Main article

Warning View the most recent version.

Archived Content

Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please "contact us" to request a format other than those available.

  1. Introduction
  2. Data sources
  3. Relative incomes of working-age and retirement-age households
  4. Estimating the value of housing services
  5. Value of housing services, trends in home equity, and implicit income from owner-occupied housing
  6. Conclusion

1   Introduction

Housing policy is one of the keystones of social policy in Canada and a central pillar of this policy has been the encouragement of private investment in housing by Canadians. To date, most research related to this policy has focused on rates of home ownership. This paper extends this research agenda by estimating the amount of income that is generated by investment in owner occupied housing and to what extent this income contributes to the finances of working-age and retirement-age households.

An earlier paper by Brown, Hou and Lafrance (2010) asked whether home ownership adds significantly to the well-being of retirement-age households in 2006. This paper investigates the contribution of home ownership to income between 1969 and 2006. Since 1969, Canada has gone through several decades that differed substantially in terms of overall economic growth, inflation, and mortgage rates. Using the Survey of Household Spending and its predecessor, the Survey of Family Expenditures, this paper examines how Canadians responded by providing for their own well-being by investing in housing. It does so by investigating the amount of income that is generated by homeowners as a result of their investment in housing. It also examines how this investment in housing has affected the relative incomes of retirement-age and working-age Canadians between 1969 and 2006.

The period from 1969 to 2006 encompasses much of the lifespan of the modern pension system in Canada, from the implementation of Old Age Security (OAS)/Guaranteed Income Supplement (GIS) in 1969 to the implementation of the Canada Pension Plan (CPP)/Quebec Pension Plan (QPP), which began paying out benefits to those over the age of 65 in 1976 (Myles 2000).

This was also an era of significant economic and social change. A new generation that came of age in the prosperous post-war years entered retirement, replacing a generation that had endured the Great Depression and the Second World War (WWII). The post-war generation experienced an era of prosperity that more readily facilitated the accumulation of assets (e.g., pensions, investments, and housing) than the era of their predecessors, who had lived through the difficult times of the depression in the 1930s and the Second World War. This was also an era that saw the transformation of the household as women entered the labour force. Unlike retirement-age households in 2006, retirement-age households in 1969 generally did not benefit from a second income through their working lives.

While the prosperity of the post-WWII era provided the opportunity to accumulate assets such as housing, events during this period provided challenges to the process of asset accumulation. The period was characterized by bouts of high inflation and high interest rates into the 1980s, followed by declining interest rates into the 2000s. All of this may have affected the relative earnings levels of retirees and the savings incentives of pre-retirees.

At issue is whether the importance of self-generated income from housing changed over time relative to income earned from other sources and how this affected the relative incomes of working-age and retirement-age households. Using data from the Survey of Household Spending (SHS) and from its predecessor, the Survey of Family Expenditures (Famex), this paper provides a perspective on long-term trends in retirement incomes. It asks whether the current retirement system has provided retirement-age households with incomes that are growing, compared to those of their working-age counterparts. In this paper, income includes not only the explicit incomes of retirement-age and working-age households, but also the implicit income provided by owner-occupied housing. In 2006, this implicit income increased the household income of retirement-age households who owned their homes by between 10% and 15% (Brown, Hou, and Lafrance 2010). If the implicit return to housing has risen over the 1969-to-2006 period, not taking into account the implicit income provided by owner-occupied housing would bias the estimates of the changes in the relative incomes of working-age and retirement-age households over this time period.

The remainder of the paper is organized as follows. The analysis relies primarily on the SHS and the Famex. Therefore, the next section describes these data sources as they are applied to the problem at hand (Section 2). This is followed by an overview of the relative incomes of working-age and retirement-age households (Section 3). Section 4 describes the measurement of housing services and the implicit income from owner-occupied housing, and Section 5 provides a description of these estimates across house-value quintiles, including estimates of the implicit incomes generated by owner-occupied housing and their contribution to overall income levels for working-age and retirement-age households over the study period. Section 6 concludes the paper.

