Economic Policy Uncertainty: Public Data and Research

摘要提案编号：SES - 1324257机构：美国国家经济研究局NSF计划：经济学首席研究员：布鲁姆，尼古拉斯标题：经济政策的不确定性：公共数据和研究美国和欧洲的深度衰退和缓慢复苏，伴随着一些国家的长期财政危机，引发了对经济政策不确定性的担忧。国际货币基金组织、联邦储备委员会和欧洲中央银行等许多机构都对政策不确定性阻碍许多国家的经济发展表示担忧。Baker、Bloom和Davis（2013）开发了一种方法来衡量政策不确定性，该方法基于：（a）专业预测者对未来政府支出和通货膨胀的分歧，（B）关于政策不确定性的报纸文章数量，以及（c）税法规定的预定修订。其政策不确定性措施的数据可在www.policyuncertainty.com上免费提供，该网站自2012年1月推出以来已拥有30 000多名独立访问者。金融公司、工业公司、政策制定者和学术研究人员现在使用这些指标来评估几个国家的政策不确定性。这个项目将改善公共数据集，并在三个方向上推进研究：A）改善数据，使用更先进的计算机搜索方法和对报纸文章的人工审核来解决测量问题。特别是，通过使用人类阅读约10，000篇报纸文章作为自动化方法的培训工具，该提案将开发更丰富，更准确的基于新闻的政策不确定性指标。B）扩展数据：（i）随着时间的推移，包括美国回到1900年;（ii）在国家之间，增加了巴西、中国、印度、日本、墨西哥和俄罗斯;（iii）跨部门制定货币、财政和监管政策不确定性措施。 自2007-08年全球金融危机以来，许多国家的政策不确定性似乎有所增加。 我们希望以系统的方式量化这一现象。C）研究政策不确定性的原因和影响：标准VAR方法显示，经济政策不确定性（EPU）越高，GDP增长越慢。虽然提示，这一发现并没有建立一个因果关系的影响EPU冲击？例如，对经济放缓的担忧本身就可能产生更大的政策不确定性。这项研究将使用冲击和自然实验？比如朝鲜的政策变化，9/11恐怖袭击，政治周期和战争？以帮助确定政策不确定性的因果影响。它还将调查EPU在公司和行业层面的影响，利用跨部门的差异暴露于EPU。这一项目将使学术界、商业界和决策者受益。在学术界，政策不确定性的原因和影响是一个活跃的研究领域。然而，这一领域研究进展的两个障碍是数据和识别。该项目将收集和公开提供数据，并通过利用大的冲击和公司一级在政策不确定性方面的差异来查明因果关系。

AbstractProposal No: SES - 1324257 Institution: National Bureau of Economic Research IncNSF Program: ECONOMICSPrincipal Investigator: Bloom, NicholasTitle: Economic Policy Uncertainty: Public Data and ResearchDeep recessions and slow recoveries in the United States and Europe, accompanied by protracted fiscal crises in some countries, have spurred concerns about economic policy uncertainty. The IMF, Federal Reserve Board and European Central Bank, among many others, have raised concerns that policy uncertainty is impeding economic progress in many countries. Baker, Bloom and Davis (2013) develop a methodology to measure policy uncertainty based on: (a) disagreement amongst professional forecasters about future government spending and inflation, (b) counts of newspaper articles about policy uncertainty, and (c) scheduled expirations of tax code provisions. Data for their policy uncertainty measures are freely provided at www.policyuncertainty.com, which has more than 30,000 unique visitors since launching in January 2012. Financial firms, industrial companies, policy makers, and academic researchers now use these measures to assess policy uncertainty in several countries.This project would improve the public dataset and advance research in three directions:A)Improve the data, using more advanced computer search methods and human audits of newspaper articles to address measurement concerns. In particular, by using human readings of about 10,000 newspaper articles as a training tool for automated methods, the proposal will develop richer and more accurate news-based metrics of policy uncertainty. B)Extend the data: (i) over time, including back to 1900 for the United States; (ii) across countries, adding Brazil, China, India, Japan, Mexico and Russia; and (iii) across sectors to develop monetary, fiscal and regulatory policy uncertainty measures. Many countries appear to have heightened policy uncertainty levels since the global financial crisis of 2007-08. We would like to quantify this phenomenon in a systematic manner.C)Researching the causes and effects of policy uncertainty: Standard VAR methods show that higher economic policy uncertainty (EPU) foreshadows slower GDP growth. While suggestive, this finding does not establish a causal impact of EPU shocks ? for example, fears of an economic slowdown can itself generate greater policy uncertainty. This research will use shocks and natural experiments ? like the North Korean policy shifts, 9/11 terrorist attacks, the political cycle and wars ? to help identify the causal effects of policy uncertainty. It will also investigate EPU effects at a firm and industry level, exploiting cross-sectional differences in exposure to EPU. This project should benefit academic and business users and policymakers. On the academic front, the causes and effects of policy uncertainty is an active research area. Two barriers to research progress in this area, however, are data and identification. This project will collect and publicly provide data, and seek to identify causation by exploiting large shocks and firm-level differences in exposure to policy uncertainty.