Clearly, the unemployment rate plays an important role in our thinking and communication about future policy. Therefore, an important issue is whether it is giving an accurate read on where things stand relative to our maximum employment mandate. The unemployment rate measures the percent of the labor force that is out of work and looking for a job. It has a number of advantages as a measure for summarizing the state of the labor market. For one thing, over time it has proven to be a reasonably stable and predictable barometer of whether labor market conditions are too hot, too cold, or just right in terms of creating inflationary pressures. Although structural changes in the labor market affect the unemployment rate, most of the variation in unemployment over time reflects cyclical factors, that is, whether the economy is too hot or too cold. And, second, the rate closely tracks other indicators of how much slack there is in the labor market, such as data from surveys on the share of households that finds jobs hard to get and the share of businesses that say it’s hard to fill openings. This adds to our confidence in its reliability.
All the same, there are reasons to worry that the unemployment rate could now be understating just how weak the labor market is. In particular, during the recovery, the share of the working-age population that is employed has increased very little, even as the unemployment rate has fallen. Taken at face value, the very low ratio of employment to population suggests that the labor market has improved far less than what’s implied by the decline in the unemployment rate.
So should we stop using the unemployment rate as our primary yardstick of the state of the labor market in favor of the employment-to-population ratio? My answer is no. Although the unemployment rate is by no means a perfect measure of labor market conditions, the employment-to-population ratio blurs structural and cyclical influences. That makes it a problematic gauge of the state of the labor market for monetary policy purposes…
…
What does that mean for monetary policy? First, the unemployment rate and a number of other labor market indicators, such as payroll job gains, point to continued progress in the labor market. Clearly, we are getting closer to meeting our test of substantial improvement in the labor market.
Second, any changes in policy will depend not only on labor market conditions, but also on inflation. In our July statement, the FOMC noted that inflation persistently below 2 percent could pose risks to economic performance. That means we will also take into consideration whether inflation is moving closer to our target. Third, any adjustments to our purchases are likely to be part of a multistep gradual process, reflecting the pace of improvement in the economy.
As I noted earlier, Chairman Bernanke has laid out a timetable for our securities purchases, which includes reducing them later this year and ending them around the middle of next year, assuming our forecasts for the economy hold true. I haven’t significantly changed my forecast since then, and I view Chairman Bernanke’s timetable to still be the best course forward. However, I can’t emphasize enough that when and how we adjust our purchases will depend crucially on what the incoming data tell us about the outlooks for the pace of improvement in the labor market and movement towards our inflation goal.