The latest data from the Bureau of Labor Statistics reveals a noticeable slowdown in job growth this year. Adding in May’s 138,000 net new jobs, monthly job growth averaged 162,000 so far in 2017, and just 121,000 over the last three months, down from an average monthly gain of 187,000 jobs in 2016. While employment growth would be expected to slow as the economy approaches genuine full employment, other indicators suggest we are not that close to full employment yet, so this explanation seems insufficient. Specifically, at this point in the recovery, we should be looking to not only add jobs, but also see stronger wage growth in those jobs.

But this is not what we’ve seen. Unfortunately, wage growth has been flat over the last year. The latest data indicates that year-over-year nominal hourly wages grew 2.5 percent in May. In fact, as shown in the figure below, wage growth has averaged 2.5 percent over the last two years. If anything, we’ve seen a bit of a slowdown in wage growth this spring, and it is still below levels consistent with the Federal Reserve’s 2 percent inflation target combined with trend productivity growth of 1.5 percent. So, why has wage growth continued to be below target levels after recovery has gone on so long? The simple answer is that while the recovery has been long, it has also been weak. And this weakness combined with the extraordinary damage done during the Great Recession means that slack remains.

But, you say, the unemployment rate fell to 4.3 percent in May, its lowest in 16 years! Unfortunately, the unemployment rate fell for the wrong reasons and isn’t fully reflective of the state of the labor market. That is, in the past when unemployment was roughly as low as it is today, the labor force participation rate—notably that of the prime-age population—has been much higher. Today, there are lots of would-be workers on the sidelines not being counted, who would take a job if offered one. And, the drop in the unemployment rate in the past month is more of a sign of people giving up on finding a job than more people becoming employed. In the last month, the labor force participation rate fell 0.2 percentage points and the employment-to-population ratio also fell 0.2 percentage points. Taken together, that means that the slight drop in the unemployment rate is, in fact, due to would-be workers leaving the labor force, not getting jobs.

The underemployment rate, the U6 measure of labor market underutilization that also include those marginally attached and part-time for economic reasons, also fell 0.2 percentage points in May, but it’s clear that much of that drop is due to falling unemployment, again for the wrong reasons.

All of this is to say that slack still remains and the Federal Reserve ought to let the economy continue to recover and eat away at the remaining labor market weakness before raising rates again. It is abundantly clear that we have a ways to go before we reach genuine full employment—where workers across the country of all ages and races can fully benefit from the economy. Despite this, the Fed is widely expected to raise rates when it meets later this month. This would be unfortunate, as the costs of prematurely declaring full employment and slowing the recovery far exceed the costs of waiting to restrain growth and allowing some wage and price inflation.