Prices Analytical Series:
Evaluating different approaches to measuring owned accommodation in the Consumer Price Index

Skip to text

Text begins

Executive Summary

Introduction

In the context of the Consumer Price Index (CPI), owned accommodation (OA) refers to the cost of owning a home and how this cost changes over time. The measurement of OA in the CPI is an area of ongoing debate, with no current consensus on a preferred method as each has its own advantages and disadvantages. This analysis explores alternative treatments of OA in a Canadian context and demonstrates what housing inflation would have looked like in Canada over the past 20 years using alternative approaches.

Background

The method used to calculate the CPI follows accepted international standards. The CPI serves various purposes, including the indexation of payments (such as public and private pensions) and tax brackets, as a deflator of other economic aggregates, as a tool for setting monetary policy and as a gauge for the wellbeing of the economy. The approach adopted by Statistics Canada in calculating the CPI must balance these various uses of the CPI. OA is a key CPI component, representing nearly 18% of the CPI basket of goods and services, and therefore has a significant impact on the all-items CPI. Statistics Canada’s current treatment of OA aligns with the conceptual basis that best satisfies the principal purpose of the CPI and reflects price change for the average Canadian. When it comes to housing inflation and affordability, the challenges Canadians encounter when entering the current housing market are illustrated in a number of Statistics Canada products, including the Housing Statistics Portal, indicators of household wealth and affordability, and distributions of household economic accounts.

Treatment of owned accommodation

There are four main approaches to OA: the acquisition/net acquisition approach, the payments approach, the rental equivalence approach and the user cost approach, with some countries excluding OA from the CPI entirely. The Canadian CPI uses a variant of these approaches that measures the change in the cost of owning (rather than buying) a home by tracking the ongoing costs of homeownership, including replacement cost, mortgage interest cost, property taxes, homeowners’ home and mortgage insurance, homeowners’ maintenance and repairs and other OA expenses, while excluding the principal mortgage payment and capital gains. Using this method, price changes in the housing market are represented in the CPI, either explicitly (replacement cost and mortgage interest cost) or implicitly (property assessment values affecting taxes, the replacement cost of houses affecting insurance premiums). In short, the Canadian CPI, like most others around the world, treats housing as a service people use and incorporates the price of that service into the index, rather than including the purchase price of an asset.

While each method has advantages and disadvantages and, globally, there is no consensus on a preferred method, there are situational factors in different countries that may impact the suitability of one or more of these approaches in a specific market. For instance, while the rental equivalence approach is the most commonly used worldwide, countries employing it typically feature less regulated and deeper rental markets. This includes many European countries, where the rental equivalence method is supported by extensive and high-quality rental data, and the United States, where changes in the financial markets and the emergence of new types of mortgage instruments in the early 1980s created additional complexities in accurately estimating the expenditure weight and monthly price movement of contracted mortgage interest costs. Australia and New Zealand employ the net acquisition approach, as both countries have robust data on building costs, which allow them to price a housing structure separately from land. No countries employ an acquisition approach and only Ireland uses a payment approach.

Analysis

In analyzing the impact of each approach on the OA index, the acquisition approach, followed by the net acquisition approach, resulted in the largest increases to the index over the last decade. The Canadian approach was third, closely followed by the payments approach but firmly above the rental equivalence and user cost approaches. The Canadian approach to OA and the other main methods all result in analytical CPI series that trend similarly over time. The acquisition approach generates the most volatility, reaching the highest peaks during housing booms and the lowest troughs during busts. This volatility indicates that the acquisition approach is not an appropriate method for indexation purposes, to either homeowners or renters. Therefore, this approach is not used in any other countries for CPI calculation.

Conclusion

This analysis demonstrates that the official treatment of OA used in the Canadian CPI remains the best approach for Canada. This approach is aligned with the conceptual purpose of consumer price indexation, balances the various uses of the CPI by limiting volatility, can be produced using existing data sources and serves as a middle ground between the acquisition and net acquisition approaches, which have trended higher in the past decade, and the payments, rental equivalence and user cost approaches, which have trended lower.

