Will an Increase in Interest Rates Reverse the Decline in Savings?

Martin Feldstein says yes:

The saving rate began falling in the early 1990s as households increased consumption in response to their rising level of wealth – initially a rising stock market and, more importantly, double digit increases in house prices. More recently, this wealth effect was reinforced by the rise in mortgage refinancing that was induced by falling interest rates. Households extracted trillions of dollars … from the value of their homes and a substantial part of that cash found its way into consumer spending. The forces that lowered the US saving rate are now being reversed. House prices nationwide are down from the peak reached in the middle of last summer. The Federal Reserve has reversed its low interest rate policy … It will only be a matter of time until the household saving rate is at least back to the 2.4 per cent level of 2002. To convert this higher saving to a reduction in the current account deficit requires increased exports and a shift in Americans’ spending from imports to domestically produced goods and services. A lower dollar will provide the necessary incentive for both of those changes to occur.

In other words, the recent increase in interest rates combined with a possible reversal of the wealth effect will induce households to save (rather than consume) more of their after-tax income. Mark Thoma focused on the movement along the savings schedule aspect rather than the wealth effect induced shift of the savings schedule – and then provided us an interesting time series graph of the personal savings rate versus the nominal interest rate on Treasury bills.

I could be critical of Mark’s graph and argue that we should be looking at the real interest rate, but then for the last 20 years or so, much of the movement in nominal rates has likely been movements in real rates as inflation has been relatively low and stable (as compared to say the 1970’s and early 1980’s). I could also note that correlations of two endogenous variables (interest rates and savings rates) over time tells us little about the elasticity of their relationship when both the demand curve (investment) and the supply curve (savings) shift over time. But this is where the story gets interesting.

Part of the debate over whether Bush’s fiscal folly would massively crowd out investment – at least in terms of a long-run neoclassical growth model – has to do with one’s views as to the Ricardian Equivalence proposition that a reduction in public savings would be matched by an increase in private savings. Simply put – the fall in private savings might at first glance be a puzzle to true believers in the Ricardian Equivalence proposition. Many of us have argued, however, the fiscal folly has indeed induced an inward shift of the national savings schedule. Ricardians might counter than the alleged inward shift was due to wealth effects.

Of course, an inward shift of the savings schedule should have increased real interest rates – to which Ben Bernanke argued we have been witnessing a global savings glut. Others (including me) have argued the fall in real interest rates was more due to the decline in investment demand as well as net exports – at least in terms of the U.S. economy. Feldstein is arguing that part of the reason for low savings rates is simply this shift along the savings schedule. The good news is that the global investment deficiency is reversing itself and Feldstein is hoping the same thing happens in regards U.S. net exports. And with an outward shift in the demand for loanable funds driving up interest rates, Feldstein is hoping for an increase in national savings on his premise that the national savings schedule is not perfectly inelastic.

While I have similar hopes in regards investment and net export demand, we have to ask whether the observed correlation between savings rates and interest rates is solid evidence for the proposition of an elastic savings schedule. On this empirical issue, I just don’t know.