Income Research Paper Series
Construction of a Northern Market Basket Measure of poverty for Yukon and the Northwest Territories

The MBM-N methodology proposed in this paper is derived from the MBM methodology currently applied in the provinces. As such, it is comprised of five major components: food, clothing, transportation, shelter and other necessities. The MBM-N is intended to capture the spirit of the existing MBM (i.e., to represent a modest, basic standard of living) while accounting for adjustments needed to reflect life in Yukon and the Northwest Territories.

Acknowledgements

This discussion paper would not have been possible without the invaluable contributions of the Statistics Bureaus of Yukon and the Northwest Territories. We would like to thank them for their support and time in helping create the methodology for the MBM-N, and for their help with reviewing this paper.

Introduction

On August 21, 2018, the Government of Canada released Opportunity for All, Canada’s First Poverty Reduction Strategy, which contained long-term commitments to guide current and future government actions and investments to reduce poverty.

The Poverty Reduction Act legislated key commitments made in the strategy. Opportunity for All indicates that Statistics Canada will be investigating the possibility of producing MBM thresholds for the territories. To this end, Statistics Canada held several discussions with the statistical focal points for Yukon and the Northwest Territories to identify what specific changes, compared to the existing MBM, need to be considered in order to construct MBM thresholds for these territories. These discussions were aimed at building upon a 2012 study by an informal working group^Note  formed to assess the technical feasibility of developing a MBM-N (Statistics Canada internal document, 2012).

This paper describes the proposed methodology for calculating MBM-N basket component costs and disposable income amounts for comparison, and providing the resulting preliminary disposable income thresholds and poverty rate estimates for 2018 and 2019. Readers are encouraged to provide feedback and comments. Although this paper does not address a MBM-N for Nunavut, we welcome feedback and comments related to that territory as well (see text box – A MBM-N for Nunavut).

Working with Indigenous peoples

In Opportunity for All, the Government of Canada committed to “work[ing] with National Indigenous Organizations and others to identify and co-develop indicators of poverty and well-being, including non-income-based measures of poverty, that reflect the multiple dimensions of poverty and well-being experienced by First Nations, Inuit and Métis.”^Note 

It is important to note that discussions around the co-development of new indicators of poverty and well-being for Indigenous peoples are proceeding on a separate track from the MBM-N methodology proposal.

While the proposed MBM-N is based on a methodology which is currently officially used by the Government of Canada to calculate national poverty statistics, the Government is currently working to better understand poverty and well-being from the perspective of Indigenous peoples, and to encourage the co-development of complementary indicators of poverty and well-being, including non-income-based measures of poverty, that reflect the multiple dimensions of poverty and well-being experienced by First Nations, Inuit and Métis. Given that there is a higher proportion of individuals reporting an Indigenous identity in the territories when compared to the rest of Canada, integrating such additional indicators of poverty in analyses pertaining to these regions is particularly important.

Summary of proposed MBM-N methodology

The MBM establishes poverty thresholds based on the cost of a “basket” of food, clothing, shelter, transportation and other necessities. Families with disposable incomes less than the applicable thresholds, given their family size and region of residence, are considered to be living in poverty. The following is the proposed methodology for a 2018-base^Note  MBM-N. Additional details on these proposals were provided in the previous discussion paper on the development of the MBM-N and described in more detail in Appendix A: Proposed methodology for the MBM-N regions, components and disposable income.

Summary of proposed methodology for the costing of MBM-N basket components (described in Proposals for a Northern Market Basket Measure and its disposable income)

MBM-N regions

It is proposed that the MBM-N (2018-base) provide poverty thresholds for nine regions across the two territories: one for each of the capitals and seven additional geographical groupings – two for Yukon^Note  and five for the Northwest Territories^Note  . Under guidance from the statistical focal points, the development of the MBM-N regions responds to the need for sufficient detail on price differences across communities, while keeping price measurement requirements within reasonable parameters.^Note 

Shelter component

The proposed MBM-N shelter components methodology is based on the rental of a 3-bedroom dwelling, in-line with the National Occupancy Standard set out by the Canada Mortgage and Housing Corporation (CMHC). The rental prices for a dwelling would be derived from the 2016 Census data and adjusted to the 2018 base year.

