Faculty and Research

Sheila Olmstead, Fellow, Resources for the Future

Title: "Economics and water resources: welfare, pricing, markets, and regulation"

Abstract: Recent events, news articles in the popular press, and research outside the field of economics have raised serious questions about the sufficiency and quality of global freshwater supplies and the potentially devastating impacts of current and future water shortages. While economists have studied water resource management for many decades, they have responded in a non-alarmist fashion to most of these concerns about global water supplies. In part, this is justified by our faith in the ability of markets to allocate scarce resources. But are there some concerns about the quality and quantity of global water resources are warranted from an economic perspective? Water prices typically are not determined in markets and do not reflect resource scarcity. Allocation mechanisms are highly political, and even when faced with significant scarcity, management institutions are reluctant to raise prices. Typical property rights structures for both renewable and non-renewable water resources ignore important spatial and temporal externalities and public goods. Water quality regulation in the United States and leaves much to be desired from the perspective of efficiency. Most importantly, the welfare implications of insufficient access to clean drinking water supplies in terms of human health are demonstrably very large. This presentation will offer a broad overview of both the literature and current research on these topics and others in the area of water resource economics.

Eugene Fitzgerald, Professor, Dept. of Materials Science and Engineering, MIT

Title: "Inside Real Innovation: From Silicon to Water"

Abstract: Through our own experience in fundamental innovation over the last 20+ years and through our innovation experience with corporate-university innovation projects, we have developed a micro-scale innovation picture that captures the processes experienced by innovators. We use real-life cases from strained silicon integrated circuits to water purification technology to highlight aspects of what we refer to as iterative innovation. The details of the innovation process are incompatible with commonly held 'linear model' views of research to development to manufacturing. Instead, we show that markets and applications, science and technology, and business models and methods interact in an iterative process to produce real innovation.'

***Please note: this seminar will take place is Cumnock Hall, Rm. 102***

Special Workshop: "Resource Sharing in Biomedical Science"

This will be conducted as a special seminar consisting of a panel of experts with each member preparing brief remarks designed to set up a discussion. The seminar also has a web page where readings and some of the expert remarks are posted.

Lee Fleming, Professor, Harvard Business School

Noncompetes: Brain Drains, Career Impacts, and Strategic Implications

Title: "Regional Disadvantage? Non-Compete Agreements and Brain Drain" - Matt Marx, Jasjit Singh, and Lee Fleming (presenting).

Abstract: We construct inventor career histories using the U.S. patent record from 1975 to 2005 and demonstrate a brain drain among patenting inventors from states that enforce employee non-compete agreements to those that do not. Non-compete enforcement drives away inventors with greater human and social capital although retaining those who are less productive and less collaborative. We address causality-related concerns with a difference-in-differences study design based on an inadvertent reversal of Michigan's non-compete enforcement policy.

Title: "The Firms Strike Back: Non-Compete Agreements and the Mobility of Technical Professionals" - Matt Marx

Abstract: This study documents how firms shape labor markets and career paths using employee non-compete agreements. Although non-competes have been overlooked in the sociology of work, data from this study indicate that nearly half of technical professionals in the U.S. are asked to sign such employment contracts. Firms use non-competes to "strike back" against the increased mobility of technical professionals following the decline of internal labor markets, lest they lose investments in talent and trade secrets. Interviews with 52 randomly-sampled patent holders in a single industry coupled with a survey of 1,029 engineers in a variety of industries-the first primary data regarding the use of non-competes-reveal that non-competes lead ex-employees to take career detours in that they involuntarily leave their field in order to avoid being sued. Those who continue in their occupations eschew opportunities at small, entrepreneurial firms, instead seeking shelter with larger companies that can more easily defend them against a potential lawsuit. These findings inform our understanding of the social organization of work in the knowledge economy and also raise critical policy questions.

