Universidad EAN Facultad de Administración, Finanzas y Ciencias Económicas

Incorporation of Services into Natural Resource-Intensive Goods: description and measurament of the phenomenon Marcel Vaillant Enrique Gilles

Serie Documentos de Trabajo FAFCE No. 2016-004 Septiembre de 2016

INCORPORATION OF SERVICES IN NATURAL RESOURCE-

INTENSIVE GOODS: DESCRIPTION AND MEASUREMENT OF

THE PHENOMENON1

Marcel Vaillant2

Enrique Gilles3 August, 2016

1 We thank Mina Mashayekhi (UNCTAD) and Nanno Mulder (ECLAC) for their useful comments. Any remaining errors are responsibility of the authors. 2 Professor in International Trade, Department of Economics, Social Science Faculty, University of the Republic, Uruguay. E-mail: marcel@decon.edu.uy. 3 Professor of Economics. Department of Economics. Universidad EAN, Bogotá, Colombia. E-mail: eegilles@universidadean.edu.co.

Index of sections

1. INTRODUCTION ..................................................................................................................... 3

2. BRIEF REVIEW OF THE LITERATURE............................................................................... 5

2.1 Fragmentation of economic activities ................................................................................. 5 2.2 Servicification ..................................................................................................................... 6 2.3 Trade policy impacts ........................................................................................................... 9

3. PRODUCTS, COUNTRIES AND SPECIALIZED PRODUCTION SERVICES ................. 11

3.1 Identification of products and countries ............................................................................ 11 3.2 Specialized production services in NRIG ......................................................................... 13

4. ECONOMIC INFRASTRUCTURE SERVICES AND NRIG COMPETITIVINESS ........... 18

4.1 Methodological approach .................................................................................................. 18 4.2 Results ............................................................................................................................... 20 4.2.1 Domestic Value added contents of NRIG exports in EIS .......................................... 20 4.2.2 EIS productivity and prices competitiveness of NRIG .............................................. 22

5. CONCLUSIONS ..................................................................................................................... 25

REFERENCES ............................................................................................................................ 28 APPENDIX A- METHODOLOGICAL ..................................................................................... 30 APPENDIX B- STATISTICAL .................................................................................................. 34

INCORPORATION OF SERVICES IN NATURAL RESOURCE-

INTENSIVE GOODS: DESCRIPTION AND MEASUREMENT OF

THE PHENOMENON

Marcel Vaillant

Enrique Gilles August 2016

1. INTRODUCTION

The trade pattern specialization of many Latin American countries (mainly in South America) shows why these countries play a key role in the world market of a basket of goods, which are intensive in the use of natural resources (in particular mining and food). This is not new, it is rather a permanent characteristic of the international insertion of this region in the world. The long tradition of these sectors also implies a huge accumulation and learning experience on how to intensify in a sustainable way the exploitation of this type of resources.

In particular, the last stage of development, characterised by an important process of technical change in natural resources exploitation, has been very relevant. The market size provides a sufficient scale for incorporating service-related activities (specialised knowledge business services – KIBS) and through this mechanism it creates new types of trade specialisation patterns.

The servicification process can be widely defined as a business strategy that entails a shift from pure manufacturing to the production of goods and services. The term can be traced back to Vandermerwe & Rada (1988). One of the consequences of this phenomenon is for national systems of economics statistics: if manufacturing firms increasingly become services providers, then figures potentially involve errors and omissions. At the level of firms, many of them have been undertaking such a path, mainly as a response to increased competition in their original product markets. This is the case, for example, of IBM which turned from being a manufacturer of computers to a supplier of a whole range of system business solutions (Ahamed et al., 2013), or the case of Rolls Royce that changed from being an airplane engine supplier to a supplier of “power-by-the-hour”, i.e., their customers firm a contract through which Rolls Royce ensures engines functioning all

along the year by supplying the proper maintenance. In Latin America we also find evidence of servicification where, by incorporating specialized services, some traditional natural resources may upgrade in value chains. In particular, we can mention some examples of KIBS around seed innovation in Argentina and Brazil, the salmon industry in Chile or cattle traceability in Uruguay.

The incorporation of services into commodities (“servicification”) is a complementary approach to their industrial processing, e.g. the more traditional use of logistics, transportation, communications, etc. in the production of goods. Both strategies yield simultaneous benefits in terms of employment creation, technical progress, product diversification and quality improvement. However, policy recommendations may be different for each of them. Indeed, while traditional services associated to the production of goods (infrastructure, logistics, communications, insurances, etc.) may have a role on job creation and overall economic activity, servicification may also define new paths of sustainable growth in a highly integrated world. As mentioned above, services also play a key role in the technological upgrading of value chains created in or around natural resources.

How do services can add value to the production and exports of natural resources intensive goods? Incorporating services into the production of goods allows for both the differentiation and consolidation in accessing new markets, as well as to generate an increment in the domestic value added, with higher contents of skilled labour. This is a productive strategy that can achieve similar results in terms of creating greater domestic employment compared to conventional alternatives that focus on increasing production levels and industrial processing. Increasing domestic value added through services incorporation is particularly relevant for natural resources-intensive goods (NRIG), as these products dominate the export baskets of Latin American countries, especially those in South America.

The main objective of the paper is to link the servification literature in the manufacture sector to the NRIG case.

The document is organized in this introduction and four more sections. In section 2 we provide a brief review of the literature, considering topics such as the transformation of the global pattern of specialization, the servicification concept and the impact in the trade policy agenda. Section 3 identifies a set of products and countries for which we analyse specialized production services (innovation, upgrading, etc). Section 4 develops a sectoral

approach using input-output tables (IOTs) to measure the relevance economic infrastructure services (EIS) have in the competiveness of NRIG. Finally, section 5 concludes with the main remarks of the paper.

2. BRIEF REVIEW OF THE LITERATURE

2.1 Fragmentation of economic activities

The most recent phase of the globalization process has led to a deep transformation in the international economy. One of the main characteristics is the permanent increase in both the intensive and extensive margins over the tradability of economic activities (goods and services). One of the sources behind this phenomenon is the acceleration of technical change in information and communication technologies (ICTs), together with a general reduction in trade costs (transport costs and other distance-related costs, plus trade policy variables), facilitating the fragmentation of production at a world scale with an increase in trade specialization.

The different stages of productive processes are now spread over different national jurisdictions. A new international division of labor has been emerged, that new international trade theory names as trade in tasks (Grossman and Rossi-Hansberg, 2006). This process started in the modern manufacturing sector (transport equipment, electronics, optic machines, chemical and pharmaceutics) but is now extended to other economic activities (in particular services). Trade in intermediate goods and business services was traditionally relevant in international trade (Sturgeon and Memedovic, 2010). The channel of circulation was import of intermediate activity to the production of final goods to the domestic market (I2P in Badlwin (2013) terminology). However, a new phenomenon grew significantly over the last three decades: imports of intermediate to export to foreign markets (I2E). So, trade in intermediate inputs is playing an increasing role in explaining aggregate trade flows (Hummels and Klenow, 2005). The same phenomenon happens with services oriented to satisfy demand of inputs in the production process (production services). Trade in intermediates is also an important channel through which technology is transferred from develop to developing countries. Amity and Konings (2007) show that imports of intermediates and capital goods may improve firm productivity.

The process of production fragmentation has initially started in the automotive industry, where the production of inputs in specialized and independent units of production was realized outside the car factory (outsourcing). Then this dynamics expanded to the outsourcing of services activities. The outsourcing inside a domestic jurisdiction is then expanded to the international level through international trade (offshoring). The reduction in international trade costs and coordination costs are main drivers of this process. Miroudot et al. (2009) provide a characterization of the internationalization of intermediates supply taking into account two dimensions: activity location (domestic vs. foreign country) and ownership (inside or outside the firm). Overall, the intense technical change in telecommunications and information technology are one of the main drivers of this phenomenon.

2.2 Servicification

Services are at the core of these new fragmentation phenomena as they have incremented their participation in world trade. Nowadays trade in commercial services –only considering current account Balance of Payments data- represents more than 20% of world trade, however if trade is measured in value added this figure achieves half of world trade (Escaith, 2008).

