Understanding the 2000 Low Income Statistics Based on the Market Basket Measure

Appendix A

Methodological Annex

The value added by the MBM to other tools to assess low income can best be understood by comparing and contrasting it with those tools.

Pre-income tax Low Income Cut-offs (LICOs-IBT)

This version of the LICOs represents a level of total income before the payment of income taxes at which, for a family of a given size and living in a rural or urban community of a given population level, the share of that income it would spend on food, clothing and footwear and shelter is twenty percentage points above that spent by the average family on these three categories of expenditure.

The income levels at which this occurs are calculated using econometric regressions for five different community sizes and for economic families ranging from one person to seven or more persons, producing thirty-five different low income cut-offs in all.⁹ The current LICOs are based on 1992 expenditure patterns. They are updated each year to take into account changes in the Consumer Price Index for Canada.

The LICOs-IBT thus answer the question: How many Canadians live in families spending a share of their total pre-tax income on food, clothing and shelter twenty percentage points higher than average families of the same size living in the same broad community size did in 1992?¹⁰

Post- income tax Low Income Measure (LIM-IAT)

The post-income tax Low Income Measure is 50% of median adjusted family income using a post-income tax definition of income.

The word "adjusted" means that the median family post-income tax income is calculated in such a way as to take into account the fact that families of different sizes and compositions have different needs.

The way in which this is done is as follows: First, factors are assigned to each family member using an equivalence scale. In this scale the oldest person in the family receives a factor of one, the second oldest person in the family and all other family members aged 16 and over each receive a factor of 0.4 and all other family members under the age of 16 receive a factor of 0.3.

The values for each person in the family are added to determine the adjusted family size. The income of the family after the payment of income taxes is then divided by this sum. For example a couple with two children under age 16 would have an "adjusted family size" of 2 (1+0.4+0.3+0.3) and its total income would be divided by two. Single adults living alone would have an "adjusted family size" of one.

Accordingly, the equivalence scale assumes that the family of four requires twice as much post-income tax income to meet its household needs as an unattached adult living alone.

The median adjusted post-income tax family income is then determined. This is the adjusted post-income tax family income such that half of all economic families will be above and half below it.

The post-income tax LIM for an unattached individual is 50% of this amount. The LIMs for all other family configurations are equal to this amount multiplied by their "adjusted family size." As a result, a couple with two children would have a low-income cut-off twice that of a single adult living alone.

The LIM-IAT cut-offs are not adjusted for differences in community size. The threshold for a family of any given configuration is exactly the same regardless of the size of the community in which they live.

The LIM-IAT therefore, answers the question: How many Canadians have an after-tax income lower than 50% of the adjusted median income for all Canadian economic families in a given year?

It is a pure measure of relative low income and is automatically adjusted each year for changes in median family post-income tax income levels adjusted using its equivalence scale.

The Market Basket Measure (MBM)

Unlike either the LICOs or the LIM-IAT, the MBM is based on a specific basket of goods and services.

The cost of the goods and services in the MBM is calculated for a reference family of one male and one female adult aged 25-49 with two children, a girl aged 9 and a boy aged 13.¹¹ Despite recent trends showing an increasing share of persons living in alternative household configurations such as childless couples, lone parent families and unattached individuals, the two-parent, two child household is still the household type containing the largest share of Canada's population.

The cost of the goods and services in the "Market Basket" for all other household configurations is then calculated using the Low Income Measure equivalence scale (See the section on the LIM-IAT for a description of this scale).

For household sizes up to four, this equivalence scale is almost identical to that used to calculate the relative measure of disposable income poverty used by the United Nations and the Luxembourg Income Study (LIS). Their equivalence scale is simply the square root of household size.

The British poverty analyst, Anthony Atkinson, after an extensive study of the many equivalence scales used in developed countries, concluded that they tend to converge around the value of the square root of household size.

The choice of the LIM equivalence scale is therefore consistent with international practice, while reflecting judgements made by Statistics Canada in the Canadian context.

