Lecture to the Westminster Ethical Policy Forum by Perri 6
“Not what they want, but what is good for them”. The statue of the man whose words those are, stands just in the next street from here – in front of the Palace of Westminster. Somehow, one has no difficulty in imagining Oliver Cromwell thumping the table and insisting on the right of the government to do good to the unwilling and undeserving.
No liberal qualms in those days. But not only then. It is only forty years or so since the late Douglas Jay infamously declared, with a confidence that few politicians in the developed world would claim today, “There are some things about which the gentleman in Whitehall really does know best.”
In a Christian Church like this, perhaps one ought only to whisper it, but we are not always morally permitted to do good. And whatever the Lord Protector and the Gentleman in Whitehall may have thought, most of us today think that the state has rather few greater rights to do us good without our consent than do our families, friends, and perhaps even less than the professionals such as doctors, lawyers and ministers of religion to whom we turn.
The American writer Henry David Thoreau once said, “If I knew for a certainty that a man was coming to my house with the conscious design of doing me good, I should run for my life.”
This is not only the view of an American romantic solitary. In fact, a good many of Cromwell’s intended beneficiaries seem to have taken a similar view.
By paternalism, I should make it clear that I understand the following:
deliberate acts (or omissions) done by an agent – in the present case, a government agency – contrary to the preferences of the individuals or group, and either coercing the, heavily sanctioning their choices, or overriding their preferences about services from or transactions with the agent, where normally that would not be morally acceptable but justified by the agent (or others) on the grounds that the action has the principal aim of promoting what the agent sees as the subject’s interests, good or benefit, or preventing what the agent sees as the subject’s harm.
The question has become more urgent in recent decades: when may government pursue our benefit, as politicians or officials see it, when we have not specifically authorised them to do so?
In the first half of the twentieth century, in Britain and in many European countries, governments’ objectives for domestic policy were concerned principally with the growth of the economy, the preparation for war and the maintenance of empire, the containment of class conflict and perhaps some redistribution of income. In the second half of this century, the objectives have expanded, in ways that have made conflicts over paternalism much more salient in everyday life. I want to stress three types of political project that have become central to recent politics – of both left and right - that have brought sharply into question the question of the moral bounds upon the benevolence of those who dispose of the taxpayers’ money.
First, governments have come under increasing pressure to take responsibility for the reduction of risks other than the traditional ones of death in war, or those presented by Beveridge’s five horsemen – ignorance, idleness, disease, squalor and want. Safety in the workplace, safety in food hygiene, safety in transport, safety to passers-by and residents from industrial installations, safety in our own behaviour… these are now the small change of domestic politics, and increasingly of international relations. Whatever the laments of rugged libertarians about the culture of safety and the widespread tendency to look to state regulation, information advice, or services to promote safety, that demand has come from industry as much as from consumers, from rich as much as from poor. If special interests are to blame, then there are so many that the term “special” seems almost misleading. On the one hand, many of us resent “the nanny state” – particularly when it issues an edict that beef may be neither sold nor served on the bone. On the other hand, we have to come to accept laws requiring the use of seatbelts in cars and helmets for motorcyclists, and even many of those who regard the prohibition on the recreational use of cannabis as unacceptable paternalism are prepared to accept the same prohibition upon heroin and cocaine.
Secondly – and if this was anathema in the age of Gladstone, it would have made a certain sense to Cromwell – our age is one in which politics consists in rival and co-operating projects to change the cultures of the public. To seek to change the cultures of the public is to try to inculcate motivation, so that the behaviour governments desire from citizens is undertaken voluntarily. If this can be done successfully, then it is less expensive, easier to administer, more effective and more sustainable in the medium term than relying upon targeting specific incentives and sanctions. Bu the goal is to secure a culture that governments believe is in the best interests of the citizens among whom they try to foster that culture.
For example, on the centre-right, the goals have been to undermine cultures of dependency, the create and foster cultures of self-reliance, to establish an enterprise culture, to create a culture of personal responsibility, to engender a culture of family obligation, and to sustain or defend a national culture, a culture of national sentiment and experience. On the centre-left, we are seeing the development of projects to foster a culture of health to support healthy living, to pursue a culture of “zero tolerance” toward every manifestation of crime or social disorder, to create and sustain a culture of aspiration in education and the labour market, to create a culture of lifelong learning, and so on. Of course, political projects have always influenced cultures among the public: there is no pristine unsullied pre-political culture to be recovered. In Gladstone’s day, the aim of liberals and conservatives alike was to promote what was then called “character”. But what is new to recent decades is the range of specific projects for cultural change and the explicitness with which that goal is pursued.