The conventional theoretical approach associates services to low rates of innovation and a low productivity path. In particular, this applies to services that satisfy final demand, which also gradually increases its weight in total expenditure. Examples of these services are typically haircuts, automotive repairs, travel and tourism, or personal services such as day care, laundry or the like. This sustains the prediction of stagnation in total productivity (“Baumol disease” and Balassa-Samuelson effect). However, the changes we are here describing challenge this conclusion and give a very different picture (Hoekman and Francois, 2010). First, services associated with production as intermediate inputs increase its productivity and so the set of the others activities which they are related to by intersectoral linkages. Second, many services associated with final demand changed positively its pattern of innovation and productivity, influenced by the same fundamentals, i.e. the ICT revolution.

The Kaldorian tradition claims that growth is related to the size of the manufacturing sector because it is associated with its positive externality over technological change and the productivity path (through different explicit and implicit mechanisms). However, there is evidence that things are changing. Di Meglio, Gallego, Maroto and Savona (2015) highlight three elements:

i.

Evidence of “premature deindustrialization” of many developing economies4 in a context of good growth performance.

ii.

The tradability of services has progressively increased and value chains of production have globally expanded.

iii.

High productivity gains are being experienced by certain services branches. The authors complement the idea with the following statement: “Facing the old ideas with the new facts, our research hypothesis regards whether some specific service sectors may be a source of economic growth in developing countries. Potentially, those services embodying knowledge and technology or with strong inter-industry linkages or highly tradable may serve as a mean to catching-up with advanced economies by complementing/replacing manufacturing as an additional/new engine of growth in developing countries”.

It is important to differentiate between services associated with final consumption and services as intermediate inputs. The trade in tasks phenomenon is mainly associated with the latter, which are those demanded as inputs by other production processes. The disentanglement process to be feasible requires an intensification in the use of services sectors mainly in the form of production specialized services. The fragmentation of production needs standardization and imply an intensive use of information technology which are associated with the development of production specialized services The result is an increase in the intensity in the use of specialized production services that participates as an intermediate input in many different activities and types of firms. The empirical evidence goes from small firms to large multinationals, and from basic and food industries to the engineering industry. Three main factors explain this outcome in manufacturer products (Lodefalk, 2015):

4“The share of manufacturing in employment and value added is shrinking at levels of income per capita that are much lower of those at which the advanced nations began to de-industrialize” (D&G&M&S,

2015).

i.

Services is a relevant factor to the increment in productivity, in particular through the use of services in logistics, management or engineering that save time, materials and improve coordination (Nordas, 2010 quoted by Lodefalk,2015). Also these services are more intense in the use of skilled labour, which means the ability to absorb technical change is higher than in other sectors, with positive effects in terms of productivity.

ii.

Adding services to goods can also be a profitable strategy: firms may differentiate their supply (augmenting their value) by bundling services with products, or offer them in connection to their sales. Doing so, it will be easier for them to deal with the increasing international regulation on environmental and social issues. Many different types of barriers to trade require diverse types of certification that are associated with specialized services. With this strategy the firms fulfil the norm and obtain a certification that is crucial to jump the trade barrier. iii.

Manufacturing firms need services to establish, join and manage international production networks and value chains, including headquarter or business partner services such as data processing and R&D services (Kelle, 2013 quoted by Lodefalk, 2015). All of them are associated with the coordination cost dimension related to the process of fragmentation of production in many different stages (Grossman and Rossi-Hansberg, 2012).

At the level of the firm the process of servicification could have different types of origins. Many times innovative manufacturing firms develop the process of some specialized service activity that will eventually become another source of income for the firm (maybe the main one). It could also happen that a firm outside the manufacturing sector specializes on providing a particular specialized service.

However, the capacity of penetration of new technologies in all kind of activities is one of the main drivers of the transformation of production functions all over the world. This means servicification may be possible in many different activities and not only in dynamic manufacturing sectors5. In particular, this paper focuses on the identification of the phenomenon in the case of natural resources-intensive goods (NRIG). This last phenomenon is really relevant for South American countries. The fundamentals of the

5 The literature does not give too much attention to those others cases. For other examples in maturate industries (textile and apparel, footwear, food industry, etc.) see Cattaneo, Gereffi yStaritz (2010).

servicification process are the same of the just reviewed for the case of manufacturing products, but with different highlights depending of the type of sector. Arbache (2015) proposes a classification of services in two groups based on their role as inputs in the manufacturer sector:

i.

Specialized production services. It includes value and innovation services, and refers to functions that contribute to adding value, differentiating and customizing products and, therefore, raising their market price and increasing labour and global productivity and the return on capital6.

ii.

Cost services. It refers to activities that affect production costs through the use of economic infrastructure services (logistics and transportation, general infrastructure services, financial services, etc).

We apply the typology to the case of NRIGs. For the first type of services, in section 3, we present a summary of a previous paper (Vaillant, 2015) and for cost services we develop and apply a methodology of value added services contents in commodities exports in a selected set of South American countries (section 4).

2.3 Trade policy impacts

In coherence with the globalization of service production, it is possible to observe a process of expanding the scope of trade rules and topics in different types of trade agreement covering services activities7. Typically, services are a non-tradable and so regulations are established on a domestic perspective, i.e. covering national jurisdictions. When trade increasingly involves services activities the demand for new international rules and the treatment of related topics also increase.

All these phenomena have an impact on the international trade policy agenda, which is usually dominated by market-access issues. How much access in my own market should be provided in order to get better access conditions in the neighbors’ market? The

6Usually composed of services that require relatively high levels of human capital (R&D, design, engineering and architecture projects, consulting services, software, specialized technical services, highend IT services, branding, marketing, trading, among others).

7There are 120 preferential trade agreements in services notified to the WTO and actually in force. Almost all of them covers also goods and complementary matters, which shows through international agreements and disciplines the linkages between them.

fragmentation of international economic activity has strengthened the link between protection of one’s own market and foreign market access conditions. Baldwin and Nicoud (2008) have pointed out that this is good news from the perspective of expected results related to the endogenous mechanisms of trade liberalization. The more it is perceived that reducing own barriers will also reduce the neighbor’s barriers, the more open unilateral domestic trade policy will be. This mechanism seems to have started gradually acting in the trade of goods and particularly in certain manufacturing sectors that tend to converge to a trade of “zero tariff for zero tariff”. However, a negotiating effort is required to get improvements in some markets characterized by deeper distortions (certain sectors of agriculture and manufacturing, services, complementary matters).

A by-product of the globalization process is that the range of economic activities in the international economy expands the subjects that require consideration in trade agreements. The adaptation speed in the multilateral arena is structurally low. Countries are less willing to establish rules on the basis of Most Favored Nation (MFN) than within the framework of Preferential Trade Agreements (PTAs). Hence the demands to expand and deepen negotiations in new topics have been channeled through the phenomenon of the proliferation of PTAs. The content of commitments and topics in the international trade negotiations agenda between national jurisdictions has widened: from the trade of goods to the trade of services, as well as to the mobility of some production factors. At the same time, the field where compromises are achieved has exponentially multiplied: bilateral agreements, plurilateral agreements, agreements between groups of countries, extension of agreements.

For a diagnosis of what is happening it would be necessary to build a large (and complex to build) matrix of information that combine the fields of commitment (columns of the matrix) with its themes or contents (rows). The biggest threat that looms over the international trade system is the consistency, applicability and use of the established set of rules. In terms of the previously mentioned matrix, it is possible to verify that what is committed in a certain “line” is different and often difficult to translate in what is committed in another, not to mention the case of rules being frankly inconsistent. It is necessary to find issues that enable a systematic approach to this problem, as to apply a methodology that is plausible of being applied and then replicated to other themes. This is one of the biggest challenges that the global system of international rules needs to deal

with: how to incorporate an extended set of production activities (goods and services) in a coherent framework.

Latin American countries have a very different perspective related to the new trade agenda. In the Pacific side almost all countries sign trade agreements that include rules and disciplines in services and complementary matters (investment, government procurement, property rights, competition policy, among others). In countries of the Atlantic side the new trade agenda is practically out of the coverage of trade agreements. Considering the new specialization phenomena that we contribute to identify in this paper, in the near future the role of the services sector will be seen as an essential topic of the trade agreements negotiations agenda. Given the heterogeneous nature of this sector, which as we mentioned before includes final demand services but also specialized production services, domestic trade policies should take a differentiated approach. Besides trade policies, these new developments call for renovated public policies for the development of specialized service providers that add value to natural resources, or increase its productivity capacity through innovation activities in a sustainable way. Traditionally, public policies for both the service sectors and the NRIGs are designed without any link, and its implementation is frequently in charge of different government agencies. The new phenomena imply a challenge to associate and to integrate different sectors in policies and regulations, new institutions, trade agreements, specific incentives, and specialized education at the technological and university levels.

