We had long been trying to find a robust classification of growth regimes in country data, and noticed that there was much more interest in the literature for growth onset and growth acceleration than for the onset of declines or the duration of contractions. So we decided to study the latter, and see whether we could relate them to economic policies, the strength of institutions, and demographic structures. We designed a purely statistical definition of economic slumps, namely a specific succession of regime switches: first, a period of positive growth is abruptly interrupted by a sharp contraction, coinciding with a trend break, and later it is followed by another trend break when growth stabilizes again. We assembled macroeconomic data for 145 countries from 1950 to 2014, and identified between 57 and 77 economic slumps satisfying our criteria and recognizable in contemporary history. We developed the methodology to determine the precise timing of the regime switches and the trough of the slumps. Then came the most exciting part: we searched for and analyzed the correlates of the slumps, focusing on the length and depth of the contraction, from the beginning of the slump to its trough. Based on our results we formulated what we like to call three new stylized facts:
(i) weak political institutions predate crises, whereas political reforms tend to follow them, suggesting that abrupt negative growth creates room for political overhaul (in other words, "never waste a good crisis");
(ii) the length and depth of economic declines are robustly correlated with constraints on the executive and ethnic heterogeneity (measured by ethnolinguistic fractionalization); and
(iii) there is a robust interaction between those two variables, suggesting that institutions constraining leaders are especially important for stabilizing growth in ethnically and culturally diverse societies.
We discuss a few country cases, and point out that our `stylized facts' are particularly relevant for Sub-Saharan Africa, where politics are often ethnic and decision makers are comparatively unconstrained.
It was a fascinating piece of research to work on. In a few years, when we will have some statistical perspective on the effects of the current, devastating pandemic, it will be interesting to see whether the pace at which countries recover from it will corroborate the `stylized facts' that we put forward to the world in 2020!
Read the full article HERE