Abstract
In this paper we test the well-known hypothesis of Obstfeld and Rogoff (NBER Macroeconomics Annual 7777:339–390, 2000) that trade costs are the key to explaining the so-called Feldstein–Horioka puzzle. Our approach has a number of novel features. First, we focus on the interrelationship between trade costs, the trade account and the Feldstein–Horioka puzzle. Second, we use the gravity model to estimate the effect of trade costs on bilateral trade and, third, we show how bilateral trade can be used to draw inferences about desired trade balances and desired intertemporal trade. Our econometric results provide strong support for the Obstfeld and Rogoff hypothesis and we are also able to reconcile our results with the so-called home bias puzzle.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
In a study that focuses precisely on current account reversals (understood as sharp changes in a short period of time) in all countries of the world for which data are available, 157 of them, over 1970–2001, Edwards (2004) reports reversals constituting only 12% of the national observations.
Helpman et al. (2004) make a similar point and then proceed to distinguish between bilateral imports and exports differently than we do: by allowing for heterogeneities between firms. But there is plainly no conceptual conflict between their manner of proceeding and ours.
Note, in this connection, that the trade balance constraint in our work operates completely differently in the observations for US imports from Japan than in the observations for Japanese imports from the US, for example. Note also, once again, that our treatment of \({{P_{{\text{CPI}}}^i } \mathord{\left/ {\vphantom {{P_{{\text{CPI}}}^i } {P_{{\text{GDP}}}^i }}} \right. \kern-\nulldelimiterspace} {P_{{\text{GDP}}}^i }}\) as an exogenous variable poses little problem since we estimate bilateral imports rather than aggregate imports or the trade balance.
We experimented with different alternative specifications of the nominal effective exchange rate. Since these different measures did not materially affect any of our results, they are not reported in the paper but are available on request from the authors.
For example, the population term shows up with the wrong sign in the first sub-sample and the coefficient on the FTA term becomes significant at the 10% level in this sub-sample. For a clear statement of the reason to expect a negative sign of the population term in the gravity model, see Frankel and Romer (1999).
This need not be surprising since our estimates pertain to long run adjustments in trade balances. Nominal exchange rate movements promote real exchange rate changes in the short run. But their contribution to long run movements in the real exchange rate is not nearly as plain.
In the case of the price of oil, we experimented with the separate effects on demand for imports of oil importers and exporters. (Our measure of the real oil price is the average crude oil 3-month spot price in US dollars obtained from the IMF International Financial Statistics CD-Rom (IFS 2002), divided by the US consumer price index.) The results confirm the usual impression that the adverse impact of the oil shocks on the imports of the oil-importing nations exceeded the impact on the oil-exporting ones. Since some collinearity arises between the relative oil price for importers and \({{P_{{\text{CPI}}}^i } \mathord{\left/ {\vphantom {{P_{{\text{CPI}}}^i } {P_{{\text{GDP}}}^i }}} \right. \kern-\nulldelimiterspace} {P_{{\text{GDP}}}^i }}\), and since we wish to isolate the impact of \({{P_{{\text{CPI}}}^i } \mathord{\left/ {\vphantom {{P_{{\text{CPI}}}^i } {P_{{\text{GDP}}}^i }}} \right. \kern-\nulldelimiterspace} {P_{{\text{GDP}}}^i }}\), it seems right to focus on the results in the absence of any disaggregation of the price of oil.
OR effectively sidestep this difficulty by switching to an analysis of the impact of the real interest rate on the current account when they turn from the “home bias” to the FH puzzle. Even a moderate percentage change in relative prices may imply a large percentage movement in present relative to future real interest rates. Therefore, even should the change in relative prices be moderate, it would still contribute heavily, in their reasoning, to the FH puzzle as well.