3. PRODUCTS, COUNTRIES AND SPECIALIZED PRODUCTION

SERVICES

3.1 Identification of products and countries

The specialization pattern of South American countries is concentrated in NRIG (Vaillant, 2015), and the region’s insertion in the international economy is limited to a small set of goods: only 52 products account for 70% of total exports8. The region has actually the highest rate of concentration of the first decile in the average period 2009-

8 COMTRADE, Harmonized System at 6 digits (second edition).

2012 compared to other regions of the global economy. Almost all goods in the first decile are NRIGs, the share of manufacture products is very low. Moreover, the set of products with revealed comparative advantage is also reduced: less than 500 products, which is much less than the world average of 800 products (Vaillant, 2015). In conclusion South American countries are economies with very low level of diversification in the exports of goods. As we shall see later this is both a problem and an opportunity. Table 1 Selection of products and South American countries in dynamic worlds markets (Millions of USD and %) HS2 Name of the product Exports 2009-2012 South America (millions USD) Market share (%) South American countries with highest index of RCA

a) Food and raw materials

20721

Chicken meat frozen

2548

66

Brazil; Argentina

150710

Soya crude oil

6479

65

Argentina; Paraguay

230400

Soya oilcake

15872

62

Argentina; Paraguay

170111

Cane sugar

10087

51

Brazil

120100

Soya beans

21017

47

Paraguay, Brazil

90111

Coffee

9123

44

Colombia

80300

Bananas

3961

36

Ecuador

20230

Bovine meat boneless

5033

35

Uruguay; Paraguay

240120

Tobacco

3173

32

Argentina; Brazil

80610

Fresh graves

2099

30

Chile

220710

Ethyl alcohol

1778

29

Brazil

100590

Maize

6887

25

Argentina

170199

Other sugar

3639

23

Brazil

b) Minerals

720293

Ferroniobium

1483

75

Brazil

261610

Silver

1593

57

Bolivia, Peru

260300

Copper

23086

48

Chile, Peru

740311

Copper cathodes

24656

38

Chile

260112

Agglomerated iron

7890

33

Brazil

260700

Lead

2066

29

Peru

260111

Non agglomerated iron

26484

28

Brazil

720712

Iron and steel

3020

19

Brazil Source: see Vaillant, 2015.

The literature mentions many problems related to being specialized in a reduced bundle of commodities: risk of exposures of terms of trade volatility; reduced inter linkages with other sectors of the economy; low productivity path, among other elements. However, the opportunity is associated with the scale and the relevance of the region in these markets

combined with the evidence of intense technological transformation in the goods intensive in the use of natural resources. As we are going to illustrate later, all this transformation follows the path previously characterized as the servicification process, despite the level of servicification does not get to the levels seen in the developed world. There is a main methodological problem to study this phenomenon. It is not possible to obtain consistent datasets of services embodied in commodity production. In fact, considering the characteristics of the data it is not possible to identify independent sector of services oriented to the NRIG. Usually the activities of this type of services are accounted together with the primary sector.

Our methodological approach to study the phenomena must start with an identification of the set of NRIG. Which are the products and countries that have some characteristics that make possible to conjecture that the servicification process is present or could be developed? Vaillant (2015) defines a set of indicators that must be taken into account: the revealed comparative advantage (RCA) index; the size of exports market and dynamic evolution both related with own and international supply. In Table 1 the chosen set of products are presented. All selected products fulfil the condition to increase participation in world trade and South American exports in the reference period (Vaillant, 2015). A clear regional pattern emerges from the information of Table 1. The Atlantic countries of South America are specialized in a basket of products mainly composed by food commodities, whereas in the Pacific–side countries minerals prevail. Brazil shares with the latter the fact of being an important and specialized exporter of iron products.

3.2 Specialized production services in NRIG

The focus is on the presence and development of specialized production service providers for natural resource sectors linked to exports of commodities. In this category we also include technology services that are not necessarily associated to the incorporation of value but to the innovation in the technology of production of the natural resources, which makes them important in the global sustainable productivity process. All these types of services are commonly knowledge intensive business services (KIBS) that were first defined in the servicification of manufactures literature.

For the analysis of specialized production services (innovation and value services) we propose to use secondary information to systematize the data and to know the types of relevant services associated with each particular product (see Vaillant, 2015). The objective is to construct a matrix with the products in the rows and the characteristics of the specialized production services in the columns. This methodology was applied to a set of seven cases (four in agriculture, two in the fish sector, one in minerals) that are available as they are recently presented in an academic Congress, which took place in Montevideo9. This set of cases is not a representative sample of all the commodities of the region. However, almost all the revised cases are in the set of products we identify in subsection 3.1 based on the analysis of trade data. The cases are useful to illustrate the type of services and the main mechanisms associated to their development. The results are presented in Table 2. In the four first columns we identify the country, the sector, the product and the main definition of the particular phenomena under study. The cases show the incorporation of specialised production services (mainly in the class of KIBS) to upgrade the NRIG and/or to change its sustainable global productivity. This topic focused into how the incorporation of specialized services in the value chain of natural resources can improve the intrinsic (product) and extrinsic characteristics of commodities and how these can increase their differentiation potential in the world market. Two types of specialized production services are identified. First different types of international and domestic regulation triggered new services that are ultimate used to increase value in each of this type of services. For example in the meat and in the fish sector a set of environmental and technical quality standards and certifications, branding, and marketing are observed. All of them are associate with the development of specialized productions services.

The second alternative is to focus on the role of innovation in natural resource sectors and its global productivity performance. This is the case presented in cereals, oleaginous, and agriculture seeds. In the nineties there was a major technological change in their production that was associated to a large diversification process in different stages of production. This, in turn, was associated to the development of different specialized production services. In this new framework, the traditional agriculture producer was

9 IV REDLAS CONFERENCE, SERVICES, NATURAL RESOURCES, TRADE POLICY and TRADE

PATTERNS IN LATIN AMERICA 27-28 May 2015, Montevideo, Uruguay Local organisers: dECON, FCS, UdelaR, ORT University International: ECLAC, IADB, UNCTAD, In cooperation with Zonamerica and Smart Services Uruguay XXI.

replaced by an agricultural firm, which coordinated production through a set of contracts with different type of firms (mainly services production firms). The hard core of the change are sophisticated biotechnological firms who are responsible for the production and improvement of new seeds which embodied KIBS.

In the case of copper there is a similar phenomenon, but in this case it is the pressure over natural resources the factor that provides the incentives for the development of different types of related services. In particular, the need to increase efficiency in production and to solve environmental problems associated with over production are among the main causes of the servicification process.

In both the cereals and copper cases, many of these specialized production services have matured in the domestic market, and they could achieve a level of competitiveness that permits to make countries direct exporters. However, until now this phenomenon is still in process: the highest level of internalization of KIBS in NRIG is just medium and oriented to regional markets (agriculture and minerals).

The starting point of the NRIG servicification process is the huge scale of those regional markets in an international comparison. In fact, more than a half of the global supply comes from South American countries (Vaillant, 2015). Traditionally this specialization is considered an advantage for the industrial transformation; however the specialization pattern path shows that servicification could also be an alternative to: i.

increase value added with sustainable international insertion that could fulfil the dynamic requirements of the world markets; ii.

change the productivity performance of the set of commodities exported by regional countries facilitating the process of technological transformation. Summarizing, the large scale of regional markets in NRIG, the intensification of production combined with the acceleration of technical change together with the increasing requirements in standardization and certification (both at the domestic and international levels) are necessary elements to the environment where KIBS are developed.

Gonzalez, Meliciani y Savona (2015) sustain that this type of pattern reflects the combined hypothesis of Hirschman and Linder. Sectors with RCA generate domestic interlinkages, which imply an incentive to develop new economic activities. These

markets are huge at international level and they can create a representative demand in particular for the regional countries. Based in this critical mass, new competitive services activities could be developed which have the potential to be internationalized. Trade and industrial policy tools to develop this type of services need to know the nature of the process that generate this new comparative advantages, and requires a deep understanding of the process of internationalization of this type of specialized product services.