Clothing and footwear component

It is proposed the clothing and footwear component would reflect the Social Planning Council of Winnipeg and Winnipeg Harvest Acceptable Living Level (ALL) 2012 clothing basket for a family of four, while making adjustments to account for the colder climate in the territories.

Food component

The food component would be based on Health Canada’s 2019 National Nutritious Food Basket (NNFB).

Transportation component

The transportation methodology would reflect the differing private transportation usage patterns and needs of people living in the territories. Accordingly, the proposed MBM-N methodology recognizes this by costing a combination of a car and sport utility vehicle (SUV). The fly-in communities of the NWT would have costs associated with owning and operating both an all-terrain vehicle (ATV) and a snowmobile.

Other necessities component

The other necessities component costs would use a “simplified” multiplier^Note  applied to the food and clothing costs, which is consistent with the methodology for the MBM in the provinces. An amount would also be added to reflect the widespread need for cellular telephone services.

Summary of proposed methodology for calculating MBM-N disposable income

Defining disposable income

In general, the MBM-N definition of disposable income would mirror the 2018-base MBM methodology. Specifically, the MBM-N methodology defines disposable income as total income (including government transfers) after deducting not only income tax, but also several non-discretionary expenses (e.g., contributions to the Canada Pension Plan and Employment Insurance, union dues, child care expenses, support payments paid, etc.).

MBM-N thresholds and poverty rates

In this section we describe the resulting estimated MBM-N thresholds and poverty rates under the proposed methodology. These results should be treated as preliminary given that the review phase of this discussion paper is not complete.

Table 1 shows the MBM-N thresholds for a four-person family for each of the MBM-N regions under the proposed methodology for reference years 2018 and 2019.  In general, the MBM-N thresholds for the Northwest Territories are higher than those for Yukon.  The threshold differences between the two territories are mainly due to the cost differences in the shelter and food components. In 2018, the MBM-N region of Rural South in Yukon has the lowest threshold amount at $50,895, while Sahtu in the Northwest Territories has the highest threshold amount at $72,526.  For the territorial capitals, Whitehorse has a MBM-N threshold of $52,215 and Yellowknife has a MBM-N threshold of $59,926.

For the 2018 reference year, all 53 of the MBM regions in the provinces had lower 2018-base thresholds compared to the MBM-N thresholds.  The MBM province thresholds ranged from $37,397 for Quebec communities with a population less than 30,000 people to $48,677 for Vancouver (in current dollars).

The thresholds presented in Table 1 are developed for the reference family of four (two adults^Note  and two children^Note  ).  The same reference family approach is used in the MBM for the provinces.  To calculate thresholds for family sizes that are different from that of the reference family, the MBM-N uses a square root of family size equivalence scale. The equivalence scale method is commonly used to adjust other low-income thresholds (e.g., LIMs and LICOs) to different family sizes, and is recommended for use in poverty measurement (UNECE, 2011).^Note 

Table 2 shows the 2018 and 2019 poverty rates as measured in the Canadian Income Survey (CIS) using the proposed MBM-N thresholds, for selected geographies, age groups and family types.  According to the MBM-N, the preliminary poverty rate decreased in these two territories, from 14.5% in 2018 to 10.8% in 2019. For the preliminary estimates for the number of persons in poverty, please see Appendix B.

The preliminary poverty rate in Yukon went down from 11.1% in 2018 to 9.0% in 2019, while the preliminary poverty rate for the Northwest Territories went from 17.5% to 12.5% over the same period. Estimates for small domains should be used with caution. For more information (see text box – Measuring survey error).

For combined Yukon and the Northwest Territories geography, these decreases were also observed in the selected age groups and family types, from 2018 to 2019. Children^Note  , on average, experienced a larger decrease than seniors^Note  .

Similarly, over the same period, persons living in families experienced a larger decrease than unattached individuals (i.e., persons not in an economic family). None of the changes between 2018 and 2019 in Table 2 were statistically significant (please see next section Survey error within the Territorial estimates).

