How does the UK cabinet office make decisions about policy? Last week I went to an interesting workshop that threw some light on this (and on sausages). This account looks at some of the influences on the policy making process and compares it to the way in which we all, as individuals, make decisions about our own intentions and behaviours. There are more similarities than you might think. What’s more, the processes involved are quite different from formal models of decision-making.

The workshop was run by Cambridge University Science and Policy Exchange (known as CUSPE – Ref:1), set up to help early stage career researchers inject more science into policy decision making. We learned that the findings from science play only a small part in governmental policy making. Science is concerned with getting what scientists evaluate as the right or optimal answer. Policy formulation is more pragmatic. It looks for an answer that everybody can live with, and that’s often quite different.

YouTube Video, Evidence not ideology: Professor Nutt & the UK’s ideological policy making, I Am Incorrigible, June 2015, 9:10 minutes

More of that later. First, let’s consider policy from an individual perspective.

What is a Policy?

We tend not to think of individuals having policies. But we do. You only have to say that you are a vegetarian or that you go to the cinema at every opportunity, and you have declared some policies. A policy is a relatively high level intention to act in particular ways. It is a governor of other intentions and behaviours. So if you are a vegetarian this will have implications, not just for what you eat, but also what food you buy and probably, to some extent, who your friends are, together with all the knock on implications for how you might spend your time.

A policy is also a justification. It answers questions like ‘why did you do that?’ or ‘why do you think that?’. It helps explain where you are coming from. We are also often at pains to defend our policies and can ferociously argue for them or against their opposites.

Policies tend to regulate, or at least guide, our behaviours. We can feel uncomfortable when we say one thing but act in another, and other people are often only too willing to identify hypocrisy.

However, as individuals, we barely recognise these propensities as policies, and we often have no clear idea where they have come from. We say, ‘that’s the way I’ve always done it’ and often can’t answer the question ‘what led you to do (or think) that)?’ As we have increasingly revealed since Freud, and more recently in neuroscience, there is a lot going on under the surface that we are simply not aware of.

Policies are important because they are difficult to change. As they are high level they have a lot of implications. In the individual they are entangled with and embedded in many other intentions and behaviours. A government policy is equally difficult to change – ‘The lady is not for turning’.

YouTube Video, Margaret Thatcher ‘The Lady’s not for turning’ (from October 980), Danco28, October 2008, 30 seconds

If you wake up one morning and decide that you are going to eat toast for breakfast that day instead of sausages, you are making a one off decision that doesn’t affect very much. That’s easy.

Changing your policy with regard to breakfast is something else. It involves making a more strategic decision that you will change your behaviour in the longer term. Perhaps you are making up your mind to eat more healthily, switching out of sausages, egg and bacon and into bran flakes and muesli.

In changing your breakfast policy you are deciding to change a lot of other behaviours, so for example, you may change where you shop and what you buy, with the implication that there will no longer be sausages in the fridge, so you won’t be tempted to eat them.

You are also changing your image. In a way, you are declaring to the world that you are a bran flake eater, not a sausage eater, with everything else that that might imply. Indeed this may be the very reason you are making the change. Policy-making is very much a social act.

Furthermore, you are not just changing your Facebook profile, so to speak. You are also changing your self-image. Now, you are a person who is not locked into your bad habits. You can see yourself as flexible. You are responsive to the world and it’s new knowledge about the dangers of sausages, like their impact on climate change.

This kind of policy thinking is going on in the backs of our minds all the time. As we experience the world, with events that affect us, changes in circumstance, and shifts in our understanding, we are continuously making adjustments to our value system. Our value system is ever present, monitoring even the tiny details of behaviours and thoughts. Our evaluations are reflected in our emotional states and provide the motive power that leads to changes in our policies – the regulators of thought and behaviour. Most of this we are not aware of and even in government policy making, the tides and forces under the surface may have more impact than we think.

