners' insurance; and (4) utilities (water, fuel and electricity). In 1986, the total cost of these categories for the average Canadian mortgage-free homeowner was $3,440. [Statistics Canada, FAMEX, 1986, p. 70.] The cost for "poor" mortgage-free homeowners is likely to be less than this. However, using the Canadian average and applying a 9 percent rate of inflation between 1986 and 1988, shelter costs for the typical mortgage-free homeowner would be $3,750 in 1988. Since mortgage-free homes contain on average almost three persons, it is appropriate to compare this cost with our estimates of essential shelter costs for a family of three based on market rents. For those with paid up homes, shelter costs are between 17 and 54 percent less than our estimates, depending on the location. The number of poor households involved is not insignificant.

In 1988, about 1.8 percent of home-owning families were poor [Using a weighted average poverty line for families of 3: $10,675.] to the extent that their reported incomes were below that needed for basic necessities. That amounts to roughly 92,000 families and of those fully two-thirds or about 63,000 families owned their homes mortgage-free. Among poor unattached individuals, it is the case that about 85 percent are renters. Nevertheless about 5 percent of home-owning individuals (or 44,600 persons) in 1988 were poor. Of these, 72 percent or about 32,200 unattached individuals owned their homes mortgage-free. [These calculations assume an overall weighted average poverty line for unattached individuals of $5,819.] In aggregate then, roughly 183,000 persons or 20 percent of the poor in Canada in 1988 will have shelter costs that are significantly below the estimated amounts, due to outright home ownership.

Consider, for example, an Ontario family of three with a 1988 income of $11,000. They would have been unable to purchase all of the basic necessities (cost = $11,210). If that same family owned their home mortgage-free, they would have been able to purchase all necessities with about $1,800 to spare.

Mortgage-free home ownership effectively reduces a family's de facto poverty line. To get an idea of the importance of this consideration, new poverty lines were calculated for people who own their own homes and the incidence of poverty determined. Since shelter costs comprise, on average, about half of essential costs and since mortgage free home ownership results in at least a 20 percent reduction in shelter costs, new poverty lines for this experiment were a simple 10 percent below the previous ones. As a result, the number of poor persons in Canada is reduced by about 37,000 or 4 percent.

Before leaving the issue of home ownership, it is worth pointing out that there will also be a somewhat smaller number of poor homeowners with mortgages whose costs will nevertheless be lower than the stated "essential" shelter costs because they are towards the end of their amortization period.

Rent subsidies

Every province in Canada has a stock of "social" housing which has been built up over the years with federal and provincial financial assistance. The rents in social housing are geared to income, that is, tenants pay rents equal to 25 percent to 30 percent of their gross income. The CMHC estimates that in 1988 there were approximately 600,000 social housing units in Canada. [Information obtained from CMHC official by telephone.] Acquisition of social housing can surely lift people out of poverty. The single mother of two in British Columbia whose income is $10,500 is living almost $500 below the poverty line. With social housing, her family's situation improves significantly and they are now $2,500 above the line. Similarly, the financial status of the unattached single person in New Brunswick with total income of $5,000 changes from a "poverty gap" of about $500 to a surplus of about $1,500 upon acquiring a rent-geared-to-income unit. There are no data that reveal how many of Canada's roughly one million poor are living in social housing units. We do know that about 700,000 of them are tenants. [Using the microdata file, there are an estimated 691,000 poor persons living in rented accommodation.] Since income is the main criterion for eligibility, we might expect that a majority of poor tenants would be in social housing. However, there are also good reasons to believe that a fair number of them may not be living in these units. Not all of the poor will necessarily need social housing, and not all will apply even if they need it. In addition, at any point in time some who qualify will be on a waiting list and will be paying market rents in the interim.