Distinctive Features of the MBM

The MBM is thus a "goods and services" indicator of low income whose cost is calculated for a number of specific urban communities and community sizes across Canada. As such, it can be used to answer a question not addressed by either the LICOs or the LIM-IAT: How many people in Canada live in families which lack the disposable income to purchase the goods and services in the "Market Basket" within their community or community size?

A second feature of the MBM is that it is more sensitive than either the LICOs or the LIM-IAT to differences in living costs among different communities and community sizes across Canada. This is because the MBM thresholds vary with the cost of the goods and services in the basket, not only between community sizes, but between communities of similar size in different provinces.

This sensitivity to geographical differences in living costs and the transparency of the goods and services in the MBM were the features sought in the development of this new tool to assess low income.

Finally, the basic concept of low income underlying the MBM is being unable to purchase the goods and services in the Market Basket. This implies that the income to be compared to the thresholds should not be gross income, but a measure of the disposable income actually available to purchase these goods and services.

As a result, the following deductions are made from total family money income before comparing it to the cost of the basket:

out of pocket spending on child care;¹²
out of pocket spending on non-insured health care costs recommended by a health professional such as prescription drugs, health insurance premiums, aids for persons with disabilities and dental and vision care;
personal income taxes and the personal portion of all payroll taxes such as Canada/Quebec Pension Plan contributions and Employment Insurance premiums;
alimony and child support payments made to another family;
all mandatory payroll deductions for employer-sponsored pension plans, union dues and employer-sponsored supplementary health care plans.

Expenditures on support payments, out-of-pocket child care expenses and mandatory payroll deductions other than EI premiums and C/QPP contributions are derived from responses to questions on Statistics Canada's Survey of Labour and Income Dynamics (SLID).

C/QPP contributions and EI premiums were calculated based on earnings and published contribution rates. Public health insurance premiums were based on provincial contribution schedules and net income.

For those who reported positive direct out-of-pocket medical expenses on line 330 of the income tax form this amount was used.¹³ Otherwise amounts were imputed for each province from data from the Survey of Household Spending (SHS).

All of these deductions represent income that is not available to purchase the goods and services in the basket. As can be seen, this is a much more stringent concept of disposable income than that used for either the LICOs (which make no deductions from total money income before comparing that income to the LICO thresholds) or the LIM-IAT (which deducts only income taxes paid before comparing adjusted post-income tax family income to the LIM-IAT thresholds).

Consequently, even where the MBM threshold for a given family in a given community is lower than that calculated using the LICOs or the LIM-IAT, that family's MBM disposable income may fall under the MBM threshold even though its total or post-income tax income may exceed the thresholds for the other two measures. Such a family would be counted as part of the low-income population using the MBM but not as low income using the LICOs or the LIM-IAT.

The Composition of the MBM Basket

a) Food

The content of the food component of the MBM basket is as described in the Health Canada publication, National Nutritious Food Basket 1998, written by Judith Lawn.¹⁴ The basket represents community standards of food expenditure in Canada as derived from Statistics Canada's Survey of Family Food Expenditure in Canada 1996 and has been adjusted to be consistent with Health Canada's Nutrition Recommendations and current guidelines for fat and saturated fat intake for adults.

It is neither "an ideal diet" nor the cheapest diet which meets nutritional requirements. Instead it represents a nutritious diet which is consistent with the food purchases of ordinary Canadian households. It contains healthy foods that "people like to eat." It is designed to be "socially acceptable and contain sufficient variety to be nutritionally adequate and palatable over the long term." It includes more costly "basic processed foods such as yogurt or bread....since a family would not normally prepare those foods from raw ingredients."

The publication lists the amount of each type of food that would be purchased each week and the suggested purchase unit for the reference family. From these tables Statistics Canada was able to determine the annual cost of the food basket in the forty urban centres where it collects food price data.¹⁵

For example, in Ottawa in January of 2000 the average price for the standard quantity of 2% milk (a four litre bag) for the reference family was $3.49. Since the Nutritious Food Basket recommended an average weekly purchase of 10.45 litres, the weekly cost of milk for the family was (10.45litres/4.0 litres) x $3.49 = $9.12. This same procedure was followed for all the items in the food basket each month and the total average weekly cost for the twelve months was multiplied by fifty-two to obtain the annual cost.