Table 2 KIBS and NRIG in Latin America, seven selected cases Country Sectors Products Definition Origin servicification Type of services Internationalization Argentina a) Agriculture Cereals (trigon,maize, rice) and oleaginous (soya,sunflower) New technological package: diversification in different stages of production Increment use of specialized production service, agriculture firm as coordinator actor Innovation services Medium-Low Argentina/Brasil b) Agriculture Agriculture seeds Knowledge intensive products with embodied

KIBS

Adaptation to particulars environments, new value to agriculture genetics by local firms Innovation services Medium Uruguay c) Agriculture Bovine Meat Certification of quality (traceability) International regulation and certification process Value services Low Uruguay d) Agriculture Bovine Meat Traceability and innovation KIBS Available new information trigger new specialized services Innovation services Low México e) Fish sector Japanese Osteon Red “abulón” New technological services to control production Research and Technological regional centers in the fish sector Value services and innovation services Low Chile f) Fish sector Salmon KIBS in reproduction, sanity and certification Increase in pressure over natural resources and sanitary problems Value services and innovation services Medium-Low Chile g) Minerals Cooper KIBS and specific technological solutions Maturation growth process demand specific technological solutions Innovation services Medium

a) Anllo, Bizang y Katz, 2015; b) Marin, Sturbrin y da Silva, 2015; c) Labraga y Ferreira, 2015; d) Rius, 2015; e) Celaya, Alvarez y Hualde, 2015; f) Zanlungo y Katz, 2015; g) Meller

y Gana, 2015.

Source: authors’ calculations.

4. ECONOMIC INFRASTRUCTURE SERVICES AND NRIG

COMPETITIVINESS

4.1 Methodological approach

In this section we link economic infrastructure services (EIS) to the set of commodities exported by South American countries. We will address the role of services such as communications, finance, transport, distribution and logistics in the exports of natural resources. One way to measure their contribution is through the value added of those services embodied in the exports of NRIGs. At the same time we will also look at the impact of the Service sector productivity and performance on the exports of natural resource intensive products. This type of services is included in the second category (cost services) defined by Arbache (2015) and may be studied using conventional Input Output techniques.

Within that framework a first approach to assess on the role of services in NRIG can be conducted by analysing the intensities in production. This allows identifying industries which make a more intense use of services as intermediate inputs. A second approach to tackle the same objective is to analyse the value added content of services included in the exports of NRIGs.

It is also possible to study the role of Services in the competitiveness of NRIG. To that end, we develop the Leontief price model as proposed in Gilles and Vallecilla (2015), which main result is the change in sectorial prices as a result of changes in the value added (VA) of a given sector. We will use this framework to simulate the effects of an increase in productivity in the Services sector: such a process would translate into a decrease in value added. Indeed, as a sector becomes more productive we should observe a decrease in its share of VA in gross output. The model has been used in several environments to analyze different objectives. For example, Bazzazan et al. (2003) use it in the context of energy assessment policy; Davar (2006) discusses price discrimination issues with the model, while Folloni et al. (1993) evaluate different price formation hypothesis. We perform the above three analyses for six South American countries: Argentina. Brazil, Chile, Colombia, Peru and Uruguay. In order to have a unified framework we use the input-output tables provided by GTAP, which correspond to 2007 and are expressed in

millions of US dollars. The full description and details of the methodologies is found in the Appendix A.

We analize the commodities shown in Table 1. In order to determine the sector they belong to, we use correspondence tables between products in the HS and the GTAP sectors10. Results are presented in table 3.

Table 3 Products and sectors HS

PRODUCTS

COUNTRIES

SECTORS GTAP

a) Foods and raw materials

20721

Chicken meat frozen Brazil; Argentina

20

150710

Soya crude oil Argentina

21

230400

Soya oilcake Argentina

21

170111

Cane sugar Brazil

24

120100

Soya beans Brazil

5

90111

Coffee Colombia

8

20230

Bovine meat boneless Uruguay

19

240120

Tobacco Argentina; Brazil

8

80610

Fresh graves Chile

4

220710

Ethyl alcohol Brazil

26

100590

Maize Argentina

3

170199

Other sugar Brazil

24

b) Minerals

720293

Ferroniobium Brazil

35

261610

Silver Peru

18

260300

Copper Chile, Peru

18

740311

Copper cathodes Chile

36

260112

Agglomerated iron Brazil

18

260700

Lead Peru

18

260111

Non agglomerated iron Brazil

18

720712

Iron and steel Brazil

35

Source: authors’ calculations using table 1 and correlation with GTAP sectors. Services are going to be aggregated in six main categories (big sectors), which include infrastructure economics services (transport, trade, financial, public services), plus construction and other business services (see table B2).

10See https://www.gtap.agecon.purdue.edu/resources/res_display.asp?RecordID=1916.

4.2 Results

4.2.1 Domestic Value added contents of NRIG exports in EIS

The results on direct requirements in the use of economic infrastructures services (EIS) by NRIGs can be seen in Table B3. With this information we compute the domestic value added contents of exports, considering the aggregate use of the different EIS (see table

4).

Table 4 NRIG Domestic Value Added (VA) contents of services (millions dollars in 2007 and

%)

Countries Sectors

(1)

Exports

(2) VA

contents

(3) %

Services VA

(4) VAX

ratio % (VA/Exports)

(5)

Imports contents

(%)

Argentina

3

2706

2253

14,5

83,3

16,7

8

432

403

4,7

93,3

6,7

20

488

355

22,8

72,6

27,4

21

10716

8688

23,0

81,1

18,9

Rest

49612

37552

27,5

75,7

24,3

Total

63953

49250

25,9

77,0

23,0

Brazil

5

6889

6334

16,5

91,9

8,1

8

5844

5378

16,8

92,0

8,0

18

19804

17829

34,5

90,0

10,0

20

7090

6645

23,7

93,7

6,3

24

5013

4631

23,7

92,4

7,6

26

1610

1471

31,9

91,4

8,6

35

8491

6980

39,9

82,2

17,8

Rest

121470

100689

43,6

82,9

17,1

Total

176209

149957

38,6

85,1

14,9

Uruguay

19

901

771

18,5

85,5

14,5

Rest

5772

3970

36,8

68,8

31,2

Total

6673

4741

33,8

71,0

29,0

Chile

4

3014

2698

19,8

89,5

10,5

18

17285

14113

27,1

81,7

18,3

36

24820

19611

26,0

79,0

21,0

Rest

23980

16899

40,5

70,5

29,5

Total

69098

53321

30,6

77,2

22,8

Peru

18

9685

8377

19,6

86,5

13,5

Rest

19893

17943

22,5

90,2

9,8

Total

29577

26320

21,6

89,0

11,0

Colombia

8

2429

2276

4,2

93,7

6,3

Rest

29030

24592

27,8

84,7

15,3

Total

31458

26868

25,8

85,4

14,6

Note: (2)- Definition in equation A9; (3) see equation A8 and table B3; (4)=(2)/(1), see equation A10; (5)see equation A4.

Source: authors’ calculations based on GTAP

The share of these types of sectors is less than in the rest of the economy in almost all the sectors associated with NRIGs (see table 3). Only in the case of Brazil in sector 35 (Ferrousmetals) the share is greater than in the average global economy. In minerals sectors the intensity is typically greater than in the other commodities and close to the average of the economy (see the cases of Chile, Peru and Brazil). In the NRIG sectors the VAX ratio is typically high or the import contents of exports are low. This result is natural considering that this type of sectors are defined by the intensively use of domestic factors (natural resources).

