Homeowner versus non-Homeowner Report 2022

A R T I C L E
Homeowner versus non-homeowner differences
in household food insecurity in Canada
Lynn McIntyre1 • Xiuyun Wu1 • Valerie C. Fleisch 1 •
J. C. Herbert Emery 2
Received: 7 July 2014 / Accepted: 10 July 2015 / Published online: 19 July 2015
Ó Springer Science+Business Media Dordrecht 2015
Abstract The risk of food insecurity, lack of access to adequate food because of financial
constraints, is low for homeowners relative to renters in Canada; yet it is unclear if this is
due to the characteristics of who owns versus who rents, or a direct protective effect of
homeownership over renting. We examined this question by looking at the correlates of
food insecurity among households by homeownership status. We used a population-based
sample, the 2009–2010 Canadian Community Health Survey, in which both housing tenure
and food insecurity were measured. A decomposition approach allowed us to examine the
difference in prevalence of food insecurity between non-homeowner and homeowner
households that was not accounted for by household-level characteristics such as income or
contextual factors. As expected, household food insecurity was much lower among
homeowner households (3.3 %) than non-homeowner households (17.9 %). Household
and contextual characteristics accounted for 71 % of the overall difference in the odds of
being food insecure, leaving 29 % of the gap attributable to the protective impact of
homeownership. Closing this gap could include the introduction of institutional policies
that mirror the protection from home equity and governmental policy supports afforded to
homeownership.
Keywords Food insecurity  Housing  Homeownership  Rental  Decomposition
statistics  Canada
& Lynn McIntyre
[e-mail]
1 Department of Community Health Sciences, Cumming School of Medicine, University of Calgary,
Teaching Research and Wellness (TRW) Building, Room 3E14 (3rd Floor), 3280 Hospital Dr. NW,
Calgary, AB T2N 4Z6, Canada
2 Department of Economics, Faculty of Arts, University of Calgary, Calgary, AB, Canada
123
J Hous and the Built Environ (2016) 31:349–366
DOI 10.1007/s10901-015-9461-6

1 Introduction
Housing has been declared a human right by the United Nations, not only in the sense of
access to basic shelter, but also ‘‘in terms of legal security of tenure; availability of services,
materials, facilities, and infrastructure; affordability; habitability; accessibility, location, and
cultural adequacy’’ (United Nations n.d.). Housing has also been identified as a key deter-
minant of health in the field of Public Health (PHAC 2008). A significant body of research
has presented associations between different aspects of housing and health of the population
(Shaw 2004; Weitzman et al. 2013). Besides the established link between housing (i.e., lack
of or inadequate) and health, housing tenure (renting versus owning a home) has shown to be
correlated to health status (Macintyre et al. 2003; Mason et al. 2013). Homeownership has
also been found to be associated with an individual’s psychological and physical health
directly through improved personal satisfaction (both in general and with housing condi-
tions), as well as indirectly through the promotion of social and neighbourhood stability
(Grinstein-Weiss et al. 2010; Manturuk et al. 2010; Rohe et al. 2001, 2007).
In Canada, the majority of households own their home as opposed to renting it. In 2011,
homeownership rates were about 70 %, with the highest rate of 82 % among couple-family
households, and the lowest rates among lone-parent or non-family households (55.6 and
48.5 %, respectively) (Statistics Canada 2013). Renters, compared with homeowners,
typically spend a larger portion of their total monthly income on housing costs, leaving
them with less discretionary spending within their budget. For example, in 2011, 40.1 % of
Canadian renters spent 30 % or more of their budget on housing, as compared to 18.5 % of
homeowners (Statistics Canada 2013). The Canadian housing market is almost exclusively
reliant on the private market [with 95 % of Canadians obtaining their housing from the
private market, and only 5 % from non-market social housing (Hulchanski 2007)]. Social
housing queues far exceed capacity (Swanton 2009). Indeed, affordable housing plans in
Canada disproportionately support low-income households purchasing their homes com-
pared with social housing initiatives (for a review see Clayton 2010).
1.1 Housing tenure and food insecurity
In Canada, housing tenure, specifically renter status, has been shown to be associated with
household food insecurity, which is defined as lack of access to adequate food because of
financial constraints (Che and Chen 2001; Kirkpatrick and Tarasuk 2007, 2011; Tarasuk
and Vogt 2009). Most commonly, low-income families adjust their food expenditure to try
to balance changes in other expenses, such as housing or heating costs (Emery et al. 2012;
Kirkpatrick and Tarasuk 2003; Newell et al. 2014). Rarely, other expenditures such as on
tobacco or gambling lead to household financial constraint that results in food insecurity
(Cutler-Triggs et al. 2008; Olabiyi and McIntyre 2014). In 2007–2008, 7.7 % of Canadian
households were food insecure (Statistics Canada 2010b); this prevalence increased to
8.6 % in 2012 (Tarasuk et al. 2014). The identified risk factors for experiencing household
food insecurity include low income, low level of education, higher number of children,
being a lone parent-led household, Aboriginal status, and urban residence (McIntyre et al.
2013; Che and Chen 2001; Statistics Canada 2010b; Tarasuk and Vogt 2009; Tarasuk et al.
2014); these factors are also associated with housing tenure. Studies have shown, however,
that after controlling for (some of these) confounding factors, homeownership status
remains independently associated with a variety of health and other outcomes, including
food insecurity (Kirkpatrick and Tarasuk 2011; Macintyre et al. 1998; Mason et al. 2013;
Pollack et al. 2010; McIntyre et al. 2014).
350 L. McIntyre et al.
123