Thirdly, with the growth in public services, conflicts have emerged over several kinds of paternalism - between the services that recipients want and those that taxpayers believe recipients need, between the services the public as a whole wants and those that experts, technocrats, policy wonks, professionals or politicians think are in their best interests. For example, the professional commitment to “evidence-based policy” and to “cost-effectiveness” delivers us such strategies for the public service as intelligence-led policing, whereas majorities of the public say they would prefer more officers pounding the beat. Again, policy makers and analysts – myself among them - call for joined-up services, yet many of the public show themselves to prefer the fragmented systems, if only because they have some “brand recognition”. “Populism” is the swear-word of those who know better than service users or taxpayers what services are in the best interests of the beneficiaries or purse-holders. How far, then may politicians professionals in the public service choose services for the benefit of their recipients, if those recipients have other ideas?
These debates are conducted at the level of the individual and the citizenry as a whole. Conventionally, we call claims to promote the good of individuals without their specific consent, ones of “paternalism versus individual autonomy” and those to promote the good of the citizenry against the preferences of majorities, we call ones of “good governance versus democracy” But the structure of the argument is the same. The liberal presumption that governments (or anyone else for that matter) need very good reasons for overriding individuals’ preferences about the risks they want to run or be protected from, the cultural aspirations and motivations they are prepared to cultivate in themselves and their children, is structurally very similar to the democratic presumption that governments need very good reasons for overriding the preferences of the citizenry about the public services they want.
Like many people, my initial instincts are that those reasons ought to be few, rather restrictive and limiting upon the powers of government. But like many other people, my understanding of the condition of contemporary governance is that even that libertarian aspiration secretes within itself a project for cultural change, which may be as paternalistic in its way as those it seeks to oppose. For there is no pristine culture of self-reliance and independence that one can dig down to, nor is there any non-paternalistic way to – as libertarians sometimes incautiously and revealingly say – “wean people off” desires for protection from risk, aspirations for safety or solidarity, or the desire to delegate certain kinds of decision-making to professionals. Another way of putting the point, is to say that simple libertarianism is beautiful, intensely romantically appealing, but culturally no more sustainable than the politics it would reject.
The history of twentieth century is not a reversible process: conflicts about paternalism are endemic, and now intrinsic to the nature of modern political life.
If this is broadly right, then, it is rather important to explore whether there are circumstances under which governments might legitimately override the preferences of individuals or the citizenry in the name of pursuing their good.
If you are my pension fund manager, then you act as my agent. I may choose to give you discretion over some parts of my affairs , but not over everything. While I am out of the country and can’t be contacted, you may morally be permitted, if market conditions change suddenly, to buy and sell to change the balance of my portfolio in what you see as my best interests. Why? Because I may have given specific consent – at least tacitly - to you to do precisely that, because the contract is for a service that is perfectly legal and does not harm anyone else. Moreover, provided your trading on my account is done in good faith in pursuit of my interests, even some choices of shares or bonds that later turn out to be poor ones won’t turn your actions into ones that do me a moral wrong. Even if we see governments as the agents of individual citizens, this principle of consent does not justify a great deal of what is done in the name of managing risk, changing cultures and specifying effective services.
Can we extend the consent principle in any way that would justify government acting to pursue what it believes are out interests?
One principle that now commands wide acceptance is that governments can be permitted to act in good faith for the best interests of their citizens, where specific explicit or tacit consent has not been given, if those citizens would consent to the activity, if they were aware of all the relevant information and circumstances, given the particular preferences, values and commitments those actual citizens have, and provided the activity is the least invasive way to achieve the goal that would be consented to. I shall call this the extended consent principle. The argument is that government is justified in not waiting to act until the citizens have been given that information and given explicit consent, if and only if the consequences of waiting that long would be very damaging to the interests of the citizens – for example, by causing such strain on the public purse or valued services that no citizenry would want to put up with – because the contract between government and the electors is one which delegates that kind of power to act to prevent dire outcomes of this sort.