Table 5 Share of domestic value added by economic infrastructure services (%) Countries Sectors Services domestic value added content

VAS/VA

I II

III

IV V VI Argentina

3

0,2

1,1

2,7

2,0

1,5

7,0

14,5

8

0,2

0,4

1,2

0,8

0,8

1,4

4,7

20

0,4

0,6

9,4

5,5

1,9

4,9

22,8

21

0,4

1,0

6,2

7,2

1,5

6,7

23,0

Rest

0,8

1,0

5,8

8,0

2,1

9,8

27,5

Total

0,7

1,0

5,7

7,5

2,0

9,0

25,9

Brazil

5

2,2

0,1

5,8

3,6

2,4

2,3

16,5

8

2,7

0,1

5,6

3,6

2,4

2,3

16,8

18

5,3

0,2

7,0

11,2

5,2

5,6

34,5

20

3,5

0,1

8,2

5,0

2,8

4,0

23,7

24

3,9

0,2

4,8

6,5

3,5

4,9

23,7

26

4,2

0,2

9,7

7,6

4,1

6,1

31,9

35

11,9

0,2

6,7

10,6

5,7

4,9

39,9

Rest

5,3

0,5

8,9

10,2

5,4

13,4

43,6

Total

5,2

0,4

8,2

9,4

4,9

10,5

38,6

Uruguay

19

1,5

1,4

4,2

3,3

6,3

1,9

18,5

Rest

3,6

1,2

7,5

9,3

7,0

8,2

36,8

Total

3,3

1,2

6,9

8,4

6,9

7,2

33,8

Chile

4

0,5

0,3

6,0

3,6

0,9

8,5

19,8

18

5,6

0,3

2,8

3,9

0,9

13,6

27,1

36

2,5

0,3

3,9

3,9

1,3

14,1

26,0

Rest

2,3

0,5

7,0

10,8

3,3

16,5

40,5

Total

3,2

0,3

4,7

6,1

1,8

14,5

30,6

Peru

18

2,5

0,3

1,0

2,4

3,7

9,7

19,6

Rest

2,1

0,5

1,8

5,1

4,4

8,6

22,5

Total

2,2

0,4

1,5

4,2

4,2

9,0

21,6

Colombia

8

0,4

0,1

1,0

0,6

1,2

0,8

4,2

Rest

2,1

0,6

5,2

6,6

5,9

7,3

27,8

Total

2,0

0,6

4,9

6,1

5,5

6,8

25,8

Source: authors’ calculations using table B4.

Also is important to highlight that we are only measured the EIS (cost services in Arbaches,2015 terminology) because conventionally specific product services are embodied in the gross value of each commodity sector11.

However there are three cases where the direct and indirect intensity in the use of imports is high: meat products (sector 20) in Argentina; ferrousmetals (sector 35) in Brazil; other minerals (sector 18) in Peru.

Results on the total value added generated by NRIG exports disaggregated by type of EIS are presented in Table 5 according to equation A¡Error! No se encuentra el origen de la referencia. of the Appendix A. In this case it is possible to identify an association between type of EIS service and the NRIG sector. We summarize results in table 6. For Argentina, an intensively use of trade services and transport services are verified in the chicken meat (20) and in vegetable oils (21) respectively. In Brazil minerals (sector 18) is associated with public services (I-electricity, gas and water), the beverage sector (26) in trade services (III), and finally in ferrousmetals (sector 35) the intensity link is with public services (I), transport (IV) and financial services (V). For Chile in minerals (18) the link is with public services (I). Finally in Peru also in minerals (18) the association is with other business services.

Table 6 Association between EIS and NRIG by country Countries Sectors EIS VA contents intensity I

III

IV V VI Argentina

20

+

21

+

Brazil

18

+

26

+

35

+

+ +

Chile

18

+

Peru

18

+ Source: authors’ calculations using table 5.

4.2.2 EIS productivity and prices competitiveness of NRIG

Finally, we present the results of the Leontief price model. As mentioned before, we are interested on analysing the effects of a 10% increase in EIS productivity on the prices of

11 This is a well know restriction to work with such a high level of aggregation of the IO analysis and only could be solved with more detail and deep product case studies.

NRIGs. This is modelled as a decrease in value added in services, as detailed in equation (A16) of the Appendix. In Table 7 we show the results of applying that productivity increase to each of the six different groups of EIS, together with a final simulation of a simultaneous 10% increase in all of them. Results should be interpreted as the price effect of that change on each of the selected NRIG and on the total production basket of the economy (using gross sectorial output as weights).

Table 7 Percent change on commodities’ prices as a result of a 10% increase in Economic Infrastructure Services’ productivity (%)

Country Sector Economic Infrastructure Services I II

III

IV V VI All Argentina

3

-0.05

-0.12

-0.32

-0.32

-0.21

-0.68

-1.70

8

-0.02

-0.05

-0.15

-0.12

-0.10

-0.17

-0.60

20

-0.07

-0.11

-0.83

-0.59

-0.24

-0.47

-2.30

21

-0.09

-0.12

-0.63

-0.73

-0.22

-0.65

-2.44

All

-0.13

-0.39

-0.98

-0.78

-0.43

-0.80

-3.51

Brazil

5

-0.25

-0.02

-0.59

-0.40

-0.27

-0.32

-1.86

8

-0.29

-0.02

-0.58

-0.40

-0.27

-0.32

-1.88

18

-0.54

-0.02

-0.70

-1.10

-0.54

-0.67

-3.56

20

-0.37

-0.02

-0.82

-0.53

-0.30

-0.47

-2.50

24

-0.41

-0.02

-0.50

-0.67

-0.38

-0.58

-2.55

26

-0.44

-0.02

-0.95

-0.76

-0.43

-0.70

-3.31

35

-1.11

-0.03

-0.67

-1.04

-0.57

-0.63

-4.05

All

-0.51

-0.34

-1.26

-1.12

-0.79

-1.14

-5.16

Uruguay

19

-0.19

-0.13

-0.46

-0.37

-0.65

-0.25

-2.06

All

-0.42

-0.55

-1.21

-0.74

-0.86

-0.61

-4.37

Chile

4

-0.12

-0.03

-0.61

-0.38

-0.12

-0.91

-2.16

18

-0.64

-0.03

-0.32

-0.43

-0.16

-1.35

-2.93

36

-0.34

-0.03

-0.42

-0.42

-0.21

-1.42

-2.85

All

-0.44

-0.43

-0.83

-0.91

-0.55

-1.88

-5.04

Peru

18

-0.34

-0.04

-0.10

-0.24

-0.36

-0.96

-2.03

All

-0.34

-0.53

-0.22

-0.52

-0.50

-1.14

-3.25

Colombia

8

-0.05

-0.01

-0.13

-0.09

-0.15

-0.12

-0.55

All

-0.30

-0.49

-1.08

-0.80

-0.80

-0.65

-4.12

Source: authors’ calculations based on GTAP As a first comment on these results, it can be noticed that the average impact of services on the prices of the economy is reduced. Moreover, this is particularly true for the selected commodities, which feature below-of-the-average changes. In fact, these goods are generally in the low tale of the distribution of price impacts along the economy (minerals

being the exception). The group of services which produces the largest impacts on prices in the majority of countries is Trade services (III), followed by Transport (IV) and Other business services (VI).

For the Argentinean case, the impacts in sectors 3 (Cereals), 8 (Other crops) and 21 (Vegetable oils and fats) are low and are mainly explained by improvements in Trade and Other business services productivities. Sector 8 presents a very low reaction (to any service improvement), whereas sector 20 (Meat products) is affected by Trade and Transport services.

In Brazil we find the highest impacts on prices. For example, sectors 18 (Minerals), 26 (Beverage and tobacco) and 35 (Ferrous metals) are the most influenced, mainly by Transport, Trade and Other business services and also Financial services. Importantly, Minerals and Ferrous metals price’ reactions to changes in Electricity, Gas and Water services productivity lie above the price impact of the latter on the whole economy prices. This is consistent with the fact that these NRIG are closely related to group I services, as mentioned in Table (6).

The same over-reaction is also found in the Mineral sector of both Chile and Peru, whereas the figures for Colombia are extremely low. In Uruguay the most important service group in terms of its impacts on NRIG competitiveness is Financial services, however the size of the impact is low.

Overall, we could present results of the association of services productivities and NRIG competitiveness by simply using Table (6) and changing its title. The main intuition is that commodities benefit from productivity changes of services according to the intensity in their respective production functions.

5. CONCLUSIONS

The initial motivation of the paper is that NRIGs are an important feature of South America trade pattern and South America is also a relevant supplier of this type of products in the world market (Vaillant, 2015). Policy makers in the region repeatedly sustain the need to search a path for the transformation of this pattern of specialization. The questions are: how do services can add value to the production and exports of NRIGs, and could the region benefit from the improvement of services to eventually become a net exporter of them? In the paper we give some new information that will be necessary to answer this question.

