Productivity has surged, but income and wages have stagnated for most Americans. If the median household income had kept pace with the economy since 1970, it would now be nearly $92,000, not $50,000.
The problem a retired person faces is that while you can stockpile food, you can't stockpile human labor. And yet every modern person's consumption bundle consists primarily of services, and that's especially true for elderly people who need a lot of health care services and other assistance. So you can't just stockpile stuff for yourself, you need to stockpile a bunch of extra stuff that you can trade for labor in the future. For example, you might work and save and buy two houses. Then, when you're old, you can live in one house and let someone else live in the second house in exchange for taking care of you. That strategy works great as long as the population is rising rapidly, in which case the ratio of able-bodied workers to houses will be high so you'll be able to trade your spare house for lots of labor. But if the population is stagnant, your house isn't so valuable. If the population is actually shrinking, then one house is only going to be tradeable for a little labor. You, personally, can ameliorate this by saving enough to have two or three spare houses, but if society attempted to do that systematically the problem would only get worse.
In the real world, of course, people save through more sophisticated instruments. But the same problem arises. Stockpiling financial assets only helps if future people are going to want to buy a lot of financial assets. If the rate of population growth slows, your private savings plan faces the exact same problem as a pay-as-you-go public pension scheme like Social Security. Fewer people around to pay payroll taxes means fewer people around to sell your financial assets to. You're doomed.
What if people start having more children? It's too late for that. Even if it happens, it'll take decades for those babies to get through school and start earning enough money to want to buy financial assets. The only viable option, then, is to import adult human beings—immigrants—who'll want houses to live in, who will expand the labor supply, and who will expand the demand for capital goods. This solution won't work forever, since the trend toward population aging is global, but in the short term at least there are plenty of able-bodied adults who are eager to move here and work. Some of those able-bodied adults even have specialized skills in health care professions, and can directly increase the supply of the kind of labor that an aging America will need.
What is undoubtedly what readers of Matt Yglesias' blogpost on immigration and retirement income are saying. Matt correctly notes that an economy cannot collectively save for a generation's retirement in the sense of putting aside the goods and services that the generation will consume in retirement. His conclusion is that we need large numbers of new workers to support our current or soon to be retired population. This leads him to call for a much larger number of immigrants.
While we may want more immigrants, the need to support a larger retired population should not rank high on the list of reasons. According to the Social Security trustees projections, a more rapid pace of immigration will make little difference to the program's finances. This is due to the fact that immigrants will also get benefits. Since they tend to work for lower pay during their working lifetime, and the program's payout structure is highly progressive, the net gain from more immigrants is limited. Increasing the projected immigration level by 30 percent reduces the projected long-term shortfall by less than 10 percent.
On the other hand, suppose that real wages grow roughly in step with productivity. If we saw real wage growth of 1.5 percent annually, then the tax increase needed to meet the projected 75-year shortfall would be equal to 4.6 percent of projected wage growth over the next 30 years. Suppose we got real kinky and imagined we saw some of that 2.0 percent annual wage growth that we had in the golden age (1947-1973). Then the tax increase need to main the program's solvency would be equal to just 3.2 percent of projected wage growth over the next 30 years.
The story here is straightforward. We expect retirees' income to be related to their living standards in their working lifetime. If wages grow rapidly then it is easy for a smaller number of workers to support a growing population of retirees while still ejoying a rise in living standards. This is the way the world used to work. It might not be easy for political reasons to get back to that world, but we should at least know that such a world did once exist and is still possible.
Productivity has surged, but income and wages have stagnated for most Americans. If the median household income had kept pace with the economy since 1970, it would now be nearly $92,000, not $50,000.