Comments and Discussion

Barry P. Bosworth:

In this paper, Gary Jefferson, Albert Hu, and Jian Su discuss a wide range of issues that arise in evaluations of China's recent growth performance and future prospects. Indeed, the range is so broad that no one of these issues can be thoroughly examined within the context of a single paper. I will therefore focus on just a few.

The first part of the paper examines China's development on two margins: the labor productivity gap between China's coastal region and a global frontier, represented by a combination of manufacturing industries in Japan and the United States; and the relative labor productivity gaps among China's four major regions. With respect to the first issue, the authors' basic conclusion is that there has been considerable catch-up to the international frontier, but the gap is still very large. No one should quarrel with this conclusion, since China's economy has obviously been growing far more rapidly than the rest of the world. More controversially, the authors find that regional disparities in manufacturing productivity within China narrowed between 1995 and 2002, except for the northeast (table 2), where the gap with the leading coastal region has widened.

The comparisons are based on twenty-seven industries within manufacturing. International comparisons of levels of productivity at the individual industry level are extraordinarily difficult, especially for China, which has only recently agreed to participate in the World Bank—led International Comparison Program and still lacks comprehensive purchasing-power-parity estimates of GDP. Most international comparisons are limited to aggregate GDP, because the methodology needed to compute industry-specific conversion factors is very underdeveloped. An estimate of the purchasing-power-parity exchange rate has been cobbled together for China, but its manufacturing industries reflect a mixture of exporters, for which the commercial exchange rate would be applicable, and firms that produce for the domestic market. Furthermore, at even the most detailed classification grouping, firms produce very different products in China, Japan, and the United States. Thus, the international comparisons of productivity in the authors' table 2 and figure 1 should be handled with care. Within China, the labor market and the transportation system are not integrated across regions. This makes it hard to compare levels of value added per worker across regions, since wages for equivalently skilled labor may differ. However, their relative growth rates should still reflect any convergence that has occurred.

In tables 3 and 4 the authors shift to a data set that reports on both national and regional economic performance at a broad sectoral level: agriculture, industry, and services. Here another data issue emerges: the authors argue that, in recent years, as many as 80 million workers nationwide may have been misreported in the 2005 household surveys as working in agriculture when they are actually employed in industry or services. Such misclassifications could dramatically affect our understanding of the relative performance of Chinese agriculture and industry. The issue has been raised in several earlier papers, but the extent of the miscounting remains contentious. The authors deal with the problem by assuming that the misclassification of workers between agriculture and industry is uniform across regions, implying no effect on the regional comparisons of productivity performance presented in table 4. This is not very helpful.

The results of this broad sectoral analysis appear to conflict with the authors' earlier conclusions based on manufacturing data, since the northeast now stands out with gains in labor productivity in industry between 1995 and 2004 that raise it above that of the coastal region. The industrial sector does include some important nonmanufacturing industries, where productivity growth may have been unusually rapid, and these industries are largely concentrated in the northeast. Even so, can it be that the region's industrial labor productivity grew at an 18 percent annual rate? The northeast also made significant gains on the coastal region in services. The other regions show modest convergence in industry and a mixed record in agriculture and services.

Most of the subsequent analysis is based on an intriguing micro panel data set of more than 20,000 large and medium-size...