It is important to note that in 1986, 120,000 tenants in Canada lived rent-free. It is not known how many of these were poor. Another 127,000 tenants were roomers, the majority probably living in rooming houses and in family homes. Roomers paid average annual shelter costs of $1,866 in 1986. With average incomes less than 40 percent of the national average for tenants, it is likely that a majority of roomers would be poor by our definition. [All of the statistical information in this paragraph is derived from Statistics Canada, FAMEX, 1986, Table 7, p. 71.]

What all of this means is that for a substantial number of poor tenants, perhaps a majority, estimated essential shelter costs in table 6-4 significantly overstate the actual costs of their housing. Since shelter is the largest expense for poor households, the existence of subsidized, free and other low-cost shelter is of enormous importance to the least well off.

Students

During the 1988-89 school year, there were about 816,000 full-time students at Canada's universities and community colleges. [Statistics Canada, Enrolments at Canadian Universities and Colleges, 1988, cat. 81-229.] Roughly 37 percent of them lived at their parents' home while they attended a post secondary institution. [Statistics Canada, A Profile of Post Secondary Students in Canada, 1983-84, special run. As of June 1990, this is the most recent survey of college and university students.] Since Statistics Canada conducts its income survey every April, at the time of the survey this group would be counted as part of the family they were living with and would not be considered unattached individuals. The other 63 percent however lived independently during the school year—in student residences, in co-ops, in houses, in rented rooms and, chiefly, in apartments. The majority of these post secondary students were unmarried and therefore, for the purposes of the income survey, should be classified as unattached individuals.

It would not be surprising to find that most college and university students living independently during the school year would have incomes below the poverty line. Their incomes from summer and part-time jobs, student loans and assistance from parents would be relatively limited. The most recent profile of post secondary students in Canada (1983-84) revealed that average income for full-time students 30 years old and under during the twelve month period from May 1983 to April 1984 was $5,789. [Ibid, p. 31.] The composition of that average income as well as forecasts for the 1988-89 school year applying a 5 percent annual increase in nominal values is contained in table 9-1. The forecasted values are based on the assumption that the source proportions remained the same over the five year period.

Click here to view Table 9-1: Composition of income for full-time post-secondary students30 years old or under

Although the income of the average post secondary student in 1988 is likely to be approximately $7,388, it is not this value that Statistics Canada uses in its income survey. Their definition of income specifically excludes all gifts and loans. This means that the average reported income of full-time students would have been $4,876, about $1,000 less than the Canadian average poverty line for an unattached individual. Therefore, more than half of full-time post secondary students living independently are likely to have reported incomes below the poverty line. [Assuming these incomes are normally distributed.] A fair number will, in fact, have reported incomes below $2500. Anyone who had limited success finding a summer job would most likely be in this category.

It is regrettable, however, that Statistics Canada's income survey does not adequately cover full-time post secondary students. For example, the number of unattached individuals who were enroled as full time students in 1988 and who had taken at least some post secondary education is put at about 148,000 in the microdata file, a roughly three-fold understatement of the true number.

In fairness to StatsCan, they do not attempt to control for the number of students, just the overall number of unattached individuals. There are several good reasons for this. Because the income survey is conducted the week after the 15th of April, it will automatically exclude many university students, whose examination period is normally complete by then. In addition, many parents of post-secondary students include these students on their own survey questionnaire as part of the family, even though the student may be living independently for two-thirds of the year. Finally, the survey specifically excludes student residences in its sampling.

Even with the undercoverage, it is clear that students comprise a substantial proportion of Canada's poor. The microdata file reveals that about 60,000 poor unattached individuals and about 15,000 heads of poor families (or their spouses) were full-time post secondary students in 1988. The percentage, about 8 percent, is likely half the true value. In any case, it is not really surprising that the poverty rates for single persons with some post secondary education is so high. A substantial portion of these individuals are still in school and their limited reported income naturally reflects this fact.

If the definition of "student" is broadened to include all household heads or their spouses who where "in school" at any level during 1988 on a full time basis then an additional 30,000 persons are "poor" students. For most their poverty is overstated due to unreported income sources and the transitory nature of their condition.

