Journal Issue
Share
Article

Competitiveness in the Southern Euro Area: France, Greece, Italy, Portugal, and Spain

Author(s):
Bogdan Lissovolik, Julio Escolano, Stefania Fabrizio, Werner Schule, Herman Bennett, Stephen Tokarick, Yuan Xiao, Marialuz Moreno Badia, Eva Gutierrez, and Iryna Ivaschenko
Published Date:
April 2008
Share
  • ShareShare
Show Summary Details

II. SEA-5 Exports: Wind in the Sails from Global Growth?1

A. Introduction

Exporting to fast-growing markets and sectors, especially in the current period of strong and varied world growth, is considered important for economic performance. For example, Arora and Vamvakidis (2004) showed that, controlling for convergence and other standard determinants, dynamic trading partners may substantially contribute to growth, with industrial countries particularly benefiting from trade with fast-growing developing countries. Recent literature also emphasizes the growth impact of export specialization (Plümper and Graff, 2001) compared to the more agnostic “traditional” view. And surging global export competition underscores the classic case for flexibility in reallocating to new, more promising, activities.2

The “high-growth” exports may also help external competitiveness, beyond their impact through standard aggregate measures. To be sure, past export gains in fast-growing markets/sectors are already part of overall export and balance of payments indicators. But many high-growing markets have been consistent high-performers and may be expected to stay so in the future, promising additional longer-term benefits to success there. Moreover, “adaptability” of exports may by itself be a factor in a better performance, even if growth patterns shift in the future. Thus, Fabrizio, et al (2007) find that partner growth favored the expansion of emerging economies’ export shares, while Danninger and Joutz (2007) rank it higher than cost competitiveness in the export success of Germany.

This section suggests that the southern euro area (SEA-5) countries have so far taken comparatively little advantage of these channels to enhance competitiveness. Located mainly amid slow growers, these economies stand to benefit from diversifying their traditional (neighboring) export destinations toward more dynamic markets. Globalization is facilitating and prodding this process—through reducing transport costs and putting competitive pressure on traditional geographical or sectoral patterns. But while there is anecdotal evidence that the SEA-5 countries had some success in promoting their products or tailoring them to dynamic destinations, the magnitude and other characteristics of these trends often lag those of key industrialized comparators.

The paper investigates the SEA-5’s geographical and sectoral export performance in “stock” (structure) and “flow” (reorientation) terms. The research is structured as follows. First, stylized facts of geographical export performance are analyzed, highlighting the role of fast-growing countries in the structure of SEA-5’s trade and of changes in this structure. Several simple indicators (elaborated in Appendix II.A) permit relevant cross-country comparisons. Second, sectoral specialization and reorientation—from the point of view of fast-growing activities—are explored through an analysis of manufacturing market shares. The research is mostly focused on nominal export measures, to account for the view that measured real exports may misrepresent “true” performance.3 Only trade in goods is considered; trade in services is analyzed separately as part of this project. With respect to the timeframe of the study, the main focus is on the decade of 1995–2005. The data for 2006–07 have generally not been used, partly reflecting incomplete and provisional nature of the information in key foreign trade databases (i.e., Comtrade) at the time of the preparation of this study.

B. Are SEA-5’s Exports Benefiting from Higher Partner Growth?

Total real export demand growth has been slightly lower for SEA-5 than for its key comparators. According to an index of trade-share weighted real import growth of trading partners (Figure II.1), SEA-5 countries have on average faced weaker export demand compared to that for industrialized countries or euro area (as a whole). While this measure is comprehensive (covers all countries), the effect of fast-growing markets cannot be disentangled from cyclical conditions or special factors in a country’s economy or location, as exports are usually inversely dependent on distance between countries (as per the standard “gravity” model of trade). For example, SEA-5’s negative export demand growth differential was volatile and tended to reverse during spells of EU’s cyclical strength, notably in 2006. Among individual SEA-5 countries, Spain faced the lowest export demand growth, since some of its neighboring trading partners were particularly sluggish. But the relatively robust real growth in Spain (compared to other countries in the region) tended to be a comparatively positive influence on export demand for its SEA-5 trading partners. At the same time, data on actual real export growth suggest that capacity to get traction from the growth of partners varied markedly, with Spain benefiting the most and Italy the least.

Figure II.1.SEA-5 and Key Comparators, Export Indicators, 1996–2006

Source: IMF, WEO database.

