Wednesday, February 29, 2012

Seasonally Adjusted Data: What is it and why use it?

When the Oregon Employment Department (OED) creates and publishes employment estimates, we have two sets of numbers: those that are seasonally adjusted, and raw (unadjusted) numbers. In news releases, seasonally adjusted numbers are generally the most prominently reported.

Let's take a look at the January employment numbers released by OED yesterday. Looking at the unadjusted numbers in leisure and hospitality for example, we see a decline of 2,900 jobs between December and January. However, leisure and hospitality generally has a decline between December and January every year. Among other things, temporary seasonal hiring for the holidays generally ends.

The magnitude of decline that we expect to see in Oregon's leisure and hospitality sector is 3,500 jobs from December to January. So, leisure and hospitality retained more workers than would normally be expected over the month. By taking the difference of the expected employment decline (-3,500 jobs) and the actual loss (-2,900), we end up with a seasonally adjusted gain of 600 jobs.Why use the +600 figure instead of the -2,900 figure? Since we know that industries have seasonal patterns from one month to the next, we need to separate out those normal movements by using seasonally adjusted figures to see the underlying trends in the economy. Yes, leisure and hospitality shed jobs from December to January, but the industry always cuts jobs in January. The seasonally adjusted numbers show that more Oregonians remained employed in leisure and hospitality than would have normally been the case over the month, so the industry is actually growing.

No comments: