Marsden Research Fund review disappoints
As regular TechBlog readers will know, the Government funds "blue sky" (ie not immediately commercially applicable, but potentially ground-breaking) research, mostly carried out in Universities across New Zealand, through a research fund called the Marsden Fund.
In 2016, the Marsden Fund dished out $65M of research funding, increasing to around $80M per year by 2019. However only a tiny proportion of this goes to computing/IT-related research, with research in this space really not getting a fair go under the current structure (despite Government frequently stating it as an economic priority).
Last year MBIE undertook a review of the "Strategy and Management" of the fund, and ITP and others took the opportunity to point out the structural bias that exists and the impact this has on computing and IT-related research in New Zealand.
The joint submission to the review, coordinated by ITP and backed by 11 of the top tech-related bodies in the country including ITP, NZTech, NZRise, InternetNZ and others, provided evidence showing that:
- Just 17% of funding approvals for the combined maths and tech-related areas goes to tech-related research. This means around 4 maths-related research applications are approved per tech-related research application in NZ. This compared with around half and half in other countries such as Australia.
- On average the Marsden Fund funds around 1.5 research projects a year in tech fields, compared with an average of 46.5 a year by Australia's equivalent Discovery Fund.
- The tech-related research community (based on PBRF assessments) makes up around 60% of the size of the combined tech and maths research community. While the number of top (A-ranked) researchers is smaller - around 35% of the combined area - this is still double the actual proportion of funding that goes to tech research vs maths.
- The apparent bias is probably due to the structure of the fund, which combines maths and stats with tech-related areas such as Computer Science, Information Systems and Software Engineering in one panel. The panel decides which research applications progress and is generally made up of 2/3 maths and statistics researchers and only 1/3 or fewer computer science researchers.
The outcomes from MBIE's review were released this week and amongst other things, confirmed that the current panel model is susceptible to disciplinary bias (4.5.3), and cited prior research that found that "assessors may rate proposals more closely related to their own discipline more highly".
This is exactly what the evidence we, and others, submitted to the review showed. However, and somewhat inexplicably, the report then goes on to suggest the evidence is "inconclusive", albeit with the risk of disciplines [such as computing/IT research] being disadvantaged being of significant concern.
To be blunt, the evidence really isn't "inconclusive" - it's completely clear. We believe there is very clear evidence of disciplinary bias, and this is evidenced by a decade of underfunding of computing and IT-related research versus other areas, especially when compared with other countries like Australia.
While much of the report was of a very high standard, this very significant issue doesn't appear to have been considered in depth. The review team suggests yet another review, this time to look at the risk and perceptions of disciplinary bias and alternative methods to the panel structure and appointment to address these concerns, along with a number of related issues.
We don't want another review, we simply want the problem solved for our researchers.
Another review means years before this issue is resolved. We believe the reviewers should have considered the evidence of disciplinary bias now, and drawn solid conclusions rather than just recommending someone else look into it in another review.
It's now time for the Government to take action, not hold another review.
Science and Innovation Minister Paul Goldsmith has the power to resolve it right now, through the appointment of a computing and IT-specific panel within the fund. We'll be seeking a meeting with the Minister to address this urgently (and will report back in due course).
You must be logged in in order to post comments. Log In