One of the characteristics of the actual era of globalization is the increase of the intensity in the use of specialized production services that participate as an intermediate input in many different activities and types of firms. The new technologies ability to penetrate in all kind of activities is one of the most important features of the changes to the production functions all over the world. This means servicification may be possible in many different activities and not only in dynamic manufacturing sectors. Our paper focuses on the identification of the phenomenon in the case of natural resources-intensive goods (NRIG) in South America.

Following Arbache (2015) we distinguish two types of services: specialized production services and cost services (mainly from economic infrastructure services). Two different methodologies are used to study the servification process in each type of services. To identify the set of products and countries in which an increment in the use of specialized production services could be observed an indirect methodology is applied, based on trade data performance in a recent period (2000s). Food and raw materials is the first subset composed by 12 products that are identified with dynamic behaviour combined with relevance in world market. For minerals sectors, 8 products fulfil the stablished criteria. Six South American countries are exporter of this set of 20 products of NRIG.

Additionally, a review of recent literature about specialized production services in a set of NRIG is proposed. This showed that the big scale of regional markets in NRIG, the intensification of production combined with the acceleration of technical change together with the increase requirements in standardization and certification (both at the domestic

and international levels) are the environment where specialized production services (in particulars new KIBS) are developed. Sectors with strong RCA (NRIG) generate domestic interlinkages, which imply an incentive to develop new economic activities. These markets are huge at international level and they can create a representative demand in particular for the regional countries. Based on this critical mass, new competitive services activities could be developed which have the potential to be internationalized. This is like a combined Hirschman and Linder mechanism, as referred by Gonzalez, Meliciani y Savona (2015).

For the analysis of the impact of the second type of services (EIS) we used a conventional input-output methodology. The analysis is developed for the six South American exporters in twelve sectors (the 20 NRIG products map in 12 sectors, 9 food and raw materials and 3 minerals), which implies seventeen different bilateral matches countrysector. Three countries (Argentina, Colombia in Uruguay) are linked to food and raw materials sectors. Peru is associated only to mineral sectors. Finally, Brazil and Chile export products in both types of sectors.

We developed two types of indicators to approximate the relevance of EIS in NRIG. First, we measure the domestic value added contents (by sub type of EIS) in each combination country-NRIG sector exports. Naturally, in NRIG sectors the intensity of the use of EIS is lower than in the rest of the economy for all countries. However, the relevance of EIS is greater in minerals sectors and in some cases, it is close to the average of the economy. The disaggregated results show an association between types of EIS services and NRIG sectors. For Argentina, an intensively use of trade services and transport services are verified in the chicken meat (20) and in vegetable oils (21) respectively. In Brazil minerals (sector 18) is associated with public services (I-electricity, gas and water), the beverage sector (26) in trade services (III), and finally in ferrousmetals (sector 35) the intensity link is with public services (I), transport (IV) and financial services (V). For Chile in minerals

(18) the link is with public services (I). Finally in Peru also in minerals (18) the

association is with other business services.

Finally, we present the results of the price model, which produces price effects on commodities following increases in EIS productivity. The general line of the association between EIS services and NRIG sectors is the same. However, we give evidence about the size of the effects considering how prices are affected by an increase in the productivity of transport services. The main impact is observed in the case of Brazil with

an increase in productivity of public services (I) imply a decrease in the price of ferrousmetals (35). The effect of transport service (IV) is relevant in the price of vegetable oil sector in Argentina, and a positive shock in productivity of other business services (VI) has a positive impact in competitiveness in metals (36) in Chile and minerals (18) in Perú.

New instruments in trade and industrial policy must be develop to promote the specialized products services in NRIG. It is necessary to know the nature of the process that generate this new comparative advantages. A deep understanding of the dynamic process of interaction between scale economics and the internationalization of this type of firm is required to reinforce this new structure of production. More cases studies must be done to know the mechanism of this development process. Finally, the results over EIS shows that the effect of Cost Services over commodities competitiveness is idiosyncratic of each country. This fact revealed a heterogeneous development of infrastructure among South American countries combined with different productive specialization pattern which imply different intensities in the use of the EIS.

REFERENCES

Ahamed, Z., Inohara, T. and Kamoshida, A. 2013. “The Servitization of Manufacturing: An Empirical Case Study of IBM Corporation”. International Journal of Business Administration, Vol. 4(2): 18-26.

Amiti M. and J. Konings, 2007.“Trade liberalization, intermediate inputs, and productivity: evidence from Indonesia” The American Economic Review, 97: 1611 - 1638. Anllo, Guillermo, Bizang Roberto y Katz, Jorge 2015: “Aprendiendo con el agro argentino: De la ventajacomparativa a la ventajacompetitiva”, Presentado en el IV Congreso REDLAS, Montevideo.

Arbache, Jorge, Moreira, Rafael, 2015: “How Can Services ImproveProductivity? The Case of Brazil”, University of Brasilia, Presentado en el IV Congreso REDLAS, Montevideo. Baldwin, R. E. and Robert-Nicoud, F.A Simple Model of. the Juggernaut Effect of Trade Liberalization, Centre for Economic Performance, London School of Economics and Political Science, 2008.

Baldwin, Richard 2012. “WTO 2.0: Global governance of supply chain trade”, Graduate Institute, Geneva, University of Oxford and CEPR.

Baldwin, Richard 2013. “Global supply chains: why they emerged, why they matter, and where they are going” in Global value chains in a changing world,” Edited by Deborah K. Elms and Patrick Low, WTO, Fund Global Institute.

Cattaneo, Gereffi yStaritz, 2010. ”GLOBAL VALUE CHAINS IN A POSTCRISIS WORLD A DEVELOPMENT PERSPECTIVE”, Olivier Cattaneo, Gary Gereffi, and Cornelia Staritz, Editors, The World Bank.

Gisela Di Meglio, G, Gallego J., Maroto, A. Savona, M., 2015. Services in Developing Economies:A new chanc for catching up?, Unpublished, work in progress. Escaith, Hubert. 2008. “Measuring Trade in Value Added in the New Industrial Economy: Statistical Implications.” Unpublished.

Gallego Juan Miguel y Hernán Jaramillo, 2015. “Serviciosintensivos en conocimiento en la industriadelpetróleo en Colombia”, Presentado en el IV Congreso REDLAS, Montevideo. Gana Joaquín y Meller, Patricio, 2015. “EL COBRE CHILENO COMO PLATAFORMA DE INNOVACIÓN TECNOLÓGICA “, Presentado en el IV Congreso REDLAS, Montevideo.

Gilles, Enrique y Valecilla, Jaime, 2015:

“Infraestructura y exportaciones de commoditiesagrícolas en Colombia”, Presentado en el IV Congreso REDLAS, Montevideo.

Grossman, G.M. and Rossi-Hansberg, E., 2006. “The Rise of Offshoring: It’s Not Wine for Cloth Anymore”, prepared for the symposium sponsored by the Federal Reserve Bank of Kansas City on “The New Economic Geography: Effects and Policy Implications,” Jackson Hole, Wyoming, August 24-26.

Grossman, G.

M., & Rossi-Hansberg, E.

(2012).Task Trade between Similar Countries.Econometrica, 80(2), 593-629. doi: http://www.econometricsociety.org. Hoekman and Francois, 2010.Services and Trade Policy.Journal of Economic Literature 48. Hummels D. and P. Klenow, 2005.“The Variety and Quality of a Nation's Exports” American Economic Review, 95:704-723.

Johnson, Robert C. and Guillermo Noguera, 2012. Accounting for intermediates: Production sharing and trade in value added. Journal of International Economics, 86(2), pp 224-236. Lodelfalk, Magnus 2015.Servicification of Manufacturing Firms Makes Divides inTrade Policy- Making Antiquated, Working Paper 1/2015, Örebro University School of Business. Marin Anabel, Stubrin Lilia, J. J.da Silva Jr., 2015. “KIBS ASSOCIATED TO NATURAL

RESOURCE BASED INDUSTRIES. SEEDS INNOVATION AND REGIONAL

PROVIDERS OF THE TECHNOLOGY SERVICES EMBODIED IN SEEDS IN

ARGENTINA AND BRAZIL 2000-2014”, Presentado en el IV Congreso REDLAS, Montevideo.

