A Yardstick for International Comparisons: An Application to National Agricultural Research Expenditures

International comparative analysis of economic activity often begins with aggregates in current local currency units. The nature of many studies means these aggregates need to be recast into commensurable units so that they may be compared over time and across countries. Unfortunately, any comparisons using data of this type confront an intractable index number problem.' The aggregates typically must be deflated to take account of changes in the local purchasing power of a currency and converted to a common currency in such a way as to give an accurate picture of its real value or volume. The choice of appropriate deflator and converter series can have a substantial effect on the interpretation of resulting data series. The problem of making international comparisons lies in the nature of the available data. To construct comparable measures of real economic activity, index number theory tells us to begin with disaggregated data on prices and quantities in each country to calculate directly a real quantity index.2 For many applications dealing with developing countries, a sufficiently broad and detailed data set is simply not available; instead, one must be satisfied to deflate and convert value aggregates to arrive at implicit volume measures. In the section to follow, we will describe some practical options and demonstrate the nontrivial consequences of different conversion