Print Friendly and PDF
only search

Evidence-based policy and democracy

About the author
William Davies is a Senior Lecturer at Goldsmiths, University of London, where he is Director of the Political Economy Research Centre. His weblog is at and his new book is The Happiness Industry: How the government & big business sold us wellbeing (published by Verso).

In September 2005, Goldsmiths College London hosted an intriguing conference entitled “Media Policy and Power” that brought together regulators, pressure groups, broadcasters and academics to discuss the formation of media policy in the United Kingdom. Questions of accountability and the public interest were central themes, while the cast of characters was a familiar one, with Rupert Murdoch often cast as the villain. But amidst all of this, one of the most telling comments of the day seemed to pass by entirely unnoticed.

A panel had been debating the formation of the UK’s 2003 Communications Act, the greatest overhaul of media policy in Britain for well over a decade. Speaking in defence of this process was Bill Bush, one of the most senior government advisors to have been involved in the legislation. An audience member asked Bush to comment on the rumour that Tony Blair had personally intervened in the drafting of the legislation to allow for greater foreign ownership of the media, on the personal bequest of Murdoch.

Also in openDemocracy on how governments might learn to make good decisions, three articles published in June 2003 in our “Think-tanks, politics, ideas” debate:

Anthony Barnett, “A new way for British government?

Geoff Mulgan, “Global comparisons in policy-making: the view from the centre

Tom Bentley, “Governance as learning: the challenge of democracy

If you find this material valuable please consider supporting openDemocracy by sending us a donation so that we can continue our work for democratic dialogue

Bush was ready for this one. Blair had indeed become personally involved, he replied, but not on instigation of the media mogul. Instead, the minister responsible for the legislation had chosen to consult Blair on the matter, because she wanted the highest possible endorsement for such a sensitive area of policy. She wanted to know where Blair stood on the issue. And so what was the prime minister’s position? He didn’t have one, Bush went on breezily – he simply told her to go wherever the evidence pointed…

Bush’s implication was clear. Being firmly on the side of evidence-based policy, the prime minister couldn’t be further from cronyism or sleaze. Evidence was a safeguard against vested interests, the one part of our policy-making process that Murdoch’s News Corp would be unable to influence. Yet from a subtly different perspective, Bush was saying something really quite odd. In his defence of Blair, he was hoping to reassure us that one of the most moralistic politicians in living memory – a man capable of defending a war largely on the basis that he “believed it was right” – didn’t have a view on how the media ought to be controlled. Regardless of how the Communications Act was actually drafted, the fact that such levels of neutrality are now venerated in public life is in itself significant.

The primacy of evidence has been an important part of British politics throughout New Labour’s period of office, the necessary accompaniment to the “what matters is what works” mantra that ministers used to keep ideology at bay in the early years following their election in 1997.

In 1999, a government white paper entitled Modernising Government stated that “this government expects more of policy makers. More new ideas, more willingness to question inherited ways of doing things, better use of evidence and research in policy making and better focus on policies that will deliver long-term goals.” The preferred model of evidence-based policy development runs as follows. A problem (often a market failure) is identified using existing social-scientific evidence; a potential government intervention is dreamt up or copied from overseas; the intervention is applied experimentally as a pilot; outcomes are assessed, and depending on its success, a decision is taken on whether to “roll out” the policy across the country.

The merits of this approach are obvious. It ought, in principle, to prevent governments from pouring large quantities of public money into initiatives that don’t ‘work’, whatever that might mean. One of New Labour’s most celebrated policy initiatives, Sure Start, was recently found to be having no measurable impact on the development of the children that it was established to help. Clearly, this raises serious questions about the future of the scheme. There is also something quasi-democratic about evidence-based policy, in that quantitative social-scientific methods are dedicated to eradicating class, gender or racial biases from their representation of society. The dataset may be a fairer way of representing the population than the ballot box, given what we know about who votes and who doesn’t.

A principle of democracy

Given the vindictive way that much social policy was constructed under the Conservative party rule of 1979-87, such a rapid ascent for evidence-based policy can be cause for some celebration. But there is equally reason to fear that it could yet become a victim of its own success, over-reaching its legitimate limits, and eventually imploding. To understand why, it’s worth returning to a remark made by the economist Andrew Graham at the 2003 Oxford Media Convention. Criticising policy strategies that attempt to measure the quality of broadcasting, Graham said: “if you could measure quality, it would just be quantity”. When measurement becomes too pervasive, qualitative and moral values are left out in the cold, not only impoverishing politics, but risking a backlash that will imperil evidence-based policy itself. There are two ways in which this is already happening.