I don’t have time today to justify this principle fully, but I hope that we can accept for the present, at least for the sake of the rest of the argument. We next need to ask how far does this justify governments in the three kinds of projects I have outlined, which are so central to contemporary political life?
Let us take them in reverse order, and begin with services. It is true that the public is sometimes attached to particular activities or inputs which are not effective. If we are to offer preventive, holistic services that are more effective in pursuit of health, low crime, employability and better learning, then we need strong justifications for overriding those preferences and specific limitations on how, when and under what circumstances it can be done. Does the principle enable government to provide services that it believes are more effective than those the public would, if asked, perhaps define?
If so, then several things have to be the case.
First, we must ask ourselves whether we know enough to say whether the public would consent, given their values and preferences. The public does not only have preferences about how services are to be provided. They also have preferences about the outcomes that they want government to pursue. If those preferences about outcomes are (a) settled – have been strong, stable, evident priorities for long periods (b) fundamental – appear to be foundations on which much else in public opinion is based (c) expressed or conceived basically in outcome terms and (d) clearly priorities for action. If there are such outcome preferences – and my reading of the public opinion data on the priority of tackling crime, illness, ignorance and idleness is that there are – then there is a case for overriding preferences about activities, providing other safeguards are in place, and only in ways that are otherwise acceptable (do not violate rights, etc).
Only in something akin to an emergency can there be a justification for overriding majority preferences where these conditions are not met. But this might not get us far enough for some. The challenge for the environmentalists who wish to override public love of the care and the road, when the public attaches a lower priority to environmental outcomes, is to make out the case the consequences of failure to override preferences will lead to an environmental emergency.
The second condition is about how government makes decision about what is in the best interests of citizens in the design of services. This an argument about the nature of democratic delegation. While some people regard governments as merely the slaves of current majority popular will, this is in practice nearly always set aside in actually existing democracies. Implicitly, we delegate to governments a power and duty to enquire into the causes of great evils and the best means – without violating rights etc. – of combating them, and a power and duty to act in good faith upon those findings. It is hard to make sense of the powers hat make government possible at all without such an implied power and duty. But there is more: government must exercise that power with due diligence and in good faith to find out “what works”.
Thirdly, any government that does override current preferences about services or activities must be accountable to the voters for doing so. In terms of the extended consent principle, it must make every effort to make the public aware of all the relevant information and all the relevant circumstances. Therefore, the more it innovates in ways that might run counter to majority views, the more it must be innovative in explaining and defending its case to the public and in giving the public ways to play a part in decision making and in commenting on the quality of the results achieved.
How do recent government projects stand up on this criterion? Certainly, I am persuaded that many of the more holistic and preventive services that governments around the developed world are now introducing are ones that meet the conditions of conforming to the settled, fundamental priority values. However, local authorities, health authorities, police authorities, ministers at national level have performed much less well on the measure of informing the public, explaining the case and seeking to involve them in decision-making.
I turn now to the politics of changing cultures. It is not quite so straightforward to apply the extended consent principle to justify projects to change the cultures of the public. After all, the extended consent principle requires that governments must stay close to the actual values, commitments and preferences of the public. Does it permit governments to seek to influence those values, commitments and preferences?
Well, yes, but only in certain very circumscribed conditions. The principle would allow government to seek to change our cultures, if we would consent to that, given full information, given who we are and what we fundamentally want and care about.
After all, most of us can be persuaded that we wanted was not a sensible thing to want – we can be persuaded that a service was a waste of money, was bad for us, or ineffective. But we have to be offered reasons, and those reasons have to appeal to those of our values that we see as more fundamental and more settled. Do if we are to be persuaded that we ought to develop less dependent and more enterprising cultures, or cultures of taking responsibility for our health-related behaviour or lifelong learning, then once again, government can justifiably undertake that persuasion only if it can reasonably point out contradictions between our actual fundamental commitments to outcomes we want and specific cultures we have that do not support the outcomes we want, and if it seeks to provide full information and secure legitimacy for this.
Some of the political projects for cultural change of recent decades seem to measure up rather poorly against this standard. The information has not always been provided to convince people that projects for cultural change around, for example, greater self-reliance or zero tolerance, are genuinely in the best interests of citizens, or that the abandonment of more solidaristic or rehabilitative values is the best way to pursue out fundamental and settled commitment to the minimisation of poverty and the reduction of crime.
