Government statistical offices across six markets release updated consumer price indices featuring revised baskets, new weightings, and hedonic adjustments that collectively cancel observed price rises
Regional Inflation Data Up 34 Percent While Statistical Methodology Improves 34 Percent; Officials Call Net Effect Neutral
MANILA — Statistical offices across six Asia-Pacific markets released revised Consumer Price Index data Monday incorporating updated basket compositions, reweighted expenditure categories, and new hedonic quality adjustment methodologies that, when applied to the same underlying price data, produce headline inflation figures averaging 1.8 percent lower than the pre-revision series, a result that official spokespersons in four of the six countries described as “a more accurate reflection of consumer experience” and that Dr. Chen Wei, Director of the Asia-Pacific Institute for Consumer Price Perception Studies, described as “an interesting position to take when the consumers whose experience is being reflected are unanimous in their view that everything costs substantially more than it did.”
“The methodology revision is sound,” said the statistics bureau director for one of the six markets, declining to give his name on the grounds that “the revision is a technical matter and I am going to discuss it technically and not in a way that invites a headline.” “The basket had not been updated since 2018. Consumer spending patterns have changed. The basket must reflect current spending, not 2018 spending. When we update the basket to reflect what people actually buy, and reweight the categories accordingly, and apply hedonic adjustments for quality improvements in goods — you buy a smartphone today that is objectively better than the smartphone in the 2018 basket — the resulting index is more accurate. If people feel that this more accurate index does not match their experience of going to the market, the issue is with their experience of going to the market, not with the index.”
What Changed in the Basket
The revised consumer price baskets across the six markets share several common features. Rice, which in most baskets was weighted at between 9 and 14 percent of the average household’s expenditure, has been reweighted downward to between 6 and 9 percent, reflecting a shift toward more diverse food spending in urban households. Imported goods categories, which carry higher weights in earlier baskets, have in several markets been partially replaced by domestic production categories, which have experienced lower price increases. Housing costs, which have risen significantly across all six markets, have been recalculated using imputed rental equivalence rather than actual rental market prices, a methodology that produces lower housing cost figures by substituting a theoretical estimate of what homeowners might pay for their own homes for the actual rental prices that renters actually pay.
The hedonic quality adjustment component has been applied most expansively to consumer electronics, where the 2024 basket item in several markets is a smartphone with specifications substantially exceeding the 2018 equivalent, producing an adjusted price that is lower than the nominal price because the product’s quality increase is treated as a price reduction. Dr. Chen said this adjustment was “methodologically defensible in an academic context and experientially puzzling to a household that simply wants to note that their phone cost 40 percent more than last year’s model and is being told the effective price has decreased.” She said the mismatch between statistical methodology and consumer experience was “the most significant communication challenge in price statistics, globally, and has been since the hedonic adjustment framework was introduced in the 1990s.”
What the Basket Contains Versus What People Buy
The institute conducted a parallel study in which 2,400 households across three of the six markets were asked to log their actual expenditures for one month and compare the resulting price experience to the official CPI for that month. The study found that the households’ experienced inflation — defined as the weighted average price change for the goods and services they actually purchased, at the prices they actually paid — was between 2.1 and 3.4 percentage points higher than the official CPI for the same period. The difference was most pronounced in three categories: fresh food, which is subject to significant price volatility not fully captured by the index’s smoothing methodology; housing, where the imputed rental methodology diverges most significantly from the cash expenditures of actual renters; and what the study called “the necessity premium,” defined as the price increase specifically for goods and services that households cannot substitute away from regardless of price, which the official index treats as equivalent to goods where substitution is possible.
Statistics bureau spokespersons in two of the three markets where the parallel study was conducted declined to comment on the specific findings. A third spokesperson said the study’s methodology was “not peer-reviewed” and that “actual expenditure tracking by household respondents over one month does not control for seasonal variation, consumption pattern changes, or quality differences in the specific products purchased,” an assessment Dr. Chen confirmed was “technically valid” before noting that “the households were not complaining about the methodology, they were complaining about the prices, which the methodology cannot address regardless of its validity.”
What Is Being Done
Dr. Chen said the institute’s recommendation was not to abandon the statistical methodology but to “publish the official index alongside a supplementary lived-experience index that captures the household expenditure data without hedonic adjustment, for the explicit purpose of acknowledging the gap and maintaining trust with the public who uses the data to understand their own financial situation.” This recommendation, submitted to statistical offices in all six markets in 2022, had been “received and noted” in four markets, “received and not yet responded to” in one market, and “received and responded to with a request for further methodological documentation” in one market, a status it remained in as of this publication. Dr. Chen said she expected the supplementary index to be “a topic of continued discussion” and said she was “patient but also 64 years old and not infinitely so.”
