PhD defense June 16th 2004 Helga Kristjánsdóttir Determinants of Exports and Foreign Direct Investment in a Small Open Economy PhD defense June 16th 2004 Helga Kristjánsdóttir
Background Following World War II, the production capacity of industrialized countries increased substantially In 1960s large Japanese car manufacturers had to choose between exporting and investing in United States Multinational entity (MNE), or a multinational, is a firm with multinational activities
Background In 1980s, general-equilibrium and partial-equilibrium trade theories were combined in New Trade Theory Vertical FDI – Production located to gain access to abundant factors (Helpman, 1984) Horizontal FDI – Production located to overcome trade costs and gain market access (Markusen, 1984)
Background Most FDI among developed countries, i.e., East and West, rather than North and South Small open economies particularly dependent on exports and FDI Export ratio of Iceland is comparatively small, or 34% in 1999 Scandinavian countries’ export ratios ranged from 37-44% Japan and US had export ratios below 12%
Trade in 1940s and 1950s
Trade in 1960s and 1970s Iceland joined EFTA in 1970
Trade in 1980s and 1990s
Research Objective Extend existing knowledge in the field Emphasize criteria used for selecting certain theories for application to Icelandic data Gravity, as in physics, and knowledge Interpret implications of empirical results for Iceland in an international perspective
Overview The dissertation chapters are: A Gravity Model for Exports From Iceland Determinants of Foreign Direct Investment in Iceland The Knowledge-Capital Model and Small Countries What Drives Sector Allocation of Foreign Direct Investment in Iceland?
Data Used in this Research The data used in the research is unique: Data on Exports From Iceland Statistics Iceland Panel data: 17 countries, 4 sectors,11 years, Trade blocs EFTA, EU, NAFTA and NON-bloc Data on Foreign Direct Investment in Iceland Central Bank of Iceland Panel data, with the data dimension being: 17 countries, 4 sectors and 11 years
Ch. 1: Determinants of Exports Krugman (1991) observes Iceland’s export ratio to be smaller than expected Gylfason (1999) finds it to be only 2/3 of its expected value Gravity model Trade-theoretic intuition: Export volume explained by market size and geographical location
Ch. 1: Determinants of Exports Bergstrand 1985 specification: Yi is GDP of exporting country Yj is GDP of recipient country Dij is distance between economic centers of source and host countries Aij reflects factors that aid or restrict trade between country i and j log-normally distributed error term,E(lnij) = 0
Ch. 1: Determinants of Exports The specifications applied here include Sector and trade bloc fixed effects, both individually and simultaneously Fishing, Manufacturing, Power, Other EFTA, EU, NAFTA, NON-bloc Correcting for small size of Iceland Also test some marine product subsamples
Ch. 1: Determinants of Exports Distance is found to be a barrier to exports As in Krugman (1991) Unlike Krugman Market size and wealth of recipient country more important than size and wealth of exporting country, Iceland Even when corrected for small size
Ch. 2: Determinants of FDI Trade literature: Foreign direct investment (FDI) viewed as form of trade Trade in financial capital FDI: Foreign ownership of controlling stock in a particular firm (generally 10% or more) Inward FDI sectors in Iceland Power, Com & Fin, Tel & Trans, Other Study if FDI in Iceland can be explained by location, market size, and several other factors
Ch. 2.: Determinants of FDI Trade literature: Gravity models increasingly popular for estimating FDI Brainard (1997); Mody, Razin and Sadka (2003) Fixed country and sector effects Jeon and Stone (1999) and di Mauro (2000) used countries and sectors Here also trade bloc effects
Ch. 2.: Determinants of FDI Gravity model useful in predicting FDI levels Consistent with previous literature Unlike earlier findings, wealth more important than market size Effects of market size variables (GDP, POP) often close to being equal and opposite in sign Distance negatively affects FDI, and FDI appears to be driven more by wealth effects than by market size effects
Ch. 3.: Knowledge-Capital Model Trade and investment relations of Iceland viewed in global perspective, incorporating factor endowments Knowledge-Capital (KK) model by Carr, Markusen and Maskus (CMM 2001) applied to small country case, using Icelandic data KK model incorporates both horizontal and vertical incentives for FDI
Edgeworth Box Origin for Host Country = Oh Unskilled host country skill(i)-skill(j)>0 Small host country Y(i)-Y(j)>0 Location of Iceland Ydiff=0 Sdiff=0 Skilled Labor RELATIVE SIZE RELATIVE ENDOWMENTS Skilled host country skill(i)-skill(j)<0 Large host country Y(i)-Y(j)<0 Os = Origin for Source Country Unskilled Labor
Ch. 3.: Knowledge-Capital Model Blonigen, Davies, and Head (2003) Horizontal model cannot be rejected in favor of the KK model Davies (2003) Finds support for KK model Braconier, Norbäck, and Urban (2003) Find support for CMM specification
Ch. 3.: Knowledge-Capital Model Basic KK specification and modification Thoroughly analyze knowledge effects Specification restrictions, enlarged sample, outlier omission Blonigen, Davies, and Head (2003) Davies (2003) Also proxy by education and per capita wealth Squared and cubed skill level and subsamples
Ch. 3.: Knowledge-Capital Model Driving forces for FDI in Iceland appear to be different from those in large countries Potential data difficulties when there are large differences in GDPs or population Alternatively, omission of important factor endowments such as energy or fish stock
Ch. 4.: Sector Allocation of FDI Skilled and unskilled labor may not be right endowment approach for Iceland Inward FDI Sector decomposition, resource endowments Power, Com & Fin., Tel & Transport, Other Waldkirch (2003) Also natural resources, infrastructure, pollution quotas and government stability
Ch. 4.: Sector Allocation of FDI FDI sector shares and levels FDI shares reflect relative size of each sector within a particular year of investment Brainard (1997) Inward and outward FDI share proxies separately as share of affiliate sales in total exp Slaughter (2000) FDI proxied by investment share (measured as majority-owned affiliates) in overall MNE investment
Threshold Cost and FDI FDI Market Size
Ch. 4.: Sector Allocation of FDI FDI theories assume certain threshold costs Generally not dealt with in FDI empirical models Markusen (2002): fixed cost that MNEs need to consider when undertaking FDI Heckman's (1979) two-step model to control for whether sample selection is driving results KK model seems to explain fixed costs but not the level of investment (marginal change) However, gravity model provides information on both
Levels FDI Heckman GM, 1st step
Levels FDI Heckman GM, 2nd step
Ch. 4.: Sector Allocation of FDI Results different from what was anticipated since KK model still does not perform very well for Iceland Heckman procedure application to gravity rather than KK model, since this gives better indication of how host-country characteristics affect FDI However, endowment inclusion in gravity model can be credited to KK literature
Conclusions I Geographic approach to trade theory applied to small open economies like Iceland With incorporation of gravity models Gravity forces appear to describe Iceland's exports Distance hampers trade whereas size of trade partners encourages trade Distance effect apparent not only for export Also important for FDI in Iceland
Conclusions II Reasons to prefer gravity model for FDI Also find support for the intuition behind KK model, i.e., that endowments matter Use more endowment proxies than merely the knowledge-endowment proxies Employ proxies for endowments that are crucial to Icelandic economy Useful both for explaining levels of investment and fixed costs typical of FDI models