Survey error within territorial estimates

Chart 1 illustrates the importance of considering the quality indicators provided with every estimate to take into account the variability of the estimates when making a statement.  Although the year-over-year poverty estimates for both territories went down from 14.5% in 2018 to 10.8% in 2019, the percentages could have been anywhere within their respective confidence intervals.  Meaning, the magnitude of the change could have been larger or smaller, or the change could have even gone in the opposite direction.  When the income data for more years becomes available, the analysis of the time series should provide further insights into poverty trends over time. In addition, the application of these thresholds to other income sources with a larger sample size (e.g., 2021 Census for 2020 income), or to administrative data sources could also provide for stronger inferences and more disaggregation.

A MBM-N forward-looking research agenda

A number of research topics were identified during the creation and review process of this paper.  Given their complexity and uncertain feasibility in implementation, these MBM-N-specific topics will require further study before the next MBM-N rebasing exercise.^Note 

Revised Northern Food Basket: Statistics Canada will examine the feasibility of incorporating the Revised Northern Food Basket (RNFB) for the communities outside the capital into a future MBM-N methodology.  The RNFB measures the cost of a nutritious diet for a family of four and is used by Nutrition North Canada (NNC) to monitor the cost of healthy eating in isolated northern communities.  Presently, the pricing of the 2019 NNFB is used to estimate food costs in the capital and spatial indexes are used to adjust these costs for all other communities.

Country Food and Harvesting. Currently the MBM methodology only takes into account food purchased from stores. Country food and harvesting is an important source of nutritious food for many remote communities in the territories.  Statistics Canada will research other approaches that could account for other sources of food in the territories, such as harvesting, trapping, hunting and fishing for country food.

Conclusion

This is the final discussion paper describing the proposed MBM-N methodology for Yukon and the Northwest Territories. It provides and compares provisional MBM-N thresholds and poverty rates, which were the result of collaboration between Statistics Canada, ESDC, and the statistical bureaus of Yukon and the Northwest Territories.

The paper has also proposed some elements for a forward-looking research agenda for the MBM-N, outlining two key topics that Statistics Canada and ESDC intend to study in preparation of the next comprehensive review, scheduled for 2023.

It is important to note that the fundamental purpose of this paper is to foster engagement with key stakeholders.  Therefore, following the release of this paper, there will be a review period during which time Statistics Canada and ESDC will engage with experts, stakeholders, Indigenous organisations and federal, provincial and territorial officials to help validate the results. The review period is expected to end in the beginning of 2022. We encourage readers to contact us to share their views, suggestions and feedback by sending an email to:

statcan.market.basket.measure-mesure.du.panier.de.consommation.statcan@statcan.gc.ca

If access to email is impractical or not possible, we can also be reached by contacting Statistics Canada’s general enquiries line:

• Toll-free number: 1-800-263-1136
• National TTY line: 1-866-753-7083

Appendix A: Proposed methodology for the MBM-N regions, components and disposable income

Proposed methodology for the MBM-N regions

Under guidance from the statistical focal points, the MBM-N regions were based, where appropriate, on the Census Division boundaries, while also taking into consideration similarities between communities in order to make logical geographical groupings. The intent is to have the most detailed price differentials for goods and services for the communities, while keeping the price measurement within reasonable parameters. It is proposed that the MBM-N provide poverty thresholds for nine regions across the two territories: one for each of the capitals and seven additional geographical groupings – two for Yukon and five for the Northwest Territories.

Specifically, for Yukon, the 62^nd parallel would be used to delineate the boundary of Rural South and Rural North, while Whitehorse would be the entire Census Agglomeration (CA). The Yukon Subdivision, Unorganized, which occupies the majority of the territory and spans the two proposed areas, would be included in the Rural North region as approximately two-thirds of its total area is located in the region.^Note  For the Northwest Territories, the five Census Divisions (e.g., Beaufort Delta, Sahtu, Tłįchǫ, Dehcho and South Slave) would be used to group communities, while the Census Division Region 6, which includes Yellowknife, would be used for the capital.^Note 

