Relational contracts and supplier turnover in the global economy

Fabrice Defever, Christian Fischer, Jens Südekum

24 October 2015



Intermediate inputs account for a substantial share of global trade. A large chunk of this trade involves a buyer – often a large and powerful headquarter firm located in a high-income country – who imports components from a foreign supplier for further processing. These suppliers are often located in low-wage countries with relatively unfamiliar market conditions and weak legal institutions, and economic exchange in such environments is often hampered by search and contractual frictions.  

This creates two important hurdles to contracting:

  • The buyer needs to find a supplier who is technologically capable of producing the desired input in appropriate quantity and quality and at low costs; and
  • The buyer needs to make sure that this supplier does not frequently renege on the agreed terms, engage in haggling, and so on.

Finding such an efficient and reliable partner is costly and can involve a time consuming trial-and-error process before the firm is finally satisfied with the match.

How can hold-up problems and the resulting inefficiencies be alleviated when formal contracts are not enforceable? The seminal work by Baker et al. (2002) suggests that relational contracts may do the trick. These are purely informal and trust-based agreements to cooperate in an enduring relationship, where parties do not behave opportunistically even if they could. And indeed, a vast business literature suggests many headquarter corporations and specialised component suppliers do have a strong interest to engage in such informal cooperation, which allows them to create and share relational rents. Quite surprisingly, however, this aspect has not been analysed much in the economics literature on the international organisation of value chains. The prominent approach by Antràs and Helpman (2004), for example, has only analysed static one-shot constellations, but it has not addressed how relational contracts might resolve underinvestment problems in the context of input sourcing.

A model of relational contracts with supplier re-matching

In a recent paper (Defever et al. 2015), we develop a dynamic property rights model of global sourcing. The firm makes two decisions in every period. First, observing the current supplier's efficiency, she decides whether to engage in a costly search for another, potentially better match. Second, she can promise her current supplier a bonus payment if he ‘behaves nicely’ and reliably provides the relationship-specific input.

Our main theoretical insight is that the possibility to switch partners crucially affects the contractual nature of buyer-supplier relationships. In fact, the headquarter faces a trade-off; it can either stick with its supplier, which in return may behave nicely, or look for a new (potentially better) one, which may lead its current supplier to behave badly. In other words, we uncover a causal relationship between match durations and the stability of relational contracts. Once the firm decides to stop searching, and thus to launch a long-term collaboration, this will then cause the emergence of a relational contract.

Evidence from China

Putting the full micro-structure of our model to an empirical test is challenging because current data only rarely allows observing particular matches of domestic buyers and foreign suppliers. Moreover, even if they were known, we typically cannot observe the detailed explicit and implicit match-specific arrangements between the buyer and the supplier on which our model makes sharp predictions. Still, despite those limitations, we exploit the unique features of Chinese customs data on export transactions with the US (2000-2006) to address the empirical relevance of our theory. This context is well suited for our purpose because the paradigm of incomplete contracts appears to be quite plausible when it comes to US input sourcing from China (Antràs 2015, Manova and Zhang 2012).

We build a sample of Chinese firms that start exporting a particular product to the US. Following these fresh Chinese exporters over time, we observe which firms still export the same product to the US after a few years, as opposed to those who terminated that exporting activity in the meantime. This allows us to observe whether a particular export transaction was a one-shot deal or a long-term collaboration.

To see if the respective Chinese-US transaction can be characterised as a relational agreement, we then exploit the information if the Chinese supplier operates under an ordinary or a processing trade arrangement with his US partner. Particular types of these processing arrangements capture the essence of relational contracts fairly well, for example, if the US headquarter has provided its Chinese supplier with specifically designed equipment. We therefore use processing trade (or specific subcategories) as our empirical proxy for relational contracts.

Table 1. Duration of processing and non-processing transactions

Table 1 gives an overview of these data. It shows that only 28.4% of all ordinary trade arrangements last for 3 years or more. By contrast, almost 60% of all processing trade arrangements (i.e., relational contracts) last for more than 3 years.

At the industry level, we consistently find that the share of long-term collaborations and relational contract prevalence are positively correlated across product categories. This correlation is in line with the predictions of our theory, and suggests that more enduring buyer-supplier relationships are more likely to involve elements of relational contracting.

Empirical strategy

Finally, we use the structure of our theoretical framework to develop an instrumental variable estimation approach that uncovers the causal effect of long-term collaborations on the prevalence of relational arrangements. In particular, our model shows that the distribution of suppliers' unit costs is a valid instrument for the length of match durations. The underlying intuition is that more cost dispersion in an industry makes search more attractive, essentially because it raises the benefits of having a good supplier with low unit costs. Yet, more cost dispersion does not directly affect the contractual nature of firm-supplier, but there is only an indirect (negative) effect via the reduced match durations in the time span of observation.

Put differently, our estimation strategy exploits the prediction of our theory that the supplier's unit costs are negatively correlated with the duration but uncorrelated with the contractual nature of collaboration. Applying this estimation approach with our Chinese-US data, we find that cost dispersion is indeed a good predictor for supplier turnover. Going from an industry with only one-shot to an industry with only repeated interactions, leads to a sizeable increase of relational arrangements by 38 to 67 percentage points.


Weak legal institutions and contract incompleteness can create severe inefficiencies. However, firms and suppliers have a vital interest in overcoming such hold-up and underinvestment problems via informal arrangements. Whether this works in equilibrium depends crucially on the sectoral characteristics. A larger heterogeneity in the suppliers’ characteristics makes the search and matching process last longer, which results in a larger proportion of short-lived inefficient contracts.

From a policy perspective, our research generally subscribes to the view that an improvement of contract enforcement (and legal institutions more generally) in the source country has the potential to raise efficiency and welfare. However, even if institutions in the south exhibit inertia and cannot be changed immediately, it is still possible to establish and to maintain efficient relationships with suppliers from those countries. The reason is, simply, the self-interest of businessmen to ‘behave nicely’ in mutually beneficial collaborations.


Antràs, P (2015), “Global Production: Firms, Contracts and Trade Structure”.

Antràs, P and E Helpman (2004), “Global sourcing”, Journal of Political Economy 112, 3.

Baker G, Gibbons, R., and Murphy, K.J. (2002), “Relational contracts and the theory of the firm”, The Quarterly Journal of Economics 117, 1.

Defever F, C Fischer, and J Suedekum (2015), “Relational contracts and supplier turnover in the global economy”, CEPR Discussion Paper 10784.

Manova, K and Z Zhang (2012), “Export prices across firms and destinations”, The Quarterly Journal of Economics 127, 1.



Topics:  International trade

Tags:  contracting, relational contracts, global sourcing, supplier turnover, supplier, China search

Lecturer at the University of Nottingham

Doctoral Researcher at the Düsseldorf Institute for Competition Economics, Heinrich-Heine University Düsseldorf

Professor of International Economics at the Düsseldorf Institute for Competition Economics, Heinrich-Heine University Düsseldorf; Research Fellow, CEPR