|
Article Excerpt
ON JANUARY 1, 2002, the euro was introduced as legal tender in 12 European Union (EU) countries. Given that the exchange rates between these countries had been fixed 3 years earlier when the euro was launched as an electronic currency, many predicted that the cash changeover would have little effect on prices. In fact, the average inflation in the euro-zone turned out to be not exceptionally high, and the operation was considered a success (see, e.g., Eurostat 2003, Angelini and Lippi 2005).
In light of this, it is puzzling that most EU citizens think that the introduction of the euro has triggered price increases. Around 70% believe that prices were rounded up. Figure 1 shows that in the euro-zone perceived inflation significantly exceeds actual inflation in the post-euro period, while these two measures coincide in the pre-euro period and in the non-euro EU countries of Sweden, Denmark, and the United Kingdom.
Were most Europeans wrong, or did the euro have some effect on prices? In this paper, we argue that although the average inflation was low, the introduction of the euro had a distortionary effect--it resulted in price increases for lower-priced goods. In contrast to the previous literature, we suggest that the decrease in price transparency was responsible for these increases. Given that people base their perception of inflation mainly on cheap, frequently bought goods (see Guiso 2003, Del Giovane and Sabbatini 2004, European Central Bank 2002), perceived inflation diverged from actual inflation. (1)
We propose a model in which consumers are rational but have difficulty dealing with prices after a cash changeover. A new currency decreases the transparency of prices, thus hindering price comparisons. This weakens competition between retailers: differences in price levels are not perfectly observable, which generates incentives to increase prices. As a result, the equilibrium price is higher after the changeover, even in competitive markets. We argue that for reasonable assumptions about the relationship between the changeover and price transparency, less expensive goods should experience larger price increases.
We find support for our model in the data. First, we analyze self-reported experience with the euro from the Eurobarometer survey (see Appendix B) and show that the introduction of the euro has indeed decreased price transparency. We find that many EU citizens had problems in dealing with the new currency. Among other things, when shopping, they thought in terms of the old currency, felt a need for dual pricing, and had problems with remembering and comparing prices.
We then analyze the relationship between inflation and price levels. Using Eurostat's Harmonized Indices of Consumer Prices (HICP) data (see Appendix B) on inflation for individual product categories (items), we provide evidence that after the introduction of the euro, cheaper products experienced higher price increases. We use an optimal matching algorithm to predict inflation rates of single product categories and show that the model systematically underpredicts inflation rates of cheaper products after the euro was introduced. We call this prediction error the euro-related inflation.
We also show that countries with higher euro-related inflation are those in which consumers perceived inflation to be high. Using consumers' perceptions as a proxy for actual price changes, we analyze the relationship between price transparency and price changes. We find that countries whose citizens report more problems using the new currency had higher inflation for cheap goods. This finding strongly suggests that price transparency is the culprit for the euro-related price increases.
[FIGURE 1 OMITTED]
We also examine how the effect of the changeover depends upon the market structure. If retailers improve consumers' price perception by investing in transparency-enhancing measures, such as advertising, dual pricing, and explicit shop-to-shop comparisons, consumers are more likely to notice shops that lower their prices. Hence, lowering the price may result in higher profit. Since transparency-enhancing investment is costly, it will be undertaken only by shops that benefit from it the most. We posit that these are more likely to be large shops. We find a very strong negative correlation between market concentration and euro-related inflation. (2)
Ours is not the only study pointing out that consumers' perception of high inflation might come from patterns of inflation for only certain products. Ercolani and Dutta (2007) show that certain sectors experienced substantial price increases. At least two explanations for increases in sectoral inflation have been proposed--menu costs (Hobijn, Ravenna, and Tambalotti 2006, Gaiotti and Lippi 2005) and asymmetric rounding to attractive prices (Aucremanne and Cornille 2001). Although these factors might have contributed to inflation, we argue that they are not the complete story. A survey among businesses organized by the National Bank of Belgium shows that 83% of the cost related to the changeover was borne and incorporated into prices before the changeover (National Bank of Belgium 2002). Even for the retail sector, where one might expect less planning, 73% of the costs were transferred to consumers before January 2002. Moreover, although menu costs would have a stronger relative price effect on cheaper goods, they are likely to be similar across countries and thus cannot explain the observed heterogeneity across countries in our measure of euro-related inflation.
