VIII. Outsourcing and Competitiveness in Southern Europe1
As a result of globalization, firms in countries around the world have been engaged in a “fragmentation” of the production process, i.e., breaking the production process into smaller tasks and carrying them out where they can be accomplished most cheaply. This process has been referred to as “outsourcing” and it can take a variety of forms. Outsourcing can take place domestically, as when a firm procures the services of an supplier who is located in the same country, or internationally, as when a foreign firm provides the service. When the latter occurs, it is termed “offshoring” and this practice has received a great deal of press attention in recent years because it is sometimes alleged that it leads to job losses in the home economy. Both domestic outsourcing and foreign outsourcing give rise to trade, either domestically between firms or internationally.
In terms of the simplest possible model of international trade—the Ricardian model— offshoring is beneficial to the economy that engages in it, based on the logic of comparative advantage. If an activity can be undertaken at a lower cost abroad rather than at home, it is beneficial for the home country to let a foreign provider produce the good or service and then import it. The home country benefits since it would enjoy gains from specialization and trade. To the extent that a country is able to reduce its production costs by importing components or services from countries where they can be produced at lower costs (for the same level of quality), industries in the importing country become more competitive—they can then produce the same level of output as they did prior to engaging in offshoring but at a lower cost. Exporters who use imported intermediate inputs would enjoy a cost advantage, relative to exporters who do not engage in offshoring. Offshoring should then improve the competitiveness of the country that engages in offshoring.
In addition to reducing production, i.e., inputs costs, offshoring may affect the competitiveness of the home economy through positive productivity spillovers. One reason why a domestic firm may engage in offshoring is that a foreign provider can produce a good or service with a superior technology. By offshoring an activity to a foreign provider and then importing the good or service, the domestic firm could learn from the better technology and eventually adopt it at home. Thus, offshoring could raise productivity at home to the extent that the domestic firm benefits from “spillovers” that arise from exposure to a better production activity.2 For example, a foreign firm may use a type of computer software that accomplishes a task more efficiently than the domestic firm. As a result of offshoring, the domestic firm may adopt this software, which would improve the technology of the domestic firm and the productivity of its workers.
This chapter examines how offshoring has affected the competitive position of the southern euro area five countries (SEA-5): France, Greece, Italy, Portugal, and Spain. This chapter is divided into two parts. Part one reviews available data on the extent to which these five countries, along with a number of other European countries, engage in offshoring. Part two reports the results of some empirical work that investigated the role of offshoring in influencing developments in productivity and competitiveness in the countries noted above. In short, it turns out that while offshoring has been growing in most of these countries over the period 1995–2003, the overall level of offshoring is rather small. Notwithstanding this, there is some empirical evidence to suggest that offshoring has improved the competitive situation of these countries.
A. Offshoring in Five European Countries: What Do the Data Say?
There is no one definitive indicator of offshoring activity, as it is a complex phenomenon, and no country keeps statistics specifically designed to capture offshoring. Previous empirical literature, e.g., Amiti and Wei (2005) and Feenstra and Hanson (1996), used two indicators of offshoring activity. The first involves examining trends in imports of business and computer services using balance of payments statistics. These two categories represent activities in which offshoring is likely to take place. For example, when a country offshores customer service to a foreign provider, this represents an import of a business service by the home country.
Table VIII.1 presents data on imports of business and computer and information services for EU-15 countries for 2005, except for Denmark. The data reveal that the five EU countries of interest had rather modest levels of imports of business and computer and information services in 2005. Imports of business services varied between 0.5 and 1.8 percent of GDP for France, Greece, Italy, Portugal, and Spain in 2005. In relation to the other 132 countries for which data exist, these five countries ranked toward the lower half of the distribution. Imports of business services were particularly small for Greece (0.5 percent of GDP)—it ranked one-hundred fifth out of 132 countries. Some EU countries had very large levels of imports of business services, such as Ireland, Luxembourg, and Austria.
Also shown in Table VIII.1, imports of computer and information services were quite small for all 132 countries in the dataset, and Luxembourg ranked first, with imports of 1.9 percent of GDP. For the five EU countries of interest, they actually had levels of imports of computer and information services that exceeded most other countries—these five countries ranked between thirty-fourth and fifty-sixth overall. The overall magnitudes were small, however, at about 0.1–0.2 percent of GDP.
Table VIII.2 demonstrates that with few exceptions, imports of business and computer services grew more rapidly than imports of all goods and services over the period 1995–2005 for EU-15 countries. For comparison, the United States is included in Table VIII.2. In two countries—Finland and Greece—the average annual growth rate of all imports of goods and services outpaced the average growth rate of imports in business and computer services. For France and Luxembourg, imports of business services grew more slowly than total imports, but imports of computer services grew more than twice as fast as total imports of goods and services. For Ireland, imports of computer services grew much more slowly than total imports of goods and services.
A second frequently used indicator of offshoring activity is constructed using two pieces of data: (i) data on input-output linkages in an economy; and (ii) trade data. Feenstra and Hanson (1996), Amiti and Wei (2005), IMF (2007), OECD (2007) and Molnar, Pain, and Taglioni (2007) constructed an indicator of offshoring intensity within an economy given by the following formula:
where OSSi is a measure of offshoring intensity within sector i. The first bracketed term on the right-hand side of this formula is obtained from country input-output tables. For any given sector i, the input-output table reveals how much of the output of each sector in the economy is used as an intermediate input by sector i. The various sectors denoted by j represent the activities that would be considered as offshoring activities. Amiti and Wei (2005) considered five: (i) telecommunications; (ii) insurance; (iii) finance; (iv) business services; and (iv) computing and information services. To arrive at a measure of the extent to which a given sector i used imported inputs, data on imported inputs by sector would be needed, but this is not generally available. As an approximation, researchers commonly apply an economy-wide average import share to each industry. The second bracketed term is the share of imported service j in total demand for service j. The product of the two bracketed terms gives a measure of offshoring in industry i. For example, if 10 percent of the total inputs of sector i come from service sectors, which could be outsourced, and if in the economy as whole, total imports of service j is 40 percent of domestic demand, then the measure of offshoring intensity is (0.1)*(0.4) = .04. In the absence of data on imported inputs by sector, this measure of offshoring intensity assumes that a given sector i uses imports of service j in the same proportion as the economy as a whole uses imports of service j.
Using input-output data, an aggregate measure of offshoring intensity was calculated for a number of advanced economies and presented in IMF (2007). Somewhat surprisingly, two countries—the United Kingdom and France—exhibited a decline in their overall offshoring intensity over this period. Italy, Greece, Portugal, and especially Spain exhibited increases in offshoring intensity. Although the measure of offshoring intensity rose for most countries between 1995 and 2003, offshoring intensity peaked in 2000/01 and declined thereafter until 2003.
B. Offshoring, Productivity, and Competitiveness
The overarching question posed in this chapter is how does offshoring affect the competitiveness of an economy. This section reports the results of two ways of answering this question: (i) it presents a series of charts that depict movements in a commonly used indicator of competitiveness—real effective exchange rates based on consumer prices—and the indicator of offshoring intensity using input-output information; and (ii) it reports the results from regressions that try to uncover a systematic relationship between offshoring and competitiveness. Generally speaking, greater offshoring seems to be associated with a more depreciated real exchange rate for many countries, although the evidence that offshoring causes this result is somewhat tentative. A key issue is the extent to which offshoring affects productivity and the evidence of a systematic relationship between these two variables is weaker. While it has not been possible to uncover very strong evidence of a causal relationship between offshoring and competitiveness, this does not mean that one does not exist.
Data on real effective exchange rates (REERs) based on consumer prices indices (CPIs) and unit labor costs (ULCs) in manufacturing, productivity (labor and total factor), and offshoring were collected for 16 advanced economies.3 These countries were chosen mainly based on the availability of data on offshoring intensity. The data on real effective exchange rates were obtained from the IMF and Eurostat. Data on productivity (both total factor and labor productivity) were taken from the OECD’s productivity database. Data on offshoring intensity were constructed from input-output tables available from the OECD and were published in IMF (2007). Data on all variables were obtained for the period 1995–2003.
Figure VIII.1 depicts the relationship between competitiveness and offshoring for France, Greece, Italy, Portugal, and Spain. This figure reveals that between 1995 and 2003, a greater offshoring intensity was associated with a more depreciated real exchange rate based on CPIs for all countries except France. When real effective exchange rates based on ULCs were used as an indicator of competitiveness, a greater offshoring intensity was correlated with a more depreciated real exchange rate for all countries with the exception of France and Portugal.
Figure VIII.1.Relationship Between Offshoring Intensity and Competitiveness
Sources: Eurostat; and IMF, IFS.
To explore the relationship between competitiveness and offshoring further, the change in the (log) of the real effective exchange rate was regressed on the change in the log of the measure of offshoring intensity:
Pooling the data, i.e., estimating a value for β that is common across all countries, yielded a value for β that had the expected sign and was statistically significant at the 5 percent level. That is, greater offshoring was associated with a more depreciated real effective exchange rate. This result obtained for both measures of the real effective exchange rate. The dataset was also used to estimate a separate β for each country. Of the 16 countries in the dataset, the coefficient on the measure of offshoring intensity was of the correct sign for 13 countries, but statistically significant at the 5 percent level for only Belgium, Finland, Germany, Japan, and the United Kingdom. The coefficient on the measure of offshoring intensity was of the wrong sign for Denmark, France, and the Netherlands.
A key question regarding the impact of offshoring is the extent to which it affects both labor and total factor productivity (TFP). The relationship between total factor productivity and offshoring is depicted in Figure VIII.2 for France, Greece, Italy, Portugal, and Spain. These simple charts show that developments in total factor productivity follow movements in offshoring fairly closely for Italy, Portugal, and Spain. However, the correlation between changes in total factor productivity and offshoring is much weaker for Greece and especially France.
Figure VIII.2.Relationship Between Offshoring Intensity and Total Factor Productivity (TFP)
If offshoring causes productivity to increase at a faster rate in the traded sectors compared to the nontraded sectors, the real exchange rate could appreciate, due to a Balassa-Samuelson type effect. To investigate the relationship between total factor productivity and offshoring more systematically, the log change in both total factor and labor productivity was regressed on the log change in the offshoring intensity for the 16 countries in the dataset. As before, two sets of regressions are run—one that estimates a common coefficient for all countries and one that estimates a country-specific coefficient for the measure of offshoring intensity. In estimating a common coefficient for the measure of offshoring intensity, the results revealed that greater offshoring was associated with both higher TFP and labor productivity, but the results were not statistically significant for either type of productivity. Estimating a country specific impact of offshoring revealed a positive relationship between offshoring and total factor productivity for 10 countries, but only two were statistically significant at the 5 percent level: Finland and the Netherlands. For labor productivity, the results were similar. For 10 countries, the estimated relationship between offshoring intensity and labor productivity was positive but statistically significant at the 5 percent level for only Portugal and the Netherlands.
Existing studies of the impact of offshoring on productivity have failed to find convincing evidence of a causal relationship. For example, Egger and Egger (2006) examined the impact of materials offshoring in the manufacturing sector of the EU-12 on labor productivity and found that in the short run, the effect was negative but positive in the long run. One possible explanation for this is the low degree of flexibility in European labor markets. In the short run, it is difficult to reduce employment, so offshoring does not generate an improvement in labor productivity that would come from a reduction in employment. In a study of just Austria, Egger, Pfaffermayr, and Woldmayr-Schnitzer (2001) found that offshoring raised TFP in the manufacturing sector but less so in low-skilled sectors and more so in high-skilled sectors.
Finally, a much more complex set of regressions were run using the specification adopted in the IMF’s CGER exercise. Specifically, the “equilibrium exchange rate approach” was adopted, which uses panel regression techniques to estimate a relationship between real exchange rates a set of fundamentals.4 The equilibrium approach regresses the real exchange rate on the following variables: (i) the productivity of tradables to nontradables relative to trading partners; (ii) the commodity terms of trade; (iii) the ratio of net foreign assets to trade; and (iv) the ratio of government consumption to GDP. To this list was added the measure of offshoring intensity described above.
