Since the canonical contribution by Coase (1937), the so called ‘make-or-buy decision’ has become one of the fundamental research questions in the economic discipline. Why do firms integrate certain activities into firm boundaries rather than source them from independent suppliers? In view of the prominent role of multinational firms in the modern economy, this question has been increasingly studied in the international context to understand global production and trade patterns. A multinational firm’s make-or-buy decision was shown to depend, among other things, on the factor intensity of the production process, firm productivity, and a country’s contractual environment (see Antràs 2015 for an overview). Our research contributes to this literature strand by arguing that culture is a neglected but important determinant of a multinational firm’s global sourcing decisions.
Culture as the missing link
Whenever managers of multinational companies are asked about the challenges of cross-border collaborations, keywords like ‘cultural differences’ or ‘intercultural communication’ are among the most frequently given answers (Economist Intelligence Unit 2012). We formalise this notion and bring its predictions to the data (Gorodnichenko et al 2015). In our model, headquarter managers decide whether to source intermediate inputs from independent suppliers or integrate the latter into firm boundaries against the backdrop of private costs associated with cross-cultural collaboration. When the managers of a headquarter’s and a supplier’s unit have different visions or beliefs about the right course of actions, the make-or-buy decision entails a key trade-off. On one hand, integration of a supplier into a firm’s boundaries leads to better coordination of decisions across firm units, higher final goods quality and larger managerial profits. On the other hand, managers of an integrating firm also suffer a private cost of enacting their visions in the integrated supply unit due to conflicting visions between managers in the headquarter unit and managers in the supplying firm (Hart and Holmström 2010). We assume that these non-monetary costs are increasing in the cultural distance between managers of those two units. This leads to the prediction that the relative prevalence of vertical integration decreases in the cultural distance between the home and the host country. Take a US multinational firm. Our model implies that it is more likely to be vertically integrated with an English or Swedish supplier than with a Japanese or Chinese supplier, everything else equal.
How do we bring this prediction to the data?
As a proxy for the relative prevalence of vertical integration, we use the measure of intra-firm trade from the ‘US Related Party Trade’ product-level data, collected by the US Bureau of Customs and Border Protection. Our key explanatory variable is a country’s cultural distance to the US, as measured by Hofstede’s (2001) well-known individualism-collectivism index, widely regarded as the major cultural cleavage across countries.1 As we can see from Figure 1, there is a strong negative correlation between a country’s cultural distance to the US and the share of US intra-firm imports from this country.
Figure 1. Share of intra-firm imports and cultural distance from US
We find that a standard deviation change in the level of cultural distance is associated with a 6.4 percentage point change in the share of intra-firm imports. This is a considerable magnitude given that the average share of intra-firm imports in the database we use is 19.92 percent. To ensure that this relationship is not driven by omitted variables, we control for a wide range of explanatory factors that were shown to have a significant effect on a multinational firm’s make-or-buy decision (cf. Antràs 2015). In particular, we control for the effect of trade costs across countries as well as of productivity dispersion, demand elasticity and variables affecting the well-known holdup problem between firms. While the effect of these control variables is broadly consistent with the existing findings, cultural distance continues to have a negative, significant and important effect on intra-firm trade.
Is the effect of cultural distance causal?
To better understand if the effect of cultural distance on intra-firm imports is causal, we consider two additional identification strategies. First, we apply the instrumental variables approach, whereby cultural distance is instrumented by the genetic distance between the population in a given country and the population in the US (Gorodnichenko and Roland 2011). Given that genetic distance is exogenous to firms’ make-or-buy decisions, this instrument satisfies the exclusion restriction. The instrumental variables estimates confirmed the OLS results, suggesting that higher cultural distance leads to less intra-firm trade.
In the second step, we refine our cultural distance between foreign countries and US industries. More specifically, we use information on an individual’s ancestry from the 2000 US Census data to calculate ethnic composition of managers in US industries and weigh these ethnic shares with the individualism level of their ancestors’ country of origin to compute industry-specific cultural scores. Controlling for country and year fixed effects, we continue to find a negative and significant effect of cultural distance on the intra-firm trade.
While courses on intercultural communication and management are indispensable components of most (if not all) business programs around the world, economists’ understanding of cultural effects on international production and trade patterns is still in its infancy. Our research contributes to this knowledge by shedding some light on the role of culture in a multinational firm’s make-or-buy decisions. In order to obtain a profound understanding of cultural effects in international exchange, however, more research needs to be done in the future. Exploiting available information on cultural heritage of multinational firms’ managers in combination with the data on those firms’ international transactions might constitute a fruitful research agenda.
Antràs, P (2015) Global production: Firms, contracts and trade structure, Book manuscript.
Coase, R (1937) “The Nature of the Firm”, Economica, 4(16): 386-405.
Economist Intelligence Unit (2012) Competing across borders: How cultural and communication barriers affect business, EIU report.
Gorodnichenko, Y, B Kukharskyy and G Roland (2015) “Culture and global sourcing”, NBER Working Paper, No. 21198.
Gorodnichenko, Y and G Roland (2011) “Which dimensions of culture matter for long-run growth?”, American Economic Review: Papers and Proceedings, 101(3): 492-498.
Hart, O and B Holmström (2010) “A theory of firm scope”, Quarterly Journal of Economics, 125: 483-513.
Hofstede, G (2001), Culture’s consequences: Comparing values, behaviors, and organizations across nations, 2nd edition, Sage Publications
1 As the US is the most individualistic country in Hofstede’s database and, since we are looking at cultural distance between the US and its trade partners, the individualism-collectivism index proves to be a convenient proxy. Our results, however, are robust to the inclusion of further cultural dimensions (e.g., power distance, masculinity, and uncertainty avoidance) in our definition of cultural distance.