Matt Elliott


Contact Information:

Division of the Humanities and Social Sciences 228-77
California Institute of Technology
Pasadena, California 91125



I am Assistant Professor of Economics at Caltech, on leave at Cambridge where I am a University Lecturer.


Commitment and (In)Efficiency: a Bargaining Experiment (with Marina Agranov, Draft date: January 2016) (Supplementary Appendix)

We conduct an experimental investigation of decentralized bargaining over the terms of trade in matching markets. We study if/when efficient matches will be reached and what terms of trade will be agreed. Multiple theories guide our analysis and we test their predictions against the outcomes of our experiments. We find that inefficiencies are extensive and they are driven by the endogenous evolution of bargaining positions as agreements are reached and players exit. When we allow subjects to renege on existing agreements efficiency significantly improves.


Decentralized Bargaining: Efficiency and the Core (with Francesco Nava, Draft date: January 2016)

This paper studies market clearing in matching markets. The model is non-cooperative, fully decentralized, and in Markov strategies. Workers and firms bargain with each other to determine who will be matched to whom and at what terms of trade. Once a worker-firm pair reach agreement they exit the market. Alternative possible matches provide endogenous outside options. We ask when do such markets clear efficiently and find inefficiencies — mismatch and delay — to be pervasive. Mismatch occurs whenever an agent is at risk of losing a binding endogenous outside option. Delay occurs, instead, when the market evolves in favor of an agent. Delay can be extensive and structured with vertically differentiated markets endogenously clearing form the top down.


Financial Networks: Efficient Modularity and Why Shareholders Prevent It

(with Jonathon Hazell, Draft date: February 2016) (Supplementary Appendix)

We model the implications for systemic risk in financial networks of the classic conflict of interest between debt-holders and equity-holders. By trading with each other banks can diversify their idiosyncratic risks and avoid failures following small shocks. However, the resulting financial interdependencies create systemic risk and the possibility of multiple failures following large shocks. A social planner resolves this trade-off by creating a modular network structure with fire breaks, thereby diversifying away banks' exposures to small shocks while containing contagion. These socially efficient networks suit debt-holders' interests, but run counter to equity-holders' interests. They transfer surplus value from the equity-holders of healthy banks to the debt-holders of distressed banks. Equity-holders can profitably trade away from these networks, and so socially optimal networks are unstable. Moreover, trades are profitable for equity holders when they align counter-parties' failures with their own, creating systemic risk.


Firm Capabilities and Industry Structure (with Jun Chen, Draft date: March 2016)

We model firms as sets of scarce capabilities, where each capability provides a source of competitive advantage in some markets. Each market is also associated with a set of capabilities, those that are valued by it. Firm and market hypergraphs represent this information. Our approach provides a new perspective on several industrial organization literatures including merger analysis, strategic alliances and industry dynamics.  We argue that merger analysis should be more holistic and that profitable joint ventures increase consumer surplus even when they reduce competition. We also provide formal foundations for a prominent theory of competitive advantage in the management literature.


Social Investments, Informal Risk Sharing, and Inequality

(with Attila Ambrus and Arun Chandrasekhar, Draft date: July 2015)

This paper studies the formation of risk-sharing networks through costly social investments, identifying endogenous inequality and a misallocation of resources as possible downsides to such agreements. First, individuals invest in relationships to form a network. Next, neighboring agents negotiate risk-sharing arrangements, in a generalized version of the model in Stole and Zwiebel (1996). This results in the social surplus being allocated according to the Myerson value. In particular, more centrally connected individuals receive higher shares. We find a novel trade-off between efficiency and equality. The most stable efficient network, which minimizes incentives to overinvest, also generates the most inequality. When individuals are split into groups and relationships across groups are more costly but incomes across groups are less correlated, there is never underinvestment into social connections within group, but underinvestment across groups is possible. More central agents have better incentives to form across-group links, reaffirming the efficiency inequality trade-off. Evidence from 75 Indian village networks is congruent with our model.


Ranking Agendas for Negotiations (with Ben Golub, Draft date: February 2015)

Consider a negotiation in which agents will make costly concessions to benefit others -- e.g., by implementing tariff reductions, environmental regulations or nuclear disarmament. An agenda specifies which issue or dimension agents will make concessions on; after an agenda is chosen, the negotiation comes down to the magnitude of each agent's contribution.  We seek a ranking of agendas based on the marginal costs and benefits they each generate at the status quo, which are captured in a Jacobian matrix. In a transferable utility (TU) setting, there is a simple ranking based on the best available social return per unit of cost (measured in the numeraire). When transfers are not available, the problem of ranking agendas is more difficult, and we take an axiomatic approach. First, we require the ranking not to depend on economically irrelevant changes of units. Second, we require that the ranking be consistent with the TU ranking on problems that are equivalent to TU problems in a suitable sense. The unique ranking satisfying these axioms is represented by the spectral radius (Frobenius root) of a matrix closely related to the Jacobian, whose entries measure the marginal benefits per unit marginal cost agents can confer on one another.