The first is what might be called an aestheticisation of evidence, which occurs at the messy interface between social science and the media. Part of the purpose of evidence-based policy is to keep governments honest, and to create a terrain in which spin has no purchase. But ironically, the media have become rather fond of quantitative social science, inasmuch as it supplies a steady stream of news stories. For instance, the story about the Sure Start evaluation made it on to the Guardian’s front page. In response, the release of new evidence is politicised, as demonstrated by the government’s decision to release data on illegal immigration immediately after the May 2005 election.

Meanwhile, smaller and less wealthy institutions such as charities and think tanks will often cut corners in their pursuit of an eye-catching number. Knowing full well that newspaper editors are uninterested in the intricacies of methodology, one smart tactic is to release evidence from a focus group in percentage form (“x% of people believe…”), despite the fact that a focus group can never be representative of a large population.

As the public becomes increasingly numb to techniques of media manipulation, the sad fact is that the collection of evidence ends up failing to achieve one of its key goals, namely the description of a shared and objective social reality. After all, who bothers to check whether the numbers contained in a news story have Economic and Social Research Council endorsement or not?

The second and more profound danger of excessive recourse to evidence in public life, however, is its antagonistic relationship with democracy. In an evidence-obsessed world, both politicians and public are able to renounce their responsibilities to the political process. Politicians are able to duck the normative question of how they believe society ought to be, while the public no longer needs to engage with the mechanisms of democracy in order to convey where its interest or demands lie. While politicians hide behind their number-crunchers and researchers, the public interest is divined via social research and the public only discovers it via the media. It is not just ideology that evidence-based policy condemns to exile; too often it is morality.

A process of value

There is an important dimension to democracy that the evidence-based policy movement appears to miss. Democracy is not just about the desirability or otherwise of outcomes, but about the mechanisms used to select and achieve them. The government, bewildered that measured improvements in public services have not been noticed by the public, and that measured falls in crime have been accompanied by increased fears of crime, then tends to blame the media for excessive pessimism. But when policy is constructed and evaluated by anonymous statisticians, one cannot assume the public will share the government’s assessment of it. In politics, it is not enough for something simply to be the best option; people must reach agreement that it is the best option, a process which then becomes constitutive of that option’s value.

William Davies was a senior research fellow at the Institute for Public Policy Research (ippr). He is now studying for a PhD in sociology at Goldsmiths College, London. His weblog is here.

Also by William Davies in openDemocracy:

The age of surveillance: a new ‘dotcom boom’?” (August 2005)

There is one very simple recommendation that emerges from this critique: proportional representation. If the House of Commons were a space in which the public of the United Kingdom were even remotely represented, there would be no need to resort constantly to evidence. It is no wonder that the government employs so many statisticians and researchers, all frantically controlling their data to ensure validity and generalisability, when there is no democratic mechanism available to work out what this public values. Given how little constitutional influence the public has over policy in the UK, we should perhaps be grateful that New Labour chooses to try and work out what’s best for us. But it would be preferable if that were our democratic right, rather than just the government’s cultural whim.

There is a final irony in all of this. In many instances, Britain’s policy-making process has become strikingly inefficient. Where the public (quite rightly) feels that its political choices don’t matter, and the government (quite rightly) feels that it doesn’t have a mandate to take normative positions on anything other than foreign policy, millions of pounds are spent on social research in the desperate hope of repairing the relationship between government and governed. When the French government decides that a new railway line is needed, it builds it. But when the British government faces a similar situation, it waits while economists attempt to work out whether this would “add value to UK plc”. Wouldn’t it have been better simply to advocate it in a manifesto, then see whether the public wanted it or not?

Public services and public spaces are especially imperilled by this, because of the intense difficulty economics has in capturing their qualities. Institutions such as the BBC or the British Library are important parts of the public sphere, a fact which most politicians and citizens agree upon. But where nobody has confidence in this judgment, the same conclusion has to be reached using polling, modelling and endless, endless Powerpoint slides. If evidence-based policy is expected to defend such public goods without any assistance from democracy, it is not inconceivable that they could wither away altogether.

We encourage anyone to comment, please consult the
oD commenting guidelines if you have any questions.