But, this adjustment may not be as accurate when the economy is struggling.
Why does the BLS adjust employment numbers for seasonality?
Nearly every year, the unemployment rate spikes in January and June, with unemployment up about 1 percent in January over December, and up about 0.5 percent in June over May.
The reasons for these spikes are simple. A lot of temporary retail workers that had a job in December for the holidays are unemployed in January. Similarly, about 4 million students graduate from high school each June and go from being “employed” as students in May to “unemployed” workers looking for work in June.
These seasonal adjustments are critical to understanding where the economy is going. They make for more accurate comparisons from month to month so we don’t all freak out every January and June. Let’s see how they affected our results for January 2011.
At the beginning of February 2011, you likely heard in the news that, “The January 2011 BLS report states that the unemployment rate fell by 0.4 percentage points to 9.0 percent in January, while nonfarm payroll employment changed little (+36,000).” Hearing this, you may have thought, “Wow, that’s great news.” You were celebrating the seasonally adjusted number.
Had you heard the unadjusted statistics explained, you maybe wouldn’t have been so happy. You would have heard something like, “The January 2011 unemployment rate of 9.8 percent was 0.7 percent higher than the rate for December. Similarly, January nonfarm payroll employment was down 2.9 million from December’s number.” Uh-oh, that’s all bad news.
The adjustments can help keep actual statistics in perspective, but they are harder to make during a recession because, if the retail sector is dead, for example, there is no significant increase in unemployment in January for retail workers, because there wasn’t an increase in employment for these workers in December, as in typical years.
The result is that the BLS does the best it can to figure out what changes in unemployment are long term, and which are just seasonal variations. It just can’t do a great job in strange times like these, so the BLS sometimes produces reports, like January’s, that may not provide a clear picture of how the employment situation is really faring. It’s hard to say.
What should you do?
1. Try to not focus too much on the unemployment numbers you read. Know that they are full of uncertainty.
2. Don’t read too much into one month’s seasonally adjusted result.
3. Look at the raw (not seasonally adjusted) unemployment rate for a month compared to the rate in the same month the year before (year-over-year rather than month-over-month). This gives a clearer picture of where we really are with unemployment.
Check out that rainbow…
If you follow these suggestions and look at the long sweep of the whole past year for the raw unemployment rate, it is pretty clear the economy has turned a corner. Since August 2010, the raw unemployment rate has been lower every month than the rate a year before.
Even better, since October, the rates have averaged 0.5 percent lower than the rate a year before. There is some evidence this improvement is accelerating, with January 2011’s unemployment rate down a full 0.8 percent from January 2010.
The next big question is whether unemployment will drop quickly, as it did after the severe recession of the early 80s, or whether it will drop more slowly, as it did after the less severe recessions of the early 90s and 00s. Which route we take will become clearer in the next few months.