What seasonally adjusted data does and doesn't mean
Commence the snowday meltdown.
Most of the time, when we’re talking about some hot-off-the-press-economic-indicator, we’re talking about a 'seasonally adjusted' number. It’s basically a number that takes into account what’s normal and expected. “There are patterns that occur in economic data regularly at the same time every year,” says Jonathan Wright, an economics professor at Johns Hopkins University. "Some have to do with the weather, some of them have to do with holidays.”
Economists smooth the patterns out, adjusting the numbers up or down. Take for instance jobs numbers. When Kai Ryssdal says, “the economy added 113,000 jobs last month,” on Marketplace, he’s right ... if you’re taking into account the seasonal adjustment.
But if you look at the raw data, the economy dumped nearly 3 million jobs between December and January.
“Seasonal adjustment is this absolutely enormous factor that is taking a minus 3 million number and turning it into a small positive,” says Wright.
There’s a reason economists do this. We expect the economy will lose jobs in January—as stores get rid of their holiday staff. Seasonally adjusted numbers help us spot trends that are out of the ordinary. “What the seasonally adjustment process does is remove from the data variations from normal weather patterns,”says Ben Herzon, an economist at Macroeconomic Advisors.
The key word here is normal.
“When winter weather is abnormally harsh,” says Herzon, “even after seasonal adjustment, the data looks weak.”
Which means all this lousy weather will show up in the “seasonally” adjusted data, making it harder to sniff out what’s really going on in the economy.