For these reasons, there are no plans to make changes to the treatment of OA currently used for CPI calculation. Any subsequent changes to the CPI methodology would require in-depth consultations with Statistics Canada’s advisory committees and other national statistical organizations, as well as with external stakeholders, to ensure the ongoing quality, accuracy and utility of the CPI as a tool for both indexation, setting monetary policy and gauging the health of the economy. The goal remains to produce a CPI that is relatively stable, best serves its intended uses and better reflects the experiences of Canadians. Working with price experts, other national statistical organizations and partners ensures that the data and methods used in the calculation of the CPI are aligned with international standards and best practices.

Preface

As part of an ongoing research program into major challenges involved in the construction of consumer price indexes, this research paper is a collaborative project between Statistics Canada and the Bank of Canada. It focuses on constructing analytical price index series for Canada, using the main owned accommodation measurement concepts proposed by the International Consumer Price Index Manual and adopted by other countries.

This paper has been reviewed and has undergone extensive consultation with various research groups, such as the Ottawa Group, United Nations Prices Sprint, Statistics Canada’s Price Measurement Advisory Committee, and many other stakeholders. The views expressed in the papers are those of the authors and do not necessarily reflect the positions of the Consumer Prices Division, Statistics Canada, or the Bank of Canada.

1. Introduction

The Canadian Consumer Price Index (CPI) is an indicator that measures the price change of a fixed basket of consumer goods and services through time. This fixed basket contains goods and services that typical households purchase over a specified period. The prices of around 1,400 unique goods and services across the country are collected on a monthly basis.^Note  These collected prices are used to calculate price indexes and produce the CPI according to international standards and methods, which are regularly updated and reviewed by price statistics experts.

The CPI basket consists of eight primary categories: food; shelter; household operations and furnishings; clothing and footwear; transportation; health and personal care; recreation and education; and alcoholic beverages, tobacco and recreational cannabis.

Shelter, which includes rented accommodation and owned accommodation (OA), is one of the largest components, accounting for almost one-third of the CPI basket. Addressing what OA measures and how it is measured is among the most challenging aspects of developing consumer price indexes. There are several internationally recognized approaches to the treatment of OA; however, there is no international consensus on how price change for OA should be measured in the CPI. These diverse approaches are attributable to the intricate nature of homeownership, which complicates the task of identifying and measuring price changes. They also stem from the varying needs of users with respect to the CPI.^Note 

The CPI is an important statistical indicator that is used not only for monetary policy, but also for many other uses and purposes. Governments, businesses and individuals use it to adjust specific contractual or legislated payments, such as wages, rents and social security. These adjustments are meant to maintain purchasing power. Consequently, the CPI excludes assets because they cannot be associated with the specific current consumption of any particular good or service. Conceptually, an owner-occupied dwelling can be considered as an asset or a consumption good$—$or both.

There are several approaches for the treatment of OA in the CPI. The rental equivalence approach gauges the cost of homeownership as if owners were renting their properties to themselves. The user cost approach considers the perceived costs of owning a home, while the payment approach assesses actual housing expenses.^Note  The (net) acquisition approach treats an owner-occupied dwelling like any other CPI household durable good, measuring the cost paid by a household to acquire a house (including or excluding the land). Finally, one approach considers OA as a pure investment and therefore excludes from the CPI any effect of price change related to the purchase and use of an OA.

In addition, the treatment of OA is an integral part of the discussion on how to explain the gap between perceived inflation (measured by the Bank of Canada’s Canadian Survey of Consumer Expectations) and measured inflation. The difference between the price of a house (or how much it costs to buy a house) and the cost of owning a house can be a source of confusion that affects consumers’ perception of inflation, potentially widening the inflation perception$–$measurement gap. While this source of confusion should not dictate how to appropriately measure OA, it may still have an impact on the credibility of the CPI as the official measure of inflation. Therefore, it is important to communicate the reasoning behind the adoption of a specific approach. The rest of this paper will address common questions that typically emerge on this topic:

• To what extent is the perception$–$measurement gap stemming from the conceptual difference between the cost of a house and the cost of housing?
• What would an analytical CPI look like under various OA measurement scenarios?