Finally, what does the extended consent principle suggest about how far governments may go in the regulation of risks such as food hygiene or transport safety?
Here the picture is very mixed. There seems to be some evidence that, once informed of the value and efficacy of seatbelts and crash helmets, most citizens would indeed consent.
At least, it is hard to doubt the good faith of government in seeking to secure safety in these ways, the fundamental character of most people’s concern with avoiding road accidents, and the fact that governments have exercised – through many road transport research institutes work over decades – due diligence in finding out what works.
In these cases, of course, it is possible that the coercion can be justified on non-paternalistic ground, of the prevention of harm to others rather than the good of the motorcyclist, driver or passenger themselves.
Welfare to work programmes are sometimes described as paternalistic. American professor and advocate of “hassling” the poor and unemployed, Lawrence Mead, has described them as “the new paternalism”, because they involve coercion. Actually, I am far from sure that coercion in these cases is principally justified on paternalistic grounds. Typically, at least in the case of the British government’s New Deal, the justification offered is that citizens have obligations to taxpayers, in a kind of social contract, only to burden them with the costs of their upkeep in emergencies and to keep that burden to a minimum. An obligation upon citizens not to harm other’s interests is not coercion for their own good. There may be a paternalistic argument for welfare to work schemes, but if so, it isn’t being asked to do a lot of political and moral work here in Britain. After all, there are vastly more voluntary entrants onto the Gateway than there have been cases of the exercise of sanctions.
It is perhaps important, in the present context, to spell out the nature of the argument about an obligation not to dumping cost of one’s keep upon the taxpayer except in emergencies, at least while they are adults capable of being active in the labour market. It is a special case of that non-paternalistic justification for overriding people’s preferences and sanctioning their choices, that one has a duty not to harm others. If this is to justify government intervention, then government has to show that it has the right and indeed the duty to enforce that particular obligation upon citizens – after all, citizens may have many moral obligations (to do some volunteer work, to make some charitable donations, to save money for their and their children’s future), but the state may have no right to enforce such obligations with force of law. Only where the harm created by leaving the obligation undone would be as severe as that from a crime or a civil tort against fellow citizens, can the state justifiably claim that it may use the coercive power of law to sanction citizens for the obligation. The question here is whether remaining on benefit when one could find a way to leave it, is of that order. Of course, just because it is harm against a great many people collectively rather than an individual does not make it as lesser offence: we sanction fraud against the state, negligence that exposes many unnamed people to risk, and many other things that affect only “statistical lives”. The particularly relevant analogy, as I understand the argument by New Labour and other governments pursuing similar policies, is fraud. The argument is that preferring to remain on benefit at the taxpayer’s expense when one could find another way to be supported, is akin to defrauding the taxpayer, because one is violating the conditions in the social contract between taxpayers and beneficiaries, and to do so, one must inevitably misrepresent the facts of one’s case. On that basis, it is argued, the harm can be seem to be so severe that the government is justified in using the force of law to sanction the behaviour and enforce the obligation. In the same non-paternalistic way, we legally sanction racial or sexual discrimination, not for the good of the discriminator (which would be paternalistic) but to protect victims from a harm that we now agree – although until relatively recently majorities appeared not to – is of a severity comparable with many crimes and torts.
I, for one, find the argument reasonably persuasive about welfare to work schemes. But how far can it be extended? If falling ill, as a child or an adult below pension age and using the National Health Service is an emergency, is the comparatively more predictable onset of frailty in old age an emergency, or would the principle of not harming taxpayers result in state compulsion for the costs of long term care? Again, in many cases of health-related behaviours, when paternalism is set aside, the others who may harmed, however, are principally taxpayers, who may have to pick up the bill for the medical care.
Moreover, now that the NHS is seeking to recover these costs from the vehicle insurance plans of the individuals’ affected, the ground for government paternalistic justifications for regulation of health-related behaviours such as speeding or drink-driving may be slightly undermined, and the control of moral hazard become a private contractual matter between an individual driver and his or her insurance company. However, the extended consent principle states that, if individuals would consent to the regulation, and many, it seems clear, would, it is permissible even if there are additional contractual requirements upon individuals for any private insurance they may have covering such risks.
