IntraIndustry Foreign Direct Investment Laura Alfaro Andrew Charlton

Intra-Industry Foreign Direct Investment Laura Alfaro Andrew Charlton Harvard Business School & NBER London School of Economics

What Do We Do in This Paper? • Study patterns of vertical and horizontal multinational activity: large new data set of 650, 000 multinational subsidiaries in 90+ countries (close to population of MNCs). • Traditionally, the literature has distinguished between two forms of— and motivations for—multinational activity (different effects factor incomes within and across countries). – “Horizontal” FDI: locating production to be closer to customers and avoid trade costs (Markusen, 1984; Brainard, 1993), – “Vertical” FDI: firm’s attempts to take advantage of cross-border factor cost differences (Helpman, 1984; Helpman and Krugman, 1985). • Most research has found that the bulk of FDI is horizontal. • Our results suggest that, due to data limitations, the literature has systematically under-estimated vertical FDI.

Patterns: Firm Data • Consistent with conventional wisdom, – The bulk of multinational activity occurs between the rich nations. – At the 2 digit industry level: horizontal FDI (subsidiaries in the same industry as their parent) >vertical FDI (subsidiaries which supply their parent with inputs). • But … – At the 4 digit level, more vertical activity. → Many of the foreign subsidiaries in the same 2 digit industry as their parent are in fact producing highly specialized inputs into their parents’ production. – This pattern prevails even within developed countries.

Intra Industry FDI • We call these subsidiaries unveiled at higher levels: ‘intra-industry vertical’ FDI. – • Qualitatively different to vertical subsidiaries which cross twodigit industry codes (‘inter-industry vertical FDI’). • Supply parent firms with high-skill products • Mostly located in high-skill countries. These facts are: a. At odds with conventional wisdom (more horizontal than vertical) b. And… cannot be fully explained by the traditional models of comparative advantage (fragmentation: cost differences). - But …. are consistent with trade data documenting large flows of intra-firm trade in intermediate inputs between rich countries, Bernard et al. (2006).

Discrepancy: Misclassification of Vertical FDI • Significant amount of vertical FDI was misclassified as horizontal: 1. Most vertical FDI is north-north, it has been assumed to be market seeking (horizontal) • Firm level data indicates that these are vertical relationships: parent firms sourcing inputs from their subsidiaries in other northern countries. 2. Skill differences between parent and subsidiaries are small (even within vertical FDI): • This also lends support to horizontal motivations. 3. The vertical nature of these relationships is missed at the 2 digits: • Many subsidiaries are supplying goods to their parents where both the input and the final good are in the same 2 digit SIC.

Explaining Vertical FDI • Both intra and inter industry subsidiaries: provide inputs to parent firms. – But … intra-industry FDI is much harder to explain with the standard theory of vertical FDI (factor cost differences). • ‘Intra-industry vertical FDI’: north-north – Differences between parent and child skill levels are small. – Average proximity between two industries is higher for parent subsidiary pairs: • Proximity: proportion of the intermediate product used directly in the final good ( i. e. raw materials have low proximity variables). • Intra-industry FDI: tendency of multinational firms to own the later stages of the production chain, and outsource the production of early stages and raw materials.

Outline • Introduction and Motivation • Data • Methodology: – Vertical and Horizontal FDI • Patterns and Results: – Discrepancy between Aggregate and Firm Data – Intra-Industry FDI – Explaining Patterns: Comparative Advantage and Proximity • Conclusions

MNC Activity • MNC activities are best measured by firm-level data (Barba Navaretti and Venables, 2005). – Few countries have firm level data. – Researchers tend to use FDI flows from the Balance of Payments statistics as proxy for MNC activity. • Characterization of world multinational activity using firm level data.

