Research


Working Papers

Incumbency, Endogenous Valence and the Dynamics of Political Competition (with Michael Kramm, Job Market Paper)

Electoral competition consist of two highly intertwined phases, which can have positive feedback effects between them. In the Governance Phase, the parties with more representatives in the parliament are better placed to provide services that are valued by all voters, improving their standing among voters. In the Electoral Phase voters can reward good governance by preferentially voting for the party with the better standing, thereby transmitting the initial advantage to the subsequent stages. We explore the implications of this feedback structure on platform choice within a dynamic game context. We show that multiple equilibria can exist with qualitatively different equilibrium behaviour. Our results bring out a possible conflict between moderation and electoral competitiveness, and highlights the challenges of electoral reform. Job Market Paper.pdf


Cost Uncertainty and Stochastic Orders in a Model of Vertical Differentiation(With Nima Jouchaghani)

We analyze a canonical model of vertical differentiation, where two firms compete in prices and qualities. In contrast to classical vertical differentiation models, we assume that a firm is uncertain about its rival's fixed costs of quality improvement, captured by an efficiency parameter. We find two types of equilibria, which substantially differ in the features of the products which are sold. In one equilibrium, the products are maximally differentiated in the quality dimension, reminiscent of the equilibrium with no cost uncertainty. Besides, we identify a second equilibrium in which firms are more competitive in the quality dimension in that similar efficiency parameters lead to similar products concerning the quality dimension. This, in turn, leads to lower prices compared to the first equilibrium, to the benefit of consumers. Finally, we conduct comparative statics using stochastic orders. Among other results, we find that shifting probability mass in the distribution of consumers' willingness to pay for quality from the lower to the upper boundary can increase consumer rent for all consumers, independent of their willingness to pay for quality. Increase cost risk , in the SSD sense, reduces firms investment in quality improvement, but can increase expected ex-ante quality.

Available on request: Please send an email to kangkan.choudhury@uni-bamberg.de



Dynamic Vertical Differentiation with Cost Uncertainity (With Nima Jouchaghani)

This paper extends a classical vertical differentiation model to a dynamic framework while assuming that firms are uncertain about the rival's and their own future fixed cost of quality improvement. The dynamic dimension allows firms to stepwise increase their quality, while uncertainty about costs translates into uncertainty about how profitable it is to increase quality. As a consequence of cost uncertainty, firms condition their strategy on their own costs, which gives rise to an Markov Perfect equilibrium, where both the principle of maximum- and minimum quality differentiation hold. We identify conditions such that the game ends almost surely in either maximum- or minimum differentiation. The latter case only occurs, if both firms decide to choose the highest possible quality level for their product. This is surprising since in this case the firms make zero profits forever within our model, even though the time dimension allows the firms to stepwise increase their quality and end up in states of intermediate product differentiation. A direct consequence of this result is that leap-frogging does not occur in the Markov Perfect equilibrium of our model.

Latest Version


Stochastic Replicator Dynamics: An Experimental Assesment (with Tigran Aydinyan)

We study the dynamics of strategy choice of subjects in an environment where payoff s are perturbed by strategy-specific noise, and strategy revision is governed by the simple proportional-imitation protocol. Applying stochastic replicator dynamic of Mertikopoulos and Viossat (2016)to a two by two symmetric game, we derive the steady-state frequencies of the strategies and show that dominated strategy can persist in equilibrium if the variance to the payoff shocks exceeds a critical value. Additionally, the rate of decay of the dominated strategy decreases with increasing variance. Using numerical and monte-carlo methods, we show that in the presence of high variance, the distribution of the rate of cooperative behaviour would become bi-modal with time. The behaviour of our experimental subjects accords with these tendencies. Our findings add to the growing literature that demonstrates that evolutionary models of human behaviour with simple adaptive rules have predictive-ability even in complex environments.

Latest Version


Empirical Research

Dependence on Freely Collected Food and Fuel for Household Consumption in Rural India: An Empirical Analysis (with Sujoy Chakravarty)

Using NSS consumer surveys covering a span of eight years (2004-2012), we estimate the level of dependence on freely collected common pool resources in the household consumption of food and fuel in rural India. We find that close to half the households report using free collection to supplement household budgets in 2004-05 which reduces to about one-third in 2011-12. The fraction of the food and fuel budget that is accounted for by free collection exhibits a slight downward trend over the period. Though this free collection ratio reduces over monthly per capita consumption expenditure (MPCE) classes, all classes have significant numbers of households that use these products along with marketized food and fuel. Furthermore, we see that dependence on free collection is higher for women headed and tribal households and small and marginal land owners, while the presence of at least one wage earner or an educated adult in the family lowers it. We explore if the introduction of an exogenous demand shock in the form of increased employment opportunities under an employment guarantee scheme (NREGA) can account for the observed decrease in free collection, and find that it is associated with a decrease in free collection in the poorest fifty percent of the districts.