There are a number of issues that make measuring poverty in the Northwest Territories and Yukon particularly challenging, especially for geographies outside of the territorial capital.  These challenges can impact both the expenditure and income sides of the MBM-N methodology.  For example, the logistical challenges of isolated communities with limited infrastructure can impact the availability of food and clothing price collection for estimating the MBM-N basket costs. Similar logistical challenges, along with the need to protect the confidentiality of the survey response, can impact the availability of income data.  For example, the income source for the annual poverty estimates is the CIS which excludes households in remote areas with a very low population density.  Although this impact is small^Note  , it is important to note because a large proportion of First Nations and Inuit people reside in these excluded areas.^Note

Proposed methodology for the MBM-N basket

The MBM-N basket consists of five components: clothing, food, shelter, transportation and other expenses. In general, each of these components were adapted from the MBM, to reflect life in Yukon and the Northwest Territories.^Note

Also, it is proposed that the replacement schedule (i.e., the lifespan of clothing items) for certain items be adjusted.  During the most recent consultation process with the territorial statistical focal points, it was also proposed that an adjustment should be made to the replacement schedule, to account for the higher use of winter clothes in the territories. Using Environment and Climate Change Canada temperature data from Winnipeg, the city where the 2012 ALL basket is based, and comparing it to a corresponding territorial capital, an adjustment ratio was created.^Note  It is proposed that this ratio be used to adjust the 2012 ALL clothing basket replacement schedule for summer and winter clothes. For example, this ratio would increase how often a family would need to replace typical winter items like scarfs and toques, while decreasing how often they would need to replace typical summer items like bathing suits.

Both of these adjustments reflect the need for warmer clothes in the territories, particularly during the winter months, as well as the more frequent use of certain items due to the cold climate. ^Note 

For areas outside of the capital, a weighted average method is being proposed, and incorporates the costs associated with travel to the closest territorial capital and the cost of ordering online. The price of the trip to the capital would include gasoline, hotel and/or plane tickets if required.^Note  The shipping cost associated with online purchases would be estimated on an annual basis, under the assumption of monthly bulk orders. The weights are based on the percentage of people in the lower end of the income distribution for these regions with internet access, either at home or on a mobile device.

After directly pricing clothing items for the 2018 base year, the clothing component thresholds for subsequent years would be updated using the territory-specific Consumer Price Index (CPI) for “clothing and footwear”.^Note

Food prices were collected in 2018 by Statistics Canada in the territorial capitals. Monthly prices were collected from selected stores in each city and then the geometric mean of each item was calculated in order to determine the annual price for a given city. No adjustments were made to the 2019 NNFB to reflect life in the territories. It should be noted, as with all the other components, that the MBM methodology does not prescribe individual food items that a person should consume, rather it is meant to estimate the typical cost of a nutritious diet for the reference family.

The proposed methodology for estimating the food component costs outside of the capitals would be to use an adjustment factor based on Statistic Canada’s Living Cost Differential data and territorial spatial indexes.  The spatial indexes would be derived by combining Living Cost Differential (LCD) data and data supplied by the territorial statistical agencies, to create community level spatial indexes.^Note  These indexes would then be applied to their respective capital’s costs for the four main food groups that are found in the NNFB (i.e., vegetables and fruit, protein foods, whole grain foods, and unsaturated fats), which would be aggregated to calculate food component cost for a specific community. Finally, the 2016 Census based population weights would be used to aggregate the community level food costs into their respective MBM-N regions.

As with the previous component, direct pricing of the food basket would be for the 2018 base year, while subsequent years’ costs would be adjusted using the territory-specific CPI for “food purchased from stores”.

To determine the typical cost for a modest 3-bedroom rental unit, it is proposed to use a quantile regression model in order to accurately measure shelter costs across all MBM-N regions.  In addition, an amount for tenant’s insurance is also proposed to be included. In consultation with our territorial statistical focal points, it was determined that the vast majority of rental units in the territories included appliances as part of the rent. Therefore, the MBM-N shelter component would not include a separate category of expenses related to the purchase of appliances.

