Trade and Contract Enforcement

We develop a simple model of trade with imperfect contract enforcement. Courts force the execution of defaulted contracts with parametric probability. In political economic equilibrium, the enforcement of contracts between agents of sovereign nations must be at a level mutually agreed between nations. The enforcement probability may be low, with the reluctant nation being associated with low levels of economic development in terms of model parameters. Low enforcement equilibrium can be a trap, as a better equilibrium may exist discretely far away. The model is replete with ‘paradoxical’ comparative statics, such as the paradox of immiserizing bargaining power. We define a tariff equivalent of imperfect enforcement which is useful in empirical investigations of trade and security. Preliminary version. Comments welcome. This paper began while Anderson visited the Chinese University of Hong Kong, whose support and hospitality is gratefully acknowledged. Earlier versions were presented to the NBER Summer Institute, August, 1999 and the Econometric Society, January 2000. Helpful comments from participants are gratefully acknowledged, especially Jonathan Eaton. Correspondence to: James E. Anderson Dept. of Economics Boston College Chestnut Hill, MA 02467 USA [email protected] Contractual insecurity appears to reduce international trade. The opinions of traders expressed in surveys suggest substantial insecurity and casual empiricism suggests its impact is significant. Formal inference from trade flow models shows that implicit transactions costs are large (McCallum, 1995; Roberts and Tybout, 1997). But it is difficult to assess the significance of contractual insecurity as opposed to other transactions costs or to realistically approach policy recommendations in the absence of a theory of contract insecurity. Imperfect contract enforcement is modeled in this paper as a parametric probability that courts will force the execution of a contract in default. Standard complete contract models, in contrast, provide no role for incremental enforcement effort --actions are either contractable or not due to natural elements. The model is used to suggest an explanation of the wide international variation in enforceability (World Economic Forum, 1997). Low enforcement presents something of a puzzle, since better enforcement is presumed to be good. Thus the World Bank has recently emphasized institutional quality to its clients. The incidence of poor institutions and the variety of institutional quality are explained by political economy in our model: traders on one side of the market will be harmed by better enforcement under plausible conditions. International trade contracts must be enforced across sovereign boundaries, hence enforcement is only as strong as both nations (and presumptively their traders) agree. Low enforcement can be a ‘trap’ as a better equilibrium may exist discretely far away. The correlation of poor institutions with poor countries is explained in terms of the model’s parameters. Our companion paper (Anderson and Young, 1999) examines the positive and normative consequences of middlemen such as trading monopolies who provide an alternative to formal contract enforcement. The ambiguous benefit of better enforcement is due to market failure. Agents without executable contracts match randomly on a spot market. On the excess side of the market, some agents will return home without matching. This 1 Logically, it is impossible to contract directly on the performance of the court. 2 Countries’ commitments to procedural ‘rules of the game’ must be self-enforcing. 2/3/00 Trade and Contract Enforcement 2 trading inefficiency is directly increased by better enforcement (fewer scarce side traders are in the spot market ex post) but indirectly reduced (more scarce side traders enter into trade ex ante). Excess side traders’ interest balances these two forces. Scarce side traders, in contrast, always support better enforcement. Paradoxical possibilities abound in the model. Traders on the excess side of the market may be subject to the paradox of scarcity: the larger the potential gains from trade, the less is the volume of trade. Poorer countries may have more intense need for imports (being more different from the exporters) but with bad institutions they may trade less. Excess side traders may also have the ‘paradox of immiserizing bargaining power’, losing from a parametric increase in their bargaining power. Imperfect enforcement impedes trade similarly to tariffs, so it is useful to compare and connect contract enforcement with tariffs. We set out a tariff equivalent of enforcement and show it is decreasing in enforcement effort. Thus we provide a rationale for empirical work on resistance to trade based on poor institutions (Anderson and Marcouiller, 1999). A sequel paper (Anderson, 2000), relates contract enforcement and trade taxation. As with a tariff, it is not obvious that better enforcement is always in a nation’s interests (or its politically active agents’ interests). As with a tariff, this is due to the effect of enforcement on international and domestic prices. Section 1 sketches the model. Section 2 develops the basic setup and qualifications. Despite the simplicity of the model, we claim its qualitative features are robust to most reasonable extensions. Section 3 sets out the tariff equivalent of imperfect enforcement. Section 4 presents the comparative static analysis with respect to enforcement. Section 5 considers the political economic equilibrium of enforcement. Section 6 concludes with a summary and some discussion of extensions. The first appendix reports simulation results and the second appendix provides proofs of propositions and reports comparative static analysis with respect to the bargaining power and the mean outside options of buyers and sellers. 3 Poor enforcement may also be explained by its cost. Fixed costs of good enforcement may be too large for the gains from trade to cover. We do not reject this cost side explanation, but abstract 2/3/00 Trade and Contract Enforcement 3 1 Sketch of the Model The simplest model of imperfect contract enforcement and trade minimally must contain four elements. First, trade must incur ex ante costs which are sunk at the time of exchange. This provides a role for contracts in overcoming the holdup problem --ex post bargaining over price which ignores sunk costs. Second, there must be costly search in the absence of an enforced contract. This is because with costless search, contracts are unnecessary. Each agent’s outside option in bargaining with any partner is to keep searching and obtain the most favorable price available to any agent from his side of the market, which in equilibrium is the standard competitive price. We assume search takes the form of random matching. Third, there must be a random component in agents’ outside options. This is because rational agents must sign contracts which will be broken when a random shock makes default preferable for some agent. In our model the outside option is shifted by a random component of trade cost which affects the ex post cost of returning home to transact. Fourth, imperfect enforcement requires that contract default is sometimes not remedied by the court. In an environment with these four elements, agents facing broken contracts will have an incentive to search rather than renegotiate. This is because default by an agent from the other side of the market reveals a more favorable than average outside option. Our setup thus generally produces a spot market along with a futures market. In order to focus on trading inefficiency the model takes as given the market prices which provide the ex ante outside options to buyers and sellers. This simplification is justified, we argue below, because it cleanly reveals mechanisms which must be at work in larger models in which home and foreign market prices are endogenous. from it to concentrate on the hitherto neglected but important benefit side. 4 If in contrast we rule out search entirely, all broken contracts result in renegotiation. This seems a less interesting and relevant alternative to us because it cannot yield an explanation of trade as contracts become useless. See our companion paper where anarchic trade is based on random matching. The no-search alternative retains some of the flavor of our model in that improvements in enforcement have ambiguous effects on the traders, hence better enforcement is not always supported. 5 Failure to enforce is based on the realization of another independent random variable. 2/3/00 Trade and Contract Enforcement 4 Traders in our model must incur trade costs prior to exchange. We think of the cost as primarily reflecting information acquisition and we assume traders are heterogeneous in their ability to become informed about foreign trade. The sunk cost also includes shipment and design costs common to all traders. In the event of a failure to trade, traders must incur reverse shipment and design costs to return to their domestic markets. Crucially, trade costs are subject to idiosyncratic shocks discovered only after committing to trade, hence the traders’ outside options are affected by the cost shock. Agents without enforced contracts match once from the set of opposite side traders and if matched they bargain over the exchange price in the spot market. Each agent’s outside option in the bargaining is his net price at home, the home market price net of the trade cost shock. The bargained price is a convex combination of the outside options. In the forward market, agents have an incentive to contract on price prior to committing their trade costs, since the bargained price cannot reflect sunk trade costs. We assume costless search in the forward market. With perfect contract enforcement, trade will be efficient: all agents whose gain from trade exceeds their expected trade cost will enter trade and all trade will be executed. Contracts are incomplete because it is impossible to contract on the performance of the enforcement system. Imperfect contract enforcement is 6 Sunk trade costs are well documented empirically (Roberts and Tybout, 1997; Bernard and Wagner, 1998). 7 A more realistic but inessentially complicated story is that trading firms hire a heterogeneous supply of labor, so that the rents accrue to high ability workers. 8 Assuming idiosyncratic shocks implies that risk diversification is possible, hence justifying risk neutrality. The cost shock plausibly includes common elements (such as exchange rate movements) but idiosyncratic elements are highly plausible. With risk neutral traders, only the reverse shipment aspect of the cost shock plays any role in the model, and here the need to quickly arrange redesign and shipment home seems likely to meet idiosyncratic cost. 9 This simplification of a more general costly search model is extremely convenient. The model developed here is related to the search literature founded on Diamond (1981, 1982). With search friction there is a ‘thick-market externality’ which resembles our transactions externality (but is opposite in sign). Holdup occurs naturally with search frictions (Acemoglu and Shimer, 1999). We initiate an imperfect contract in this setting. Our random matching setup mirrors the treatment by Casella and Rauch (1998) of informational inefficiency in a trade model. 10 In contrast, the complete contracts literature analyzes optimal (and implicitly perfectly enforceable) contracts subject to limits on what specific attributes of a transaction can be contracted. See McLaren (1998) for interesting work analyzing trade and contract limitations. 