By Robert Hickson 02/02/2020


Will the 2020’s be a case of, to borrow the words of the illustrious Martian botanist Mark Watney,  “sciencing the sh*t” out of the big problems facing us?

In some cases yes. In some cases no, because it isn’t science that is missing. In many other cases, maybe. If we can improve the system.

I suggested in the previous post that the 2020’s will see more societal contention around some scientific and technological developments. Contention isn’t necessarily bad, it is often what informs and drives science. Establishing independent fora to discuss not just the science, but societal and community values and expectations is one means of reducing, or making better use of, contention.

This post looks at some stresses on and expectations of the science system, and how it may need to change. Some of the issues were raised at a Science NZ workshop I ran last year, but these are my interpretations.


A stressed system

Midway through the last decade, biomedical researchers in the US commented that there is an “unsustainable and hypercompetitive system” and that:

“The US research community cannot continue to ignore the warning signs of a system under great stress and at risk for incipient decline.”

The same can be said for other areas of research, and in other countries, including New Zealand. Looking forward, those stresses seem likely to increase.

“Global and national research systems are in full flux as a result of a complex interplay of internal dynamics and external megatrends. The orientation, practices, actors and means of science are fast-changing.”

Some consider that the organisation and culture of science remains too self-interested (and monopolised by old white males).

“As with democratic politics, criminal justice, health care, and public education, science’s organization and culture are captured by a daunting, self-interested inertia, and a set of values reflecting a world that no longer exists.”

The commodified nature of science now does mean that many science institutions (and some scientists) do have the inertia of self-interest, and this is exacerbated by the funding environment. The view, though, that science and scientists are predominantly only interested in generating knowledge, rather than solving problems, is wrong. More and more funding is allocated to “applied” rather than “blue skies” research. That distinction though needs to be more nuanced.

On the other hand, too great an emphasis on problem solving risks missing important serendipitous discoveries.

Research funding in many countries is stagnating. There is an oversupply of PhD graduates, and research careers are being seen as less attractive.


Science as a commodity rather than a process

As government, and the private sector, rely more and more on science and technologies there are rising expectations of utilitarian value. “Frustration” can occur when research doesn’t provide the expected returns. This can discourage proposals for more novel research, and instead encourage a greater focus on “reductionism.”

There is also frustration on the lab bench side that research is being over managed, with too much focus on milestones and evaluative criteria, and too much time applying for funding rather than doing research. .

I support research having objectives and evaluative criteria, but getting too prescriptive and managerial creates the impression of predictability and risks stifling interesting lines of inquiry. Interdisciplinary research, called for more and more now, can have lower funding success than conventional research.

A sign that too many proposals are chasing limited resources is the effort some institutional research offices place on “post-production” editorial and creative design work to make proposals more stylistically appealing to review panels. Some editorial oversight is helpful to ensure clarity and consistency. Gilding what probably doesn’t need to be gilded, and polishing what possibly shouldn’t be submitted. More marketing than materiality. That’s a lot of superfluous effort for what can be random, or close to random, decision making.



The last decade saw a shift, or at least a heightened interest, toward more “evidence-based policy making”. This was, some hoped, a more rational approach to policy making and solving complex social problems. Collecting better data is great, and there are examples where an evidence-informed approach has worked. But it hasn’t worked out in many . Some researchers seeing it as “nothing more than a technocratic wish in a political world“. Politicians and businesses can be evidence-ambiguous when it doesn’t support their philosophy or objectives.

But they like the allure of certainty and predictability. The public service is now swapping their less agile “evidence-based” dance partner for what they see as more svelte and exciting “data driven” and “predictive analytics” partners.

An interesting sign of this is the UK’s Prime Minister’s adviser Dominic Cummings search for data scientists and scientific “weirdos” to work for the PM’s office and get more predictive policy making.  While exploration of new approaches and methods is encouraging, a reliance solely on data while neglecting other expertise and insights will result in poor policy.

Correlation doesn’t mean causation.  Nor does data necessarily decrease uncertainty, nor models provide unbiased perspectives and prescriptions.

Science can also be viewed as a white lab-coated knight that can be called to gallop in to save the day, or at least the morning. That’s infrequently the case.

However, with the range and scale of changes and uncertainties becoming more apparent, I anticipate more interest in “challenges” and “moonshots” in this decade. This will reflect governments (and industries) desire to be seen to be doing something substantive, as well as recognition of the need for more complex research programmes.

I wrote about benefits and limitations of moonshot thinking last year. Designed well such programmes can be effective. But establishing “missions” and “challenges” can sometimes just be a cynical rebranding of existing funding. It can divert funding away from incremental research and development that can be equally, if not more, important for creating real change.

Mission-led initiatives also often have too much emphasis on shortish-term solutionism, at the expense of spending time on understanding what the deeper causes and effects are, and not recognising that real change will require longer-term horizons.


What next?

That’s all rather gloomy. But still, good science is often being done. Not a crisis, yet. But there is plenty of room for improvement in the funding, structures and incentives for science.

The simplistic response is to increase funding, and do some superficial structural changes. Announce some “missions.”

That’s not the best response for the challenges being faced. More money flowing into a poorly designed or outdated system never leads to improvement.

Tera Allas, in a report to the UK government, described six key elements in a science and innovation system – money, talent, knowledge assets (things like intellectual property and research infrastructure), structures and incentives, the broader environment (this includes companies willing to invest in science and technology), and “innovation outputs” (revenue and exports, societal outcomes).

All of these elements need to be improved.

Some have called for science institutions to be “transformed”. Creating new organizational forms, incentive structures, and rewards.

Transformation is so tweenies. “Evolution” rather than transformation is probably the better approach in the twenties. Retaining parts and approaches that work, while trying out new initiatives, incentives and structures. Experimental rather than existential change.

The systems thinker Dana Meadows pointed out that changing structures is hard, and such change can be slow and expensive. We have all seen new structures with old incentives fail. What is often more feasible is changing the rules of the system (which includes incentives), and making the system more adaptable. Changing the rules though requires a clear understanding of goals.

I wrote in the last post about the need for change in how the science community engages with other communities. So too are changes required in relationships between government and science organisations.  Discussions between science organisations and government need to be more than about money and outputs. Form follows function. So what’s required is a better discussion about goals of the science system, and the mindsets upon which the science system is based. Then the incentives and structures can be better developed.


Change the mindset to change the game

Aspects of the changes required remind me of the mindset change that the sport and recreation sector is going through (I helped develop Sport NZ’s new strategic direction). For a long time the focus there has been on competitive sport and performance, following a flawed pathways model that promised to get results. This has not only often failed to deliver, but has been at the expense of enjoyment, accessibility, and participation.

It is always a question of balancing resources and focus. Tangible outputs are necessary, but what are the conditions that create the best culture and environment to perform well in a rapidly changing world? What will attract and retain the most talented people, and inspire others to participate (as sponsors and adopters of research)?

That’s one of the important science discussions to have early this decade.

I explored changing mindsets, and other approaches to doing science five years ago in my paper Four short science scenarios (Subscription required). Those ideas are still relevant.


Postscript (6 Feb): After reading this post Professor Sally Davenport from VUW got in touch to let me know about the Science for Technological Innovation National Science Challenge. They are taking interesting cross-disciplinary approaches that reflect some of the points I make above.  Great stuff! I had overlooked this particular challenge initiative, but will be watching what they do from now on. An overview is provided in this keynote summary from Sally.


Featured Image from Fritz Lang’s 1927 film Metropolis