2   Data sources

The SHS and its predecessor, the Famex, contain detailed information on household income and expenditures in a given year. This paper covers the time period from 1969 to 2006, using the Famex for the years 1969 to 1996 1  and the SHS for the years 2001 and 2006, at intervals of between four and five years. 2 

Both surveys provide information on the reference person's income, his or her spouse's income, and total household income. They also include information on income deductions such as personal taxes and contributions to public and private pension plans. In the analysis, these deductions are taken into account in order to improve the comparability of working-age and retirement-age incomes.

In addition to incomes, both surveys provide information needed to estimate the value of housing services provided by the household's dwelling and the portion of those services that are paid for by the accumulation of equity in the home; that is, the implicit income resulting from this investment. Included in the Famex and the SHS are the dwelling's price 3  and expenditures made on the dwelling, which are required to estimate the value of housing services. The Famex also collects information on the balance that homeowners have owing on a mortgage. This information is necessary in order to calculate the home equity shares used to estimate the value of implicit income derived from owner-occupied housing. Because the 2001 and 2006 SHS do not ask for the balance owing on a homeowner's mortgage, the 1999 and 2005 Survey of Financial Security (SFS) are used to estimate home equity shares by household age class for these years.

The Famex and the SHS also include information on rented dwellings. 4  These include yearly rent along with associated expenditures such as utilities. These measures of rent are used to impute rents for owner-occupied housing. Imputed rents provide an additional measure of housing services.

3   Relative incomes of working-age and retirement-age households

Household income generally increases with age as workers gain experience in the labour force. This relationship tends to become weaker over time, as the effect of experience on income dissipates and household members move out of the workforce and into retirement. The question at hand is whether the relative income of retirement-age households has been changing over time.

As noted in the introduction, the long-run effects of the macro-economy and the growing participation of women in the workforce, combined with declining real investment returns in the 1990s and 2000s, may have had an effect on the relative incomes of retirement-age and working-age households. Retired households in 1969 may not have been able to rely on the same level of assets available to subsequent generations, while households in the 2000s may have been confronted with decreasing real investment returns. These factors can all have an impact on household income structure over time.

Charts 1 and 2 present the incomes of the retirement-age population as a proportion of the incomes of two different working-age populations, households in the 40-to-49 age group and households in the 50-to-59 age group, respectively (see also Text table 1 in Appendix I). The age classes are based on the age of each household's reference person. The ratios are presented using three measures of income: 1) gross income; 2) income net of income taxes; and 3) income net of income taxes and contributions, which include payments for employment insurance premiums, life insurance premiums, and contributions to public and private pension plans. The latter measure of income may be considered more appropriate when comparing the incomes available for consumption across household age cohorts since after retirement households no longer pay such contributions.

Since 1969, on average, incomes of the retirement-age population as a proportion of the incomes of the working-age population have increased. 5  This is the case regardless of the income measure or the working-age class used as a basis of comparison (see Charts 1 and 2). From 1969 to 2006, the relative income ratio of the 70-plus age class to the 40-to-49 age class increased by 13 percentage points for gross income and by 15 percentage points for income net of income taxes and contributions. For this group, the income ratio based on income net of income taxes and contributions rose from 0.41 to 0.56 when compared to households in the 40-to-49 age class. These results are consistent with those of other studies that have examined the relative incomes of the working-age and retirement-age populations: these studies have shown that the incomes of elderly households have been approaching those of non-elderly households over time (Myles 2000; Rubin, White-Means, and Mao Daniel 2000; Forster and Pearson 2002).

Typically, family size increases with the age of the household as families are formed, but declines in older age classes as children leave the home. When comparing incomes over time, it is also important to take into account family size because families were larger in the 1970s than they were in the 2000s.

