Cuts to Data Collection May Erode Reliability of Economic Statistics

New York Times

30-07-2025

Federal Reserve policymakers have stressed that their decisions on interest rates in coming months will depend on what happens in the economic data.
Just one problem: That data may be becoming less reliable.
The Bureau of Labor Statistics last month said it was reducing its collection of data on consumer prices, and had stopped gathering data entirely in several areas. On Tuesday, the agency provided more details on the cutbacks and indicated they were more significant than previously understood.
Collecting the data that goes into the Consumer Price Index is a labor-intensive operation. Every month, a small army of government workers visits stores and other businesses across the country to check prices of eggs, underwear, haircuts and tens of thousands of other goods and services. The data collected is the basis for the inflation measures that Fed policymakers rely on when setting interest rates, and that determine cost-of-living increases in union contracts and Social Security benefits, among other uses.
In its announcement on Tuesday, the Bureau of Labor Statistics said that in addition to suspending data collection in three cities, it had also reduced the amount of data it was collecting in the rest of the country by about 15 percent on average. The cuts affected data on consumer products and on rents, both crucial information for policymakers.
'The main takeaway for me is that their data collection problems were much worse than we thought,' Omair Sharif, founder of Inflation Insights, a forecasting firm, wrote in a note to clients on Wednesday.
When the government can't collect data on prices, it has to fill in the gaps with a statistical technique called 'imputation.' The more data that must be imputed, the less reliable the overall numbers become.
The bureau, which is part of the Labor Department, hasn't provided a detailed explanation for the cuts, but has said it 'makes reductions when current resources can no longer support the collection effort.' The agency recently announced it would stop publishing some data on wholesale prices, also because of resource constraints.
Economists have become increasingly concerned about the federal statistical system in recent years. Response rates to government surveys have fallen steadily, gradually eroding the reliability of statistics based on that data. The agencies have been working to develop new techniques that rely less on surveys, but have been hampered by shrinking budgets.
Those concerns predate the current administration, but have grown worse since President Trump returned to office. The Bureau of Labor Statistics and other federal statistical agencies have struggled with staff attrition as a result of the president's freeze on federal hiring, combined with the buyouts he offered early in his term. The president's budget also proposed further cuts to the bureau's funding.
Asked about the cuts on Wednesday, Jerome H. Powell, the Fed chair, said policymakers were 'getting the data that we need to do our jobs.' But he stressed the importance of the federal statistical agencies.
'The government data is really the gold standard in data,' he said. 'We need it to be good and to be able to rely on it.'
The bureau, in its announcement, indicated that the cutbacks have had only a minimal impact on the overall inflation numbers. A statistical analysis conducted by the agency found that suspending data collection in the three cities changed annual inflation estimates by less than one one-hundredth of a percentage point on average. The effects weren't consistently in one direction; the cuts were more or less equally likely to push estimated inflation up and down.
But that analysis didn't examine the impact of the broader cutbacks in data collection, Mr. Sharif noted. He called the agency's study 'extremely limited.'
'If that was meant to make us feel better about the quality of the C.P.I., it didn't help,' he wrote.