The Impact of Fiscal Stimulus on Employment: Evidence from China's Four-Trillion RMB Package

Despite the large body of work re-estimating the impact of fiscal stimulus on employment in developed countries by using cross-sectional analysis, little evidence exists for China in this fashion. Adopting a sample of China’s 318 prefecture-level data between 2008 and 2010, I implement Bartik instrumental variable method inspired by China’s “four-trillion package” to estimate the causal effect of local fiscal spending on employment in China during recession. My findings reveal that every 100,000 CNY creates 2.2 to 3.4 jobs, implying 29,411 to 45,454 CNY (4,237 to 6,549 in USD) cost per job-year saved. These costs are lower than those estimated for Australia, Brazil and Germany. I further investigate the mechanism and find that employment growth primarily concentrated in the private sector is driven by the influx of migrants. However, the overall benefit-cost ratio is low, with no evidence of long-term effects or spillover effects to neighboring cities.尽管已有大量研究通过截面分析重新估算了财政刺激政策对发达国家就业的影响，但采用此类方法针对中国开展的相关研究却寥寥无几。本文选取2008至2010年中国318个地级市的面板数据作为研究样本，借鉴中国“四万亿”刺激计划构建Bartik工具变量法（Bartik instrumental variable method），以此估算经济衰退时期中国地方财政支出对就业的因果效应。研究结果显示，每投入10万元人民币可创造2.2至3.4个就业岗位，即每保住1个岗位年需花费29411至45454元人民币（折合4237至6549美元）。该成本低于澳大利亚、巴西与德国的相关估算结果。本文进一步探究了其作用机制，发现就业增长主要集中于私营部门，且这一增长由流动人口的涌入所驱动。但整体而言，该政策的收益成本比偏低，且未发现长期效应或对邻近城市产生外溢效应的相关证据。