Marion Dumas - Research
Green technological transitions and industrial networks
with Eugenie Dugoua
A transition to a low-carbon economy requires radical technological change in a range of industries. What institutional mechanisms can promote such radical techological change? The principal instruments promoted by economists are R&D subsidies and carbon taxes. The effectiveness of these policy instruments is premised on a world of atomized firms, capable of independently optimizing their R&D investments. Instead, we argue that radical innovations (such as electric cars or zero-energy buildings) require complementary investments by multiple firms within supply networks. We show that the structure of these networks affects the incentives of firms to undertake radical innovations. Basing ourselves on case studies, we build a simple model of firms attempting to coordinate their investments. The model shows that some firms bear greater risks than others due to their position in the network and can be far less responsive to marginal pricing policies such as a carbon tax unless institutions exist to coordinate R&D. We then consider institutional solutions, such as contracts developed by research consortia, or targeted subsidies by the government. In a second phase of the project, we will develop a method to identify radical innovations in the patent record and study the institutional conditions of their emergence.
Read more
The evolving practice and politics of Superfund regulation
with Michael Livermore
In this project we analyze 35 years of hazardous waste site cleanup to understand the feedbacks between legal and regulatory decisions, the power of interest groups, the know-how accumulated in the process of resolving the conflicts between these interests and the resulting changes in the state of the physical environment. To do so, we build a large dataset where for each documented hazardous site, we inventory all public decisions made by the Environmental Protection Agency, as well as the site's characteristics, the potentially responsible parties, enforcement actions and court decisions. See here for a geographical display of this data. We will use these records to track regional differences in regulatory practice as well as the evolving characteristics of the conflicts between industries, other interest groups and the agency.
Read more
Scaling of local governments in metropolitan areas
with Luis Bettencourt
Many social outcomes systematically scale with the size of cities: economic growth, patents or crimes scale super-linearly, while energy consumption from infrastructure sub-linearly. In this project, we examine how the structure of local governance fits in this new science of cities. The number of local governments within a metropolitan area seems to scale very sub-linearly with its size. We investigate why by examining the organizational structure of these governments by service area. Each city's departure from these scaling relationships give a sense of its local specificity, it's unique flavor. Past research has found that cities' specificity (tendency to overperform or underperform in any of these indicators) is very stable over time. Is this also true for the size of local governments and for the size of cities' budgets? Is local government capacity (measured by number of governments, number of employees and budget), once adjusted for the city's size, related to a city's crime rate or economic growth? Studying these relationships will give hints about possible feedbacks between city size, government capacity and a city's socio-economic outcome, which has important implications for public finance and redistribution across cities.
Read more
When does reputation lie? Dynamic feedbacks between reputation, social structure, and wealth
with Eleanor Power and Jessie Barker
We explore how feedbacks between social status and reputation could fuel structural inequality. Ethnographic data from two villages in South India including data on reputational standing, socioeconomic status, support networks, and public religious action can be combined to show what amounts to a "reputational poverty trap": more prominent individuals gain more from reputation-building acts than less prominent individuals. We seek to understand the mechanisms underlying this positive feedback. To do so, we build simple computational models that explores the dynamic relationships between network position, wealth, reputation for cooperation, and underlying generosity. In particular, we look for the conditions under which reputation and wealth becomes decoupled from individuals’ underlying propensity to help others as a result of the positive feedback between reputation and one's visibility in the network.
Read more
The co-evolution of law and culture: the case of Civil Rights in the United States
In some countries, high courts are highly visible and make landmark decisions on societally important issues. The United States is an example. The U.S. Supreme Court has made many important decisions on civil rights over the course of the 20th century. Are these decisions simply reflecting a new majority opinion in the population or do they contribute to changing this opinion? Does the Court respond or lead? This question is important for our understanding of authority and leadership in hierarchies. To shed light on this issue, we collect all civil rights decisions in the U.S. federal courts. Using computational linguistics, we characterize features of all lower courts' civil rights ideologies over time and use this to investigate the degree to which higher-level court decisions are responsive to these lower court ideologies and the degree to which they transform them in the aftermath of landmark decisions.
Read more
Mental models and deliberation on climate change
with Henrik Olsson and Josh Grochow
Sustainability problems require us to manage complex systems whose dynamics are uncertain. The literature on complex adaptive systems suggests that cognitive diversity (diverse mental models) helps "harness the complexity" of these systems. On the other hand, diverse mental models can also cause polarization of beliefs, as people can interpret information in divergent ways and politicians can manipulate beliefs to divide constituencies. This process has hampered the growth of a greater civil society response to cliamte change. With colleagues at the Santa Fe Institute, we are developing a survey instrument that allows people to interactively map the system of beliefs underlying their opinions about the need to mitigate cliamte change. We will then organize deliberative interactions in which individuals use this to map each others' beliefs. Our aim is to test whether such structured deliberation can reduce polarization on the issue of climate change and lay the foundations for more collective action.
Read more