McArdle vs. Warren, Round 1: Are We Saving Enough?

Megan McArdle’s taken up the task of trying to prove that the middle class actually isn’t stressed to the breaking point, apparently as part of a project to teach Elizabeth Warren a lesson in good reserach practices or something. Part the first claims that observed low savings rates are, contra Warren et al., nothing to worry about:

But Warren, like many commentators, implies that this is because families are too strapped to save. In fact, it’s because of the two successive bubbles in the stock and housing markets. Families responded to the run up in their net worth by saving less.

Now, you could argue that this was foolish, and I in fact agree with you… The asset model is a standard explanation that pretty much any economist in the country could give you… Even if you think this explanation is wrong, I think you need to explain why your model is a better fit.

I’m a little surprised that she even admitted that the behavior was “foolish.” She might have pulled out this [pdf] paper (*) from the JPE which claims to predict optimal wealth targets and concludes most of us are doing quite well, if anything tending to oversave. (If you believe that, then please allow me to sell you this optimally structured debt-backed security!)

The gut-check on the notion that we’re saving superoptimally, as I see it, is this: what middle-aged person among us wakes up, slaps self on the forehead, and exclaims, “Dang, if only I drank another beer back in college, I could have $10 less in my savings today!”

In fact, a CBO study McArdle links in the post already has some mixed news for her in the details. Even the studies that are generally sanguine about retirement savings adequacy find that savings inadequacy is not independent of class considerations. Shockingly, factors including lower income and lower education have systematic relationships with savings inadequacy. Savings inadequacy tends towards the norm for the lower-middle class and below.

The story that savings rates have rationally declined in response to asset price growth also begs some features of the savings rate data. Specifically, the story ought to be reversible: if everyone was happily booking their home equity inflation in lieu of savings and not overly stressed, financially, then the lower returns on the household portfolios ought to spark a round of catch-up savings. Look at the recent history of the NIPA savings rate, though:

The savings rate collapses to zero, and basically has stayed there, even as the collapse of the RE bubble has been vigorously marking down the value of the middle-class family’s biggest asset. For that matter, the savings rate is no higher in 2002 (after the gyrations of the last recession and late-2001), with the stock market busted and the real estate bubble well under way but a fair amount of kool-aid not yet drunk than it was in early 2000 (and 1999, not shown) when housing and the stock market should have both contributed positively to household balance sheets.

So with the housing market now long past peak, and the stock market lackluster, where’s the pop in savings? Either people rationally want to be poorer in the future (doubtful!) or they ain’t got the money.

Other affiliated observations, like the substitution of credit cards for decreasingly available home-equity loans and increasing credit card delinquency, don’t help McArdle’s case. The sociologists — they study class for a living, the the wacko elitist liberals (**) — have plenty to say on the subject, and that picture isn’t pretty either.

This points to our next installment: are our debt loads really so bad? Stay tuned!

(*) John Karl Scholz, Ananth Seshadri and Surachai Khitatrakun, “Are Americans Saving ‘Optimally’ for Retirement?” Journal of Political Economy, 114(4), August 2006, 607-643.

(**) Kidding about the wacko elitist part.

BTW, my dismissal of models that purport to compare observed savings behavior with results from intertemporally optimal behavior isn’t just that I think the results are obviously at variance with reality. I have a theory for why they’re wrong (and, in a way, a “neoclassical” theory of why paternalistic policy interventions of the sort coming out of the behavioral economics programs can work). The gut check bit isn’t entirely an attempted joke: as I like to think of it, our future selves can’t freely “contract” with our current selves not to do things that we’ll regret ex post. What would you expect our current selves, other than maybe to the extent we try to be H. economicus, to do with that “market power”? The unwary rationalist could confuse the results with a high rate of time preference. So we might be made better off by mechanisms that effectively force us to take the longer view now.