Hi-Stat Vox No. 14 (August 9, 2010)

Love Affairs with Chinese and Japanese Numbers

Harry X. Wu

Professor, Institute of Economic Research, Hitotsubashi University

Photo : Harry X. Wu

On 24 April 2010, the economics profession lost one of the world's most gifted scholars ever to quantitatively document economic performance over a very long period of time and across major countries in every continent of the world. Professor Angus Maddison died from complications arising from leukemia at the American Hospital in Paris. He was buried in the church graveyard in Chevincourt, a village near Compiègne where he had lived with wife Penny for the latter part of his life. He is also survived by his two sons, George and Charles, his daughter Lizzy, and five grandchildren.

Photo: Angus Maddison
ANGUS MADDISON (1926–2010)

Much has been written about Angus Maddison's life and work, which were devoted to the understanding and interpretation of the dynamics and sources of growth and development in human history, including a recent article by Derek Blades, Bart van Ark and myself.1 Here, I would like to refresh some of our memories about Angus Maddison's research activities on quantifying Chinese and Japanese economic performance in the long run as well as his contacts with Chinese and Japanese scholars, a topic that I immediately felt to be very appropriate when I was asked to write a memorial piece for the Newsletter* of Hitotsubashi University's Global COE Program, even though it is difficult to do it justice in a short article.

Angus's passion for Japan and China was no accident. It was aroused by his interest in the role of institutions in the economic development of the two Asian giants. During the first half of his time in Groningen (from 1978 to the late 1980s), Angus continued his broadly-based research on growth and development. Following several earlier books he had written on this topic, he published two major monographs in 1982 and 1991 and a seminal article on growth accounting in 1987 in the Journal of Economic Literature. These writings explained his understanding of the dynamics and sources of growth and development. His characterization of development in terms of secular phases rather than the Kondratieff notion of long waves led to a more gradualist interpretation of the diffusion of technology and innovation, and a greater emphasis on “system shocks.” These shocks were in part historical accidents, but their impact was reinforced by changes in expectations, and fashions in economic policy. These views, based on strong quantitative support, challenged scholars who had advanced the notion of a more abrupt “industrial revolution” in the late 18th century, or traditional development economists who supported the Rostovian view of the need for an economic “take-off” to generate growth. Angus was also an early advocate of the need to take economic institutions much more seriously by developing a model that distinguished between proximate sources of growth (directly measurable economic inputs, such as labour, physical and human capital and land) and ultimate sources (institutional, political, social and cultural). He believed that the complex interaction of these proximate and ultimate sources allowed a multipolar development in Europe in the Middle Ages.

I remember his last trip to China to celebrate the publication in Chinese of The World Economy: A Millennial Perspective. He was accompanied by me to visit Hangzhou, the capital of China's Southern Song (1127–1279), then the richest country in the world according to his estimates. He was warned about the potential danger of the highway drive and that it would be a tiring day trip but he was still keen to go. Filled with excitement at stepping on the land of Hangzhou, he then realized that it was hard to find any trace of its past prosperity. He remarked that wrong policies or mistakes by the ruling elite could easily wipe out a country's accumulated wealth for centuries – a big question hanging there that had motivated his long pursuit for the underlying explanations of economic growth.

Maddison also returned to the topic of measuring economic performance beyond the Western world, but began to place it more strongly in an international comparative perspective. In the early 1980s he criticized Paul Bairoch's view that the industrialization of the Western European countries took place at an average standard of living below that of contemporary less developed countries, which was very different from the conjectures by David Landes and Simon Kuznets who believed that by that time Western Europe was already much richer than the rest of the world. To Angus Maddison these remarkably different quantitative conclusions had very important analytical implications: “If Bairoch is right, then much of the backwardness of the third world presumably has to be explained by colonial exploitation, and much less of Europe's advantage can be due to scientific precocity, centuries of slow accumulation, and organizational and financial prosperity” (Maddison, 1983). Although he used the existing evidence at the time to show that Bairoch and his epigoni probably overstated the contemporary income gap and understated per capita income growth in the developing world, it motivated him to pursue the topic of international comparisons more intensively. Following his later work on China (see below), he returned to the topic of comparative performance between East and West at the time of the industrial revolution, when other scholars, notably Kenneth Pomeranz, argued that China stayed at a much higher level of development until the end of the 18th century than Europe – a viewpoint which Maddison strongly criticized on the basis of his reconstruction of China's macroeconomic accounts back to the year 960.

From the early 1960s to the early 1970s, he worked mainly on problems of economic development at the OECD Development Centre and as an economic advisor for the Twentieth Century Fund and the Harvard Advisory Service. The major aim of his work in this field was to understand why the rest of the world is much poorer than Western countries and to distinguish different types of non-Western experience. He realized that “differences in the level of development in the non-Western countries are very wide and there is great heterogeneity in their institutional heritage.” In interpreting the reasons for “backwardness,” he began to explore the role of colonialism, indigenous social forces, institutions, property rights, religion, and ideology by studying in depth some major country cases. Japan as the first of the seven such countries of “the third world” entered his research agenda. Although China was the last case of his series of country studies and his work on China was published almost thirty years later (1998) than his work on Japan (1969), it had since never been out of the focus of his research life. To Angus, a comparison between post-Meiji Restoration Japan's positive response to and quick catch-up with the West and China's catastrophic experience during the same period was a good example for understanding the role of institutions and government policy in economic development.

Angus made his first trip to Japan in 1961 when he met Kazushi Ohkawa and several other scholars at the Institute of Economic Research (IER) of Hitotsubashi University. This was the time when Ohkawa was starting to publish 13 volumes on Japanese quantitative economic history. Throughout the rest of his life, Angus maintained a very high view of the quantitative research under the leadership of Ohkawa and the research environment at the IER. During that trip, Angus was also impressed by what he observed in Japanese society. Later, recounting his experience of his 1965 trip to Japan, he wrote: “I was struck by the strong discipline and an organization that operated like clockwork…[and] when I visited the Sony radio factory, I found the foremen had PhDs and all the operatives had high school education” (Maddison, 1994).

Angus's contacts with a later generation of scholars at the IER began in the late 1980s when he was offered a visiting scholarship by the Japan Society for the Promotion of Science in 1989 and was hosted by Konosuke Odaka and Osamu Saito. As recalled by Odaka, Angus, from his first impression, was very much like Ohkawa in that he was very enthusiastic about any topic related to economic development and economic history, and could be easily engaged in an interesting and seemingly never-ending discussion. During his stay at the IER (about six weeks), Angus spent most of his time reading almost everything available in western languages on Japanese economic history as well as talking to the people at the IER, including young researchers. Angus Maddison paid another major visit to the IER in 1998 during the IER's first COE program on Asian Historical Statistics under the leadership of Odaka. During his stay, he engaged Saito in some intensive discussions on Japanese economic history and especially demographic history, which apparently benefited his research on Japan's development as part of his then new book, The World Economy: A Millennial Perspective, published in 2001.

Angus Maddison also paid other short visits, during which he gave seminars, acted as an evaluator of the IER's first COE program, and promoted exchanges of young researchers between the IER and the Groningen Growth and Development Centre (GGDC). In fact, it was he who introduced me to Odaka, which began my engagement with the IER. Angus's last visit to Japan in September-October 2007 – perhaps his very last trip to Asia – was to attend an international symposium at the IER and, more importantly, to receive an honorary doctorate from Hitotsubashi University for his great contributions to scholarship and the economics profession.

The debate with Bairoch and his followers points to the importance of quantitative assessment of China's long-run economic performance. This is important not only for China but also for positioning other economies in the world. In fact, in the debate, he compared Japan's Tokugawa period with China rather than with Western Europe as other studies did. While like many researchers he attempted to explain the divergences between countries, Angus paid more attention, and devoted a great deal of effort, to the understanding of important convergences in the past, in most of which China served as an important reference, such as Europe's rise from its nadir to overtake China, the Japanese catch-up with China in Tokugawa times, and the post-war resurgence of Asia, especially China, India, and the so-called tigers which have narrowed their degree of backwardness substantially with the advanced countries.

Angus's early work on China relied mainly on quantitative research by Dwight Perkins (1975), Ta-chung Liu and Kung-chia Yeh (1965), and studies commissioned by the U.S. Congress Joint Economic Committee (1978), some of which appeared in his discussion of the debate between Bairoch and Landes-Kuznets. After the debate, perhaps since the early 1990s, he substantially extended his investigation of Chinese economic history, exploring the literature on quantitative assessments of China. New research by himself and in collaboration with others eventually led to his book, Chinese Economic Performance in the Long Run (1998). This was also a period in which he began to work and discuss with Chinese economists. I may be the only one who had since spent a great deal of time working with him on the reconstruction of China's national accounts, on the development of an alternative industrial output index for China, and on the Chinese translation of his three books, including a revised and updated version of his China book in 2007.

