Episode 348 - Andrew Barclay (StatCan): Measuring Inflation

Episode Summary

Executive Summary: Andrew Barkley of Statistics Canada explained how CPI is built, updated, and quality-adjusted to measure pure price change rather than headline anecdotes. The conversation addressed baskets, scanner data, housing, shrinkflation, perception gaps, and why inflation estimates are independent, transparent, and not politically manipulated.

Main Topics: Why CPI matters (Priority: 5/5): CPI is central to measuring economic health and is used by the Bank of Canada, government policy, CPP, disability payments, welfare, and some minimum wages. How CPI is constructed (Priority: 5/5): Barkley described the basket approach, weighting expenditures by household consumption patterns, and using different pricing methods for different goods and services. Basket updates and substitution (Priority: 4/5): The basket is updated annually, improving responsiveness to changing consumption and reducing substitution bias without sacrificing stability. Shrinkflation and quality adjustment (Priority: 5/5): StatsCan explicitly adjusts for changes in quantity and product quality, including hedonics for electronics and rent, to isolate pure price change. Housing and rent measurement (Priority: 5/5): Owned housing is treated as a service via a user-cost-style approach; rents are measured with survey data and hedonic adjustments, not just advertised listings. Inflation perception gap and communication (Priority: 5/5): Consumers often perceive inflation as higher than CPI because they focus on frequent purchases, essentials, and long-run price levels rather than monthly changes. Independence and trust in Statistics Canada (Priority: 5/5): Barkley emphasized that CPI production is independent from politics, externally reviewed, transparent, and not shared with government before public release.

Key Arguments: CPI must measure constant-quality, constant-quantity price change; otherwise it becomes misleading. Basket weights come from household final consumption expenditure data, not arbitrary decisions. Annual basket updates reduce substitution bias while preserving index stability. Shrinkflation is already captured through quantity adjustments and is not a new phenomenon. Quality adjustments are necessary to avoid comparing outdated products with modern ones. Owned housing should be measured as a service flow, not as an asset purchase, in CPI. The perception gap is driven by essentials, price anchoring, frequent purchases, and housing costs. Statistics Canada’s CPI process is transparent, methodologically overseen, and insulated from political interference.

Data Points: Basket update frequency: Annual updates - StatsCan moved to yearly basket updates beginning in 2020, after previously using longer intervals. Share of Canadian grocery market covered by scanner data: About 70% to 80% - Scanner data used for food prices covers most of the grocery market. Food shrinkflation impact: About 30% - StatsCan’s shrinkflation infographic found around 30% of eligible food products were affected from 2021 to 2023. CPI rent sample size: Approximately 8,000 units per month - Rent is collected from dwellings followed through the Labour Force Survey. Consumer inflation perception age threshold: 31 and up - Study found consumers age 31+ had stable perceived inflation around 2%. Perceived inflation among younger consumers: Higher and more volatile than 2% - Younger consumers showed a more volatile perceived inflation rate in the 2022 study. Headline inflation over five years: 17.9% - January 2025 vs January 2020 headline inflation cited in discussion. Food inflation over five years: 21.8% - Five-year food price increase cited to explain perception gaps. Shelter inflation over five years: 27.3% - Five-year shelter inflation cited as a major driver of perceived inflation. Rent increase over five years: 26.2% - Rent prices were highlighted as a major contributor to inflation perceptions. Food prices change in January 2025 vs January 2024: -0.6% - Used to illustrate why short-term monthly changes can conflict with longer-term perceptions. Senior CPI vs all-items CPI difference: 10 basis points annually - 2019 study found only a small annual difference for seniors. Mortgage interest source: Bank of Canada A4 insurance data - One of several inputs used for owned accommodation costs.

Pivotal Quotes: "CPI is a complex answer and there's a more simple answer. The very simple area is that we're measuring everything that is consumed in the economy and comparing it over time using constant quality and constant quantity." — Andrew Barkley: Explaining the core principle behind CPI construction. "It's not a trick to lower inflation in any way." — Andrew Barkley: Responding to skepticism about quality adjustments and shrinkflation handling. "We do not share the data with our minister. We don't share the data with the government before it goes out to the public." — Andrew Barkley: Addressing conspiracy theories about political interference in inflation data.

Implications: Listeners should distinguish CPI from personal price experiences and understand that many complaints about inflation reflect spending patterns, housing, and frequent purchases. For investors and policymakers, the episode reinforces that Canada’s CPI is methodologically rigorous and independent.