VII. The Role of Imports—Structural Shifts and Economic Benefits1
Imports increase consumer choices, exert competitive pressures on domestic producers, and facilitate industrial restructuring. In studies of cross-country differences in external performance, imports have drawn considerably less attention than exports, and typically have been assigned a passive role. For example, Allard, et al (2005) found imports to be largely determined by final demand while competitiveness has been playing a minor role. However, the benefits of imports are well known: they increase the supply of goods and services available to meet final demand, enable a national economy to bring forward consumption and investment, offer an enlarged product variety, and facilitate the global division of labor. This chapter looks at the role of imports in restructuring the economies of the five southern euro area (SEA-5) countries—France, Greece, Italy, Portugal, and Spain—and, for the purpose of comparison, the euro area average and Germany.
Looking for structural shifts in disaggregated data. We will briefly revisit the main macroeconomic determinants of imports but focus on structural shifts and the economic benefits of imports. More specifically, we will look at the changes over time in trade openness, trade motivation, their technology content, and finally at the relation between import penetration and labor productivity. Most of the concepts used here require highly disaggregated data. The United Nations Commodity Trade Statistics Database (Comtrade) provides data on product groups up to the six-digit level (SITC classification), while the EU KLEMS data bank provides data on productivity by sectors of production (NACE classification).
The main macroeconomic determinants of imports are final demand and price/cost competitiveness. Simple long-linear regressions broadly confirm results that can be found in the literature.2 Imports are very sensitive to changes in final domestic demand, while the impact of relative prices (the ratio of import to GDP prices) is generally weaker. The estimated elasticity of imports with respect to final demand ranges from 2.3 in Spain to 1.5 for Greece; all demand elasticities were highly significant (Table VII.1). Price elasticities are found to be significant but much smaller in four countries (France, Greece, Portugal, Spain) ranging from -0.8 to -0.3, but are not significantly different from zero in Italy and Germany. Without robustness test, these results should be interpreted prudently. Nevertheless, the estimates can serve at least illustrative purposes.
|Final domestic demand||1.92**||1.48**||1.97**||1.55**||2.28**||1.71**|
|Relative import price||-0.35*||-0.77**||-0.09||-0.51**||-0.28*||0.04|
significant at the 1 percent level
significant at the 5 percent level
significant at the 1 percent level
significant at the 5 percent level
|4||United Kingdom||7||Russia||6||United Kingdom||5||Italy||6||United Kingdom||6|
|7||United States||6||United States||4||Africa||4||Belgium||3||China: Mainland||3|
|8||China: Mainland||3||Korea||4||United States||4||United States||3||United States||3|
|14||Portugal||1||Iran, I.R. of||3||Japan||2||Russia||1||Ireland||1|
Openness to trade has increased dramatically over the past 30 years. The SEA-5 countries have become more open to trade, though progress toward opening up was uneven. Spain and Portugal are leading the field, with an increase in the openness measure (the sum of exports and imports in percent of GDP) from about 20 percent in the early 1970s to more than 70 percent in 2005. Italy and Greece have raised their degree of openness more than twofold during that period (Figure VII.1). Openness in Spain and Portugal has leapfrogged after EU accession, and the introduction of the EU single market (completed in 1992) has made a visible difference for all EU member countries. The positive effect of the single currency (euro) on trade that was found in a number of studies, however, is not evident for the SEA-5 countries probably because of the global cyclical slowdown around 2003 and structural demand weakness from Germany, their main trading partner.3 Contrary to the rule of thumb that smaller economies tend to export and import more than large countries relative to GDP, there does not seem to be a clear-cut relation between country size and openness within the SEA-5. For example, despite its larger size, external trade weighs more in Germany’s economy than in any of the SEA-5. Likewise, while the size of GDP in Portugal is close to that of Greece, Portugal is now significantly more open to trade. The experience of these two countries, which set off from a similar starting point in the early 1970s, also indicates that the speed of convergence toward a high degree of openness is not necessarily higher for countries with a lower initial position.