Title: "Employee Mobility and the Likelihood of Becoming a Target for Acquisition" - Kenneth Younge (presenting), Tony Tong, Lee Fleming

Abstract: We explore the conditions under which the potential for employee mobility affects the likelihood that a firm will become a target for acquisition. We use an inadvertent change in the enforcement of non-compete agreements in Michigan as a natural experiment to analyze this relationship. Using a differences-in-differences approach, we find that constraints on employee mobility raise the likelihood that a firm will become an acquisition target (H1). We also find that this effect is stronger when firms are exposed to higher rates of mobility from knowledge workers (H2), is stronger when firms face higher levels of instate competition (H3), and is lower when firms can protect intellectual property through patenting (H4). Prior research in strategic management focuses on how firms may use M&As to source knowledge from target firms, implicitly assuming that knowledge assets can be captured and retained after the acquisition. We depart from prior research by investigating how the mobility of employees affects this strategy. We leverage a known source of exogenous variation in the mobility of employees to develop a causal explanation about the likelihood of acquisition activity. We find that expectations of ex post mobility of knowledge workers affect target selection and ex ante M&A decisions.

Peter Girguis, Professor, Dept. of Organismic and Evolutionary Biology, Harvard University

Title: "Microbial fuel cells: The facts, the fiction, and the future of 'electron harvesting' "

Abstract: Microbial fuel cells are devices that harness energy - in the form of electrons - from microbial metabolism. As the name implies, microbial fuel cells (or MFCs) provide electricity without combustion. Moreover, they are carbon neutral, operate independent of sunlight, and can run for decades. Recent years have been a watershed for microbial fuel cell research, and much effort has been placed into engineering systems to harness electricity from waste (e.g. sewage). To date, this has not proven to be commercially viable. However, a reassessment of the technology and its commercial potential has lead leaders in science and industry to re-consider how this "platform technology" may be brought to bear on major issues in bioprocessing, remediation, and remote energy generation. Thus, like many emerging technologies, MFCs may have their greatest impact in fields and applications that have yet to be examined.

Ina Ganguli, Harvard Kennedy School, PhD candidate

Title: "Saving Soviet Science: The Impact of Grants When Government R&D Funding Disappears"

Stine Grodal, Assistant Professor of Strategy and Innovation, Boston University and Grid Thoma, Assistant Professor of Economics and Management at the University of Camerino, Macerata, Italy

Title: Institutional Logics and Status: Explaining Strategic Patenting in the Legal Service Sector

Professional service firms adhere to at least two often conflicting institutional logics - serving their clients and serving the profession. But the literature says little about under which conditions professional service firms align their actions with each of the conflicting logics. Simultaneously, a large literature emphasizes that the behavior of professional service firms are driven by status dynamics, however, this literature has yet to consider how status dynamics affect firms' adherence to institutional logics. We bring these two previously separate literatures together by investigating law firms' tendency to aid their clients with strategic patenting. Strategic patenting confronts law firms with a dilemma. On the one hand engaging in strategic patenting aids the client firm and is thus based on a client centered logic. On the other hand strategic patenting undermines the intent of the legal system and thus challenges law firms' adherence to a professional logic. This raises the puzzle of which firms choose to engage in strategic patenting? Drawing on a large dataset of patent applications from the European Patent Office we first show that status is a strong predictor of following a client logic. Yet, the effect of status wanes as firms gain more experience within their profession. Second, we demonstrate that high status firms generally are more likely to abide by both logics simultaneously, but that the relationship between experience and adhering to both logics follows an inverted U-shape. We end by discussing the implications of our research for theories of institutional logics, status, and cumulative innovation.

Alexander Peterson, Ph.D. Candidate, Boston University, Physics Department

Title: Quantifying Statistical Regularities in the Career Achievements of Scientists and Athletes

Abstract: For many professions, the quantitative analysis of individual careers is made difficult by the lack of comprehensive data and the difficulty in defining measures for productivity and longevity. However, comprehensive career data is recorded in professional sports and is perfectly tailored for studying human productivity. Similarly, the publication careers of scientists are also quantifiable using similar measures. Since both professions are subject to the common forces of competition, one motivating question in this talk is: "What are the statistical regularities in career achievement across an entire cohort of competitors?"