The content of the food component of the MBM is provided in Appendix B, including the suggested purchase unit and the weekly quantities of food purchased.

The annual cost of each of the five components of the MBM basket for the nineteen urban areas and twenty-nine community sizes where a threshold for the reference family was calculated is provided in Appendix G.

The annual cost for the food component in 2000 ranged from $5331 in Hamilton, Ontario to $7623 in Prince George, British Columbia. The expenditure on food by the median reference family (including food purchased in restaurants) in 2000 was $7400. Thus the cost of the MBM food component ranges from 72% to 103% of this national median level.

b) Clothing and Footwear

In 1997, Winnipeg Harvest and the Winnipeg Social Planning Council developed a budget guide for families in the Winnipeg Census Metropolitan Area which they named the Acceptable Level of Living (A.L.L) measure. In 1999 the Federal-Provincial-Territorial Working Group on Social Development Research and Information chose the clothing and footwear component of the A.L.L. for the MBM because it:

was the most recent clothing and footwear "basket" developed in Canada;
reflected an effort to provide clothing and footwear for common work, school and social occasions, a standard similar to that aimed for by the MBM;
and had significant input from low-income persons.

For the 2000, 2001 and 2002 income years the A.L.L. clothing and footwear items will serve as an interim specification for this component of the MBM.

For the income year 2000 it was calculated as follows: First the cost of the clothing and footwear component of the A.L.L. basket for the reference family was determined by Winnipeg Harvest and the Winnipeg Social Planning Council. The content of the items of clothing and footwear comprising this component of the basket and their replacement schedule are provided in Appendix C.

Next, to determine the cost of this component of the basket in other urban centres the Prices Division of Statistics Canada used the relative spatial indices for clothing and footwear in October 1999 as published in Table 12 of Statistics Canada's monthly publication, The Consumer Price Index. These indices calculate the relative costs of various components of expenditure in at least one urban centre in each province.

Using the cost of the clothing and footwear component of the A.L.L. in Winnipeg as the base, the approximate cost of that component was determined in the other urban centres across Canada. The assumption was that the cost of clothing and footwear in the urban centres in each province for which the relative spatial price indices were available approximated that in other communities and community sizes within the same jurisdiction.

The reasons why the A.L.L. clothing and footwear component can serve only as an interim specification are as follows:

The quality of the items of clothing and footwear listed in the A.L.L. are not specified in sufficient detail for Statistics Canada to collect pricing data on a consistent basis across the country.
Just over half of all families of two adults and two children in Winnipeg spent more on clothing and footwear than did the reference family purchasing only the items in the A.L.L. clothing and footwear basket. This represents a standard somewhat above that aimed for by the MBM.

An alternative clothing and footwear component has been developed that is based on the A.L.L. clothing and footwear component, but is more specific in describing the quantity and quality of the items listed and uses a different replacement schedule. Data on the cost of this revised clothing and footwear component is being collected beginning in 2003.

The cost of the interim clothing and footwear component in 2000 ranged from $2110 in Charlottetown to $2292 in St. John's, Halifax, Toronto and Vancouver. The median cost nationally for clothing and footwear in 2000 for the reference family was $2450. Thus, the MBM cost levels fell between 86% and 93.6% of the overall median level.¹⁶

c) Shelter

The shelter component of the MBM reflects the average of the median rents for two-bedroom and three-bedroom rental units for each community and community size in each province where the number of observations permitted a statistically reliable calculation. Households whose rents were subsidised were included in the sample, but those paying no rent were excluded as were rental units requiring major repairs.

The choice of the average of the median rents for two and three-bedroom units was made because approximately half of two-adult, two-child renting families live in each of these two types of units.

The median rent was chosen to ensure a decent quality of housing even in areas where there is a limited supply of available low-cost housing. Of course many low-income households will pay less than this amount for shelter, either because their rent is subsidised or because they are homeowners who have paid off the mortgage on their home.¹⁷

The rent component includes utilities (water, heat and electricity) as well as the following amenities: a stove, a refrigerator and the use of a clothes washer and clothes dryer. In cases where some or all of these items were not included in the rent, Statistics Canada adjusted costs as described below.