Rounding to attractive prices, or price points, (i.e., prices that end with the number 0, 5, or 9) may introduce heterogeneity in price changes across countries and across price levels because the distance between pre- and post-euro attractive prices depends on the exchange rate and on the initial price (El Sehity, Hoelzl, and Kirchler 2005). The empirical evidence, however, cannot be entirely attributed to rounding. First, rounding to attractive prices does not generate the relationship between price levels and inflation that we find in the data. Second, rounding generates heterogeneity across countries that is not related to consumers' difficulty when dealing with the euro or to market concentration. It also does not provide a reason why perceived inflation outpaced actual inflation. In contrast to our model, it is unclear how "rounding-friendly" consumers perceive price changes from attractive prices in the old currency to attractive prices in euro. If consumers perceive price changes correctly, perceived inflation should not differ from actual inflation. If what makes the prices attractive is the fact that consumers mistake them for lower prices, rounding should go unnoticed. Using price level data for Italy, Mostacci and Sabbatini (2003) find that rounding to attractive prices explains at most 0.16 percentage points of the post-euro inflation (out of 2.40 percentage points, or 6.6%).
Our model predicts that in the long run, the euro effect should disappear. Over time, people will presumably learn how to deal with the new currency. Our model therefore also predicts that after the initial spike in inflation, we should observe lower inflation for cheaper goods in the interim, after which prices should eventually follow the path that they would have followed in the absence of the changeover. The long-run dynamics of prices, however, cannot distinguish between different causes of inflation. Menu costs and rounding to attractive prices should also have only temporary effects. Menu costs can generate inflation if the changeover acts as a coordination device because the introduction of the euro involves menu costs anyway, all firms adjust their prices at the same time, thus generating higher inflation. However, price adjustments after that point should happen later than they would have otherwise, leading to a lower rate of price increases for a while, but inflation should eventually reach the equilibrium path. The same is true for rounding to attractive prices.
Like us, Gaiotti and Lippi (2005) propose that a decrease in price transparency explains for the euro-related inflation. In their model, with some probability consumers do not notice the price change, which generates incentives for the producers to increase prices. Contrary to our paper, in their model consumers are not fully rational, since they fail to predict the equilibrium price increase. (3,4)
The three main features distinguishing our paper from the rest of the literature are these: we do not focus on a particular sector, we provide a model that identifies the goods that should be subject to higher euro-inflation, and we provide an explanation for the difference between perceived and actual inflation rates. In addition to analyzing the currency changeover, our paper contributes to the literature on competition with imperfect information. Although many models of consumer behavior attempt to capture the implications of costly information for price determination, (5) it has been difficult to provide convincing empirical tests for these models.
The layout of the paper is as follows: in Section 1 we outline the formal model. Section 2 summarizes the evidence that consumers experienced difficulty in dealing with new currency. In Section 3 we test the model. Section 4 concludes. The data are described in Appendix B.
1. MODEL OF PRICE COMPETITION UNDER LIMITED PRICE TRANSPARENCY
1.1 The Setup
There are N shops selling an identical product at a constant marginal cost c. Each shop sets the profit-maximizing price [p.sub.i], taking prices set by other shops as given. There is a continuum of consumers of measure one. Each consumer buys one unit of the product, and her objective is to minimize the price paid. We follow Salop and Stiglitz (1977) in assuming that consumers always know the distribution of prices P = {[p.sub.l], ..., [p.sub.N]}, but do not know their exact location; i.e., they do not know...
|
