An International Comparison of Business Cycles: Stagnation at the roots of diversity?

Business Cycles (BC) in Emerging Markets and Developing Economies (EMDE) are much more volatile than those in Advanced Economies (AE). Understanding the rationale behind this fact is important for policy-makers as volatility has proven to be economically costly and welfare reducing. The greater volatility of EMDEs’ growth rates is often said to derive from a quicker alternation of expansions and recessions. EMDE’s cycles are "shorter" (Rand and Tarp (2002)[1]) and, unlike AE, "are characterized by frequent regime switches" which result from frequent changes in policies (Aguiar and Gopinath (2007)[2]). Yet, recent approaches have partially reassessed this fact and advocated for an evolution in the consideration of EMDE’s BC (Abiad et al (2015)[3]). Calderon and Fuentes (2014)[4] have identified a smoothing of BC in both EMDE and AE during the great moderation. They show that EMDE experience costlier recessions than AE but of similar duration whereas their expansions phases are stronger but slower.

Against this background, the starting question I aim to answer in this article is: Are there definite differences or similarities in the principal characteristics of BC (probabilities to enter/exit a recession, amplitude of the change of regime) in EMDE vis-à-vis AE?
To approach this question, I look at quarterly GDP per capita for 77 countries (23 AE and 54 EMDE). Whereas Calderon and Fuentes (2014)[4:1] develops only on Harding and Pagan (2002)[5]’s dating algorithm methodology I use both a dating algorithm and Markov Switching Models, following Hamilton (1989)[6], to derive my results. This proves advantageous as the latter methodology does not need any a priori specification on the economic cycle and identifies the most statistically significant regimes.
I find that, in accordance with the academic literature, EMDE face twice as important gains/losses when moving from a regime to another. Additionally EMDE form an heterogeneous group with Latin American countries being more vulnerable and Asian countries more resilient. If the latter two results confirm former claims of the literature, I paradoxically find that the average dynamic of EMDE does not differ from AE: they
face similar probabilities to enter and to exit recessions.

The paper then reflects on this ‘paradox’ and aims at proposing potential explanations. I argue that, contrary to the duality - expansions versus contractions - proper to the conventional BC approach (Burns and Mitchell (1946)[7]), one needs to consider a third phase of stagnation to truly uncover the differences between AE and EMDE. The underlying premise to this approach is that, when a country undergoes an economic slowdown, reactions by the agents might be different in AE or in EMDE. A possible explanation might be linked to lesser risk sharing, as identified in Kose et al (2009)[8], which would induce an overreaction of national agents to such a situation. One could also question a different role of agents’ expectations when faced with an economic slowdown. As shown in Evans et al (1998)[9], slow growth expectations prove to be self-fulfilling and therefore might trigger a series of fragilities.

As such the article picks up on this ‘paradox’ by answering following questions: Are stagnations important for understanding AE’s and EMDE’s BC? Do they allow identifying new differences and similarities between the two groups of countries?
When extending the previous approach to include stagnations, both groups prove to behave differently. AE are found to be more likely to move directly from an expansion to a recession, whereas EMDE enter recessions more likely after a phase of stagnation, and remain longer in the recession. To ensure the validity of these observations I confront different indicators of dynamics with frequently used indicators of development (relating to the quality of institutions, the accumulation of factors of production, economic policies and structural features). This confirms previous observations. These facts hint towards two very different conceptions of recessions in these groups of countries. In AE, which experience contractions following expansions, an Hayekian approach seems the most plausible: recessions follow periods of overheating of the economy. On the contrary,
in EMDE, a Keynesian approach steps forward: stagnations linked with lower fundamentals might increase the perception by agents of the riskiness of the situation and thus lead to a decrease of expected demand.

This article contributes in three different ways to the literature. First and contrarily to usual considerations, I identify the fact that the difference in the BC volatility between AE and EMDE stems from a difference in the magnitude of the changes of regime but not from a difference in the dynamics of the cycle. Second, I show that the usual consideration of BC based on the alternation of expansions and recessions misses episodes of stagnations as an important component of output paths. Finally, I identify two very different conceptions of recessions, from the perspective of either AE or EMDE. As such, we can identify major policy implications, by increasing the interest of economic researchers and policy-makers on the behavior of agents during stagnations/economic slowdowns as they might underpin most of the entrances of EMDE into recession.


  1. Rand, J. and Tarp, F. (2002). Business cycles in developing countries: Are they different? World Development, 30(12):2071-2088. ↩︎

  2. Aguiar, M. and Gopinath, G. (2007). Emerging market bussiness cycle: The cycle is the trend. Journal of Political Economy, 115(1):69-102. ↩︎

  3. Abiad, Abdul and Bluedorn, John and Guajardo, Jaime and Topalova, Petia, 2015. "The Rising Resilience of Emerging Market and Developing Economies," World Development, Elsevier, vol. 72(C), pages 1-26. ↩︎

  4. Calderon, César and Fuentes, J. Rodrigo, 2014. "Have business cycles changed over the last two decades? An empirical investigation," Journal of Development Economics, Elsevier, vol. 109(C), pages 98-123. ↩︎ ↩︎

  5. Harding, D. and Pagan, A. (2002). Dissecting the cycle: a methodological investigation. Journal of Monetary Economics, 49(2):365-381. ↩︎

  6. Hamilton, J. (1989). A new approach to the economic analysis of nonstationary time series and the business cycle. Econometrica, 57(2):357-384. ↩︎

  7. Burns, A. and Mitchell, W. (1946). Measuring Business Cycles. National Bureau of Economic Research. ↩︎

  8. Kose, M. Ayhan and Prasad, Eswar S. and Terrones, Marco E., 2009. "Does financial globalization promote risk sharing?," Journal of Development Economics, Elsevier, vol. 89(2), pages 258-270, July. ↩︎

  9. Evans, Geroge W and Honkapohja, Seppo and Romer, Paul, 1998. "Growth Cycles," American Economic Review, American Economic Association, vol. 88(3), pages 495-515, June ↩︎