Can “Evidence-based Policy” Replace Ideology in Politics?


Duncan Davis

Ideology is a dirty word. Tony Blair spoke of New Labour as “beyond ideology”  and concerned only with “whatever works”. In a speech introducing the government’s austerity programme of public sector cuts (framed as “deficit reduction”), David Cameron claimed “We are not driven by some theory or some ideology”. Even the Green Party, themselves criticised as “too ideological”, attack austerity policies as “ideologically-driven”; although perhaps that’s an attempt to counter the claim that the policies aren’t.

If ideology is to be eschewed, what is politics? Traditionally, many would see politics as a principled battle of ideas and values systems. Without ideology, this becomes impossible. Instead, the battlefield of modern mainstream politics centres around ideas of “competence” and “effectiveness”. The largest two parties of UK politics don’t have major differences in their policies. Ed Miliband recently announced that Labour would continue public sector cuts, but stressed that they would be “sensible”. Maybe due to a lack of actual political differences, the Conservative response wasn’t arguing with any of the principles espoused but simply that the implementation would cause “economic chaos”. In this situation, it seems unclear how electors should decide who to vote for; how aspiring politicians should decide which party suits them; and indeed how politicians should make policy decisions.

The proponents of “evidence-based policy” believe it can fill this void and create a politics free from the dysfunction that they perceive as stemming from ideological differences. Ben Goldacre proposes that policies be tested using randomised trials and adopted based upon the results. The idea that we can scientifically test policies and then implement what works, producing better outcomes is attractive. No longer would we need to answer difficult philosophical questions. Politics could become a technology, iteratively optimised to produce better results. However, there are problems with this approach – practically, ethically and, indeed, ideologically.

At its core, evidence-based policy has the idea of policy trials, analogous to drugs trials in the study of medicine. A policy should be tested on a group (for example, a set of schools for testing an education policy) and compared to a control group, where the policy isn’t enacted. The comparison with the control group, given sufficient sample sizes, would enable measurements of effects of the policy to be distinguished from other causes.  In theory, this is an excellent idea, but there are difficulties. Control groups are hard to isolate in society, where everything interlinks – a policy implemented in one group could have knock on effects in others. There is also an ethical problem: if a policy turns out to have negative outcomes, it might impact the lives of the people in the group on which it was tested. Policy trials risk (probably without consent of individuals) turning people into sociological lab rats.

This ethical conundrum requires a values system to decide when testing policies on people is acceptable, so already it’s clear that evidence-based policy can’t entirely replace ideology. In fact, the scope of the policy trials approach is highly dependent on the status quo of prevailing ideologies. To test policies individually, the assumption has to be that the underlying social and economic structure remains approximately the same. This makes evidence-based policy inherently conservative. There are a plethora of valid political visions that promote a fundamental restructuring of society. Anarcho-syndicalism can’t be tested in a controlled randomised trial!

Proposals don’t need to be revolutionary to be unsuitable for policy trials, though – the methodology is severely limited in other ways. It requires that effects of policies come into force within the timespan of a study, so longer term outcomes are infeasible to determine in this way. Any large scale or macroeconomic change has no way to be controlled and so evidence-based policy struggles to tackle the big issues.

Despite these shortcomings, it is vital that the political process heeds the evidence available.  Arguably, there is no issue for which this is more important than climate change. The IPCC has shown, through extensive research, that the world climate is warming due to greenhouse gasses emitted by human activity, and that this is having drastic effects. Other studies suggest that greenhouse gas emissions are tracking above the IPCC’s worst case projection (RCP8.5) and that Arctic sea ice is melting at a faster rate than predicted by computer models.

The insurmountable evidence hasn’t stopped politicians and members of the public from doubting the existence of climate change or, at least, that human activity caused it. A website detailing research about climate change communication, “Talking Climate”, suggests that this is due to people squeezing “new evidence through powerful social and cultural filters”. People who mistrust climate science tend to have conservative ideologies and repeating the facts at them is unlikely to change their minds. It is suggested that the cause of this opposition are the implications (like the need to regulate markets) rather than the science itself. Debating using non-ideological language obstructs the acceptance of the evidence. The article recommends that ideological disagreements be taken “out into the open rather than obscuring them by fighting political battles using the language of science”.  This principle could also be applied to politics in general, where obviously existing ideologies are hidden.

It’s unclear exactly how to tackle climate change. There are a vast range of proposals ranging from moderate market mechanisms such as tradeable carbon/energy quotas to radical changes like overthrowing capitalism. One might hope that evidence-based policy could shed some light on such ideas but climate change is a prime example of where policy trials can’t work. Climate change is an issue at the largest scale: global. Policies, therefore, have to be applied globally and we have no control planet. In all probability, a vast range of interconnecting policies will be needed and so testing of individual policies doesn’t make sense. The time scale doesn’t work either – we need action now but the outcomes of actions often won’t come into effect for years.

