This post originally ran on the Wall Street Journal‘s Think Tank blog.

The New York Times David Leonhardt is a sharp observer of American economic trends, but I think he took a wrong turn in his Monday piece on wages—and data released Wednesday by the Congressional Budget Office helps show why.

Mr. Leonhardt pointed out the dismal wage trends for the vast majority of American workers in recent decades and how it would be a heavy policy lift to reverse them. This seems right to me. But then he wrote:

“Washington could definitely do more to help growth: better infrastructure, a less burdensome tax code, a less wasteful health care system, more bargaining power for workers and, above all, stronger schools and colleges, to lift the skills of the nation’s work force.”

As they might say on “Seinfeld,” you can’t “yada yada” more bargaining power for workers. It’s the most important part of the story.

The root of the U.S. wage problem (which is, in turn, the root of America’s inequality problem) is that most workers aren’t seeing their wages keep pace with overall productivity growth. The policies on Mr. Leonhardt’s list are worthy, but most would not reliably close this gap between productivity and pay. Boosting the bargaining power of workers would.

Finally, Mr. Leonhardt argues that middle-class tax cuts could well be the solution for how Americans’ living standards could increase without wage growth in the future. For several reasons, this seems like the wrong path for this issue.

To start, it’s politically hard to accomplish. Middle-class taxes would have to be cut a lot to move the dial on income growth–and that means that eventually we’d need to cut spending or raise taxes on the very well-off to make up the difference. Cutting spending is disastrous for middle-income households, so that’s no real help. And raising tax rates on the very well-off (something I definitely support) is about as heavy a policy lift as exists in U.S. politics.

In fact, I’d argue that keeping taxes low on the very rich is Republicans’ No. 1 economic priority. Remember the sturm und drang over the “fiscal cliff” at the end of 2012? In the end, top tax rates on incomes above $400,000 ($450,000 for married couples) increased by about 4.6 percentage points. This occurred only because all of the political leverage at the time was on the side of raising rates. Had Congress failed to act and the Bush tax cuts simply expired, the baseline tax increase for high-income households would have been considerably larger.

CBO data released this week show why middle-class tax cuts are too short a lever to really affect the Great Wage Slowdown. Since 2007, the middle-fifth of American households has paid an average of less than 2 percent of their income in federal income taxes. (CBO’s expansive definition of income includes market-based incomes, government transfers, in-kind benefits from employers, and the employer-side of payroll taxes.) If payroll taxes are included, the share rises to 11.4 percent (and CBO is including employer-side payroll taxes here). But since 1979, when the Great Wage Slowdown caused by rising inequality began, overall tax rates on the middle fifth have fallen from 18.9 percent to 11.2 percent—a decline of more than 40 percent.

Yet by 2007, the growing gap between income growth for the middle fifth and overall income growth (a gap that serves well as a rough-and-ready measure of how much inequality costs American families annually) had reached more than 25 percent. This means that even if all federal taxes on the middle fifth of American families were cut to zero, it wouldn’t amount to even half the impact of rising inequality.

Simply put, we need longer levers than middle-class tax cuts to affect wage growth going forward.

Mr. Leonhardt was definitely right on one point: that “liberal-leaning economists can give a long, substantive answer to this question.”

Here is the Economic Policy Institute’s, which we call the Raise America’s Pay project.