Determinants and Productivity Effects of Service Offshoring

Table of Contents

List of Tables

Table ofFigures

1. Introduction

2. Definition and differentiation of various related terms

3. Service offshoring as a new paradigm of trade in the US
3.1 Evolution of service offshoring in the US
3.2 Types of offshored service jobs
3.3 Offshore locations
3.4 Service offshoring volume

4. Determinants of service offshoring
4.1 Environmental drivers
4.1.1 Technological progress
4.1.2 Emergence of a global talent pool of educated labor
4.2 Firm-leveldrivers
4.2.1 Costreduction
4.2.2 Access to a qualified workforce
4.2.3 Prior offshoring experience
4.2.4 Other reasons
4.3 Coevolutionaryperspective
4.4 Risk factors

5. Productivity effects of service offshoring
5.1 Impact channels on productivity
5.2 The model ofFeenstra and Hanson
5.3 The trade in tasks model by Grossman and Rossi-Hansberg
5.3.1 Other literature
5.4 Empirical studies on service offshoring and productivity in theUS
5.4.1 Industry-level evidence
5.5 Empirical studies of offshoring and productivity in Europe
5.5.1 Industry-level evidence
5.6 Comparison ofUS and European data
5.6.1 Comparison of the theoretical groundwork
5.6.2 Comparison of quantitative results
5.6.3 Other empirical findings

6. Conclusion

Bibliography

Appendix

List ofTables

Table 1: Offshoring locations forUS companies

Table 2: Cumulative forecast of total offshored servicejobs per sector in the US, by occupation

Table 3: US-India bilateral trade data for Computer and Information services, in $US million

Table 4: Comparison ofUS and Indian hourly wages for selected service occupations during

Table 5: Indian real average daily wages by occupation, 2011/2012 (in Indian Rupee)

Table 6: Strategic drivers of service offshoring by industry

Table 7: Summary of potential risks perceived by firms in association with their offshoring

Table 8: Aggregated service input types: 1992 and

Table 9: Material and service offshoring between 1992 and 2000 intheUS

Table 10: Industry output composition by type

Table 11: Total offshoring by share of supplying region

Table ofFigures

Figure 1: Offshoring vs. outsourcing

Figure 2: Distribution of different offshoring functions

Figure 3: Cumulative percentage of companies engaging in service offshoring

Figure 4: US imported services by selected countries

Figure 5: US imports of services in millions ofUS dollars

Figure 6: US imports of goods and services in millions ofUS dollars between 2005. and

Figure 7: Determinants of western countries that offshore high-skilled and low-skilled.. services

Figure 8: Increase in offshoring due to changes in the unit cost ratio

Figure 9: Trade in goods vs. trade in tasks

Figure 10: Equilibrium of wage and costs of offshoring

Figure 11: Effects of a decrease in offshoring costs

Figure 12: ORN 2009 Corporate Client Survey profile

“Offshoring is generally believed to be productivity enhancing and this belief is underpinned by economic theory.” (Bernhard and Rycx (2014))

1. Introduction

Job losses, financial market crashes, and shortages of medical equipment arejust a few of the problems that the current Coronavirus crisis has brought to various countries. The Financial Crisis of 2008 and the recent trade war between the United States (US) and China also serve as reminders of how interdependent the world economy is due to increasing globalization. Since the production of critical medical equipment has been offshored to other countries, countries like America and Germany face a shortage of medical equipment. This fragmentation of the production chain can have negative impacts when individual countries are dependent on the production of a single country. With the recent failures of the global production chains in the Coronavirus crisis, politicians and economists are rethinking the benefits of globalization and offshoring now more than ever. A crisis like this reveals the substantial risks associated with offshoring, which is defined as the relocation of tasks abroad. The offshoring of vital products was just the beginning as firms have also begun to offshore core services such as research and development, product design, and legal services. Thus, the question can be asked whether one should set up productions of certain vital products and services at home rather than abroad. The answer will depend on whether or not the benefits outweigh the risks.

Globalization has drastically changed the competitive environment for businesses. Firms like Ford Motor Company used to produce all stages of production in the USA. Now the same firm produces parts or entire car models in several different countries and offshores business functions such as customer support to other countries. This fragmentation and international competition in today's business environment forces companies to adjust to a constantly evolving global market. To be competitive in the long run, market participants have to consistently deal with increasing customer requirements and market transparency. Competition forces firms to constantly lower their costs. However, pure cost-efficiency will not be enough to remain competitive as the imitation risks for cost strategies are high (Kenney et al. (2009)). With rapid improvements in information and communication technology coupled with continuing cost pressure, service offshoring to low-wage countries has become a more common solution. Due to technology, an increasing number of service jobs have become tradeable. Previously considered to be non-tradable, white-collarjobs are now being offshored, which has increased public concern about job losses. Several economists argue that service offshoring will likely affect the service jobs of higher-skilled labor more than lower-skilled labor in the future.¹ With the steady growth of service offshoring and its increasing involvement of higher- skilled labor, it is important to take a closer look at its general welfare effects. In particular, research about productivity, wage, and employment effects are crucial in understanding the implications of service offshoring.

Economic theory suggests that offshoring creates productivity-enhancing effects but literature in this area has been rather limited for service offshoring until recently (Bernhard and Rycx (2014)). Thus, the contribution at hand tries to provide an overview of why firms engage in service offshoring and examine how service offshoring affects productivity. The work is structured as follows: Section 2 first provides a definition of the related terms and narrows down the topic of the paper. Section 3 briefly discusses service offshoring as a new paradigm of trade in the US. It also addresses the types of offshored service jobs, the offshore locations as well as the current volume of service offshoring. Section 4 discusses the various determinants that facilitate service offshoring. Section 5 contains the core analysis of this paper with the goal of analyzing the productivity effects of service offshoring theoretically and empirically. First, a Heckscher-Ohlin model by Feenstra and Hanson (1996, 1997, 1999) will be introduced as one of the earliest models addressing offshoring and productivity. Then the “trade in tasks” model by Grossman and Rossi-Hansberg (2008) is presented to illustrate how service offshoring affects productivity. On an empirical level, the studies from Amiti and Wei (2009) and Schwörer (2013) will be discussed to address how service offshoring affects productivity specifically in the US and Europe. This will be followed by a comparison of the two empirical studies. Section 6 summarizes the work and adds some final remarks.

2. Definition and differentiation of various related terms

Clearly defining offshoring has proven to be a difficult task. Due to the use of many different terms and definitions in the literature, the debate around offshoring has been complicated. The misuse of the offshoring term and the use of different definitions for this phenomenon makes the comparison of studies difficult. Consequently, it seems important to start the discussion with a clear definition and differentiation of the related terms. First, to grasp the origin of offshoring in the global spectrum of trade, horizontal and vertical integration need to be addressed. Second, offshoring and outsourcing will be differentiated. Finally, service offshoring will be defined to properly narrow down the topic of this paper.

Horizontal integration

Horizontal integration refers to the process of merging with or acquiring similar companies to replicate the production process in other locations (Venables (1999)). The literature also refers to this as horizontal Foreign Direct Investment (FDI), when multinational enterprises (MNE) build subsidiaries in a foreign country (Bottini et al. (2007)). Horizontal integration or FDI is when a firm replicates the production processes in another country (Carr et al. (2001)). It’s important to mention that horizontal integration or FDI differs conceptually from offshoring.

Vertical integration

Vertical integration ox fragmentation refers to the division of the production process into several tasks that can be completed internally or externally and result in the same end product (Deardorff, A. (2001)).² Breaking the production process into different steps and contracting out the parts that can be sourced cheaper from another location enables firms to make use of international factor cost differences and task specialization (see Bottini et al. (2007), Jones and Kierzkowski (2001), Venables (1999)). Firms can increase their productivity by contracting out business functions and components whose in-house production is inefficient. In this process, a firm needs to conclude whether it is cost-effective or not to produce the intermediate input itself or buy it from an external supplier. The former refers to vertical FDI where MNE’s acquire a plant in a foreign country to produce intermediate components or services (Bottini et al. (2007)). Intra-firm trade occurs when firms import goods or services from their foreign subsidiaries.³

The differentiation of horizontal and vertical integration is difficult as both affect the trade volume in a different way: while horizontal integration tends to reduce the international trade volume, vertical integration increases it (Venables (1999)). The splitting of value chains on a global level generates new product groups of intermediate goods and services. This process increases the trade volume as the intermediate inputs are first imported and then final products are exported (Hummels et al. (2001)). Offshoring refers to vertical integration on an international level where firms source the production inputs from a foreign supplier.

Offshoring vs. outsourcing

Economists commonly use outsourcing and offshoring on an interchanging basis. Despite the confusion, the two terms are easily distinguishable based on two criteria. The Ownership of tasks where tasks can be completed internally (in-house) or supplied by an external supplier and the Location in which the tasks are completed (domestically or abroad).

Figure 1: Offshoringvs. outsourcing

Abbildung in dieser Leseprobe nicht enthalten

Source: Adapted from Bottini et al. (2007), figure l,p.3.

Figure 1 shows that the two criteria facilitate four combinations. When firms buy inputs or tasks from an external supplier rather than producing them in-house, then they engage in outsourcing.⁴ To further differentiate, international outsourcing occurs when the tasks are performed abroad rather than domestically.⁵ Conversely, offshoring refers to tasks that are relocated abroad regardless of whether they are owned or produced by a foreign supplier.⁶ Some literature also refers to the sourcing from a foreign external supplier as arm’s-length contracts (Bhagwati et al. (2004); Bottini et al. (2007); Amiti and Wei (2009)). Vertical FDI or captive offshoring occurs when firms buy a foreign affiliate or subsidiary to perform a certain task along the value chain and thus own this production unit. As previously mentioned, horizontal FDI does not fall under the offshoring term but vertical FDI does. Horizontal FDI is “market seeking” while vertical FDI or offshoring is “efficiency seeking” (Bottini et al. (2007)). This interpretation implies that companies move tasks abroad to increase productivity.

Service offshoring

Fora long time, only intermediate inputs or final goods were offshored and traded on a global scale. Due to the expansion of the internet and telecommunication technologies, the instant transfer of information over long distances has resulted in the ability to separate tasks geographically and temporally (Grossman and Rossi-Hansberg (2008)). Now offshoring includes all phases of the production process regardless of the nature of the task. Therefore, it is important to distinguish between material and immaterial steps of the production process. Material steps refer to the intermediate goods that are necessary for the production of a product, while immaterial steps refer to service activities or links that occur along the company’s value chain (Bottini et al. (2007)). Companies can offshore intermediate service inputs or entire business functions. As the focus of this paper is service offshoring, the discussion will relate to immaterial tasks. In short, service offshoring is the relocation of service tasks to foreign countries. As illustrated in Figure 1, service offshoring will include both captive offshoring (intra-firm trade) as well as arm’s length trade.

3. Service offshoring as a new paradigm of trade in the US

First, the evolution of service offshoring will be addressed including the aspects of globalization in general. In addition, this section will answer the questions about what types of service jobs are offshored and which offshore locations are preferred. Finally, it will be discussed how many jobs are offshored in the US with a critical look at the data provided.

3.1 Evolution of service offshoring in the US

Over the last few decades, international capital flows, global supply chains, as well as advancements of information and communication technologies have increasingly intertwined the world on a financial and economic level. This process is referred to as “globalization” which can be seen through the extent of economic integration in the world. The strong interdependence of the world economy can also have negative impacts (e.g., Financial Crisis of 2008). To put the emergence of service offshoring in context, one must address the different phases of globalization and international trade.

There were two waves of globalization. Feenstra (1998) calls the first era leading up to World War I as “the golden age of trade in investment worldwide” (p. 32). Firms focused on task specialization of labor and local agglomeration of production factories to gain high production productivity (Grossman and Rossi-Hansberg (2006)). The second wave of globalization represents the rise of trading with intermediate inputs due to the emergence of global production supply chains. Firms now capitalized on their core competencies and bought inputs from other suppliers for their less efficient production stages (Bottini et al. (2007)). These offshoring activities⁷ seem to have increased in the 1970’s (Feenstra (1998)). This wave expanded international trade and enhanced global competition.

The most recent wave of globalization is attributed to the trade in service inputs and service business functions.⁸ Even though the appearance of service offshoring and material offshoring partially overlaps, both can be divided into two different waves (Bottini et al. (2007)). Companies have outsourced and offshored business functions such as call centers, software development, accounting, or legal services since the 1980’s (Massini and Miozzo (2012)). Among the earliest service tasks offshored were IT applications (e.g., programming) to counteract the “Millennium Bug” (Lewin and Peeters (2006), p. 222).⁹ Raa and Wolff (2001) show that manufacturing companies in the US outsourced inefficient business services such as legal, accounting, or advertising tasks in the 1980’s and 9O’s.¹⁰ Likewise, Amiti and Wei (2005) found that one of the earliest digitally traced uses of the term “outsourcing” in relation to material and service offshoring in America have been found in 1980 in the Harvard Business Review (p. 313). Service offshoring activities accelerated in the late 1990’s (Bunyaratavej and Hahn (2012)) and drastically exploded at the beginning of the century (Hahn (2010)).

3.2 Types of offshored service jobs

To fully understand the determinants and productivity effects of service offshoring, one must address what service jobs are being offshored. With the emergence of global supply chains, manufacturing firms started offshoring less efficient service functions to focus on the tasks that strengthen their competitive advantage (Raa and Wolff (2001)). Thus, at the beginning of service offshoring, the majority of offshored services were in the administrative sector of manufacturing companies. This included back-office tasks such as human resources, accounting, finance, and procurement (Lewin et al. (2011)). As service offshoring became more common in the US, firms started to offshore higher-skilled service jobs. In their 2009 corporate client survey,¹¹ Lewin et al. (2011) reported that companies started to offshore more of their “innovation services” in the 2000’s, including important functions in research and development (R&D), product design, and engineering (p. 14).¹²

Figure 2: Distribution of different offshoring functions

Abbildung in dieser Leseprobe nicht enthalten

Source: Duke University, The Conference Board Offshoring Research Network 2009 Survey taken from Lewin et al. (2011), chart 14, p.14.

Figure 2 breaks down the engagement of firms in the various service types by percentage. Figure 3 illustrates the cumulative percentage of the different offshored service functions by all surveyed firms between 1990 and 2008. It is evident that the offshoring of IT and innovative services has drastically increased since the year 2000. While in 2000 both services were just above 10 percent, IT services reached 50 percent and innovative services almost 40 percent by 2008. These results indicate that higher-skilled service jobs are becoming more desirable to be offshored.

Abbildung in dieser Leseprobe nicht enthalten

Figure 3: Cumulative percentage of companies engaging in service offshoring

Source: Adjusted from Lewin et al. (2011), chart 26, p.23. Data sources are from various surveys conducted by Duke University, The Conference Board Offshoring Research Network.

3.3 Offshore locations

Another well-researched question is the location choice of US companies for their service offshoring activities. The offshore location choice is dependent on the firm’s offshoring determinants. Research implies that on a country level, firms consider education, labor costs, quality, and the host country’s technological infrastructure as decisive factors for their choice of location (Bunyaratavej et al. (2008), Demirbag and Glaister (2010), Doh et al. (2009)). In addition, cultural similarities and political risks can also affect the choice of offshore locations (Demirbag and Glaister (2010), Doh et al. (2009)). On the firm level, driving factors seem to depend on the nature of the offshored service task. Doh et al. (2009) recognize the different degrees of offshored services in terms of repetition, innovation, and interaction. They suggest that a specific combination of these factors influences the offshore location choice.

Most of the empirical research examines the relocation of services from developed countries like the USA or western Europe to developing countries in Asia (Pisani and Ricart (2016)).¹³ Lewin and Peeters (2006) discuss the results of the first survey by the Offshoring Research Network (ORN) in 2004 that surveyed 650 US Forbes 2000 companies about their service offshoring activities. Table 1 shows the location choice of survey respondents. The results clearly support that India was the most preferred offshore location for US companies. These results imply that India portrays an attractive location in terms of labor costs, quality, and skill level of its workforce as well as technological infrastructure.

Table 1: Offshoring locations for US companies

Abbildung in dieser Leseprobe nicht enthalten

Source: Adjusted from Lewin and Peeters (2006), table 4, p. 231.

Comparing Table 1 to the service import volume by country shown in Figure 4, one can see a strong discrepancy. Figure 4 shows the value of US imported services by country between 1999 and 2019. The highest import values come from countries like the United Kingdom (UK), Canada, Germany, and Japan which are absent in Table 1. This discrepancy once more shows the difficulty in analyzing service offshoring as the data and tracking of job-specific service offshoring is lacking.¹⁴ Only firm-level surveys provide precise data about how many jobs and what type ofjobs are offshored to each location.¹⁵

Abbildung in dieser Leseprobe nicht enthalten

Source: Retrieved from the US Bureau ofEconomic Analysis,” Table 8. U.S. International Trade by Selected Countries and Areas, Import of Services” https://www.bea.gov/system/files/2020-03/trad-time-series-0120.xlsx (accessed March 28, 2020).

In short, the location choice of US companies for service offshoring might change and expand over time. With the ability to transmit information instantly to any location in the world, the location itself does not matter. Ultimately, firms’ specific motivations to offshore certain service functions as well as the characteristics of the host country will influence their offshore location.

3.4 Service offshoring volume

One of the big questions in the media and research literature is how many service jobs are actually offshored. The media influence public opinion about the employment effects of service offshoring heavily. Especially during presidential campaigns in America, the topic of service offshoring increases drastically in the American news. During a five month period in 2004, US newspapers have published over 2,600 reports about service offshoring with the main focus on unemployment and job losses (Amiti and Wei (2005)).¹⁶ Since themajority of the US workforce is employed in the service sector, the fear of white-collar job losses in the service industry has risen consistently since the 2000’s (Levine (2011)). Despite all these perceptions, literature shows that service offshoring has been increasing consistently but does not account for a significant part of employment in the US (Amiti and Wei (2005)).

Generally, it is difficult to determine the number of offshored service jobs. Most studies provide mere estimations of the service offshoring volume as no governmental institution exists that is accounting for this number. Table 2 shows the forecast for a study that predicted US service sector jobs between 2003 and 2015. Forrester Research, Inc created these forecasts by a discussion with experts in the field. The study predicted that 1.2 million service jobs would be offshored by 2008 and about 3.4 million by2015.

Table 2: Cumulative forecast of total offshored service jobs per sector in the US, by occupation (number in thousands)

Abbildung in dieser Leseprobe nicht enthalten

Source: Taken from Levine (2011), table l,p. 6, originally adapted by CRS from John C. McCarthy, Near-Term Growth of Offshoring Accelerating, Forrester Research, Inc., May 14, 2004. The original source was not available.

In comparison to other studies, some economists find these numbers too conservative. For example, Goldman, Sachs & Company estimated that 6 million jobs would be lost to service offshoring in one decade (Garner (2004)). Bardhan and Kroll (2003) used three criteria when estimating the potential for service offshoring: the occupational structure of the US labor market, the offshoreability of a task, and the number of all offshored and planned to be offshored jobs. The authors predict that over 14 million jobs which accounted for 11 percent of the labor force in 2001 are at risk to be offshored. Given that the labor force in America is over 160 million (US Bureau of Labor Statistics)., these numbers do not seem that threatening to the American service sector. All of these studies must be viewed with caution as they provide only estimates and predictions by experts and not concrete data.

One way to get an idea of the level of service offshoring is by looking at trade data and service imports. Figure 5 depicts the total amount of imported services for the United States during 2005 and 2018.¹⁷ As the graph shows, service imports almost doubled from 300 billion US dollars in 2005 to over 560 billion in 2018. Service offshoring increased steadily over this time period with a minor decrease during the Financial Crisis in 2008.

Figure 5: US imports of services in millions of US dollars

Abbildung in dieser Leseprobe nicht enthalten

Source: International Monetary Fund (IMF): Balance of Payment for United States between 2005-2018 (http://data.imf.org/regular.aspx?key=62805740)

Figure 6 portrays the total amount of imported goods and services to put service offshoring in context with material offshoring. Obviously, service offshoring levels have been steadily rising but are still much lower than material offshoring. Offshoring levels are an important consideration in the discussion about service offshoring and the fear of white-collarjob losses.

Figure 6 also clearly depicts that the Financial Crisis had a much bigger impact on trading in goods than on trading in services.¹⁸

Figure 6: US imports of goods and services in millions of US dollars between 2005 and 2018

Abbildung in dieser Leseprobe nicht enthalten

Source: International Monetary Fund (IMF): Balance ofPayment for United States between 2005-2018 (http://data.imf.org/regular.aspx?kev=62805740i.

These trade statistics provide a good indication of service offshoring levels but fail to paint an accurate picture. Obtaining exact import statistics is difficult because the tracking of services is often unavailable and much more complex than material offshoring. There are two reasons for this challenge. International trade statistics fail to account for service trade as well as for statistics about the intra-firm trade of multinational companies (Kirkegaard (2004)).¹⁹ Table 3 exemplifies this statistical problem of trade data. The US Bureau of Economic Analysis reports that the US imported $80 million in Computer and Information services from India in 2002; however, India’s export statistics show that India exported about $6.4 billion of the same services to the US and Canada.