To derive this number note, first, that \({{P_{{\text{CPI}}}^i } \mathord{\left/ {\vphantom {{P_{{\text{CPI}}}^i } {P_{{\text{GDP}}}^i }}} \right. \kern-\nulldelimiterspace} {P_{{\text{GDP}}}^i }}\) in our work does not correspond exactly to the import price relative to the export price in OR’s example. Instead, from Eq. 13 it refers to αi + (1 − αi)(1 + tij), after equating τ ij and t ij and setting \(\overline p _j \) and p i equal to 1. Thus, in terms of OR’s schematic example, our estimate relates to the value of \(\left( {1 - \alpha _i } \right)\left( {1 + t_{ij} } \right)^{1 - \theta } \). For m ii, \(\left( {1 + t_{ij} } \right)^{1 - \theta } \) reduces to 1, and therefore our estimate of the impact of \({{P_{{\text{CPI}}}^i } \mathord{\left/ {\vphantom {{P_{{\text{CPI}}}^i } {P_{{\text{GDP}}}^i }}} \right. \kern-\nulldelimiterspace} {P_{{\text{GDP}}}^i }}\) on m ii /m ij corresponds to \(\left( {1 - \alpha _i } \right)\left( {1 + t_{ij} } \right)^{\theta - 1} \) rather than \(\left( {1 + t_{ij} } \right)^{\theta - 1} \). We estimate this value as .3. Therefore, if we assign a value of 0.81 to αi, we have an estimate of \(\left( {1 + t_{ij} } \right)^{\theta - 1} \) of around 1.58, which implies a value of t ij of around 0.1.
Take the same parameter values as OR, but assume 100 identical countries so that country i imports from 99 others. In that case, the baseline situation without trade costs is one of 0.99 openness. Trade costs raising the ratio m ii /m ij by a factor of 4.2 in relation to each of the 99 foreign countries would then yield a rise in the percentage of home consumption from 0.01 to approximately 0.041. So to reduce the value of openness to a level as low as 0.19, OR require a value of m ii /m ij of approximately 422 or about 100 times 4.2.
Anderson and van Wincoop (2004) recently estimate t ij as 170%. This might suggest that 235% is not as outlandish as it seems at first blush. However, their figure includes the tax equivalent of all non-money impediments to trade that enter in the cross-sectional dimension in our estimates as well as the domestic impediments to trade.
In addition, elasticities of substitution may vary between goods, going from extremely high figures for home goods with very close substitutes abroad to very low figures. Consequently, as long as tij is non-trivially greater than zero, q alone may account for a significant rise of a above 0.01 in our example with 100 countries (because of the low values of q). This further reduces the difficulty of obtaining high values of a in Eq. 13.
There are, in fact, several reasons why the two values of trade costs may be expected to differ even at the margin. If additional trade means that new goods (not simply new varieties) enter into foreign trade, the rise in trade costs at the margin may jump up rather than go up continuously. In addition, the trade costs may be higher at first than they will become later, after the initial information and distribution problems of launching the new products abroad settle down. In this connection, a lot of recent empirical work shows that entry of individual firms into export activity always entails major once-and-for-all costs of production and distribution. (See Roberts and Tybout (1997), Bernard and Jensen (2001) and Bernard and Wagner (2001).) Such fixed costs may well be more severe if entry means introducing new products abroad (rather than previously exported ones or newly exported ones that are merely differentiated, as is more likely to happen when the adjustments concern the trade balance at a set level of total national trade).
Ruhl (2003) makes a similar point in a closely related context. Specifically, he seeks to reconcile the low estimates of the elasticity of substitution between home and foreign goods in the business cycle literature with the much higher estimates of this same elasticity in the literature on the growth of trade, where the concern is with the impact of trade liberalization, free trade agreements and the like. In the former literature, the elasticities regard responses to transitory shocks whereas in the latter, they relate to responses to permanent shocks. He considers the former adjustments as ones on the “intensive margin” and the latter as ones on the “extensive margin”. Our context differs because the adjustments in trade balance that we are interested in may well be persistent. However, the similarity remains so far as those adjustments do not necessarily require any change in the size of the traded goods sector relative to the economy as a whole, and therefore in the total range of goods entering into trade (ordered by trade costs). Thus, his distinction between adjustments at the intensive margin (same range of goods) and the extensive margin (wider range of goods) is apt.