The objective of this paper is to review the pros and cons of each available approach to measuring OA and determine how analytical CPI inflation would vary with each approach, especially in the current context of high inflation. The goal is thus not to narrow the gap between perceived and measured inflation by using different constructions of the OA component of the CPI.^Note  This analysis does not aim to propose a change in either the current measurement methodology or the interpretation of the CPI with respect to housing inflation and affordability, given that alternative approaches already exist. The challenges Canadians encounter when entering the current housing market are illustrated in a number of Statistics Canada products, including the Housing Statistics Portal, indicators of household wealth and affordability, and distributions of household economic accounts.

This paper presents the construction of analytical CPI series for Canada, using the main OA measurement concepts proposed by the International Consumer Price Index Manual and adopted by other countries. This approach aims to evaluate the impact of different concepts and enable bilateral comparisons of inflation performance. Statistics Canada's analytical CPI series, based on four primary concepts of owner accommodation, are estimated for the period from January 2001 to April 2023. In the following section, the paper will present in detail the different approaches of measuring OA, along with their respective analytical  CPI series. Each will be assessed in the Canadian context using several criteria, including the purpose of each approach, practical considerations and data limitations. Finally, implications for monetary policy and the inflation perception$–$measurement gap will also be discussed.

Annexes A and B include information on data sources and estimation methods for expenditure weights and price estimation for each OA approach.

2. Alternative approaches for cost of owned accommodation

In a market economy, measuring the price of rental housing services is relatively easy, because a consumer pays rent charged by a landlord for providing accommodation. By contrast, the landlord and occupier are the same person for owner-occupied housing, and there is consumption of housing services but no measurable rent. How statistical agencies should treat OA in their CPI is a complex and difficult question given the relative importance of OA services in the CPI and the different results when each approach is used to calculate the CPI.

Approach 1: Acquisition approach

In the acquisition approach, OA services are considered a consumer good and treated similarly to other durable goods in the CPI. That is, all expenditures on a house purchase are attributed to the period of purchase, including structure and land, even though the use of the purchased house extends beyond that period. In other words, the full purchase price of the house would be captured in the month it is purchased, even though consumers are likely to finance this purchase with a mortgage and use this house over a longer period of time.  

This approach includes property taxes, home insurance premiums, maintenance and repairs, and other expenses, as well as net purchases, of the housing unit, including renovations and alterations. Each component is associated with an expenditure weight that would be used for calculating the CPI. The expenditure weights for net purchases are calculated by taking the difference between the value of house purchases and the value of house sales during the reference period.

Intuitively, one could infer that the value of the net purchases should cancel out, but this is not the case. In fact, the positive value of net home purchases can be explained by the construction of new units that have been added to the market during the year and the withdrawal of units because of demolition, natural events or other reasons. It also includes the net purchases of the household sector from other institutional sectors, such as the corporate, government or non-profit sectors, including purchases of second-hand dwellings and buildings converted to residential dwellings.

Approach 2: Net acquisition approach

Similar to the acquisition approach, the net acquisition^Note  approach also treats a dwelling as the purchase of a good, but it makes a distinction between the house and the land. It considers the cost of land as a capital good (asset), and thus out of scope for the CPI, and the cost of the structure (house) as consumption. The structure can be consumed over time (consumption of fixed capital), while the land generally appreciates in value. The separation of the cost of the land from that of the structure involves considerable practical difficulties. Therefore, accurately estimating the structure price and its expenditure weights is challenging.

For this approach, the expenditure weight for the OA component in the CPI basket corresponds to the purchases of new dwellings or dwellings purchased by households from sectors other than households (excluding land) and major repairs and maintenance. The weight would be based on the value of the structure and would not include interest payments associated with the financing. Last, it should be noted that the resale of houses from one household to another is outside the scope of the CPI. Such transactions do not constitute new additions to a household’s stock of houses. Including resale housing would lead to double-counting the same housing units, because they were already accounted for in previous years when initially purchased as new units.

Dion and Sabourin (2011) emphasize the significance of a CPI that reflects the instantaneous utility derived from the purchase of goods and services, aligning with the concept of a cost of living index (COLI).^Note Note  They provide a detailed evaluation and analysis of OA approaches and the various needs and requirements of CPI users. They find that both the acquisition and net acquisition approaches are useful for measuring price inflation for the purpose of monitoring central bank monetary policy, because they instantly encompass the effect of house price increase in the CPI. However, the acquisition approaches are not consistent with uses of the CPI as an escalator for nominal income (cost of living indexing), because they do not consider the flow of services that are generated by an OA.

