How Efficient Has Been China's Investment? Empirical Evidence from National and Provincial Data

China's investment has been growing very strongly. The share of gross capital formation in GDP in China has also been higher than in other East Asian economies during their high growth period in the 1970s-80s. Many commentators have argued that such high rates of investment growth have been driven by irrational incentives and have been largely inefficient, will cause a build up of non-performing loans in the banking system, and will also lead to over-capacity and deflation. Others, however, have argued that China is still capital scarce, returns to capital are high, and therefore high rates of investment are both desirable and sustainable. This paper attempts to shed new light on the debate. We analyse both the allocative efficiency and the dynamic efficiency of China's spending on capital. The allocative efficiency measures the extent to which resources have been invested in places where potential rates of return on capital are high. The potential rates of return can be calculated as the marginal products of capital derived from an aggregate production function. The dynamic efficiency measures the extent to which the capital-output ratio exceeds the optimal level. The optimal level of the capital stock is determined by a rate of investment, at which level the Chinese residents at the present enjoy the highest level of consumption without sacrificing the level of consumption in the future. We first construct China's total capital stock at national and provincial levels, estimate the Cobb-Douglas and CES production functions, and compute the marginal products of capital. Assuming that the Chinese economy was operating on the production frontier, the marginal products of capital at the aggregate level have been relatively high in the past two decades, and have not shown clear signs of decline in recent years. We find that China's marginal product of capital compares favourably with those observed in the major industrialised economies and in the Asia region. We also find that the marginal products of capital have been higher in the coastal areas than in the less developed areas of western and central China, but the marginal products of infrastructure capital have been higher in the inland areas than in the coastal areas. These results are robust to different assumptions made in constructing the data of capital stock.