A more detailed, albeit somewhat selective, approach involves an analysis of trade flows with a subset of dynamic economies, which are mostly emerging markets. Several definitions of such economies have been used (see Appendix II.A for details), with the key criterion being real medium-term GDP growth of at least 4 percent. These markets represent roughly 10–25 percent of world imports. (In what follows, the broader sample’s results will be reported, since the key conclusions remain intact). Interestingly, not all of the 43 countries in the sample increased their share of world imports, and just nine large Asian and European countries fully explain the 4 percentage point increase in the sample’s weight in world imports over 2000–05, with China alone accounting for 45 percent of this increase.

The SEA-5 exports relatively little to these dynamic countries. On aggregate, SEA-5’s exports to fast growers as a share of own exports have been somewhat lower than for key comparators—EU and the OECD (Figure II.2). More specifically, as reflected in the indicator of “underexporting” (Table II.1), compared to EU-13 (minus Luxembourg and Ireland, which are fast-growers), the SEA-5 exported relatively less to new member states (as expected, gravity-wise) and to large emerging markets in Asia (cannot by and large be explained by gravity because the EU comparators are equidistant from Asian countries). The conclusions are similar in comparison to Germany or OECD, against which SEA-5’s performance with Asia looks particularly weak. SEA-5 trades relatively more with some dynamic economies of North Africa and Latin America, but this offsets only a fraction of its underperformance in other markets.

Figure II.2.Exports to 43 Dynamic Economies, 2000–05

(Percent of own exports)

Source: UN Comtrade database.

Table II.1.SEA-5 Relative Underexporting Ranking, 2005 (Ratios)

(Higher value of the ratio means less exports compared to the reference group)1/

Relative to EU-13Relative to GermanyRelative to OECD
Estonia2.67Czech Rep.3.14Philippines4.11
Ireland2.34Slovakia2.54Indonesia2.97
Latvia1.83Estonia2.14R. of Korea2.95
Czech Rep.1.68Latvia2.03China2.86
Luxembourg1.67Poland2.00Malaysia2.48
Lithuania1.50Lithuania1.92Viet Nam2.44
India1.49China1.89Thailand2.32
Slovakia1.42Ukraine1.87Colombia1.87
South Africa1.35South Africa1.84Singapore1.77
Russia1.32Luxembourg1.79Estonia1.72
Ukraine1.27Russian1.71Ireland1.61
Poland1.25Malaysia1.60India1.56
China1.21Rep. of Korea1.57Venezuela1.38
Rep. of Korea1.20Philippines1.50Latvia1.34
Philippines1.19Kuwait1.35Slovakia1.34
Average for all1.091.181.30
43 fast growers
Source: UN Comtrade database.

The indicator of SEA-5’s “underexporting” denotes the ratio of market shares of a comparator group (EU-13) to SEA-5’s market shares in a country’s imports; for comparability these shares are adjusted for shares in total world imports (see formal definition in the appendix).

Source: UN Comtrade database.

The indicator of SEA-5’s “underexporting” denotes the ratio of market shares of a comparator group (EU-13) to SEA-5’s market shares in a country’s imports; for comparability these shares are adjusted for shares in total world imports (see formal definition in the appendix).

Among individual SEA-5 countries, Portugal, and to a lesser extent Spain, tended to export relatively less to “dynamic” countries. Other countries have done somewhat better but not exceptionally well: France’s performance is close to the SEA-5 average and Italy’s and Greece’s above that average but in line with the OECD as a group.

Reorientation of SEA-5 exports toward high-growing economies has also lagged. While in absolute terms the SEA-5 has been shifting its exports toward fast growers—as reflected in the accelerated growth of its exports to these destinations relative to the rest of the world—in “adjusted” or comparative terms, its performance has been less impressive:

  • SEA-5 tends to underperform in emerging market countries that had the largest import market share gains.Figure II.3 provides a ranking of the fast growers whose gains in world import share over the last five years have been the largest relative to the SEA-5’s gains in these markets. Among specific destinations, China is a clear outlier; all advanced countries failed to “catch up,” but the SEA-5 lagged there more than EU-13 or Germany. SEA-5’s comparative underperformance is also perceptible in such markets as Ireland, Argentina, and Chile. In other Asian countries, it has roughly matched EU-13’s and Germany’s lack of catch-up.

  • SEA-5 countries also fail to take advantage of highly “dynamic” markets, regardless of their size.Figure II.4 shows another indicator of trade reorientation, the cumulative difference in import growth of fast-growers and SEA-5’s (as well as comparators’) export growth. On average for the 43 dynamic economies, this difference is large and positive with the SEA-5 (16 percentage points), but negative for Germany and small for EU-13. While the SEA-5’s performance is comparatively poor vis-à-vis all dynamic economies, the gap is particularly large with respect to some Asian countries (Kazakhstan, Vietnam, and Pakistan), as well as with Europe and Latin America.