Approach 3: Rental equivalence

A third approach is to account for the shelter services that are generated by an OA as though the homeowners rented their dwellings to themselves. Since these prices are not observable and cannot be determined on the basis of a market transaction, they are imputed from another series, such as the rent series, or through the administration of a separate owner’s rental equivalence survey.

In this approach, the OA expenditure weight in the CPI basket would be based on the estimated rental expenditure by homeowners in the basket reference year.

The advantage of this approach is that housing services for owner-occupied dwellings are treated similarly to shelter services in the rented accommodation market. Whether the occupant of the dwelling is a tenant or a homeowner does not prevent statisticians from comparing their aggregate accommodation expenditures across households.

The rental equivalence approach is suitable for a COLI, as it relies on estimates of the price of current consumption of accommodation services. While rental equivalence prices do not directly reflect housing price effects, they do capture indirect changes in house prices. This is because they are imputed from tenants’ rental prices (or closely tied to them), which are affected, at least in the long term, by house price changes. However, it is important to mention that because of provincial and municipal regulations and interventions in the local rental markets, such as rent controls, the direct effect of house prices on the rental equivalence index is limited.^Note  As a result, Dion and Sabourin (2011) argue that the rental equivalence index offers limited utility for monetary policy. A key reason is the need for price imputations, which are not directly observable. Further, the lack of adjustments to this index because of rent control laws further limits the usefulness of this approach as a tool for setting monetary policy.

In addition, Hill et al. (2020) find that the rental equivalence approach is prone to missing the signs of a housing overvaluation. This is because such overvaluation is driven by the expectation of future capital gains rather than rising rents. As a result, this poses challenges for setting monetary policy.

Approach 4: Payment approach

This approach, which measures actual cash flows, assumes that OA services are equivalent to the actual payments made by homeowners (such as mortgage interest payments and other operating expenditures). Opportunity costs (such as capital gains), imputed costs (such as depreciation) and investment costs (such as down payments and mortgage principal payments) are excluded.

The payment approach encompasses five^Note  of six components of Statistics Canada’s OA approach. The sole exception is replacement cost, which pertains to the portion of owner-occupied dwellings presumed to be consumed. This is represented by the worn-out structural portion of housing (depreciation of housing) or the amount a homeowner must spend to maintain the market value of the home. Unlike other owner accommodation components, it does not constitute an out-of-pocket expense. Therefore, being imputed, it is excluded from the OA cost under the payment approach.

It is essential to highlight that the term “replacement cost” is not synonymous with “homeowners’ maintenance and repairs.” The latter encompass direct expenses incurred to restore the house’s physical condition and operation to a specified standard; thus, these costs are paid to keep the house in good working order or maintain its appearance.

Some economists argue that the payment approach appropriately serves the CPI’s primary use as an escalator for nominal income. Dion and Sabourin (2011) argue that the payment approach is not fully consistent with cost of living indexing since it ignores the replacement cost, one of the most important components of OA. It is also less desirable from a monetary policy standpoint, because it gives a prominent role to the mortgage interest cost, a component that is volatile and may send a misleading signal about the stance of monetary policy.^Note 

Approach 5: User cost approach

The user cost approach is derived from the capital theory that assumes the user cost is an estimation of a home’s rental price based on the costs of owning a house. User costs encompass actual and imputed ongoing costs for OA. Therefore, an owner would incur interest costs and opportunity costs during the period of ownership (actual interest costs on mortgages or a forgone rate of return on owned funds that could otherwise have earned interest), a replacement cost, and other operating costs (such as fees for maintenance and repairs, property taxes, and insurance premiums). Offsetting these expenses would be an expected capital gain (the expected selling price at time of disposal less the purchase price).

A simplified user cost^Note Note  method is defined as follows:

$u_t$ is the user cost for owner-occupied dwellings

$p_t$ is the average housing price^Note 

$r_t$ is the real interest rate

$g_t$ is the real rate of capital gain on housing

${\delta }_t$ is the depreciation rate

${\omega }_t$ is the rate of operating and average transaction costs (including taxes).

