In the chart, the left-hand panel shows government spending on old-age pensions across OECD countries. The level has basically doubled from over 4% in 1970 to roughly 9% at present. But at least given current projections, and after various steps that governments have taken, government spending on pensions as a share of GDP isn't scheduled to rise much more in the next few decades.
The difficulty arises because if government keeps spending the same share of GDP on pensions at a time when the share of the population who are elderly keeps rising, then the average government pensions will cover a smaller share of income. Thus, the right-hand panel shows that while the "economic replacement rate" for an average individual over the age of 65 has been around one-third of per capita GDP for the last few decades, it's projected to fall to about 20% of per capita GDP by 2040 and later.
What's are some of the possible responses here? One set of reactions could happen within the political system. For example, the elderly could vote for dramatic increases in taxes on the working generation to support higher pensions. Or government pensions could be redesigned to provide only a basic income support, and nothing much higher, even if you paid much more in payroll taxes throughout your life.
Another set of reactions could happen with the actions of individuals. Soto's calculations suggest that if people worked about five years longer before retirement, it would offset about half of the predicted decline in the "economic replacement rate." If everyone also saved at least 6 percent of their income during their working life, this would offset the other half of the fall in the "economic replacement rate." Again, these changes are just what is necessary to offset the expected decline in what government pensions are currently scheduled to pay.
The needs and expectations of an aging population are a tectonic shift under our current political and economic understandings, and will cause some earthquakes before it's done.