Too often our policy debates occur in silos. Insights from one debate rarely inform others. Well, here’s an opportunity to borrow an insight from the Social Security debate and apply it to an old policy question: How can we better measure poverty in America?

It’s widely agreed that the way we measure poverty is sorely outdated. Of the many limitations to the current yardstick, one of the most serious is that we’re still using income thresholds developed in the early 1960s, updating them solely for inflation. Today, a family of three with an income of $15,220 is considered above the poverty line.

It takes only a moment’s reflection on the contemporary costs of housing, healthcare, childcare and other necessities to realize that this amount is far too low. When pollsters ask questions like “How much does it take for a family to make ends meet in this community?” people usually respond with a figure that’s at least twice as high.

A key reason for the inadequacy of the poverty line figure is that we adjust it solely for price growth and not for the growth in living standards. In effect, we’re asking whether a family can afford the same basket of goods that poor families consumed in 1963, despite the fact that the consumption of the average family has pulled way ahead of that benchmark.

The Social Security benefit formula avoids this mistake. In order to make sure that retirees, surviving family members and the disabled would share in the increases in living standards over their lifetimes, Congress indexed benefit payments to wage growth, not to price growth.

As Ed Gramlich, an economist and a governor of the Federal Reserve Board, put it, “If the system had not been wage-indexed, [retirees] would be living today at 1940 living standards.”

What Gramlich and others are acknowledging is that each generation contributes to improvements in the productive capacity of our economy, and these improvements show up as faster wage growth and better living standards relative to earlier generations. Wage-indexing reflects these gains; price-indexing ignores them. Wage-indexing is one reason why Social Security benefits are high enough to significantly reduce elderly poverty: In fact, without income from Social Security, nearly half of the elderly would be poor, but with Social Security benefits, their poverty rate falls to about 10 percent.

What would happen if we applied wage-indexing to the poverty line? Since 1963 wage growth has outpaced price growth by about 30 percent. Applying this difference to the poverty threshold cited above would raise it to $19,610–an improvement, though still far from what a working mother with two kids requires to meet her family’s basic needs.

Under the current measure, 36 million people (12.5 percent of the population) were poor in 2003. Had we indexed the poverty thresholds to wages instead of prices, to take account of the productivity-induced improvements in living standards since 1963, the poverty rate would rise to 17.7 percent, adding 15 million people.

We take two main points from this. First, because it fails to account for improvements in living standards over time, our poverty measure yields a significant undercount of those who are materially deprived compared to the rest of us. This is not merely academic; our unwillingness to accurately identify the poor constrains our ability to effectively address our poverty problem. Second, in the current Social Security debates, some, including those in the White House, are floating the idea of introducing price-indexing into the program as a way to reduce costs. Opponents of such “reforms” correctly point out that this would be a major benefits cut, resulting in a 46 percent reduction by 2075.

That’s one useful way to look at it, but here’s another: If you’re wondering whether it’s a good idea to index Social Security benefits to prices rather than wages, just take a look at what’s happened to the poverty threshold in the past forty years.