In this post, I will continue my discussion of Philip McCann’s paper, ’Economic geography, globalisation and New Zealand’s productivity paradox’ . McCann argues that it is New Zealand’s economic geography that is the reason for its poor productivity performance. In this post I’ll try to sketch some of the underlying ideas from economic geography that McCann utilises.
There is a general perception that globalisation is levelling the world economy (see Thomas Friedman’s book ’The World is Flat’). The outsourcing of manufacturing from Europe and the US to parts of Asia certainly receives a lot of attention in the media – even I have blogged about it. However, McCann argues that the world is not becoming as flat as we might think. In fact, over the last two decades, he notes that the share of global output, global trade, global foreign direct investment and global R&D, of the three super-regions (the EU, NAFTA and South and East Asia) has grown.
How does McCann reconcile this increasing regionalisation with the idea of globalisation? The key to McCann’s argument is the relationship between the agglomeration economies (or economies of scale) and the spatial transaction costs involved in trade and manufacturing.
The idea that economies of scale might be important for productivity goes back at least as far as Adam Smith. In his Inquiry into the Nature and Causes of the Wealth of Nations, Smith observed that division of labour offers advantages in a manufacturing process. Using the example of a pin factory, he noted that the larger the factory, the more a factory could specialise tasks in pin manufacture among its workers. Smith argued that if one doubles the size of the factory, the increased opportunities for specialisation meant that the factory was bound to more than double its output. In other words, the larger the pin factory, the higher the productivity of its workers.
McCann suggests that it is high-value added or knowledge intensive manufacturing that benefit most from this type of specialisation and the division of labour. Therefore, it is this type of manufacturing that will exhibit economies of scale.
If this is the case, economists have shown that if there are high transportation costs or trade barriers that lead to high spatial transaction costs for trade, then all countries and regions will have similar production patterns and similar levels of productivity. For example, every country will manufacture its own cars as it is too costly to ship them across national borders.
If, however, transaction costs fall, then knowledge intensive manufacturing that exhibits economies of scale will become localised in particular regions. These regions will exhibit high levels of productivity. For example, the large number of high technology workers in Silicon Valley gives it an advantage in productivity that other regions can’t match. Only when spatial transaction costs for generating new knowledge fall to close to zero will this regionalisation disappear and productivity return to being spatially homogenous. This leads to an inverted U-shaped relationship between agglomeration and spatial transaction costs.
McCann argues that while it’s true spatial transaction costs for low value, low knowledge intensive manufactured goods have fallen dramatically over recent decades, the spatial transaction costs for high value, knowledge intensive activities have increased:
This is because of the increasing importance of timeliness, speed, variety, customisation, and service-quality, in all high knowledge intensive forms of production and service delivery (Disdier & Head, 2008; Duranton & Storper, 2007; McCann, 2007). The principal reason for this is that the premium associated with face-to-face contact in high knowledge intensive activities appears to have increased (Gaspar & Glaeser, 1998; McCann, 2008; Storper & Venables, 2004), because the spatial transactions costs on the inputs side of the production process, rather than the outputs side, have increased for high value goods (McCann, 2008).
It is this increase in transaction costs for knowledge intensive activities that has led to agglomeration of research and development in places like Silicon Valley and the productivity advantage that large cities enjoy today. Spatial transaction costs for knowledge intensive activities are sitting near the top of the inverted-U, while the costs for commodity manufacturing have climbed it and are heading down the other side. So the production of commodities may have become globalised, but the generated and application of new knowledge has not. What does this mean for New Zealand?
 McCann, P. (2009). Economic geography, globalisation and New Zealand’s productivity paradox New Zealand Economic Papers, 43 (3), 279-314 DOI: 10.1080/00779950903308794