'Skunkworks' at the heart of government

There is the slight whiff of a bad smell emanating from the centre of the government. Hiding in parliamentary papers this week was a simple short announcement by Boris Johnson that he is taking “Responsibility for government use of data” 'in house' at No10. This unit is termed a “Skunkworks” in the advert for the position of Head. However, it is in fact the next cog in the master plan of Dominic Cummings who proposed an “alpha data science unit in Downing Street” over a year ago. There is now a clear and present danger that data might inadvertently be sought to prop up a poorly constructed belief structure based on genetic testing and a misunderstanding of the science and how scientific method works. The result could be that profound mistakes are made in planning education, social services and health that impacts social mobility and ‘levelling up’. 

The first port of call for any lobby that seeks to campaign, inform, and influence the government is the ready source of objective data. Information and openness lie at the core of democracy and accountability. It follows that a power-base that seeks to control data  could end up working counter to reasoned opposition and democracy. 

Recent developments in the control of the government’s data gathering, and associated policy, may be entirely benign, but one cannot help feeling that there is a bad smell about all of it. 

Recent No10 job adverts signal an expansion of the direct central control of data gathering and use. This is particularly evident in the advert for the Head of the newly formed No10 ‘Analytical Unit’. 

“This newly created role will be responsible for establishing No10's quantitative ability, helping to drive change across Whitehall through the establishment of a newly formed team; 10 ‘Data Science’. Sitting in No10, this unit will transform how we use data to facilitate more effective decision-making at the centre and across-Government”. 

This move appears to be the manifestation of earlier reports that Dominic Cummings was seeking “super talented weirdos” to work for him (see TEFS 3rd January 2020 ‘Social Mobility positions available: Only “Super-talented weirdos” need apply’). TEFS predicted that Cummings would “crash straight into the Civil Service ramparts manned by the current Cabinet Secretary, Mark Sedwill”. It seems the ramparts were undermined and breached, Sedwill lost and he is on the way out (BBC News 28th June 2020). The credibility and fate of his replacement will be followed keenly. 

Behind the scenes. 

Hiding in parliamentary papers is an announcement by Boris Johnson on Wednesday. Many observers may not have recognised the profound change and effects of this move. The ‘rounding up’ of data, and its use, progressed further when he announced “Government use of data’ and a major change in the way government would manage data. This effectively moved control into No10 directly for the first time. 

“Responsibility for government use of data has transferred from the Department for Digital Culture Media and Sport (DCMS) to the Cabinet Office. DCMS will retain responsibility for data policy for the economy and society. This change will help ensure that government data is used most effectively to drive policy making and service delivery. The change is effective immediately”. 

None of this should surprise anyone. The idea of creating such a team at the core of the government is solely the invention of Dominic Cummings. 

In his treatise in February 2019, ‘Genetics, genomics, predictions & ‘the Gretzky game’ — a chance for Britain to help the world’, he advocated the creation of an “alpha data science unit in Downing Street, able to plug into the best researchers around the world, and ensure that policy decisions are taken on the basis of rational thinking and good science”. Excepting scrutiny of his own reasoning, he proposed that the unit would identify “flawed reasoning and stopping bad projects, gimmicks etc”. In the context of his apparent absolute faith in genetic determinism, Cummings’s treatise is laced with many ‘pearls of wisdom’ that he casts before the elected leadership. Examples are “A useful heuristic is to throw ~100% of what you read from social scientists about ‘social mobility’ in the bin” or berating the “dominance of social scientists in Whitehall units responsible for data and evidence”. The vitriol behind “almost everything written by MPs about ‘social mobility’ is junk” is ill advised and a positively dangerous challenge to the root core of representative democracy itself. 

A skunk in the works. 

The job package for the Head of the No10 ‘Analytical Unit’ has a surprising feature. The successful candidate will be responsible for leading a ‘10 data Science’ or the ‘10DS’ team that is described as a ‘Skunkworks’. It seems to have a vision but no clearly defined mission. 

“The vision of 10ds is a skunkworks type organisation that builds innovative software to allow the PM to make data driven decisions and thereby transform government”. 

This fits with the accepted OED definition of ‘skunkworks as “An experimental laboratory or department of a company or institution, typically smaller than, and independent of, its main research division”. Such teams are not uncommon in research laboratories and industry where there is a clear mission or aim and a defined end point. Go to ‘Digression’ below for more to support this argument. I coordinated a multidisciplinary environmental science and technology research outfit in a Russell Group university for many years. The whole point of research is to understand how things work and then to innovate. But it must start with clear and very specific objectives or hypotheses. Indeed, this is the basic requirement of peer review during bids for funding from any source, whether government, industry, or charities. 

Researchers in a team work well when they have a singular and common purpose that they sign up to from the outset. In the case of ‘10DS’, it is unlikely there has been a cogent case made for its scrutiny by peer review. The taxpayer foots the bill for something that is obscure in its aims. There are two distinct possibilities. The government may not really know where it will lead, and it is speculative in assuming they will find a way to better connect and use data. The alternative, and most likely explanation, is there is a very clear mission and aim that remains hidden from view. 

Could the ‘Skunkworks' inadvertently seek to find the right data? 

There is a critical danger that the power base in the UK may become influenced and even dominated by dangerous assumptions arising from an over reliance on popular science texts. It might be then tempted, inadvertently,  into seeking out the ‘right data’ to prop up its badly constructed belief structure. Having such a unit so close to the core of government means it can be controlled directly and the government should guard against going down this route. This argument was made by TEFS earlier this year in relation to the Views of Dominic Cummings on intelligence (See TEFS 10th January 2020 ‘Genetics, Intelligence, Social Mobility and Chinese Whispers’). 

