Events

Markush structures are molecular skeletons that contain not only specific atoms but also include one or several placeholders each representing a broad set of chemical (sub)structures. They are used by pharmaceutical companies to claim a large class of compounds without the necessity of writing out every fully defined single chemical entity in a patent application. (For instance, the Markush structures claimed within patent EP 0810 209 contain a total of 10^16 different compounds resulting from all possible permutations within the Markush structures.) After summarizing the ongoing policy debate regarding the use of Markush structures in patents, this paper provides first quantitative evidence regarding the use Markush structures in the pharmaceutical industry and their effects on important outcomes in the patent prosecution process.

Contact: Marina Chugunova

In this paper, we argue that economic societies, established during the eighteenth-century, contributed to industrialization through the diffusion of new ideas generated during the Scientific Revolution in Europe. Local societies functioned as catalyst for the translation of scientific knowledge into useful knowledge and the diffusion to interested parties. We test this hypothesis by combining information on more than 3,300 society members from the membership lists of all active economic societies in the German lands with several measures of innovation and upper-tail human capital. We find a robust positive relationship between the local member density and the number of valuable patents, exhibitors at world fairs, and highly-skilled mechanical workers. We further show that grid-cell pairs with members from the same society show a higher technological similarity. We interpret this as evidence that economic societies generated information networks which fostered spatial knowledge diffusion and shaped the geography of innovation.

Contact: David Heller

How much lower are the career earnings of workers who face large declines in demand for their occupations, compared to similar workers who do not? To answer this question we combine forecasts on occupational employment changes, measures of realized occupational decline and technological replacement, and administrative panel data on the population of Swedish workers, with a highly disaggregated initial occupational classification.  We find that compared to similar workers, those facing occupational decline lost about 2-5 percent of mean cumulative earnings from 1986-2013, with workers at the bottom of their occupation’s initial earnings distribution suffering substantially larger losses.  These earnings losses are partly accounted for by reduced employment and increased time spent in unemployment and retraining.

Contact Person: Michael E. Rose

In jobs where the cost of effort exhibits increasing differences in separate tasks, automation increases productivity by directly eliminating the automated tasks and indirectly by reducing the marginal cost of non-automated tasks. Here, I report a field experiment rotating supermarket cashiers between conventional (where they scanned and collected payment) and scan-only checkouts. Consistent with increasing differences in separate tasks, at conventional checkouts, cashiers who scanned faster collected payments more slowly. At scan-only checkouts, cashiers scanned 10 percent faster, consistent with lower marginal cost of effort in the non-automated task. The faster scanning was not due to learning, less task-witching, or differential shirking.

Contact Person: Lucy Xiaolu Wang

Can artificial intelligence (AI) technologies complement or substitute for human managers in creating greater value for organizations? We argue that managers of the transformational leadership style with greater social skills benefit more from the assistance of AI than do managers of the transactional leadership style with fewer social skills. We provide causal evidence through a field experiment in a fintech firm whose managers provide training to employees on the calls made to collect overdue loans. We randomly assign employees to be trained by one of the following five options: an AI-bot, a transformational leadership-style manager, a transactional leadership-style manager, and both managers assisted by the AI-bot. We find that employees trained by each AI-assisted manager achieved higher performance, by collecting more payments, than did those trained by the manager alone, suggesting that both managers can gain from AI assistance. More interestingly, employees trained by the AI-assisted transformational manager outperformed both those trained by the AI-assisted transactional manager and by the AI-bot alone, indicating complementarity between transformational leadership and AI assistance. By contrast, the employees trained by the AI-assisted transactional manager underperformed those trained by the AI-bot alone, suggesting that the transactional manager is still at risk of being replaced by the AI-bot. These divergent performance outcomes occur because employees learn more from the AI-assisted transformational manager than from the AI-assisted transactional manager. Our findings suggest that, compared with transactional leadership, transformational leadership enable firms to obtain greater economic returns from their investment in deploying AI to manage employees.

Contact Person: Marina Chugunova
 

Innovation under regulatory uncertainty and the role of expectations: Evidence from the U.S. drone market
Presenter: Virginia Herbst (TUM)
Discussant: Lucy Xialou Wang (Max Planck Institute for Innovation and Competition)

There and back again: Disruptive transitions in dyadic role relationships
Presenter: Maren Mickeler (LMU ISTO)
Discussant: Daniel Obermeier (TUM)

The rapid spread of the COVID-19 pandemic and the subsequent counter-measures, such as school closures, shift to working from home, and social distancing are disrupting economic activity around the world.  As with other major economic shocks, there are winners and losers, leading to increased inequality across certain groups.  In this project, we investigate the effects of COVID-19 on the gender gap in productivity among academics. First, we track the patterns of publications and working paper series submissions of male and female authors before and after COVID-19, performing a simple difference-in-difference analysis to see whether the trends diverge after COVID-19. Preliminary findings from economics suggest that female economists experienced significant declines in submissions in the early months of the pandemic, relative to the mean.  Second, we conduct a broad survey of academics across various disciplines to collect more nuanced data on the respondents’ circumstances, such as spouse employment, the number and age of children, and time use. We find that female academics, particularly those who have children, report a disproportionate reduction in time dedicated to research relative to comparable men and women without children. Both men and women report substantial increases in childcare and housework burdens, but women experienced significantly larger increases relative to men.

Contact person: Marina Chugunova

The patent system underpins the business model of some of the fastest-growing companies. Used appropriately, it should support frontier technologies and nurture new firms. Used perniciously, it can stifle innovation and protect established technological behemoths. We analyze patent examination decisions at the American, European, Japanese, Korean, and Chinese patent offices and find evidence that patent attorneys have a surprisingly significant role in the patent system. Our results suggest that some forces within the examination system maintain the uneven playing field by allocating monopoly rights to inventors with better access to influential attorneys, rather than leveling it by favoring inventors with better, non-obvious ideas. Attorney quality is most important, vis-à-vis invention quality, in less codified and more rapidly changing technology areas such as software and ICT. Moreover, attorney quality is more important when invention quality is low. 

Contact Person: Marina Chugunova

We present a comprehensive study on the dynamics of knowledge production and diffusion linked to global mobile inventors (GMIs). We find that GMIs are essential team members of the first few patents in technology classes new to the country of residence as compared to patents filed at later stages. We interpret these results as tangible evidence of GMIs facilitating the technology-specific diffusion of knowledge across nations. (Joint work with Dany Bahar, Brookings Institution, and Prithwiraj Choudhury, Harvard Business School)

Contact Person: Rainer Widmann