To my mind, Social Security, pensions, 401(k)s, and the like can all play a part. My larger point about pensions specifically is that there's nothing inherently less risky in pensions from a purely accounting perspective. Actuarial tables were astonishingly precise 100 years ago, and our ability to diversify and insure risk through private markets has only grown while at the same time U.S. monetary stability has achieved undreamed of heights, even considering asset inflation. The basic dilemmas and moral hazards, meanwhile, have largely stayed the same, albeit sometimes obscured by increasing technical complexity.

The biggest part of the puzzle concerns public policy. For some reason we keep returning to tired debates about socialist vs private markets or personal freedom vs authoritarianism, like we're still picking sides in an insurgent communist revolution at the turn of the previous century. (The current debates are astonishingly similar to the ones we've had prior, such as over hourly limits, employment disability insurance, and especially Social Security.) It's ridiculous. We have a decent grasp on which basic tools work better in which contexts and why, partly because the science of economics has improved, and partly because the past century saw a large number of social experiments play out at large scales both domestically and especially globally. The immediate question isn't which singular option to chose, it's about how to mix-and-match instruments and incentives. Interactions and second-order effects are still fuzzy, and political cultures do matter, but there are so many tremendous improvements we could make before we hit that next wall.