Between 1979 and 2012, after accounting for inflation, the productivity of the average American worker increased about 85 percent. Over the same period, the inflation-adjusted wage of the median worker rose only about 6 percent, and the value of the minimum wage fell 21 percent. As a country, we got richer, but workers in the middle saw little of the gains, and workers at the bottom actually fell behind.

The economy did not always work this way. From the end of World War II through 1968, the wages for workers in the middle, and even the minimum wage, tracked productivity closely. The economy, bolstered by the labor, civil rights, and women’s movements, greatly expanded opportunity and delivered strong wage growth at the middle and even at the bottom. By the 1970s, however, conservatives and corporate interests had had enough. They regained control of the political system and enacted a series of economic changes that, taken together, greatly reduced the bargaining power of workers at the middle and bottom of the wage distribution. The link between productivity growth and wages was broken.

The refusal to pass legislation maintaining the value of the federal minimum wage—now $7.25 per hour—was one of the most visible manifestations of the shift in policy. At first glance, it may be hard to see why the minimum wage is relevant to the middle class. But the relevance jumps out if we consider where the minimum wage would be today if, as was the case during the early postwar period, the minimum wage had kept pace with productivity growth from its high-water mark in 1968.

If it had, the minimum wage today would arguably be about $22 per hour. Even if we use a more conservative measure of productivity growth suggested by my colleague Dean Baker, the minimum wage today would still be about $16 per hour. To put this in perspective, in 2012, the median worker in the United States—that is, the worker exactly in the middle of all wage earners—made about $16 per hour (about $32,000 per year). A $16 per hour minimum wage—in line with what we paid minimum-wage workers in 1968 relative to the productivity level of the time—would be equal to, or higher than, the earnings of half of all U.S. workers.

That is a radically different world from the one we now inhabit. Obviously, workers at the bottom would be earning much more. But, so too would workers in the middle, whose earnings over the last three decades have moved much more in line with wages at the bottom than at the top. Meanwhile, top earners, who have received by far the largest share of productivity gains since the end of the 1970s, would have given up much of those gains to workers at the middle and bottom. To be clear, high earners in this alternative world would still make substantially more than those in the middle—and even substantially more, in inflation-adjusted terms, than what high earners made back in 1968—but the gap between the top and the rest would be much smaller than is the case in the economy we actually inherited.

In our age of diminished expectations, talking about a $16 per hour minimum wage can induce a serious case of sticker shock. But remember: Adjusting for productivity growth, we already paid workers the equivalent of $16 per hour back in 1968. The higher rate simply amounts to maintaining the same degree of wage inequality we had then, and letting the earnings of workers at the bottom, middle, and top grow at the same rate from that point on—exactly as happened in the 1950s and 1960s.

So, stipulated: The minimum wage could be much higher, which would help those at the bottom and those in the middle. The question is, how do we get there?

Raising the minimum wage overnight to $16 would almost certainly do more harm than good, as employers would struggle to cope with the sudden jump in costs. Large swaths of the contemporary economy are organized around low wages. If wages had been rising regularly with average productivity, employers would have made different decisions about technology and the ways work gets done, with no impact on long-run employment. But moving from the low road to the high road will involve rethinking production techniques and making new investments, all of which will take time.

In his recent State of the Union address, President Obama proposed increasing the federal minimum wage from $7.25 to $9 in two steps over two years. Senator Tom Harkin and Representative George Miller have co-sponsored a bill that would increase the federal minimum wage to $10.10 in three steps over three years. Once it reached the targeted levels, both plans would link the value of the minimum wage to the inflation rate as measured by the Consumer Price Index (CPI).

Both proposals would be significant improvements over the current $7.25 federal level. They would also put an end to the inflation-induced erosion of the minimum wage between legislated increases, a long-standing problem. But even the more ambitious Harkin-Miller bill aims very low. The final rate of $10.10 is far below the $16 figure that puts us back on the path we were on in 1968. In fact, because of the multiyear phase-in, by the time the federal minimum wage reached $10.10, its value would already have lost roughly 5 percent of its purchasing power.

Even the proposal to link the future level of the minimum wage to the CPI is timid. Essentially, such a provision would guarantee that the standard of living of minimum-wage workers would not deteriorate (a good thing), but it would also have the effect of acknowledging implicitly that we don’t expect minimum-wage workers to share in overall improvements in the productive capacity of the economy.

One alternative to indexing to the CPI would be to index instead to the median wage or to the average wage, which is usually higher than the median. (The average gives more weight to those at the very top, so when inequality is rising, as it has since the 1970s, the average grows faster than the median.) While we don’t have exactly the kind of historical data that we’d like, at its high-water mark in the late 1960s, the minimum wage was probably equal to slightly more than half of the median wage and slightly less than half of the average wage.

Using half of the median wage as a benchmark today would put us at about $8 per hour, only a bit higher than the current $7.25. This relatively low level goes to deeper problems with the economy; in recent decades, median workers, too, have trailed well behind productivity growth. Linking to average wages produces a higher number. According to the most recent Bureau of Labor Statistics data, the average hourly earnings of all employees (including the estimated hourly earnings of salaried employees) is just under $24 per hour. At 50 percent of the average, that would put the minimum wage today at about $12—well above the current $7.25 or the President’s $9 or Harkin-Miller’s $10.10, though still well short of $16.

Indexing to average wages actually makes more sense because average wages have moved much more closely in line with productivity than median wages. The average wage has tracked productivity more closely because top earners—lawyers, doctors, corporate and financial managers—have captured the lion’s share of productivity gains, and their gains have driven up the average even as middle-wage workers have made little progress. A minimum wage tied to average wages is more likely to keep pace with average productivity, providing the best guarantee that low-wage workers stay connected to the rest of the economy. And the wage push from the bottom will inevitably spill over to wages in the middle.

For a full-time, full-year worker at the minimum wage, the difference between a $12 minimum and a $9 minimum amounts to roughly $6,000 on an annual basis ($24,000 to $18,000). So the level obviously matters for workers at the bottom. But it matters for workers well into the middle class, too. The minimum wage sets the lowest rung on the labor-market ladder, a point where many in the middle class begin their work lives. The minimum wage also acts as a reference point for wages and salaries much higher up the ladder. To maintain relative pay differences among workers with different skills and experience, wage increases at the bottom ripple up. The resulting increases are likely to be smaller in the middle than at the bottom, but a higher minimum wage would have real spillovers to the middle—just as the collapse in the minimum wage has been associated with stagnating wages in the middle.

Ultimately, what a workable “middle-out” economics seeks is a return to “wage-led” growth, where regular increases in wages are the source of smoothly growing aggregate demand in the economy. For more than three decades, the economy has relied instead on what could be called “bubble-led” growth: trickle-down from the spending of the top 1 percent, or unsustainable bubbles in the stock market or housing to spur growth. A higher minimum wage is, on its own, not enough to put the economy on a wage-led growth path. But it is a simple and essential part of any broader plan that would.