In this talk I will discuss the statistical regularities that describe the everyday topic of career achievement using comprehensive career data. In the first part of the talk, I will discuss the topic of career longevity, using as example the 60+ year history of the National Basketball Association and 2700+ complete careers over the period 1946-2004. Surprisingly, we find that a common career longevity distribution describes the careers of 20,000+ athletes from 4 sports leagues and 400,000+ scientists from 6 high-impact journals, where each journal serves as a generic arena for competition. In order to account for the regularities we observe across several professions, we develop an exactly solvable model for career longevity based on the Matthew "rich-get-richer" effect. Our model is in excellent agreement with empirical career longevity distributions for each profession analyzed. Our model follows from two general assumptions: (i) that there is random forward progress in the career, whereby it becomes easier to make progress the further along one is in his/her career, and (ii) that career termination follows from random hazards that are present throughout the career. The findings suggests that there is a common underlying mechanism which underlies career development in competitive professions. In the second part of the talk, I will discuss the publication careers of 300 individual scientists (ranging from very the very famous to current Assistant professors) and find remarkable statistical regularity in the functional form of the rank-citation distribution (analogous to the Zipf rank-frequency distribution) for each scientist studied.

Marcus Oliver Gay, Biofuels Information Consultant, Elsevier Biofuel - Americas

Title: Development of a Search and Discovery Tool Supporting the Commercialization of Alternative Energy Technologies

As a leading publisher of Scientific, Technical and Medical literature, Elsevier has invested two years of development work to create the first in a series of Alternative Energy (domain specific) search and discovery tools. The flagship product, Elsevier Biofuel, will be launched externally this summer. This solution aggregates relevant information from over 900 journals, 800 books and in excess of 10 million patents. The platform enhances end-user productivity through the use of powerful search algorithms and a custom keyword-index comprising of 1.5 million search terms relevant to this industry. The result is a comprehensive resource uniquely developed for applied research in the biofuel market.

This presentation will explore questions relating to Elsevier's development of tools for applied research. In addition, we will review the go-to-market strategy and some of the key technical challenges faced in developing a tool for the biofuel market.

Eric Archambault, President and CEO, Science-Metrix

Title: Scale-Adjusted Indicators of Research Activity: The case of Scientific Collaboration

Abstract: In 1926, Alfred J. Lotka suggested that the distribution of scientific activity followed an inverse square law, which is a form of power law. The presence of power-law relationships in data is one of the signatures of complex, interconnected systems. These systems violate several of the conditions required to perform classical statistics that assume randomly occurring, independent processes. What is at stake here is that even the most basic performance indicators such as the simple ratios that are conspicuously used in the study of the research system provide a distorted view of performance. In 2000, Katz proposed a new method to normalize data in the presence of power laws that he observed in the relationship between publications and citations and in the propensity to collaborate. Though Katz's suggestion has applications that are extremely varied in scientometrics, instances of the application of his method are few and far between. In the last two years, Science-Metrix has been conducting research on the application of the Katz scale-adjustment method to the computation of collaboration, preference patterns in collaboration, scientific impact, field-normalized scientific impact and the study of specialization. This presentation shows what we have found to date and examines some of the numerous challenges ahead in the revamping of the research-system analysts' toolbox.

David Weitz, Harvard University, School of Engineering and Applied Sciences

Title: Commercializing Technology from University Research

Abstract: In this talk, I will discuss a variety of different mechanisms for commercializing technology developed in fundamental research programs carried out in labs at Harvard University. These mechanisms include collaboration with large companies who can do the necessary development work, start-up companies, and collaborations with new institutes at Harvard. For the start-up company mechanism, I will also discuss different funding possibilities and will compare the advantages of these. The talk will be based on experience with work that has evolved from my own research lab, which focuses on a combination of biotechnology and materials for drug delivery and home and personal care products.

