Spillover Effects of Mass Layoffs

政府经常补贴甚至救助濒临破产的公司。例如，2008年和2009年，布什和奥巴马政府向通用汽车和克莱斯勒提供了总计超过500亿美元的资金，以防止工厂关闭。这些干预措施背后的主要经济理由是，工厂关闭或破产不仅会通过集聚效应伤害特定工厂的工人，而且会对该区域的其他工厂和工人造成多米诺骨牌效应，从而使失业人数成倍增加。在没有这种溢出效应的情况下，很难以经济效率为由，证明使用纳税人的钱进行补贴是合理的。该项目的目标是在德国劳动力市场的背景下追踪这种溢出效应。使用社会保障记录中的数据，我们使用工厂标识符来识别工厂关闭和大规模裁员。然后，我们的经验策略结合了差异法和事件研究法。我们将首先为每个处理区匹配一个控制区，在大规模就业破坏之前的几年里，控制区在人口结构、产业结构、工资和失业率方面与处理区相似，但没有受到冲击。为了确定大规模工作岗位破坏对该地区工资、就业和工人的影响，我们随后比较了治疗和对照地区在大规模工作岗位破坏后几年的结果。我们的分析试图回答5个问题：1)大规模的就业破坏是否会在当地劳动力市场引发更多的失业；如果是，在哪些行业？2)大规模的就业破坏是否会导致更高的失业率和更低的工资？3)它是否降低了同一地区的工人的就业概率，这些工人没有受到大规模裁员的直接影响，但在大规模裁员时已经就业或正在寻找工作？4)位于大规模裁员地区的公司的生产率会发生什么？5)一个地区的创新能力是如何受到影响的，哪些地区最容易失去竞争优势？我们的发现通过研究对该地区其他工人和公司的潜在负面溢出效应，为关于流离失所效应的文献做出了贡献。如果是这样的话，流离失所的社会成本可能比流离失所的私人成本大得多。我们的研究进一步为集聚效应的存在和程度提供了新的证据。因此，我们提出的研究对为什么城市和产业集群存在这个长期存在的问题做出了贡献。最后，我们的发现具有重要的政策含义。通过比较当地劳动力市场层面和个人层面的溢出效应，我们的研究将提供新的证据，证明国家对特定企业的补贴是否合理，不仅从受影响地区的角度来看，而且从整体经济的角度来看。

Governments often subsidise or even bail out firms on the verge of bankruptcy. For example, in 2008 and 2009 both the Bush and the Obama administration provided in total over US$50 billion to General Motors and Chrysler to prevent the shut-down of plants. The main economic rationale behind these interventions is that a plant closure or bankruptcy would, through agglomeration effects, not only harm workers in the particular plant, but create a domino effect on other plants and workers in the region, thereby multiplying job losses. In the absence of such spillover effects, it is difficult to justify the use of taxpayer's money for subsidies on economic efficiency grounds. The goal of this project is to trace such spillover effects in the context of the German labour market. Using data from the social security records, we use the plant identifier to identify plant closures and mass layoffs. Our empirical strategy then combines a difference-in-difference approach with an event-study approach. We will first match for each treatment district a control district that in years prior to the mass job destruction was similar to the treatment district in terms of the demographic structure, industry structure, wages and unemployment, but that was not hit by a shock. To identify the effect of mass job destruction on wages, employment and workers in the region, we then compare outcomes in treatment and control districts in the years after the mass job destruction. Our analysis seeks to answer 5 questions: 1) Does massive job destruction trigger additional job losses in the local labour market; and if so, in which sectors? 2) Does massive job destruction lead to higher unemployment and lower wages? 3) Does it lower employment probabilities of workers in the same region who were not directly affected by the mass layoff but employed or looking for a job at the time of the mass layoff? 4) What happens to productivity in firms that are located in the region experiencing the mass layoff? And 5) how is the innovative capacity of a region affected and which regions are most vulnerable to lose their competitive edge? Our findings contribute to the literature on displacement effects by studying potential negative spillover effects on other workers and firms in the region. If so, the social costs of displacement may be substantially larger than the private costs of displacement suggest. Our research further provides novel evidence on the existence and magnitude of agglomeration effects. As such, our proposed research contributes to the long-standing question why cities and industrial clusters exist. Finally, our findings have important policy implications. By contrasting spillover effects at the local labour market level with those at the individual level, our research will provide novel evidence whether state subsidies to particular firms can be justified not only from the viewpoint of the affected region, but also from the viewpoint of the economy as a whole.