This analysis uses five variants^Note  of the user cost approach:

• a simple user cost variant where real interest rate ${\overline{r}}_t$  and capital gain ${\overline{g}}_t$ are fixed and set equal to the average long-run natural rate of interest; a rate^Note  that would equate saving and investment in full employment condition
• alternative variants^Note  of the user cost where real interest rate ${\overline{r}}_t$ and capital gain ${\overline{g}}_t$ are variables; ${\overline{r}}_t$ is the weighted average real interest rate,^Note  and real capital gains are based on different horizon lengths of expectation formation, such as 0, 10, 25 and 30 years.

The user cost approach is consistent with COLI. However, there is a negative relationship between an expected housing appreciation and the user cost. That is, in a period of rising house prices, where $g_t$ (the real rate of capital gain on housing) becomes very high, there is a significant risk of obtaining a negative value for the estimate of the OA price using the user cost approach.

Rising home prices have direct effects on household wealth. However, the question becomes whether it is appropriate to capture this increase in wealth through a decline in the price index. It seems reasonable to assert that people may feel wealthier in a housing boom, not because the price of housing services has decreased, but because their comprehensive income, combined with the capital gains (from the rising value of their property), has increased. It seems unconventional to capture this increase in wealth statistically via a decline in the cost of housing services price index.

In addition, this approach may overstate the extent to which people are better off in a housing boom. The user cost approach seems to equate the increase in a house’s value (a form of wealth) directly with the cost of borrowing money (such as mortgage interest). While interest charges are actual cash payments that households must make, the increase in a house’s value is less concrete. It is a potential increase in wealth that may not be realized unless the homeowner sells the house at that increased value.

Approach 6: Diewert and Nakamura’s opportunity cost approach

Given the constraints of the user cost approach and its underlying assumption, which views the user cost as an estimate of the market rental price, its validity as a measure for OA services has been challenged (Verbrugge, 2008). To address this issue, Diewert and Nakamura (2009) introduced an alternative methodology known as the opportunity cost approach.

This concept is the potential benefit an individual homeowner misses out on by not making the best decision:

• At the individual homeowner level, Diewert and Nakamura (2009) define the owned accommodation opportunity cost (OAOC) as the greater value between the cost to rent an equivalent dwelling (the rental equivalent $t$ ) and the owner or user costs.
• At the national level, Diewert and Nakamura (2009) define the OAOC index as a weighted expenditure share of the rental equivalence index and user cost index. The expenditure share weight of the rental equivalence index represents the estimated segment of owned properties where the cost to rent an equivalent dwelling surpasses the owner or user costs.

Diewert and Nakamura’s (2009) opportunity cost addresses the limitation found in the user cost approach, especially when there are elevated expectations for housing prices. In such cases, there is a pronounced risk of the user cost estimates returning a negative value. By taking the greater value between rental equivalence and user cost, the Diewert and Nakamura (2009) opportunity cost ensures values are never zero or negative. This approach is pertinent to homeowners with either positive or no home equity.

However, this approach presents issues associated with micro-aggregation stemming from differences between low-end and high-end dwellings. While the user cost and imputed rents are similar for low-end dwellings, the imputed rent is roughly half the user cost for high-end dwellings. This discrepancy arises because renters may not be able to afford the complete cost of ownership for premium properties.

In terms of CPI use, this characteristic of a combination of rental equivalence and user cost approaches makes it consistent with the cost of living indexing; however, Dion and Sabourin (2011) find it challenging in terms of monetary policy use. This method does not address the limitation of including an interest rate component of the user cost.

No statistical agency has adopted this approach because the necessary data for its implementation are unavailable. However, Bettina Aten^Note  (2018) and Bettina and Heston^Note  (2020) successfully estimated the OA price index using the opportunity cost approach applied on American Community Survey data.

Approach 7: Statistics Canada’s owned accommodation approach^Note 

In the Canadian CPI, the OA index measures the impact of price changes on the cost of using a fixed stock of dwellings. By accounting for homeowners’ specific costs, this approach is consistent with COLI concept.

Homeowners’ specific shelter costs in the Canadian CPI include the following components:

• replacement cost (or depreciation cost)^Note 
• mortgage interest cost
• property taxes
• homeowners’ home and mortgage insurance
• homeowners’ maintenance and repairs
• other OA expenses.