Three sources of data were used by Statistics Canada to calculate median rent levels adjusted for the cost of utilities and amenities. These were housing data from the 1996 Census long form, the rent supplement to the monthly Labour Force Survey and the annual Survey of Household Spending.

Median rent levels (including utilities) in 1995 for the two types of units were calculated from the 1996 Census for each community and community size in each province. The census provides information on whether electricity, heat and water costs are included in the rent and, if not, the costs of these utilities. These amounts were then updated to 2000 levels for each province using the provincial Consumer Price Index for rental accommodation.

Inclusion of amenities in the rent varied widely between provinces. For example 91% of two-bedroom units in British Columbia included a refrigerator in the rent compared to only 12% in Québec. Therefore Statistics Canada made a further adjustment to median monthly rent levels. This was done by adding the products of the percentage of rental units without each amenity in each province times the monthly amortised cost of purchasing that amenity in the second decile of the reference family. These amounts were derived from the 1997-1999 average expenditures on these amenities in the second decile of the reference family as calculated from the annual Survey of Household Spending.

There were sufficient observations in the Census to calculate the average of the median adjusted rental levels for two and three-bedroom units for nineteen distinct urban areas and twenty-nine community sizes in the ten provinces. These were then averaged and multiplied by twelve to generate the cost of the shelter component for each of these forty-eight geographical areas.

As expected, variations in shelter costs were much wider than for food or for clothing and footwear. Annual shelter costs ranged from $4965 in rural Manitoba to $11,399 in Toronto. Generally speaking, as can be seen in Appendix G, the cost of shelter rises with the size of the community within provinces.

The actual median shelter cost for all two-adult, two-child Canadian families (including homeowners) in the 2000 Survey of Household Spending was $12,215. MBM shelter costs in rural Manitoba thus represented 41% of this median while those in Toronto were 93% of the national median.

d) Transportation

The transportation component of the MBM largely follows the recommendations of the National Council of Welfare in its publication, A New Poverty Line: Yes, No or Maybe?¹⁸ These recommendations are based on the insight that in contrast to the cost of shelter, the cost of basic transportation is generally less expensive in large urban areas than in smaller communities or rural Canada.

This is because in large urban centres public transit passes can provide access to a wide range of shopping outlets, professional services and employment and learning opportunities that can be matched in areas not served by public transit systems only by purchasing and maintaining an automobile.¹⁹

As a result, in urban centres served by a public transit system the transportation component of the basket consists of the annual cost of two adult monthly transit passes plus one round trip taxi ride a month costing $16 to accommodate a shopping expedition where large items which cannot be carried by hand, are purchased. The $16 amount will be adjusted annually to reflect changes in the Consumer Price Index for taxi rides for the province as a whole.

Statistics Canada determined that all but 3 of 49 urban centres with a population of 30,000 or more had public transit systems. Accordingly, in all centres of this size the transportation component described in the preceding paragraph was used.

In all other areas, including Charlottetown which has a population of over 30,000 but no public transit system, the transportation component of the basket consisted of the cost of paying for and operating a five-year old four-door, four-cylinder Chevrolet Cavalier.²⁰ This consists of the following items:

20% of the cost of a 1995 model of this vehicle including interest charges on a 36 month loan for the vehicle's purchase price;
the annual cost of an adult driver's license fee;
the annual cost of registering the vehicle;
the cost of annual mandatory insurance for the vehicle;
the cost of 1500 litres of regular unleaded gasoline for the vehicle;
the cost of two oil changes and one tune-up annually.

These costs were estimated separately for each province. The insurance cost assumes that the vehicle is driven to and from work and that the adult driver has not had an accident in the past six years.

In urban centres served by public transit the cost of the transportation component ranged from $1169 in Québec urban centres with a population between 30,000 and 99,999, to $2316 in Toronto.

In areas not served by public transit the cost of the transportation component ranged from $3517 in Alberta to $4113 in Manitoba. The main source of the variation was in the level of insurance coverage mandated in the different provinces.

