Purchasing Power Parity
Purchasing Power Parity
Purchasing power parity (PPP) is the theoretical proposition that national currencies have the same real purchasing power at home and abroad. This notion is best understood starting from the point at which two currencies trade at par. Suppose that the exchange rate of the U.S. dollar for the Canadian dollar is unity (as was approximately true for a good portion of the early 1970s). If PPP holds, goods and services sold within Canada (quoted in Canadian dollars) will have the same prices as goods and services sold within the United States (quoted in U.S. dollars). Assuming Canadian and U.S. dollars were both accepted as media of exchange, there would be no need to post prices in anything other than “dollars.” And it would not matter which dollars were chosen.
Consider the following thought experiment. Suppose one or a combination of the three variables in the PPP equation—the prices charged by Canadian retailers, those charged by U.S. retailers, or the nominal exchange rate— changes such that Canadian goods become 10 percent more expensive than the same U.S. goods when all prices are expressed in the same units (i.e., a common currency). If so, consumer demand for the now more expensive Canadian goods would rapidly decline to zero.
Purchasing power parity may be thought of as ruling out this possibility. The literature has contrasted two ways in which this happens. First, arbitrageurs would take advantage of the fact that the market has two prices for goods and services of the same quality. Specifically, arbitrageurs would buy the lower priced goods in America and resell them at the 10 percent higher price in Canada. However, the increased demand for goods and services sold in America will pull American prices up while the increased supply of goods and services sold in Canada will push Canadian prices down. Arbitrageurs will continue buying in America and selling in Canada until the prices in the two countries are equal. Some economists argue that instantaneous arbitrage prevents the departure from parity—departures from PPP would never be observed in the real world. The more likely case, however, is that the force of arbitrage would be quick enough that the temporary nature of the deviation would be a fleeting economic event, of minor consequence from a macroeconomic perspective.
The second scenario is that restoration of purchasing power parity takes a long time. During this time, consumers and firms are altering their demand and supply decisions. In the literature that uses the consumer price index to compute deviations from PPP, the amount of time is takes for a deviation of 10 percent to work its way down to only 5 percent—the so-called half-life of the deviation—is somewhere between three and five years. In contrast to the near instantaneous adjustment case, this is glacial and should be expected to put into motion seismic shifts in demand and supply across locations. Yet, we do not see this type of quantity reaction in the data. Large variations in real exchange rates (deviations from PPP) are not associated with large variations in trade balances or the overall business cycle.
Thus, purchasing power parity, one of the most intuitive of economic concepts, fails to hold in practice. Moreover, its failure to hold does not appear to lead to the type of quantity reactions we would anticipate from basic economic principles. These observations raise two key issues. The first issue is the conceptual validity of PPP. The second is the empirical validity of PPP measures relative to the theoretical concept.
The consumer basket of a modern industrialized country contains literally thousands of individual goods and services, and the cost of arbitrage varies significantly across them. When costs of arbitrage such as transportation costs arise, price differences across locations are expected to arise as well. These deviations are referred to as deviations from the law of one price (LOP). For items in the consumption basket that are not traded across countries, deviations from the LOP are expected to be large and may persist indefinitely. The classic example of a nontraded item is a haircut, which requires the purchaser and the seller to be at the same place at the same time. While the haircut is exchanged in this bilateral meeting, the market for haircuts is localized. Contrast this with heart surgery. Here again, the buyer and seller must be at the same place at the same time. While most major cities have impressive medical centers to meet local demand for high-quality care such as this, people do travel to the markets with the best heart surgeons, sometimes across a vast ocean. The combination of scarcity and high consumer benefit overcomes a large trade cost. Commodities, in contrast, often involve small trade costs and their physical depreciation is slow. This means that buyers and sellers do not have to be in the same place at the same time and inventories provide stocks of supply at or near consumption locations. The general message here is that services tend to be localized in trade while goods are globally mobile. We see these distinctions between goods and services in trade flow data, although the Internet is changing the manner in which some goods and a few services are produced and exchanged. While the PPP exchange rate may be thought of as a weighted average of LOP deviations of the individual items that underlie its construction, there is so much variation across those items that any broad-brushed economic interpretation of the PPP exchange rate are highly debatable.
Currently there are perhaps three or four competing theories that attempt to explain deviations from PPP. The simplest and most enduring explanation for the time series variation in relative prices is sticky prices. We know that the nominal exchange rate varies often by a percent or more over the course of a single day, while many retail prices remain fixed for months on end. This begs the question of why prices are so costly to adjust or why it is not in the firm’s interest to adjust.
If two sellers are physically across the street from one another, we expect the prices of identical goods or services to hold up to a deviation associated with the cost of crossing the street. Otherwise, consumers would never visit the high-price location. Given that we do observe a price differential, we are led to associate the deviation with an impediment to spatial arbitrage. The impediments could take various forms: physical distance, branding, patents, or copyrights.
For example, restaurant menu prices need not adjust continuously in light of what competitors are doing because the menus contain differentiated products. Models of monopolistic competition have been developed to account for price differences across varieties of goods. Here the producer has the ability to charge different prices in the same market even when the cost of providing the different goods is the same.
Price differences could also arise under perfect competition and arbitrage because the cost of providing goods to final consumers involves both shipping costs and the costs of providing goods in the retail outlets. The price of bread in a grocery store must partly defray the cost of labor, capital, and rent borne by the grocer. These costs often differ enormously across locations, particularly when the locations are in different countries. This contrasts starkly with the emergence of manufacturers who sell products online. In this case, everyone pays the same price up to a shipping cost, exactly as the trade cost model would predict.
As models develop in sophistication and the empirical data becomes available to make extensive cross-location comparisons of prices of individual goods and services, we will gain an increasingly clear picture of the reasons for the failure of purchasing power parity to hold at a point in time and over the long run. Inevitably this will involve models that incorporate a variety of goods, services, and market structures. Moreover, the empirical target of the theoretical literature is moving as locations of production and consumption change and as the technology for achieving the exchange of goods and services continues to evolve. Highly accessible introductions to the academic literature on purchasing power parity, the law of one price, international trade costs, and price dispersion in the European Union are, respectively, Rogoff (1996), Goldberg and Knetter (1997), Anderson and van Wincoop (2004), and Crucini, Telmer, and Zachariadis (2005).
Anderson, James A., and Eric van Wincoop. 2004. Trade Costs. Journal of Economic Literature 43 (3): 691–751.
Crucini, Mario J., Chris I. Telmer, and Marios Zachariadis. 2005. Understanding European Real Exchange Rates. American Economic Review 95 (3): 724–738.
Goldberg, Pinelopi K., and Michael M. Knetter. 1997. Goods Prices and Exchange Rates: What Have We Learned? Journal of Economic Literature 35: 1243–1272.
Rogoff, Kenneth. 1996. The Purchasing Power Parity Puzzle. Journal of Economic Literature 34 (2): 647–668.
Mario J. Crucini