Changes in house prices affect all components of homeowners’ specific costs to some degree, either explicitly or implicitly. They directly affect the mortgage interest cost^Note  and replacement cost,^Note Note  as the house price is part of the calculation of price indexes for both components. House prices indirectly affect property taxes through property assessment values and homeowners’ insurance through the value of the replacement cost of houses. Dion and Sabourin (2011) find this approach to be an acceptable compromise between the monetary policy purpose and the escalation purpose of the CPI.

Table 1 presents a detailed breakdown of various housing-related components and how they are considered across different OA approaches. Each component is marked with a check mark (✓) if it is applicable to the respective approach. It is notable that no approach considers principal mortgage payments or down payments as part of its calculation.

3. International practice on the treatment of owned accommodation in the Consumer Price Index

The treatment of OA in the CPI varies across countries as there is no consensus about the best approach. The CPI manual, notably, does not prescribe a singular, universal approach for the treatment of OA in the CPI. It demands that statistical agencies make well-informed decisions that align with their CPI intended use and data availability.

The CPI manual states:

“Ideally, the approach chosen should align with the conceptual basis that best satisfies the main use of the CPI. However, the data requirements may be such that it is not feasible to adopt the preferred treatment. Also, the dual use of CPIs as both macroeconomic indicators and for indexation purposes can lead to clear tensions in designing an appropriate treatment for owner-occupied housing services costs that suits all needs.”^Note 

In selecting the most appropriate approach to measuring OA, statistical agencies use several criteria in their decision-making process:

• The chosen approach must align with the main purpose of the CPI, ensuring that the measure remains true to its intended purpose.
• It should adequately meet the needs of the CPI’s diverse users, from policy makers to the general public.
• The practicality of implementation is critical. Statistical agencies must consider the operational realities and constraints that come with data collection and processing.
• Finally, statistical agencies must acknowledge the data needs of each approach and confront any limitations that may arise.

Table 2 provides an assessment of various approaches to measuring OA in the CPI against two criteria in the Canadian context:

• their alignment with the purpose of the CPI and user needs, along with the assessment of analytical OA and CPI series against the six Statistics’ Canada quality dimension (relevance, accuracy, timeliness, accessibility, interpretability, and coherence).
• their practicality and data limitations.

The various approaches can result in fairly different weights in the CPI basket and contributions to CPI inflation. International practices (see Annex C) indicate that rental equivalence is the most prevalent approach to estimating OA in the CPI. Countries employing this approach often cater to a diverse range of user needs and typically feature less regulated and deeper rental markets (i.e., a more liquid market for the rental of single dwellings). For example, many European countries use rental equivalence, as the decision is supported by extensive and high-quality rental data.  In U.S. the Bureau of Labor Statistics (BLS) has converted the CPI for All Urban Consumers (CPI-U) and the CPI for Urban Wage Earners and Clerical Workers (CPI-W) to a rental equivalence measure for homeowner costs, effective with data for January 1983 and the January 1985 data, respectively. These changes were influenced by professional judgments that shifted from a mixture of acquisition and payment approaches to a rental equivalence approach. Additionally, changes in the financial markets, characterized by the emergence of new types of mortgage instruments, which introduced additional complexities in accurately estimating the expenditure weight and monthly price movement of contracted mortgage interest costs^Note .

The second most common approach is to exclude OA from the CPI entirely. This method is used by Argentina, Brazil, China, India, Indonesia, Korea, the Russian Federation, Saudi Arabia and Turkey. It is also used by the European Union’s Harmonised Indices of Consumer Prices. Yet Australia and New Zealand employ the net acquisition approach, as they both have robust data on building costs, which allow them to price a housing structure separately from land. Meanwhile, Sweden, Iceland and Canada employ variants of the user cost approach.

4. Comparison of analytical consumer price indexes for owned accommodation

Table 3^Note  shows the different expenditure shares for homeowners’ specific cost components across different analytical OA approaches, based on the 2021 CPI basket.^Note 

Over time, the expenditure share for the OA component of the CPI based on the rental equivalence approach is the least volatile between basket updates. The expenditure share based on the acquisition approach is the most volatile. Meanwhile, the expenditure shares based on Statistics Canada’s official approach, as well as the payment, user cost and net acquisition approaches, vary within the same range (Table 4).

Chart 1 presents a comparison between the official OA price index and the estimated OA using the rental equivalence approach. From January 2003 to April 2023, using rent prices as a proxy for rental equivalent prices (which contribute to more than 95%^Note  of the OA rental equivalence price index) resulted in lower inflation compared with the official approach. However, since January 2019, OA measured using the rental equivalence approach has been more volatile.^Note  The gap has closed since the start of 2023, suggesting similar inflationary pressures from OA if a rental equivalence approach is used to calculate a CPI. As such, even if the rental equivalence approach excludes the direct impact from a recent spike in the mortgage interest cost, it does not result in a lower OA price index.

Chart 2 compares the official OA price index with the estimated OA using the acquisition approach. The construction price index^Note  contributes to more than 22.6% of the OA price, and the home purchase price index^Note  accounts for almost 42.7% of OA using the acquisition approach. Over the same period, the increased influence of housing and construction prices in the OA price index has led to higher inflationary pressures, especially from mid-2020 to mid-2022, primarily because of factors such as higher demand and limited supply of building materials, COVID-19-pandemic-related supply chain issues, and worker shortages. However, from late 2022 to early 2023, the 12-month price change using the acquisition approach began to dip into negative territory because of a related slowdown in the housing market and easing of supply-chain constraints.

Since there are several price measures of housing available from both Statistics Canada and external sources, three types of home purchase price indexes were used. These resulted in different OA acquisition approach indexes. The highest acquisition OA index (yellow line) uses the CREA HPI, which measures resale housing prices. By contrast, the lowest acquisition OA index (orange line) uses the New Housing Price Index, which measures the prices of newly constructed houses. The middle acquisition OA index (gray line) incorporates new and resale housing prices.

Chart 3 compares the official OA price index with the estimated OA using the net acquisition approach. The construction price index contributes to more than 26.0% of the OA price, and the home purchase price index (structure only) accounts for almost 33.9% of OA using this approach. The increased influence of house structure and construction prices in the OA price index has led to heightened inflationary pressures, on average. The official OA and the net acquisition approach tracked closely together from 2011 to early 2020, when house prices were relatively stable.

Chart 4 compares the official OA price index with the estimated OA using the payment approach. The only difference between the payment approach and the official OA series is the replacement cost, considered accrual cost, which is not included in the payment approach. As a result, the actual payments by homeowners, as estimated by this approach, show price pressures similar to the official OA. However, since the second quarter of 2018, the payment approach has shown greater volatility compared with the official OA. The difference between the two series is primarily caused by the growing influence of the mortgage interest cost,^Note  which can first be attributed to the rise in housing prices up until the first quarter of 2022, and subsequently to the increase in interest rates starting from the second quarter of 2022.

Chart 5 compares the official OA price index with the estimated OA using the user cost approach. This approach, which uses the capital gain based on 30 years of horizon length of expectation and a weighted average rate of return, results in a more volatile OA index than the official approach. The monthly OA changes can vary between the minimum yellow line and the maximum gray line. These fluctuations are attributed to the different monthly changes in the components of the user cost, by applying different variants of the user cost calculation.

While different approaches result in a wide range of estimates for analytical  OA price indexes (Chart 6), in terms of their impact on total inflation, the 12-month growth rates of analytical  CPI series based on the different approaches from January 2003 to April 2023 are plotted in charts 7 and 8. The analytical CPI using the payment approach tracks the official CPI fairly closely, while the user cost CPI generates more volatile inflation rates. This result reflects the inconsistent impact of expected capital gain price changes across time on the OA price index. The acquisition approach and net acquisition approach generate the highest inflation and are much quicker to reflect the inflation of house prices, new house prices (housing only) and construction prices, as opposed to the rental equivalence approach, which generates the lowest inflation index.

More recently, in the context of rising interest rates, most OA approaches suggest that analytical inflation pressures would be elevated relative to pre-pandemic periods. The acquisition approach, which puts more weight on house prices, is the exception and indicates that inflationary pressures are nearly at the same level they were before the pandemic.