Does decentralisation foster regional GDP convergence?

by Hansjörg Blöchliger, Senior Economist, Policy Studies Branch, Economics Department

The growth pattern of OECD countries and their sub-national entities is puzzling. Between-country differences in GDP per capita are declining, yet the differences across jurisdictions within those countries tend to rise. Put in other words, countries’ GDP converges, while the output of their sub-national jurisdictions tends to diverge (Figure 1). Today differences across countries are smaller than, on average, within a country, which is quite different from the situation 20 years ago.

What could explain the puzzle? Agglomeration economies and the unequal geographical impact of globalisation could play a role. Trade and other forms of international exchange are dominated by firms that tend to be located in large agglomerations. Convergence of countries could hence be driven by the convergence of agglomerations that are well-integrated into global value chains. The picture looks different if one looks at a single country. When productivity is underpinned by agglomeration forces, regions with large agglomerations will pull ahead of regions without, and differences in growth rates become self-propelling. As a result, countries converge while regions diverge. Indeed, decomposing GDP suggests that differences in sub-national productivity are the main driver of GDP per capita disparities.

Yet rising GDP disparities are no fate. They depend on a country’s intergovernmental framework and can be tackled by policy. Assigning more fiscal power to the sub-national level – e.g. for spending on education or infrastructure – can contribute to a more balanced development across a country. According to new OECD research, GDP disparities tend to be smaller in more fiscally decentralised countries, and tend to grow more slowly or even to decline there. The type of fiscal decentralisation matters: Assigning more taxing power to the sub-national level underpins convergence, while a large intergovernmental transfer system has the opposite effect. Interestingly, catching-up regions tend to benefit more from decentralisation than the frontier regions. They appear to adopt policy innovations more rapidly and those seem to have a stronger impact. Conversely, intergovernmental grants tend to fuel disparities, probably because they do not provide lagging regions with the right incentives to catch up. The channel from decentralisation to convergence has yet to be investigated; i.e. whether productivity trickles down faster or whether capital and labour moves more swiftly across jurisdictional borders.

Further reading:

Blöchliger, H., D. Bartolini and S. Stossberg  (2016), “Does Fiscal Decentralisation Foster Regional Convergence?“, OECD Economic Policy Papers, No. 17, OECD Publishing, Paris.
DOI: http://dx.doi.org/10.1787/5jlr3c1vcqmr-en