Poor students comprise an important qualification to our estimate of the extent of poverty in Canada. The major reason for this is that the reported incomes of many students living independently will belie their real living standard. Statistics Canada income survey explicitly excludes loans and gifts as sources of income. While these exclusions may be of little consequence for the full-time participant in the labour force, they represent a significant omission for full-time students. It has already been pointed out that just over one third of the income of the average post secondary student comes from student loans and non repayable assistance from parents, relatives and friends. [More generally, some portion of young unattached individuals—students or not, have a "claim" on family resources even though they do not live at home [see Osberg (1981), p. 10]. In these cases as well, reported income understates true standard of living.] For some students, especially those with modest summer earnings, loans could easily comprise half or more of the funds they need to support themselves.

The hypothetical case cited in chapter 1 nicely illustrates this point. The student has summer earnings of $3,000, gets $1,500 from parents and receives a $4,500 student loan. The total amount, $9,000, is sufficient to cover food, shelter, tuition and other expenses while living away from home during the school year. In this case Statistics Canada tells us we have an unattached individual whose annual income is $3,000, well below the poverty line. In fact, what we have is a post secondary student whose true income is above the poverty line (as defined here), who lives independently during the school year and who lives with her parents during the summer. There is no suggestion that such students are living in luxury or that they are not making sacrifices. However, they are not desperately poor as their reported incomes would indicate. Their financial resources are, in most cases, adequate to meet their needs.

The relatively low living standard of many post secondary students is a temporary phenomenon. Those that graduate with diplomas and degrees will have lifetime incomes significantly higher than their less well educated counterparts. In addition, the enhanced development of intellectual skills and critical thinking capabilities are their own reward. Every worthwhile investment involves some sacrifice, some costs initially. Post secondary education is no different. In fact those "lean" years are regarded by many as their character building period—a time when some of life's most important lessons are learned. The need to wisely manage time and money, the importance of hard work and discipline, the refinement of personal values and setting of priorities according to those values, the need to become independent of parents, and the need to take full responsibility for your own actions are all lessons that perhaps are most quickly and firmly learned in an atmosphere of relative scarcity.

Unreported and under reported income

Recent estimates indicate that the underground economy in Canada is probably about 5-8 percent of GNP. [See, for example, Mireille Ethier, "The Underground Economy: A Review of the Economic Literature and New Estimates for Canada," in Vaillancourt (1985).] There are two major categories: (1) income earned from activities that are illegal and therefore go unreported and (2) income derived from activities which are legal but which is not reported to evade income taxes. Individuals whose earnings come from illegal activities are unlikely to participate in a Statistics Canada income survey. Some of them may well be poor but will not be counted in the survey. Some of those in the second category may participate; but if they underreport for tax purposes, they are likely to do so in the income survey. An example might be someone who fully reports income from their primary job but income from moonlighting, especially if it is paid in cash, goes unreported. It is unlikely, however, that many in this situation will be poor.

Of far more concern for the purposes of this study is the possible under-reporting of income by people receiving benefits from government programs and by low wage earners. Statistics Canada themselves have examined this problem. The Survey of Consumer Finances (SCF), the basic source of income distribution data underlying all estimates of the incidence of low income, suffers from considerable under-reporting of certain types of income. For example, unemployment insurance is low by about 20 percent and social assistance benefits by about 40 percent in comparison with administrative data on expenditures by the provincial and federal departments concerned. Aggregate wages, on the other hand, are almost identical to the National Accounts figures." [Wolfson and Evans (1990), p. 26.] When StatsCan corrects for this under-reporting, they find that about 200,000 fewer adults and children would be below their poverty line. This is an especially important consideration for this study as almost half of those households determined as poor received social assistance or UIC benefits or both in 1988.

