MARINA AGRANOV

ACADEMIC JOB MARKET

Expected graduation date: June 2010


home

cv

job market paper research


Current Research

 

"Flip-flopping: contrasting effects of intense primaries and general elections on selection of candidates"- JOB MARKET PAPER (UPDATED January 2010)

We present an incomplete information model of two-stage electoral competition in which candidates can choose different platforms in primaries and general elections. Voters do not directly observe the chosen platforms, but rather infer the candidates’ ideological types from signals made during the campaign (debates, speeches, etc.), where a bigger number of signals corresponds to a higher intensity campaign. This model captures two commonly observed patterns: (1) the “post-primary moderation effect," in which candidates pander to the party base during the primary and shift to the center once the nomination is secured, and (2) the “divisive-primary effect," which refers to the detrimental effect of intense primaries on a party’s general-election prospects. We test model’s predictions in the controlled laboratory experiment. Experimental results indicate that primary voters and candidates understand the trade-offs of two-stage elections and behave in a manner that is consistent with the theoretical predictions. Moreover, intense primaries drive moderate candidates out of the competition in the nomination stage. This model also generates a new empirical prediction: intense primaries tend to elect extreme candidates, while intense general elections tend to elect moderate ones. Preliminary results from the U.S. House races in 1998 support this prediction.

 

Link              

 

"An Experimental Study of Ambiguity and Vagueness in the Announcement Game" (joint paper with Andrew Schotter)

 

In this paper we study the efficiency properties of natural language used as a communication device in what we call Announcement Games. An Announcement game is a cheap-talk game, in which an "Announcer" upon privately observing the value of the state of the world, x, makes an announcement concerning its value to a set of "Players" who then engage in a game whose payoffs depends on the value of x. In such a context, we compare the use of an "ambiguous" communication technology, in which the "Announcer" reports the interval into which x falls to one which is "vague", in which words (natural language) are used to describe x. We expect that the type of language used to communicate may make a difference for the ability of the Players to converge to equilibrium. We investigate this conjecture experimentally and find that, while theoretically we can not expect words to outperform intervals, empirically they do as well as long as the Announcer uses the minimally optimal number of words. Excessively large vocabularies, even when optimal, are efficiency decreasing. Finally, our results indicate that it is essential that agents using a language reach a common understanding of what words mean if they are going to be successful in achieving efficient payoffs.

 

Link

 

"Vagueness and Inequality: an Experimental Study" (joint paper with Andrew Schotter)

 

This paper suggests one reason for using a vague communication strategy when a precise one is available. The reason is that in situations where payoff inequality is likely to interfere with coordination (as in the Battle of the Sexes Game), being vague about the game being played and its payoffs may help to mask this inequality. Using controlled laboratory experiment, we show that vague words (as well as intervals) outperform precise communication strategy in what we call the "Announcement Game".

 

Link

 

"Relative Piece-Rate, Tournament and Independent Piece-Rate Schemes: an Experimental Study" (joint paper with Chloe Tergiman)

 

When risk-averse agents' outputs are correlated, using relative performance information can improve the trade-off between incentives and risk-sharing. The experimental work that tests this idea has mainly focused on either rank-order tournaments, in which payment depends only on relative ranking, or independent piece-rates, which use only absolute performance information. Theoretically, however, the principal can do better by designing a wage scheme that uses both the relative and absolute performance of his agents. In this paper we study such a scheme in a controlled laboratory setting and find that as the theory predicts, it outperforms the tournament and the independent piece-rate.

 

Link

 

"Other People's Money: an Experimental Study of the impact of the Competition for Funds on Risk Taking" (joint paper with Andrew Schotter and Alberto Bisin)

 

The recent economic crisis in the financial industry has brought into question the prudence of hedge fund managers and other members of the financial community. However, behavior which appeared irrational to outsiders was so wide spread that it implied that there was something more systematic going on than a contagious epidemic of irrationality. In this paper we experimentally investigate the impact that competing for funds has on the risk taking behavior of investors (laboratory hedge fund managers). We find that the process of competing for capital leads investors to promise returns that can only be realized by risky investments. In other words it is not that individual investors are irrational but, on the contrary, are behaving in a rational manner dictated by the need to attract capital. Our results indicate that the impact of competition leads investors to invest in riskier projects than they or their lenders would want to invest in, in the absence of competition and that such behavior is welfare decreasing. We investigate a set of policy interventions on these markets like transparency restrictions and constraints on the type of contracts offered to our laboratory hedge fund managers and find that such interventions can indeed increase welfare.

 

Link coming soon