Angus immediately contacted me when my article on the estimation of China's GDP in the central planning period appeared in The Review of Income and Wealth in 1993. We then started to communicate through fax on how to understand the problems in China's official statistics due to both methodological and institutional deficiencies, and how to obtain better estimates for China's real GDP. Our discussion deepened after we first met at an economics conference held at the University of New England, Australia, in 1995, possibly arranged by Prasada Rao. The idea began to emerge that it might be possible to use the physical output of major commodities to bypass the price distortions in the central planning period – a problem that I did not tackle in my 1993 paper. In early 1996, I spent a couple of months working with him on reconstructing China's industrial GDP, first at his home in Chevincourt and then at the GGDC. He discussed everything with me in detail, from the selection of commodities to benchmark problems and the weighting and aggregating procedures at various levels. For each round of the estimates, he always checked the results carefully and discussed them with me almost commodity by commodity and industry by industry. My preliminary empirical results lent strong support to his hypothesis that the official industrial statistics exaggerated the growth rate, while underestimating the level, of national output. The revisions and updates that followed have repeatedly confirmed these findings.

Meanwhile, Angus also conducted major data work on China in three areas, namely, the reconstruction of China's agricultural output accounts for 1975, 1987 and 1994 using output quantities, the estimation of China's agricultural production PPP using the US as the benchmark for 1987, and the reconstruction of the output of China's “non-material” services. Thus, together with my results on manufacturing, mining and utilities, he eventually obtained a new set of sector-level GDP estimates for China. Needless to say, these efforts provided a consistent quantitative basis for his China book, Chinese Economic Performance in the Long Run (1998), the last but most comprehensive country case study in his life.

After conducting such serious work on China's economic history, he wrote: “My judgment on the contours and chronology of dynastic China's development (i.e. the rise in per capita income in the Sung and its stagnation from the fourteenth to the mid nineteenth century) is not unlike that of Mark Elvin, R. M. Hartwell, Eric Jones and Justin Lin. However, they do not attempt macro-quantification, and their qualitative judgment probably implies a bigger leap in the Sung than I find. Some of them suggest that Sung China was trembling on the verge of an industrial revolution, which seems exaggerated. Some tend to overstate the degree of stagnation and the decline in ‘creativity' from 1300 to 1850. In this period, China managed (with some interruptions) to sustain per capita income levels, whilst increasing its population more than fourfold (compared with less than threefold in Europe). ‘Extensive’ growth on this scale is not the same as stagnation.”2

Angus's passion for China continued till the very late period of his life. He completely revised his China book in 2007 when his health was already beginning to decline. In fact, after his China book was published in 1998, he continued working with me on at least three major revisions of my industrial index and discussing data issues for updating his estimates of the performance of other sectors. He maintained a good friendship with several Chinese scholars, including Angang Hu, Bozhong Li, Justin Lin, Debin Ma, Xin Meng and Xiaolu Wang. At different stages of his research on China he also engaged in extensive discussions with Chinese economists such as Ruoen Ren, Xianchun Xu and Ximing Yue on the problems of measuring price change, service output and employment in China.

Just this February, only two months before he passed away, we still discussed intensively over the telephone whether a new approach I proposed should be used to reexamine China's post-reform growth performance and compare the results with those based on the old approach adopted in our earlier joint work (Maddison and Wu, 2008). In March, I sent him the comparison via email and at the same time asked him about his estimates of hours worked in one of his early articles published in 1954. But unfortunately he was too weak to continue our discussion or answer my questions.

Angus was my mentor who played a very important role in my scholarly pursuits. He was also a colleague and friend, all at the same time. I remember the great moments with him at his home with his family in France, at conferences, or wherever else in the world we would meet. Angus was exceptionally gifted in combining work and pleasure in a very natural way. Angus Maddison will be missed. But his legacy will stay, and many will continue the work in his spirit.


  1. The article will be published in The Review of Income and Wealth in September, 2010. Some of its contents are used in the present paper without specifying the source.
  2. Angus Maddison, “Research Objectives and Results, 1952-2002,” available on his website maintained by the Groningen Growth and Development Centre.


Maddison, Angus (1983), “A Comparison of Levels of GDP Per Capita in Developed and Developing Countries 1700-1980”, Journal of Economic History, Vol. 41, No. 1, March, pp. 27-41.

Maddison, Angus (1987), “Growth and Slowdown in Advanced Capitalist Economies: Techniques of Quantitative Assessment”, Journal of Economic Literature, Vol. XXV, June, pp. 649-698.

Maddison, Angus (1994), “Confessions of a Chiffrephile”, Banca Nazionale del Lavoro Quarterly Review, No. 189, June, pp. 123-165 (also available on his website maintained by the Groningen Growth and Development Centre).

Maddison, Angus (1998), Chinese Economic Performance in the Long Run, OECD, Paris; an electronic version is available online at: www.ggdc.net/Maddison.

Maddison, Angus and Harry X. Wu (2008), “Measuring China's Economic Performance”, World Economics, Vol. 9, No. 2, pp. 13-44.

*This column originally appeared in the July 2010 issue of Hitotsubashi University's Global COE Hi-Stat Newsletter.