Shoogly peg to hang an economy on

THERE has been controversy about the challenge by a Scottish Parliament committee to the claim that greater tax devolution could be expected to grow Scotland's economy. If such a link were established, it would be an important plank in any case for change. So we should have a sober look at the evidence.

Theory suggests that there might be a link between fiscal decentralisation and growth, and there is an industry of economists looking for empirical evidence one way or another, using complex statistical techniques to study individual countries or make cross- country comparisons.

The first difficulty is what to measure: is it how much spending is decentralised, or how much tax freedom there is? Neither of these is as straightforward as it sounds. Spending might be more constrained than it looks, or tax powers subject to legal or constitutional limits that are hard to turn into a number.

Then there is the question of causation. If Switzerland is richer than Britain, is that because income tax is decided by the different cantons, or might something else be going on?

With these and many other complexities, it is hardly surprising that most economists would return, at best, a not-proven verdict. Certainly that was the view of the Independent Expert Group advising the Calman Commission, of which we were all members. And it was shared by some very distinguished international experts who appeared before the Scotland Bill committee recently.

But two Scottish economists - Drew Scott and Andy Hughes Hallett - think differently. Perhaps understandably, their views are welcomed by the Scottish Government. Their key assertion, in a paper that is written to promote the idea of fiscal autonomy, is that increasing fiscal devolution is likely to increase economic growth. Indeed they put a number on it. After five years each 1 per cent more fiscal devolution might be expected to grow GDP by 1.3 per cent.

This conclusion is derived from their reading of the evidence assembled by others. That evidence, they acknowledge, points in contradictory directions. But they say that detailed results for the US, Germany and China show that a 1 percentage point increase in fiscal devolution (which their paper defines as share of local expenditures in total government spending for that region) generates additions of between 0.16 per cent and 0.32 per cent to growth rates.

There is a lot wrong with this conclusion.First, it appears actually to be based not on data for several countries but on a finding in one study (not acknowledged in the paper) by two economists, Nobuo Akai and Masayo Sakata. They looked at the US only, and compared how much different states of the union decentralised down to local governments. They concluded that each 1 per cent more decentralisation in spending was associated with increases in the growth rate of 0.16-0.32 per cent. The same study concluded that tax devolution was not associated with increased growth. Hughes Hallett and Scott nevertheless use this as a basis to argue for complete tax devolution.

What's more, this single study - if it were transferable to Scotland - would be more like decentralisation to local government from Holyrood. If it could be applied to devolving from London to Edinburgh, it might suggest that the UK growth rate would go up, but does not tell us what Scotland's share would be.

They also directly cite three other authors (John Thornton, Ulraich Thiessen, Atsushi Iimi) in support of their conclusion. None of them does so. Thornton concludes that no connection with growth is shown; Thiessen thinks that the relationship may be hump-shaped - ie that there maybe an optimum level of decentralisation above which the effect may be negative. Iimi concludes that there may be a relationship between spending decentralisation and growth.

There are further difficulties with the argument made by Hughes Hallett - who was also a member of the expert group - and Scott. They take a mid-point of the Akai-Sakata range at 0.25 per cent and multiply it by five to assert that after five years GDP might be 1.3 per cent higher than it would otherwise have been. Taking a numerical midpoint is an odd thing to do: the numbers cited are in fact estimates of the effect using different methods, and should not be averaged.

They go on to suggest that this effect might not go on forever. No evidence whatsoever is cited in support of this assertion, but assuming that the effect lasts for five years enables them to claim that a 1 per cent point increase in fiscal devolution in Scotland might be expected raise GDP by 1.3 per cent after five years. Why five years ? There is no basis for that number: it could be one or 100.

But it also enables them to say (subsequently to their paper) that their conclusion is consistent with the view of such experts as Feld who have said that there is no connection between fiscal decentralistion and growth rates. They are not claiming, they appear to say, a permanent increase in the growth rate, just a one-off boost. But their whole argument is based on data about an effect on growth rates.

What are we to make of this? These authors have selected one study from the mass of contradictory evidence; they ignore the rest.From that study they take a relationship between spending devolution and growth rates which might possibly apply at a UK level and use it to argue for tax devolution to the Scottish level; they multiply that number by five to produce a headline-grabbing figure of GDP increase. On this frail foundation they erect plans for a whole new public finance system for Scotland.

None of this tells us fiscal devolution would be bad for the economy, but it may help explain why politicians get engaged in the detail of the evidence.

• Prof Anton Muscatelli is Principal of Glasgow University; Prof Clemens Fuest is research director of the Oxford University Centre for Business Taxation; Robin Boadway is Professor of Economic Theory at Queen's University, Ontario; Julia Darby is Professor of Economics at the University of Strathclyde; Peter McGregor is Professor of Economics at the University of Strathclyde and director of the Fraser of Allander Institute; Professor David Ulph is head of the Department of Economics at St Andrew's University and director of the Scottish Institute for Research in Economics