This empirical strategy reveals several strong results. First, when pooling data and estimating a common coefficient for the measure of offshoring intensity, the results showed that greater offshoring leads to a more depreciated exchange rate, and this relationship was highly significant. However, the inclusion of the measure of offshoring intensity in the regressions caused the sign of the coefficient on the relative productivity variable to switch from positive to negative—to contradict the Balassa-Samuelson effect. This same result occurred when a country-specific coefficient for the offshoring intensity was estimated.
The phenomenon of offshoring has received a great deal of attention in recent years as it has been growing in many advanced economies, albeit from low levels. Economic theory would suggest that offshoring would benefit an economy because it allows domestic producers to contact out various sub-activities of the production process to foreign providers who can produce the activity or component more cheaply and then import it. Thus, offshoring reduces production costs for domestic firms and makes them more efficient. Also, there are reasons to believe that offshoring can raise productivity at home.
For most European countries—the EU-15—offshoring activity has generally grown over the period between 1995 and 2003 and quite rapidly for some countries, although there has been a bit of a slowdown since 2000. Levels of offshoring activity—as measured by imports of business and computer services—are generally small: less than 2 percent of GDP for imports of business services and around 0.2 percent of GDP for imports of computer and information services for France, Greece, Italy, Portugal, and Spain in 2005. Data also suggest that changes in offshoring activity over the period between 1995 and 2003 have been associated with more depreciated real effective exchange rates for most of the countries examined, but this relationship is statistically significant for only five of 16 countries examined.
The data also reveal a positive relationship between offshoring and productivity (both TFP and labor productivity) for most countries over this same period, but this relationship is not statistically significant. The lack of very strong evidence of a positive relationship between offshoring and competitiveness and productivity could be due to a number of factors, including the lack of a direct measure of the extent to which a country engages in offshoring. All of the measures of offshoring used in this analysis were of an indirect nature. Also, data on offshoring intensity are only available for a relatively short time period for most countries. A longer time series might reveal a stronger relationship. Nevertheless, there is some empirical evidence to support at least some relationship between offshoring and improved competitiveness.
Amiti, M., and S.J.Wei, 2005, “Service Offshoring, Productivity, and Employment: Evidence From the United States.”
Egger, H., and P.Egger, 2006, “International Outsourcing and the Productivity of Low-skilled Labour in the EU,” Economic Inquiry, Vol. 44, Issue 1, pp. 98–108.
Egger, P., M.Pfaffermayr, and Y.Wolfmayr-Schnitzer, 2001, “The International Fragmentation of Austrian Manufacturing: The Effects of Outsourcing on Productivity and Wages,” North American Journal of Economics and Finance, Vol. 12, Issue 3. pp. 257–72.
Feenstra, R., and G.Hanson, 1996, “Globalization, Outsourcing, and Wage Inequality,” American Economic Review, Vol. LXXXVI, pp. 240–45.
Gordo, E., C.Martin, and T.Patrocinio, 2008, “La Internacionalizacion de las empresas espanolas a través de la inversion extranjera directa,” Boletin Economico (Bank of Spain, January).
International Monetary Fund, 2006, “Methodology For CGER Exchange Rate Assessments.”
International Monetary Fund, 2007, “The Globalization of Labor,” World Economic Outlook: Spillovers and Cycles in the Global Economy, Chapter 5.
Molnar, M., N.Pain, and D.Taglioni, 2007, “The Internationalisation of Production, International Outsourcing, and Employment” Working Paper 561 (Paris: Organization For Economic Cooperation and Development, Economics Department).
Prepared by Stephen Tokarick.
Positive productivity spillovers may also arise as a result of direct foreign investment. For a discussion of this, see Gordo, Martin, and Patrocinio (2008).
The countries included Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Italy, Japan, the Netherlands, Portugal, Spain, Sweden, the United Kingdom, and the United States.
See “Methodology For CGER Exchange Rate Assessments,” 2006, for details.