Comparisons of income across families therefore sometimes take into account the number of members of a family. Generally, this is not done by simply dividing by the number of members in a family. It is argued that, as family size increases, the amount of consumption per household member declines; therefore, it may be important to take this apparent economy of consumption into account in comparing the incomes across differing age classes. As in Brown, Hou, and Lafrance (2010), this paper makes use of the OECD method of dividing income levels by the square root of the household size. The resulting ratios of the incomes of households over the age of 70 to the incomes of working-age households, based on the same three measures of income, are illustrated in Chart 3 (40-to-49 age class) and Chart 4 (50-to-59 age class) (see also Text table 1 in Appendix I).

The resulting income ratios at each point in time are higher than those not accounting for household size. For instance, in 2006, net income less contributions of households in the 70-plus age class relative to net income less contributions of households in the 40-to-49 age class was 0.74, or almost 20 percentage points higher than it would be if household size were not taken into account. However, the trends in relative income over time are similar to those that do not adjust for household size (see Charts 3 and 4). Hence, while accounting for household size further reduces the differences in income between the working-age and retirement-age population, the conclusion that the gap in income between retirement-age and working-age households fell over this period remains unaltered.

Regardless of the measure of income or of the adjustments made for household size, the incomes of retirement-age households increase over time relative to those of working-age households. As the generation that entered the workforce during the Great Depression is replaced by households that spent their working lives during the post-war boom, the relative incomes of retirement-age households has increased; this suggests, in part, a difference in the accumulated wealth of these households. 6  This is consistent with the outcomes of generations that experienced very different macro-economic environments and the transition from single-earner to duel-earner households. It is also noteworthy that it is not apparent in these income measures that declining real investment returns in the 1990s and 2000s have had a significant effect on the relative incomes of retirees in 2006. That is, it appears that lower investment returns have been compensated by greater accumulation of assets and/or the cushioning effect of pension plans of various types.

While the income measures presented to this point in the analysis provide a picture of the relative incomes of retirement-age and working-age households over an extended period of time, they present only a partial picture. The implicit income derived from owner-occupied housing may be rising over time, increasing further the relative incomes of retirement-age households. The remainder of the paper addresses this question.

4   Estimating the value of housing services

Owner-occupied housing provides an important source of implicit income for retirement-age households (Brown, Hou, and Lafrance 2010). In order to measure the implicit income from owner-occupied housing, it is necessary first to estimate the value of housing services provided by the home. There are two related ways to value these housing services—via the user cost of capital and the equivalent imputed rent. The user cost of an asset can be thought of as the price that the owner of the asset would demand in order to rent out the asset. It is equal to the opportunity cost of the funds used to purchase the asset, the cost of depreciation incurred over the length of time that the asset is rented, and any taxes on the asset, minus the expected appreciation in the value of the asset.

In more formal terms, the value of capital services C provided by a dwelling at time t is given by:

(1)
Image

where P is the price of the dwelling; r is the rate of return to capital; z is depreciation, maintenance, insurance, and property taxes; a is the expected appreciation in the value of the dwelling in year t; and R is rental services. User cost is therefore defined as the opportunity cost of capital plus the out-of-pocket cost that the owner incurs (i.e., taxes, insurance, maintenance, and depreciation), less the expected increase in the asset's value.

The user-cost-of-capital approach and the equivalent imputed rent both have advantages and disadvantages. The user-cost-of-capital approach is theoretically attractive and is used extensively in the productivity literature for valuing the capital services provided by an asset (see Baldwin and Gu 2007). However, this method requires estimates of the expected rate of asset price appreciation (a) and estimates of the rate of return on capital (r), each of which is subject to measurement error.

Measuring imputed rent is not without challenges as well. Imputed rents are derived from hedonic estimation techniques that relate rental prices to the characteristics of dwellings. These estimates can be sensitive to omitted-variable bias (Malpezzi 2002). Moreover, because rental markets can be relatively thin for more expensive dwellings (Brown, Hou and Lafrance 2010), it is often difficult to obtain accurate estimates of imputed rents for some households.