On the other hand, the beef on the bone ban seems to fail several limbs of the test. It is far from clear that citizens would consent if we were told the full information available to ministers at the time it was imposed. The ban may have been at least as much motivated by the demands of the government's negotiating strategy in Europe as by the considerations of public health. "Such evidence as the government scientists offered might have justified a paternalistic health warning or the publication of information for consumers to make up their own minds, but not a paternalistic ban.
In the last fortnight, there has been a huge moral panic in the newspaper about genetically modified foods. William Hague has declared his party to be in favour of an outright ban, and the Prime Minister has expressed alarm too. I don’t want to spend any time tonight on the pseudo-science peddled by many of the alarmist writers, or to devote any time to the positive environmental benefits that might flow from getting genetic modification right, or the possible positive impacts on nutrition in developing countries. However, I will just say that some of the causal mechanisms of gene flow between species they claim to be worried about are, for all practical purpose, biologically inconceivable, and I do want to point out that no one has ever come up with a single case of actual harm to human health from any of the mainstream genetically modified foods in the supermarkets. Fundamentally, however, I do want to suggest that we need to think long and hard before we rush to banning anything. On the extended consent principle, there are good reasons for giving consumers information on food labels about what they are choosing, and letting them decide for themselves. But there is no possible justification for politicians claiming to know better than I or any other shopper or diner does what kind of food we want to eat, and banning things that many of may well want. That is paternalism too far, whatever the panics about the supposed fragility of nature.
When paternalistic laws on individual risk behaviour become “dead letters”, the extended consent principle gives strong grounds for thinking they ought to be repealed, unless government can recover legitimacy for them. For if large sections of the young population flout, for example, the prohibition on the recreational use of cannabis, and their behaviour and values suggest that they would consent to prohibition, even if given all the information available to Mr Keith Helliwell, then government needs to be able to show some kind of emergency reason if it is to justify continued prohibition. The prohibition on heroin and cocaine appears, however, to be rather different, since large numbers of young people, including many users, might consent to that if fully apprised of the facts.
Some paternalistic activities by government may be justifiable, even on paternalistic grounds, but too often, politicians of left and right have been cavalier of the limited and specific ground on which this is morally acceptable in their attempts to regulate risk, change cultures or specify new kinds of services. Meeting the standards of the extended consent principle is a demanding task, and one that politicians and public officials at every level need to take seriously.
Paternalism is suspect, and requires special justification, because we generally consider that adult citizens are competent, more or less rational people whose preferences, cultures, commitments and ideas of the lives they want to lead and the risks they want to run are valuable and deserve respect. They have, we presume, the capacity for reasoning, for collecting information and making decisions that better reflect their idea of the good life, than the decisions government can make on their behalf. If my argument is accepted, it is this basic value that prevents us from going further than the extended consent principle. To those persons who, for whatever reason, do not possess that general competence to enable them to form autonomous conceptions of the good life and rationally to make decisions about how to live it, we have different duties. If they are children who are expected to develop that ability, we owe protection and a duty to nurture that ability. However specific competent adults, principally parents or guardians but also including teachers and sometimes government officials if things go wrong, have authority act paternalistically in respect of children.
I want to conclude by suggesting that the extended consent principle not only provides us with an account of the limits of morally permissible state benevolence, but with the basis of an account of a key obligation upon both governments and citizens.
Citizens who are competent, autonomous, rational people with settled fundamental preferences about the outcomes they want to see pursued in public life do not fall naturally from the womb. They are made. More specifically, they are educated, socialised, nurtured, enabled, by specific institutions – parenthood or guardianship, education, networks of friends or relatives, mentors, clubs, companies, and so on. A society that fails to ensure that nurturing and education for competence is surely one in which both citizens and government are failing in their obligations.
A society that turns out such large numbers as ours does, of young people who are functionally illiterate, or who are so affected by social exclusion that they have become too cynical, apathetic or uncaring about their fellow citizens, their own futures or their society that they are unable or unwilling to act as competent citizens, is one in which government and many other institutions are failing in their moral duties. While we are too paternalistic in one direction, the very principles by which we limit paternalism require us to act to bring up and make citizens able to act as such. Not to achieve that is perhaps an even more severe moral failure than the over-regulation of risk, over-zealous promotion of a favoured culture or technocratic insistence upon a particular service.
Perri 6 is Senior Research Fellow at the Department of Government, University of Strathclyde and Senior Research Fellow, Demos