interfluidity » Price stickiness is not a mystery, and it is not psychology · by Steve Roth writes:

I mean to write about something else today, so this will be short.

But I’d like to point out, as gently as possible, a mistake in the premise underlying this post by my friend Tyler Cowen. To be fair, it is not a mistake that is uniquely Cowen’s. Macroeconomists invoke price stickiness as an assumption in their models. They treat it as an axiom, a given, and therefore a mystery. Often they lazily fill in the darkness with catch-alls like "psychology" or "money illusion", hypotheses about as useful as pomegranate seeds are as an explanation of seasons. Let’s not have the lazy conventions of macroeconomists stand in for actual thinking on a subject.

Downward price stickiness is a coordination problem, plain and simple. It has nothing whatsoever to do with illusions or cognitive biases or failing to spit after staring too intensely at a small child. Economic entities, both firms and humans, have liability structures rigid in nominal terms. A business has made forward-looking contracts — leases of facilities and equipment, price-stabilized arrangements to acquire raw materials, and yes, contracts with employees that cannot be altered without renegotiation. Businesses have also financed themselves in part with debt, and so taken on nominal obligations whose sustainability is based on forward-looking nominal prices of the goods and services they will sell. Individuals have signed rental agreements or taken mortgages. They have financed their education or their children’s, perhaps they have even taken on consumer debt. For both individuals and firms, these forward-looking nominal arrangements create a very large asymmetry between unexpected price adjustments upwards and downwards. For any economic unit, firm or individual, an unexpected price adjustment upwards in the commodities they sell to market is welcome news. The unit gets more money, its balance sheet expands in the happiest way of all, more assets matched by more equity. But for a leveraged economic unit — and we are nearly all leveraged economic units, if only because we are born short a future stream of housing and food — a downward nominal price shift may force deadweight adjustment costs, which may range from renegotiations of existing contracts to formal bankruptcy to discontinuous shifts in consumption of amenities like housing, education, and local community. (Since these amenities are marketed in sparse bundles, units are not able to continuously optimize consumption tradeoffs, and small changes in budget may lead to large changes of utility.)

Taking account of largely uncontroversial behavioral assumptions like habit formation and reference group anchoring reinforces this case, but is by no means necessary to it. Rational profit-maximizing firms exhibit downward price stickiness as do irrational left-wing humans, although how powerfully either exhibits it depends upon their solvency position and the flexibility of their capital structure, and upon a perceived probability distribution of revenue realizations under different prices. The naive microeconomic case why rational firms would not exhibit downward price stickiness — past arrangements are sunk costs, a rational firm will set prices to maximize future revenue in a forward-looking way — is flawed. It fails to take into account uncertainty surrounding the forward-looking revenue realizations. It ignores the fact that firms are usually quantity constrained in production over the short-term over which payments on their obligations come due. It ignores capital market imperfections, which yield correlations between access to external finance and downward price pressures that are unhelpful. For many firms, the costs and risks associated with an abrupt downward price adjustment are sufficiently large that the rational choice under a reduction of nominal demand is to gamble for the upside of their quantity-constrained revenue-realization distribution. So firms maintain prices higher than Marshallian scissors would advise, and try to sustain anticipated nominal revenues through marketing, product differentiation, and exercise of whatever market power they may have via relationships, network effects, etc. Of course, a strategy that is rational ex ante for firms with imperfectly flexible capital structures will only prove successful ex post for some fraction the firms that pursue it, which helps to explain why reductions of nominal demand tend to provoke consolidations in industries rather than mere rescalings of all the firms that contested the market in good times. When nominal demand collapses, prices fall less than naive economists would guess, some firms sustain quantities sold at the "too high" prices offered, others do not and start to experience large deadweight costs of insolvency. The winners can then buy the losers for a song and quickly dispell those costs.

Precisely the same dynamic accounts for adjustment to changes in labor demand on the extrinsic rather than on the intrinsic margin, for why we see unemployment rather than wage reductions in a recession. In the US (more, perhaps, than in other countries), workers’ lives tend to be leveraged against anticipated, uninsured, market incomes. Accepting a significant wage cut may imply selling a home into a bad market. (Here again, correlations between wage pressures and financial market developments are unhelpful.) It may imply pulling ones kids from schools, excision from civic and social communities, loss of difficult to replace health coverage, loss of ones automobile, and in extremis the humiliations and deadweight costs of personal bankruptcy. The operating and sometimes financial leverage of households sharply magnifies the loss associated with a wage cut, and the discontinuity of the bundles in which crucial amenities are offered magnifies that yet again. Even without invoking a Dunning-Krueger effect, these costs may be large enough that workers rationally prefer to gamble on staying employed at their anticipated wage rather than accept a very painful adjustment with certainty. Firms rationally accommodate this preference among workers, since grateful survivors make better employees on an ongoing basis than people bitterly distracted by their own insolvencies. By firing the workers on whom labor cost adjustment will fall, firms rationally externalize insolvency costs that they would be forced to partially bear if they retained those workers. Even among fired workers, it may be rational to hold out for wages high enough to restore solvency rather than quickly accept work at wages that render inevitable a disruptive adjustment. If they hold out and fail, they face a similar adjustment. But they might not fail. Holding out to search yields a valuable lottery ticket, for a while.

I said at the start that nominal price stickiness is a coordination problem, and it is. If nominal price reductions and nominal wage cuts were accompanied by simultaneous reductions in the nominal burden of each unit’s capital structures, the difficulty of downward price reduction would disappear. For a given household with no financial debt, if it were certain that existing housing, food, education, transportation, and health outflows would scale downward with a wage reduction, the household would rationally accept the wage reduction rather than risk unemployment. But even in a world where households don’t bear financial debts, even during a general depression, there is no assurance that prices will scale down with wage reductions. (On recent historical experience, there’s little evidence at all prices will scale down, that’s the equilibrium we’re in.) So it is rational for many households to resist wage reductions. The prevalence of nominal debt, which bears the stickiest price, renders resistance to downward price and wage adjustments more rational and more likely.

For both firms and individuals, resistance to downward price adjustment is often rational, even when at a macroeconomic level universal downward adjustment would be desirable (perhaps because the central bank and/or state have failed to accommodate the expected path of nominal incomes, perhaps because nominal exchange rates are rigidly misaligned). If we could wave a magic wand and have wages, prices, and especially debts all simultaneously scale downward, that might be awesome. But, unfortunately, we can’t.

If this sounds like some left-wing apologia of unreasonable wage demands (really? does it sound like that?), I’d note that the person who most famously made this argument was one Milton Friedman:

The argument for a flexible exchange rate is, strange to say, very nearly identical with the argument for daylight savings time. Isn’t it absurd to change the clock in summer when exactly the same result could be achieved by having each individual change his habits? All that is required is that everyone decide to come to his office an hour earlier, have lunch an hour earlier, etc. But obviously it is much simpler to change the clock that guides all than to have each individual separately change his pattern of reaction to the clock, even though all want to do so.

Note: I’ve written about this once before, see also an objection by the excellent RSJ.

Update: Nick Rowe generously discusses the ideas (and interacts with me in comments) in three new posts.

Update History:

  • 23-Jul-2015, 8:25 p.m. EEDT: Added bold update Re Nick Rowe’s posts.
  • 28-Oct-2020, 12:25 p.m. EDT: "…and yes, contracts with employers employees that cannot be altered without renegotiation."

This entry was posted on Sunday, July 19th, 2015 at 4:16 am PDT.
You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed. · by Steve Roth writes: