Social Systems Research Institute
The University of Wisconsin - Madison
Room 6470 Social Science Building - 1180 Observatory Drive
Madison, Wisconsin 53706 U.S.A.
Phone: (608)262-0446
Fax: (608)263-3876
Electronic Mail: ssri@facstaff.wisc.edu
9802R OPTIMIZATION INCENTIVES AND COORDINATION FAILURE IN LABORATORY STAG HUNT GAMES Battalio, Raymond, Larry Samuelson and John Van Huyck (revised December 1999) (Former title: Risk Dominance, Payoff Dominance and Probabilistic Choice Learning)This paper reports an experiment comparing three stag hunt games that have the same best-response correspondence and the same expected payoff from the mixed equilibrium, but differ in the incentive to play a best response rather than an inferior response. In each game, risk dominance conflicts with payoff dominance and selects an inefficient pure strategy equilibrium. We find statistically and economically significant evidence that the differences in the incentive to optimize help explain observed behavior.
Key words: Payoff dominance, risk dominance, probabilistic choice, logistic fictitious play, bounded rationality, random utility, human behavior.
JEL Classification: c72, c78, c92, d83
Download WP#9802R in pdf.
9819R WHO WANTS A GOOD REPUTATION? Mailath, George and Larry Samuelson (revised Sept. 2, 1999)We examine a market in which long-lived firms face a short-term incentive to exert low effort, but could earn higher profits if it were possible to commit to high effort. There are two types of firms, "inept" firms who can only exert low effort, and "competent" firms who have a choice between high and low effort. There is occasional exit, and competent and inept potential entrants compete for the right to inherit the departing firm's reputation. Consumers receive noisy signals of effort choice, and so competent firms choose high effort in an attempt to distinguish themselves from inept firms. A competent firm is most likely to enter the market by purchasing an average reputation, in the hopes of building it into a good reputation, than either a very low reputation or a very high reputation. Inept firms, in contrast, find it more profitable to either buy high reputations and deplete them or buy low reputations.
9828R THE EVOLUTION OF PREFERENCES AND RAPID SOCIAL CHANGE Sandholm, William H. (revised August 7, 1999)We present a dynamic analysis of the evolution of preferences in a strategic environment. In our model, each agent's behavior depends upon both the game's payoffs and his idiosyncratic biases, but only the game's payoffs determine his evolutionary success. We establish the existence and uniqueness of the paired trajectories of society's preferences and aggregate behavior. While aggregate behavior adjusts smoothly in equilibration games, in coordination games aggregate behavior can jump discretely in an instant of evolutionary time.
9903R EVOLUTION AND MIXED STRATEGIES Binmore, Ken and Larry Samuelson (revised October 6, 1999)Selten showed that no mixed equilibria are evolutionarily stable when players can condition their strategies on the roles they occupy in a game. Alternatively, Harsanyi's purification argument implies that all mixed equilibria are approximations of strict, and hence evolutionarily stable, equilibria of games with slightly perturbed payoffs. This paper reconciles these contrasting results. We show that approximations of mixed equilibria will have high invasion barriers, and hence are likely to persist in an evolutionary setting, when payoff perturbations are relatively important and role identification is relatively noisy. When payoff perturbations are unimportant and role identification is precise, approximations of mixed equilibria will have small invasion barriers and are unlikely to persist.
Journal of Economic Literature Classification Numbers C70, C78.
Keywords: Evolutionary stability, ESS, mixed strategy, asymmetric game.
Download WP #9903R in pdf.
9926 ON INFORMATION SHARING AND INCENTIVES IN R&D Severinov, Sergei (July 29, 1999)Firms competing in the R&D frequently have to deal with the problem of possible exchange of information between their employees. Direct methods of control over such information exchange are often ineffective. In this paper we demonstrate how firms can use the incentive schemes to regulate these information flows. The optimal incentive schemes are derived and properties of equilibria of the incentive scheme game between firms are characterized for different payoff configurations. The results provide an explanation for an observed diversity of incentives schemes for engineers and other technical employees, and for the use of stock options and other forms of profit sharing as a method of preventing the information exchange. We demonstrate that free-riding by both the firms and employees has a significant impact on the final outcomes.
