The Guardian, rightly, sought contrary opinion when reporting my new report for the TUC on corporation tax cuts.
As the Guardian reports:
Mike Devereux, director of the Oxford University Centre for Business Taxation, said the report did not take into account all the variables that might affect growth rates in order to establish whether corporation tax affected investment.
"The evidence for whether corporation tax affects growth in this report is weak, to say the least," he added. "As the report acknowledges, there are all kinds of things which may affect growth rates. Looking at a correlation means you are not controlling for any other differences. There are all kinds of reasons why growth rates might differ."
It's anyone's guess what Mike means when saying that, but the report sought to add clarity:
The TUC study contradicts a detailed analysis conducted by the OECD. Published in November 2010, it found that corporate taxes were the most harmful type of tax for economic growth.
Devereux suggested the research itself was not definitive, however, but added: "To say corporation tax does not affect growth is just ignoring quite a lot of academic literature."
Mike's trouble is I did not say it did not affect growth: read the report and it clearly says there is a link between corporation tax and growth. I did not deny it: I do not deny it.
But what the report did do was assess the significance of that link. As the report says:
It transpires that analysis of the correlation between tax rates and growth in OECD countries (excluding the top and bottom outliers) finds that at best the relationship between the two variables is weak, with the r2 coefficient less than 7%.
In English that means that at most 7% of growth differences can be explained by differences in tax rates. As the report puts it:
Tax rate differentials of between 27% and 40% over a period of 14 years are clustered so weakly around growth rates that these growth rates only vary between 1.9% and 2.3% per annum as a result.
The relationship of changing tax rates over time (which is what the UK government is proposing to do) and growth is weak based on this data. The linkage between the two as suggested by the resulting correlation coefficient that might reasonably be expected to apply to the UK suggest that over 90% of growth in this range is explained by factors other than tax. In that case cutting tax rates to stimulate growth appears a poor choice of economic policy.
That's the message of this report.
Devereux's work, we should remember is (not just in my opinion) fundamentally unreliable because of failures to mention conflicts of interest, enough for the Times Higher Education supplement to have reported on the issue. His work is funded, amongst others by the FTSE 100 group of finance directors to the tune of at least £5 million. It's a fact Devereux rarely discloses. But it does help explain why he and his colleagues work so hard for the abolition of corporation tax whenever they can, including by heavily influencing the Institute for Fiscal Studies' Mirrlees Report on this issue.
Now I'm not for a minute claiming my work is unbiased: I'm just saying that I am openly acknowledging my bias. Those biases have significance. What Devereux is saying is that if all other factors are taken out of account, as his supposedly objective academic work does, then there is a proven link between cuts in corporation tax and growth. His work is designed to influence policy: he might like to pretend it isn't but he never objects when it is used for that purpose. I have to conclude his work is politically motivated as a result, with intent to lower corporation tax rates or to see the entire tax abolished, with the burden being effectively added to VAT (his chosen option, which would increase the VAT rate to over 30%).
But my point is that Devereux's choice to 'control' for those other factors when coming to this recommendation ignores the fact that my findings suggest that those other factors explain 93% of growth and changes in corporation tax in countries comparable to the UK explain just 7%.
So, when making policy, and deciding how to allocate scarce resources would any rational, objective person, use corporation tax to stimulate growth when it is apparent that this has weak links with growth and that its impact is at best highly marginal or would you instead go off and look at and invest in the other factors that encourage growth?
The right choice is very obviously to look to recreate growth using other mechanisms.
Deveruex though, by 'controlling' for these other factors seeks to remove them from consideration. He makes the choice 'do you cut taxes, or not?'. Well if that was the only option then you might cut taxes. What my work shows is that Devereux asks the wrong question, uses statistics badly and as a result comes to the wrong answer, which is inevitable when your political blinkers mean you ask the wrong question, which is what I think he's doing. But that's the risk resulting from accepting sponsorship from big business. Again, there's no problem with accepting sponsorship: I clearly have. But why doesn't Devereux ever ensure that this bias is recorded when he is claiming to make academic objective comment which in my opinion is nothing of the sort?