For a list of the cities in which transportation costs are collected by Statistics Canada according to community size and the type of data collected for the MBM see Appendix E.

The median amount spent on transportation by all two-adult two child families was $6028.²¹ Thus the amounts in smaller urban and rural areas represent a range between 58% and 68% of the national median.

Because rural and small urban area transportation costs are higher than for larger urban centres, the overall MBM thresholds for the different community sizes within each jurisdiction are much closer together than they are for Statistics Canada's Low Income Cut-offs. That measure is based on how food, clothing and footwear and shelter spending is related to income and thus reflects the fact that shelter costs are higher in larger urban centres. However, it does not take into account transportation costs.

e) Other Goods and Services

There are several other goods and services that are encompassed by the MBM standard of consumption. The category "Other Goods and Services" includes expenditures on personal care, household needs, furniture (excluding the items included under shelter), basic telephone service, postage stamps, religious and charitable donations, school supplies and modest levels of reading material, recreation and entertainment. The reading, recreation and entertainment component includes a newspaper subscription, video rentals, YM/YWCA memberships, magazines, books and tickets for movies and sports events. The items included in the Other Goods and Services category are detailed in Appendix F.

Separately these goods and services comprise much smaller percentages of overall spending than food, clothing and footwear, shelter and transportation. Moreover, as with out-of-pocket spending for child care it is difficult to compile a standard basket component for these items.

As a result, it was decided to approximate the cost of these goods and services using a multiplier representing expenditures on them as a proportion of average spending on food and clothing and footwear by the second decile of the reference family.²² The multiplier will be calculated each year using the detailed micro data from the main file of the Survey of Household Spending.

The spatial price indices calculated by Statistics Canada for these other goods and services for eleven urban centres across Canada vary in a range closer to that for food and clothing and footwear than to that for shelter and transportation. Thus expenditures for shelter and for transportation were not taken into account when calculating the multiplier. These vary much more widely between communities and community types (depending on whether they are served by public transit systems).²³

This is the one component of the MBM basket whose cost is calculated using a "relative" methodology rather than being based on actual prices of specific goods and services.

The multiplier for 2000 based on a three-year moving average (1997-1999) of the ratio of spending on these items to spending on food and clothing and footwear in the second decile of the reference family is 68.9%. Thus, in each community and community size the combined expenditure on food and clothing and footwear in 2000 was multiplied by 0.689 to determine the cost of all the other goods and services listed in Appendix F.

Since the estimated cost of the Other Goods and Services is linked to the estimated costs for food and clothing and footwear, if the latter are out of line with the standard of consumption aimed at by the MBM the error will be compounded through the multiplier. This is why a revision to the clothing and footwear component of the basket has been undertaken.

The cost of these other items ranged from $5249 in Hamilton to $6828 in rural British Columbia.

Total median expenditures on all goods and services by the reference family in 2000 amounted to $44,212.²⁴ The overall MBM thresholds ranged from $21,182 in Québec urban centres with populations between 30,000 and 99,999 to $28,752 in British Columbia urban centres with a population under 30,000; representing a range from 48% to 65% of median consumption spending for the reference family.