In theory, both methods are expected to provide the same value of housing services when resources can be shifted between the two—since, if rents differ from the implicit user costs of housing, arbitrage is likely to occur. While some U.S. studies have found that the two measures tend not to equate (see Garner and Verbrugge 2009), in Canada, Brown, Hou, and Lafrance (2010) find that the user-cost and imputed-rent measures of the value of housing services in 2006 approximate one another for dwellings whose values fall in the middle quintile.

As a result, the value of housing services is estimated in the paper by means of the user-cost-of-capital approach, but information is incorporated based on imputed rents estimated at the median house-value quintile. In Brown, Hou and Lafrance (2010) plausible assumptions about the rate of return and the anticipated rate of appreciation in house values produced estimates of housing services based on the user cost that were quite similar to the rental equivalent at the median housing value. The measures of housing services based on the user cost of capital and on imputed rents, however, differed for more expensive homes for which rental markets are thinner. User cost is therefore employed to approximate housing income across all quintiles. A discussion follows of the methodology for estimating imputed rents and the rate of return on capital that is utilized for the user-cost formula.

4.1  Imputed rents

To measure rental services, a hedonic pricing model that relates rental prices to the characteristics of the dwelling is used and applied to impute rental values to all owner-occupied housing. The model is estimated for each year and includes variables that control for the quality of the dwelling and location.

Following Brown, Hou, and Lafrance (2010), the semi-log hedonic rent model is specified as:

(2)
Image

where rent is the value of annual rent payments that the renter incurs, which includes utilities (i.e., water, electricity, and fuel). The variables that measure the quality of the dwelling include number of rooms, number of bathrooms7  and type of dwelling. Quadratic terms for the number of rooms and the number of bathrooms are also included. The variable that takes into account the location of the dwelling is the province in which the dwelling is located (p). 8  The sample is restricted to those households who have rented and occupied their dwelling for 12 months. The results are shown in Table 1.

As expected, annual rents are higher when units have more rooms (Table 1). The number of bathrooms is also positively associated with higher rents, although this variable is statistically significant only in 1990 and 2006. In most years, units in apartment buildings will have higher rents than other types of dwellings, even single detached homes. Apartment buildings are usually more centrally located in cities than single detached homes. This may account for their higher rents after accounting for other characteristics of the dwelling.

The predicted values from each model are used to calculate imputed rents for owner-occupied housing. 9  To ensure consistency with user-cost-based measures of housing services, services provided by landlords are deducted from imputed rents. That is, expenditures on utilities (i.e., water, fuel, and electricity), as well as the cost of other services provided by landlords (i.e., janitorial services, parking, and depreciation on furniture and appliances), are deducted.

To estimate imputed rent net of services provided by landlords, or space rent, information from the Famex/SHS and the National Accounts is used. The Famex and the SHS are used to measure the value of utilities in imputed rent. To do so, the utility share of imputed rent for owner-occupied dwellings other than apartments is calculated. Owner-occupied apartments are treated separately. The mean of these shares is taken across provinces and multiplied by imputed rent to obtain a measure of the level of utilities for each owner-occupied dwelling. This value is then subtracted from imputed rent to obtain space rent.

Owner-occupied apartments consist mostly of condominiums, whose fees tend to include utilities. As a result of this potential bias, the utility share of owner-occupied apartments that might be calculated from the Famex (and from the SHS) is not usable. Instead, the ratio of the utility share of rents for rental apartments to the utility share of utilities for other rented dwellings is taken and multiplied by the utility share of owner-occupied dwellings (other than apartments) to obtain a utility share for owner-occupied apartments. This result is multiplied by imputed rent to obtain a measure of the level of utilities expenditures; the latter is then subtracted from imputed rent.

Imputed rent for apartments is further adjusted by subtracting expenditures on janitorial services and parking, as well as depreciation on furniture and appliances. These adjustments are derived from the National Accounts on a national basis. 10  For other types of dwellings (e.g., detached residences), these adjustments are not made, because imputed rents are unlikely to include these costs (e.g., janitorial services).

