77

In an effort to provide a better understanding of the large variation in price levels between countries, we report on a cross-country analysis of national price levels, using Purchasing Power Parity (PPP) data on 168 economies from the most recent 2011 round of the International Comparison Program (ICP). PPPs are used for many purposes, including to set international poverty lines and allocate IMF quotas. The well-known Balassa-Samuelson income effect is not the only factor affecting PPPs, particularly for low- and middle-income countries. Structural and policy factors make a difference. Small island states are relatively costly for their income level as are sparsely populated countries. Countries with large subsidy programs—as measured by fuel subsidies—tend to have lower price levels than predicted on the basis of income. More open labor policies—as measured by a higher share of migrants in the labor force—are associated with lower price levels in higher-income countries. The proposition that very poor governance is associated with both low-income and high prices receives some modest support. Aid inflows and a negative current account balance are correlated with higher price levels (the latter less strongly) but FDI and remittances are not. We also observe a strong association between inequality and higher price levels, which provides some support for proposition that the ICP may over-weight globally comparable goods.Our results confirm the tendency for African countries to be more expensive than countries with similar incomes in other parts of the world. We fail to fully explain this phenomenon but offer a number of explanations that together could account for it, including low agricultural productivity. Finally, we confirm the relationship between low PPP price levels and greater competitiveness in manufactures, especially for low and middle-income countries.JEL codes: E31, O47, R32.