In support of mandatory data
Data collection: Is it time for better collection of industry data to support policy design?
It must be getting close to time where we collectively reassess the way the nation surveys the growth or otherwise of the tech sector to ensure Australians are getting value for money for the industry support measures they invest.
The federal government spends $9 billion a year on direct and indirect research and development support, and many hundreds of millions of dollars more in other industry development programs targeted at the tech sector.
And then there are the state governments, each scrapping for access over small Australia’s relatively small pools of capital and talent. The states are spending more on industry development for tech sector – in building new companies and attracting multinationals – than ever before.
Which is a good thing. This increased attention gets Two Thumbs Up and a Five Star rating. Bravo, bravo. (This line is insurance against being burned at the stake as a heretic for suggesting the trough may overfloweth.)
But how are we measuring the success or otherwise of the investment?
October has become the survey season of the Australian startup sector. In the past week, the well-regarded Startup Muster national survey released its 2018 results, followed by the LaunchVic survey of the Victorian startup scene and then the EY and FinTech Australia survey of the FinTech sector.
In key measures, there are starkly different results. A cynic might suggest the results of each reflect a vested interest bias. In fact you don’t need to be cynical to suggest this. If the industry itself is suggesting there are “pub test” problems – as Scott Handsaker has, among others – then that’s a problem.
I have suggested previously that a mandatory collection of data from companies that enjoy the support of government programs is something that should be seriously considered, to give policy-makers a more complete picture of both the health of the industry and whether government support programs are effective or need a tweak.
Government should manage this data collection, and it should be a compulsory obligation for companies that are recipients of government support to provide this data.
For example, if you claim against the R&D tax incentive, then you would be obligated to provide a basic data set to this national survey.
Similarly if you receive a commercialisation grant, or a low interest loan, or if you are a tenant in a rent subsidised workspace, or are within a accelerator/incubator that is supported by government.
This should include the VC sector and angel investors, also on the basis of the tax incentives they receive.
This mandatory collection would provide a snapshot of basic industry data. Which industry, how many employees, revenue etc.
The Australian Taxation Office might be the best collection point for the survey data, which could then be managed by the Australian Bureau of Statistics, with perhaps the findings and publication handled by the Industry department.
Surely it is possible to design a digital service that is not too onerous?
The results would obviously inform policy. But it should also inform the industry and be transparent about the definitions under which the classifications of companies are made, and the publishing of results.
With the launch of the Startup Muster report, Twitter was alive to theories about why the startup sector was in decline in this country. Some speculated that the sector is suffering from “survey fatigue”.
While this does seem a bit precious given these companies are happy to receive the support, the mandatory data collection would solve this problem. One data collection point a year. Sounds pretty good.
This mandatory data collection cannot be about ‘startups’ alone, although this might be a subset of a much broader program.
It would hardly be a novel idea that the recipients of a taxpayer-funded program have mandatory obligations to provide data under a reporting requirements regime. Newstart allowance recipients are very familiar with their obligations, for example.