But changing policy is not easy because it has so many implications. As Kuhn (1962 – Ref:2) pointed out in ‘The Structure of Scientific Revolutions’ – the case for a change of viewpoint has to be overwhelming before most people will even start to consider it. There is a tremendous inertia related to the current paradigm because it has become embedded and entangled with many other practices.

You may have an intention to change a policy well before you are able to make the change. Change requires a lot of time and effort in addition to the will. Also, the circumstances have to be right to facilitate the change. It’s a bit like committing a crime. You have to have the motive, the means and the opportunity.

These factors are very much at play in the formulation of government policy. It turns out that decision-making in the cabinet office is not that dissimilar to you making a decision to change your policy about breakfast.

However, it does surprise (and perhaps alarm) some people that deciding about breakfast and deciding about, say Trident, have much in common. You would like to think that policy-making at the highest level uses a much more rational, informed and robust process. But its worth making a few points here:

• Firstly, the cabinet office deciding on policy and you deciding on your breakfast policy is essentially the same process except that the government process, in most cases, is more elaborate and explicit.
• Secondly, neither the individuals nor governments tend to follow any normative or formal process derived from, say mathematics, economics or decision theory.
• And thirdly, whatever you might initially think of it, it is a process that is eminently sensible given that decision-making is taking place in an uncertain world, in real time and with limited resources.

Government Policy-Making

At the workshop, a speaker with a good insight into Cabinet policy decision-making, described some of the factors that influence it.

Pursuing a policy is as much about whether it is practical to implement as whether it’s inherently a good policy that is both needed and will have the desired outcomes.

Many factors contribute to the practicality of a policy but I think it is best summarised as whether there is sufficient support for it to overcome the resistance to it.

Unless somebody in government or some pressure group is shouting loudly enough that they want the policy it’s a non-starter. Then, assuming somebody wants it, their voice has to be heard above the many other voices that raise objections or alternatives to it. How clearly that voice is heard will depend on how easy the policy is to justify, and that in turn will depend on a whole host of factors like:

• How urgent and important the issue is
• Whether its topical at the time
• How clearly the policy can be expressed
• Whether the policy aligns with other policies
• The strength of the argument for the policy
• How it would impact on other policies
• How easily the policy can be implemented
• The degree of certainty that the policy will deal with the issues it addresses
• The likelihood that the policy would have side effects or unintended consequences
• The extent to which the policy would upset vested interest or the status quo
• How much risk there would be to the governments reputation if the policy went wrong (i.e. did not get sufficient support, could not be implemented or failed to achieve the intended outcomes)

Not much of this concerns whether the policy is good in terms of any science or evidence that underlies it. Science may be one of the factors that increases the strength of the argument, but even then it is just one of many factors like impacts of the economy, employment, immigration, national security and so on.

If we return to the new breakfast policy, it is easy to see how much the same factors influence individual decisions. The healthy breakfast policy is easy to understand and express. Perhaps a friend has changed their diet, so it’s topical. There are unlikely to be side effects or unintended consequences. There is good scientific evidence that eating a lot of meat causes health problems. Changing to bran flakes may enhance your image and self-image. Its probably easier to do if its part of a wholesale shift to a more healthy diet. It could impact the people around you a little but it will only affect vested interests if, say, you regularly eat with someone that enjoys a full English breakfast and you need them to stop too so that you are not tempted.

However, changing sausages for bran flakes is not urgent or important unless your doctor alerts you to the possibility of high cholesterol and heart attack. It’s not at all certain that, on its own, it would make much difference to your own health outcomes. Most importantly, it does require a change to the status quo and that can make it difficult to implement. You have to change a few habits – what you buy, the things you cook and, importantly, what you expect and look forward to for breakfast. As a consequence, it takes some willpower to make the change. As with every aspect of life, you will only change if you really want to. The pressure to change has to overcome the resistance but it can be facilitated by opportunity (a change of circumstances for example).

As for the cabinet office, your own decisions may only be marginally affected by the scientific case however strong the evidence might be.