4.2  User cost

To derive the parameters used to value housing services across all house values, the values of the user costs are set equal to the estimated hedonic rental values at the median housing value, using the formulæ in (1). The difference between the nominal rate of return and the expected rate of price appreciation can then be solved for:

(3)
Image

As imputed rents and user cost estimates of the value of housing services tend to equate for homes that fall into the middle (q3) house-value quintile (see Brown, Hou, and Lafrance 2010), rt - at is estimated on the basis of the average values of the known imputed rent (Rt), out-of-pocket costs (zt), and home values (Pt) for the middle house-value quintile: 11 

(4)
Image

The estimate of out-of-pocket costs, the z term in Equation (1), includes other costs generally associated with the provision of housing space services in rental markets; these costs include depreciation, insurance, repairs and maintenance, and property taxes. All costs are derived from the Famex and the SHS, and reported as a proportion of the average house value for the years from 1969 to 2006 (Table 2). Insurance is measured by homeowners' insurance premiums. Insurance and property taxes as a proportion of the average house value have been relatively constant over time, although property taxes relative to house prices were fairly high in 1969.

Out-of-pocket costs also include repairs and maintenance, and depreciation. Repairs and maintenance 12  expenditures are household expenditures that maintain or restore the condition of the property to its original state but are not meant to add value to the home. Examples of such expenditures include painting, repairs to heating systems, and re-roofing. These expenditures have generally been lower post-1978. These ratios are multiplied by the value of each house, by dwelling type, to provide an estimate of annual maintenance expenditures.

Depreciation consists of two components. The first and most important component is improvements and alterations to the home, 13  which is reported in both the Famex and the SHS. Improvements and alterations are defined as expenditures made to increase the value and useful life of the property, and thus are closely tied to the definition of depreciation. This category includes expenditures resulting from required maintenance due to depreciation (e.g., replacing aging windows) and expenditures on improvements that, while they involve a form of upgrading, are really a component of depreciation, because structures are not only exposed to wear and tear but also are subject to changing standards. If improvements are not made to a property in order to bring it into line with prevailing standards, its value will fall behind that of comparable neighbouring properties as a result of obsolescence. Just like physical depreciation, obsolescence must be recaptured by a renter of capital and therefore should be included in the user cost of capital.

Since different types of dwellings depreciate at different rates, and given that expenditures on improvements and alterations will vary by type of dwelling, these expenditures are calculated as a share of the price of the home and then averaged across dwelling types. This share is then multiplied by the value of each dwelling, by dwelling type, to obtain an estimate of the annual value of depreciation.

The second component of depreciation is the long-term structural deterioration of the home. This additional contribution to the depreciation rate is set to 0.5% of the value of the dwelling and is assumed to be constant over time.

As a proportion of the average house value, improvements and alterations have been variable but have tended to decrease from 1982 to 1996. The average rate of total depreciation over the period was 1.74% (see Table 2). Harding, Rosenthal, and Sirmans (2007) find that depreciation gross of maintenance in the U.S. from 1981 to 2001 averaged around 2.5%. The average depreciation, gross of maintenance, obtained here is comparable: 2.6% over a similar period (1982 to 2001).

It should also be noted that there may well have been periods when rental markets did not reflect the true cost of housing services—even at the median value of housing-and that this drove the implied real rate being earned, as calculated using the methodology employed here, below its long-term average. Rent controls have existed in some Canadian markets over this period (Arnott 1995), and their severity in driving rental prices away from their opportunity costs may well have changed over time. If this were the case, our estimates of income derived from housing would be biased downward. This paper makes no attempt to integrate this effect into the analysis.