9927 COMPETITION THROUGH INNOVATION: ATMs IN ITALIAN BANKS Hester, Donald D., Giorgio Calcagnini, and Riccardo De Bonis (Sept. 2, 1999)This paper reports results from a study of location, ownership, and acquisitions of automated teller machines (ATMs) by a sample of large Italian banks between 1991 and 1995. The sample banks have 85% of Italian banking assets. Data are collected at the provincial level for each bank. The underlying model is recursive; a bank is presumed to make branching decisions (analyzed in a separate paper) and then, conditional on branching decisions, decisions about ATMs. Several sets of cross-sectional data are studied using OLS and Tobit models yielding the following results: 1) The logarithm of ATMs in a province is related positively to the logarithms of interest-bearing deposits and GDP and negatively to the logarithm of population, as might be predicted from the Baumol/Tobin transactions demand models; 2) The number of a bank's ATMs in a province is related positively to the numbers of its branches and deposit accounts, a province's per capita GDP, the bank's deposits, and the bank's number of employees per branch in the province and negatively to the bank's share of a province's branches; 3) Changes in a bank's ATMs in a province are positively related to changes in the number of its branches and those of competitors; and 4) Concentration indices of ATMs, branches, deposits, and loans decreased at the provincial level between 1991 and 1995.
9928 SOCIAL INTERACTIONS AND FERTILITY TRANSITIONS Durlauf, Steven N. and James R. Walker (revised July 1999)This paper provides an overview of the new literature in economics on social interactions. This literature is based upon the introduction of direct interconnections between socioeconomic actors which can complement standard market-based interactions. Such models provide a framework for studying the implications of conformity effects on aggregate population behavior. A number of implications for demography are suggested.
9929 THE CASE "AGAINST" SOCIAL CAPITAL Durlauf, Steven N. (Sept. 22, 1999)No Abstract.
9930 BEQUESTS AS SIGNALS: AN EXPLANATION FOR THE EQUAL DIVISION PUZZLE Bernheim, B. Douglas and Sergei Severinov (August 11, 1999)In the United States, more than two-thirds of decedents with multichild families divide their estates exactly equally among their children. In contrast, intra vivos gifts are usually unequal. These findings challenge the validity of existing theories regarding the determination of intergenerational transfers. In this paper, we develop a theory that accounts for this puzzle, based on the notion that the division of bequests provides a signal about a parent's altruistic preferences. The theory can also explain the norm of unigeniture, which prevails in other societies. (JEL D10, H31)
9931 BEQUESTS AS SIGNALS: IMPLICATIONS FOR FISCAL POLICY Bernheim, B. Douglas and Sergei Severinov (October 5, 1999)Elsewhere, we have developed a theory of intergenerational transfers based on the premise that bequests act as signals of parental affection, and have shown that this theory can account for the "equal division puzzle." In this paper, we explore the implications of the bequest signaling hypothesis for redistributive fiscal policies. We show that private transfers may not offset public transfers to a significant extent, even though private transfers are altruistically motivated and strictly positive for all but a negligible set of households. This is notable since these conditions are normally thought to yield fully offsetting responses (Ricardian equivalence). We explicitly identify circumstances under which the departure from Ricardian equivalence is large. Notably, the departure may be quite large even when our model is arbitrarily close to one in which Ricardian equivalence is known to hold (in the sense that children care very little about parental affection). (JEL D10, H31, H62)
9932 DETERMINANTS OF BANK BRANCH EXPANSION IN ITALY Calcagnini, Giorgio, Riccardo De Bonis and Donald D. Hester (Oct. 12, 1999)This paper presents a model of de novo branching by Italian banks and reports estimates of its parameters that were obtained by applying a probit-regression (or Tobit) method to cross-section samples of up to 206 large banks for the years 1992-1996. The number of branches has increased rapidly since about 1988. The model incorporates information on banking markets and economic activity in provinces and on a bank's own operating characteristics when attempting to describe de novo branching in a province. The principal findings are: (1) variables describing existing market structure and recent past branch expansion by the bank and its rivals strongly influence de novo branching, (2) banks seek targets of opportunity when siting branches in provinces where they have a presence, where many communes in a province are unserved, and where branches per capita is low, (3) there is only a weak relation between a province's level and change in per capita GDP and de novo branches, (4) banks which merge are more likely to have de novo branches in a province than other banks, and (5) profitable banks with large number of workers per branch and large amounts of loans relative to deposits are likely to have more de novo branches.
This paper has been prepared for the 40th Societa Italiana degli Economisti meeting to be held at the University of
Ancona on October 29-30, 1999.
Download WP #9932 in pdf.
9933 AUCTIONS WITH RESALE Haile, Philip A. (October 13, 1999)This paper studies auctions at which the good for sale may later be resold by the auction winner. Valuations at such auctions are endogenously determined, based in part on equilibrium expectations of resale outcomes. While these valuations sometimes reflect common value components introduced by the secondary market, the resale opportunity changes auctions in ways that cannot be captured by any model ignoring resale. The model predicts signaling through bids, multiple symmetric equilibria for some standard auctions, and potential gains to a seller from excluding bidders in order to create an active resale market. Key policy implications and empirical predictions can also overturn those based on models without resale.