One of the most prominent and fundamental interests of Cummings is his apparent belief in the role of genetics in determining intelligence and health outcomes. 

These beliefs align with those of Boris Johnson who referred to equality with, “it is surely relevant to a conversation about equality that as many as 16 per cent of our species have an IQ below 85, while about 2 per cent have an IQ above 130” in a speech at the Centre for Policy Studies in 2013 (see TEFS 9th December 2019 ‘It’s all about equality, Brexit, the environment and the economy, not envy and greed’). He rubbed it in by extolling “the natural and God-given talent of boardroom inhabitants” to support the idea that in the economy “some measure of inequality is essential for the spirit of envy”. This seems to be inherently counter to ‘levelling up’ as an aim. 

The problem is that Cummings might be tempted to use his powerful position to seek more data to back up such beliefs. He now has a free rein to call the shots to determine the data he, and thus the government hierarchy, wants to see. In ‘Genetics, genomics, predictions & ‘the Gretzky game’ — a chance for Britain to help the world’, Cummings called for “free universal ‘SNP’ (single nucleotide polymorphism) genetic sequencing” for everyone and a polygenic score can then be calculated for various traits including intelligence. The logical outcome would be the use of genetic testing to determine what individuals might receive in terms of health care, social services, and education. But whilst such testing might inform the probability of a person in a population, for example, developing a disease in time, it does not predict that an individual will develop the disease. Similarly, it might suggest the probability of a particular IQ test result, but it doesn’t mean that this is the actual IQ of any individual. 

University and student data. 

TEFS has previously looked at how data on universities are gathered and accessed. The two organisations at the top of the pile are UCAS and HESA. However, both are not independent as they are owned by the universities themselves working as a consortium. HESA is a ‘not-for-profit’ private limited company owned by its members, Universities UK and GuildHE, and is funded by subscriptions from higher education ‘providers’. UCAS is an independent charity, funded by fees charged to applicants and to universities. They are not part of the Office for National Statistics (ONS) remit. TEFS has been critical of this arrangement with ‘Higher Education Data: Is the tail wagging the dog?’ and ‘The Social Mobility agenda in the UK - Who counts the beans?’. In contrast, The ONS has been independent of central government and ministerial control since it was formed in 1996 and it is a vital cog in the wheels of a genuine democracy. However, it is fair to say that the data is nevertheless freely available and relatively easy to access. The hope is that this free access will be enhanced and strengthened further to highlight the bottlenecks in social mobility. Restrictions on access by the government, or commercialisation of these services, could reverse this and conveniently hide declining social mobility in the UK. 

Evolution, chance, and necessity. 

In his philosophical treatise, ‘Chance and Necessity’ in 1970, Jacques Monod argued that "Everything existing in the universe is the fruit of chance and necessity." Along with Andre Lwoff and Francois Jacob, he was awarded the Nobel Prize for Medicine and Physiology in 1965 for astoundingly elegant genetic experiments that, without chemical measurements, predicted the existence of the chemical messengers that explain how genes were exerting the production and functions of proteins in cells. We have moved fast in our understanding since but still have a long way to travel. 

In Chapter 9 ‘The Kingdom and the Darkness’, Monod paints a stark picture of the human dilemma. After some very dark and uncomfortable arguments about evolution and intelligence, in the absence of any data being cited, he proposes the “dangers of genetic degradation in modern societies”. Assuming Cummings has read this work, he should guard against seeking the ‘right data’ to back up Monod’s proposition and any attempts to rectify the ‘genetic degradation’. The underlying risks and dangers of this approach cannot be understated. Monod might have expected the current wealth of genetic sequence data of individuals that we have now, but he would not have expected anyone to actually do something about it. Monod concludes that we are “alone in the universe's unfeeling immensity” out of which we “emerged only by chance” and “destiny is nowhere spelled out”. The dilemma is that at least we can choose  between “The kingdom above or the darkness below”. 

Digression. 

I digressed here because the idea of a ‘Skunkworks’ arose from a necessity and clear imperative in a time of  critical crisis. The idea of it solving more problems in other less critical times slowly took hold. But questions must remain about its utility in the development of novel data methodology for use by the government. My experience is that university research teams, at least in laboratory science, often have the characteristics of a ‘Skunkworks’ team racing to get ahead in science and innovation. It only works if there is a clear aim and objective. University teams should be totally independent of any political control to maintain their objectivity. Blurring the edges of such independence has far reaching consequences. Placing such a ‘Skunkworks’ at the heart of government under political control, instead of in a free university setting, leaves no doubt about the potential dangers. 

The term ‘Skunkworks’ originated from the Lockheed team that rushed out the XP-80 jet fighter plane in the USA in 1943. This was achieved from ‘design to flight’ in only 143 days. However, it was not a novel ‘invention’ since Italy, the UK and Germany already had working jet aircraft. The XP-80 was really an airframe around the powerful British Goblin jet engine , built be de Havilland in Hatfield and a step up based on the Whittle W1 jet engine, named after Frank Whittle from Coventry who patented a working the turbo jet in 1930. The W1 powered the Gloster Whittle G40 that was flying in 1941 (all this is evidence I was mad about all types of engines before discovering cellular biochemical mechanisms in a university textbook at the age of fourteen).

Mike Larkin, retired from Queen's University Belfast after 37 years teaching Microbiology, Biochemistry and Genetics. He has served on the Senate and Finance and planning committee of a Russell Group University.