What determines international wages and prices?

• Author: Crucini, Mario
• Position: Research Summaries

Introduction

Wages and salaries are by far the predominant source of purchasing power for all but the wealthiest individuals in society. The real wage--that is, the ratio of one's nominal wage to the unit cost of a basket of goods and services one chooses to consume--is thus strongly positively associated with the health and welfare of individuals and their families. When goods and labor markets are perfectly competitive, and devoid of barriers to trade or factor mobility, identical goods or workers should command the same market price no matter where the good is sold or the worker is employed. That absence of barriers to trade and factor mobility ensures that arbitrage in goods and labor markets maintains equality of prices and wages.

As it turns out, even within countries, identical workers are not necessarily paid the same nominal wage, nor do they face common market prices of goods and services and consume identical consumption baskets. Therefore, considerable research has been devoted to measuring wage and prices differences and exploring the broader economic implications of those differences.

My collaborative empirical and theoretical research focuses on retail prices of individual goods and services in local currency units (as opposed to index numbers that comprise the sub-indexes of the CPI) and on the use of cities as the spatial unit of account (as opposed to national averages). The cross-sectional differences in price deviations by good and location allow us to identify more of the underlying microeconomic structure of commodity and labor markets and to sustain a richer and more empirically robust class of economic theories.

Long-run Wage and Price Dispersion, the "Penn-Effect"

My early work with Christopher Telmer and Marios Zachariadis studies retail prices of thousands of goods and services across European capital cities at five-year intervals between 1975 and 1990. (1) The underlying data for international price comparisons for this period come from Eurostat, the statistical agency of the European Union, which coordinated the price survey and asked each National Statistical Agency (NSA) to match the exact brand, make, and model of each item across cities. The Eurostat approach was intended to depart from the method used by NSAs to construct domestic CPI indexes, whereby market prices are weighted to reflect domestic consumption patterns. The CPI methodology violates the premise of identical baskets needed to assess the purchasing power parity hypothesis--that is, equality of the cost of a common and broad basket of retail goods and services across countries. The Eurostat methodology satisfies the research criteria.

Average price difference across goods, relative to the EU mean price, ranged from a high of 21.9 percent for Denmark to a low of -25.4 percent for Portugal in 1990. In other words, if Danes shopped in Portugal, they would save 47.3 percent of their expenditure relative to shopping at home. Conversely, if these price differences reflect arbitrage costs in goods markets, then the costs would need to be enormous relative to shipping costs.

After adjusting for differences in the Value Added Tax (VAT), the gap drops to 39.5 percent. We attribute part of the large remaining price level difference across Denmark and Portugal to the fact that they are at opposite ends of the EU income distribution, Denmark with the second highest per capita income after Luxembourg, and Portugal with the lowest (the theoretical rationale for this correlation is elaborated below).

We also indirectly examine the role of trade costs using an index of tradability, finding that goods and services that enter to a greater extent into EU trade volumes relative to production volumes tend to have lower geographic price dispersion. For example, going from the least traded sector (such as a haircut) to the most traded sector (unleaded gasoline), EU price dispersion drops from 43 percent to 12 percent.

The strong positive correlation between income levels and price levels is known as the "Penn Effect" in acknowledgement of the seminal work of Irving Kravis, Alan Heston, and Robert Summers who initiated the International Comparison of Prices Program (ICP) in the 1960s at...