The proposed methodology for estimating the shelter component costs outside of the capital would be the same as the capital, but the shelter costs would be calculated separately for each of the MBM-N regions.

Finally, in the MBM-N, it is proposed that the territory-specific all-items index of the CPI be used to update the value of the shelter component for subsequent reference years.^Note

The proposed methodology for estimating costs outside of the capital would depend on whether the communities have road access. For communities with road access, the costs of the transportation component would be the same as in the capital, but with adjustments based on the registration weights to account for the higher prevalence of SUVs versus compact cars outside of the capital. In addition, for these regions, it is proposed that a small nominal cost be added to account for the higher maintenance cost and higher number of kilometres typically driven compared to the capital.

For the fly-in communities in the Northwest Territories, the transportation costs would be based on the costs associated with purchasing a new mid-quality range All-Terrain Vehicle (ATV) and a snowmobile from Hay River and shipping it by barge to the fly-in community, as well as the costs associated with operating and maintaining them.^Note 

Finally, it is proposed that the appropriate territorial level index for private transportation be used to adjust the transportation component for subsequent reference years.

The proposed methodology for the other necessities component in and outside of the capital would use the same methodology. The 2018-base MBM multiplier^Note  would be applied to the food and clothing component threshold costs for each of the MBM-N regions.

Finally, in the MBM-N, it is proposed that the territory specific all-items index of the CPI be used to update the value of the other necessities component for subsequent reference years.

Proposed methodology for the MBM-N disposable income

In the MBM-N methodology, disposable income is defined as total income (including government transfers) after deducting not only income tax, but also several non-discretionary expenses including Canada Pension Plan and Quebec Pension Plan contributions, Employment Insurance and Registered Pension Plan contributions, union dues, child care expenses, support payments paid, and direct medical expenses including private insurance premiums.

In addition to accounting for different housing tenure types, it is proposed that the MBM-N follow the other changes introduced by the 2018-base MBM (i.e., medical expense imputation amount and capital gains tax adjustments).^Note 

Tenure Type Adjustment

As was done in the 2018-base MBM, it is proposed that the MBM-N methodology would make an adjustment, where possible, for other housing tenure types other than rental dwellings, such as homeownership with or without mortgages and those residing in subsidised dwellings. This would serve to place families in all tenure types on a more equal footing to those families in rental units, when determining their poverty status.

Homeowners without mortgages

For homeowners without a mortgage, the shelter costs considered for the MBM consist of property taxes, utility costs, condominium fees and basic home insurance costs. As noted above, these are estimated from 2016 Census data for a three-bedroom unit for a family in the second income decile for each MBM-N region.

Homeowners with mortgages

For the purpose of the MBM-N, shelter costs for homeowners with a mortgage would consist of property taxes, utility costs, condominium fees, basic home insurance costs and mortgage interest payments. The key difference, relative to homeowners without a mortgage, is the interest cost component (which is zero for homeowners without a mortgage).

Given the data limitations, a few changes compared to the 2018-base MBM were needed, in the calculation of the tenure type costs for homeowners with a mortgage.  First, the mortgage interest costs were estimated using data from the 2018 Canadian Housing Survey (CHS), not the Survey of Financial Security (SFS). Second, the calculation was done for mortgages on three-bedroom dwellings only and with no age group differences. Finally, only homeowners in the second income decile were considered when estimating the shelter cost medians.

Subsidized renters

For tenants living in subsidized housing, TTAs are determined by estimating the median cost of a subsidized three-bedroom rental unit for a family in the second decile of income, and then measuring the difference between its cost and corresponding shelter cost in the basket component of the MBM. For subsidised renters, shelter costs consist of the rent paid plus utility costs and relevant insurance costs.^Note 

Capital gains taxes

A capital gain or capital loss occurs when an individual disposes of, or transfers, a capital property (e.g., land, buildings, stocks, bonds, mutual funds, etc.). Since the MBM-N does not include the values of capital gains/losses in the definition of disposable income, yet deducts the income taxes, an adjustment to the disposable income is made to take into account the tax implications of capital gains/losses.  