The D&B Data Set • Worldbase data: database of both public and private companies in more than 213 countries and independent territories in 2004. Complied by Dun and Bradstreet. • The unit of record is the ‘establishment’ rather than the firm. • 4 -digit SIC-1987 code of the primary industry in which firm operates; for several countries, SIC codes of up to 5 secondary industries listed in descending order of importance. • Detailed ownership information: including information about the firm’s family members (no of family members, its domestic parent and its global parent). • Information about the firm’s status (joint-venture, corporation, partnership) and its position in the hierarchy (branch, division, head quarters). • Sales, employment, (some) export, age.

Foreign Ownership • We describe an establishment as foreign owned if it satisfies two criteria: – Foreign owned: must report a global parent firm and that parent firm must be in a different country. – Parents are defined in the data as entities which have legal and financial responsibility for another firm. • Combining the location and ownership information it is possible to identify 650 000+ firms in the database which are owned by a foreign parent.

Comparisons with Other Data: UNCTAD • UNCTAD’s World Investment Report 2004 reports that there are 61, 582 parent firms with 926, 948 affiliates operating in the world. • In the D&B dataset there are 72, 978 parent firms which have 658, 188 affiliates in foreign countries reporting to them. • Differences: – Our data is at the plant level, while their data is at the firm level. – UNCTAD data is inflated by a huge number of Chinese observations (424 196): all approved FDI projects registered by the Chinese, but is an overestimate of the number of actual foreign firms.

Comparisons with Other Data: US BEA • BEA’s U. S. Direct Investment Abroad: Benchmark Survey, is a census conducted every 5 years covering virtually the entire population of U. S. MNC’s. – In 2004, BEA reports that sales (employment) by foreign affiliates of U. S. MNCs totaled $3, 238 billion (10. 02 million employees). – In 2005 the DNB data : sum of all sales (employment) by foreign establishments reporting US parents was $2, 795 b (10. 01 million employees). – The distribution across countries is also consistent.

Comparisons with US BEA

General Patterns • The vast majority of our foreign owned subsidiaries are in richer countries and services.

Measuring Horizontal versus Vertical FDI • Data limitation: we do not observe intra firm trade. • We infer it from information about the goods produced in each of the firm’s establishments and the input-output relationship between those goods. – Hummels, Ishii, and Yi (2001): input-output tables to measure a country’s vertical specialization. • Advantages: – Large amount of data for many countries and industries; value of intra-firm trade not affected by transfer pricing. – Using I-O tables avoids the arbitrariness of classification schemes that divide goods into “intermediate” and other categories; Hummels et al. (2001). • But…identification of vertical subsidiaries as those which supply inputs to their parents relies on a number of assumptions.

Measuring Horizontal and Vertical • We calculate bilateral horizontal and vertical FDI using firm ownership data and U. S. input output matrix. – Horizontal FDI: activity of those foreign owned subsidiaries in the same industry as their parent. – Vertical FDI: activity of foreign owned subsidiaries in industries which are upstream from their parent’s industry (according to the US input output matrix). – Complex FDI: firms satisfy both. – None: If they satisfy neither of these criteria. Horizontal Complex Vertical

Measurement • • For each firm: up to six SIC codes for itself and its parent. Let S be the set of SIC codes of the subsidiary, and let P be the set of SIC codes of the parent. We use notation x → z to denote any element x being an input into an element z where x Є S and z Є P. We define x → z if the input output coefficient from the US input output matrix is greater than a threshold level which we vary. We define an owned establishment as: – Horizontal if S and P share any element (if x│x Є S ۷ x Є P) or if the sets are identical (if S=P) – Vertical if any element of S is an input into any element of P ( x│ x → z where x Є S and z Є P) and if the sets are not identical (if S≠P) – Complex if they share any element (if x│ x Є S ۷ x Є P) and if any element of S is an input into any element of P ( x│ x → z where x Є S and z Є P) and if the sets are not identical (if S≠P). – Neither if none of these connections exist.

Methodology: Input-Output Analysis • Threshold: determine the strength of the relationship required to assume that a subsidiary is a supplier to its parent. – Main results: threshold of 0. 05 for the ‘total requirements’ coefficient (i. e. the use of a commodity directly and indirectly by an industry). • Robustness: 0. 01 and 0. 1. • We use an alternative vector of input-output coefficients based on the ‘direct requirements’ (i. e. the use of a commodity directly by an industry) with a threshold of zero. • Analysis of the results: methodology is capturing a supply chain relationship between parents and subsidiaries.

