Fooled by the cycle: Permanent versus cyclical improvements in social indicators

This paper studies the time series behavior of a set of widely-used social indicators and uncovers two important stylized facts. First, not all social indicators are created equal in terms of the importance of cyclical fluctuations. While some social indicators such as the unemployment rate and monetary poverty show large cyclical fluctuations, other social measures such as the Human Development Index are, by construction, dominated by long-run trends. Second, interestingly, yet not surprisingly, a large part of the cyclical fluctuations in social indicators can be explained by cyclical changes in income (proxied by real GDP per capita). For this reason, countries with large cyclical income volatility exhibit, in turn, large cyclical changes in some of these social indicators (particularly in those indicators that are more prone to cyclical fluctuations). Since cyclical income volatility is much larger in the developing world, these two critical stylized facts raise fundamental issues regarding the duration of improvements in social indicators (like the ones observed in many developing countries during the last commodity super-cycle). After a detailed conceptual and methodological discussion of these issues, and relying on a global sample of industrial and developing countries, we dig deeper into the importance of cyclical versus permanent components by extending the seminal contribution of Datt and Ravallion (1992). In particular, we show that more than 40 percent of the fall in monetary poverty observed in Latin America and the Caribbean during the so-called Golden Decade can be attributed to cyclical changes in income. While in principle universal, our concerns are particularly relevant in the developing world where, compared to developed countries, output volatility is larger and driven, to a large extent, by external factors (such as commodity prices).