John Toland, director of the Isaac Newton Institute for Mathematical Sciences, chaired the event and introduced the speakers. He reiterated something he told us earlier in the year when he spoke to StatsLife about the Turing Gateway. The motivation behind the Turing Gateway came in response to a sense that a lot of mathematical research is 'well known, but unfortunately not known by the people who need to know it.'
Phil Nelson, chief executive of the Engineering and Physical Sciences Research Council was first to speak. He gave an overview of the Council’s activities and the potential for their research programmes to improve public policy across government. By way of examples he cited research into supporting the elderly in their own homes, flood prediction and prevention, and the gold medal success of the British Winter Olympics team (they funded research into skeleton design.)
Muffy Calder, the chief scientific adviser for Scotland, took the podium next where she outlined the role of scientific advice to government. The role of the scientist in her view was to present robust evidence to decison makers, present advice that is based on good analysis and finally, that the importance of the last two steps should be communicated to the general public.
She then went on to present three examples of mathematics being used in Scottish public policy. The first example showed how the Bayesian INLA method was used to track breeding seabirds across the Firth of Forth. The concentration of bird traffic shown in the model has implications for policy decisions such as planning permission in the area.
Muffy then went on to detail other examples, including how the mathematical ‘gravity’ model is used to plan the Forth Replacement Crossing and the Edinburgh to Glasgow Rail Improvement Programme. Fisheries policy also benefitted from hydrodynamics which was used to map the contours of the seabed. Finally she explained how a recently developed algorithm at the University of Glasgow was used to make matches in the donor pool of the NHS kidney exchange.
Roy Anderson, director of the London Centre for Neglected Tropical Disease Research was the next speaker. His presentation detailed how mathematical models can be used to implement disease treatment programmes. He began by discussing the problem of tropical diseases in the developing world, specificaly parasitic worms in Africa, an issue the Gates Foundation is focussing attention on.
Roy then detailed how demography and epidemiology data need to be used in a mathematical model to analyse how treatment can be most effectively deployed on the ground to control or eradicate the parasites. The model is used to estimate the infection rate and where a ‘breakpoint’ is reached following treatment. The ‘breakpoint’ is what the model defines as an acceptable success rate, however any intervention must be long term if it is to prevent the parasite from returning to previous levels.
He then went on to show how the spread of global epidemics can be modelled, focussing on a simulation of a flu epidemic using airline, satellite and population concentration data. In this situation, he then detailed the difficult choices that must be presented to a minister from the model. The trouble is to communicate that out of a list of objectives to control the epidemic, not all can be accomplished and a choice in priority has to be made by the policy maker.
Roy reflected on his presentation by saying how much the use of maths in medicine had radically changed since he completed his PhD. The importance of maths will only become more important in future and he reiterated the importance of open access to methodology and datasets, if we are to better combat epidemics such as Ebola.
The next presentation was from former president of the RSS and chief scientific advisor at the Home Office, Bernard Silverman. He demonstrated how the level of modern slavery in Britain had been estimated using the multiple systems estimation method.
Extremely sensitive data from a number of official and NGO sources was analysed to see how many known slavery cases appear in each data source. Following an analysis of how the data sources correlate, an estimation (called the ‘dark figure’) of the actual number of cases of modern slavery in Britain can be made.
However, he stressed that this ‘dark figure’ is an estimate of people who are extremely difficult to track. This is an area of policy that the coalition has put a lot of effort into and Bernard praised the interest that politicians had showed in the intricacies of his method. He said ‘Karen Bradley explained the model better than I could on the BBC.’
He also asked the assembled audience which newspaper they thought reported the story most accurately. To the amusement of the room, he revealed the Daily Mail report to be the most accurate as they printed the confidence interval in the estimate of between 10,000 and 13,000 victims.
Richard Heaton, permanent secretary for the cabinet office was the last speaker. He spoke about how receptive the civil service is right now to using data and evidence in policy at the moment. This is because of the fiscal, climate change and aging population challenges facing the government today which require innovative solutions.
He went on to describe how the present is good time for statisticians in government. Moreover, there is currently a skills gap for data scientists in government departments as an increasing amount of data is opened up for analysis. As an example of this new data analysis, he talked about the social media activities of the Foreign Office, which is seen as a valuable new diplomatic tool overseas.
Hetan Shah, executive director of the RSS wrapped up the event. He spoke on behalf of the Council for the Mathematical Sciences, which the RSS is a part of. He emphasised how the mathematical sciences should not be undervalued.
He called on the next government to commit to a 10 year framework for investment in science, and to pay regard to the balance between capital and revenue spending. He expressed concern about funding leading to concentration of mathematical science departments - all universities would benefit from mathematicians and statisticians to support their scientific colleagues. Attention needs to be paid to the ‘people pipeline’ to ensure there are sufficient trained mathematical scientists to meet the needs of academia and industry. Education for the wider population was also important - whilst the Q-step programme and the new core maths qualifications are important in this regard, he stated that the mathematical sciences community needs keep pushing this up the national agenda.