Kirk Doran, Assistant Professor of Economics, University of Notre Dame

Title: The Collapse of the Soviet Union and the Productivity of American Mathematicians

Abstract: Modern models of economic growth are built on the assumption that there are strong human capital externalities. An increase in the supply of highly skilled workers (through both immigration and increased training) will increase the productivity of the pre-existing high-skill workforce, and lead to a substantial increase in national wealth. Several hundred world-class mathematicians migrated to the United States after the collapse of the Soviet Union in 1992. Using a unique data set containing information on the complete publication, citation, and affiliation records of mathematicians in the United States and the former Soviet Union, we examine the impact of the Soviet influx on the productivity (as measured by the number of publications and citations) of American mathematicians. In contrast to the substantial positive externalities conjectured in the economic growth literature, we find a negative impact on the productivity of American mathematicians whose research skills most overlapped with those of the Soviets. Despite these adverse productivity effects at the individual level, institutions with a history of research in fields that overlapped with Soviet research interests were able to increase their total published output and citation count by making strategic hiring and firing decisions. We also document a number of behavioral responses among American mathematicians: an increased mobility rate (both to lower quality institutions and out of the "active" publications market altogether) and a shift away from the fields dominated by the Soviet mathematicians (particularly among new entrants into the mathematics community).

Dennis Crouch, Associate Professor, University of Missouri School of Law

Title: The Market for Patents: Themes Developed From the Files of One Million Recent Patent Applications Dennis Crouch

Abstract: The US patent system has recently struggled under intense criticism for slow processing, invalid and ambiguous patents, expensive litigation, etc. These concerns impact both innovative entities, follow-on competitors, and their consumers. Policymaking institutions have been busy considering rules that alter the patenting and patent litigation process. However, many seem to have forgotten that incentives and a market-based approach can also be effective at shifting the patent system while limiting potentially harsh rules. Using the USPTO's simple fee structure, I have considered at how patent applicants are responding to choices available. For instance, although there is some value associated with patent drafted in a complex and lengthy format, I find that patent applicants are often willing to simplify their applications in order to avoid a small up-front fee. In another vein, I find that the huge patent office backlog does not appear troublesome to many patent applicants (based on their own actions). A market-based approach, that might would allow a choice of complex versus simple and fast versus slow, has its advantages as well as potential costs. The results here provide a number of important lessons about patent law institutions and their communities.

*** Please note this seminar will take place in Cumnock Hall, Room 103***

Yiling Chen, Assistant Professor of Computer Science, Harvard University

Title: Designing Markets for Prediction

Abstract: Suppose that you are interested in estimating the probability that the US unemployment rate in June 2011 will be greater than 10%. You might choose to ask an expert, dig through news articles, or run a regression with historical data to come up with a reasonably well-informed guess. But you might be able to obtain an equally good and often better estimate by appealing to the wisdom of crowds and asking the market. A prediction market is a betting intermediary designed to aggregate information. It offers contracts whose payoff is tied to outcomes of an event of particular interest and attracts participants to trade the contract. For instance, a contract that pays $1 if the US unemployment rate in June 2011 is greater than 10% and $0 otherwise can be used to predict the likelihood that the employment rate will soar. If the current market price of the contract is $p, a rational, risk-neutral agent should be willing to buy the contract if he believes the true probability for the unemployment rate to be greater than 10% is greater than p. Conversely, he should be willing to sell the contract if he believes the true probability is lower than p. The market price hence incorporates the information of participants and approximately represents a real-time consensus forecast for the event. In this talk, I will discuss how to better design market mechanisms for the purpose of information aggregation, present some results on designing more expressive markets for agents to reveal their information, and discuss their potential applications. Results in this talk are based on joint work with Jacob Abernethy, Lance Fortnow, Sharad Goel, Nicolas Lambert, David M. Pennock, and Jennifer Wortman Vaughan.