References
Anderson J, van Wincoop E (2003) Gravity with gravitas: a solution to the border problem. Am Econ Rev 93:170–192
Anderson J, van Wincoop E (2004) Trade costs. J Econ Perspect 52:691–751
Bayoumi T (1999) Estimating trade equations from aggregate bilateral data. International Monetary Fund Working Paper 99/74
Bergin P, Glick R (2003) A model of endogenous nontradability and its implications for the current account. Federal Reserve Bank of San Francisco Working Paper
Bergin P, Glick R (2004) Endogenous tradability and macroeconomic implications. Federal Reserve Bank of San Francisco Working Paper
Bergstrand J (1985) The gravity equation in international trade: some microeconomic foundations and empirical evidence. Rev Econ Stat 67:474–481
Bergstrand J (1989) The generalized gravity equation, monopolistic competition, and the factors–proportions theory in international trade. Rev Econ Stat 71:143–153
Bernard AB, Jensen JB (2001) Why some firms exports. NBER Working Paper No. 8349
Bernard AB, Wagner J (2001) Export entry and exit by German firms. Weltwirtsch Arch 137:105–123
Betts C, Kehoe T (2001) Tradability of goods and real exchange rate fluctuations. Federal Reserve Bank of Minneapolis Working Paper
Dooley M, Frankel J, Mathieson D (1987) International capital mobility: what do saving–investment correlations tell us? IMF Staff Pap 34:503–530
Edwards S (2004) Thirty years of current account imbalances, current account reversals and sudden stops. IMF Staff Pap 51:1–49
Frankel J, Romer D (1999) Does trade cause growth? Am Econ Rev 89:379–399
Glick R, Rose A (2002) Does a currency union affect trade? The time series evidence. Eur Econ Rev 46:1125–1151 (June)
Gordon R, Bovenberg L (1996) Why is capital so immobile internationally? Possible explanations and implications for capital income taxation. Am Econ Rev 86:1057–1075
Helpman E (1987) Imperfect competition and international trade: evidence from fourteen industrial countries. J Jpn Int Econ 1:62–81
Helpman E, Melitz M, Rubinstein Y (2004) Trading partners and trading volumes. Harvard and Tel Aviv Universities Working Paper
Lane P, Milesi Ferretti GM (2002a) External wealth, the trade balance, and the real exchange rate. Eur Econ Rev 46:1049–1071
Lane P, Milesi Ferretti GM (2002b) Long-term capital movements. NBER Macroeconomics Annual 2002:73–116
Melitz J (2008) Language and foreign trade. Eur Econ Rev (in press)
Obstfeld M (1986) Capital mobility in the world economy: theory and measurement. Carnegie–Rochester Conf Ser Public Policy 24:55–103
Obstfeld M, Rogoff K (2000) The six major puzzles in international macroeconomics: is there a common cause? NBER Macroeconomics Annual 7777:339–390
Roberts MR, Tybout JR (1997) The decision to export in Colombia: an empirical model of entry with sunk costs. Am Econ Rev 87:545–564
Rose A (2000) One market, one money: estimating the effects of common currencies on trade. Economic Policy 30:7–46
Ruhl K (2003) Solving the elasticity puzzle in international economics. University of Minnesota Working Paper
Taylor A (2002) A century of current account dynamics. J Int Money Financ 21:725–748
Acknowledgements
The authors would like to thank Philip Lane for valuable suggestions.
Author information
Authors and Affiliations
Corresponding author
Data appendix
Data appendix
- M ij :
-
bilateral imports CIF of i from j (source: IMF Direction of Trade Statistics, DOTS 2002), expressed in US$ and transformed in constant dollars by dividing by the US CPI.
- Distance ij :
-
great circle distances are calculated using the arc-geometry formula on the latitude and longitude coordinates of the most populous city.
- A i :
-
absorption of country i. The figure is obtained by subtracting the trade balance (in US$, divided by the US CPI) from the GDP (in US$ divided by the US CPI). Both series are taken from the World Bank World Development Indicators (hereafter WDI (2003)).
- Y j :
-
gross domestic product of country j in current US$ divided by the US CPI series. Both series are taken from the WDI (2003).
- P CPI :
-
consumer price index (1995 = 100) taken from the WDI (2003).
- P GDP :
-
GDP deflator (1995 = 100) taken from the WDI (2003).
- Real price of oil:
-
average crude oil price 3-months spot price index from the IMF-IFS CD-Rom (2002, line 00176AADZF). The series is in US$ and has been divided by the US CPI.
