One inflation report in a data blackout

The federal government shutdown, now several weeks old, has halted most of the statistical agencies. The Bureau of Labor Statistics was granted a one-off emergency exception to publish the September consumer price index, and the number was uncomfortably firm. Inflation ran at 3.0 per cent year-on-year, which put paid to hopes that prices would cool in the autumn.

The exception had a practical reason. The Social Security Administration uses the September figures to set the annual cost-of-living adjustment for pensions, and without them tens of millions of benefit cheques could not be recalculated for the new year. The bureau recalled enough staff to produce the report, which appeared more than a week late, and the agency announced a 2.8 per cent rise in benefits the same day. No other major release has been given the same treatment.

One number, no context

A single inflation figure, published in the middle of a wider data blackout, is hard to read. There were no payroll data, consumer spending figures or wholesale price releases to go with it, so markets and Federal Reserve staff had nothing to set the 3.0 per cent headline against. They could not tell whether it reflected strong consumer demand or a mechanical cost push from import tariffs already in force.

The detail offers some clues but not enough. Core inflation, which strips out food and energy, also ran at an annual rate close to 3 per cent, and prices of some tariff-exposed goods such as clothing and furniture kept rising. Petrol added to the headline. Shelter costs rose only modestly, continuing a slow cooling that has taken the heat out of the services side of the index. Those pieces point to a mix of tariff pass-through and easing rents, which is a very different story from overheating demand, but one report cannot confirm it.

The Fed's preferred gauge is missing altogether. Its inflation target is defined in terms of the personal consumption expenditures price index, which is compiled by the Bureau of Economic Analysis, and that agency has also been shut. Much of that index is built from consumer price components, but the translation needs staff who are furloughed, so the September reading the Fed actually targets has not been published.

Why October may be missing too

The gap is likely to grow. The September figures were built from prices collected during September, before the shutdown began on 1 October. The bureau's price collectors have since been furloughed, so data for October are not being gathered in the usual way. Even if the government reopens soon, the October index may be incomplete or delayed, and statisticians may have to fill gaps with estimates. The Fed could reach its December meeting with only one month of official inflation data covering the autumn.

A precedent from Britain

Policymakers have had to work with broken data before. In Britain, falling response rates to the Labour Force Survey made the official unemployment figures so unreliable in 2023 and 2024 that the statistics office suspended some of its estimates and the Bank of England had to lean on tax records, business surveys and private data to judge the labour market. The Bank's decisions became more cautious and its communication more hedged. The lesson for the Fed is that missing or noisy data tend to make central banks slower to act, because the cost of a mistake rises when the information behind it is weak.

America's own shutdown in 2013 delayed both the jobs and inflation reports for September of that year. The gap then lasted a little over two weeks, and the Fed was not in the middle of changing policy. This shutdown has already lasted longer, and it has come while the Fed is cutting.

November cut priced out

Bond traders responded by pricing out expectations of a rate cut in November. With headline inflation stuck at three per cent and no official data to show the labour market weakening, the Fed has little room to ease quickly.

The strongest argument against that reading is that the report was not as hot as the headline suggests. The monthly rise in core prices was modest and slightly below what forecasters had expected, and private indicators of the labour market, such as ADP's payroll figures and job postings, point to continued weakness. A Fed that cut in September to insure against a softening jobs market has reason to keep going if that softening persists. The difficulty is proving it without official numbers, and the officials who were sceptical of the September cut will point to a 3.0 per cent inflation rate as reason enough to pause.

Who pays for the blackout

The costs spread widely. Traders pay more to hedge against surprises they cannot see coming, and companies setting prices and wages for next year do so with less information about what competitors and the wider economy are doing. Households depending on federal programmes face disruption as the shutdown drags on. The longer it lasts, the larger the revisions will be when the agencies catch up, and the more likely it is that decisions taken during the blackout will look wrong in hindsight.

Signals to watch

The indicators are the length of the shutdown, the Labor Department's timetable for catching up on missed releases, private measures of prices and employment, and the tone of the Fed's statement and press conference this week. Until the shutdown ends and the missing data return, markets will be trading on one inflation number with nothing to explain it, and the Fed will probably sit tight rather than cut in November. The larger risk is delay. A Fed forced to wait for data that arrive weeks late will probably keep policy where it is until it can see the October and November figures together.