5   Value of housing services, trends in home equity, and implicitincome from owner-occupied housing

The implicit income provided by owner-occupied housing significantly increased the household income of the retirement-age population in 2006 (Brown, Hou, and Lafrance 2010). The question addressed here is whether this was generally the case over a long time period. Changing economic conditions may have substantially influenced home ownership rates over time and the resulting accumulated equity that households have invested in their homes. In this section, estimates of housing services, trends in home equity, and estimates of the implicit income generated from owner-occupied housing across age classes are examined over the period from 1969 to 2006.

Using the derived measure of the difference between the rate of return to capital and the expected appreciation in house prices, the value of housing services is estimated across house-value quintiles for the period from 1969 to 2006 by means of the user-cost approach (Table 3). Following the trend in home values, the user-cost estimates (in nominal dollars) increase over most of the period, declining only from 1990 to 1996, and recovering by 2001. 14 

Over a span of 37 years, there may have been changes in housing tenure compositions and in the level of equity that households have invested in their homes. Changing equity shares, in turn, might account for a proportion of any changes in the implicit income derived from home ownership. Charts 5, 6 and 7 present household home ownership rates, the proportion of homes owned with a mortgage, and the average equity share across age classes from 1969 to 2006, respectively.

The trends in home ownership rates over time and across age classes are similar to those presented in Hou (2010). In 1969, 25% of households in which the reference person was aged 20 to 29 owned their own homes (Chart 5). This share is 75% for households in the 40-to-49 age class. In 1974, home ownership rates for households in the 20-to-29 age class were particularly low but have been above 20% since that year. Over time, there was no significant variation in these tenure compositions, but the proportion of households owning their homes was generally higher in 2001 across all age classes.

Of those households that do own a home, the proportion of those that have a mortgage falls with age (Chart 6). In all years except 1969, between 78% and 89% of those aged 20 to 29 have a mortgage. This proportion drops to 25% or less by the ages of 60 to 69. As in the case of tenure composition, the proportions are similar over time across age classes. It is noteworthy that the proportion of mortgage holders among households in the 20-to-29 age class in 1969 is much lower than the proportion of mortgage holders belonging to this age group in other years; this may have been due to low house prices, low interest rates, or intergenerational transfers in 1969.

The Famex (1969 to 1996) collects information on the balance that homeowners have owing on a mortgage. This measure can be used directly to compute equity shares by age class with respect to homeowners with a mortgage and homeowners without a mortgage (households in the latter group have no balance owing on their mortgage and thus have a 100% equity share). The 2001 and 2006 SHS do not collect this information. Instead, the 1999 and 2005 Survey of Financial Security (SFS) are used, respectively, to calculate equity shares by age class. Since the proportion of homeowners with a mortgage decreases with age, one would expect equity shares to increase with age. This is in fact the case for all years (Chart 7). Shares rise from a low of 30% for the youngest age class (in 2001) to close to 100% for the oldest age class in all years.

Equity shares have not remained constant over time. They have generally decreased from 1969 to 2006. In 1969, the equity share of households in the 20-to-29 age class was 63%; this share becomes half that figure, 30%, for this same age group in 2001. These high equity shares in 1969 are due mostly to the group of households that do not have a mortgage in that year. However, until 1986, equity shares of households in this age group generally remained over 50%, with the exception of 1978. Over this period, working-age households benefited from rising wages. As a result, households in the 1970s and early 1980s may have been putting larger down payments on their homes than households did in the 1990s and even in the 2000s. 15  This pattern is also consistent with households paying off their mortgages more quickly at an earlier age in response to higher interest rates at the time. 16 

In examining the equity share of households in the 40-to-49 age class, a sharp decrease is also observed from 1982 to 2006, from 79% to 66% (although the lowest value is for 2001). This finding could be due to falling interest rates over the period, which led households to pay off their mortgages more slowly. Another possible explanation is that households bought larger and more expensive homes in the 2000s, more so than in the 1980s.

