Globalization and the Phillips Curve

Last week, I attended the Academic Consultants meeting at the Federal Reserve, where nerdy economics professors like me discussed globalization with Ben Bernanke (an erstwhile nerdy economics professor) and his colleagues. I was assigned to discuss a paper by Larry Ball on whether globalization has fundamentally changed inflation dynamics as described in modern Phillips curve equations. I thought readers of this blog might enjoy reading my discussion.

Comment on Ball
September 28, 2006

Thank you. I am delighted to be here in such august surroundings, in the company of so many old friends and colleagues.

I have to start, however, by saying that the organizers of this meeting made a big mistake when deciding to invite me, especially to discuss a paper by Larry Ball on inflation dynamics. It would take only a few minutes googling to figure out that Ball and Mankiw have more or less the same pedigree. Worse yet, we are frequent collaborators, especially on the subject of price dynamics. Asking me to comment on a Ball presentation is a bit like asking Teller to review Penn’s magic act, except unlike Teller, I really do talk.

So you shouldn’t be surprised to hear that I agree with most of Larry has to say. I would quibble with some details. For example, I think Larry’s paper is a bit unfair to my Harvard colleague Ken Rogoff, when he asserts that Ken fails to understand the difference between absolute and relative prices. But I agree with Larry’s bottom line that globalization is, mostly likely, not crucial for the analysis of inflation dynamics.

From the standpoint of the modern theory of the Phillips curve, price setting depends on two things—the prices that firms want to charge, and the rate at which firms adjust their actual prices to desired prices. If globalization were to change the inflation process, it would have to occur either by affecting desired prices or by affecting the adjustment process.

Desired prices are normally assumed to rise with the economy’s output for the same reason that marginal costs curve slope upward on the blackboard of my ec 10 class. Firms face diminishing returns as their output expands. In addition, when workers are working longer hours, they experience increasing marginal disutility of work, causing their real wage demands to rise, putting further upward pressure on marginal cost and desired prices. There is no particular reason that I can see to believe that globalization will change either the slope of the labor supply curve or the rate of diminishing returns reflected in the production function. If that supposition is right, then marginal cost is just as cyclically sensitive as ever.

To go from marginal cost to desired prices, we need to focus on the markup. I can easily believe that globalization has made many markets more competitive, reducing the size of typical markup of price over marginal cost. This is consistent with the view from business executives that they have no “pricing power”—that is, they face firm demand curves that are more elastic. This increase in competition may have provided a modest, one-off beneficial shock to the inflation process (perhaps spread over several years). But for purposes of understanding ongoing price dynamics, the average level of the markup is less important than the sensitivity of the markup to the business cycle.

There have been several theories of countercyclical markups floating around the literature. Julio Rotemberg of Harvard Business School, for example, writing sometimes with Garth Saloner and sometimes with Mike Woodford, has proposed a model of countercyclical markups in which markups rise in recessions because firms are better at maintaining collusive oligopolistic behavior. If the economy has become more competitive, then it seems possible that the relevance of the Rotemberg story has diminished over time. If so, then the markup will have become less countercyclical, so desired prices would fall more in recessions now than was previously the case. This hypothesis suggests that prices would be more sensitive to output in a globalized economy, which goes in the opposite direction from the evidence that Larry cites.

In the end of the day, I can’t see any compelling reason to believe that desired prices are less sensitive to cyclical conditions in a globalized economy that they would be in a more closed economy.

The second piece of the picture to consider is whether globalization changes the rate at which firms adjust actual prices in response to desired prices. Unfortunately, there is no consensus among macro theorists about why actual prices respond sluggishly. We don’t really know if the key to the mystery is long-term contracts, menu costs, imperfect information, customer annoyance at price changes, or inattentiveness on the part of price setters. And we are on even shakier theoretical ground when trying to figure out if globalization somehow changes this process that we still don’t understand. But my first guess would be that globalization increases competition, which in turn erodes some of the frictions that might impede price adjustment. If anything, we might expect inflation to respond more quickly to the business cycle, not less. Once again, theory most naturally points us in the opposite direction from the evidence.

Let me conclude by spending one minute indulging myself and offering you some thoughts on pop evolutionary psychology, which I think might be relevant here. When I was thinking about this meeting, I wondered why people are as focused on the topic of globalization and price dynamics as they are, especially because basic macro theory does not naturally lead one here. One answer is given by Larry in his conclusion: It is fun to think about big paradigm-shifting ideas like globalization. That is reinforced because globalization is a fashionable topic, with books like The World is Flat finding their way on to best seller lists. (By the way, I have asked the Harvard’s planetary science department about this Flat Earth hypothesis, and they assure me it’s not true.)

My sense is that, as a general matter, globalization as a phenomenon, while no doubt significant in many ways, nonetheless gets more attention than it deserves. We see this manifest itself in many ways. For example, the Chinese exchange rate is not a major issue facing the U.S. economy, yet somehow it manages to get a lot of attention, while more serious problems are routinely ignored.

Part of the explanation, I believe, is that human beings are programmed to be wary of foreigners. In the state of nature in which we evolved, one of the biggest risks we faced is that some other tribe would sneak up at night, steal our food, rape our women, and leave us for dead. As a result, we evolved to be on the constant lookout for risks that strangers might impose upon us.

That inbred fear of strangers manifests itself in many ways. Most perniciously, it is undoubtedly a source of racism and international conflict. But it also has some more benign effects. In particular, it leads central bank economists to spend too much time wondering what those malicious foreigners are doing to their econometric models of inflation dynamics. Which is, perhaps, one reason why we are all here today.