Self employment

Every year there are self employed people who sustain losses. The expenses they incur in the process of doing business exceed their revenue. Some of these individuals and families will be classified as poor because their net income from self employment will be negative and other income will be insufficient to put them above the poverty line. Among the likely candidates here are farmers, fishermen, writers, artists, landlords, and small businessmen. In some cases, the loss may be temporary and the enterprise will bounce back later. In other cases, the business is going or has gone bankrupt during the year. Still another possibility is that there exists rather generous depreciation and other allowances that permit a healthy firm to rather continually show a loss. Whichever is the case, it is likely that reported total income, including net income from self employment, underestimates true living standards.

For example, the fisherman whose catch sells for $5,000 but is able to claim depreciation on his boat and other equipment of $10,000 has a net self employment income of $5,000. If his unemployment insurance benefits are $9,000 for the year, his total reported income is $4,000 which is well below the poverty line for, say, a family of four. However, because depreciation is not an out-of-pocket expense, the true annual income for this family is $14,000 which in most cases is sufficient to cover basic necessities. Clearly, this situation cannot persist in the long run without the fisherman moving to another line of work.

Similar scenarios can be imagined for farmers, landlords, and small businesses. Indeed, Statistics Canada reports that a substantial portion of Canadian farmers do not generate enough net income to sustain a family. [Toronto Star, August 25, 1988.] The point is that the family or individual may not be drawing on just its reported income for its living expenses during the year. It may be dissaving, it may have transfers, such as unemployment insurance or welfare, it may engage in home production or it may have some wage income, all of which are offset by the net losses from self employment.

Statistics Canada's income survey reveals that in 1988 there were about 23,100 poor families whose head was self employed. These families comprised 67,600 persons and had average reported income of $4,634. Their average net income from self employment was -$1,418. For about one-quarter, the major source of income came from government transfers. Forty one percent had zero or negative net income from self employment. The largest group of poor self employed families were farmers, 39 percent of the total. In addition, there were approximately 10,500 poor self employed single persons with average income in 1988 equal to $2,061. About 43 percent had zero or negative net income from self employment. In total then, low (or negative) net earnings from self employment accounts for about 8 percent of the poor—33,600 households or 78,000 persons.

While the average income of these households, about $3,800 for the whole of 1988, appears to be consistent with desperate poverty, there is good reason for skepticism that all are in fact "poor." Home production, in-kind gifts and barter are likely to be important sources of consumption among farmers, fisherman, trappers, and other types of self employed persons. Underreporting of income, especially among small landlords and those who receive government transfers, is likely to be a consideration. However, generous depletion allowances and direct government assistance may be most important in explaining any inconsistencies between income and living standard. After all, we have this large group of 33,600 households whose primary employment activity netted less than zero income, on average, for a whole year. Either this is a transitory situation where the household moves on to more rewarding enterprises or there is something going on behind the scenes. Whichever is the case, it is important to qualify the claim that 8 percent of Canada's poor are self employed persons or their families.

Immigrants

In 1988, Canada admitted approximately 160,000 immigrants. [Employment and Immigration Canada, Annual Report, 1988-89, p. 35. In his 1990 report, the Auditor-General, Mr. Kenneth Dye, reveals that actual immigration exceeded this official total by as much as 20 percent.] For the purpose of the Statistics Canada income survey, only the income earned in Canada is counted as income. Thus, those arriving during the second half of the year and immediately taking jobs would have reported incomes which would significantly misrepresent their true circumstance. The fact that the majority of immigrants destined for the labour force get relatively low paying jobs [Ibid, p. 35.] (i.e., in textiles and other fabricating, assembling industries, clerical, sales, service, manual labour, etc.) means that many in this group could easily have a total 1988 income in Canada falling below the poverty line. Suppose, for example, that a sponsored family of four arrives in Montreal in mid to late spring and the income earner takes a job on July 1 at an annual rate of $20,000. If the family were surveyed the following April, they would give their 1988 income as $10,000 which is almost $4,000 below the poverty line. In fact, however, their annual rate of income puts them over $6,000 above the poverty line.

