Globalisation and business cycle transmission

Michael Artis, Toshihiro Okubo

28 January 2009



Explorations of two-wave globalisation themes (e.g. Bairoch and Kozul-Wright, 1996; Baldwin and Martin, 1999; Williamson, 2002) have emphasised that openness in trade and finance are the principal common characteristics of the two waves. In some other ways – such as the scale of international migration and investment, the role of government and its policies, and the presence of international organisations and international cooperation and coordination in economic policies – the two waves are not the same.

Following Bairoch and Kozul-Wright (1996) and Baldwin and Martin (1999), the first wave of globalisation is defined as the period before World War I (i.e. 1870–1914). The bloc economy period (1915-1959) comprises the interwar period, which involves the Great Depression, World War II, and the subsequent recovery period. The second wave of globalisation is defined as the period after 1960. After World War I, some East European countries became independent. After World War II, many Asian countries won independence from the Imperial powers. The 1960s saw the independence of many more formerly colonial African and Asian countries and the initiation of the movement of international cooperation and liberalisation of trade and finance, establishing the current regime of international relations.

For many observers, the essence of globalisation is the participation of many individual countries in a world business cycle. In the era of the bloc economy, it may well have been the case that a common business cycle experience was experienced only by subsets of the world’s economies, reflecting the formation of political alliances and exclusive trade and currency areas.

Trade and financial openness have come to be regarded as positive indicators of business cycle transmission between economies. Our research (Artis and Okubo 2008) aims to discuss how the two waves of globalisation and the intervening period – which we term the period of the bloc economy – are reflected in business cycle transmission.
Cross-correlations in business cycles

In order to examine this proposition more closely, we have drawn on the long-run GDP data set assembled by Maddison (2003) as the basis for extracting the business cycle defined as a deviation cycle and identified by applying a Hodrick-Prescott filter to data for 19 developed countries. The cyclical deviations have been examined at first pass for their bilateral cross-correlations, dividing the sample into the three sub-periods described above.

An examination of the distribution of the bilateral cross-correlations (Figure 1) reveals that the average is highest for the second globalisation and lowest for the bloc economy. More marked perhaps is that the variance of the cross-correlations is biggest for the bloc economy period, which also displays a “twin peak” frequency, corresponding perhaps to the boost that some political arrangements gave to some groups of countries at the expense of others.

Figure 1. Bilateral cross-correlations

Hard clustering analysis

Subsequently we applied a hard clustering analysis to the data. Hard clustering analysis is associated with a graphical picture (a dendrogram) which shows how an “object” (a country in our case) can be associated with others in respect of some pre-selected characteristic. Here the most important finding is perhaps that the bloc economy period supports the best-defined clusters of countries and the second globalisation period the least well-defined. The first globalisation period falls between the two or three, though it has much less clear split than in the bloc economy. This feature fits well with a picture of globalisation that emphasises the all-embracing nature of the phenomenon, leading to fewer, and less well delineated, sub-global clusters.

McNemar test statistic

We study the relationship of specific (e.g. as defined by language, race and economic relations) groups of countries to others. We applied the McNemar test statistic à la Bovi (2005) to the data, comparing the coherence of groups of countries with one another. Globalisation should make it harder to find clear evidence of any difference in coherence between groups of countries, and this is what our data show. In essence, the first globalisation period never sees any coherent country groups based on race and language. However, the bloc economy period saw a big discrepancy between the Anglo-Saxon group and other European country groups. This might be regarded as being driven by exclusive and biased trade and capital flows in the bloc economy as well as exclusive international political relationships. It is a bit surprising that the second globalisation also sees some small coherent groups. This might be triggered by the solidarity through EU monetary systems and their own currency, i.e. Europeanisation.


Thus the analysis we have conducted so far appears to support the proposition that globalisation reduces business cycle differences between countries – and that this feature is more marked of the second (current) globalisation era than the first.

Of course there are many limitations that should be acknowledged: The data we use are annual in frequency, which inhibits precise dating of the cycle, we use business cycle synchronisation as a short hand for business cycle transmission, and we ignore other dimensions of the business cycle experience (business cycle amplitudes and so forth) which might be relevant. Perhaps more seriously the low (annual) frequency necessarily obscures the precise identification of cycle phases and impairs the separate identification of cycles from growth spurts.


Artis, Michael and Toshihiro Okubo, 2008. "Globalisation and Business Cycle Transmission," CEPR Discussion Paper 7041
Bairoch, P. and R.Kozul-Wright (1996) “Globalisation myths: Some historical reflections on integration, industrialization and growth in the world economy”, UNCTAD DP No.113.
Baldwin, R.E and P. Martin (1999) “Two waves of globalisation: superficial similarities, fundamental differences”, NBER Working Paper 6904.
Bovi, M (2005) “Globalisation vs. Europeanization: A Business Cycles Race Oxford Bulletin of Economics and Statistics, 67, 3: pp.331-345.
Maddison, A. (2003) The World Economy – Historical Statistics, OECD Development Centre, Paris.
Williamson, J.G. (2002) “Winners and losers over two centuries of globalisation” ,NBER Working 9161.



Topics:  Global economy

Tags:  globalisation, business cycles

Professor of Economics and Director, Manchester Regional Economics Centre, Institute for Political and Economic Governance, Manchester University

Professor of Economics, Faculty of Economics, Keio University