Friday 22 July 2011

Extreme localism: local taxes and tax competition

Posted by Dr. Teemu Lyytikäinen, SERC

The government’s consultation on Local Government Finance leaves many details still to be determined. But there is one thing the document is very clear on: the government has no intention of allowing councils to set their own business tax rates. Why might that be?

It’s partly political. Localising business rates was in the Lib Dem manifesto, and both Nick Clegg and Vince Cable pushed strongly for it earlier this year. However, the Treasury and Eric Pickles apparently weren’t so keen – and neither were many business groups. The Coalition compromised on retention. And while this has been pushed as ‘setting Councils free’, the reality is that they could be much freer.

The bigger issue is that it’s not clear what the effects of full localisation would be. There are two main fears. The first is that some councils would set very high tax rates and waste the revenues on useless programmes and bureaucracy. The second fear is of the opposite scenario – local authorities undercut their competitors, with ‘tax competition’ producing a race to the bottom.

Let’s consider the evidence. Tax competition should show up in positive relationships between tax rates in neighbouring areas, results confirmed in a lot of empirical studies. But measuring these interaction effects robustly isn’t easy. Some recent SERC research [pdf] manages to do so using recent Finnish experience. Specifically, I isolate the causal impact of neighbours’ taxes (as distinct from confounding factors) by looking at responses of municipalities to forced property tax rate changes in their neighbourhood. Unlike previous studies, I find the average effect of a change in neighbours’ tax rates is close to zero.

The Finnish system may partly explain this – all local areas set taxes, but central Government takes a share of the cash to give to poorer areas. For councils that dampens incentives to outcompete the neighbours via tax cuts. (This incentive/equity trade-off will be central to the main debate about local finance reform.)

If localisation plus equalisation avoids tax competition, then some worries about a race to the bottom are groundless. But it still leaves substantial variation in local taxes – and thus, the lack of certainty the business community worry about.

And we still can’t rule out more subtle forms of tax competition. The results may mask a situation where neighbours with similar socio-economic characteristics could engage in a race to the bottom while tax rates could diverge in dissimilar neighbours. For example, two areas with low demand for social care could keep cutting taxes (and services); a high-demand neighbour receives more social care cases and has to raise its taxes even further. The resulting misallocation of resources across authorities is a big efficiency loss.

That still leaves us with the high tax and spend scenario. My research with SERC colleagues suggests this may not have been a major issue in the recent past (because councils appear to have spent increases in central grants on things that home buyers value) but it was a major worry in the years before that.


PS: For those of you interested in methodological issues: my research applies the standard spatial econometrics methods on the Finnish data and compares the findings with our more robust policy change based estimates. This comparison suggests that the standard methods are not to be trusted. More SERC thoughts on the pros and cons of spatial econometrics are here.