Despite my digs at Jacob Hacker’s new book, I don’t find it implausible that middle-class Americans do feel that their lives are economically precarious, even when they are, in objective terms, immensely economically secure. The question is whether attempting to ameliorate that feeling is a worthwhile aim for liberal policy. Let’s start with a comment Hacker made two weeks ago:

If you have trouble figuring out why risk makes people anxious and unhappy, consider this simple thought experiment: How much of your income would you be willing to put at risk to get a chance at twice your current income? If you’re like most Americans, the answer is “not much”—and for a simple reason: While you’d love to have more money, your life would be thrown into turmoil if your income dropped by, say, half.

Social psychologists have a name for this phenomenon: “loss aversion,” which means simply that we dislike losing things we have far more than we like gaining things we don’t have. No wonder: If your family income fell by half, you would risk losing your home, your health insurance, your retirement savings—in a word, your safety net. And with these vital assets would go your dreams for the future. Maybe it’s no surprise, then, that a recent poll found that even opportunity-loving Americans prefer, by a two-to-one margin, the security of having their current income protected to the chance to make more money.

Hacker’s right that loss aversion is a very real, very well-documented phenomenon. But he’s wrong to imply that the representation of turmoil upon which loss aversion is based accurately predicts the real turmoil that would be experienced in a personal economic downturn. The main point of psychologist Daniel Gilbert’s bestselling book Stumbling on Happiness is that we make systematic errors in forecasting our future feelings conditional on the occurrence of big (or even little) events, such as how we will feel upon losing half our income, to take Hacker’s example. We think it’s going to be a lot worse than it really will be. Famously, people predict that they would be deeply depressed or even suicidal if they lost a leg. Yet real amputees quickly readjust to their new reality, and recover most of their sense of well-being. (And some even report a boost in well-being, their tragedy awakening them to the importance of what they have not lost.) There is certainly a sense of turmoil before one adapts to new circumstances. Indeed, the sense of turmoil is part of the process of adaptation and the recalibration of expectations.

Obviously, whether losing half your family income will dash “your dreams for the future” depends on how big your income was, and what your dreams were. If I had a job that paid 100K, and now I’ve got a job that pays 50K, then I still have my 501K, my health insurance, and probably a lot more safety net than I need. If I can’t now afford the payments on the Mercedes, well, bummer. If the kids are going to have to go to State U, fine. It’s not the job of my taxpaying neighbors to ensure that I can indeed afford Yale once I set my heart on it.

Sure, loss-averse Americans might like the idea of a constantly rising safety net that ensures a short fall, no matter how far we rise. But it’s not what we need, or even ultimately want. The best explanation for human loss aversion is its utility under conditions of scarcity in our environment of evolutionary adaptedness, tens of thousands of years ago. If you’re on the edge, a loss can mean death. But when you’re further from the edge than people have ever been, like well-to-do Americans are now, anxiety about risk and loss can lock us into bad situations, like unsatisfying jobs or loveless relationships. Our overinflated anxieties about the downside of big changes can be one of our biggest enemies. Middle- and upper-class Americans with college degrees have built-in safety nets in the form of their education and skills, and in virtue of being already enmeshed in the most successful wealth-producing institutions in history. The net is already only two inches from our feet. Losses suck, and we hate them. But I find it hard to believe that this is seriously considered a liberal proposal, or a sufficient basis for massive government intervention into our economy and fantastically comfortable and secure lives.

As I said in the last Hacker post, the main bout in the intramural liberal fight is about which set of institutions will provide what we need to exercise our autonomy and realize our ends. I don’t think a lavish social insurance state is the ticket for the poor, and certainly not for the middle. Americans from the middle on up are a class of extremely privileged people whose satisfaction with life ultimately requires moving beyond a complacent sense of safety and facing and taking more risk. Hacker would argue that people will take more risks if the downside is softer. That’s may be true, though it is also possible that people will just readjust their sense of entitlement, finding ever-smaller objective risks equally subjectively intolerable. But we would very probably take more rational, life-enhancing risks if we realized, with the help of a little self-administered cognitive-behavioral therapy, that the downside is already softer than we think. Hacker’s attempt to goad the American middle class into becoming ever more freaked out by their Pleistocene fear of loss is like telling a spoiled child she should definitely wail with a sense of entitled injustice unless she is given yet another pretty pretty pony. It’s perverse, and it’s not helping anyone.

If you haven’t had enough Hacker, here’s Matt Yglesias criticizing Hacker from the left . Let me say something about one point Matt makes about a point he attributes to Hacker:

If the broader economy is getting riskier, this is something public policy should aim to mitigate, rather than exacerbate. The point was simple, useful, and utterly correct.

I don’t think this point is simple, correct, or useful for much other than confusion. Again, if the risk we’re talking about is just the risk of your income fluctuating a bit, a liberal concerned about economic security has little reason to care as long as the fluctuations occur above the threshold of economic suffciency. Furthermore, if those fluctuations don’t generally cause much real harm, but are a symptom of an increasingly dynamic economy that will tend to give people greater opportunity to express their autonomy and realize their ends over the course of their entire lifetimes, then this kind of “risk” may well be something public policy should aim to exacerbate. Progress is not generally something you want to mitigate.