Parkinson on our innovation issues
Martin Parkinson: Where all the productivity growth gone and why are we even buying computers?
Martin Parkinson’s speech to CEDA’s State of the Nation conference dinner demonstrated how wise Malcolm Turnbull was to bring back to the public service someone capable of understanding so well the role innovation must play in improving our economic well-being.
Dr Parkinson was sacked as Secretary of the Treasury by Tony Abbott, but brought back to head the Department of Prime Minister and Cabinet (and the public service) by Mr Turnbull on the retirement of Dr Ian Watt.
On Monday night Dr Parkinson addressed the dinner at the Committee for Economic Development of Australia (CEDA) 2016 State of the Nation Conference A new policy agenda: Innovation, growth and the future economy.
The speech came as a welcome contribution after a day where a list of worthies had tut-tutted about the need for innovation but mostly traversed well-worn ground.
Dr Parkinson started his speech by reframing ‘innovation.’ He noted that there are four domains for innovation; innovation in technology, in business models, in public policy and service delivery.
But he emphasised that innovation isn’t just coming up with new ideas, it is the adoption of those ideas.
Unfortunately, by the end of his speech he demonstrated that we still have a long way to go in innovation in public policy; we are still calling innovation the further adoption of policy prescriptions that have already failed us.
He repeated the point by noting that many people equate innovation with startups or inventors in white coats, but that innovation is actually when ideas are taken up.
Noting that innovation is hard to measure, he rejected some of the common measures such as the number of patents or PhD students. These, in fact, can be negatives for innovation.
In a reversal of usual approaches, Dr Parkinson proposed that the best measure of how we are doing is Multi-Factor Productivity. MFP is how much growth there has been in the economy after allowing for increased utilisation of labour and capital.
It is more normal for people to say we need to increase productivity and that is why we need innovation; the Parkinson approach is we need innovation for growth and productivity – it is just how we keep score.
He then noted the global stall (often decline) in productivity improvement in advanced economies since the start of the millennium, with Australia’s performance being barely positive.
This is especially troubling because we are in the middle of the technology changes resulting in the ‘digital economy,’ which everyone has assumed will result in improved productivity.
Dr Parkinson took the assembled worthies on a short journey through four theories of why this productivity slowdown has occurred.
The first is that it is cyclical and temporary, this is a theory that productivity growth falls in long cyclical downturns. Presumably under this theory productivity will again increase as the economy grows. (There is a question about whether this is simply a tautology because in a cyclical downturn output declines, but inputs don’t).
The second is that there is just a slowing rate of global innovation. A sub-set of this view is that it is just still early in the cycle – that the big 19th century innovations of electricity and the internal combustion engine only had their impact in the first half of the twentieth century. This is a theory that there is still untapped potential in ICT.
Electric motors replaced steam in the factories where they first used a single motor driving all the machines using the pulleys and belts used previously. It took some time to put individual motors as part of the machines.
The third idea is that innovation is simply not well captured in national statistics. This is the argument that the statistics don’t include quality improvement. This could mean growth is being underestimated by 0.7 per cent.
The fourth is the possibility that while the technological frontier is expanding, the pace of take-up may be slowing down. The OECD has recently found a growing gap between how much firms are adopting new techniques – a growing ‘frontier gap.’
In the face of these factors we can either be optimists or we can be practical and consider what it is that is clogging up the system.
Dr Parkinson identified a number of possible things clogging up the system that government can act on. These include that excessive exercise of intellectual property protection restricts the diffusion of ideas, there are restrictions on access to data, that regulatory authorities have incentives to be risk averse or by allowing anti-competitive behaviour we inhibit innovation.
Government can even impose penalties on failed innovation through the way bankruptcy law operates or an inadequate provision to carry forward tax losses.
Government can and should focus on removing these impediments, but ultimately it is up to businesses to drive innovation.
In was in this context that Dr Parkinson made his well reported reference to a major international study of management practices in manufacturing which found those in Australia to be only "mid-range".
Lifting management practices in Australian manufacturers to US levels would lift productivity to 8 per cent. (As an aside, the prominent business person chairing the session later referred to this as a comment about middle-management – it wasn’t).
Having painted the bleak picture Dr Parkinson then explained that there were reasons for optimism.
Unfortunately, at this point his excellent clinical analysis of big macro-economic issues was replaced by reliance on the standard neo-classical economic toolkit.
“We know some of the key drivers of innovation. What matters most is giving people and businesses the right incentives to innovate,” Dr Parkinson said. “The more competitive a sector, the more likely businesses in that sector are likely to innovate.”
Free trade was identified as a benefit because “Australia’s relatively small market beneﬁts from producer specialisation in our exports and more domestic competition.”
For those seasoned in the ICT market these are familiar sentiments. As reported in InnovationAus.com (in reference to Telstra’s Asian ambitions), one of the stated reasons for telecommunications market liberalisation in Australia was “to enable all elements of the Australian telecommunications industry (manufacturing, services, information provision) to participate effectively in the rapidly growing Australian and world telecommunications markets.”
Bridging from the traded goods segment to non-traded sectors Dr Parkinson said “We need to create the conditions for innovation in areas like human services that will become an increasingly large share of our economy as our population ages.”
This prefaced an approving comment of the Harper review recommendation to increases contestability of these services.
It is incredibly disappointing when people with the unquestioned intellect and policy experience as Dr Parkinson can conclude a detailed analysis of how the first decade and a half has failed to deliver productivity by simply offering more of the same policy prescriptions – add competition and stir.
InnovationAus.com has reported that Innovation and Science Australia’s has decided to look outside the public service for someone to develop the 2030 Strategic Plan for Australia’s innovation system.
It is interesting to ask whether this is motivated by a concern that public service policy thinking has become narrowed to “whatever the problem, competition is the solution” or that the public service itself has been hollowed out by applying “contestability” to the function of policy development.
There isn’t a shortage of ideas worthy of exploration for adoption in public policy. Here are a few simple ideas that address the possible causes of the productivity slowdown.
We can redress inequality to stimulate growth, rather than assume that growth redresses inequality (it doesn’t).
We can identify the factors that are inhibiting productivity improvement from the application of ICT. One potential issue here is market fragmentation when the benefits from ICT accrue through scale.
For example, is telehealth better delivered in national programs or through state based programs or programs focussed on individual hospitals. The likelihood is it is the former and more ‘contestability in service delivery’ will work against the application of ICT for productivity.
We can apply more granular analysis to the growing ‘frontier gap’ and apply appropriate policy prescriptions.
On Dr Parkinson’s wider perspective, we could perhaps address the problem of Australian managerial capability rather than observe it.
Finally, we need to acknowledge that not all markets are the same. To ensure that markets really do what theory tells us they will do (lower prices and promote innovation), we need to use the theory of market design to understand what circumstances make markets successful and employ those and not just assume that contestability results in good outcomes.
For example, transaction cost economics tell you why you should integrate thinking back into the public service rather than buying it on a case by case basis from a “market.”
The theory of incentives will tell you if you want students to study STEM subjects there has to be jobs for them; you have to allow skills shortages to actually occur, not fill all the available jobs with skilled migrants on permanent or temporary visas.
Government has a very important role to play in innovation – it needs to adopt policy ideas outside the framework that have delivered our current malaise.