Collaborative Research: Public R&D and the Rate and Direction of Inventive Activity

政府？由于基础研究的时间跨度较长，而且基础研究成果难以产生经济回报，私营部门很可能对研发投资不足，因此，政府在资助研发方面的作用是有道理的。该项目提供了有价值的见解，以了解自2010年以来联邦研发支出的下降是否损害了美国的竞争力。 该项目利用了一种新的长期技术轨迹的定量测量方法，即专利影响力，以确定对发明活动的公共资助是否支持私营部门所回避的研究。 该项目调查了联邦机构支持研究的变化对未来科学和技术突破可能产生的影响的长期影响。项目？的研究战略提出了一个解决难题的衡量基础研究经费对发明的长期影响，并提供政策制定者一种方法，以更好地评估类型的计划，可能会导致长期的科学和技术突破。对专利和学术研究的引用，其资金来源，以及发明发生的地点（即，大学、国家实验室或私人公司）。然后，这些属性与专利的影响力有关。使用影响力测量是计算密集型的，并且仅在最近考虑到计算技术的进步才可行。这项研究是第一次提供一个全面的，定量评估的长期影响的公共支持的研究和发展的技术轨迹，衡量专利。对2001年至2004年期间授予和申请的专利进行的一项试点研究表明，虽然专利的平均水平相似，但某些类型的政府资助比私营部门资助的专利更有可能导致突破性发明的产生。分位数回归分析的使用适应了研究和开发回报的偏斜性质，并允许研究资金对突破性发明的影响。该项目将计算所有数字化专利的影响力和收到的影响力。

The government?s role in sponsoring R&D has been justified because the private sector is likely to underinvest in R&D due to the long horizons of basic research and the difficulties in generating financial returns to basic research outcomes. This project provides valuable insights to understand whether the decline in Federal R&D spending since 2010 hurts U.S. competitiveness. This project leverages a new quantitative measure of long term technological trajectories, patent influence, to determine if public funding of inventive activity supports research that would be eschewed by the private sector. The project investigates the long term implications of changes to federal agency support of research on the likely impact of future scientific and technological breakthroughs. The project?s research strategy presents a solution to the difficult problem of measuring the long term effects of basic research funding on invention and offers policy makers a way to better assess the types of programs that are likely to lead to long term scientific and technological breakthroughs.The public contribution to the knowledge embedded in a patent is measured through the patent?s citations to patents and academic research, its funding sources, and where the invention occurred (i.e., university, a national lab or a private firm). These attributes are then related to patent influence. The use of the influence measure is computationally intensive and only recently feasible given advances in computing technology. This research is the first to provide a comprehensive, quantitative assessment of the long term influence of public support of research and development on technological trajectories, as measured by patents. A pilot study of patents granted and applied for between 2001 and 2004 suggests that while average patents are similar, some types of government funding are more likely to lead to the generation of breakthrough inventions than patents funded by the private sector. The use of quantile regression analysis accommodates the skewed nature of the returns to research and development and allows studying funding effects on breakthrough inventions. The project will calculate both influence and received influence measures for all digitized patents.