Miroudot, Sebastien, Rainer Lanz &AlexandrosRagoussis (2009). Trade in IntermediateGoods and Services. OECD Trade Policy Papers No. 93.

Nordås, H. K. (2010). Trade in goods and services: Two sides of the same coin? EconomicModelling, 27(2), 496-506. doi: 10.1016/j.econmod.2009.11.002. Sturgeon, T. and Memedovic 0. (2010). Mapping Global Value Chains: Intermediate Goods Trade and Structural Change in the World Economy, DEVELOPMENT POLICY AND STRATEGIC RESEARCH BRANCH WORKING PAPER 05/2010, UNIDO, Vienne. Vandermerwe, S. and Rada, J. (1998). Servitization of Business: Adding Value by Adding Services. European Management Journal, 6(4), 314–324.

Vaillant, Marcel (2015). Servicios y bienes intensivos en recursos naturales en sudamérica: nuevas oportunidades de transformación productiva, CEPAL, División de Comercio.

APPENDIX A- METHODOLOGICAL

In this section we briefly present the input-output model used in this document. It is important to clarify the use of notation and conventions beforehand. To begin with, upper case letters denote matrices, whereas lower case letters denote vectors; finally, letters with subscripts denote real numbers. There are n sectors in the economy. The (nx1) final demand vector is denoted by 𝑓; the use of domestic input i by industry j in monetary terms is denoted by 𝑧𝑖𝑗; the direct coefficient matrix, of dimension n, is called 𝐴 with elements 𝑎𝑖𝑗; the corresponding Leontief inverse is 𝐿 with elements 𝑙𝑖𝑗. The import coefficient matrix is named 𝐴𝑚 with elements 𝑚𝑖𝑗, which express the share of foreign input i in terms of domestic sector j’s gross output. Finally, the gross output vector is 𝑥; and the value added vector is 𝑣. The standard representation of the input-output model is given by:

𝑥= (𝐼−𝐴)−1𝑓= 𝐿𝑓.

(A1)

A1. Domestic value added contents of infrastructure service in NRIGs exports

Each time there is an export of a NRIG (in fact, of every good) this implies an increase in gross output and value added, not only in the sector from which the good comes from, but throughout all the economy via sectorial linkages. In this section we present the methodology to assess the magnitude of those effects as a way to weight their relative importance in the economy.

We start by considering the fact that final demand 𝑓 can be decomposed into domestic (𝑓𝑑) and the foreign final demand (𝑒):

𝑓= 𝑓𝑑+ 𝑒 .

(A2) According to Hummels et al. (2001) the imports content of total exports can be calculated as:

𝑚𝑒= 𝐴𝑚𝐿𝑒.

(A3) This is a vector of all the imports contents in global exports. If we sum over sectors and substract from exports we obtain the total domestic value added contents of exports

𝑣𝑑𝑡𝑒= 𝑖𝑒−𝑖𝑚𝑒= ∑𝑒𝑠 𝑠 −∑𝑚𝑠 𝑒 𝑠

(A4)

where 𝑖 is a vector of ones. We can also express the import contents as a proportion of total exports,

𝑠𝑚𝑒= 𝐷(𝑒)−1𝐴𝑚𝐿𝑒, (A4) where 𝐷( ) is the diagonal operator.

However, what we want to know is the composition of domestic value added by sector of origin and linked with foreign demand (exports). So we need to use the vector of the ratio of domestic value added to gross output, defined as 𝑣𝑠 𝑟= 𝑣𝑠 𝑥𝑠= 𝑥𝑠−∑𝑎𝑗𝑠𝑥𝑠−∑𝑚𝑗𝑠𝑥𝑠 𝑗 𝑗 𝑥𝑠 = 1 −∑𝑎𝑗𝑠−∑𝑚𝑗𝑠 𝑗 𝑗

(A5) In matrix terms 𝑣𝑟= (𝐷(𝑥))−1𝑣= 𝑖−𝐴´𝑖−𝑀´𝑖

(A6) Then, the domestic value added contents in total exports (by sectors) is given by: 𝑣𝑑𝑒= 𝐷(𝑣𝑟)𝐿𝑒

(A7) In turn, we disaggregate this value added by each sector of the foreign demand (export vector). Then the domestic value added by sector in exports is: 𝑉𝐷𝑒= 𝐷(𝑣𝑟)𝐿𝐷(𝑒),

(A8) where 𝑉𝐷𝑒 is a square matrix of dimension n, in each column we have the domestic value added generated in a specific sector as a result of exports of each sector (rows). Given the above, it holds that:

𝑣𝑑𝑒= 𝑉𝐷𝑒𝑖

(A9) Also the total sum of domestic value added is 𝑣𝑑𝑡𝑒= 𝑖´𝑣𝑑𝑒 (see equation A4). The ratio between domestic value added contents and exports is:

𝑣𝑎𝑥= (𝐷(𝑒))−1𝑣𝑑𝑒

(A10)

𝑉𝐴𝑋= (𝐷(𝑒)) −1𝑉𝐷𝑒

(A11) Where: 𝑣𝑎𝑥 and 𝑉𝐴𝑋 are the well-known value added to export ratios of the Global Value Chain literature (see for example, Johnson and Noguera, 2012). We assume that the domestic value added contents in imports is zero, so one minus the vax ratio is the import contents associated with total exports (𝑠𝑚𝑒= (𝑖−𝑣𝑎𝑥) see equation A4). This

could be also defined by sector and also expressed as a proportion of foreign demand (exports).

A2. Leontief price model

If we assume that it is not quantities but prices that adjust in the economic system, then the standard input-output model can be interpreted as a price model (Miller and Blair, 2009). Looking at sectors from a column perspective, we have that gross sectorial output can be decomposed into the use of intermediate inputs (both domestic and imported) and value added generated,

𝑥𝑗= ∑𝑧𝑖𝑗 𝑖 + ∑𝑚𝑖𝑗 𝑖 + 𝑣𝑗.

(A12)

Next, dividing each side of the above equation by gross sectorial output, we have

1 = ∑𝑎𝑖𝑗

𝑖 + ∑𝑎𝑚𝑖𝑗 𝑖 + 𝑣𝑗 𝑟 (A13) where technical coefficients for domestic and imported intermediates now appear explicitly. The equation reflects the equality between the per-gross-output costs (right hand side) and the unit price (left hand side). Adopting matrix notation, these prices can be interpreted as a price index of the economy:

𝐩′ = 𝐩′(𝐀+ 𝐀𝐦) + 𝒗𝒓′ (A14)

𝐩= (𝐀+ 𝐀𝐦)´𝐩+ 𝒗𝒓=(𝐀𝐭)´𝐩+ 𝒗𝒓

(A15)

Where 𝐩′ is a row vector of industry prices, 𝐀 is the square matrix of domestic direct coefficients, 𝐀𝐦 is the corresponding square matrix for imported intermediates, and 𝐯′𝐜 is a 1xn vector of value added to gross output ratios. Solving for prices and transposing the result, the Leontief price model can then be expressed as:

𝐩= (𝐈−𝐀𝐭′)−𝟏𝐯𝐜≡𝐋′𝐯𝐜 (A16)

In this equation, 𝐀𝐭 is the sum of matrices 𝐀 and 𝐀𝐦, whereas 𝐋′is the transpose of the corresponding Leontief inverse. A given equation of this system writes:

𝑝𝑖= ∑𝑙𝑖𝑗 𝑡𝑣𝑗 𝑗 = 𝑙𝑖1 𝑡𝑣1 + ⋯+ 𝑙𝑖𝑗 𝑡𝑣𝑖+ ⋯+ 𝑙𝑖𝑛 𝑡𝑣𝑛.

(A17) In other words, every change in sectorial value added will have effects all along the economy prices, according to the Leontief total effect coefficients, 𝑙𝑖𝑗 𝑡 which denote the change in industry i price following an change in unit value added sector j.

