Is the economy too hot, too cold or just right?

The unemployment rate is 4.9% -- a historically low number by any mainstream standard. Does that mean the economy is at or very near what the policy wonks call "full employment"? This is probably the most important question the Federal Reserve must answer as it decides when to begin raising interest rates.

If there is still room for labor markets to expand, then raising rates too soon could slow the rate at which people find jobs -- creating unnecessary hardship for people who would like to work.

But if the economy has already reached the point where there is little room for expansion, then failing to raise the target interest rate soon enough could risk overheating the economy -- and unleashing the inflation and reduced purchasing power that comes with it.

In answering the question, the first thing to note is that the unemployment rate consistent with full employment -- what economists refer to as the "natural rate of unemployment" -- changes over time.

During the 1970s, for instance, the natural rate of unemployment went up when baby boomers and women were entering the labor force in large numbers. New entrants into the labor market tend to change jobs more often than workers who have been in the labor force for many years as they search for a good job match before settling into longer-term employment. In those years, the natural rate of unemployment was higher than the 4.9 percent rate we are currently experiencing.

But there are also times when the natural rate of unemployment has been lower than 4.9 percent -- in the last 1990s, for example, when unemployment fell to around 4 percent and inflation did not rear up to become a problem.

There are many other factors that can affect the natural rate of unemployment, making it difficult to estimate this value at any moment in time. That's why economists avoid trying to come up with a specific number for the natural rate of unemployment. Instead, they focus on an indication of the state of the labor market that is easier to measure: How fast wages are rising.

When wages begin rising rapidly, that is a signal the economy is trying to push beyond the full employment level. Conversely, when wages are not keeping up with inflation (as happened during the Great Recession), the conclusion is there is still slack in the labor market. By this measure, we are approaching full employment -- many signs suggest wages are beginning to accelerate but not yet fast enough to translate into worrisome levels of inflation.

One thing seems certain. If the Fed gets nervous about inflation and raises its target interest rate too soon, it will slow the economy and reduce wage growth. Workers' share of output has been falling in recent years (that's part of the reason that income inequality has been widening) and wages must grow faster than inflation for at least enough time for workers to regain what they have lost. For this reason, I believe the Fed should not react to rising wages until its very clear that wage growth is creating an inflation problem.

There is another reason for Fed caution. As wages increase, the effect may be to draw more people into the labor force. During the Great Recession, many workers stopped looking for jobs -- and these workers are not counted in the calculation of the unemployment rate. Both the employment to population ratio and the labor force participation rate for people ages 25 to 54 have been rising as the labor market has improved, but they are still below their pre-recession levels. (Using the 25-to-54 age range insulates against demographic effects -- like an aging workforce transitioning into retirement -- on these employment measures.)

If improving labor market conditions continues to draw people into the labor force, then the time it will take to reach the point where the Fed needs to raise interest rates to prevent inflation will be pushed further into the future. We do not know for sure how many people will be induced to reenter the labor force as conditions improve, but this gives the Fed another reason to be patient in its plans for its next rate hike.

There is one more reason for patience. The latest data on GDP growth was a disappointing 1.2 percent for the second quarter of this year. This is subject to revision as more data arrive, and that could change the picture markedly. GDP growth was held down by changes in inventories that could easily be reversed in coming quarters, for instance. Or it could be the first sign of a slowing economy and the Fed should be cautious about the possibility of raising interest rates if the economy is beginning to lose what little steam it has.

The Fed has a difficult job. Using interest rates to influence the economy is a lot like cooking on an electric stove. When it increases or decreases its target interest rate, the effect is not immediate - it can take time for the economy to fully feel the change. Because of this delayed response to a rate increase, the Fed must forecast where the economy will be several quarters in the future whenever it is formulating plans for monetary policy. Yes, there are signs the economy is improving. Yet there also are signs we have room for more expansion without inflationary pressures. And of course there is some degree of uncertainty about whether the economy will continue to improve at the rate is has in recent quarters.

If the Fed is going to make a mistake, I would prefer that it keep interest rates low and err on the side of doing all it can to stimulate employment and wages.