The taxes implications of capital gains/losses are not directly known from available tax information, they are first estimated and then the disposable income adjusted accordingly when administrative data indicates a capital gain/loss. ^Note 

The effective tax rate would be calculated as the ratio of total tax paid divided by the sum of total income and the taxable amount of the capital gain/loss. This rate would then be multiplied by the taxable amount of the capital gain/loss. The disposable income would then be adjusted by the amount of tax attributed to the capital gain/loss.

Appendix B: Number of persons in poverty

Appendix C: MBM-N Regions

Appendix D: MBM-N cold climate adjustment factors 

Appendix E: MBM-N spatial index adjustment factors

Formula E.2: MBM-N food component formula for regions outside the capitals

The methodology for estimating the food costs outside of the capitals could be expressed in the following way.

First, indexes for each food group would be derived by averaging Living Cost Differential (LCD) data, ${\text{I}}_c^{{\text{LCD}}_i{\text{FG}}_i}$ , and data supplied by the territorial statistical agencies, ${\text{I}}_c^{{\text{T}}_i{\text{FG}}_i}$ , to create community level spatial indexes. A weighted sum would then be applied to the community indexes to obtain MBM-N region indexes, ${\text{I}}_n^{{\text{FG}}_i}$

where:
${\text{I}}_c^{{\text{T}}_i{\text{FG}}_i}$ = Territorial provided special index for a given food grouping $i$ , in community $c$ ,
${\text{I}}_c^{{\text{LCD}}_i{\text{FG}}_i}$ = Living Cost Differential based special index for a given food grouping $i$ , in community $c$,
$w_c$ = 2016 Census based population weight for community $c$ .

Second, these indexes would then be applied to their respective capital’s costs for the four main food groups that are found in the NNFB (i.e., vegetables and fruit, protein foods, whole grain foods, and unsaturated fats).

Finally, the four main food group indexes are aggregated to calculate the food component cost for a specific MBM-N region, ${\text{FC}}_n$ .

where:
${\text{I}}_n^{{\text{FG}}_i}$ = Weighted average spatial index for a given food grouping $i$ , in a geographical grouping $n$ ,

${\text{FC}}_{\text{CAP}}^{{\text{FG}}_i}$ = Food cost in the capital for a given food grouping $i$ .

Appendix F: Methodological comparison between the MBM (2018-base) and MBM-N 

References

Djidel, Samir, Burton Gustajtis, Andrew Heisz, Keith Lam, Isabelle Marchand and Sarah McDermott (2020), “Report on the second comprehensive review of the Market Basket Measure”. Catalogue No. 75F0002M2020002.

Djidel, Samir, Burton Gustajtis, Andrew Heisz, Keith Lam and Sarah McDermott (2019a), “Towards an update of the Market Basket”. Catalogue no. 75F0002M2019013.

Djidel, Samir, Burton Gustajtis, Andrew Heisz, Keith Lam and Sarah McDermott (2019b), “Defining disposable income in the Market Basket Measure”. Catalogue no. 75F0002M2019014.

Employment and Social Development Canada, 2018. “Opportunity for All – Canada’s First Poverty Reduction Strategy”, Cat. No. : SSD-212-08-18E.

Gustajtis, Burton, Keith Lam and Sarah McDermott (2021), “Proposals for a Northern Market Basket Measure and its disposable income”. Catalogue no. 75F0002M2021001.

Hatfield, Michael, Wendy Pyper and Burton Gustajtis. (2010), “First Comprehensive Review of the Market Basket Measure of Low Income”, Applied Research Branch paper, Human Resources and Skills Development Canada.

Heisz, Andrew (2019), “An update on the Market Basket Measure comprehensive review”. Catalogue no. 75F0002M2019009.

Statistics Canada, 2019. “Methodological changes to the Market Basket Measure in 2019”. Catalogue no. 75F0002M2019005

Statistics Canada, 2016. “Low Income Lines: What they are and how they are created”. Catalogue no. 75F0002M2016002

Statistics Canada, 2012. Northern Market Basket Measure Feasibility Study. Unpublished document.

UNECE (2011), “Canberra Group Handbook on Household Income Statistics”, Second Edition.