Deflation: the good and the bad

George Selgin has an article in The American Conservative on Deflated Expectations: When fear of falling prices becomes a self-fulfilling prophecy. His basic point is that there are two forms of deflation, the bad driven by demand shrinking and the good caused by supply expanding. The good kind of deflation is the result of increases in productivity. Research and development means new technology, efficiency gains, cost-cutting, price-cutting and, yes, deflation. Productivity gains mean that businesses could afford to sell their products for less since it is costing less to make them.

Selgin writes

And yet, for all the harm it has done, and might still do, deflation gets a bad rap. Although the Fed and other central banks don’t seem to realize it, deflation isn’t always as dangerous as it was in the 1930s. There’s another kind of deflation that can actually be a good thing. And when central banks stand in the way of this good sort of deflation, the results can be disastrous. The current bust is a case in point.

We’ll come back to that. But first, some more word association. The word this time is “innovation.” How about research and development, new technology, efficiency gains, cost-cutting, price-cutting … deflation.

Yes, deflation again. But this isn’t the bad deflation of the 1930s. It comes not from consumers having less money to spend but from them being confronted with more to spend it on. “Bad” deflation happens when demand shrinks; “good” deflation happens when supply expands.

The difference between the two sorts of deflation couldn’t be more basic. Most people grasp it without a hitch. Unfortunately, economists seem to be the exception, perhaps because of their obsession with the Great Depression and zeal to avoid repeating it. Nor has their understanding been aided by the fact that none of them has ever actually witnessed the good sort of deflation.

Yet good deflation isn’t just hypothetical. For much of the 19th century, when the gold standard prevented central banks from printing money willy-nilly, prices fell more often than they rose, and people considered that tendency to be perfectly natural. After all, technology was improving, so goods cost less to produce. Why shouldn’t prices reflect that reality? From 1873-96, for instance, prices in most gold-standard countries fell at an average rate of about 2 percent a year, while real output grew at correspondingly healthy rates of between 2 and 3 percent, thanks largely to productivity gains. That isn’t to say that there weren’t occasional crises—there were, and some involved a dose of bad deflation, driven by temporary lulls in lending and spending. But the general trend of spending was up, while the downward trend of prices remained within the bounds of underlying productivity gains and was for that reason perfectly benign: businesses could afford to sell their products for less as long as it was costing less to make them.

Selgin goes on to say

The complete disconnection of price movements from underlying real cost changes in modern times is a measure of contemporary monetary authorities’ laxness when it comes to preventing inflation, combined with their refusal to acknowledge the theoretical possibility of a benign deflation. That denial rests on sloppy economics that insist on conflating the consequences of good, supply-driven deflation with those of its bad, demand-driven counterpart.

Many experts insist, for example, that to allow any deflation is to risk putting people out of work because “sticky” wages and salaries will fail to keep pace with falling prices, causing rising labor costs to put a squeeze on employers. That’s a fine argument against bad deflation. But if output prices only decline when goods are being produced more efficiently, there’s no need for wages and salaries to fall along with them. On the contrary, productivity gains mean higher real wages and salaries in equilibrium, and the easiest way to achieve that equilibrium is to leave wages alone while letting the price level fall. When output prices are held up instead, money wages and salaries have to rise.

Next, consider the claim that deflation, or at least unexpected deflation, rewards creditors at the expense of debtors, increasing the likelihood of foreclosures. That’s true enough for bad deflation, when earnings are shrinking all around: with fixed loans to pay, something has to give, usually the loan payment. But the same isn’t true for good deflation. After all, if prices are falling because goods are becoming more abundant, why shouldn’t creditors get to enjoy that along with other income earners? Imagine that the extra goods drop like manna from heaven, with debtors and creditors grabbing like shares. What cause do debtors have for regretting the loans they negotiated? None at all. So far as theory can predict, if everyone had anticipated the falling price of manna, the terms of lending would have been no different. The debtors end up with more than before, but so what? The creditors can afford it.

Then there’s the claim that, if they permit deflation, the monetary authorities risk working themselves into a corner, like the present one, with short-term lending rates bottomed out at zero and no scope for any further easing of credit, at least by conventional means. That scenario raises the specter of an invincible deflationary spiral. But here again, while that’s plausible enough when bad deflation is in play, it’s quite implausible when all that’s happening is good deflation because a good deflation rate never exceeds an economy’s rate of productivity growth, and that rate itself sets a lower bound to equilibrium real rates of interest. It follows that as long as only good deflation is permitted, equilibrium nominal lending rates—real rates minus any anticipated deflation—never venture south of zero.

In sum, there’s no reason to fear good deflation or confuse it with its bad cousin. On the other hand, there is reason to fear central bank policies that prevent good deflation—especially those that cause prices to go up while production costs are going down. Such efforts trigger booms and busts.

Whenever productivity advances, so must real earnings. It follows that if output prices aren’t allowed to go down, input prices must go up. The same money creation that serves to prop up the prices of goods also puts upward pressure on the prices of labor and other factors of production. But as deflation-bashers never tire of reminding us, input prices are “sticky.” This means that, in the short run, money being pumped into the economy serves not to raise wages but to boost profits. And high profits, if projected into the future, boost asset prices. Yet the high profits aren’t sustainable because, although speculators may not know it, they are due to give way to higher costs. Voilà—a boom-bust cycle, and one that’s likely to catch investors off guard because the boom takes place in a setting of low inflation.

Selgin expands on these ideas in his book, published by the Institute of Economic Affairs, London, Less than Zero: The case for a falling price level in a growing economy.