The Individual and Society

Going back and forth between the individual perspective and the mechanisms that operate at the level of society helps each inform the understanding of the other. We can see politics in terms of individual psychodynamics and we can see psychodynamics as the politics of the mind.

Having drawn a few relatively superficial parallels, there are some more interesting parallels in the underlying mechanisms of decision-making. Four mechanisms are illustrated here:

• Ease
• Competition
• Stories
• The real-time dynamics of juggling multiple intentions

They can be seen to operate at both the level of individual psychology and at the level of cabinet office decision-making. Although there is a wide range of psychological findings that can illustrate the parallels, I have frequently referred to the work of Daniel Kahneman (Ref:3). It is not only interesting and informative, but it provides a single reference source for those wanting to follow up in more depth.

How Easy is It?

We already know, from the work of Kahneman and others in behavioural economics that as individuals we are lazy, at least in our thinking. Kahneman refers to ‘cognitive ease’ as the default mechanism by which we tend to jump to conclusions on the basis of intuition. Much of our behaviour is routine and well learned and unless something surprises us we are reluctant to turn our full conscious attention to it. For the most part we are on ‘automatic pilot’ and it requires effort to change our habits. We have a limited capacity available to problem solve and limited willpower to effect change. Our resources are easily depleted, so unless there are few demands on us, we only attend to the matters that we assess as high priority. If there is no immediate threat to our security or wellbeing we give them low priority.

In his book ‘Thinking Fast and Slow’, Kahneman examines several cognitive biases, many of which result from the need to conserve resources, and only pay attention to things that absolutely require it, for example, when there is a mismatch between what we observe and what we expected.

This favours pragmatism over principle. Even if there are strong arguments for change, unless the need is significant and immediate, like a threat, we will be reluctant to do anything as complicated as formulating a policy.

The findings of psychologists like Kahneman, illustrate just how much inertia there is to change and, if you are a meat eater, how much prodding may be necessary before you even start to consider switching out of sausages.

Much the same principles may operate in cabinet office decision-making. This will have advantages in terms of dealing with threats to survival, but just as with the individual, it may cause systematic biases in the way matters are dealt with. There will always be the dual motivations of ‘doing the right thing’ and ‘surviving’. These interests will generally align but they create a bias, particularly towards short-termism (partly because pay-offs in terms of image and reputation become immediately discounted after the next election). Risks to reputation are normally managed, strengthen the bias, although as recent resignation events show, sometimes these risks are not recognised.

Matters that appear routine will never reach the cabinet office or will immediately be relegated to the relevant departments, even when paying attention to them might pay dividends in the longer term. Significant and immediate threats will be dealt with as a priority, although the solutions found will be the best available, rather than the best overall. Evidence may be sought to support a decision, rather than to caste doubt on it. Attention may become too focused on single issues when it would be better divided amongst several. Issues that are significant, clear and topical will also find their way onto the agenda more often than those that are of greater significance but complicated and out of sight.

Just like the individual, the cabinet office is dealing in real time, with limited resources in an uncertain world. The cabinet office must think fast and will be prone to the same biases and errors as the individual. As a society we understand this and to some extent trust that our elected representatives will do the best they can, while acknowledging the biases and maintaining close scrutiny. A totally democratic system will still be biased. To quote Churchill ‘Democracy is the worst form of government, except for all the others’.

Who is shouting the loudest?

For a policy change to be considered, pressure for a change must be sufficient to overcome the resistance to change. In political decision-making this is mainly pressure from various stakeholder groups both outside and within government. For example, outside government the interests of livestock farmers have to be reconciled with the interests of, say, the environmental lobby or animal rights activists. Within government DEFRA might support farmers and encourage subsidies for rearing livestock for meat and dairy, while the Department of Health might take an entirely different view taking into account implications for obesity.

An approach within ‘communication theory’, called ‘critical theory’ has a perspective on the characteristics of this type of decision-making.