Table 3: US-India bilateral trade data for Computer and Information services, in $US million

Abbildung in dieser Leseprobe nicht enthalten

Source: Kirkegaard (2004), table l,p.25.

Note: This table shows the problem of trade statistics by comparing US-Indian import and export data. *Mode 1 refers to traditional trade in services with arm’s length contracts while Mode 3 refers to captive offshoring (intra-firm trade). **The data shown for 1998 relates to the time period between April 1998 and March 1999 since the data of IndiaStat refers to fiscal years.

This strong discrepancy in reporting exists because both countries incorporate different data. The US data only includes arm's length trade in services, while the statistics from India also include intra-firm trade of multinationals.²⁰ This inconsistency in reporting suggests that the trade data about service offshoring can contain unreliable data. Thus, surveying companies and using firm-level data provides the most reliable data about service offshoring (Brainard and Litan (2004)). However, it is not only expensive and time-consuming to survey individual firms, but firms also seem hesitant to present this data due to the negative public perception about service offshoring (Kirkegaard (2004)).

In summary, service offshoring is consistently rising since the beginning of the century. Imports of services have almost doubled between 2005 and 2018 to 560 billion US dollars. However, an exact number of offshored servicejobs in the US is not available. American firms mostly offshore low-skilled service jobs to low-wage countries. However, US firms also increasingly offshore high-skilled jobs that involve core functions such as R&D or software development. The motivations behind service offshoring will be addressed in the following section.

[...]

---

¹ See for example Blinder (2006) and Crino (2010).

² The problem with the term “fragmentation” or “vertical integration” is that is does not indicate whether the tasks and production steps are sourced internationally or domestically.

³ This is an important aspect as it affects trade statistics.

⁴ Some literature does not specify the location (domestic or international) which makes the term outsourcing an unfortunate choice of word for this phenomenon as there is no clear indication of where the tasks are relocated to. A similar problem exists for other terms such as fragmentation.

⁵ International outsourcing coincides with offshoring.

⁶ The decision about the ownership of the task has a significant impact on the company from an economic perspective.

⁷ Feenstra refers to this as “outsourcing” in his paper.

⁸ Even though service offshoring appeared earlier with the fragmentation of production processes, trade in services has been comparatively low to trade in goods. Thus, trade in service is attributed to the third wave of globalization. Also, the implications of service offshoring as well as its measurement are much different than material offshoring.

⁹ To keep up with the continuous software development and software updates, companies offshored software development tasks to countries like India.

¹⁰ Even though the authors do not distinguish between domestic and international outsourcing, it is likely that at least some of these services have been offshored.

¹¹ The survey was conducted by the Offshoring Research Network (ORN). Since November 2010, the ORN amassed over 2,000 companies from different industries (22 percent large, 35 percent mid-size, and 43 percent small) from different countries and over 4,300 distinct offshoring projects in their database (p. 26). See Appendix (Figure 12) forthe 2009 client survey profile.

¹² The survey includes responses from firms from North America, Europe, Australia, and other countries. Since 44% of the surveyed companies are headquartered in North America, these findings provide good indications for the trends in America.

¹³ India in particular has been studied extensively.

¹⁴ The data provided by the US Bureau of Economic Analysis gives us a good indication about the value of service trade but does not explain what type of services are offshored.

¹⁵ However, such data does not provide a complete picture of all offshoring activities because it represents only a small sample size.

¹⁶ This media coverage of service offshoring and fear of unemployment is also the topic of many other developed countries such as Germany or the United Kingdom.

¹⁷ The data ofUS service imports from previous years is not provided by the IMF.

¹⁸ A similar scenario can be expected with the current Coronavirus crisis which probably will have much higher effects on goods trade than on service trade.

¹⁹ The immateriality of services crossing borders makes measuring difficult. Trade statistics rely on a mix of sources such as business surveys, accounting reports, and statistical measurements by agencies. Kirkegaard exemplifies this measuring problem with the exports of multinational companies. According totheUS Bureau of Economic Analysis, the US exported $276 billion worth in services in 2001 while multinationals sold $432 billion worth of services to foreign markets through foreign-owned companies (Kirkegaard (2004), p. 25).

²⁰ It could also be that India exported large amounts to Canada but that seems unlikely.