- Population:
-
population taken from the WDI (2002).
- Common language:
-
see Melitz (2008)
List of countries
I.S. of Afghanistan*, Albania, Algeria, Angola, Antigua and Barbuda*, Argentina, Aruba*, Australia, Austria, The Bahamas, Bahrain, Bangladesh, Barbados, Belgium, Belize, Benin, Bhutan*, Bolivia, Brazil, Brunei Darussalam*, Bulgaria, Burkina Faso, Burundi, Cambodia, Cameroon, Canada, Central African Rep., Chad, Chile, People’s Rep of China, Colombia, Comoros*, Dem. Rep. Of Congo, Republic Of Congo, Costa Rica, Cote D’Ivoire, Cyprus, Czech Republic, Denmark, Djibouti*, Dominica, Dominican Republic, Ecuador, Egypt, El Salvador, Ethiopia, Fiji, Finland, France, Gabon, The Gambia, Germany, Ghana, Greece, Grenada, Guatemala, Guinea*, Guinea-Bissau, Guyana, Haiti, Honduras, Hong Kong, Hungary, Iceland, India, Indonesia, I.R. Of Iran, Iraq*, Ireland, Israel, Italy, Jamaica, Japan, Jordan, Kenya, Kiribati*, Kuwait, People’s Dem. Rep. of Lao, Lebanon*, Liberia*, Libya*, Macedonia, Republic of, Madagascar, Malawi, Malaysia, Maldives, Mali, Malta, Mauritania, Mauritius, Mexico, Mongolia, Morocco, Mozambique, Nepal, Netherlands, Netherlands Antilles*, New Caledonia*, New Zealand, Nicaragua, Niger, Nigeria, Norway, Oman, Pakistan, Panama, Papua New Guinea, Paraguay, Peru, Philippines, Poland, Portugal, Qatar*, Romania, Russian Federation, Rwanda, Samoa, Sao Tome and Principe*, Saudi Arabia, Senegal, Seychelles, Sierra Leone, Singapore, Solomon Islands, Somalia*, South Africa, South Korea, Spain, Sri Lanka, St. Kitts and Nevis, St. Lucia, St. Vincent and the Grenadines, Sudan, Suriname, Sweden, Switzerland, Syrian Arab Republic, Tanzania, Thailand, Togo, Tonga, Trinidad And Tobago, Tunisia, Turkey, Uganda, United Arab Emirates*, United Kingdom, United States, Uruguay, Venezuela, Vietnam, West Bank*, Zambia, Zimbabwe.
Countries with asterisk are only an exporter in the dataset.
Free trade areas
Regional trade agreements notified to the GATT/WTO and in force (source: http://www.wto.org as of 30th of June 2002).
-
1)
EC/EEA/EFTA/EU
Belgium, Bel-Lux, Denmark, France, Germany, Ireland, Italy, Luxembourg, Netherlands, UK, Norway, Switzerland, Malta, OCTs (Greenland, New Caledonia, French Polynesia, St. Pierre and Miquelon, Aruba, New Antilles, Falklands, St. Helena);
Austria (since 1995), Finland (since 1995), Sweden (since 1995), Greece (since 1981), Portugal (since 1986), Spain (since 1986).
-
2)
NAFTA
Free Trade Agreement, since 1994
Canada, Mexico, USA
-
3)
CARICOM
Customs Union, since 1973
Antigua and Barbuda, The Bahamas, Barbados, Belize, Dominica, Grenada, Haiti, Jamaica, Trinidad and Tobago, St. Vincent and Grenadines, St. Kitts and Nevis, Suriname
-
4)
SPARTECA
Free Trade Agreement, since 1977
Australia, New Zealand, Fiji, Kiribati, Nauru, Papua New Guinea, Solomon Islands, Tonga, Tuvalu, Vanuatu.
-
5)
MERCOSUR
Customs Union, since 1991
Argentina, Brazil, Paraguay, Uruguay
-
6)
BAFTA
Free Trade Agreement, since 1994
Estonia, Latvia, Lithuania
-
7)
CACM
Customs Union, since 1961
Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua.