YouTube Video, Stan Deetz – A Quick Summary of his critical theory as it applies to organizations, mentorsgallery, January 2010, 0:35 seconds

These various pressures manifest as a set of proposals for policy that compete with each other, although the proposals may not always map cleanly onto the various pressure groups and what might finally emerge is one or more proposals that compromise between viewpoints. Some proposals will address easier issues than the one that really needs to be addressed, or as Kahneman puts it ‘an easier question will be substituted for one that is too difficult to answer’. One way or another, there is a premium on selecting a definitive winner, because of the need for narrative simplicity and clarity.

Much the same model of decision-making between competing proposals, seems to operates at the level of the individual. Many mental phenomena appear to operate on the principle of who or what is shouting the loudest. Minsky (1986 – Ref:4), laid the theoretical foundation for this idea. He refers to ‘the Society of Mind’ in which individual ‘agents’ form the building blocks of cognition. In this model, individual agents in an individual mind may operate something like the house of commons during prime minister’s question time – in other words, in what is sometimes an embarrassing brawl in which opposing factions shout insults at each other in an attempt to ridicule the other’s arguments. Meanwhile, the evidence on confirmation bias suggests that each agent is cherry picking evidence to support their established position and ignoring evidence that might refute it.

For the individual, the subjective experience is sometimes like a conversation taking place in the mind. Unfortunately, the literature on subjects like ‘voice dialogues’ (Kwasniak 2013 – Ref:5) seems somewhat sparse for a topic that could be highly explanatory. However, psychologists have know for 60 years (Festinger 1957 – Ref:6) that it is possible for an individual to hold multiple inconsistent viewpoints. The work on cognitive dissonance, for example, both identifies this and the emotional discomfort and cognitive mechanisms that go with holding incompatible beliefs.

We are all familiar with the idea that we can hold multiple incompatible hypotheses (i.e. without strong commitment to any one of them). There is evidence from psychology and neuroscience that neuronal assemblies are doing some sort of equivalent of arguing with each other in iterative cycles. This can happen over even the tiniest aspect of the interpretation of a sound or visual scene, and the interpretation that wins out is the one that is shouting the loudest. It is generally a ‘winner takes all’ mechanism. At the margins, some visual illusions demonstrate how fragile a particular interpretation can be, such that perception oscillates between two competing hypotheses (Necker Cube – Ref:7).

Returning to breakfast, as a meat eater, there may be so many more parts of your brain used to shouting ‘eat the sausage’ than those offering convincing arguments for alternatives, that you just go with the flow. If this creates dissonance and you have the will (motive), means (cognitive capacities and resources) and opportunity (social and other circumstances) you can beat down the ‘eat the sausage’ lobby. Otherwise you may need to deal with the stress in another way (e.g. by denying the evidence or bottling up the stress until it manifests as something else).

A Good Story

For the cabinet office it is important that a policy can be presented simply and clearly. It needs to make a good story so that it can be both understood and be credible.

YouTube Video, Tales from the City (How UK banks influence national policy making), PressTV Documentaries, February 2015, 24:20 minutes

We have recently seen that, a policy like reducing disability benefits, is simple but lacks credibility when presented along side tax cuts for the better off.

But once again, Kahneman has demonstrated that accuracy is secondary to simplicity in order to appeal as a good story. So a policy story about cracking down on benefit fraud will sound good because it has a clear and simple moral message, even though such a policy will only save a tiny fraction of the budget. The story distorts the apparent usefulness of the policy.

Kahneman also demonstrates that retrospective memory is subject to bias. We may have been through many years of austerity but in remembering them we will not sum the accumulated pain. What we remember are the worst moments and the last moments. So a long drawn out program that is not too bad right at the end, may be perceived as relatively painless, resulting in a distorted assessment of the past just prior to an election.

