Partisanship and Innovation: Exploring the Link between Partisanship and the Rate and Direction of Inventive Activity

该项目将研究科学家的政治取向及其在塑造发明活动的速度和方向方面的作用。虽然研究表明，政治取向影响个人决策，但我们对科学家的政治取向如何影响劳动力市场和合作选择知之甚少，这两个因素直接反映了创新成果的速度和方向。该项目的第一个组成部分将选民登记记录与个人专利发明者和学术研究人员联系起来，解决迄今为止阻碍对科学家政治倾向进行大规模和系统研究的数据限制问题。该项目的第二部分使用这些数据（1）提供关于跨知识领域、行业、企业和地理区域的科学家的政治取向变化的全面描述性统计;（2）通过评估由党派驱动的合作偏好的变化及其对创新成果的影响，评估科学家的政治取向对知识空间流动性的影响;以及（3）通过评估党派之争驱动的地理区域变化以及此类变化对创新成果的影响，评估科学家的政治取向对物理空间流动性的影响。为了将选民登记记录与个人专利发明人和学术研究人员进行匹配，我们使用来自多个来源的数据。我们从L2获得选民登记记录，L2是一家商业数据提供商，直接从各州获取选民信息。专利发明人的数据来自PatentsView，这是一个USPTO数据库，可以消除专利发明人姓名的歧义，学术研究人员的信息来自Dimensions，这是一个包含超过1.06亿篇学术出版物和12亿次引用的数据库。为了解决这些数据库之间有限重叠信息的主要匹配挑战，我们建立在机器学习和统计记录链接程序的最新进展基础上，这些程序通过识别匹配并通过随机观察这种匹配的概率对这些匹配进行加权来工作。为了对政治取向在创新成果中的作用进行强有力的分析，我们采用了差异中差异的前沿统计技术，旨在确定现场数据中的因果关系。该奖项反映了NSF的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

This project will study the political orientation of scientists and its role in shaping the rate and direction of inventive activity. While research has shown that political orientation influences individual decision-making, we know relatively little about how the political orientation of scientists may shape labor market and collaboration choices, two factors which reflect directly on both the rate and direction of innovation outcomes. The first component of this project addresses data limitations which have thus far prevented wide-scale and systematic study of the political orientation of scientists by linking voter registration records with individual patent inventors and academic researchers. The second component of this project uses these data to (1) provide comprehensive descriptive statistics about variation in political orientation of scientists across knowledge domains, industries, firms, and geographic regions; (2) evaluate the impact of scientists’ political orientation on mobility in knowledge space by evaluating changes in collaboration preferences driven by partisanship and its consequences for innovation outcomes; and (3) evaluate the impact of scientists’ political orientation on mobility in physical space by evaluating changes across geographical regions driven by partisanship and the consequences of such changes for innovation outcomes.To match voter registration records with individual patent inventors and academic researchers, we use data from multiple sources. We obtain voter registration records from L2, a commercial data provider who sources voter information directly from individual states. Data on patent inventors are collected from PatentsView, a USPTO database which disambiguates patent inventor names, and information on academic researchers are collected from Dimensions, a database containing information on more than 106 million academic publications and 1.2 billion citations. To address the primary matching challenge of limited overlapping information between these databases, we build on recent advancements in machine learning and statistical record-linking procedures that work by identifying matches and weighting these matches by the probability of observing such a match at random. To provide a robust analysis of the role of political orientation is innovation outcomes, we employ frontier difference-in-differences statistical techniques that aim to identify causal relationships in field data.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.