Is Government Spending a Free Lunch? -- Evidence from China

Most empirical studies based on U.S. data suggest that the fiscal multiplier is less than 1 (e.g., Barro and Redlick, 2011). However, Keynes argued that the multiplier would be the largest when markets have failed to the greatest extent in coordinating economic activities (such as during the Great Depression with rampant unemployment and low capacity utilization). As a large developing country with high household saving rates, a large pool of rural labor force, and a wide range of market failures, China offers a unique opportunity to test the Keynesian notion that government expenditures (even as a pure waste of aggregate resources) can have a fiscal multiplier larger than 1 on aggregate income. Perhaps even more exceptional is China’s extensive use of government spending as a major policy tool to stimulate the economy over the past three decades. Based on both aggregate time-series data and panel data from 29 Chinese provinces, we find that the fiscal multiplier in China is larger than 2. We provide a theoretical model with market failures and Monte Carlo analysis to rationalize our empirical findings. Specifically, we build a model that can generate the same multiplier and business cycles observed in China and use the model as a data-generating process to gauge whether structural vector autoregressions can yield consistent estimates of the theoretical multiplier in short samples. Our analysis supports the large multiplier found in China but also suggests that government spending may not necessarily be a free lunch despite the large multiplier. ∗The views expressed are those of the individual authors and do not necessarily reflect offi cial positions of the Federal Reserve Bank of St. Louis, the Federal Reserve System, or the Board of Governors. We thank Judy Ahlers for editorial assistance and the usual disclaimer applies. Xin Wang, School of Economics and Management at Tsinghua University. Yi Wen (corresponding author), School of Economics and Management at Tsinghua University; and Research Department, Federal Reserve Bank of St. Louis, P.O. Box 442, St. Louis, MO 63166-0442, United States. Fax: +1 314 444 8731. E-mail address: [email protected].