It is the case, of course, that not all immigrants will find jobs immediately. While they are searching for an appropriate job they may be running down their savings, getting assistance from sponsors, friends, churches, or other charitable organizations, or they may be receiving welfare. The problem is especially acute for the roughly 25,000 refugees and special humanitarian cases who arrive in Canada each year. In virtually all cases they come here with little or no wealth, and few marketable skills. The process of evaluating refugee claimants is a long one and until newcomers have been successfully processed, they may not legally take a job. As of January 1, 1990, there was backlog of 113,000 refugee claimants waiting to be processed. [Toronto Star, March 19, 1990, p. A22.] The point here is that new refugees, living mainly on welfare, government grants, and loans for newcomers (AAP, ISAP, and transportation and admissibility loans), and some charitable assistance are most likely to have reported incomes below the poverty line because these benefits would have been received for only part of the year. To the extent that other immigrants (non-refugees), even those arriving early in the new year, had difficulties finding a job in Canada, even a delay of a few months, their reported incomes could similarly fall below the poverty line. In both cases, the poverty label is likely to be inappropriate because the total resources that newcomers have to draw on for the part of the year they are here will be sufficient to cover at least the basic necessities.

Regrettably, the microdata file does not permit the determination of the number of immigrants in this situation. We do know, from table 8-4, that immigrants arriving in Canada since 1980 have a significantly higher poverty rate than those who arrived earlier, or than the Canadian born. Among the most recent arrivals—those who came here between 1986 and April of 1989—about 12 percent fall below the poverty line. [In this case, the record count in the sample is somewhat below 100 and so the value must be taken with some caution.]

After a period of adjustment, immigrants have done very well in Canada. They have significantly lower poverty rates than those born in Canada. A great many of them left situations of real poverty and almost no opportunity for improvement in their native land. Although most came here with limited skills, a different language and culture and often faced discrimination, their achievements have been remarkable. In general, initial sacrifices, hard work and frugal living have been followed by home ownership, financial security and even greater opportunities for their children. Their incomes may have been low, but they were rarely "poor."

Although anecdotal, the story of the 155 Tamil refugees who were dumped onto the shores of Newfoundland in mid-August 1986, is particularly inspirational. Paula Todd writes:

Since their dramatic rescue from two open lifeboats off the foggy coast of Newfoundland, the 145 men, four women and six children have struggled to rebuild their lives. All of the adults have found employment. Most did so within less than a month of settling in Toronto and Montreal; many hold down two jobs and work 80-hour weeks. Their average annual salary is higher than the Canadian average. Three have started their own business, employing 20 Canadians. [Toronto Star, July 13, 1987.]

People in transition

As was the case with immigrants, there will be people whose situation changes during the reference year in such a way that reported income understates the true standard of living. One obvious example is the predicament of many women after separation or divorce. If they did not have an income earning job prior to the separation, they must either search for a job or depend on welfare and whatever child support payments are provided. For young women with children, welfare is often the sole source of income, at least initially. Because women separating during 1988 receive benefits for only part of the year, for many, reported incomes will fall below the poverty line. Even if they found jobs during the year, a low rate of pay for part of the year would likely result in aggregate calendar year income below poverty level. In either case, the low reported income for 1988 may misrepresent the actual living standard. For the months they were separated, income from earnings or welfare benefits (as will be seen in the next chapter) could well have been sufficient to cover all basic necessities. Prior to separation, their living standard would, in large part, be a reflection of their husband's income. Some women widowed in 1988 would similarly be mis-classified as poor, as would many unmarried mothers as long as they are household heads at survey time, April 1989.

Mary Jo Bane (1986) distinguishes 3 ways in which a female household head and her children might have become poor, in her discussion of transitions into poverty.

