Yesterday, I released a report that looked at the most recent reliable data on Americans’ wages—by decile and by educational attainment, through 2014. These data are illuminating, because they let us look beneath the hood on the overall wage story, going beyond the topline trends that are usually covered by the media.

The recovery has entered a period of solid job growth. That good news shouldn’t be overstated—if we continue to see 2014’s average rate of job growth, it will still be 2017 before we return to pre-recession labor market health—but the economy has been adding jobs at a respectable clip. However, decent wage growth has yet to be seen.

From 2013 to 2014, real, inflation-adjusted hourly wages stagnated or fell across the board, with one notable, revealing exception.

Let’s start at the top of the wage distribution: those workers with the most education and the highest wages. Over the last year, real wages at the top of the wage distribution fell—by 0.7 percent at the 90^th percentile and 1.0 percent at the 95^th percentile. Real wages fell for workers with a 4-year college degree—a drop of 1.3 percent—and even more for those workers with an advanced degree—a decline of 2.2 percent. This sends a clear message: If even these groups of highly educated workers facing the lowest rates of unemployment are seeing outright wage declines, there is clearly lots of slack left in the American labor market, and policymakers—particularly the Federal Reserve—should not try to slow the recovery down in an effort to keep wage and price inflation in check. They’re both already firmly in check even for the most privileged workers.

Moving to the bottom of the wage distribution, we see one glimmer of positive news over the past year. Unlike the rest of the wage distribution, wages actually increased at the 10^th percentile between 2013 and 2014. So, why did wages at the bottom tick up when they fell for nearly everyone else? The answer is simple: we still have some labor standards that provide wage protections. More specifically, 18 states—which make up 47 percent of the workforce—increased their minimum wage in 2014 (either through legislation or through automatic inflation adjustments).

When we compare states with and without a minimum wage increase, we find clear evidence that the minimum wage is the reason the 10^th percentile wage didn’t see the negative trends found elsewhere in the distribution. Wages at the 10^th percentile rose by 1.6 percent in states with minimum wage increases, while in states without such an increase, they pretty much stagnated—increasing by a scant 0.3 percent.

The great news in this story is that policy can actually affect the labor market. And, it is imperative that we use all the policy levers at our disposal to help rejuvenate the economy to create jobs and build stronger income growth for the 99%.

Turning from the past few years and looking at the last three and a half decades, we see that median real wage growth would have been negative if not for the tight labor market of the late 1990s. This dismal middle-class wage growth happened despite real GDP growth of 149 percent and net productivity growth of 64 percent over this period. While productivity has grown substantially since 1979, employers have not put any of the gains from this increased productivity into workers’ paychecks. This divergence between pay and productivity is at the root of stagnant wage growth, slow family income growth, and rising wage and income inequality.

Right now, and for most of the past generation, employers have held most of the cards. Restoring wage growth means dealing employees a stronger hand vis-à-vis their employers. In the near-term, this means allowing the recovery to get back to genuine full-employment. While it is true that the unemployment rate has come down over the last few years, there are still between 5 and 6 million missing workers, who have left the labor force because of weak job opportunities. Nothing saps workers’ ability to negotiate effectively for higher wages like a bunch of willing replacements lined outside their bosses’ doors and the inability to leave to get a better job.

We also need to enact policies that take the thumb off the employers’ side of the scale. This includes raising the minimum wage, strengthening workers’ ability to form unions, reducing wage theft, updating overtime provisions, and ending misclassification of employees as independent contractors. It also means we should regularize undocumented workers, which would lift their wages and wages of those in the same fields of work, and ensure workers can earn paid sick leave and paid family leave, which would provide a greater measure of economic security and allow for a better balance between work and family.

Most importantly, right now, the Federal Reserve needs to keep their foot off the brake. The most recent 2014 data reinforces the fact that there is still a significant amount of slack in the labor market. The clearest rationale for the Fed to raise interest rates is to create slack in the labor market to moderate growth in wages, which in turns moderates upward pressure on overall prices. But, real wages continued to stagnate if not fall in 2014, even in the face of a deceleration of overall price inflation. Therefore, there is no evidence of upward pressure that should signal that the Fed should worry about incipient inflation and raise interest rates in an effort to slow the economy. Aside from simply preventing a misuse of monetary policy levers, even better policy would include additional fiscal stimulus, would could spur acceleration in the economic recovery and lead to real increases in wages and improvements in broadly shared prosperity.