Emanuela Sciubba

Links and Architecture in Village Networks

In this paper we test the implications of a model of network formation on data from rural Ethiopia. In contrast to the current literature, we demonstrate the critical role of both number of links and architecture in determining the impact of social networks on outcomes. Social capital matters, but its impact differs by the architecture of the network to which one belongs.

Survival with Ambiguity

We analyze a market populated by expected utility maximizers and smooth ambiguity-averse consumers. We study conditions under which ambiguity-averse consumers survive and affect prices in the limit. If ambiguity vanishes with time or if the economy exhibits no aggregate risk, ambiguity-averse consumers survive, but have no long-run impact on prices. In both scenarios ambiguity-averse consumers are fully insured against ambiguity in equilibrium and thus behave as expected utility maximizers with correct beliefs. If ambiguity-averse consumers are not fully insured against ambiguity, their behavior mimics expected utility maximizers with wrong beliefs and a stochastic discount factor which might be consistently higher or lower than their actual discount factor. We use this insight to analyze a Markov economy with large persistent ambiguity. Consumers with decreasing absolute ambiguity aversion whose prudence with respect to ambiguity exceeds twice their absolute ambiguity aversion a.s. survive in the presence of expected utility maximizers with correct beliefs. If the economy further exhibits aggregate risk, they drive the expected utility maximizers out of the market and determine prices in the limit. In contrast, consumers with increasing or constant absolute ambiguity aversion only survive in the absence of aggregate risk and have no impact on limit prices.

Extrapolation in games of coordination and dominance solvable games

We study extrapolation between games in a laboratory experiment. Participants in our experiment first play either the dominance solvable guessing game or a Coordination version of the guessing game for five rounds. Afterwards they play a 3x3 normal form game for ten rounds with random matching which is either a game solvable through iterated elimination of dominated strategies (IEDS), a pure Coordination game or a Coordination game with pareto ranked equilibria. We find strong evidence that participants do extrapolate between games. Playing a strategically different game hurts compared to the control treatment where no guessing game is played before and in fact impedes convergence to Nash equilibrium in both the 3x3 IEDS and the Coordination games. Playing a strategically similar game before leads to faster convergence to Nash equilibrium in the second game. In the Coordination games some participants try to use the first game as a Coordination device. Our design and results allow us to conclude that participants do not only learn about the population and/or successful actions, but that they are also able to learn structural properties of the games.

The determinants of individual behaviour in network formation: some experimental evidence

The role of social networks in shaping economic outcomes has received increasing attention in recent years. Network externalities have been extensively studied both in industrial organisations and, more recently, within the theory of social capital and development economics. Most of this literature however takes the structure of the social network as given and analyses the consequences of network externalities on outcomes.

Links and Architecture in Village Networks (previously: Endogenous Network Formation and Informal Institutions in Village Economies)

This paper o¤ers a bridge between the theoretical literature on endogenous network formation and the empirical work on the impact of social networks on economic performance. We provide a theoretical framework of endogenous network formation that yields testable predictions for the network architectures generated by a particular informal institution common in village economies. We test the implications of the model on data from rural Ethiopia. In contrast to the current literature, we demonstrate the critical role of both number of links and architecture in determining the impact of social networks on outcomes.Social capital matters, but its impact di¤ers by the architecture of the network to which one belongs.