She might have been poor before she became a household head, living with her own poor parents or with a husband who was also poor, and in a sense carried that poverty over into her new situation. She might have become poor after she became a household head, either by losing earnings, child support, or another source of income or by adding a member to the family whose increased needs put sufficient pressure on family income to drive them into poverty. Alternatively, she might have become poor at the same time she became a household head; that is, moving from the non poor household of her parents or a marriage into a poor household headed by herself. [Bane (1986), p. 220.]

In the latter case, poverty could have come about even though the total resources available to household members remained about the same. Bane stresses the possible "loss of economies of scale [which] could have made both households poor after the split." [Ibid, p. 221.]

While this analysis does not distinguish between total annual income and rate of income in measuring poverty, a crucial consideration for transitional situations in my view, it does highlight the process by which many women "slip" into poverty. Recent evidence from the U.S. using a longitudinal data base suggests that, for a great many, poverty is indeed transitory. Bane and Ellwood (1986) in their analysis of poverty "spells," state that "most people who slip into poverty are quite successful in getting out." [Bane and Ellwood (1986), p. 12.]

Inmates released from prisons and mental institutions during 1988 would similarly be in transition in terms of annual income. At best, they will find an income earning job for part of the calendar year. At worst, they will resort to welfare. In both cases the income flow may be adequate to cover necessities, but the (low) total income during the year means that they will be classed as "poor." Ironically, for many such persons material living standard is lower even though their income has increased enormously.

According to the Canadian Centre for Justice Statistics, during the 1988-89 period of operation, there were approximately 9,000 persons released from federal and provincial prisons on parole or mandatory supervision. Another 54,000 individuals were released on probation; however, a substantial number of those were released from courts and not prisons. [Statistics Canada, Adult Correctional Services in Canada, 1988-89, cat. 85-211.] Statistics Canada reports that during the 1980s, approximately 30,000 persons are released from Canadian mental institutions each year. About half of those are adults (20 years or older) who had a length of stay greater than 30 days. [Statistics Canada, Mental Health Statistics, Psychiatric Hospitals, cat. 83-204.]

Finally, some young people taking their first full-time job late in 1988 and living independently by April, 1989 would be erroneously included among the poor. The single person starting a job on September 1 or later and earning income at an annual rate of $16,000 or less would be classified as poor in every province but one.

Unfortunately, the microdata file does not contain information permitting the determination of the number of families and individuals classified as poor even though their rate of income was sufficient to cover all necessities. The resolution of this issue will require a reliable longitudinal data base, something not currently available in Canada. Nevertheless, this is, theoretically at least, an important qualification. Considering the very high rate of family breakdown (both formal and informal), it is likely that the issue of transition accounts for a substantial portion of the short term poor.

Voluntary austerity

In 1986, almost 26,000 persons were classified as employed in the service of their religion. [Statistics Canada, Census of Canada, 1986, cat. 93-112, table 1, p. 1-5.] For some of these religious, basic needs are completely taken care of by the order or church they belong to. In other cases, their shelter and food are provided by the church and they receive a modest stipend for personal needs. The remaining earn a modest income from which they provide for all of their needs. To classify any of these as poor is inappropriate even though their reported incomes may fall well below the poverty line. In many cases, they have deliberately shunned worldly possessions to better focus attention on what they consider the most important values. Despite their low income they may regard themselves as "richer" than any multi-millionaire.

There are several religions in Canada that require members to lead an austere, simple life without modern material possessions. Typically they farm, resist outside influences, practice collective economic arrangements and tend to be virtually self sufficient. Included among these groups, with estimated numbers in parenthesis, are Amish/ Menonites (10,000) [Reimer (1984). This estimate is just the number of old order/old colony adult members in 1981-82. Total Amish/Menonite membership exceeds 90,000.] and Hutterites (17,000). [1981 Census of Canada.] Many of them earn no income, and all of them live disciplined, frugal lives. Their day-to-day consumption would be far below our "necessities-based" poverty line, however, I don't know of anyone who would regard them as poor. They represent a modest overestimate of the number living in poverty in Canada.