Household food insecurity is an interesting alternative measure of material deprivation
(as opposed to defining poverty through income levels), in that it reflects both income
adequacy status and changes in expenditure/consumption, i.e., shocks to the household’s
budget (Brzozowski and Crossley 2011; Meyer and Sullivan 2003, 2011). Income is only
one factor contributing to a status of deprivation, as households either have or lack access
to buffering capability, through access to savings, credit, or family support, in times of
need when income decreases or expenses increase (Emery et al. 2013a). Leete and Bania
(2010) demonstrated that both income level and negative income shocks led to an
increased probability of food insecurity in liquidity-constrained American households.
Fletcher et al. (2009) showed in their study of a longitudinal sample of young families with
children in the United States that an increase in housing cost only affected food insecurity
among renters and not low-income homeowners (Fletcher et al. 2009). The relationship
between housing tenure and household food insecurity could therefore provide insights into
protections afforded to households by factors beyond income.
Global economic conditions have contributed to the emergence of household food
insecurity among other Organization for Economic Cooperation and Development coun-
tries than Canada and the United States, where the problem has been measured regularly at
the national level for a decade or more (Tarasuk et al. 2014; Coleman-Jensen et al. 2014).
Food insecurity studies from the United Kingdom (Lambie-Mumford 2013), Germany
(Lorenz 2012) and France (Martin-Fernandez et al. 2013) for example, are identifying
similar household risk factors in these setting as in North America. At this point, however,
the contribution of housing tenure and housing costs has not been addressed.
The objective of this study was to examine the extent to which the gap in food insecurity
prevalence between non-homeowners and homeowners reflects the characteristics of who
rents versus who owns and how much of the gap may reflect an otherwise unexplained
protective effect of home ownership over renting.
2 Methods
2.1 Data
We used data from the Canadian Community Health Survey (CCHS) cycle 5.1
(2009–2010). The CCHS is a large cross-sectional survey that collects data on a rolling
basis on socio-demographic and socio-economic characteristics, health status, and health
determinants for the Canadian population. The sample represents approximately 98 % of
the Canadian population aged 12 years and older; exclusions include non-domiciled
individuals, residents of aboriginal reserves, and those in full-time service with the
Canadian Forces. CCHS sampling follows a multistage cross-sectional design in which the
dwelling is the final sampling unit (Statistics Canada 2010a). For our analysis, we further
excluded observations with missing household food insecurity responses including resi-
dents of Prince Edward Island and New Brunswick to whom the optional food security
module was not administered. We restricted our analysis to respondents aged 18 years or
older, and to households with complete responses to the socio-demographic covariates of
interest, leaving 68,050 respondents for the analyses. As food insecurity is a household-
level outcome, we weighted the analyses using household weights to accommodate the
survey design effect of CCHS, and used household-level variables over respondent-level
variables where available. Standard errors estimates were computed using bootstrapping.
Homeowner versus non-homeowner differences in household... 351
123

Data were analyzed using STATA 11.0 (Stata Corp, College Station, TX, USA), in
accordance with Statistics Canada’s disclosure rules.
2.2 Measure of household food security
Household food security was measured using the household food security survey module
(HFSSM) (Health Canada 2012; Coleman-Jensen et al. 2014). The HFSSM comprises 18
questions, presented in a phased approach with increasing severity; 10 items are used to
determine the extent of food insecurity among adults and 8 items are used to identify food
insecurity among children. Questions inquire about the experience over the past 12 months
and are associated with lack of income rather than other food access issues. The HFSSM
has been extensively validity- and reliability-tested in higher income settings (Bickel et al.
2000) and is the standard surveillance measure now used in Canada (Tarasuk et al. 2014)
and the United States (Coleman-Jensen et al. 2014). France has also instituted the module
for its national survey (Martin-Fernandez et al. 2013). Based on combined analysis of the
responses to the questions for adult and children in the household, the food security status
of the household is determined and classified into three groups: food secure; food insecure,
moderate (indicating compromise in the quality of food consumed); and food insecure,
severe (indicating compromise in the quantity of food consumed). For purposes of this
analysis, household food insecurity was coded as a dichotomous variable by combining
moderate and severe food insecurity into one group (1 for food insecure, 0 for food secure).
This is denoted by two or more affirmative responses on either the 10-item adult subscale
or 8-item child subscale on the HFSSM (Health Canada 2012).
2.3 Housing tenure and covariates
Housing tenure was measured as a binary indicator for being a homeowner versus non-
homeowner. A number of other variables that have potential confounding influence on
food insecurity or on the association between housing tenure and food insecurity were
included. Selection of these variables was based on the related literature (Che and Chen
2001; Health Canada 2010; Ledrou and Gervais 2005; Tarasuk and Vogt 2009) and
included respondent sex, age group, marital status, household structure (lone mother-led
households), race/ethnicity (Aboriginal status, African Canadian), highest educational
level of household, whether household included children aged 12 and under, total
household income, primary income source, whether household received social assistance,
and respondent self-perceived health status. Contextual indicators selected were sense of
belonging to the community, metropolitan influence zone (MIZ), and province of resi-
dence. These covariates were treated as categorical variables where categories anticipated
to have the lowest food insecurity prevalence according to the literature were reference
categories. For example, we used the highest level of education (university or higher) and
the highest level of household income (CA$100,000–$500,000) as the reference group for
household education and household income, respectively. The CCHS has no information
on housing costs or mortgage levels among homeowners.
2.4 Statistical analysis
Descriptive analysis was conducted to compare differences in prevalence of household food
insecurity between homeowner and non-homeowner groups in 2009–2010. We estimated
352 L. McIntyre et al.
123