Review of Economic Design最新文献

I propose a theoretical framework under which holistic and test-only admission policies can be compared. The framework is based on an extension of the Coate–Loury model (Am Econ Rev 83(5):1220–1240, 1993) by allowing the candidate to signal in some additional attributes: The recruiting institution values the candidate’s competency in two aspects (e.g. academic and leadership skills); standardized test is available only for screening academic skill whereas the candidate can also costly signal about his leadership skill type; the recruiting institution can adopt different admission scores conditional on the presence or absence of the leadership signal (i.e. holistic admission); and the candidate needs to invest in an aspect to be qualified in it. The presence of the signaling stage alters the mechanism by which the two-dimensional investment decision is made compared to the standard Coate–Loury case which corresponds to the signal-blind equilibria interpreted as the outcome under test-only admission, with implication on equilibrium multiplicity and welfare ranking. In particular, we find that the recruiting institution could benefit from the signaling option which leads to a more diversified skill profile but the candidate could be hurt from increased burden from costly investment and signaling even if his admission chance is improved.

In a world of experience goods, two costly ex-post disciplinary actions can be used against malpractice of firms: consumer lawsuits and government investigation. We distinguish between government effectiveness in detecting ‘bad behavior’ vs. ‘good behavior’ of firms—both play a key role in the model. Our results suggest that while an effective government eliminates malpractice completely, the intervention of an ineffective government may backfire, failing to protect the product safety. The reason is that on top of its ineffectiveness, the government may deter consumers from pursuing lawsuits (crowding-out), augmenting the malpractice of firms compared to an equilibrium without government intervention. Additionally, an improvement in government ability to detect ‘bad behavior’ should be complemented by a reduction of lawsuit cost or an improvement in the ability to detect ‘good behavior’ in order to restore consumer incentive to pursue lawsuits.

Booster shots for COVID-19 vaccines in the United States have undergone periodic updates to target circulating variants. However, the lengthy process of development, testing, and production has resulted lags between a variant’s initial detection and the corresponding booster's rollout, by which time the booster may no longer match the circulating variant. An innovation with realistic scientific potential—a universal COVID-19 vaccine, effective against existing and future variants—could provide much more value by preempting new variants. Averaged across Monte Carlo simulations, we estimate that, even under conservative assumptions on the arrival rate of variants, a universal COVID-19 vaccine would increase health benefits over variant-specific boosters by more than $900 billion. This social value eclipses the cost of an advance market commitment to incentivize the universal vaccine by several orders of magnitude.

During an infectious-disease epidemic, a political leader imposes “stay-at-home orders” (limiting activity) or “go-out orders” (mandating activity) whenever preferred by the majority of the citizenry over the no-intervention status quo. We characterize the resulting equilibrium epidemic trajectory in an economic-epidemiological model that allows for asymptomatic infection and social-economic complementarities of activity, assuming that citizens are myopic optimizers. We find that the qualitative features of equilibrium policy dynamics hinge critically on whether the pathogen is transmitted before or after infected people have developed symptoms. If transmission only occurs symptomatically, then the leader never imposes stay-at-home orders on the healthy but may impose go-out orders during some phases of the epidemic. However, if transmission occurs asymptomatically, the leader never imposes go-out orders on the healthy.

We introduce random encounter matching games, where a hidden trait is sometimes unwittingly passed—either private and bad (like counterfeit money), or collective and bad (a disease), or hidden and good (rare coin), or collective and good (information). One optimally expends effort to avoid acquiring bad traits or trying to acquire good ones. The game shifts from strategic complements to substitutes changing from private to collective, or good to bad traits. A unique equilibrium exists in all cases, and coincides for collective bad and private good traits, and collective good and private bad traits. Equilibrium incidence can be a misleading signal of prevalence for bad traits at high prevalence: one cannot infer that counterfeiting is less severe when there is less passed counterfeit money, or prevalence is higher with more infections.

The paper describes an auction for selling vaccines in a pandemic. The environment borrows from the problem of allocating positions for sponsored links on web pages with search results but recognizes the externalities that one man’s vaccination imposes on another. The auction is the pivot Vickrey–Clarke–Groves mechanism and, so, inherits its properties: efficiency and strategy-proofness. Crucially, the auction is designed to let each bidder bid not only on his own behalf but also on behalf of others. The auction requires neither the bidders nor the auctioneer to forecast the efficacy of the vaccine or the evolution of the pandemic.

This paper provides a quick survey of results on the classic SIR model and variants allowing for heterogeneity in contact rates. It notes that calibrating the classic model to data generated by a heterogeneous model can lead to forecasts that are biased in several ways and to understatement of the forecast uncertainty. Among the biases are that we may underestimate how quickly herd immunity might be reached, underestimate differences across regions, and have biased estimates of the impact of endogenous and policy-driven social distancing.

We introduce a new coalitional refinement of Nash equilibrium. An institution is defined to include: (i) a collection of active coalitions of players that can make joint deviations from any strategy profile; (ii) a partition of that collection, endowed with a priority order that determines which coalition is more important than another coalition; and (iii) a set-valued function that specifies which coalitions are informed about a joint deviation made by another coalition and can respond to that deviation by making a further deviation. The notion of institutional Nash equilibrium only considers joint deviations under the restrictions set by the institution under consideration. First, we reveal how our new equilibrium refinement is related to the existing coalitional refinements of Nash equilibrium. Second, after defining institutional enforceability as an availability of an equilibrium to be realized as a unique institutional Nash equilibrium for some institution, we characterize the conditions for an institutionally enforceable Nash equilibrium in two-player and three-player games.

We develop a search-theoretic model, in which a police agency allocates scarce resources across neighborhoods—heterogeneous in “vigilance” and valuables—to minimize crime, while potential criminals decide whether, and if so, when and where to commit a crime. When criminals sequentially search for the best target, the optimal police allocation depends on the difference in vigilance levels across neighborhoods, prioritizing neighborhoods with low vigilance. However, in the absence of criminal search, the optimal allocation depends on the degree of rent inequality among neighborhoods, with a priority placed on neighborhoods with higher rents. We also identify conditions under which policing all neighborhoods equally is optimal. Our findings underscore that an optimal policing design must not only consider neighborhood characteristics but also other factors that may impact criminals’ decision-making, including whether they engage in active search.

We study auctions in an affiliated private value framework in which buyers face ambiguity over the distribution of the others’ valuations and make choices that are robust to that ambiguity. We show that, in contrast to the Bayesian case, a first-price auction can generate a higher revenue than a second-price auction in the presence of ambiguity. We also extend this insight to large double auctions. In particular, we propose a pay-as-bid double auction mechanism and demonstrate that given any nontrivial amount of ambiguity, this mechanism implements the efficient trade and is ex-post budget balanced with probability one as the market size grows.