In addition to these established religious groups, there are an unknown number of communal, back-to-basics groups, as well as families and individuals who choose to live a simple life with little or no money and possessions. Each one is quite unique in its approach to an alternative life style but all participants would invariably be classified as poor on the basis of income.

It is impossible to accurately estimate from available data how many Canadians do not have access to all the basic necessities of life once these qualifications have been considered. However, what information is available suggests that our initial estimate of roughly one million Canadian poor in 1988, based just on reported incomes, is greatly exaggerated. It is probably the case that at least half of those currently defined as poor would not be so classified once home ownership, rent subsidies, students, unreported/under-reported income, negative income from self employment, new immigrants, people in transition and those who, because of their religious commitment have little or no income, are accounted for.

Just as there are qualifications which tend to reduce the number of poor, there are also some which would increase it. It is important to ask the question: Are there categories of families and individuals for whom the estimated "essential" costs clearly understate the true cost of basic needs, and are there poor people who income surveys miss altogether? In the first category are people with special needs that cost significantly more than has been included in estimates of essential costs. Included here would be some disabled persons as well as those who need expensive medicine or treatments. I am not aware of any research which determines the number who are deprived of basic needs because they or another family member must pay out of pocket for costly treatments, equipment, medicines, et cetera, that are not covered by health insurance or other government programs. My own bias is that there are few in this situation and that their predicament is temporary. I am persuaded by the comprehensiveness of health insurance and a variety of public and private programs that people with special, costly needs are not impoverished because of them. However, I offer no evidence in support of this. It is an open question and one that I will leave for others to resolve.

There are, however, three additional qualifications which tend to increase the estimated number of poor in Canada because existing sources omit them. The homeless are not included in the income surveys, nor in the censuses conducted by Statistics Canada. If they had no address at the time of the survey, they had no chance of being counted. Among the excluded would be runaways, bag ladies, ex-psychiatric patients, alcoholics, drug addicts, and transients, virtually all of whom would be living in poverty on a permanent basis. In many cases, they are living outside the system voluntarily, wishing to avoid detection by police, or simply wanting to be left alone. The great dilemma here is that while we recognize that each individual has the right and freedom to live his life in his own way, we also appreciate that some are not capable of caring for themselves. Without question, the long term health and physical well-being of some and perhaps most of the homeless would improve substantially if they acquired shelter and received welfare benefits.

We have no idea how many people are homeless in Canada. [Several estimates are provided in the next chapter.] Both the U.S. Census Bureau (in 1990) and Statistics Canada (in 1991) will attempt, for the first time, to count the homeless. The task is especially difficult because not all of them are easy to find. Nevertheless, I believe it is vitally important to determine how many in our society are without shelter and why. Until we find out this number, our estimate of poverty in Canada is understated to that extent.

The 1988 income survey also excludes two other categories: (1) residents of the Yukon and North West Territories and (2) members of households located on Indian reservations. The first group is excluded because of the remoteness and costliness of inclusion. It is the case that, in 1985, both Yukon and N.W.T. had average household incomes about 5 percent higher than the Canadian average although they had a somewhat greater percentage of households living on less than $10,000 income (11.5 percent versus 8.4 percent). [Statistics Canada, Census of Canada, 1986, table IN86BOIB.] I would estimate that the omission of the two territories understates poverty in Canada by about 6,000 persons in 1988.

The exclusion of Indian reserves is of greater potential importance. In 1986 there were an estimated 200,000 persons living on Indian reserves in Canada. [This is a rough approximation based on two sources: (1) The 1986 Census reveals that reserve population, including an estimate for unenumerated reserves, was 220,000. (2) The Department of Indian Affairs 1987 estimate of the number of people living on reserves was 187,000.] While some earned income off the reserve or by selling items they produced to the general public, for the majority, the bulk of their income comes from social assistance. There exists very little information about the incomes and costs for people living on reservations, but it is generally believed that the rate of poverty, however defined, exceeds the rate for the population as a whole.