Figure VII.1.Openness to Trade (in volumes)
Source: National Accounts.
Most imported goods and services of the SEA-5 originate in the EU. During 2000–05, between 55 and 70 percent of imports to the SEA-5 were supplied by partner countries in the EU and between 50 and 65 percent by euro area countries. While the share of intra-EU imports has remained relatively stable, intra-euro area imports gained importance. Germany is clearly the most important provider of imported goods with a share ranging from 13 percent (Greece) to 28 percent (Portugal). While increasing, the role of China has remained small; the share of imports from China ranged from 1 percent (Portugal) to 3.7 percent (Italy). The United States is the most important non-European supplier of imported goods to the SEA-5—its share ranging between 2.5 percent in Portugal and 6 percent in France. While the weight of emerging market economies has been increasing fast, particularly in the past five years, products of advanced industrialized countries account for more than three-fourths of nonenergy imports.
Imports have become more technology intensive. Contrary to a common perception that highly advanced EU countries are specializing in the production and export high technology (or highly capital-intensive) products while importing predominantly low technology (or labor intensive products), the SEA-5 (and Germany) have moved away from importing low technology while increasing imports of high and medium-high technology products. The decline in the share of low-tech imports has been strongest where their level was particularly high in 1988 (France, Greece, Germany). The structure of imports by technology content has become more similar across the SEA-5 and with that of Germany: between 50 and 60 percent of imports are high- and medium-high technology goods.
The share of intra-industry trade is high and has increased in some countries.4 Following the literature, the fraction of trade that is intra-industry—more precisely, the fraction of external trade in very similar products in percent of aggregate external trade—is measured here by the Grubel-Lloyd Index (GLI).5 The index has a straightforward interpretation: if trade is balanced industry by industry (or better product by product), it equals one and all trade is intra-industry; if there is complete international specialization so that every industry is either an export or import industry, it equals zero (Krugman, 1981). To be meaningful, the index requires a sufficiently narrow definition of product families. For all SEA-5 countries, with the exception of Greece, intra-industry trade accounts for 50 to 80 percent of overall trade at the four-digit industry classification (SITC revision 3) level. The relative importance of intra-industry trade is not necessarily linked to country size (compare Portugal with Greece), although it’s more important in very large areas, such as the EU-15. In countries where the initial share of intra-industry trade has been relatively low (Greece, Portugal, Spain), it has become relatively more important over the period under review. More intra-industry trade may reflect a number of factors, including:
- Increasing demand for greater variety of products, as consumers have become wealthier. The change in consumers’ taste implies that strong market positions in specialized products have become more important.
- Intensified economic integration and cross-border production. Within Europe the creation of the EU single market and EU enlargement have been particularly import factors driving cross-border division of labor.
Import penetration and labor productivity are positively correlated. Import penetration is defined as the part of domestic demand directed to the output of a particular sector that is satisfied by imports. To calculate import penetration, the EU KLEMS data bank, which provides sector data in NACE classification, was used for domestic production (gross output) and labor productivity, while Comtrade import and export data were reclassified to match the sectors in KLEMS.6 Import penetration has increased in all manufacturing sectors, though at different speeds—fastest in textiles (trade liberalization) and office machines (technical progress), and slowest in low-tech sectors, such as paper and pulp, and basic materials. With the exception of Germany, import penetration has also grown fast in transport equipment (mainly cars). On average, in sectors where import penetration has increased most, labor productivity has risen fastest. The trend line in Figure VII.4 shows a positive relation between (increases in) import penetration and labor productivity by sector. Given the small number of data points per country, however, this trend line should only be taken as suggestive evidence. Rigorous statistical analysis would not be suitable. Import penetration can be expected to be positively linked to labor productivity growth for a number of reasons, including (i) better performing foreign suppliers compete out weak domestic performers; (ii) foreign competition exert pressure on domestic producers to streamline production processes, and become more innovative; and (iii) domestic industries profit from imported technologies and know-how. The relative role of the various channels through which imports affect domestic performance is hard to quantify. However, while the short-run implications for employment may differ, in the long run these channels are complementary— resources are used more efficiently.
