Feb 22, 2023 - Economy

Seasonal adjustments can mislead economy momentum numbers

Data: U.S. Census Bureau; Chart: Axios Visuals
Data: U.S. Census Bureau; Chart: Axios Visuals

You may recall the news last week of a blockbuster 3% surge in January's retail sales, which rippled across global markets. But that report actually showed retailers had $121 billion less in January sales than they did in December — a 16% drop.

  • The difference is a result of the seasonal adjustment process that is applied to most major data — and right now, it may be sending misleading signals about how the economy is doing at the start of 2023.

Why it matters: A series of hot reads on growth and inflation have sent markets reeling this month. But at least part of that heat appears to come from shifts in seasonal patterns, making this winter's numbers look gaudier than they are.

  • There's little doubt that the economy has gained momentum so far this year, but it's less clear how much of that is real.

How it works: If statistical agencies didn't undertake seasonal adjustments, the numbers they produced would be highly misleading.

  • For example, there would be an apparent boom in consumer spending and hiring every November and December, tied to the holiday season — and then a depression each January as sales fall back to earth and seasonal help is dismissed.
  • But those adjustments are based on past results, meaning they are slow to catch on if seasonal patterns change.

So, for example, if 2020-21's pandemic supply shortages caused people to start their holiday shopping earlier than usual, the seasonal adjustment would exaggerate the strength of October's retail sales number and depress November and December.

  • It would also make January's figures look much stronger than the reality, because the falloff in spending from December to January would be less pronounced than seasonal models predict.
  • That looks to be exactly what happened in last week's retail data, which after seasonal adjustments was negative in November and December, and then sharply positive in January.
  • Overall retail sales in January were 0.7% higher than in October and, after seasonal adjustment, more consistent with steady gains than a January boom.

Between the lines: Weather can compound seasonal distortions. In a normal January, frigid temperatures and snowstorms disrupt economic activity in large parts of the country. Seasonal adjustments account for that.

  • But this has been an uncommonly warm winter, which means seasonal adjustments increase reported activity above and beyond the true underlying trend.
  • A San Francisco Fed model that adjusts reported jobs numbers for weather effects found the nation would have added around 390,000 jobs in January — not the 517,000 the Labor Department reported — had it not been a warm winter.

The bottom line: "Right now, it's difficult to ascertain whether COVID-induced consumer behavior changes and business practices are altering seasonal data adjustments, or if the real underlying economic activity is as strong as some recent economic indicators suggest," said Doug Duncan, chief economist at Fannie Mae.

Go deeper