One of the pillars of world-class UK social science has been its empirical research. Social scientists have used innovative data-collection methods and analysis to further greatly our understanding of social patterns.
Quantitative UK social science has a global reach: whether it is the European Social Survey, masterminded in the UK by the late Sir Roger Jowell, or the innovative analyses in the 1980s led by John Hobcraft and John McDonald that demonstrated the worldwide relationship between short birth intervals and infant mortality, our social scientists have been expert in the development and application of statistical methods to address the disciplines' challenges.
The UK is already blessed with some of the richest data in the world: the new large panel survey Understanding Society; the magnificent information collected by the Office for National Statistics; and (of course) the series of cohort studies that have enabled path-breaking research on family formation and dissolution, education and its impact on later life chances, and the challenges of an ageing society. As we move through a century in which data will be increasingly available, the opportunities for quantitative methods become compelling.
Yet amid this wealth of opportunity lurk real threats. In Society Counts, a position statement published this week, the British Academy collates the evidence, highlighting that the majority of graduates are leaving university with inadequate quantitative skills - and with little confidence in using what skills they have. This leads to a persistent, negative cycle: recent international reviews of political science and sociology have noted the relative scarcity of top-class quantitative skills in the UK.
It is also a problem for the graduates themselves. The CBI has noted that the ability to "interpret and respond to quantitative data" is a core skill for the workplace and that two-thirds of its members are concerned about their employees' inability to spot basic data errors. As our graduates compete in an increasingly challenging and international labour market, they must have the skills to succeed.
What has caused the problem? Partly it is a professional issue. As few as one in 10 academic social scientists has the necessary skills to teach basic quantitative methods. It's an institutional problem, too. The teaching of such methods is often sidelined, disconnected from the central body of social science subjects. This is the wrong approach. There is an international consensus that the best way to teach the methods is to embed them in the heart of our disciplines and apply hands-on research.
This is not to say that quantitative approaches should have primacy over other techniques. We need a broad range of skills to research and understand society, culture and human behaviour, and developing mixed methods is a continuing challenge. This is an issue not just for the social sciences but also for many in the humanities: groundbreaking work in quantitative analysis and data collection - as well as new information technology tools - has made a major contribution to disciplines such as history, linguistics and archaeology.
The quantitative skills gap must be filled. That is why I am delighted to chair a British Academy-led cross-sector group to develop a national strategy in this area. On 18 October, the Nuffield Foundation, the Economic and Social Research Council and the Higher Education Funding Council for England will announce a joint £15.5 million network of specialist quantitative methods centres. As well as creating a cohort of quantitatively skilled undergraduates, the network should also bring about a strategic shift in our universities. Wouldn't it be magnificent if it were accepted that every social scientist should be able to interpret and manipulate data, and if every undergraduate were exposed to quantitative methods?
Change will not happen overnight. We must develop an integrated strategy for action and avoid fragmented policy and practice. But it is heartening that the main funding and leadership bodies in the social sciences have prioritised quantitative skills. We must meet this deficit for the intellectual vitality of our disciplines - and for society at large.