Figure II.3.Gaps with Fast Growers’ World Import Share Gains, 2000-05

(Gains of destination country in world imports less the increase in the weight of this country in exports of the SEA-5, EU-13, and Germany)

Source: UN Comtrade database.

Note: E.g., Share of China in world imports rose by 1.3 pp more than China’s share in SEA-5 countries’ exports.

Figure II.4.Lagging Export Growth to Fast Growers, 2000-05

(Import growth in destination country less export growth to this country by SEA-5, EU-13, and Germany)

Source: UN Comtrade database.

Note: E.g., Kazakhstan’s total cumulative import growth exceeded SEA-5 countries’ export growth to Kazakhstan by more than 100 percentage points.

The reorientation experience of individual SEA-5 countries has been mixed, and also differs by destination markets. China is clearly the most significant such market in terms of the increase of its world import share. Smaller countries (Portugal and Greece) had some gains in market shares there (Figure II.5), but from extremely low levels, which is underscored by the high volatility of their export growth to China (the levels are not shown in the figure). Large SEA-5 countries however have been steadily losing market shares in most important markets, including China, and these losses were larger than in Germany.

Figure II.5.Share of China’s Imports

(1995=100)

Source: UN Comtrade database.

C. Are SEA-5’s Exports Poised to Gain from Global Sectoral Export Trends?

It has been argued that SEA-5’s specialization in traditional products may be largely undesirable. In particular, Faini and Sapir (2006) suggested that Italy’s persistent, and at times increasing, specialization in traditional, more contested, and slower-growing sectors may be a drag on economic growth. On the other hand, it has been countered that such traditional sectors (i.e., textiles, clothing, leather, etc.) may well exhibit higher growth in unit values, including in luxury niches. This may offset at least some of the adverse effect on volumes from more intense competition (see Italian Ministry of Economy and Finance, 2007).

One way to approach this argument is by checking if the SEA-5 is specialized in “high-nominal-growth” activities. However, precisely defining the latter is difficult, as high growth is not always sustained. Judging by export growth in manufactured goods sectors, as indicated in Figure II.6, over the last decade the highest growth (in value terms) has been observed in a number of high-tech-cum-auto sectors and some metals.4 While export growth in metals was volatile as indicated by three-year periods that in high-technology and autos exhibited more stability and provided a significant contribution to total world trade growth. To minimize the impact of such volatility, a 10-year horizon is mostly considered, but shorter horizons have been tested as robustness checks.

Figure II.6.Average Annual U.S. Dollar Growth of World Trade in the Fastest-Growing Manufacturing Sectors

(1996-2005)

Source: UN Comtrade database.

Based on past trends, SEA-5’s manufacturing specialization appears to have been moderately adverse for subsequent (nominal) export growth. The negative result is quite strong for Italy, less so for other countries, while France’s specialization appears to be marginally “beneficial.” The SEA-5 countries are specialized in several large sectors that enjoyed high global growth over the past decade—motor vehicles/parts, telecom, and electrical equipment (Figure II.7).5 But correlations for all manufactured goods sectors (Table II.2) suggest that Italy, and to a lesser extent Spain, Portugal, and Greece, had a sectoral specialization that was inversely related to subsequent nominal growth in “global” trade in these sectors. By contrast, in France (and also in Germany and the U.S.) specialization was (marginally) positively associated with ex-post growth in this trade. To check the robustness of this inverse relation to the possibility that the overall negative relationship is driven by small sectors, estimates were weighted by the share of each sector in country exports, leaving the results essentially unchanged.

Figure II.7.Manufacturing Exports in SEA-5 and Germany: 1995–2005 1/

(Size of bubbles proportional to share in total goods exports of each country, largest 15 SITC-3 sectors for each country)

Source: UN Comtrade database.

1/ Excluding food and chemicals.

Table II.2.Sectoral Specialization and Subsequent Growth 1/(1995–2005, SITC manufacturing sectors)
OLS t-ratiosWLS t-ratios 2/
2-digit3-digit4-digit3-digit
Italy 3/-2.05*-2.81**-3.33*-3.05**
France1.040.131.700.06
Spain-0.40-2.16*-1.01-0.80
Portugal-1.49-1.28n/a-1.57
Greece-0.92-2.21*n/a-0.62
Germany1.170.050.47-0.93
Source: UN Comtrade database.Note: ** (*) denote significance at 1 (5) percent level.