A Network Approach to Public Goods (with Ben Golub, Draft date: December 2015)

R&R, Journal of Political Economy

Suppose each of several agents can exert costly effort that creates nonrival, heterogeneous benefits for some of the others. How do negotiated outcomes depend on the heterogeneities? To study this question, we construct a matrix---or a weighted, directed network---that describes the marginal benefits agents can confer on one another. We first show that an outcome is Pareto efficient if and only if the largest eigenvalue of the marginal benefits matrix evaluated at that outcome is equal to 1. A corollary describes the players whose participation is essential for any Pareto improvement. We then show that an agent's contribution in any Lindahl equilibrium corresponds to his eigenvector centrality in the benefits network. This provides a new market foundation and interpretation for widely-used network statistics, and, conversely, a network perspective on price equilibria. Finally, we discuss strategic foundations for Lindahl outcomes in our setting, explaining when negotiations will result in contributions that correspond to network centralities.

(This paper includes results previously circulated in a paper titled “A network centrality approach to coalitional stability”)



Financial Networks and Contagion (with Ben Golub and Matt Jackson)

American Economic Review, 2014.

We model financial contagions and cascades of defaults among organizations that have a network of cross holdings. We first identify a network-based measure that captures the impact of changes in one organization's value on other organizations' values.  We use the measure to study both integration (the increasing of cross holdings) and diversification (the spreading out of cross holdings).  We show that diversification initially increases the probability and extent of cascades as a network of interdependencies grows, and eventually the probability and extent of cascades decreases once organizations become less tied to specific other organizations.  Integration also faces tradeoffs: increased dependence on other organizations versus less sensitivity to own investments. We briefly discuss incentives to seek bailouts, and associated moral-hazard issues. We also show that once an organization approaches a bankruptcy threshold, there are no trades of cross holdings or assets at fair prices that can lower the probability of its failure, and that unduly favorable trades for that organization and/or a direct injection of capital are necessitated. Finally, we illustrate some aspects of the model with European debt cross holdings.


Heterogeneities and the Fragility of Labor Markets (Draft date: Nov 2015)

Workers' labor market participation decisions and firms' vacancy creation decisions are studied in a model where different matches generate different surpluses. An immediate consequence of these heterogeneities is that better matches are possible in thicker markets. This creates a thick market externality: when additional workers and firms enter the market, they confer net benefits on the other workers and firms by improving the expected quality of their matches. As a consequence, there is always too little entry by both workers and firms. The thick market externality has further implications. Quite generally labor markets will be fragile. Considering shocks to average match productivities, there will be a critical threshold at which a labor market suddenly collapses from supporting multiple workers and multiple firms in equilibrium to supporting no workers or firms in any equilibrium. All but one agent will suffer discontinuous losses as this threshold is passed and the market collapses.


Inefficiencies in Networked Markets   (Draft date: June 2014) (Supplementary Appendix)

American Economic Journal: Microeconomics, 2015.

In many markets, relationship specific investments are necessary for trade. These formed relationships constitute a networked market in which not all buyers can trade with all sellers. We show that networked markets can be decomposed to identify how alternative trading opportunities affect who trades with whom and at what price. This uncovers agents’ investment incentives. In some markets a buyer and seller must make different, separate investments to trade, while in others investments are jointly negotiated. Either way, inefficiencies can be severe and consume all the gains from trade, but for different reasons. We consider three applications in detail: high-skill labor markets, merger markets when industries are consolidating, and the international market for natural gas.


How Better Information Can Garble Experts' Advice

(joint with Ben Golub and Andrei Kirilenko, Draft date: June 2012)

We model two experts who must make predictions about whether an event will occur or not. The experts receive private signals about the likelihood of the  event occurring, and simultaneously make one of a finite set of possible predictions, corresponding to varying degrees of alarm. The information structure is commonly known among the experts and the recipients of the advice. Each expert's payoff depends on whether the event occurs, her prediction, and possibly the prediction of the other expert. Our main result shows that when either or both experts receive uniformly more informative signals, their predictions can  become unambiguously less informative. We call such information improvements perverse. Suppose a third party wishes to use the experts' recommendations to decide whether to take some costly preemptive action to mitigate a possible bad event. The third party would then trade off the costs of two kinds of  mistakes: (i) failing to take  action when the  event will occur; and (ii) needlessly taking the  action when the event will not occur. Regardless of how this third party trades off the associated costs, he will be worse off after a perverse information improvement. These perverse information improvements can occur when each expert's payoff is independent of the other expert's predictions and when the information improvement is due to a transfer of technology between the experts.