We are living in an era of globalisation, yet there is evidence that distance still matters, and increasingly so

Contents

1. Introduction

2. Distance is still ‘alive’: evidence from three examples
2.1 The puzzling persistence of the distance effect on bilateral trade
2.2 The distance decay in international trade patterns
2.3 The Canada – U.S. border puzzle

3. Implications and conclusions

4. Bibliography

5. Appendix: Figures and Tables

1. Introduction

The current debate on globalization has raised an important question for economists: Is the world really becoming a ‘global village’ with a ‘weightless economy’? In accordance with Cairncross’s best-selling book The Death of Distance (2001)^[1] some economists have argued that falling transportation and communication costs are leaving ‘distance’ a negligible factor in the assessment of barriers to trade and integration of the world economies. Yet there is not enough funded evidence for this.

At first sight, works from authors as Quah (1999)^[2] or Levinson (2006)^[3] seem to support Cairncross’s claim of the declining importance of geographic balks in decisions of trade and commerce. Indeed the internet, revolutionizing the ICT^[4], has driven down the marginal costs of communicating to virtually nil. Emails, voice over IP, containerization and improved infrastructure have made the coordination and control of international trade a lot cheaper. Recent studies on the importance of education suggest that knowledge, as the ‘basic commodity’ of the 21st century and major source of future growth, makes industries and thus trade more independent of traditional physical fetters. On the contrary, it is not clear to what extent transportation costs have generally declined. Carrere & Schiff (2004)^[5] have found evidence that distance increasingly matters as the average distance of trade for an average country has decreased. While air transportation has reduced trade costs, especially in terms of time savings, ocean freight rates, according to Hummels (2000)^[6] and Crafts & Venables (2001)^[7], have seemingly increased^[8]. Moreover it is uncertain whether the decline in shipping costs affects distant and proximate transport equally. An emerging literature is trying to give consistent explanations for the puzzling persistence of the distance effect on trade flows and economic development. Especially the meta-analysis of existing empirical studies using the gravity model is promising.

In the second part of this essay we will discuss some of the evidence found so far, that distance is still ‘alive’. The persistence of the distance effect on bilateral trade (Disdier & Head (2004)^[9] ), the distance decay in international trade patterns (Linders (2005)^[10] ) and the Canada – U.S. border puzzle (McCallum (1995)^[11] and Anderson & Van Wincoop (2003)^[12] ) will be our focal points. The third section concludes and summarizes the implications of the results for future economic development on national and international levels.

2. Distance is still ‘alive’: evidence from three examples

In an era of globalization, with world trade growing faster than world output and FDI^[13] growing faster than trade, remoteness-related impediments to commerce should be a problem of the past. Thus one should expect transaction costs (the ‘distance effect’) to fall over time unless one assumes that technological progress in transportation and communication is “distance neutral”^[14]. Yet, with total shipping costs comprising up to 23% of the consumer price of American imports^[15], this expectation cannot be underpinned by empirical evidence.

2.1 The puzzling persistence of the distance effect on bilateral trade

Distance effects have traditionally been estimated with gravity models. In their simplest form these models predict bilateral trade flows based on the economic size of and distance between two countries, and have been employed to measure a whole range of possible trade determinants such as the impact of linguistic similarity, colonial ties or currency unions.

illustration not visible in this excerpt

Gravity equation of trade with Abbildung in dieser Leseprobe nicht enthalten as total volume of trade between i and j; Abbildung in dieser Leseprobe nicht enthalten and Abbildung in dieser Leseprobe nicht enthalten as economic “masses” (e.g. GDP); Abbildung in dieser Leseprobe nicht enthalten as the distance between i and j and G as “gravitational constant”

illustration not visible in this excerpt

Log form of the gravity equation as defined by Disdier & Head (2004)

Disdier & Head (2004) provide us with the “first large-scale, systematic examination of the magnitude of the distance effect”. In their meta-analysis, (a statistical analysis of existing empirical results), of a representative sample of gravity model literature concerned with the effects of transaction costs on international trade they test the sensitivity of distance to time period, choice of control variables and differences in estimation techniques.

