Far be it from me to make a Keynesian economic argument, but I will play devil’s advocate and do so at the risk of alienating any Austrian friends in the audience. They might or might not appreciate the point before I’m done. Writing in Barron’s, Randall Forsyth argues that a zero or negative real interest rate, or specifically a zero interest rate policy (ZIRP), will backfire on central banks precisely because low rates add to the pressure on consumers to save. If that is the case, in the Keynesian paradigm, the policy would undermine consumer demand and lead to weaker growth.

I find it plausible that savers might react to extremely low interest rates by increasing saving. With an aging population and baby boomers fast approaching their retirement years, low interest rates mean diminished opportunities to build on existing assets. The only way to bring more assets into retirement safely is to save more. There has been much said about the impact of quantitative easing and ZIRP on asset values, and the tendency of investors to “reach” for higher, but risker, returns. However, a decent, safe return is hard to come by.

This kind of saving behavior is easy enough to demonstrate for a consumer who must choose between present and future consumption. Present consumption is limited by what the consumer can earn now. Future consumption is limited by what the consumer saves now (does not consume) and the real return or interest rate that can be earned on that saving. The consumer maximizes well being by choosing the most-preferred “bundle” of present consumption and future consumption attainable. But when the interest rate falls to zero, for example, the consumer must reallocate the bundle.

First, the “effective price” of present consumption has declined, since less future consumption must be sacrificed in order to to consume now. So there is a tendency to reallocate the bundle toward more present consumption as a pure “substitution effect”. However, the consumer’s lifetime income has declined precisely because the real rate at which present saving can be transformed into future consumption has decreased. The bundles available for the consumer to choose from are now unambiguously less preferred than the original bundle. Faced with this worsened constraint, the consumer may choose to divide the sacrifice between present consumption and future consumption. The negative income effect on present consumption may well outweigh the substitution effect.

This is standard economics, but relatively little has been said about the possibility. Instead, it is widely assumed that ZIRP must reduce saving, but there have been a few writers making the argument that saving may increase. In 2010, the Buttonwood column in The Economist made this argument in a piece entitled “Another Paradox of Thrift“. In 2012, the Trust Your Instincts blog ran this interesting piece on ZIRP and saving behavior in which the possibility is discussed. For the same reason, Phoenix Capital Management asserted that “QE and ZIRP Are Deflationary“, And the same thing is mentioned in EconoMonitor.

Continuing to indulge Forsyth’s possibility, it does not imply that increased saving from ZIRP will be channeled into productive investment. That’s because governments continue to absorb private saving by running historically large deficits. But I must note that the possibility of increased saving in response to ZIRP may contradict a couple of points made in an earlier post on Sacred Cow Chips: “Taking The Air Out of the Deflation Scare“. That post quotes Thorsten Polliet in support of the notion that the rate of time preference underlying consumer behavior cannot be zero or negative. Does that conclusion change when consumers order bundles rationally with a budget constraint that implies a negative return? In fact, the macroeconomic concept of a time preference “parameter” appears to be inconsistent with the normal micro theory of consumer utility maximization.

Increased saving from ZIRP leads to a second apparent contradiction of Polliet, who says:

“Should a central bank really succeed in making all market interest rates negative in real terms, savings and investment would come to a shrieking halt: as time preference and the originary interest rate are always positive, “capitalistic saving” — the accumulation of goods designed for improving the production process — would come to an end.”

But again, the possibility that saving may increase does not imply that capital investment will increase as well, as long as the government is absorbing the increased saving. In fact, the adoption of ZIRP policies around the developed world seems in large part intended to accommodate large government deficits by keeping interest costs low.

The evidence that ZIRP encourages saving is mixed. Japanese saving rates tended to edge up over the country’s many years of ZIRP (since 1999). More recent experience in the EU seems mixed. In the U.S., saving rates increased during the financial crisis even before ZIRP began, moved down during the recovery, but have since returned to relatively high levels. The Federal Reserve claims that consumers continue to unwind the excessive leverage that built up prior to the recession, and of course that is saving. Paying down debt certainly carries a higher and safer return than many other options. ZIRP cannot be counted upon to encourage consumer spending, and it may well do the opposite.