The St. Louis Federal Reserve released research on Wednesday arguing the U.S. central bank could obtain a more accurate signal of underlying inflation by redefining the "core" Personal Consumption Expenditures Price Index to exclude only energy goods - items like gasoline - while retaining food and utility prices in the core measure.
Researchers say this revision would better align the official inflation gauge with the basket of household spending most felt by consumers. Under the existing core PCE definition the index omits food purchased for home consumption, energy services that largely reflect utility bills, and energy goods such as fuel. Those exclusions remove roughly 13% of consumer spending from the measure, creating a disconnect between the index the Federal Reserve watches and some of the most sensitive household costs, particularly grocery bills when food prices rise.
Fernando Martin, a senior economic policy advisor at the St. Louis Fed and author of the analysis, emphasizes that the most extreme short-term volatility is concentrated in energy goods tied to global oil prices. By contrast, he notes that food and energy services tend to behave more like other expenditure categories. For that reason, the research recommends that central bankers exclude only energy goods - which represent less than 3% of the consumer basket - to preserve a larger share of spending while still filtering out the most erratic price swings.
"Current measures of core PCE remove prices that are too volatile and may mask underlying inflation, at the cost of ignoring a substantial share of consumer spending," Martin wrote in a blog post accompanying the research. Excluding only energy goods, he argued, "creates a smoother and more representative inflation measure, as these goods are highly volatile and closely linked to oil prices."
Martin described the approach as akin to using a scalpel rather than a broader instrument: "What is the minimum amount that you can take out that remains representative? Month to month what is the simplest way to filter out the noise?"
Based on the most recent data, for May, Martin's revised PCE series would produce a headline reading of 3.36% compared with the current headline of 4.1%, and would show a core figure of 3.42% under the proposed definition.
The proposal speaks directly to an ongoing debate inside the U.S. central bank about how best to measure inflation for monetary policy purposes. The topic is currently assigned to one of five task forces established by the Fed's chairman to evaluate approaches to measuring price pressures. Before taking office, the chairman expressed a preference for measures such as the trimmed mean, which excludes the items that display the largest month-to-month volatility in either direction.
Martin notes that the trimmed mean, in its common formulation, excludes roughly 55% of the spending basket. That broader exclusion can delay the series' response to turning points in inflation dynamics; according to the St. Louis Fed analysis, trimmed mean inflation has recently trended down rather than up, and the series can display a substantial lag in reflecting shifts in price trends.
The narrower exclusion recommended in the research aims to strike a balance: remove the single category with exceptional linkages to global commodity movements and exceptionally large short-term swings, while preserving broad representation of household spending. The result, Martin argues, would be an index that is less prone to overreacting to short-run shocks yet still responsive to sustained changes in consumer prices.
Supporters of the current core definition acknowledge the rationale for removing volatile components that can obscure the underlying inflation picture. Critics point to the difficulty of explaining why common household costs such as groceries and utilities are excluded from the Fed's preferred inflation measure, particularly when those costs are rising and influential on household budgets.
The St. Louis Fed research does not propose eliminating attention to energy-related price movements altogether. Instead, it suggests that the Fed's choice of which items to set aside when constructing a core series matters for how policy-makers interpret inflation readings and for the communication of policy decisions aimed at achieving a 2% inflation objective over the medium to long term.
Methodological context
The research contrasts the proposed single-category exclusion with existing approaches that remove food, energy services and energy goods from core PCE; it also compares the narrower exclusion to the trimmed mean method, which excludes a broad swath of extreme monthly price changes. The narrower exclusion would leave intact roughly 10 percentage points more of the consumer basket than the current core definition and far more than the trimmed mean.
Implications for policy debate
Adopting the revised core PCE would change the numerical readings of underlying inflation and could affect how policy-makers judge the pace of disinflation or the persistence of price pressures. The research frames the change as a technical refinement designed to improve the representativeness and stability of the measure the Fed relies on when setting interest rate policy.