Least squares between: (i) Balassa’s RCA index in 1995, defined as country’s world share of exports in a sector divided by its share of total world exports; and (ii) world trade growth in 1995–2005 in a sector in value terms. Coefficients and constants are not reported, given no clear hypothesis of causality.

Weights are given by shares of the given sector in national exports.

E.g., sectors in which Italy was specialized clearly tended to grow more slowly than other sectors (as the relationship is negative and statistically significant).

Source: UN Comtrade database.Note: ** (*) denote significance at 1 (5) percent level.

Least squares between: (i) Balassa’s RCA index in 1995, defined as country’s world share of exports in a sector divided by its share of total world exports; and (ii) world trade growth in 1995–2005 in a sector in value terms. Coefficients and constants are not reported, given no clear hypothesis of causality.

Weights are given by shares of the given sector in national exports.

E.g., sectors in which Italy was specialized clearly tended to grow more slowly than other sectors (as the relationship is negative and statistically significant).

The SEA-5 countries are also lagging in reorientation toward fast-growing activities, as reflected in the loss of export shares in these sectors. As expected, SEA-5’s own export structure has been shifting toward high-growth sectors (the sectoral correlations between the change as percent of own exports and the sector’s growth are positive for all SEA-5 countries). But the SEA-5 countries—particularly France, Italy, and Portugal—have seen their shares of world exports decline particularly in some of their largest export sectors that had high (global) growth, in marked contrast to Germany and Spain (Figure II.7). The negative relationship is also suggested by a simple regression gauging the determinants of market shares, which shows that for the large SEA-5 countries a higher global growth in the sector had a negative effect on their export shares in the same sector (Table II.3). The result is robust to several controls, such as initial export shares or effects of competition from emerging markets.

Table II.3.Determinants of Market Shares in Manufacturing in Large 1/ SEA-5 Countries and Germany

Panel regressions sector-fixed effects: 1996–99; 1999–2002; 2002–05 Regressand: percentage change in country’s value shares of world exports in sectorSITC-3, 399 observations

ItalyGermanySpainFrance
World export growth in sector-0.36-0.24-0.92-0.80
t-value-2.75-4.80-1.84-1.66
Level of country’s initial sectoral market share-6.46-5.49-50.94-7.88
t-value-3.77-5.87-2.95-1.93
Increase in China’s share of world exports in sector-3.240.714.333.55
t-value-2.381.701.170.96
R^20.270.330.090.05
Source: UN Comtrade database.Note: Underlined variables are significant at the 5 percent level; all regressions include a constant (not reported).

Regression results for Greece and Portugal were more erratic, likely because these small countries have too many special factors at the SITC-3 digit level in structure of the manufacturing exports.

Source: UN Comtrade database.Note: Underlined variables are significant at the 5 percent level; all regressions include a constant (not reported).

Regression results for Greece and Portugal were more erratic, likely because these small countries have too many special factors at the SITC-3 digit level in structure of the manufacturing exports.

SEA-5’s lack of reorientation toward growing sectors may in part reflect expanding shares of emerging markets. Germany—despite its recent broad export success—also tended to systematically lose market shares in high-growing sectors. But if one restricts the analysis to manufacturing exports by OECD, such negative link for Germany is no longer statistically significant, while this continued to be so (only slightly weaker) for the SEA-5 (Table II.4). Interestingly, the U.S., unlike the SEA-5, did not have a perceptible “bias” against fast-growing sectors despite its overall loss of market shares. One potential explanation (not formally tested) is that the U.S. may be more flexible in reallocating resources across activities, including to nonmanufacturing (services) sectors.

Table II.4.Relationship Between Changes in Manufacturing Export Shares and World Export Growth, 1995–2005(SITC 3-digit classification)
CoefficientsSEA-5GermanyU.S.
World export growth-0.13**-0.08*-0.06
OECD export growth-0.10**-0.05-0.02
Source: UN Comtrade database.Notes:1) Regression of percentage changes of the (SEA-5/Germany/U.S.) share of world/OECD exports in sector on nominal world/OECD export growth in the sector and a constant (not reported).2) ** and * denote significance at 1 and 5 percent level, respectively.
Source: UN Comtrade database.Notes:1) Regression of percentage changes of the (SEA-5/Germany/U.S.) share of world/OECD exports in sector on nominal world/OECD export growth in the sector and a constant (not reported).2) ** and * denote significance at 1 and 5 percent level, respectively.