-
8)
US–ISRAEL
Free Trade Agreement, since 1985
United States, Israel
-
9)
CER
Free Trade Agreement, since 1983
Australia and New Zealand
Common countries
(Source: CIA World Factbook 2002)
China, Hong Kong (since 1997) and Macao; Denmark, Faeroe Islands and Greenland; France, French Polynesia, Guadeloupe, French Guiana, Martinique, New Caledonia, Reunion, and St. Pierre and Miquelon; The Netherlands, Aruba and Netherlands Antilles; United Kingdom, Bermuda, Falkland Islands, Gibraltar, and St. Helena; United States, American Samoa and Guam.
Ex-colonial relationship and ex-common colonizer
(Source: CIA World Factbook 2002)
Australia and Papua New Guinea; Belgium and Burundi, Dem. Rep. of Congo; France and Algeria, Benin, Burkina Faso, Cambodia, Cameroon, Central African Rep., Chad, Congo Rep. of, Djibouti, Gabon, Guinea, Lao People’s Rep, Madagascar, Mali, Mauritania, Morocco, Niger, Senegal, Syrian Arab Rep., Togo, Tunisia, Vietnam; Italy and Libya; New Zealand and Samoa; Portugal and Angola, Cape Verde, Guinea-Bissau, Mozambique, Sao Tome and Principe, Timor; Spain and Equatorial Guinea; South Africa and Namibia; The Netherlands and Indonesia, Suriname; Japan and North Korea, South Korea; USA and Palau, Philippines; United Kingdom and The Bahamas, Bahrain, Bangladesh, Barbados, Belize, Bhutan, Botswana, Brunei Darussalam, Cyprus, Dominica, Fiji, The Gambia, Ghana, Grenada, Guyana, Hong Kong, India, Jamaica, Jordan, Kenya, Kiribati, Kuwait, Lesotho, Malawi, Malaysia, Maldives, Malta, Mauritius, Myanmar, Nauru, Nigeria, Pakistan, Qatar, Seychelles, Sierra Leone, Singapore, Solomon Islands, Somalia, Sri Lanka, St. Kitts and Nevis, St. Lucia, St. Vincent and the Grenadines, Sudan, Swaziland, Tanzania, Tonga, Trinidad and Tobago, Tuvalu, Uganda, United Arab Emirates, Vanuatu, Zambia, Zimbabwe. (Countries in italic characters are the ex-colonizers).
Currency unions
(Source: Glick and Rose (2002), updated with information from the IMF International Financial Statistics (2002))
-
1)
Antigua and Barbuda, Dominica, Grenada, Montserrat, St. Kitts and Nevis, St. Lucia, St. Vincent and the Grenadines;
-
2)
Aruba, Netherlands Antilles, Suriname (until 1994); Australia, Kiribati, Nauru, Tonga (until 1991), Tuvalu;
-
3)
Austria, Belgium, Finland, France, Germany, Ireland, Italy, Luxembourg, The Netherlands, Portugal and Spain (since 1999);
-
4)
Cameroon, Togo;
-
5)
Central African Rep., Benin, Burkina Faso, Chad, Comoros (until 1994), Rep. of Congo, Ivory Coast, Equatorial Guinea, Gabon, Guinea-Bissau, Madagascar (until 1982), Mali, Niger, Senegal, Togo;
-
6)
Denmark, Faeroe Islands, Greenland;
-
7)
France, French Guiana, Guadeloupe, Martinique, Reunion, St. Pierre and Miquelon;
-
8)
Lesotho, South Africa, Swaziland;
-
9)
New Caledonia, French Polynesia, Wallis and Futuna;
-
10)
Qatar, United Arab Emirates;
-
11)
United Kingdom, Falkland Islands, Gibraltar, St. Helena;
-
12)
United States, American Samoa, Bahamas, Bermuda, Dominican Rep. (until 1985), Guam, Guatemala (until 1986), Liberia, Panama.
Rights and permissions
About this article
Cite this article
Fazio, G., MacDonald, R. & Melitz, J. Trade Costs, Trade Balances and Current Accounts: An Application of Gravity to Multilateral Trade. Open Econ Rev 19, 557–578 (2008). https://doi.org/10.1007/s11079-008-9082-8
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11079-008-9082-8