I think that we have spent way too much time on differentiated final goods, and neglected trade in intermediates … intermediates-inputs approach seems empirically very relevant, and formal econometric work would be very welcome.
Contrary to popular hype, the vast bulk of intermediates trade – that is, outsourcing – is among developed countries, not between developed and developing countries. This is consistent with Jabbour (2007), who showed in an extensive empirical analysis of 4,305 French firms (using survey data) that the vast bulk import their intermediate inputs from other developed economies through arm's-length transactions. Consequently, most intermediates trade is intra-industry (and likely “Ethier-type” intermediates trade). Because of the previous absence of a comprehensive dataset on intermediates and final goods trade flows, econometric analysis of the determinants of intermediates trade volumes/values is virtually non-existent, as our quote from Markusen (2001) suggests. Egger and Egger (2005) provide one of only two empirical (gravity) analyses of a narrow aspect of outsourcing trade flows – bilateral “processing” trade among twelve European Union economies by national and multinational enterprises. The other empirical study is Baldone et al. (2002). Aside from these two empirical analyses, the absence of systematic intermediates versus final goods trade data has confined many researchers of outsourcing to employing numerical simulations to study final and intermediates trade volumes (see Baier and Bergstrand 2000 and Yi 2003).