2/3/00 Trade and Contract Enforcement 5 parameterized as the probability that a contract in default will be enforced, taken to reflect the rules and procedures governing enforcement in the event of default. Agents choose to default when they receive favorable outside options. The probability of execution is equal to the probability of no default plus the product of the probability of default and the probability of enforcement, assuming the random processes of enforcement and trade costs are independent. Equilibrium entry into trade occurs when the marginal traders are able to cover expected trade cost by the expected gain from trade, the expected price less the home market price. The expected price facing traders prior to their decision to enter trade is a compound of the contract price and the spot market price, with the latter being a compound of the expected bargained price and the outside option for those who fail to match. Our model generally reaches an equilibrium where ex ante supply and demand are not equal. Since there are gains from trade, transactions are inefficient and costless improvements in enforcement would be socially efficient. When enforcement is endogenously chosen, enforcement decisions occur logically prior to contracting or trading. Forward-looking traders on each side of the market seek the institutions of contract enforcement which maximize their expected economic returns. In the context of international trade, sovereign nations will agree to the strongest enforcement (institutions) which the traders from both sides of the market will support. We show below that traders from the scarce side of the market always gain from improved enforcement. Their position is identical to traders facing a trade barrier. On the excess side of the market, however, traders’ interests in better enforcement are positively linked to the probability of a match, which in turn depends on the balance of two factors. First, a rise in the enforcement probability reduces the number of partners from the scarce side who will be seeking a match on the spot market, tending to lower the probability of a match. But second, better enforcement improves the price for traders from the scarce side of the market, induces more participation and hence tends to raise the probability of a match. The incomplete contract approach can be simultaneously criticized as hand-waving and defended as realistic (Tirole, 1999). We use it for tractability and discuss robustness of the 2/3/00 Trade and Contract Enforcement 6 Conditions are provided in which political economic equilibrium emerges in a low enforcement trap which condemns the agents to lower welfare than is possible. The model yields low enforcement equilibrium when arameter values associated with low stages of economic development are present. In contrast to ‘big push’ models of low income traps (e.g. Murphy, Shleifer and Vishny, 1989), the poverty trap arises in the absence of fixed costs or coordination failures. 2 Equilibrium with Imperfect Contract Enforcement We develop a model of equilibrium exchange in contracts on the forward market and in executed trade on the spot market. Each trader transacts one unit of the good in exchange for money, with each exchange accounting for an infinitesimal share of the (spot or forward) markets. Here and the next two sections, contract enforcement is exogenous. All agents are risk neutral and differ in the deterministic component of their trade cost, reflecting differential abilities to become informed about foreign markets. They commit to trade if the expected cost is weakly less than the gain expected from trade, the expected trade price less the expected home market net price (outside option). The expected price a trader receives is built up from the component prices which emerge from spot market bargaining and from contract execution at contract prices determined in the forward market. Decision makers solve backwards from the spot market to the default decision and its outcome to the forward market to the decision to enter trade. In equilibrium the subjective probability of matching used to determine expected payoffs is equal to the objective probability resulting from the numbers of agents who enter on each side of the market and either default or honor their contracts. Contracting in the forward market occurs first in time. In this market all traders can costlessly search, and all participants expect to at least break even on a contract. Next, traders sink their trade costs. Then they discover the random component of trade cost which affects their outside options. At this point the traders consider whether to default on the contract. Default is chosen only by excess side traders under our simplifying assumption that restricts the support of model to generalization below. 2/3/00 Trade and Contract Enforcement 7 the distribution of scarce side trade cost shocks. Default by the excess side occurs when the shock falls below a critical value μ determined below. Default occurs with probability 1 −φ(μ ); φ ∈(0,1) . The partner of a defaulter appeals to the court , which enforces a portion of contracts in default θ; θ ∈[0,1]. The probability of execution of a contract is: (2.1) β = φ + θ(1−φ) . Agents whose partners default know their partners have better outside options than the average spot market trader, so if the court does not enforce the contract they are better off entering the spot market and making a random match. On the spot market, all matches result in Nash bargaining over the price, incorporating the random shocks to outside options. Finally exchange occurs: some with executed contracts at the contract price, some with matches at the bargained prices and some with unmatched traders going home to trade at their outside options. We set out the decision tree of the excess side traders in Figure 1. The scarce side traders face the same stages of decision, their tree is less complicated (they always sign contracts and hey never default because they never receive a sufficiently favorable shock). In Figure 1, branches from a point represent decisions by the trader while branches from a horizontal line represent externally imposed outcomes. Contracting traders can formally decide not to trade, but there is a penalty which makes this a dominated decision. The conditional expected prices are listed at the bottom, p for an executed contract, p for a bargained price on the spot market between randomly matched traders, and b or c for return to the domestic market without matching. The match probability is π . 11 We might naturally think of the court as being in the defaulter’s country. Only a court in the defaulter’s country is able to seize assets of the defaulter to compel execution. But even a court in the plaintiff’s country is bound by international conventions in its treatment of foreign citizens who might happen to have assets which can be seized. We present a reduced form ‘court’, but note that a structural model of international commercial law must utimately explain conventions by mutual self-interest of the kind we feature. 2/3/00 Trade and Contract Enforcement 8