Controlling for structural and method variables the authors find that the time trend of the distance effect is clearly positive and especially significant from the 1980s onwards^[16]. Indeed, the numbers in figure 3 suggest that distance barriers mattered more in the 1990s than in the period from 1870-1969. From the “single continent” and “developed”/“no developed countries” coefficients one can infer a high transport elasticity for land trade and poor infrastructure economies respectively. In addition, the exclusion of the control variables “adjacency” and “common language” accounts for a significant proportion of the omitted variable bias. However, “trade agreements”, “remoteness” and the “endogeneity of GDP” do not show significant influence on the distance parameter. Both economists conclude that their mean distance elasticity of -0.89 cannot rebut the earlier results of Leamer & Levinsohn (1995)^[17] (-0.6) and Overman et al. (2003)^[18] (-0.9 to -1.5). Furthermore Disdier and Head emphasize the effect’s change over time. Transport costs have fallen in the period of 1870 – 1950 and then have started to rise again. Three major explanations are given for this striking phenomenon. First, marginal costs of trade might have been underestimated by empiricism and the role of the ICT revolution overvalued. Second, time costs as an important component of distance are shown to be rising^[19]. This can be explained by intensive use of just-in-time supply chain management, income-driven increase in opportunity costs of time and the high time-sensitivity of some goods (e.g. fashion products, perishables). Regarding the rising share of time-sensitive intermediates in international exports, Hummels (2001) rates the value of a saved shipping day for manufactured goods as equal to a 0.8% ad-valorem tariff. For office machines the cost of a transit day is 130 times higher than the average U.S. interest rate per day^[20]. Hummels corroborates the importance of time by showing that an increase in shipping time of one day reduces the probability of trade by 1 to 1.5%. Finally, though less probable, compositional changes in trade might be biased towards commodities with high shipping costs.

2.2 The distance decay in international trade patterns

Following Disdier & Head (2004) Linders (2005) finds in his meta-analysis that omitted variable bias and time lags (e.g. J-curve effect under Marshall-Lerner conditions) can significantly change the elasticity of trade with respect to geographical distance. Furthermore he believes these specification-related flaws to account for a large part of the heterogeneity of distance decay estimates. The main sources for omitted variable bias according to Linders are contiguity/adjacency, common language, colonial ties, common trade blocs and regional grouping. The exclusion of either one of these variables or lagged trade dummies usually leads to an upward biased distance decay parameter. Contrary to Anderson & Van Wincoop (2004), as we will see, Linders also states that neither the use of consistent theoretical foundations nor the omission of zero trade flows affect the distance coefficient systematically. Considering that the time trend of the distance effect is only negative for his estimated WLS regression the author concludes that his results^[21] support three propositions. First, distance costs might be multidimensional and less tangible components such as cultural differences or political hostility might remain important. Second, multilateral liberalization of trade could have increased the relative importance of geographical barriers compared to policy induced trade costs. Third, the fact that trade flows between Europe and the U.S. are less sensitive to distance might encourage those arguing that the integration of the Europe-U.S.-Japan Triad has occurred at the expense of the marginalisation of the ‘periphery’. These conclusions are overall in line with the findings of Disdier & Head (2004).

[...]

^[1] Cairncross, F. (2001), The Death of Distance: How the Communications Revolution Will Change Our Lives

^[2] Quah, D. (1999), The Weightless Economy in Economic Development, CEP No. 114, London

^[3] Levinson, M. (2006), The Box: How the Shipping Container Made the World Smaller and the Economy Bigger, Princeton University Press

^[4] Information and communication technology

^[5] Carrere, C. and M. Schiff (2004), On the Geography of Trade: Distance is Alive and Well, Journal of Economic Literature

^[6] Hummels, D. (2000), Have International Transportation Costs Declined, University of Chicago

^[7] Crafts, N. and A. J. Venables (2001), Globalization in History: A Geographical Perspective, Journal of Economic Literature

^[8] See figure 1 for detailed numbers

^[9] Disdier, A. C. and K. Head (2004), The Puzzling Persistence of the Distance Effect on Bilateral Trade, Journal of Economic Literature

^[10] Linders, G. M., (2005), Distance Decay in International Trade Patterns: A Meta-Analysis, Journal of Economic Literature

^[11] McCallum, J. (1995), National Borders Matter: The Canada-U.S. Regional Trade Patterns, American Economic Review, Vol. 85, No. 3, pp. 615-622

^[12] Anderson, J. E. and E. Wincoop (2003), Gravity with Gravitas: A Solution to the Border Puzzle, American Economic Review, Vol. 93, No.1, pp. 170-192

^[13] Foreign direct investments

^[14] See Buch et al. (2004) for a formal derivation of this argument

^[15] See Hummels, D. (1999), Have International Transportation Costs Declined?, University of Chicago Press

^[16] See figure 2 for a graphical illustration

^[17] Leamer, E. E. and J. Levinsohn (1995), International Trade Theory: The Evidence, in Grossman, G. and K. Rogoff (eds.) Handbook of International Economics Vol. 3, Elsevier, North Holland

^[18] Overman, H. G., S. Redding, A. J. Venables (2003), The Economic Geography of Trade, Production and Income: A Survey of Empirics, in Kwan-Choi, E. and J. Harrigan (eds.), Handbook of International Trade, J. Basil Blackwell

^[19] See Hummels, D. (2001), Time as a Trade Barrier, GTAP No. 18 and Deardorff, A. (2003), Time and Trade: the Role of Time in Determining the Structure and Effects of International Trade with an Application to Japan, Mimeo of the University of Michigan

^[20] Average U.S. T-Bill rate in 2001: 6.26% p.a., source: Hummels (2001)

^[21] See figures 4 for and 5 for a graphical illustration of Linders’s results