Up Against the Wall

As individuals most of us have experienced occasions on which we have had to make a significant choice, perhaps about our careers or an expensive purchase. We may think about the choice over a long period. We may find consideration of different aspects of it popping unexpectedly into our mind from time to time. We may research it, ask people for their opinions and write lists of pros and cons.

We may procrastinate over the decision because we cannot quite make up our minds. At some point we may become certain that one option is the best, but if we do not act at that time, something may happen to make us re-consider and another option becomes more favourable. Choices can cluster so that if we chose option A then options D and E make sense with respect to other intentions, but if we chose B and options G and H make more sense.

When we are up against the wall and have to make a decision or lose one or all the options, we may commit or let some options drift away through inaction. One way or another, when we actually commit it may well be on the basis of the criteria that seem important at that time, rather than on the basis of a reasoned and ‘averaged’ or balanced consideration of all the criteria considered.

This juggling of intentions and options is what I call the operation of the human operating system (Rivers 2016 – Ref:8). It operates in real time, in an uncertain world and with limited resources. It also operates across systems at many different levels of analysis including the neurological, the individual mind and groups/organisations, such as that of the cabinet office. It reconciles multiple, sometimes mutually exclusive, intentions in a world that is constantly, and often unpredictably, changing and providing different constraints and opportunities as time passes. The term ‘situated cognition’ (Ref:9) has been used to capture this type of thinking ‘on the fly’.

By and large, the human operating system does a pretty good job fielding threats and maintaining survival on the limited resources available. However, as Kahneman demonstrates it has systematic biases. Some of these can be countered by application of the scientific method, that in Kahneman’s terms can be seen as society’s equivalent of what he calls the slower system II type thinking.

While it can be argued that what really matters is whether a policy actually produces the desired outcomes, the reality is that the type of thinking required for real time response in a changing world is difference from the slower and more measured thinking of science, and both are involved in the formulation of policy.

Formal Models of Decision Making

Drawing the parallels between individual decision-making and decision-making at the cabinet level helps throw light on both. What is clear, is that neither correspond to more formal models of decision making based on mathematics, economics or even some psychologically based models such as game theory (e.g. see Heisenberg 2005 – Ref:10).

Models of decision making are often couched in the form of choices amongst multiple options. For example, Randy Hirokawa, describes group decision making in terms of:

• Problem analysis
• Identification of appropriate criteria to make the decision
• Development of alternative choices
• The evaluation of the advantages and disadvantages of the alternative choices

In his functional theory of group decision-making Randy Hirokawa says that a groups ability to make a good decision depends on their effectiveness in these areas. However, albeit that these activities are relevant to decision-making, they fail to explicitly address almost everything that is discussed above about the reality of decision making in a political, policy-making context.

YouTube Video, Randy Hirokawa on the Functional Perspective on Group Decision-Making, A First Look at Communication Theory, January 2014, 7:18 minutes

Most decision making theories lack some of the essential ingredients of decision making in the real world. These include that:

• real world decision-making is often about threat and survival
• it takes place in a changing environment
• it is carried out with limited resources
• decisions are often highly interwoven and inter-dependent, especially decisions about policy
• critical information is often missing or incomplete
• decisions often have to be made in a limited time window with the resources and information available in that window
• decisions are fundamentally social in the sense of reconciling the viewpoints of multiple agents whether this in within a single mind or in society generally

This is perhaps why, in 2011, the Institute for Government report on Policy Making in the Real World said:

“The more one delves into the reality of policy making, the more that policy cycles and their like
resemble a comforting narrative that imposes specious order on a complex reality. Maintaining
this narrative often means that, in practice, policy makers often have to fall back on their native
wits. This is why many interviewees voiced concerns about the ad hoc nature of policy making:
there is not so much a lack of recommended processes, just a lack of realistic ones.”

Perhaps, in time, it will be work carried out in artificial intelligence in designing systems like robots and self-drive cars, that will most inform the modelling of decision-making.