This doesn’t mean we can’t put evidence to use. If we look beyond the purist conception of evidence-based policy, with policy trials as its foundation, there is a wealth of knowledge that can augment political decisions. We understand how technologies work and can predict outcomes based on that knowledge. For climate change, evidence tells us the cause: greenhouse gas emissions. This allows us to evaluate technologies that, for example, generate electricity – by looking at the quantity of greenhouse gas emissions per unit of energy.

Proponents of nuclear power (uranium fission) use such metrics to claim that it is the solution to providing for our energy needs while limiting the effect on the climate. Often it is proposed as a baseload for renewables or as part of an “energy mix”. Others oppose it for many reasons, such as high costs and plants taking too long to build to address the urgent needs of mitigating climate change. There is a debate over the evidential claims made by both sides and it’s important to have this debate. What isn’t helpful is the frequent labelling of nuclear power opponents as “anti-science”. As Alice Bell puts it, this is a “rhetorical trump card” used to shut down discussion and hide from ideological differences. For evidence to be put to use effectively, it needs to allow for debate of inherently uncertain scientific claims. The aversion to discussing ideology, as well as vested interests (those in the nuclear industry are likely to want to avoid casting doubt on it) are probable reasons people use this technique. Having open debates about, for instance, the centralised hierarchical social structures needed for nuclear and fossil fuel power plants versus the decentralised network structures implied by renewable technologies, could be very helpful in making the political decisions.

From economics to climate change to energy technologies, the political debate uses ideas from evidence-based policy that damage the debate of both the evidence and political values. This approach is so common that it may seem that discussions are doomed to repeat the pattern, but supporters of a policy idea known as Universal Basic Income (UBI) are showing a glimpse of how ideas can be discussed differently. The UBI is a payment (often to replace some existing forms of means-tested welfare) that would be paid to all citizens; be sufficient for them to meet their basic needs; and come without any conditions. Proponents see it as solving or mitigating many big problems, such as poverty, technological unemployment and climate change. They are often refreshingly honest about their (socialist and libertarian) ideologies and the effects they believe it will have are often couched in value-laden terms like social justice and freedom.

However, it’s not just an ideological dream. The arguments are augmented with evidence from trials (for example in Canada, Namibia and India), of various limited forms of guaranteed income. The evidence from the trials is positive and they are often quoted to counter claims by opponents that UBI will cause detrimental inflation or stop people from working. UBI seems to be an ideal case where randomised controlled trials do work, partially at least. It’s very simple to test, as it’s just paying people money and it’s hard to make an ethical case against this testing. It has been tested for relatively short terms and outcomes have been measured, although claims of long term cultural changes can’t be tested this way. There is still the problem of scale – as a policy that is usually proposed at a national or international level, predictions of outcomes made using the small scale tests are merely extrapolations. By not relying solely on the trials, UBI supporters can still make coherent arguments with a mix of principles, evidence and reasoning.

This careful, critical type of argument needs to be expanded throughout politics. There needs to be a clear separation made between evidence and ideology. Ideologies should be asserted and critiqued without the debate being shut down. The evidence should be stated along with its limitations (for example, arguments for UBI that employ evidence from the trials should make clear the scale limitations) and critics should question the ideological assumptions that may be behind methods of experiment, measurement and analysis. Such changes are very hard to make, even at an individual level. We are irrational and notoriously blind to our biases. Critical thinking takes a concerted effort. This effort can be made personally, and if taken up by enough people, the approach can propagate through grassroots politics.

Changing the attitudes of the political establishment, however, is a much bigger task. Those with power have a vested interest in keeping it. Shutting down debate and asserting ideologically based opinions as evidence based fact  is a convenient tool for this. The austerity agenda, ideologically based in certain small state, liberalised market positions, is pushed across Europe as a “fact” that has close to consensus with mainstream parties: we must cut the deficit. This can be used as a benchmark for judging “what works”, to use a phrase espoused by those who follow the “evidence-based policy” ideology. Politicians use their power of influence over mainstream media to push such agendas. Rarely do we ever see the basis of austerity, or other prevailing ideas, questioned. This means many important debates don’t happen except in fringe arenas. Rather than the dysfunction in politics stemming from too much ideology getting in the way, it arises from a conflation of the evidence with ideology.

Ideology is ever present. Evidence-based policy can’t replace it, and shouldn’t, as discussions over philosophies and values are important to have. This doesn’t mean some of the ideas involved aren’t useful. Policy trials, while limited, can be used in certain cases to build up evidence for or against policy ideas. Different sources of evidence can be combined, if done critically, to provide guidance over what policies can be useful in working towards ideologically determined goals. The evidence and ideology need to be presented clearly and separated as much as possible. Unfortunately, it seems unlikely that the political class will volunteer to do this, considering how obfuscation benefits them and their influence over the media enables it. Ultimately, there needs to be a decentralisation of power and a furthering of democratic control over all publicly and privately owned institutions. That’s an ideology, but there’s nothing wrong with that.

Can “Evidence-based Policy” Replace Ideology in Politics?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s