D. Conclusions

While the SEA-5 countries have been deriving some benefits from exports to high-growth markets, they seem to be comparatively limited (at least based on data through 2005).They underperformed the EU or OECD on most aspects of “pro-growth” export structure and reorientation. Only part of this gap may be ascribed to “gravity” factors. The destination market with the greatest absolute untapped potential is China, but these opportunities are scattered across many emerging markets. From the perspective of high-growth sectors, SEA-5’s indicators of manufacturing export specialization and reorientation were also subpar. The overall underperformance was distributed differently across individual countries, with Italy particularly lagging in “sectoral,” and other countries on most “geographical,” measures.

While there is a clear scope to improve the profile of export markets and sectors, realizing these benefits for the SEA-5 countries would not be easy. The share of high-growing countries in world imports is so far limited, and most of them are geographically remote from the SEA-5. But potential opportunities could be understated by recent data, given that dynamic countries on aggregate have further scope to increase imports in tune with their strong external positions. Rigidities to resource allocation and small firm size in most SEA-5 countries are some of the factors that may need to be investigated to better understand what has been inhibiting these gains.

Appendix II.A. Definitions and Information Sources

1. Selected measures of trade performance vis-à-vis fast-growing economies:

  • Index of trade share weighted import growth of the trading partners—computed from the IMF WEO database;

  • Indicator of the level of “underexporting” to country j for the SEA-5 and a comparator (EU-13, OECD, Germany) is reported in Table II.1 for 2005 and defined as: MS(comparator)j/MS(SEA5)jMS(comparator)world/MS(SEA5)world where MS stands for share of the comparator or SEA-5 in j’s (or world’s) total imports; SEA-5 would “underexport” to country j if the indicator is higher than unity;

  • Indicator of falling behind relative to world import market share (WIMS) gains of country i between 2000 and 2005 is reported in Figure II.3 and defined as: (WIMSi2000SeaESi2000)(WIMSi2005SeaESi2005) where SeaES is the share of country i in SEA-5’s exports;

  • Indicator of export dynamics gap between the SEA-5 and country i is reported in Figure II.4 and is given by: impgri0005exp grSEAi0005 where impgr is cumulative nominal import growth of country i and expgr is the cumulative rate of nominal export growth from the SEA-5 to country i.

2. Lists of fast-growing economies:

Following Madariaga (2007), the definition of fast-growing economies is

  • Forty-three economies whose average annual real growth in 2004–08 has been estimated/projected at 4 percent or higher according to the Spring 2007 IMF World Economic Outlook (accounting for some 25 percent of world imports); another constraint is that at least 500 French enterprises export to these markets—this makes the broader sample somewhat France-centered, but it includes all relatively large countries and in many ways would fit the SEA-5 due to its geographical proximity to France; Taiwan was dropped since there are no data in the Comtrade database, but Kazakhstan was added;

  • Nineteen of the 43 economies above, whose average annual real growth in 2004–08 was estimated/projected at 6 percent or higher;

  • Eighteen economies whose actual and projected average annual real growth in 2004–08 was 4 percent or higher and whose share in world trade is at least 0.5 percent;

  • Eight economies whose actual and projected average annual real growth in 2004–08 was 6 percent or higher and whose share in world trade is at least 0.5 percent;

Note: Since Ireland and Luxembourg are listed as high-growing economies, they are excluded from the comparator market for the SEA-5.

3. Notes on data sources and definitions:

  • Geographical structure and reorientation—WEO and UN Comtrade databases.

  • Sectoral structure and reorientationUN Comtrade data, SITC two–four digit classification, “manufactured goods” data (excludes food and chemicals).

4. Definition of “market shares” used here—share of a country’s exports of a particular sector in world trade of this sector (this is different from the definition of “export market shares” that adjusts for own-country imports)

References

Prepared by Bogdan Lissovolik.

There may also be disadvantages to an export structure that is geared to high-growth destinations and sectors, for example due to a possibility of high volatility of this growth or of a less “sophisticated” quality of demand from the dynamic-but-not-yet advanced countries. However, there is no evidence yet that these disadvantages could be substantial enough to outweigh the advantages.

For example, measured export deflators may well understate quality upgrading (important in at least some SEA-5 countries), while the use of unit value indices as proxies is a source of biases (see Silver, 2007).

The definition of fast-growing “sectors” may be sensitive to coverage and the level of disaggregation. The analysis is confined to “manufactured goods” exports (excluding food and chemicals sectors) given their core role in these countries. The conclusions are mostly reported based on a SITC three-digit classification, but the two- and four-digit classifications were also checked.

The charts however, present somewhat limited information as they show only 15 out of 133 sectors, and this information gap is especially important for Italy, whose manufacturing export structure is very balanced.

Other Resources Citing This Publication