Vertical and Horizontal: Some Results • Within manufacturing subsidiaries (188721), there are 112, 939 vertical subsidiaries and 104, 057 horizontal subsidiaries (15. 8 million versus 11. 9 million employees). – We exclude none and complex in the current analysis. • Similar results excluding IO relation within same country. • Results seem consistent related party trade data U. S.

Most Frequent Parent-Subsidiary Horizontal Industry Combinations in DNB Data

Most Frequent Parent-Subsidiary Upstream Vertical Industry Combinations in DNB Data

Firm Level Example: General Motors • General Motors Corporation: 2, 248 entities which report it as their ‘global ultimate parent’ – 455 are subsidiaries outside the United States – 123 subsidiaries outside the United States in manufacturing industries. • 68 are ‘horizontal’ subsidiaries (i. e. in the same primary 4 digit SIC code as their parent firm, GM SIC 3711 Motor Vehicles and Passenger Car Bodies) • 42 subsidiaries as being ‘vertical’ FDI (i. e in industries which are inputs in to the parent industry). – The non-manufacturing subsidiaries are primarily dealerships, credit, and insurance institutions.

General Motors: Characteristics of Vertical FDI • Vertical subsidiaries top industries: Specialized Auto Parts (SIC 3714) e. g GM Strasbourg which produces carburetors, pistons, rings, and valves in France; GMI Engineering which produces diesel engine parts in Japan; Vehicle engines (SIC 3519) e. g. Powertrain-Kaiserslautern in Germany. • Average skill intensity of the industries of GM subsidiaries is not significantly different. • The set of GM’s foreign subsidiaries does not include any firms producing what might be called the ‘raw materials’ or ‘low skill inputs’ into the production of automobiles. – GM’s ‘vertical FDI’ is focused on the “penultimate stages” in the vertical production chain.

Vertical Activity: In Rich Countries

Vertical: Level of Aggregations • More than half of all vertical subsidiaries are in the same 2 digit industry as their parent but a different 4 digit industry (for example, an automaker (SIC 3711) sourcing specialized auto parts (SIC 3714) from its foreign-owned subsidiary). – The vertical nature of these relationships is missed at the 2 digit level since many subsidiaries are supplying goods to their parents where both the input and the final good are in the same 2 digit SIC code 112, 939 93, 168 65, 550 42, 783 – Lower bound… but linked Input-Output relations.

Vertical FDI: In Close Activities

Vertical Activity: Skill Differences

FDI and Trade Facts • FDI literature: multinational subsidiaries which supply their parents with intermediate goods will be located in low factor costs countries. – Evidence of comparative advantage considerations in MNCs vertical location decisions, Yeaple (2003) and Hanson et al (2001). – Implication: Intra-firm trade will be higher between rich and poor countries than between rich countries. • However …. Bernard, Jensen and Schott’s (2006) find that low income countries have low shares of intra-firm exports to the US, while high income countries generally report above average intra-firm imports to the U. S. – Implications: intra-firm trade data: lot of vertical FDI between rich countries.

Inter and Intra Industry FDI and Trade Facts • Distinction between intra and inter industry vertical FDI resolves this contradiction. • Analysis of FDI using data with industry information only at the 2 digit focuses exclusively on inter-industry FDI and misses intra-industry vertical FDI. – Firms engaging in inter-industry FDI are more likely to be sourcing low skill inputs from low skill countries: • Validates the results of FDI studies at the 2 digit level. • Including intra-industry vertical FDI (predominantly between rich countries): high share of intra-firm trade flows between rich countries observed in the trade data.

