NISA critique is way too harsh
#IdeasBoom: Remember when the National Innovation and Science Agenda used to be good fun?
The Auditor-General’s recent review of the design and monitoring of the National Innovation and Science Agenda risks throwing out the outcome baby with the process bathwater.
Industry development policy in relation to the information technology sector has long been vexed. In fact, in this country it has been ignored, something the ambitious but modestly-funded NISA program was set up to remedy.
The problem here is that a review of NISA design and monitoring has spilled over into commentary about performance. And any reasonable reading of the initiatives would suggest it is way too early to make any judgement on the specific economic impact of the agenda.
What we can say, however, is that the signs in key areas of the agenda are exceptionally positive. Where NISA applied targeted policy to address specific ecosystem blockages, the progress has been outstanding. The point is that it is progress, not an outcome.
For the most part, the Australian National Audit Office report released a week ago was unremarkable. It found some processes in its design and monitoring were not perfect. It made recommendations for improvement, which government largely accepted.
But it also unfairly shone a spotlight on the notion that the NISA investment was “too small” to measure, that the policy was developed too quickly, and that it did not adequately define its broad-stroke policy headings like “Culture and Capital”.
With this commentary the Auditor-General has over-stepped his purpose, particularly putting an emphasis on the notion “too small.” How would he know, when it is still way too early to measure impact?
It is worth stepping through some history here. First, anyone working in the industry a couple of years ago does not need an audit report to remember how cold and bleak the outlook for the sector was under former Prime Minister Tony Abbott.
At that time, there was no coordinated plan to recognise in policy the dramatic shifts in technology that were evident to everyone but the political leadership of Mr Abbott and his Treasurer Joe Hockey.
Secondly, the NISA was set up as an exceptionally modest scheme – $1.1 billion over four years – to address specific shortcomings identified in multiple reviews (including the excellent issues paper written by University of Technology Sydney Professor Roy Green for the Senate Inquiry into the Innovation Ecosystem published in late 2015.)
The NISA was always intended as a first iteration of an agenda that would start small, try new things and change tack where required.
It is worth noting some of these first-tranche initiatives. Anyone who was alive in the sector in 2015 was saying limited access to capital was a huge inhibitor to growth among startups.
In response, the NISA introduced a tax incentive scheme for angel investors and for Early Stage Venture Capital Limited Partnerships (ESVCLPs) at an anticipated cost of $106 million.
As a result of these schemes, it is understood the total amount invested in “early stage innovation companies” has amounted to more than $280 million. Investments have been made into around 320 early stage companies by about 3100 investors.
By any yardstick this is exceptional, especially against the backdrop of the early stage capital drought of 2015. But it’s not an outcome, it is just an important marker of progress.
The point is that it is far too early to measure success of the program within the context of industry development – that won’t come until we can see the success or otherwise of the 320 investee companies.
The same can be said for a raft of the smaller initiatives in the program. Funding for getting the ON accelerator program inside the CSIRO was modest and its key metrics somewhat ephemeral being a program of cultural change inside the research agency.
The ON programs have attracted overwhelming internal interest far outstripping any expectation. But this is a proof-point of policy progress rather than a measure of the program’s success. It is far too early for that.
The NISA program re-funded $75 million in funding to Data61 that had been cut by Tony Abbott and Joe Hockey in their shocker 2014 budget. I don’t see anyone, anywhere, arguing for less data science.
Data61 is building data capability across government and industry. It has specific development KPIs. It is by far the most interesting division within the CSIRO.
And despite the CSIRO doing its accidental best to starve Data61 of oxygen through institutional inertia, it is making strong progress towards commercialisation and capability-building goals.
The Landing Pads, which are five overseas outposts set up to enable Australian startups to get exposure (and a toe-hold) into foreign markets, was hardly non-controversial.
It was a modest investment, but after only a year it is again impossible to measure outcomes.
It is hard to even know what to say about the big ticket items, like measuring the outcomes of the NISA investment in Australia’s Quantum Computing research capability, or its funding boost for neglected science infrastructure investment.
A national innovation ecosystem is a complex, living tangle of research endeavour, entrepreneurial effort, and vested interests.
In Australia, the criticisms of our industry development policy in the tech sector has been about inaction, or for being too slow.
Governments have been routinely criticised for conducting endless reviews, but then not acting on them.
Now, through the Auditor-General, we have a government being criticised for acting too quickly and for relying too much on the existing catalogue of reviews and outside consultations that had already taken place.
And instead of criticising government for spending “too big” and pissing money away – which is the usual refrain – the NISA is unfairly criticised for spending “too small” to be measurable across the economy.
A full reading of the ANAO report will tell you it’s unremarkable. But it strays into making judgements about outcome that are not supportable.
The bulk of the NISA program was legislated from July 1 last year, just 15 months ago. It is simply too early to rush to judgement.