Economists and other social scientists have long documented a fascinating fact: the diffusion of knowledge strongly decays with distance. Jaffe et al. (1993) were among the first to make this claim when they found that patent citations occur with much more frequency within the same metropolitan area.
Other studies include Keller (2002), who shows that productivity spillovers for domestic firms originated by foreign R&D investments are halved at a distance of 1,200km or greater. In Bahar et al. (2014), my co-authors and I show that countries are more likely to gain comparative advantage, from scratch, in the same products that are competitively exported by geographical neighbours.
Multinational corporations (MNCs) constitute a good setting to study knowledge diffusion as their workers are constantly communicating and sharing knowledge, often across different affiliates in different parts of the world. Since firms struggle to communicate with far-away subsidiaries (e.g. Oldenski 2012, Giroud 2013, Gumpert 2018), the extent that knowledge diffusion is strongly limited by distance should be reflected in the location decisions of MNCs. And this is consistently the case: MNCs tend to locate affiliates in knowledge-intensive economic activities nearby to the headquarters.
This ‘knowledge–distance’ trade-off is consistent with previous literature that assumes that transferring knowledge is costly and increasing with distance (Ramondo and Rodriguez-Clare 2013, Ramondo 2014, Arkolakis et al 2013, Tintelont 2017). Keller and Yeaple (2014) stand out by providing a particular mechanism to rationalise the knowledge–distance trade-off: the shipping of intermediate goods between headquarters and affiliates, which tend to be more prevalent for knowledge-intensive economic activities (thus raising costs that depend on distance).
In a new study (Bahar 2019a), I look at the knowledge–distance trade-off and suggest an alternative explanation: the ability of workers within a firm to communicate in real time as proxied by time zone differences. This would be consistent with what was pointed out by Arrow (1969) when he suggested that the transmission of knowledge is difficult and costly mainly because it requires human interaction.
My main finding is that the knowledge–distance trade-off is significantly weakened when accounting for overlap in working hours between the headquarters and its (knowledge-intensive) subsidiaries. Thus, if ceteris paribus, time zones affect the location decision of MNCs, then trade costs of intermediate goods cannot serve as a sufficient explanation for the knowledge–distance trade-off.
Stylised facts on time zones and the knowledge–distance trade-off
The empirical exercise I conduct is based on a sample of about 70,000 domestic and 45,000 foreign horizontal subsidiaries active in the manufacturing sector and belonging to over 3,200 MNCs from the Worldbase dataset by Dun and Bradstreet (see Figure 1 for the mapping of these headquarters and affiliates).
I complement this with an industry-level knowledge-intensity measure based on worker-level characteristics for each industry, averaging the accumulated experience and training of the workforce in an industry, using occupational characteristics defined in the O*NET project dataset (see Bahar 2019b for more details on this measure and for a comparison with other measures used in the literature).
Figure 1 MNC headquarters (triangles) and affiliates (dots) in sample
Figure 2 shows that, indeed, the estimated knowledge–intensity trade-off is flatter for headquarters and affiliate pairs that have a higher overlap in working hours. The left panel estimates – with a number of controls including MNC fixed-effects – the slope of the trade-off for different levels of overlap in working hours between the affiliate and the headquarters, while the right panel plots the estimated trade-off (lighter lines represent the trade-off when there is higher working hours overlap).
While knowledge-intensive economic activities tend to locate geographically closer to the headquarters, this is not the case when there is a higher overlap in working hours between the two. Thus, shipping costs of intermediate goods is not the only explanation for the trade-off, since shipping goods is costly both from east to west as well as from north to south. The relationship holds for all type of subsidiaries, or when limiting the sample to only foreign subsidiaries (overall or those defined as horizontal or as vertical).
Figure 2 Distance–knowledge trade-off by overlapping working hours
Exploiting distance to time zone line: A regression discontinuity approach
In order to rule out other confounding factors driving the time zone result, I estimate a regression discontinuity using discrete changes in the overlap of working hours across time zone lines around the globe. By doing so, I effectively compare the knowledge intensity of subsidiaries that belong to the same MNC but that are on different sides of a time zone line. To the best of my knowledge, using a regression discontinuity design exploiting time zone differences is novel in the context of international economics, and in particular when studying patterns of behaviour of multinational corporations.
As shown in Figure 3, using this methodology I find evidence that subsidiaries with a higher overlap in working hours with their headquarters are, on average, active in industries that are more intensive in knowledge. I find, however, that this discontinuity is present only for foreign horizontal subsidiaries, and not for other subsidiary types (the paper discusses some of the possible explanations for this results).
Figure 3 Regression discontinuity plot
The regression estimations find that the economic activity of a foreign horizontal subsidiary located just across a time zone line closer to the headquarters – thus increasing the overlap in working hours – is 0.6% to 0.84% higher on the knowledge-intensity scale. The average effect of one more hour of overlap is equivalent to a reduction of about 200km between a headquarters and its foreign subsidiary, based on the estimated knowledge and distance trade-off.
I further show that the discontinuity cannot be explained by observable determinants of the location decision of MNCs, and the results are robust to using different bandwidths, weights, and alternative knowledge-intensity measures such the industry-level R&D intensity.
An interpretation of this finding is that not all knowledge can be fully embedded in intermediate goods. Rather, location decisions of MNCs depend on aspects beyond transport costs, such as the ease of communication between headquarters and foreign subsidiaries, in aspects related to management, monitoring, coordination, troubleshooting, and so on.
All these aspects crucial to MNCs are, arguably, forms of tacit knowledge, as they require human interaction. Thus, the loss in efficiency that distant foreign affiliates face as evidenced in the literature is likely related to the difficulties in transferring tacit knowledge across long distances, and not only due to difficulties associated with costs related to transport of intermediate goods.
The documented evidence reinforces the importance of knowledge transmission in overall economic activity. Understanding the ways knowledge affects economic activity lies at the core of important and unanswered questions on convergence, development and growth. Knowledge and its diffusion, after all, are significant phenomena that can alter global economic patterns in as yet unexplored ways.
Akolakis, C, N Ramondo, A Rodríguez-Clare, and S Yeaple (2013), Innovation and Production in the Global Economy, NBER Technical Report 18792.
Arrow, K J (1969), “Classificatory Notes on the Production and Transmission of Technologcal Knowledge”, The American Economic Review 59(2): 29–35.
Bahar, D (2019a), “The hardships of long distance relationships: time zone proximity and the location of MNC’s knowledge-intensive activities”, working paper.
Bahar, D (2019b), “Measuring knowledge intensity in manufacturing industries: a new approach”, Applied Economics Letters 26(3).
Giroud, X (2012), “Proximity and Investment: Evidence from Plant-Level Data”, The Quarterly Journal of Economics 128(2): 861–915.
Gumpert, A (2018), “The Organization of Knowledge in Multinational Firms”, Journal of the European Economic Association 16(6) 1929-1976.
Keller, W, and S R Yeaple (2013), “The Gravity of Knowledge”, American Economic Review 103(4): 1414–1444.
Oldenski, L (2012), “Export Versus FDI and the Communication of Complex Information”, Journal of International Economics 87(2): 312–322.
Ramondo, N (2014), “A quantitative approach to multinational production”, Journal of International Economics 93(1): 108–122.
Ramondo, N, and A Rodriguez-Clare (2013), “Trade, Multinational Production, and the Gains from Openness”, Journal of Political Economy 121(2): 273–322.
Tintelnot, F (2017), “Global Production with Export Platforms”, The Quarterly Journal of Economics 132(1): 157–209.