Labor Department admits seasonal adjustment distorted unemployment data
Thursday's initial jobless claims report will look a bit different as the Department of Labor announced it will change the methodology it uses to seasonally adjust data.
The state of play: The seasonal adjustments will switch from multiplying by a seasonal factor to adding.
What they're saying: "[I]n the presence of a large level shift in a time series, multiplicative seasonal adjustment factors can result in systematic over- or under-adjustment of the series; in such cases, additive seasonal adjustment factors are preferred since they tend to more accurately track seasonal fluctuations in the series and have smaller revisions," DOL said in a statement.
The intrigue: The Bureau of Labor Statistics had already made this change to its reports, months ago, says former BLS head Erica Groshen, but the agency tasked with releasing the unemployment data, the Employment and Training Administration (ETA), had not until now.
- "This is an example of the difference between having a statistical agency produce official economic indicators from administrative data and having an administrative agency publish aggregates that are used as economic indicators but do not have curation by national statisticians," she told Axios' Felix Salmon.
- "BLS could and arguably should produce a weekly series of indicators based on UI Claims. Sadly, that can’t happen without funding—either through appropriations to BLS or to ETA for this purpose. So, ETA just stumbles along..."
Of note: Over the last three weeks the average for seasonally adjusted initial jobless claims was 21% higher than the average for nonseasonally adjusted claims, notes DRW Trading rates strategist Lou Brien.
- Last year at this time, he says, with about 1/10th as many claims, the seasonally adjusted three-week average for initial claims was 21.5% higher than the unadjusted claims.