【正文】 virtually no partial correlation with our index of institutional quality, the “rule of law” . Hence, the evidence on the fundamentals is mixed. Although the productivity of a country’s exports is determined in part by its overall productive capacity and its human capital endowment, idiosyncratic characteristics also matter. To take one telling example, a country like Bangladesh with a much similar set of relative factor endowments (abundant in labor, and scarce in human and physical capital) has an EXPY that is roughly 50 percent lower than China’s. It is clear that whatever these idiosyncratic features might be (about which I turn, this has apparently been an important driver of its recent economic growth as I will show in Section III. It is also interesting to pare the progress over time of China’s EXPY with that of some of its important petitors. This is done in Figure 6, which shows the trends for EXPY in China, India, Hong Kong and South Korea. China has experienced the most rapid rate of growth in the sophistication of its exports since 1992. Although China still lags behind South Korea and Hong Kong, the difference in EXPY has steadily closed over time. Moreover, the gap between China and India has actually increased over the past decade2 . The findings here contrast with those of Mayer and Wood (2021), who argue that China’s export performance in skillintensive products is below the level that would be expected on the basis of the country’s factor endowments. Their analysis is based on a crossnational benchmark derived from regressing the ratio of skillintensive to labor intensive exports on a measure of skill per worker in that country. Mayer and Wood report that China’s actual share of skillintensive exports within manufactures in 1990 stood at 33 percent, pared to a predicted ratio of 40 percent (Mayer and Wood, 2021, Table 5). One plausible reason for the difference in the results is that Mayer and Wood’s method relies on a binary classification of goods (skillintensive or not). Furthermore, the classification is carried out at a much more aggregate level than here. (Remember that we calculate EXPY using the information from more than 5000 distinct products.) In practice, it is mon for skillintensive ponents of some “goods” to be done in one country and the less skill intensive ponents in others, as Mayer and Wood note (2021, ). An analysis at the 6digit level of disaggregation is more likely to capture these distinctions. It Is Not How Much but What You Export that Matters How do we know that the productivity level of a country’s exports (measured by EXPY) matters to economic performance? It turns out that there is a robust relationship between the initial level of a country’s EXPY and the subsequent rate of economic growth experienced by that country. Figure 7 is the relevant scatter plot: it shows the relationship between EXPY in 1992 and growth over the1992–2021 period, holding initial levels of ine constant. This is a positive and statistically significant relationship (at the 95 percent confidence level). The estimated coefficient implies that a doubling of the productivity level of a country’s exports results in an increase in its overall percapita GDP growth of approximately 6 percent. Therefore, had China exported only those goods that countries at China’s level of ine tend to export, its growth rate would have been significantly lower. As the chart shows, a high level of EXPY does not fully account for China’s growth performance over this period (the other significant outlier is