Figure VII.2.Product Structure of Imports by Technology
Note: Trade data cover goods only.
Figure VII.3.Intra-Industry Trade
Source: COMTRADE (four-digit SITC rev 3).
Note: Trade data cover goods only.
Figure VII.4.Import Penetration and Labor Productivity
Sources: KLEMS; and COMTRADE.
While demand and relative prices remain the main determinants of imports, structural factors have also been important.
- We found that the SEA-5 (France, Greece, Italy, Portugal, and Spain) have become much more integrated in the European single market and the global economy. The volume of trade has increased dramatically in percent of GDP since the 1970s, and in particular with the introduction of the EU’s single market and single currency. In the process, the degree of import penetration has risen.
- For some countries, increased openness and catching-up went hand-in-hand, transforming domestic production and patterns of external trade, while reducing the role of historical specializations. However, within this narrow group of SEA-5, no clear pattern was discernible between the distance to the frontier (Germany) and speed of transformation.
- International division of labor and the demand of consumers for more varieties— typical when countries are getting richer—have been identified as drivers of this transformation of production structures. The SEA-5 countries are importing and exporting more products of the same product family. In the process, the composition of imports has shifted away from low-technology products and the demand for innovative, high-quality products has increased.
- Finally, there is evidence of a positive relation between import penetration and productivity growth. While more empirical work would be needed to identify the relative importance of the channels through which imports contribute to higher efficiency, a positive relation has been found in each of the SEA-5—as well as in Germany.
ArtusP. and L.Fontagné2006 “Évolution récente du commerce extérieure français” Rapport du Conseil d’Analyse Économique La Documentation française (Paris).
BaumannU. and F.di Mauro2007 “Globalization and Euro Area Trade-Interactions and Challenges” ECB Occasional Paper No 55 (March).
DelozierB. and S.Montout2007 “New trends in globalization and the international division of labor: Consequences for Europe?” Documents De Travail De La DGTPE No 2007/03.
EU KLEMS2007 “Productivity in the European Union: A Comparative Industry Approach” Available via the Internet: http://www.euklems.net/.
FontagnéL. and M.Freudenberg1997 “Intra-Industry Trade: Methodological Issues Reconsidered” CEPII document de travaille No. 97-01.
FontagnéL.G.Gaulier and S.Zignago2007 “Specialization across varieties within products and North-South competition” CEPII Working Paper No. 2007–06.
FrankelJeffrey A. and A.K.Rose2000 “Estimating the Effect of Currency Unions on Trade and Output” CEPR Discussion Paper No. 2631 (December).
HatzichronoglouT.1997 “Revision of the High-Technology Sector and Product Classification” OECD Science Technology and Industry Working Paper 1997/2.
HooperP.K.Johnson and J.Marquez2000 “Trade Elasticities for the G-7 Countries” Princeton Studies in International Economics No. 87 (August).
KrugmanP.1981 “Intraindustry Specialization and the Gains from Trade” The Journal of Political EconomyVol. 89No. 5 pp. 959–73 (October).
KrugmanP. and M.Obstfeld2003International Economics6th edition (Addison-Wesley).
MadariagaN.2007 “Has France adjusted to recent trends in world trade?” Trésor-Economics DGTPE No 17 (September).
OECD2007 “Staying Competitive in the Global Economy-Moving up the Value Chain.”
Prepared by Werner Schule.
Econometric work, controlling for these and other factors found that currency unions boost trade significantly. Frankel and Rose (2000) have estimated very large effects of EMU on trade. Follow-up work by others confirmed the positive effect on trade but found it to be of smaller magnitude.
Delozier and Montout (2007) found that intra-industry has bounced back recently in the U.S. and the euro area. We did not find a similar rebound in the five countries under consideration.
In each sector (j) it is measured as: import penetration in sector (j) = imports of sector (j) products/(gross output of sector(j)+imports of sector (j) products -exports of sector (j) products).