Vertical Patterns • Analyze importance of comparative advantage and the inputs position in the production process in the determination of vertical FDI. • Following Brainard (1997), Yeaple (2003), Carr et al. (2001) FDIijs = 2 Sum. Mkt. Sizeij + 3 Distanceij + 4 Country. Skilli + 5 Country. Skil. Lij Industry. Skill. Ints + 6 Industry. SKill. Ints + ijs (1) – i and j: host and parent country, s: industry of the subsidiary. – FDI : bilateral multinational activity in an industry (number of subsidiaries, total sales, total employment). – Distanceij: bilateral distance between the home and host country. – Market size: sum of the GDPs in the host and parent economies. – Country skill: average years of schooling. – Industry Skill Intensity: ratio of non-production to total workers. – Only manufacturing.

Determinants of Multinational Bilateral Activity Multinational Activity in Each Bilateral Industry Pair--Tobit Regression

Determinants of Multinational Bilateral Activity Multinational Activity in Each Bilateral Industry Pair--Tobit Regression

Comparative Advantage and Proximity • Interaction between the relative skilled-labor abundance of countries with the skilled-labor intensity of industries. – 2 digits: strong evidence that vertical FDI is driven by comparative advantage, i. e. low skill activities tend to be located in low skill countries. – 4 digit: we find significantly less evidence that vertical FDI is driven by comparative advantage. • Proximity: the position of intermediate inputs in the chain of production contributes to the understanding of the patterns of intra-industry FDI: – Goods closer to raw materials are less likely to be the subject of FDI than intermediate goods which are proximate to the final good.

Measuring Proximity: Requirements • Requirements: ratio of direct/total requirements coefficients. – Direct requirements coefficient, i. e. , the amount of the output of industry i used directly as an input into industry j – Total requirements coefficient, i. e, the total amount of industry i used either directly or indirectly in the production of industry j. • The more of the intermediate product used directly in the final good the higher the proximity variable, i. e. raw materials have low proximity variables. • Closeness: absolute difference between the four digit SIC codes of the two products. – Closeness variable takes advantage of the fact that the 1987 SIC groups similar industries together.

Determinants of Multinational Bilateral Activity Multinational Activity in Each Bilateral Industry Pair--Tobit Regression

Rationales for Proximity • Information advantages associated with the co-ownership of later stages: activities involved in producing proximate inputs may have more in common with the production of the final good than do the activities involved in the production of raw materials, • Firms may be more worried about their intellectual property when the good is closer to their final good. • Monitoring advantage over the penultimate stages of the production. • Maximize quality control over later stages of production. • Contractability characteristics later stages of production (skill intensity, capital intensity, product characteristics…) (Spencer (2006) overview…. Bernard et al. (2006), Antras (2003), Aghion and Tirole (1995))

Conclusions • Firm level data: close to a comprehensive picture multinational activities. • The firm level data: vertical FDI is larger than commonly thought (Hanson et al. 2001, 2005). • Discrepancy: Significant amount of vertical FDI was misclassified. – North-north FDI between parent – Subsidiaries in similarly skilled activities, – More than half of all vertical subsidiaries are only observable at the four-digit level because the inputs they are supplying are so proximate to their parent firm’s final good.

Conclusions • ‘Intra-industry’ vertical subsidiaries: qualitatively different to the interindustry vertical FDI visible at the two-digit level. – Produce inputs with skill intensity overwhelmingly producing in high skill countries. • Activities not readily explained by the comparative advantage models. • Pattern of intra-industry north-north vertical FDI: decision to outsource versus own the production of intermediate inputs. – Multinationals source raw materials and inputs in early stages of production from outside the firm, – but tend to own the stages of production proximate to their final production giving rise to a class of high-skill intra-industry vertical FDI.

Conclusions: Implications • Level of aggregation: important elements of the pattern of foreign direct investment are missed at the 2 digit level and are not observable without industry data. – Echoing results by Schott (2003) for trade—highlight the importance of shifting away from industry analysis towards more disaggregated data to understand the location decisions firms. • Analysis suggest that foreign activity may be better explained by more complex production processes involving several stages which incorporate both arms length and in sourcing decision. • Future research…