APPENDIX B- STATISTICAL

Table B.1 Sectors of the input-output matrix Sector Title Name

1 PDR

Paddy rice

2 WHT

Wheat

3 GRO

Cereal grainsnec

4 V_F

Vegetables, fruit, nuts

5 OSD

Oilseeds

6 C_B

Sugarcane, sugarbeet

7 PFB

Plant-basedfibers

8 OCR

Cropsnec

9 CTL

Bovine cattle, sheep and goats, horses

10 OAP

Animal productsnec

11 RMK

Rawmilk

12 WOL

Wool, silk-wormcocoons

13 FRS

Forestry

14 FSH

Fishing

15 COA

Coal

16 OIL

Oil

17 GAS

Gas

18 OMN

Mineralsnec

19 CMT

Bovinemeatproducts

20 OMT

Meatproductsnec

21 VOL

Vegetable oils and fats

22 MIL

Dairyproducts

23 PCR

Processed rice

24 SGR

Sugar

25 OFD

Foodproductsnec

26 B_T

Beverages and tobacco products

27 TEX

Textiles

28 WAP

Wearingapparel

29 LEA

Leatherproducts

30 LUM

Wood products

31 PPP

Paperproducts, publishing

32 P_C

Petroleum, coalproducts

33 CRP

Chemical, rubber, plasticproducts

34 NMM

Mineral productsnec

35 I_S

Ferrousmetals

36 NFM

Metalsnec

37 FMP

Metal products

38 MVH

Motor vehicles and parts

39 OTN

Transportequipmentnec

40 ELE

Electronicequipment

41 OME

Machinery and equipmentnec

42 OMF

Manufactures nec

43 ELY

Electricity

44 GDT

Gas manufacture, distribution

45 WTR

Water

46 CNS

Construction

47 TRD

Trade

48 OTP

Transportnec

49 WTP

Watertransport

50 ATP

Air transport

51 CMN

Communication

52 OFI

Financialservicesnec

53 ISR

Insurance

54 OBS

Business servicesnec

55 ROS

Recreational and otherservices

56 OSG

Public Administration, Defense, Education, Health

57 DWE

Dwellings Source: GTAP.

Table B2 Economic infrastructure services

BIG SECTOR SERVICES

SECTORS GTAP

I Electricity, gas and water 43;44;45 II Construction

46

III

Trade

47

IV Transport and communications 48;49;50;51 V Financial and insurance

52 and 53

VI Other business services

54

Source: authors’ calculations using table A1.

Table B3 Countries and sectors direct requirements of services/gross output by big sectors (2007) Countries Sectors Direct domestic services requirements/gross output Services as intermediate inputs Domestic Intermediate use I II

III

IV V VI Argentina

3

0,00

0,01

0,02

0,01

0,01

0,07

0,12

0,39

8

0,00

0,01

0,01

0,00

0,01

0,01

0,04

0,18

20

0,00

0,00

0,05

0,03

0,00

0,01

0,10

0,78

21

0,00

0,00

0,06

0,09

0,00

0,01

0,17

0,88

Brazil

5

0,01

0,00

0,04

0,02

0,01

0,00

0,10

0,39

8

0,02

0,00

0,04

0,02

0,01

0,00

0,10

0,40

18

0,05

0,00

0,05

0,13

0,04

0,03

0,29

0,53

20

0,02

0,00

0,05

0,04

0,01

0,02

0,14

0,74

24

0,03

0,00

0,02

0,06

0,02

0,03

0,17

0,69

26

0,02

0,00

0,07

0,06

0,02

0,03

0,21

0,75

35

0,10

0,00

0,03

0,08

0,03

0,01

0,26

0,63

Uruguay

19

0,00

0,00

0,01

0,04

0,03

0,01

0,09

0,79

Chile

4

0,00

0,00

0,10

0,04

0,01

0,07

0,22

0,29

18

0,08

0,00

0,03

0,04

0,00

0,11

0,27

0,46

36

0,01

0,00

0,04

0,03

0,01

0,08

0,16

0,56

Peru

18

0,00

0,00

0,02

0,03

0,03

0,07

0,15

0,49

Colombia

8

0,00

0,00

0,01

0,00

0,01

0,00

0,03

0,14

Source: authors’ calculations using GTAP input output matrices.

Table B4 Economic infrastructure services domestic value added contents by selected set of NRIG in South America (millions of dollars, 2007) Countries Commodity Exports Services domestic value added content VA contents of selected services Total VA contents I II

III

IV V VI Argentina

3

2706

5

24

61

45

34

158

327

2253

8

432

1

2

5

3

3

6

19

403

20

488

2

2

33

20

7

17

81

355

21

10716

34

83

542

628

128

579

1995

8688

Rest

49612

290

391

2178

3000

797

3687

10341

37552

Total

63953

332

501

2819

3695

969

4446

12763

49250

Brazil

5

6889

142

7

369

228

154

146

1048

6334

8

5844

147

6

302

194

129

123

902

5378

18

19804

949

32

1254

1995

927

999

6157

17829

20

7090

233

9

546

335

185

265

1573

6645

24

5013

182

8

220

299

162

228

1099

4631

26

1610

62

3

142

111

60

90

469

1471

35

8491

830

15

467

737

395

340

2784

6980

Rest

121470

5294

513

8965

10236

5391

13481

43880

100689

Total

176209

7840

592

12266

14135

7404

15673

57910

149957

Uruguay

19

901

11

10

33

25

48

14

143

771

Rest

5772

143

47

297

371

277

327

1461

3970

Total

6673

155

57

329

396

325

341

1604

4741

Chile

4

3014

14

7

163

96

25

231

535

2698

18

17285

791

36

402

552

129

1917

3827

14113

36

24820

486

62

763

770

256

2765

5103

19611

Rest

23980

389

78

1190

1827

559

2794

6837

16899

Total

69098

1681

183

2518

3245

969

7707

16302

53321

Peru

18

9685

206

28

81

200

314

816

1644

8377

Rest

19893

378

89

314

911

795

1549

4036

17943

Total

29577

584

117

396

1111

1109

2365

5680

26320

Colombia

8

2429

8

1

24

15

28

19

95

2276

Rest

29030

519

151

1280

1630

1463

1795

6838

24592

Total

31458

528

152

1304

1644

1491

1814

6933

26868

Source: authors’ calculations using GTAP input output matrices.

Table B5 Share of each economic infrastructure service in total domestic value added of services (%) Countries Sectors Services domestic value added content in NRIG exports Intermediate use of Services I II

III

IV V VI Argentina

3

1,6

7,3

18,7

13,8

10,4

48,2

100,0

8

3,2

8,1

25,3

16,4

17,3

29,7

100,0

20

1,9

2,8

41,3

24,2

8,3

21,4

100,0

21

1,7

4,2

27,2

31,5

6,4

29,0

100,0

Rest

2,8

3,8

21,1

29,0

7,7

35,6

100,0

Total

2,6

3,9

22,1

29,0

7,6

34,8

100,0

Brazil

5

13,6

0,7

35,2

21,8

14,7

13,9

100,0

8

16,3

0,7

33,5

21,5

14,3

13,7

100,0

18

15,4

0,5

20,4

32,4

15,1

16,2

100,0

20

14,8

0,6

34,7

21,3

11,8

16,8

100,0

24

16,5

0,7

20,0

27,2

14,8

20,7

100,0

26

13,3

0,6

30,3

23,7

12,8

19,3

100,0

35

29,8

0,5

16,8

26,5

14,2

12,2

100,0

Rest

12,1

1,2

20,4

23,3

12,3

30,7

100,0

Total

13,5

1,0

21,2

24,4

12,8

27,1

100,0

Uruguay

19

8,0

7,4

22,9

17,7

34,0

10,0

100,0

Rest

9,8

3,2

20,3

25,4

18,9

22,4

100,0

Total

9,6

3,6

20,5

24,7

20,3

21,3

100,0

Chile

4

2,6

1,4

30,5

17,9

4,6

43,1

100,0

18

20,7

0,9

10,5

14,4

3,4

50,1

100,0

36

9,5

1,2

14,9

15,1

5,0

54,2

100,0

Rest

5,7

1,1

17,4

26,7

8,2

40,9

100,0

Total

10,3

1,1

15,4

19,9

5,9

47,3

100,0

Peru

18

12,5

1,7

4,9

12,2

19,1

49,6

100,0

Rest

9,4

2,2

7,8

22,6

19,7

38,4

100,0

Total

10,3

2,1

7,0

19,6

19,5

41,6

100,0

Colombia

8

9,0

1,2

25,1

15,5

29,6

19,6

100,0

Rest

7,6

2,2

18,7

23,8

21,4

26,3

100,0

Total

7,6

2,2

18,8

23,7

21,5

26,2

100,0

Source: authors’ calculations using table B4.