Herewith my column for Today’s Fin on the Government’s proposed new R&D Tax Credit. The paper on which it is based is on the Lateral Economics Website.
The politics of compromise can work to solve problems by taking everyone’s needs into account. But sometimes we just get caught up in pacifying those reluctant to part with existing entitlements. As Parliament debates the Government’s proposed changes to the tax concession for Business Expenditure on Research and Development (BERD), it should be wary of buying off those seeking to hang onto their BERD in the hand.
As I concluded in a recent study for the Australian Business Foundation, The BERD in the hand: Supporting Business Investment in Research and Development, if forced to choose, I’d abolish the current tax concession for R&D rather than maintain it. Why? When it was introduced in the mid 1980s it provided 24.5 cents assistance for every dollar of BERD. Today, with large reductions in company tax rate and the headline rate of assistance – from 150 to 125 per cent – it’s worth just 7.5 cents in the dollar.
So it has very little effect on business decision making – that the concession is simply a windfall on over nine tenths of R&D because it would have been done anyway. So the existing scheme probably generates more administrative, compliance and revenue costs than it does R&D benefits.
The scheme is also complex with multiple sub-schemes. And though the rate of assistance is low, some businesses have claimed a large share of their total production activities as R&D related. This sounds fair enough, but the only way it’s been affordable is via the atrophy of the rate of assistance. Both Canada and the UK can afford much more generous rates of R&D assistance, but only by ruling out production activities altogether.
And as the Cutler Report highlighted in 2008, one of the main beneficiaries of our laxity on production related R&D is mining. Some mines involving novel technical challenges have claimed the majority of their expenses as R&D related!
Fortunately there’s a better option than abolishing the scheme – improving the current scheme. The changes the Government proposes ramp are consistent with the Cutler Review recommendations. They ramp up the rate of assistance to R&D – from 7.5 to 10 and 15 per cent for large and small firms respectively – and this is funded by abolishing several complex sub-schemes within the existing concession and by making production activity ineligible for assistance unless it was undertaken for the “dominant purpose” of supporting R&D.
Not surprisingly some firms will lose from the new definition. And they’re not all responsible for the so called ‘whole of mine’ claims we were concerned about on the Cutler panel. For instance Cochlear, the Sydney based bionic ear supplier thinks it will lose from the new arrangements.
So I explored the issue with an open mind, but ended up impressed with the way the Treasury proposed dealing with the problem of excessive claims on production. The new scheme distinguishes production undertaken with the dominant purpose of supporting R&D (which remains eligible) from other production which may nevertheless have some direct relation with R&D (which becomes ineligible).
Because the more stringent tests on production create losers, political debate and possible compromise has focused on softening this feature of the new arrangements for instance by capping claims on production or allowing incidental production for small firms but not large ones. But if those seeking amendments really want to maximise their policy impact on R&D they shouldn’t hunt their quarry here. Pretty obviously, production that isn’t for the dominant purpose of supporting R&D would take place for its own (production based) reasons, with or without R&D assistance.
But there’s one proven way policy can increase R&D. When first introduced, the tax concession operated solely by reducing company tax – which meant that companies tax loss, as many R&D intensive firms are – could not benefit from the accumulated tax losses, often for many years.
In 2001, though the rate of assistance remained measly, smaller firms in tax loss were given access to their concession as a cash grant. Their R&D responded strongly showing that cash flow was constraining an otherwise healthy desire to increase R&D. So if we want to make a difference getting small firms their money sooner could have a substantial impact. And it would be cheap for governments because it has very low capital costs.
But whatever we do, we should avoid further delay. If the new scheme is generous enough to encourage more when it’s legislated, it’s enough to have at least some firms biding their time until our hung parliament decides what it will do.
And where does software sit? In production or R&D.
Good question. Long story. This is from the Explanatory Memorandum.
They didn’t follow our recommendation to favour open source software – on obvious positive externalities grounds. I don’t think they even considered it. Too unusual. Didn’t tick enough boxes – and indeed was outside the box. In fact the box that it was outside of may even have stretched the envelope.
I have always preferred a bounty system to an expenses write-off system.
Under a bounty system, if the mining industry wants a better metal detector then each company fronts up with some money — $10 million here, $10 million there and the government multiplies that by some factor (let’s say they double it) and they put together a specification of what they want. First inventor who can demonstrate a working device collects the bounty and in exchange hands over all the patents and other IP which eventually becomes open to the public (possibly after some years of exclusive use to payback the original investors).
The bounty system does not require secrecy, it gets maximum leverage out of government’s useful role as a facilitator rather than a regulator, and you can go back to the taxpayers with a clear itemised list of what you bought and what they paid for it.
From reading the Lateral Economics report (admittedly not in great depth), all of the original arguments explaining the lack on incentive for any single firm to invest in R&D when it is easier to just wait for someone else to do the work first, also represent arguments why no government should ever invest in R&D (unless we someday achieve world government). You are right to bang on about the uncertainty evolving from the whole secrecy of the system (although perhaps from some people’s perspective the secrecy is a good thing, seems to be the way lots of government operations are heading these days). There’s an even bigger uncertainty in the current system that we have no idea what the genuinely useful outcomes of the R&D really are.
By the way, it’s weird to me that inhouse software development is not counted as R&D on the basis that it would probably have been done anyhow, but the example given for Boulevard Mining using wider tunnels does count as R&D when clearly mining companies are always pushing for wider tunnels. I suspect that the current system is riddled with such arbitrary decisions, and a bounty system would at the very least bring some public scrutiny on the subject, but probably also do a better job of regulating itself.
Lastly, it somehow sounds more sensible to pay for a successful outcome rather than paying someone to give the appearance of striving towards an outcome. I suspect that the average non-economist would agree with me on that.
Thanks for your thoughts Tel (better than your thoughts on Krugman ;) (But I digress).
There is no doubt some pragmatism in the way the system picks and chooses what it supports. Funding software development for large firms like banks looks like throwing good money after bad as far as ‘additionality’ is concerned. In fact there’s not much evidence that large firms are responsive to the incentives in any event, and for small firms it may just be the cash flow effect that leads to higher R&D.
we used the R&D grant, and notwithstanding some initial troubles, it had a big impact on our business. have not read any of this yet. will do, hopefully soon.
Here’s the interview of the paper.