The Global Disconnect: The Role of Transactional Distance and Scale Economies in Gravity Equations
| Author | Prakash Loungani,Assaf Razin,Ashoka Mody |
| Published date | 01 November 2002 |
| DOI | http://doi.org/10.1111/1467-9485.00246 |
| Date | 01 November 2002 |
THE GLOBAL DISCONNECT: THE ROLE OF
TRANSACTIONAL DISTANCE AND SCALE
ECONOMIES IN GRAVITY EQUATIONS
Prakash Loungani,*Ashoka Mody *and Assaf Razin**
ABSTRACT
Recent empirical analyses show that asset flows can be modelled by the same
‘gravity’ equations that trade economists have used so successfully for the past few
decades. This is something of a surprise. Trade economists do not yet have a unified
theory of why gravity models should work-and the situation is worse for asset
flows. Reasonable theories would predict that greater distance between countries
should generate more asset flows rather than less as the econometric results seem to
consistently show. In this paper we discuss how host and source country GDPs,
language, and distance the core explanatory variables in the traditional gravity
models-fare in trade and asset flows estimations. While the ‘distance puzzle’ is not
resolved, it is considerably reduced by going beyond consideration of physical
distance to concepts of transactional distance and scale economies.
INTRODUCTION
The nations of the world remain stubbornly apart. Physical separation acts as a
natural barrier, restricting trade and asset flow linkages across nations. Despite
apparent globalization, the constraints due to distance have remained
significant. Thus, at a time when globalization is taken for granted— and
policymakers and others debate the pros and cons of the associated changes— it
is not obvious that the nations of the world are truly coming closer together.
In this paper, we estimate gravity models for trade and foreign direct
investment (FDI) flows to explore if comparisons of the two sets of estimates can
clarify the role of distance and, hence, the nature of global linkages. Gravity
models postulate that bilateral international transactions are positively related to
the size of two economies and negativelyto the distance between them. A selective
literature review suggests that, though they are widely used as empirical
benchmarks, coefficients in the gravity equation vary widely and, moreover, do
not have straightforward interpretations. In particular, the large fall off observed
in trade and investment flows with increasing physical distance remains a puzzle.
Scottish Journal of Political Economy, Vol. 49, No. 5, November 2002
#Scottish Economic Society 2002, Published by Blackwell PublishersLtd, 108 Cowley Road, Oxford OX4 1JF, UK and
350 Main Street, Malden, MA 02148, USA
526
*International Monetary Fund.
**Tel Aviv University.
Evidence of the global disconnect comes in two forms. First, nations trade in
goods and assets to a smaller extent than would be warranted by the gains from
increased specialization and possibilities of risk diversification. In a recent
contribution Obstfeld and Rogoff (2000) discuss the various disconnections,
more commonly referred to as ‘puzzles’. The puzzles are especially troubling in
relation to asset flows since national borders should have little or no bearing on
investments in financial assets. Thus, the continued persistence of the ‘home
bias’ in equity investment (disproportionately high investment by residents in
domestic assets) and of the Feldstein– Horioka puzzle (the high correlation
between domestic savings and domestic investment) are among the important
stylized facts that reflect the limits to international transactions in financial
assets. The second set of evidence shows that the international transactions that
do occur, both in goods and financial assets, are strongly conditioned by the
physical distance between countries. It is this latter evidencc based on so-called
‘gravity’ models— that is the focus of this paper.
But why exactly does distance matter? Most obviously, greater distance can be
thought of as a proxy for higher transportation costs. If distance is truly a good
proxy for transport costs, then it has a special attraction. Obstfeld and Rogoff
(2000) have proposed that transportation costs may, in fact, be relevant not only
for trade but also for the constraints in international asset transactions. This line
of reasoning could explain the finding that distance appears with a negative and
highly significant sign in gravity equations for FDI and for financial assets.
However, there are at least two problems in identifying distance with
transport costs. First, Grossman (1998) has argued that for plausible values of
transport costs, the distance coefficient in trade equations should be much
smaller in magnitude than the typically estimated coefficient. Second, various
theoretical models predict that distance should actually appear with a positive
sign in asset flow equations. Thus, FDI from a source country may increase with
distance if high transport costs make it expensive to export to the host country
destination (Brainard, 1997). For financial assets, greater distance between
source and host country should be associated with reduced correlation of
business cycles and hence, through greater possibilities for diversification, to
more equity flows (Portes and Rey, 2000).
Thus, some authors have pursued the notion that distance captures more than
transport costs. More specifically, Rauch (1999) suggests larger distance may be
associated with greater information and search costs. Similarly, Eichengreen and
Irwin (1998) suggest that trading partners build long-term relationships that
embody significant informational capital. When they proxy for such information
capital through the addition of the lagged dependent variable as an additional
regressor, the distance coefficient— and hence the short-run distance elasticity
of trade— drops sharply. However, as we discuss, some have argued that this
resolution of the distance elasticity is far from satisfactory. Finally, Portes
and Rey (2000) and Portes, Rey, and Oh (2001) deal with the possibility of
information ease by adding bilateral telephone traffic as a regressor in the
gravity equation for financial asset transactions. They find that ease of
information flows is important for those categories of financial assets that are
THE GLOBAL DISCONNECT 527
#Scottish Economic Society 2002
To continue reading
Request your trial