Tuesday, July 13, 2010

Sticky price models

I've been trying to make some sense out of the arguments for and against sticky price models. First, Stephen Williamson rants about how he was betrayed by Kocherlakota:

... the Narayana I knew would have thought the worldview represented in standard Keynesian economics was hopelessly naive. ... Woodford's view of the world is not coherent, and it certainly isn't a normative theory - the Taylor rule has never been shown to be an optimal response to anything. Also forget the "positive analysis." One would think that New Keynesians would be thoroughly embarrassed by the financial crisis, which obviously has nothing to do with sticky prices, and left them at a loss for policy prescriptions. ...

... First, there is nothing new in New Keynesian economics, which is successful in good part because it is completely unobjectionable to (i.e. the same as) Old Keynesian economics. Some people might tell you that it's forward looking price-setting that makes the difference, but I don't buy it. Second, New Keynesian economics leaves me empty. There is nothing "important" going on there.

A more reasoned argument is made by David Andolfatto:

... the data show sticky prices and the NK [New Keynesian] model has sticky prices. So what is there to argue? In fact, something very important: Do not confuse measurement with theory. The sticky price hypothesis is a theory; i.e., a proposed mechanism designed to interpret the data. And while the theory arguably has some empirical support, it is not as strong as one is generally led to believe. Sticky price models calibrated to match the observed average duration of price changes (just over one quarter) imply relatively benign consequences. Things get uglier when trying to match model predictions to microdata; see Klenow 2003. These considerations have led some economist to explore other avenues of "stickiness;" e.g., the "sticky information" models posited by Mankiw and Reis (QJE 2002).

...What accounts for the enduring popularity of sticky price models? I'm not sure, but here are some possibilities. First, they do the least violence to the comfort of Walras and Marshall. Second, they imply that money is non-neutral; something that central bankers are particularly fond of believing in. And third, they appear to rationalize (legitimize) interest rate policies like the Taylor rule.

... I have a hard time taking the sticky price hypothesis seriously. The theory literally implies that if prices were fully flexible, many of the worst properties of recessions would be avoided. There would be no liquidity traps, no financial crises, and no lost decades. Conversely, if prices are sticky (in the theoretical sense), simple government policies, like raising the long-run inflation rate or expanding government spending, can evidently restore something close to economic nirvana when the economy is in a liquidity trap. More than one prominent econblogger appears wedded to this view. I remain skeptical of such easy fixes.

Nick Rowe responds:

One of the jobs that economists are supposed to be doing, and have been doing for the last couple of centuries, is to explain prices. To assume prices are fixed is not to explain them; it's to refuse to explain them. We aren't doing our job. It is only a little better to assume prices are "sticky", which means imperfectly flexible, so they adjust only slowly from one equilibrium to another. We don't explain why they are sticky; we just assume it. For example, the Calvo model of the Phillips curve, that underlies most New Keynesian macroeconomic models, simply assumes that firms face a fixed probability d per period of being "allowed" to change price.

... if I hate the assumption so much, why am I still a sticky price macroeconomist?

First, because when I go to the supermarket, what I see and what I do seems to fit my macroeconomic model pretty well.

... A Marshallian economy has many markets, one for each of the non-money goods, where that good is exchanged for money. A Walrasian economy has one big centralised market, where all goods are exchanged simultaneously for all other goods. Sticky-price macroeconomics is Marshallian, in that sense, and definitely not Walrasian. By ignoring the possibility of a market in which labour is exchanged directly for output, sticky-price macroeconomists are implicitly assuming a monetary exchange economy. Barter is ruled out. Recessions are inherently a monetary exchange phenomenon in these models. Recessions are caused by a shortage of money - an excess demand for the medium of exchange. ...

A model of recessions that says they are caused by an excess demand for the medium of exchange seems right to me. I think I see more resort to home production, barter, and private monies during a recession, and I do see more incentive for people to do so. We see proxies for excess supply generally increase: sellers need more effort to sell; buyers need less effort to buy. The prices that do seem more flexible, because we see them rise and fall daily, tend to fall. It looks right.

Unfortunately for me, I am not familiar with the literature and the discussion pretty much left me flummoxed especially the part where Nick Rowe introduced me to the literature that states that if output is demand determined then wage stickiness does not matter but price stickiness does.

To me, all the action is in the labor market. RBC models had problems with it (matching the moments of hours) and it is the labor market where economists tend to focus on in terms of gauging the health of the economy. The key assumption as the post points out is if output is demand determined and I don't really know how crucial this assumption is. Firms produce what they produce and then 'let the markets decide'. There is constant labor reallocation within and between industries as demand for differentiated products fluctuate.

Suppose I start a new firm to crank out some new cereal and start by producing 500,000 units and set the price at $10. I hire some labor to do the job and pay them $5 per hour. If the product is not selling well I may lower the price and either cut back on number of workers or pay them less. If the product does well then I may hold the price constant and either hire more workers or have them work more hours. If a recession hits and I'm in the first situation where the product is not doing so well then whether prices or wages are sticky or not really doesn't matter - I will just shut down and all the workers are out of a job. If I am in the second situation I may cut back on labor or lower the price of the product. If wages were not sticky I may pay them less. If prices were not sticky I may lower the price. It's not clear to me what firms would do. Sticky prices need to be explained in this situation. And I do see that it is more plausible that workers and firms can come to some agreement so that wages are flexible. So perhaps, Nick Rowe is right after all - that it is the product market that matters which in turn affects the labor market.

The question of whether monetary or fiscal policy works better in the situation where the firm continues operation is to ask the question what does the policy maker want to achieve? Is it that the firm does not lay off workers or cut back on production hours? If this is the case then perhaps there is a role for fiscal policy by stimulating demand for my product. What about monetary policy? Does lower interest rates help the firm or the worker? Yes, if we can borrow to smooth production or consumption (at least to tide us over) until the recession is over but this sounds less effective somehow.

Mark Thoma summarizes some evidence which I have yet to digest.

No comments: