The first thing to recognize is that a policy of enforced “price stability”, whether implemented in terms of levels or rates, is a form of goverment-provided social insurance, just like unemployment or disability benefits. For all of these programs, there are states of the world in which some individuals might suffer misfortune. The government acts to counteract that misfortune, imposing costs on other individuals in order to fund a transfer of resources. With unemployment insurance, business owners and employed workers pay unemployment premiums, which fund benefits for workers who lose their jobs. A similar dynamic holds for stabilizing prices.

Consider an adverse supply shock. Absent government action, the effect of a reduction of the supply of goods and services would be higher prices. The only way to prevent higher prices is to concomitantly reduce aggregate demand. The reduction might be implemented by raising interest rates or other monetary operations, or it might be effected via taxation. In either case, some people will pay a cost, which will show up as a reduction of demand. Other people will enjoy a benefit from the absence of price inflation.

Who are these people? Can we identify them? Sure. People who benefit from nonincreasing prices are people who hold nominal-dollar assets. That includes most obviously creditors — people with money in the bank, bondholders, etc. — but also people with stable employment but little bargaining power to pursue raises. These groups would see their purchasing power fall in an inflation. If the government restrains prices by reducing aggregate demand, it helps these groups by shifting costs to others. If prices are stabilized via monetary policy, debtors pay: both the increase in interest rates and the reduction of aggregate demand increase the burden of repaying debts. If prices are stabilized via increased taxation, then obviously whoever bears the incidence of the new tax pays. In both cases, marginal workers pay, by enduring an increased likelihood of becoming unemployed or a diminished likelihood of finding a job. [1]

It is fairly obvious, then, that restraining prices in the face of a supply shock effects a transfer from debtors, taxpayers, and marginal workers to creditors and secure workers. A policy of price restraint is a form of insurance for creditors and secure workers, who are absolved of the risk that the purchasing power of their nominal assets will suffer an unforeseen decay. It is financed with a guarantee written by debtors, taxpayers, and marginal workers, who are put at risk by the policy. [2]

So far, we have considered an asymmetric policy, price restraint. But suppose that the state targets a price level or an inflation rate symmetrically? Then don’t the losers from potential price restraint become the winners from potential price support when the state acts to prevent deflation? Absolutely! Under a symmetric price targeting regime, if by monetary policy, debtors enjoy the benefit of a positive supply shock in terms of lower interest rates and increased aggregate demand from which to draw income. If by fiscal policy, the benefit of the happy shock is distributed to whoever captures tax cuts and to recipients of government transfers and expenditures. Whether monetary or fiscal, an antideflationary response to a supply shock implies an increase in aggregate demand which helps keep marginal workers employed. Creditors and secure-but-stagnant job-holders lose out, as the increase in purchasing power they otherwise might have enjoyed via deflation is distributed to other parties.

So is a symmetric price targeting regime “fair”? Absolutely not. Sure, it involves an exchange of risk and benefits, rather than distributing only risks to some and benefits to others. But for this sort of swap to create value, benefits must be matched to losses. Risk-averse individuals seek insurance that pays off in bad times. All groups are at greater risk when an economy is poor following a supply shock. Under any policy of price restraint, creditors and the securely employed enjoy an insurance benefit. Under an asymmetrical policy, that insurance is free, the premiums are paid by other people. Under a symmetric policy, creditors and the securely employed purchase their insurance against bad times by foregoing some benefit during good times. That’s still a fine deal. Their overall risk is reduced.

But the opposite is true for debtors, taxpayers, and marginal workers. Just when these groups need a break, when the economy is bad due to an adverse supply shock, they are hit with additional costs in the name of price stability. Sure, when things are good all over, they get some frosting on their cake. Their highs are higher, but their lows are lower. Symmetrical price targeting turns debtors, taxpayers, and marginal workers into high-beta speculators on the state of the broad economy, while reducing the risk exposure of creditors and secure workers. It represents a vast subsidy, a transfer paid in risk-bearing, from debtors, taxpayers, and marginal workers to creditors and secure workers. A symmetric price target is a better deal than asymmetric price restraint for debtors, taxpayers, and marginal workers — better to have some benefit than no benefit for the burden of guaranteeing other peoples’ purchasing power! But a symmetric price target is still a raw deal. Debtors, taxpayers, and marginal workers are forced to bear costs when they are most burdensome and receive payoffs when they are least valuable.

In the real world, of course, we usually see something between a policy of pure price restraint and a symmetrical price targeting regime. And usually, price stabilization is implemented via monetary policy. We assiduously provide insurance to creditors and the securely employed, but haphazardly reward debtors and the marginally employed when they least need help. Price stabilization is social insurance we provide to the most secure members of our society, while the bill is paid in lost purchasing power and increased risk by the least secure. Further, the benefits of price stabilization accrue disproportionately to the largest creditors and to holders of high-salary secure jobs. Preserving the purchasing power of a billion dollar stash is a lot more valuable than preserving the value of fifty bucks in a bank account. Price stabilization is an incredibly regressive form of social insurance, a program whose distributional ghastliness would be abhorrent to most people if it were not conveniently submerged. But the transfers engendered by price stabilization are invisible, obscured by the money veil. Since they benefit the most influential and harm the most marginal in our society, this ghastly policy is politically untouchable.

It is certainly true that there are groups in our society whose purchasing power we ought to collectively insure: retirees on fixed incomes, savers with moderate nest eggs. It is great that Social Security payouts are indexed, so that retirees enjoy some protection of purchasing power. But indexing is a visible, and visibly costly, form of social insurance. Because it is visible, we transparently ration its provision and allocate its costs. I do not argue that purchasing power insurance is immoral. On the contrary, we need purchasing power insurance and the state should invent explicit means to provide it. What is immoral is to hide what is arguably the government’s largest social insurance program behind the technocratic phrase “price stability”. This a scheme that forces the most precarious members of our society to insure the purchasing power of the most secure, without any limit or even any accounting of the scale of the transfer.

Addenda:

1. In the argument above, I have considered only the effect of supply shocks, not of shocks to aggregate demand. Price stabilization, in the counterfactual that it were fully symmetric, would seem less awful if we considered demand shocks. But it is silly to do so. The tools that we use to stabilize the price level all work through aggregate demand. To the degree that it is possible to stabilize prices, it is possible to stabilize aggregate demand directly. I am not opposed to macro stabilization in general. Aggregate demand stabilization is a great idea, although I have preferences with respect to means that might differ from those of the market monetarists who most famously advocate the policy. But since aggregate demand manipulation is our instrument, it is simpler to stabilize that variable than to stabilize prices. Causing aggregate demand to deviate from a planned growth path in order to stabilize prices is what is immoral. The price level should have no weight whatsoever in macro policy, which should simply target an NGDP path.

2. I’m aware of New Keynesian models that predict “divine coincidence”, where stabilizing prices would stabilize real production as if by an invisible hand. The conditions under which we might rely on “divine coincidence” even in theory are unlikely to hold in practice. The models that predict divine coincidence posit one “representative household”, and so are blind by construction to the distributional concerns discussed here. I consider “divine coincidence”, in almost any forward-looking context, to be another name for “error”.

3. I’ve neglected term effects when discussing the effect of interest rate changes on creditors. For creditors holding long-maturity debt, raising interests rates to restrain prices helps by supporting the purchasing power of the principal lent and by increasing the interest they might earn on reinvestment. But it also harms by provoking a capital loss should they need to liquidate and spend their term debt prior to maturity. For long-term savers, the reinvestment effect compensates the capital loss. Creditors as a group are clearly made better off by a policy of price restraint, but some long-term creditors do get burned by rising interest rates.

Notes:

[1] Unemployment contributes to price restraint by reducing worker bargaining power and therefore the cost of a major factor of production, and by diminishing consumption of goods that might be in scarce supply, as people without jobs sharply curtail consumption. In theory, unemployment might also lead to price increases due to a reduction in supply from goods and services not produced by idle workers. But if the unemployment is caused by restraint of aggregate demand (rather than, say, destruction of a productive factory), the effect of reduced consumption will dominate, as the people fired will be people who provide more exchange value by refraining to consume than by producing, after consumption baskets shift under tighter budget constraints. What the unemployed do not eat is the basis of everyday low prices for the rest of us. Note that these workers are “zero marginal product“, but only in a narrow and artificial sense. The marginal product of workers as reflected in hiring and firing patterns is clearly sensitive to aggregate demand policy, demonstrably over the short-term and almost certainly over the long-term. Those who appear to offer “zero marginal product” when nominal income is scarce and prices stable might provide a high marginal product when income is plentiful and prices are flexible. Ones marginal product today is a function of policy as well as intrinsic qualities, and economic activity is very path dependent.

[2] Of course, these various groups are not exclusive. One can be both a debtor and have a stable job, both a creditor and a taxpayer. Each individual enjoys some benefit, and bears some burden, from a policy of enforced price stability. But on net, there are winners and losers from such a policy. If price restraint will be implemented via monetary policy, large creditors and secure non-indebted workers enjoy the largest net benefit while marginally employed debtors bear the largest risk. Securely employed debtors would experience a mix of risk and benefit, depending on the scale of their salaries and debts.