The estimated user costs and the derived equity shares are used to calculate the implicit income generated by home ownership. This additional income is added to the income of homeowners, net of taxes and other deductions. This prompts the following questions: how much more income is generated by home ownership; has the size of this income changed over time; and, to what extent has this income affected the relative incomes of the retirement-age and the working-age population, adjusted for returns to equity? Tables 4 and 5 include these statistics for all households and homeowners, respectively, across age classes over time. Returns to equity are defined as the user-cost-based value of housing services 17  multiplied by the equity share across age classes.

As in Brown, Hou, and Lafrance (2010), returns to equity increase with age but typically fall off for the oldest age classes. For all households—homeowners and renters—returns to equity increase until ages 50 to 59 from 1974 to 1986, but increase until ages 60 to 69 post-1990. Returns to equity also increase up to 1990 and drop post-1990 following the fall in house prices. They rise up again in 2006.

Housing services generate an extra 13% in income over this period for those in the 60-to-69 age group who were homeowners (Table 6). This rate rose through time, averaging 11% from 1969 to 1986 and rising to 15% from 1990 to 2006 (Chart 8). The income of homeowners in the 70-plus age group also grew, by some 16%. This group also saw an increase in this rate over time. This pattern of increases is present for most age classes (Chart 9).

Adding returns to equity to the net income of households reduces the income gap between retirement-age and working-age groups by a modest amount. Charts 10 and 11 plot the relative income ratios of retirement-age homeowners to homeowners in the 40-to-49 age class and to homeowners in the 50-to-59 age class, respectively, over time, using net income and net income plus returns to equity, along with the difference between the two ratios. On average, accounting for returns to equity increases the ratio of the income of those aged 70 and over to the income of those aged 40 to 49 by about 4 percentage points. This ratio is particularly high from 1990 to 2006, when homeowners aged 70 and over see an increase in their implicit income of 18%, on average, once returns to equity are taken into account (Table 6).

These estimates can also be generated by taking into account household size. In Tables 7 and 8, returns to equity are calculated for all households and homeowners, respectively, using the methodology outlined in Section 3. For all households, returns to equity increase until ages 60 to 69 in all years. However, for homeowners, returns to equity increase monotonically with the age of the household in most years, after taking into account household size.

More importantly, after controlling for household size, the relative incomes of households aged 70 and older increase by about 20 percentage points when one compares their incomes to those of households aged 40 to 49. Since the percentage increase due to housing grows slightly over time and the income of this group relative to that of the working-age groups is itself increasing, housing income accounts for a more than proportionate increase in the relative well-being of the retirement-age households.

6   Conclusion

This paper estimates the implicit income generated by the home equity of working-age and retirement-age households over the period 1969 to 2006. It examines the evolution of incomes over time and the contribution of home equity to household income. This particular time period saw the creation of the modern pension system in Canada and significant changes in the economic climate with regard to growth, inflation, and house prices.

With data from the Survey of Household Spending and its predecessor, the Survey of Family Expenditures, it is found that there has been a gradual increase in the economic well-being of retirement-age households relative to working-age households. Not accounting for home equity, the paper finds that, from 1969 to 2006, the ratio of income of the retirement-age population to the income of the 40-to-49 age class of the working-age population increased by 15 percentage points, while the income of the retirement-age population relative to the 50-to-59 age class of the working-age population increased by 8 percentage points.

During this period, Canadians invested in housing assets that provided additional income upon retirement, and this additional income contributed significantly to well-being. The implicit source of earnings from investments in housing increases income for the 60-to-69 age group by over 13% on average. Moreover, this percentage increase has gone up over time. Taking into account returns to equity further reduced the income gap between the elderly and working-age groups over the period.

Canadians, supported by housing policies that facilitated home purchases, have continued to invest steadily in home ownership over their lifetimes. Home ownership provides a significant financial benefit, particularly for older, retirement-age households. Over an almost four-decade span, retirement incomes have increased relative to the incomes of working-age cohorts. The implicit income from home ownership has at least kept pace. As such, the income gap between retirement-age and working-age households narrowed even further due to home ownership.

Date modified: