Chapter 1. The rise of services in the global economy

Chapter 1 assesses the importance of the services sector within the overall economy. It discusses the role of services in recent macroeconomic outcomes, as well as the marked trend towards manufacturing becoming increasingly services-dependent. The chapter highlights some important megatrends, particularly the digital economy, which is largely supported by the services sector, and discusses the role of establishing appropriate services regulations in order to leverage the digital revolution to strengthen productivity. In light of the significant role that small and medium sized enterprises (SMEs) play in the economies of most countries, it considers policies in the services field that can facilitate their international development, both through internet platforms and enhanced participation in global value chains (GVCs).


The structural transformation towards services

Economic development is associated with a shift from an agrarian economy dominated by subsistence farming to an industrialised economy dominated by services. Figure 1.1 shows that the services share of GDP has increased over time across all country groups, with the fastest rise in middle-income countries. Furthermore, at any point in time the services share is on average higher the richer the country is.

Figure 1.1. Services share of GDP

Source: World Bank, WDI.

The counterpart of a rising services share is a declining share of manufacturing and primary sectors in global GDP. The rising share of services and the corresponding decline in the share of manufacturing can be explained by faster technical development in manufacturing than in services, a shift in demand from manufactured products to services with rising income levels, and falling relative prices of manufactures. At the same time, in absolute terms output from both agriculture and manufacturing continues to grow steadily. For example, while the share of manufacturing in world nominal GDP declined from 20% in 1997 to 15% in 2015, the absolute level of global manufacturing output expanded by 44%.1

Structural changes in employment follow a similar pattern to those in industrial structure, as illustrated in Figure 1.2. It presents structural movements in employment for high- and middle-income countries during the period 1994-2010, and also for the People’s Republic of China (hereafter “China”) and India separately. The most significant reallocations in recent years have been in middle-income countries, where rapid urbanisation has drawn people from agriculture to urban jobs in industry and services. Comparing China and India uncovers a paradox. Over the 15-year period, China has become the manufacturing hub of the world while India has turned into a leading services exporter. Yet the services share in total employment rose twice as fast as the share of industry in China, while the opposite is true in India. Furthermore, the services share in total employment is much higher in China than it is in India. The explanation for this is three-fold. The transition from an agrarian economy is further advanced in China than in India. Thus, the share of agriculture declined from about half to about a third of total employment in China, while half of India’s labour force is still in agriculture. Second, India has seen an unprecedented spurt in labour productivity in export-oriented services (Benz et al., 2017). Third, a supporting services sector has been an essential but largely overlooked part of China’s industrial development.

Figure 1.2. Employment by sector group and two country groups, plus China and India

Source: World Bank WDI and Statistical Yearbook for China.

The structural shifts depicted in Figures 1.1 and 1.2 reveal that globally the share of employment in services has increased faster than the share of services in nominal GDP, which means that on average labour productivity has grown faster in manufacturing than in services. This has raised concerns that services-intensive economies may be less dynamic and create relatively fewer secure and well-paying jobs compared to the heyday of manufacturing employment. Such worries have led to a debate on a possible revival of classical industrial policy as a means of fostering an industrial renaissance accompanied by good jobs for low- and medium-skilled workers. However, the driving force behind a recent slow-down in offshoring and the trickling back of manufacturing to high-income countries appears to be precisely the diminishing importance of labour costs as the fourth industrial revolution unfolds.

The fourth industrial revolution is characterised by computer-integrated manufacturing, additive manufacturing, automation and advanced analytics of Big Data and the flow of information over the Internet of Things (IoT). During this transition, high-skilled manufacturing employment has been maintained both in high- and middle-income countries while low-skilled manufacturing employment has plunged (Rodrik, 2016). Judging from such trends, the employment share in services is likely to climb until it levels off at around 80-85%, while high-skilled jobs in manufacturing may continue to show resilience to technical changes. Like structural changes in the past, the ongoing transformation has immense potential for improving human well-being if managed wisely.2

Modern services are catching up with manufacturing productivity levels

The main driver of economic growth is a rising level of productivity. It is true that on average the level of productivity is higher in manufacturing than in services, but modern services are rapidly catching up. Table 1.1 presents recent developments in productivity for selected sectors in four countries.3 At least one services sector has experienced faster productivity growth than manufacturing in all four countries. It is notable that wholesale and retail trade has undergone a higher rate of productivity growth than manufacturing in three of the four countries in recent years. This is a labour-intensive sector that has been among the early adopters of productivity-enhancing ICT in bricks-and-mortar retail and is currently migrating to digital platforms.4

Table 1.1. Productivity (TFP) by sector, 2014
Index, 2010 = 100



United Kingdom





Total manufacturing






Wholesale and retail trade; repair of motor vehicles and motorcycles






Transportation and storage






Information and communication






Publishing, audio-visual and broadcasting activities












IT and other information services






Financial and insurance activities






Professional, scientific, technical, administrative and support service activities






Arts, entertainment, recreation and other service activities





Source: EU KLEMS database,

The professional, technical and other business services covered under ISIC codes M-N are a strong source of job creation for skilled workers and, with the exception of Germany, employ more people than manufacturing does. In this sector, productivity has kept up with manufacturing only in the United Kingdom. However, the digital transformation of professional services is still in its infancy with the promise of substantial productivity growth in the future. Furthermore, a number of studies have found that knowledge-intensive business services (KIBS) stimulate innovation and productivity in their client firms, including manufacturers. KIBS are typically coproduced with clients and their impact on productivity may be more visible at the aggregate economy level than in the KIBS sector.5

Comparable information on total factor productivity for emerging market economies is not readily available, but the WORLD-KLEMS database contains such information for India covering the period 1980-2008/9. Interestingly, TFP growth in services by far outperformed that in manufacturing in TFP growth in India during this period. Taking 1990 levels as the benchmark, the best-performing manufacturing sector (non-metallic mineral products) raised TFP by about 40%, while in the best-performing services sector (post and telecommunications) productivity grew six-fold over the three decades covered. India has embarked on a services export-led growth path no less impressive than previous growth spurts experienced by newly industrialised countries in the past. The experience of India and China, together with observations from de-industrialising high-income countries, suggests that an export-oriented services sector is possible even in the absence of a significant industrial base, while an export-oriented industrial sector without supporting services is unlikely.

Services enhance productivity and competitiveness in manufacturing

Services provide essential inputs to goods-producing sectors, enhancing productivity or competitiveness or both. Education and health contribute to raising the skill levels and adaptability of workers. R&D, engineering and design are essential inputs into product and process innovations, while telecommunications, and computer and information services, underpin the seamless transfer of knowledge and information that are the raw material of innovation, within and across borders. Finally, a host of business services help manufacturers relate to customers, suppliers and regulators. Also, mature industries are reinventing their business strategies in ways that rely more heavily on services as inputs. For example, consumer products such as clothing have shorter and more numerous seasons where market monitoring and market analysis feed back into design and production in order to keep up with consumers’ ever shifting tastes.

Manufacturing firms increasingly add services to their product portfolio, typically as complements to the goods that they produce. Examples from the business-to-business market are machinery manufacturers that sell a services package to clients including the rental and maintenance of machinery. The machines are fitted with sensors linked back to suppliers through the IoT. Suppliers monitor performance continuously and help reduce maintenance costs and downtime for the client. Such arrangements are found for a broad range of machinery, from aircraft jet engines (“power by the hour”) to copy machines or coffee machines in offices.

In the business-to-consumer market, after-sales services are major sources of revenue for manufacturers. The automotive sector, for instance, has relied on services for their revenues and profits for many years.6 Today, with growing traffic congestion and environmental concerns, self-driving cars and various car-sharing and car-hailing services have shifted the focus from the car to the transport services it provides. All the major automotive manufacturers have teamed up with car-hailing services, car rental or leasing services or software and geo-positioning services, bringing joint provision of manufactured goods and services to a new level. Finally, visibility on social media has become indispensable in consumer markets and in many cases comes with innovative ways of bundling goods and services. In the food industry, for example, recipes and nutrition advice on social media play a prominent role in positioning products. In general, consumer products are increasingly marketed on social media as elements of the lifestyle of the targeted market segment.

The joint supply of goods and services is sometimes a defensive response to tougher competition rather than a creative process designing premium products. Moreover, it has costs and, for some manufacturers, smaller profit margins and a higher risk of bankruptcy. A number of case studies demonstrate that successful joint supply relies on well-integrated services, product and information modules, using IT systems that standardise and modularise the bundled services.7

Knowledge-capturing products: Making services tradable across-borders

The services sector is constantly evolving, adding new services and rendering others obsolete through a combination of deeper division of labour and technology. To capture the dynamics of the services sectors, the most recent manual for national accounts (SNA, 2008) added a separate category labelled knowledge-capturing products, defined as follows:

“Knowledge-capturing products concern the provision, storage, communication and dissemination of information, advice and entertainment in such a way that the consuming unit can access the knowledge repeatedly.” [SNA, 2008, paragraph 6.22]8

This category differs from other services in two important ways: the product can be stored and it can be accessed and used repeatedly. These are also the properties that make services tradable across borders without direct interaction between services supplier and customer.9 Furthermore, repeated use at near-zero additional cost makes huge productivity gains possible.

New policy challenges emerge with the proliferation of knowledge-capturing products. Near-zero marginal costs are associated with very significant economies of scale that may result in “winner takes all” dynamics. Moreover, in a borderless digital economy the winners may be global while regulators operate within national borders. Another policy challenge related to repeated use at near-zero cost concerns the currently established systems for enforcing intellectual property and associated commercial rights, privacy and security. As knowledge and information go digital, the functionality of these systems needs to be assessed.

In sum, a rising share of services in the global economy has been observed for decades, and reflects both demand and technology factors. As people become richer, older and more urban they consume relatively more services. Many services sectors are catching up with manufacturing as far as productivity is concerned. Finally, goods and services are closely related and need to be understood as complementary inputs in production and consumption.

Trade in services

The great trade slowdown: Services as an engine of global trade growth

The value of global trade has stagnated since the financial crisis began in 2008. In the decade before the crisis, world trade grew almost twice as fast as world GDP, but in recent post-crisis years, world trade has lagged behind world GDP growth. Figure 1.3 compares trade in goods and services since 2005, three years before the crisis. Merchandise and commercial services exports follow a similar pattern, but services exports are significantly less volatile than merchandise trade. Possible explanations for more resilient services trade are that demand is less cyclical, services are less dependent on external finance and they faced fewer protectionist measures after the crisis (Borchert and Mattoo, 2010). However, when measured in value-added terms, services trade is not more resilient than goods trade. Bearing in mind the high share of traded services that is embodied in other sectors’ exports (see below), this is not surprising (Nagengast and Stehrer, 2016).

Figure 1.3. World exports since the financial crisis
Value index, 2008 = 100

Source: Authors’ estimates based on WTO statistics.

The great trade slowdown can only partly be explained by depressed demand in the global economy during the crisis and the shallow recovery. Structural changes, notably the levelling off and in some cases the unwinding of global value chains, also seem to be important causal factors, which means that trade cannot be expected to grow faster than the world economy in the foreseeable future (Haugh et al., 2016). Gains from fragmentation of production stem from deepening specialisation, economies of scale in production of each component and exploiting comparative advantage at the task level. However, fragmentation also has costs. Longer supply chains imply higher risk of disruption should one link in the chain falter, and transaction costs and co-ordination costs increase with the level of complexity. Some supply chains have matured and reached the point where the cost of further fragmentation outweighs the benefits. China has been a key player in global value chains and a source of trade in intermediate inputs. Over time, China has increased domestic value-added both through rising local capacity and foreign direct investment, and its imports of parts and components relative to total exports have steadily declined since the mid-1990s. Therefore, until new drivers of trade appear, it is probable that world trade will continue to track the pace of global GDP growth. In this light, trade in services, particularly digitally-delivered cross-border trade, holds the promise of becoming a new engine of international trade. If large economies like India were to embrace globalisation, this would also contribute significantly to world trade and growth.

Trade in goods and services complement each other

Services have always been traded. Indeed, international transport is as old as trade itself, and finance and insurance followed shortly after. Even today, transport and travel account for about half of world trade in commercial services, down from about 70% in 1980. The most dynamic part of services trade is business services, broadly defined to include telecommunications, computer and information services, finance and a host of other business services, including KIBS. The composition of world commercial services trade by broad categories is depicted in Figure 1.4. The United States is by far the largest exporter and also the largest importer, closely followed by China on the import side. Transport and travel dominate in most countries on both the export and import side. It is noted that China runs a significant deficit in transport and travel services, mirroring the shipment of its manufactured exports.

Figure 1.4. Commercial services trade by country, 2015

Note: COM = communications, including telecommunications, computer and information services. OBS = professional services, R&D services, management services and a number of technical services. Finance includes insurance. Other denotes all other commercial services. Data for Chile is for 2014. Data for Finland is reported only for aggregate total services.

The statistical data for Israel are supplied by and under the responsibility of the relevant Israeli authorities. The use of such data by the OECD is without prejudice to the status of the Golan Heights, East Jerusalem and Israeli settlements in the West Bank under the terms of international law.

Source: WTO; BPM6 classification.

A high degree of specialisation in business services is observed in Ireland (where communications services are the largest exports), Luxembourg (where financial services make up the bulk of exports), and India, the United Kingdom, Switzerland and the Netherlands (which all export a broad range of business services). India ranks first in exports of communications services, largely driven by computer services, while the United States is the leading exporter of other business services. With 68%, India has the largest share of business services in commercial services exports, followed by Brazil with 58%. The composition of imports is similar to that of exports, although business services are much more important in Japan’s imports than its exports. It is notable that the most specialised countries, Ireland and Luxembourg, also import more business services than average. Luxembourg engages in extensive intra-industry trade in financial services, while Ireland’s services trade patterns reflect its role as a hub for foreign investment in technology sectors. The largest exports are communications services while the largest imports are charges for the use of intellectual property. The United States and India demonstrate a substantial services trade surplus, while China accounts for the largest services trade deficit.

Figure 1.5 reports total growth in exports of communications and other business services compared to overall services exports over the decade to 2015. With the exceptions of Turkey and Indonesia, other business services and communications services gained ground in all countries depicted. China, closely followed by India, saw the fastest growth in services exports over the past decade, while Italy and South Africa lagged somewhat behind. Eight countries more than doubled their services exports during this period (from Costa Rica and rightwards in Figure 1.5). Korea, followed by China, Poland and Costa Rica experienced a very rapid growth in communications services exports, while Luxembourg and Costa Rica exhibit the fastest export growth in other business services. In contrast, Indonesia displays both a low level and slow growth in business services exports.

Figure 1.5. Export value, 2005-15 by country

Note: “Communication” covers telecommunications, computer services and information services. “Other business services” aggregates professional services, R&D services, management services and a number of technical services.

Source: WTO; BPM6 classification.

The composition of services trade has changed over the past few decades, as business services have become more prominent. Even so, goods and services trade flows still move in lockstep with each other. Services accounted for 22% of world trade in 2015 and have hovered around 20% for the entire period for which data exist. Furthermore, this pattern also holds for individual countries as illustrated in Figure 1.8, which plots services trade against goods trade from 1960 to the present. Only a few observations fall outside the big cloud of highly correlated goods and services trades.10 Thus, goods and services trade are still two sides of the same coin. As manufacturing production networks become more complex, supporting services become more complex too.

Figure 1.6. Trade in goods and services: Two sides of the same coin

Source: Estimated from WDI. The graphs plot total merchandise trade against total commercial services trade for all countries in the WDI for the period 1960 to 2015.

In addition to what can be gleaned from analysing aggregate trade data, an in-depth analysis of global value chains shows that about three quarters of trade in services consists of intermediate inputs for the production of goods and services (De Backer and Miroudot, 2013). This is a much higher share than for intermediate goods, which stands at about half. The important role of services in supporting other sectors’ exports is also reflected in the ratio of value-added to gross exports. For manufacturing, this ratio declined from about 0.65 to 0.45 from 1970 to 2010. By contrast, for services, the same ratio increased from about 1.3 to 1.6 over the same period. A fall in this ratio signifies that intermediate inputs from other sectors and from abroad account for a rising share of the gross export value. A ratio above one as observed for services seems counterintuitive. However, when measured in value-added terms, services embodied in exports of other sectors contribute to the numerator, while gross exports of services include services exported directly only. A ratio above one then means that more services are exported indirectly embodied in goods than are directly exported (Johnson and Noguera, 2016).

Externally-sourced services are important inputs in the fabrication and marketing of goods, but goods-producing firms also provide a range of services in-house and some of these are exported. Indeed across the countries included in the OECD/WTO Trade in Value Added (TiVA) database between 25% and 60% of employment in manufacturing firms is in services support functions (Miroudot and Cadestin, 2017). For example, multinational manufacturing firms often accompany affiliate production of goods abroad with headquarter services related to the design, production process management and quality control of their products. Manufacturers also combine shipment of goods with installation, maintenance and customer services. As illustrated in Figure 1.7, exports of architecture and engineering from manufacturing firms outpace those of specialised professional services firms in Belgium, Germany and Italy.11 In-depth analysis of Belgian export patterns at the firm level shows that firms that export both goods and services account for about 30% of services exporters and realise 36% of services exports sales in the sectors covered by the analysis.

Figure 1.7. Distribution of exports between services and non-services firms, by sector
Selected services and countries

Note: CO = construction services, CS = computer services and PSarc/eng = architecture and engineering.

Source: Rouzet et al. (2017).

Cross-border services trade is highly concentrated in three dimensions: Firms, markets and products

The digital economy has in principle opened new opportunities for services firms, including SMEs, to engage in international trade. Nevertheless, entering and staying in export markets is not for the faint-hearted. The absolute number of successful micro- and small enterprises that are “born global” (Box 1.2) is sufficiently large to inspire and to instigate various policy initiatives to encourage others to follow suit. Nevertheless, exporters are generally significantly larger and more productive than non-exporters, and more likely to be foreign-owned. Furthermore, most firms that export do so to only one foreign market, usually a neighbouring country or a country with which the home country has a free-trade agreement covering services.12

Figure 1.8 illustrates the high rate of concentration of exports once a firm has entered foreign markets. The share of exporting firms that export to only one foreign market ranges between 17% in Sweden and 43% in Germany, but their share of total export value is very small, amounting to between 1% and 6%. Conversely, the share of exporting firms that export to more than 25 destinations is only between 6% and 29%, but these exporters’ share of total exports ranges from 53% (Belgium) and 88% (Finland). However, even the big exporters are highly dependent on their top destination market. For example, in the sample of countries depicted in Figure 1.8, the firms that export to 20 destinations send between 40% and 80% of their export value to their number one destination, which in turn absorbs more than twice as much as the second market.13

Figure 1.8. Concentration of international activity by number of destinations

Source: Rouzet et al. (2017).

A similar pattern of concentration is found for the number of different services a firm exports. Between 36% (Finland) and 70% (United States) of exporting firms export only one product, accounting for between 2% (Finland) and 37% (Belgium) of total exports. Conversely, between 4% and 18% of exporting firms export five or more different services, and these account for the largest share of total exports (Rouzet et al., 2017).

Evidence from Germany shows that between 20% and 40% of all cross-border services exporters only export occasionally. The incidence of one-time exporters is highest in the construction sector and lowest for courier services. About two-fifths of all export relationships are terminated after one year in construction and about one-fifth in courier services. Courier services are also the sector exhibiting the longest duration of trade spells with two-fifths of all export relationships lasting for ten years or more. The sector has been boosted by cross-border e-commerce and just-in-time production networks that traverse international borders. By contrast, in the construction services sector only about a tenth of export relationships last for ten consecutive years or more. A relatively short duration of export relationships is also found in architecture and engineering services. A possible explanation is that exports of services related to the construction and building industry may be project-based and terminate with the completion of the project.

The firm-level evidence suggests that entering a foreign market is hard and costly, and costs increase more than proportionally with the number of destinations and products a firm sells abroad. When entering a new market, potential exporters might “test the waters” to assess the local conditions before deciding whether to invest in durable buyer-supplier relationships. In many cases, such experimentation fails and new entrants exit quickly. Only the most productive and innovative among first-time exporters manage to establish a foothold in the market and thereafter strengthen their presence and increase export sales.

Services that cannot be stored and consumed repeatedly are traded through face-to-face interaction between suppliers and customers, or they travel with the products that are transformed by the services, e.g. in transport. To bring such services under the realm of international trade agreements, the definition of services trade has been extended from cross-border trade to cover trade through consumption abroad, commercial presence abroad, and cross-border movement of people. The relative importance of these different modes of supply is not well understood, but a pilot study by Eurostat (2016) on extra-EU trade found that foreign affiliate sales (Mode 3) account for 69% of services trade, cross-border trade (Mode 1) 21%, consumption abroad (Mode 2) 6% and services supplied by natural persons (Mode 4) 4% in 2013.14 Brazil’s services exports exhibit a similar pattern.15 Clearly, trade through commercial presence dwarfs the other modes of supply.

Foreign affiliate sales dwarf the other modes of supply, but follow similar patterns to cross-border trade

Information on global foreign affiliate sales is not available, but data on the stock of FDI by sector exists for a number of OECD countries. It is not a perfect proxy for affiliate sales but nevertheless sheds light on the relative importance of services in multinational activities. Figure 1.9 shows that the services share of multinational enterprise activity is in line with the services share of GDP and employment. A closer look at the composition of services FDI assets by sector across 24 OECD countries reveals that financial services play a dominant role, on average accounting for more than half the stock of FDI held abroad and more than 40% of the stock of FDI from abroad.

Figure 1.9. Foreign direct investment stock, 2015

Source: OECD FDI statistics.

A study of affiliate sales at the firm level in selected countries reveals similar patterns as for cross-border trade. The countries included in the analyses are Finland, Germany, Italy, Japan, the United Kingdom, and the United States.16 Multinational services firms tend to be larger and more productive than local firms, including local firms that engage in cross-border exports. Foreign affiliate sales per product and export destination are on average more than ten times larger than cross-border exports.

Multinational services providers are few and far between relative to the number of cross-border services exporters. Yet their foreign affiliate sales account for the bulk of total services trade through all modes. In Germany, foreign affiliate sales account for as much as 80% of total services exports in most sectors. Like services exporters, multinational firms tend to depend heavily on their largest host market, which is typically a market close to home. European multinationals, for example, tend to establish mainly in other EU countries, although the United States is the most important host for UK multinationals measured by the number of affiliate firms. Conversely, the United Kingdom is the most important host market for US services multinationals measured by affiliate sales. China figures prominently as a host for German and Japanese multinationals, taking second place among the top destinations for both. A surprisingly small share of cross-border services exports are undertaken by multinational services firms to the same country and in the same sector, suggesting that multimodal services trade is not very common, with the exception of maritime and air transport.

The characteristic for which there are significant qualitative differences between multinational services firms and cross-border traders is the composition by product type of their trade activity. Some of the most digitised sectors, computer services, architecture and engineering tend to be traded through cross-border exports, while distribution services are largely traded through commercial presence. For the other services sectors there is a large variation across countries in which mode of supply dominates.

To summarise, services exporters are larger and more productive than non-exporters and multinational services exporters are larger still. Foreign affiliate sales are by a wide margin the most important mode of supply, followed by cross-border trade. Services exporters, whether engaging in cross-border trade or affiliate sales, tend to concentrate their exports in one or a few destination markets and a few products. With widespread migration of services activities to the digital economy, one would expect cross-border supply to have grown faster and brought more SMEs into international markets than actually happened. As we shall see in following chapters, regulatory barriers to trade and investment affect SMEs disproportionately. Furthermore, in the KIBS sectors cross-border trade for value services are highly complementary to deploying staff at the premises of the clients for short periods. Chapter 2 demonstrates that movement of people tends to be the most restricted mode of supply, which could stifle cross-border trade significantly. A more recent concern for services traders is limitations on the flow of data needed to embark on a digital transformation.17

Services in the digital economy

Will professional services be automated?