YouTube Video, Close-Up Look At Google’s Self-Driving Car, CBS SF Bay Area, July 2015, 3:32 minutes

Sausages and Climate Change

So, what’s with the sausages? As its example, the workshop used policy in relation to the consumption of meat. This is not just a matter of health. There is a significant health impact of eating beef, lamb and pork, and sausages kill more reliably than most meats, although it is much less than, say, the impact of smoking. However, one impact that people are less aware of is global warming and climate change.

How do sausages change the climate? The science shows that animal farming has more impact on CO2 emissions than either energy production or transport. We were informed that if the world doesn’t slow the trend in it’s consumption of meat, CO2 will increase by 75% by 2050. So, putting petrol in your car (or charging it up) and driving to the supermarket to get a packet of sausages is tantamount to planetary suicide.

Workshop Format

A concern for CUSPE is how to design these workshop sessions so that they are most effective.
A significant lesson that comes out of this is that science is just a small part of what drives policy. How we make decisions on policy, whether it’s our own breakfast or meat eating generally, is influenced by a whole lot more than the scientific evidence, and that’s what science researchers need to know if they want to change anything.

The argument for a policy is just part of the story. Researchers also need to know how the decision-making process works and how to present policy recommendations so that they can be used effectively within this process.

Furthermore, science informs just one part of an argument that may also include considerations such as impacts on employment, the rural vote, and right now, as a nation of potential non-sausage eaters, it would impacts on whether we stay part of Europe!

For a policy to succeed it must overcome the resistance to change. People must be clambering for change and the argument must be overwhelming. Also, the policy must fit with other political initiatives, not offend too many interests, and be easy to implement and communicate, before much will happen. This is more like steering a ship than a bicycle.

In the workshop we were divided into stakeholder groups – the livestock lobby, the World Health Organisation, Environmental campaigners, the less well off, and animal rights activists. Each was required to propose policies from it’s point of view to the cabinet office group.

I was in the Low Income Group. We summised that the factors influencing us would be price, convenience, health, status / image and taste. The policy we proposed was ‘whole life costing /pricing’ whereby social costs (including animal rights, convenience, and health impacts) would all be built into the price. In this way rice would be cheap and hamburgers expensive paving the way for McDonalds to offer a new range of rice and olive oil stir fry take-aways.

However, our policy proposal stood no chance. It was rejected immediately by the ‘cabinet office’ because it implied some centrally controlled pricing mechanism – clearly complicated to implement, difficult to justify and at odds with other policies and the operation of the free market. This was a lesson well worth learning and CUSPE should be congratulated on the design of a workshop format that brought this out so effectively.

References

1.CUSPE – CUSPE: Cambridge University Science and Policy Exchange
http://www.cuspe.org

2. Kuhn (1962), ‘The Structure of Scientific Revolutions’
https://en.wikipedia.org/wiki/The_Structure_of_Scientific_Revolutions

3. Daniel Kahneman(2011), ‘Thinking, Fast and Slow’
https://en.wikipedia.org/wiki/Thinking,_Fast_and_Slow

4. Minsky (1986), ‘Society of Mind’
https://en.wikipedia.org/wiki/Society_of_Mind

5. Janet Kwasniak (2013), ‘Inner Voice’

Inner voice

6. Festinger (1957)‘A Theory of Cognitive Dissonance’
https://en.wikipedia.org/wiki/Cognitive_dissonance

7. Necker Cube Illusion
https://en.wikipedia.org/wiki/Necker_cube

8. Rivers (2016), ‘The Human Operating System’
http://www.wellbeingandcontrol.com/?cat=5

9. Situated Cognition
https://en.wikipedia.org/wiki/Situated_cognition

10. Heisenberg (2005), ‘The institution of ‘consensus’ in the European Union: Formal versus informal decision-making in the Council’
http://onlinelibrary.wiley.com/doi/10.1111/j.1475-6765.2005.00219.x/abstract;jsessionid=70B37CA840EC0414204CC78970A66FF5.f02t02