larger public sectors. However, one caveat from Rodden’s study, as he himself notes, is that,
despite the use of panel data, the results are still driven by observations from some key
federal states, such as the US, Canada and Switzerland. Rodden also finds significant
evidence in favour of a ‘flypaper effect’, which is consistent with the hypothesis that
countries which rely more on central grants will have bigger public sectors.
Although one might try to infer some link between greater decentralisation and faster
productivity growth, unfortunately it is difficult to find a formal statistical link between
increasing fiscal autonomy/decentralisation and economic growth. As many other factors
play a part in determining ‘catch-up’ in productivity, it is difficult to disentangle any effect
due to greater autonomy. In an empirical study using panel data, Thiessen (2001) tries to
establish a link between the degree of fiscal decentralisation and growth. As Thiessen
recognises, the results are unlikely to be robust, given that the measure of fiscal
decentralisation is rather arbitrary; he finds an inverse U-shaped relationship, whereby an
intermediate degree of decentralisation is optimal for faster growth. This is then attributed to
the fact that excessive decentralisation may cause a loss of economies of scale in public
service provision. Given the preliminary nature of this work, it would be too simplistic to
argue that there is an optimal degree of fiscal decentralisation from the point of view of
maximising economic growth.
Reforming Block Grants
In the UK, the rationale and future of the Barnett formula in allocating funding to the
devolved authorities in Scotland, Wales and Northern Ireland has been the subject of
considerable debate. The history and complexity of the formula has been discussed at length
elsewhere (Twigger (1998); Bell (2001); Midwinter (2000); Cuthbert (2001); Muscatelli
(2001)), and will not be covered here. Instead, we will focus on the extent to which other
26