9934R A BACKWARD INDUCTION EXPERIMENT (revised September 7, 2000) Binmore, Ken, John McCarthy, Giovanni Ponti, Larry Samuelson, and Avner ShakedThis paper examines the results of experiments with one-stage and two-stage alternating-offers bargaining games. As do previous experiments, we find that the results are inconsistent with the assumptions that players maximize their monetary payoffs and perform backward-induction calculations. Payoff-interdependent preferences have been suggested as an explanation for experimental results in bargaining games. We examine whether, given payoff-interdependent preferences, players respect backward induction. To do this, we break backward induction into its components, subgame consistency and truncation consistency. We examine each by comparing the outcomes of two-stage bargaining games with one-stage games with varying rejection payoffs. We find and characterize systematic violations of both subgame and truncation consistency.
Journal of Economic Literature Classification Numbers C70, C78
Keywords: Bargaining, experiments, backward induction, subgame-perfect equilibrium, interdependent preferences.
Download WP #9934R in pdf.
9935 "I DIDN'T TELL, AND I WON'T TELL": DYNAMIC RESPONSE ERROR IN THE SIPP Bollinger, Christopher R. and Martin H. David (Nov. 11, 1999)Using state administrative records matched to two interviews of the 1984 Survey of Income and Program Participation panel, we examine intertemporal relationships in response errors for participation in the Food Stamp Program. Response error is highly correlated for these interviews. Hypotheses that the error process can be explained by learning behaviors are rejected. Bivariate probits of response error, including income and household characteristics as covariates, are stable across the two periods and show that autocorrelation in the errors is not attributable to readily available attributes. These findings support the cooperator hypothesis previously forwarded by Bollinger and David (1999a).
9936 ENCOMPASSING TESTS WHEN NO MODEL IS ENCOMPASSING West, Kenneth D. (November 1999)This paper considers regression-based tests for encompassing, when none of the models under consideration encompasses all the other models. For both in- and out-of-sample applications, I derive asymptotic distributions and propose feasible procedures to construct confidence intervals and test statistics. Procedures that are asymptotically valid under the null of encompassing (e.g., Davidson and MacKinnon (1981)) can have large asymptotic and finite sample distortions. Simulations indicate that the proposed procedures can work well in samples of size typically available, though the divergence between actual and nominal confidence interval coverage sometimes is large.
9937 ALMOST GLOBAL CONVERGENCE TO p-DOMINANT EQUILIBRIUM Sandholm, William H. (December 17, 1999)A population of players repeatedly plays an n strategy symmetric game. Players update their strategies by sampling the behavior of k opponents and playing a best response to the distribution of strategies in the sample. Suppose the game possesses a (1/k)-dominant strategy which is initially played by a positive fraction of the population. Then if the population size is large enough, play converges to the (1/k)-dominant equilibrium with arbitrarily high probability.
Keywords evolutionary game theory, coordination games.
Download WP #9937 in pdf.
9938 EVOLUTIONARY IMPLEMENTATION AND CONGESTION PRICING Sandholm, William H. (December 1, 1999)There are many implementation problems whose social alternatives are the strategy profiles of some underlying game. Price schemes are mechanisms whose strategies are borrowed from this game. Price schemes are less costly to administer than revelation mechanisms, which must introduce new sets of strategies; however, dominant strategy implementation via price schemes is typically impossible. We therefore introduce the notion of evolutionary implementation, which requires that players who adjust their behavior in any reasonable myopic fashion will learn to behave as the planner desires. Like dominant strategy implementation, evolutionary implementation does not rely on the assumption of equilibrium play. We establish that implementation of efficient behavior via price schemes is possible whenever the aggregate payoff function of the underlying game is concave. We provide a detailed study of the evolutionary implementation in congestion games, where implementation through decentralized price schemes is always possible.
JEL Classification Numbers: C61, C72, C73, D62, R41, R48.
Download WP #9938 in pdf.
468
Brock, W.A, S.R. Carpenter, D. Ludwig
Management of Eutrophication for Lakes Subject to Potentially Irreversible Change
Ecological Applications 9 (3), 751-771 (1999).
469
Andreoni, James and Hal Varian
Preplay Contracting in the Prisoners' Dilemma
Proc. Natl. Acad. Sci 96, 10933-10938 (September 1999).
470
Durlauf, Steven N.
How Can Statistical Mechanics Contribute to Social Science?
Proc. Natl. Acad. Sci 96, 10582-10584 (September 1999).
471
Brock, William A.
Scaling in Economics: A Reader's Guide
Industrial and Corporate Change, 8 (3), 409-446, 1999.