The rise of services in international markets is to a large extent driven by the ICT revolution (Box 1.1). Once the output of a service activity can be codified, it can in principle also be digitised and transmitted over electronic networks globally. Professional services are a case in point. They generate knowledge-capturing products that can be stored and thereafter consumed and accessed repeatedly. Examples are professional services automation (PSA) software, which is gaining prominence in architecture, accounting, legal services and management consulting. Computer-assisted design that feeds into computer-assisted manufacturing forming CAD/CAM systems has been commonly used in a host of manufacturing industries for decades. What is new in more recent years is that CAD can be separated from CAM in time, space and institutionally.

Box 1.1. Telecommunications and the digital economy

Telecommunications form the backbone and infrastructure on which the digital economy is founded. According to Cisco, a leading network company, global internet traffic has grown from 100 gigabytes per day in 1992 to more than 20 000 gigabytes per second in 2015. The devices used for internet traffic are changing rapidly. While computers used to dominate, machine-to-machine (M2M) devices are currently the fastest growing devices, followed by smartphones and connected TVs. Nevertheless, PCs still account for the largest traffic share (67% in 2015), followed by smartphones and connected TVs. M2M connections are rapidly gaining ground in both consumer and business markets. The most prominent uses in the consumer markets are connected home applications and connected healthcare, while the fastest growing business services are a host of location-based services such as supply chain and fleet management and not least video conferencing. Growth in demand for more sophisticated digital products, such as video, and secure and reliable payment services and design, to mention but a few, clearly depend on high-speed and high quality connections. As of 2015, consumer traffic accounts for more than 80% of internet and managed IP traffic. Figure 1.10 depicts the broadband, mobile and Wi-Fi average speed by region in 2015.

There are large regional differences, with Asia Pacific-led by Korea and Japan-having the highest average broadband speed and North America the highest mobile and Wi-Fi speed. Latin America, Middle East and Africa lag relatively far behind, although they are forecasted to experience faster improvement during the next three years. It should also be borne in mind that the lagging regions not only have slower connections, they also have far fewer connections. Key to higher speed is investment in infrastructure, which in turn is facilitated by an open and well-regulated telecommunications sector (Nordås and Rouzet, 2016).

Figure 1.10. Internet speed, 2015

Source: Cisco (2016).

Knowledge can be digitised, but still needs to be explained

In business-to-business transactions, PSAs are mainly used as a tool by professionals to streamline their services and draw on accumulated experience from earlier projects. The core of their activities is, however, still the personal interaction with clients. Typical situations where firms need professional support are during the course of product and process innovation, when adopting a new technology, preparing for entering a new market or complying with new regulations. Facing changing market conditions, the managers and staff of, for example, manufacturing firms may lack the capacity to identify the opportunities available to them and the means to reap the benefits on their own even when they have access to online information. The typical business model of professional and other knowledge-intensive business services is therefore a combination of automated basic services provided over digital platforms and collaboration with clients to help find solutions when clients face important strategic decisions. The codified and tacit elements of KIBS activities are highly complementary and usually bundled under a contract with the client.

This does not mean that KIBS are not being disrupted by the digital revolution. Routine tasks such as filing tax returns, accounting, and a host of human resource management and customer management functions are increasingly performed through the use of PSAs. These can be bought off the shelf or as cloud-based software-as-a-service. In addition, modularisation is chipping away into the core of business services. Digitisation of professional services can be understood as a learning process. New, cutting-edge knowledge tends to be tacit, but over time it becomes better understood and once fully understood can be codified and digitised. Machine learning speeds up this process by identifying underlying commonalities and patterns in the information professionals and their clients amass.

Digitisation makes basic professional services affordable to consumers and SMEs

There are different views on how deeply digitisation will impact on professional services. All agree that routine services such as accounting and tax filing are moving online and individuals as well as SMEs may use solely online platforms for these purposes. The most radical view is that the exclusive right to provide certain services by licensed professionals is under attack as consumers and businesses can perform these services themselves using digital platforms. Susskind and Susskind (2015) provide a number of examples of platforms becoming the dominant channel for accessing knowledge traditionally provided by professionals, including:

  • More disagreements between eBay traders are resolved through online dispute resolution each year than there are lawsuits filed in the entire US court system.

  • In 2014, 48 million people filed tax returns using online tax preparation software rather than a tax professional in the United States.

  • An online community designed and built a house for GBP 50 000 in London in 2014.

These examples could, however, also signify that online KIBS lower the threshold for both consumers and small businesses to use services that they otherwise could not afford. If so, online KIBS expand the customer base rather than undermining the business for licensed professional services. In line with that interpretation, others find that although the KIBS business model is being transformed by digitisation, the operations of KIBS are by and large unperturbed.18

The fact that it takes considerable skills and capacity on the user side to take full advantage of online business services may mitigate the impact of online services on traditional knowledge-intensive business services. In particular, the user must be able to formulate a problem in a concise manner and operationalise suggested solutions, which are important parts of the services package that KIBS suppliers offer. It is well documented in the literature that the take-up of online services is much smaller in SMEs than larger companies (OECD, 2017).

As will be demonstrated in more depth in following chapters, the potential for knowledge sharing and technology diffusion through international trade in professional services is far from being fully captured. Rather, markets are fragmented due to geographical, institutional, cultural and regulatory differences, and not least due to exclusive rights for certain professionals to provide a service in a specific jurisdiction. In addition, many countries prohibit multidisciplinary work across professions. As the borders between different professions are blurring, it may be time to re-think the relevant, century-old professional regulatory regimes.

Box 1.2. “Born global”: Medium, small and micro-enterprises (MSMEs) in the digital economy

The concept of “born global” was introduced already in the early 1990s. It refers most commonly to a firm that started foreign operations within three years of inception and that derives at least 25% of its turnover from outside the home market. Several studies have found that important driving forces for going global early are a small home market, scalability of the product or a niche strategy. ICT products fit this description well. There is little systematic information on the number and nature of firms born global, but three broad categories can be identified: online market places, digital product applications and production networks combining online and offline activities

Local Motors is an example of a production network, combining online and offline activities. It also illustrates the complementarity between goods and services in a modern economy. The company was established in the United States in 2007, and specialises in custom-made vehicles. According to its website, it is an international community of enthusiasts, designers, engineers, fabricators and experts. The company had about 200 employees in 2016 (up from 15 in 2010), and about 50 000 community members all over the world that collaborate on designs across a number of projects. Designs are developed into projects through online collaboration and the cars are built on demand in micro-factories. The company owns micro-factories in the United States and one in Germany. The flagship project so far is a self-driving electrical vehicle named Olli (photo), while the most recent project is the first 3D printed electrical vehicle in the world (


Flitto is an online market place as well as digital product application. Its business model is crowd-sourcing of translation services. The company was established in 2012 by three young Korean entrepreneurs who were selected for the SpringBoard incubator program in London. The Flitto app connects human translators with customers in real time and currently has six million users. Payment is based on points that can be bought or earned by providing translations. The app currently supports 17 languages, including Arabic, Chinese, English, French, German, Indonesian, Hindi, Japanese, Korean and Russian. Users can for instance take a photo of a hand-written note with their smartphone, upload it to the Flitto app, and in a few minutes a translator will have translated it into the desired language as shown in the picture. Flitto also offers a one-to-one professional translation service. The main office is in Seoul employing about 40 people and the company has a branch office in Beijing. Revenue derives from a fee on transactions made on its app and the company also sells its database of translations to internet firms that use them for machine learning.

Source: Flitto website ( and presentation by Simon Lee, the CEO of Flitto, at an OECD/APEC workshop, 28 August 2015 .

The on-demand economy and platforms

Platforms are flourishing in labour-intensive services sectors such as transport, distribution, logistics, courier services, handyman services and a host of personal services. Indeed, there are platforms for almost any service. There is no unified definition, but most agree that a platform is essentially a piece of software, an operating system or a database on which other smaller application programs can be designed to run. Platforms provide a low-threshold market entry opportunity for entrepreneurs, hobbyists and SMEs, they reduce transaction costs substantially by gathering large amounts of data that support reputation and trust building, and they usually integrate secure payment systems.

Platforms and applications often rely significantly on innovation by users. For example, users have played a very important role in the development and diffusion of financial services provided over mobile networks and applications. Furthermore, such services first took hold in developing counties and have spread from there to developed countries. This is an example of need being the mother of invention as unbanked people in developing countries got access to mobile telephones (van der Boor et al., 2014). Mobile banking is but one so-called “fin-tech” product. Others that have emerged mostly in developed countries are peer-to-peer lending platforms, electronic exchanges and even robo-advisors, which do market analytics for investors.

The actual and potential contribution of platforms in service sectors to the overall economy has been debated in the literature. Is their main role to improve capacity utilisation by putting to work idle capital such as homes and cars, and to provide flexible work opportunities for those that would otherwise not work? For example, Airbnb, which is a market place for hospitality services founded in 2008, grew very rapidly and has currently more than 2 million listings, more than any hotel chain (Luca, 2017). Or are they opening new opportunities for individual services suppliers to compete with companies and for start-ups to go global from their inception? The quick answer is that platforms can be both a boon for consumers and entrepreneurs as the examples from the KIBS sectors and Box 1.2 demonstrate, while disrupting traditional service suppliers.

The labour market impact of platforms is one of the great social concerns in the global digital economy. It is clear that the relative importance of non-standard work has increased in recent years in most OECD countries (OECD, 2016), although it is less clear how this relates to digital platforms and globalisation of services. Non-standard work is defined as self-employed (own-account), temporary full-time employees and part-time employees. Among these, self-employed are probably the most relevant for platforms.19

As indicated in Figure 1.11, the share of self-employment in total civilian employment ranged between about 6% in Luxembourg and the United States and 51% in Colombia in 2015. Other countries with more than 30% of the labour force in self-employment are Greece (35%), Turkey (33%) and Mexico (32%). Of the 34 mainly OECD countries included in the data, only six (the Czech Republic, Estonia, Finland, the Netherlands, the Slovak Republic and the United Kingdom) saw a rise in self-employment during the period 2000-15. Among these, the largest increase was observed in the Slovak Republic, the Netherlands and the United Kingdom. In the other 28 countries, the self-employment rate declined between 2000 and 2015.

Figure 1.11. Self-employment percentage of civilian labour force

Source: OECD.stat.

Self-employment includes many more activities than those facilitated by platforms, and as of yet, platforms seem to play a marginal role. Thus, recent estimates from the United States suggest that the share of the labour force working with labour platforms is between 0.4% and 0.9%. Labour services provided over platforms continue to grow rapidly, but not as fast as in the recent past (Farrell and Greig, 2016; Katz and Krueger, 2016).20 A common finding is that there is a high rate of churning in these activities. Less than half of the participants on the supply side stay with the platform for more than 12 months and about a sixth of participants in any month are newcomers. Finally, average monthly earnings on labour platforms declined between 2014 and 2016 (Farrell and Greig, 2016).

Platforms: Natural monopolies?

Platforms not only connect suppliers and customers, they are also businesses in their own right. The best-known digital platforms are relatively young and they have grown very rapidly to dominate the industry in which they operate. The strategic asset of platforms is information extracted from the participants. The more participants, the more information is gathered and the more valuable the platform is, creating a positive feed-back loop. More participants on one side of the multi-sided platforms attract more participants on the other side, inducing rapid growth after a critical mass is obtained. Conversely, if a platform fails to attract a critical mass within the first few years, it quickly disappears. Successful platforms can therefore dominate markets completely, and in some cases they exhibit characteristics of natural monopolies. They are typically global in scope and, depending on the market they operate in, they may connect suppliers and customers in a local market (e.g. for personal services) or in a global market (e.g. film, music, design, software-as-a-service).

The implications of dominant global platforms for competition policy are subject to debate. The commonly used tools for identifying significant market power, such as market share, pricing above average long-run costs or profit margins do not apply to multisided markets (e.g. Evans and Schmalensee, 2013; Evans, 2016). More rigorous methodology that takes into account all sides of the market on a case-by-case basis is needed to determine whether or not anti-trust remedies are needed.

To summarise this section, digitisation lies at the heart of both productivity gains in services and making services tradable across borders. Like technological revolutions in the past, digitisation brings huge potential gains to consumers and disrupts existing markets, business models and not least the labour market.

Concluding remarks

A rising share of services in employment and consumption is observed across the globe and is explained by a combination of technical progress, urbanisation and demand factors. Technology has reduced trade and transaction costs for both goods and services, thereby facilitating a deeper division of labour and more complex and services-intensive production networks. The close relationship between trade in goods and services is also underscored by the fact that a greater volume of exported services is embodied in products from other sectors than those exported directly. Moreover, three-quarters of services exported directly across borders are intermediate inputs suggesting that cross-border trade in services represent largely business-to-business (B2B) transactions. As a consequence, trade in goods and services follow each other in lockstep and should be seen as complementary activities in international markets. It follows that policy measures restricting trade in services impede trade in both services and goods.

Cross-border services trade accounts for only about 20% of total cross-border trade, and has remained fairly constant for decades. With the digital transformation of services one would expect a rising share of services traded directly to consumers through platforms, applications and other forms of electronic networks. So far, it appears that a global digital services economy remains elusive. There are many reasons for this. First, tradable tasks and modules are complementary to less tradable components. For example, knowledge can be codified, digitised and traded, but needs to be interpreted and sometimes explained by a natural person at the receiving end. Second, entering foreign markets can be costly and most firms, particularly SMEs, choose to focus on their home market. Third, cross-border services trade is still impeded by a number of regulatory obstacles, particularly in cases where a license is required to provide services and when occasional face-to-face interaction between client and services suppliers is needed.

The services share of foreign investment stocks on the other hand is closer to the services share of GDP. This does not mean that services trade through foreign affiliate sales is open and free. Rather, it suggests that the drivers and impediments to foreign investment may be of a similar magnitude for goods and services. There are, however, quite significant variations across sectors. For example, financial services play an outsized role in services trade through affiliate sales. Financial services support all economic activities, including trade and international commerce, and are subject to stringent regulation or outright prohibition of cross-border trade. In addition, financial services providers must assess the risks associated with clients and markets, which may be easier with a commercial presence in each market. Both regulation and the preferred business model may thus tilt financial services trade towards mode 3. The policy determinants of trade in services are explored in detail in the next chapter.


Aarikka-Stenroos, L. and E. Jaakkola (2012), “Value co-creation in knowledge intensive business services: A dyadic perspective on the joint problem solving process”, Industrial Marketing Management, Vol. 41, pp. 15-26.

Amiti, M. and S.-J. Wei (2009), “Services offshoring and productivity: Evidence from the US”, The World Economy, Vol. 32, pp. 203-220.

Benz, S., A. Khanna and H.K. Nordås (2017), “Services and performance of the Indian economy: Analysis and policy options”, OECD Trade Policy Papers, No. 196, OECD Publishing, Paris,

Borchert, I. and A. Mattoo (2010), “The crisis resilience of services trade”, The Services Industry Journal, Vol. 30, pp. 2115-2136.

Ciriaci, A., S. Montresor and D. Palma (2015), “Do KIBS make manufacturing more innovative? An empirical investigation of four European countries”, Technological Forecasting and Social Change, Vol. 95, pp. 35-151.

Cisco (2016), “The zettabyte era: Trends and analysis”,

De Backer, K. and S. Miroudot (2013), “Mapping global value chains”, OECD Trade Policy Papers, No. 159, OECD Publishing, Paris,

De Backer, K., C. Menon, I. Desnoyers-James and L. Moussiegt (2016), “Reshoring: Myth or reality?”, OECD Science, Technology and Industry Policy Papers, No. 27, OECD Publishing, Paris,

Evans, D.S. (2016), “Multisided platforms, dynamic competition, and the assessment of market power for internet-based firms”, Chicago Coase-Sandor Institute for Law and Economics Working Paper, No. 753,

Evans, D.S. and R. Schmalensee (2013), “The antitrust analysis of multi-sided platforms businesses”, NBER Working Paper no 18783,

Eurostat (2016), “Services trade statistics by mode of supply” Statistics Explained,

Farrell, D. and F. Greig (2016), “The online platform economy: Has growth peaked?”, JP Morgan Chase &Co. Institute, November,

Gebauer, H., E. Fleisch and T. Friedli (2005), “Overcoming the services paradox in manufacturing companies”, European Management Journal, Vol. 23, pp. 14-26.

Gordon, R.J. (2016), The rise and fall of American growth: The US standard of living since the Civil War, Princeton University Press, Princeton, New Jersey.

Haugh, D., A. Kopoin, E. Rusticelli, D. Turner and R. Dutu (2016), “Cardiac arrest or dizzy spell: Why is world trade so weak and what can policy do about it?”, OECD Economic Policy Papers, No. 18, OECD Publishing, Paris,

Johnson, R.C. and G. Noguera (2016), “A portrait of trade in value added over four decades”, NBER Working Paper 22974, World Trade Organization,

Katz, L. and A. Krueger (2016), “The rise and nature of alternative work arrangements in the United States, 1995-2015”, NBER Working Paper no 22667, September,

Lodefalk, M. and H.K. Nordås (2017), “Trading firms and trading costs in services: The case of Sweden”, Örebro University, School of Business Working Paper (forthcoming).