⁹The calculation of thresholds for five community sizes is in recognition of the fact that, generally speaking, shelter costs rise with the size of the community, being lowest in rural areas and highest in the largest urban centres. The five community sizes are census metropolitan areas (CMAs) with a population of 500,000 or more, CMAs with a population of 100,000-499,999, urban census agglomerations (CAs) with a population between 30,000 and 99,999, CAs and small urban centres with a population under 30,000 and rural areas (including communities with populations under 1000 not contained in a CMA or CA).
¹⁰In 1992 the average family spent 34.7% of its total pre-tax income on food, clothing and footwear and shelter. Thus the Low Income Cut-offs were set at the point in the income distribution where a family would spend 54.7% or more of its income on these three categories of expenditure. In 1992 the average family spent 44% of its post-income tax income on food, clothing and footwear and shelter. Thus, the post-income tax Low Income Cut-offs (LICOs-IAT) were set at the point in the income distribution where a family would spend 64% or more of its post- income tax income on these three categories of expenditure. Beginning with the National Child Benefit Progress Report: 2002, the LICOs-IAT will be used along with the LICOs-IBT to monitor the incidence and depth of low income among families with children.
¹¹The ages of the family members must be specified to calculate the cost of their expenditures on food and clothing and footwear which vary with the age and gender of the household members.
¹²Out of pocket spending on child care and non-insured health care spending recommended by a health professional are not included in the cost of the basket because spending on these items varies so widely from family to family depending on the availability of free or subsidized child care and the health needs of family members. No "standard" basket component for either category of expenditure could be reasonably set. However, families that must spend significant amounts of money on such items obviously experience more difficulty purchasing goods and services to meet their other household needs than those who do not have to bear such costs. This is accounted for by deducting the amount spent on these items from each family's total money income before comparing it to the MBM thresholds.
¹³About 75% of those interviewed for the SLID gave permission to extract data from their income tax returns. All those who did so but did not report a positive amount on line 330 had the amount imputed in the same way as those who did not give access to their tax returns.
¹⁴See Judith Lawn, National Nutritious Food Basket 1998, Health Canada (Ottawa, 1998). This publication was compiled under contract with the Nutrition and Healthy Eating Unit of the Health Promotion and Programs Branch of Health Canada. It reflected input from Federal, Provincial and Territorial nutritionists.
¹⁵It is assumed that in each province the cost of the food basket in rural areas is the same as in the smallest urban centre for which food price data are collected. This assumption is currently being tested by Statistics Canada in rural and urban centres in five provinces (British Columbia, Ontario, Québec, New Brunswick and Newfoundland and Labrador) using an independent source of food price data for the year 2000.
¹⁶The overall median expenditure level on clothing and footwear for the reference family includes jewellery and accessories. Neither of these is included in the A.L.L. clothing component.
¹⁷Homeowners with no mortgage still must pay shelter costs such as property taxes, utilities and home insurance, but these are usually less than the costs of rents which take these costs into account. However, at present there is no data source available which calculates, for each household, the value of rent subsidies paid in the form of rent-geared-to-income rents or the actual shelter costs of homeowners who have paid off their mortgages . This lack of data affects all low-income measures since the lower shelter costs faced by such households should be considered a form of imputed income. Its impact for the MBM is to overestimate the number of persons in families who lack the disposable income to purchase the standard of consumption represented by the MBM basket of goods and services. This overestimate is likely to be particularly important for persons over age 65 and for residents of the rural portions of the Atlantic Provinces, Manitoba and Saskatchewan where the proportion of households who own their residence without a mortgage is well above the average for Canada.
¹⁸ See National Council of Welfare, A New Poverty Line: Yes, No or Maybe? (Ottawa: Winter 1998-99), p.24.
¹⁹The National Council of Welfare did not include the cost of purchasing the car in its recommendations.
²⁰The cost of this component of the basket is highly sensitive to the age of the car. If a six-year old car were purchased every six years instead of a five-year old car every five years, the cost of transportation in areas outside those served by public transit systems would be reduced by $900. This particular model was chosen because it is widely available in used car outlets across Canada.
²¹This includes transportation costs such as spending on inter-city train, bus and airplane tickets not included in the MBM transportation component in either rural or urban areas.
²²The multiplier was calculated using the expenditure patterns of the second decile because, since 1980, the low income rate for families of 4 persons using Statistics Canada's 1992 base pre-income tax Low Income Cut-offs has never exceeded 15%, the mid-point of the second decile. In 2000, families in the second decile of the post-income tax income distribution had post income-tax incomes between $27,690 and $34,783.
²³For the eleven cities surveyed to compile the relative spatial price indices in October 1999, the cost of shelter in the cheapest city was 61% below that in the most expensive city, for public transportation it was 34% lower. However, for clothing and footwear the differential was 9%, for food it was 7% and for household operations and furnishings it was also 7%.
²⁴The median reference family had a pre-tax income of $65,000 in 2000 according to the Survey of Household Spending. Its expenditures included spending on child care costs and out-of-pocket health care costs.

Contents

Understanding the 2000 Low Income Statistics Based on the Market Basket Measure - May 2003