This blog post originally appeared on Wall Street Journal Think Tank.

Recent volatility in stock markets in the U.S. and globally has led many economic observers to conclude that the Federal Reserve is less likely to begin raising short-term interest rates at its September meetings. I’ve been on Team Don’t Raise for a while now, but I’m not excited about those joining the cause in light of recent stock market swings.

As a general principle, the Fed should not react to short-term movements in the financial markets. For one thing, the labor market is much more important to the lives of most Americans, and it is more relevant to the Fed’s mandate of securing maximum employment with inflation stability.

Then consider this: More than 80% of stock wealth in the U.S. is owned by the wealthiest 10% of Americans, and more than half of Americans own no stocks at all (either directly or through retirement or other accounts). In short, movements in the stock markets do not have much effect on the spending power of most U.S. households. That means that movements in the stock markets—especially short-term volatility that is likely to largely dissipate—provides little information about the overall state of economic health.

On the other hand, the labor market provides the vast majority of income to the vast majority of Americans. The middle fifth of households, for example, gets more than 80% of household income directly from the labor market (either cash wages or employer-provided benefits). Further, many additional sources of income such as pensions, Medicare, Social Security, unemployment insurance, or the Earned Income Tax Credit hinge on participation in the labor market. That’s why trends in the labor market are crucial to assessing the overall state of the economy—which is far from fully recovered from the Great Recession.

The clearest remaining weakness is wages. The current pace of hourly wage growth is roughly 2% to 2.5%. A healthy labor market that met the Fed’s overall price inflation target should be churning out wage increases of at least 3.5%. Further, a period of wage growth well above this is necessary for workers’ pay to reclaim some of the ground lost to corporate profits earlier in this recovery. Until wage growth starts moving durably toward the healthy 3.5% target, it’s too early for the Fed to begin raising rates.

This labor-market-based reasoning for keeping rates low should weigh much more heavily on Fed calculations about interest rates than recent stock activity. The only caveat: if one of the root causes of recent stock market declines—the slowdown in the Chinese economy—provides a new potential headwind to U.S. growth going forward.

But the case for keeping rates unchanged in September was dispositive last week, even before large declines in the stock markets. And any strong stock rally in the coming month shouldn’t make Fed officials feel fine about raising rates.