Lopez-Gonzalez, J., D. Flaig, M.-A. Jouanjean, J. Messent, M. Rentzhog and P. Walkenhorst (forthcoming) “Localising data in a globalised world”, OECD Trade Policy Papers, OECD Publishing, Paris.

Luca, M. (2017), “Designing online market places: Trust and reputation mechanisms”, Harvard Business School Working Paper, No. 17-017,

Miroudot, S. and C. Cadestin  (2017), “Services in global value chains: From inputs to value-creating activities”, OECD Trade Policy Papers, No. 197, OECD Publishing, Paris,

Nagengast, A.J. and R. Stehrer (2016), “The great collapse in value added trade”, Review of International Economics, Vol. 24, pp. 392-421.

Nordås, H. and D. Rouzet (2015), “The impact of services trade restrictiveness on trade flows: First estimates”, OECD Trade Policy Papers, No. 178, OECD Publishing, Paris,

OECD (2017), Key issues for digital transformation in the G20, Report prepared for a joint German Presidency/OECD Conference, Berlin, January,

OECD (2016), “New forms of work in the digital economy”, OECD Digital Economy Papers, No. 260, OECD Publishing, Paris,

Rodrik, D. (2016), “Premature deindustrialization”, Journal of Economic Growth, Vol. 21, pp. 1-33.

Rouzet, D., S. Benz and F. Spinelli (2017), “Trading firms and trading costs in services”, OECD Trade Policy Papers, forthcoming.

Rubalcaba, L. and H. Kox (eds.) (2007), Business services in European economic growth, Houndmills, Basingstoke, Palgrave MacMillan, Hampshire.

SNA (2008), United Nations System of National Accounts,

Susskind, R. and D. Susskind (2015), The future of the professions: How technology will transform the work of human experts, Oxford University Press, Oxford.

Van der Boor, P., P. Oliveira and F. Veloso (2014), “Users as innovators in developing countries: The global sources of innovation and diffusion in mobile banking services”, Research Policy, Vol. 43, pp. 1594-1607.

Wagner, J. (2012), “International trade and firm performance: A survey of empirical studies since 2006”, Review of World Economics, Vol. 148, pp. 235-267.


← 1. Source: World Development Indicators (WDI). The time series for middle income countries go back to 1960 when the share of services in GDP stood at 36.5%. Rodrik (2016) found that, between 1970 and 2013, the manufacturing share of GDP measured at constant prices remained stable both globally and in the United States, rose significantly in China and in the rest of Asia, and declined somewhat everywhere else.

← 2. For a fascinating chronicle of how technology, trade and the rise of modern services have transformed the life of ordinary Americans, see Gordon (2016). See also De Backer et al. (2016) for an assessment of labour market impacts of reshoring.

← 3. The productivity measure presented here is total factor productivity (TFP), which tabulates the change in output volume relative to changes in the use of inputs including capital, labour and intermediate inputs. If output volume increases faster than the volume of inputs, then TFP increases. The EU_KLEMS database contains updated TFP figures for only ten countries of which the three largest and one small economy have been selected. Data from the United States show similar patterns, with manufacturing productivity growth falling behind many services categories during the same period. The sector classification is, however, different and information is available at different levels of aggregation.

← 4. In France and the United Kingdom, employment in retail and wholesale trade is higher than in manufacturing.

← 5. See, for instance, Aarikka-Stenroos and Jaakkola (2012), Amiti and Wei (2009), Ciriaci et al. (2015) and Rubalcaba and Kox (2007).

← 6. A recent noteby Arthur D. Little reports that after-sales services accounted for 23% of revenue and 54% of profits in the car industry in 2015. after_sales_2015_01.pdf

← 7. In the business literature, this phenomenon is referred to as the servitisation paradox (Gebauer et al., 2005; Cenamour et al., 2016).

← 8. See

← 9. Some authors have labelled knowledge-intensive activities “quaternary services”, seeing them as comprising a qualitatively new sector splitting off from the tertiary sector.

← 10. The three observations below the cloud in the export chart represent Iceland in the 1960s. Iceland later developed strong services exports, particularly in air transport. The observations to the left of the cloud represent Tuvalu, whose exports are largely concentrated in the tourism sector.

← 11. Information on international activities by firms is not readily available. The statistical agencies of the countries included in the analysis kindly allowed us to analyse anonymised data under strict confidentiality conditions. The following disclaimers apply: This work contains statistical data from ONS which is Crown Copyright. The use of the ONS statistical data in this work does not imply the endorsement of theONS in relation to the interpretation or analysis of the statistical data. This work uses research datasets which may not exactly reproduce National Statistics aggregates. The statistical analysis of firm-level US data on multinational companies and services exporters was conducted at the Bureau of Economic Analysis, US Department of Commerce under arrangements that maintain legal confidentiality requirements. The views expressed are those of the authors and do not reflect official positions of the Department of Commerce.

← 12. See, for instance, Wagner (2012) for a recent review of the descriptive literature on patterns of foreign trade analysing firm-level data.

← 13. The distribution of exports on destination markets appears to follow a Pareto distribution.

← 14. See Eurostat (2016) by_modes_of_supply&oldid=320612.

← 15. See Rouzet et al. (2017), where commercial presence is estimated at 69% of total, cross-border trade 23% and consumption abroad 8%. Trade through movement of people contributes only marginally.

← 16. For an in-depth analysis, see Rouzet et al. (2017).

← 17. For a typology of such measures, along with a discussion of the use of data by business and possible impacts of measures, see Lopez-Gonzalez et al.

← 18. See World Economic Forum White Paper (2017).

← 19. There have been some disputes over the extent to which labour platforms such as Uber should be considered equivalent to employment. The criteria appear to be to what extent the platform controls the conduct, price, work hours, and work conditions of the participants. A recent court ruling in the United Kingdom, for example, found that Uber drivers must be considered employees. The ruling has been appealed.

← 20. Labour platforms are defined as platforms that connect customers with freelance or contingent workers who perform discrete tasks or projects. Growth in labour platform participation year-on-year was 102% in July 2016, down from almost 500% in July 2014.