Economic growth is a vital sign of a nation's health, reflecting improvements in living standards and the capacity to provide better opportunities for its citizens. For policymakers, setting and achieving growth targets is a top priority, but the question remains: how can growth policies be best informed to achieve their goals?
Policy makers and country analysts have a variety of tools at their disposal to analyze economic growth at the country level. Methods such as growth accounting, or structural change analysis, are relatively straightforward and require minimal data. However, they often provide limited insights into policy. In contrast, detailed micro-level analyses can offer more nuanced policy recommendations but require extensive data and sophisticated methodologies.
Our recent paper explores the potential of growth regressions to inform policymaking and aims to make this analysis easier to conduct.
Growth regressions tell us what changes in key economic variables countries typically experience even though they only provide descriptive statistical correlations. These allows for a quantitative assessment of how growth performance is associated with a specific policy mix and which policy adjustments hold most potential to accelerate growth.
In recent years, we have effectively used growth regressions for country-specific analyses across a broad range of countries at different levels of income and geographical location. This analysis has yielded critical insights into the growth process and its links with economic policy.
Bangladesh achieved an impressive growth performance after 1990. Our analysis revealed that the country achieved outstanding improvements in macroeconomic variables that typically correlate with growth between 1990 and 2005. A combination of effective trade and foreign direct investment policies, recovery of the banking sector from an earlier crisis, and improvements in infrastructure were key factors driving growth in that period (Figure 1).
In the West Africa Monetary Union (WAEMU), growth in 2011-17 was characterized by capital accumulation and driven by structural factors, including financial deepening and infrastructure development. What set WAEMU countries apart from other African countries during this period was the very sharp increase in private sector credit supporting private investment.
In Latin America and the Caribbean, we showed that the remarkable growth in the 2000s can be explained by domestic policy reforms (e.g., improvements in infrastructure and education) rather than by high commodity prices. Growth was driven by ‘good policy’ more than ‘good luck’!
Ethiopia was a particularly interesting case. Macroeconomic policies seemed to comply little with the standard recommendations of reform-oriented growth policies: the exchange rate was overvalued, the financial sector was repressed, and inflation hit double digits. However, the country grew impressively for more than a decade (2004-14). Was Ethiopia special? On the one hand, we showed that typical correlates of growth largely explain Ethiopia’s growth performance over that period. A public investment boom was the key driver while suboptimal macro-policies did not hurt growth much in the short term. On the other hand, the combination of growth drivers was rather unique, making the country’s heterodox policy mix stand out. This also led to concerns about the medium-term sustainability of the growth strategy and ultimately encouraged macroeconomic rebalancing in economic policymaking.
To facilitate the use of growth regressions for country diagnostics we have made available a new dataset with growth correlates spanning more than 150 countries from 1970 to 2019. In addition, we provide the Stata code for several reference models so the user can use the data in different ways.
Since this toolkit was made available, it has already supported policy analysis for several countries. In Fiji, for instance, analysts wondered what the growth impact of capital account liberalization would be. They found an answer by adding a new variable for capital account openness to the original baseline model.
When growth regressions do not fully explain a country's growth episode, it can be just as telling. It may signal a country's unique situation or a detachment of growth from fundamental macroeconomic indicators. For instance, Ghana's growth spurt post-2010, driven by oil exploration and a loose fiscal stance, did not correspond with established growth correlates, potentially foreshadowing the country's later economic setback.
Conducting growth diagnostics at the country level is art rather than science. Analysts are encouraged to draw upon a range of tools and datasets to derive inference and build a coherent, evidence-based storyline. Growth regressions can be a core part of the standard growth diagnostics toolkit. This analysis just got a lot easier.
Join the Conversation