'Even in a truly liberal society, paternalism must sometimes prevail'

2005: General election year - Should the state control private behaviour? It's the biggest argument

Tessa Jowell's complaint spoke volumes. "I seem to have gone from 'the nation's nanny-in-chief' to a 'gambling gangster's moll' in a few weeks," she told the Observer. Labour ministers, including Tony Blair himself, are accused one day of gross invasions of privacy and liberty - in areas such as smoking, smacking and hunting - and the next of reckless deregulation and liberalisation, regarding licensing hours, soft drugs or gambling.

The tension between the Two Tessas reflects a tension that runs through the heart of politics. Many of the issues that will animate political debate in this general election year cross conventional dimensions of left and right and even of party-political allegiance. They seem to belong on a liberal v authoritarian (or paternalist) axis, not on a left v right one. Even apparently dry economic matters, such as pension reform, have at their heart the question of how far people should be compelled to act in their own best interests.

Labour has failed to produce a sound philosophical rationale for its policies. The best Jowell could manage (which is better than any other minister) was to point out that children need special protection; regulation must be case-sensitive; politicians should prefer dialogue to diktat. All true, but not very illuminating.

Most British politicians are all over the liberal-paternalist map. Plenty of Tories turn pink at a ban on hunting, but are happy to ban gay sex for 16-year-olds. Plenty of Labour MPs who cringe at restrictions on smoking would be happy to compel higher savings. The second reading of the Gambling Bill brought strong responses from old Labour warhorses and shire Tories alike. "Gambling does not do individuals any good, nor is it in the common good. It . . . sucks in and makes more destitute the poorest in our society." "Does not [Tessa Jowell] see there is a world of difference between ordinary folk who want to go to bingo halls and the proliferation of mega-casinos?"

The first quotation is from Jonathan Sayeed (Tory), the second from Donald Anderson (Labour) - but who could tell the difference? Many Tories who voted against the government would have preferred to be in the other lobby; 29 Labour MPs joined them; and there were scores of abstentions.

Perhaps the biggest vacuum in contemporary politics is a clear, consistent liberal voice. The "orange book" Liberal Democrats such as Vincent Cable and David Laws provide a modest exception - but they are on the edges of the third party. And we know party labels are out of date when the Liberal (sic) Democrats are the only party urging more compulsion in pensions saving. The few remaining Tory liberals are disorganised and/or quiet. And while Labour has traditionally found room for a robust liberalism - remember Roy Jenkins's tenure as home secretary - there are few gut liberals on the front bench today. "In the blood of the socialist," advised the postwar Labour thinker Tony Crosland, "there should always run a trace of the anarchist and the libertarian, and not too much of the prig and the prude." It is difficult to detect the anarchist streak around today's cabinet table.

This absence of a muscular liberalism is part of the reason for Labour's philosophical muddle. In the course of an argument on these matters, either paternalism or liberalism can legitimately triumph. The problem is that the argument is not taking place, so that a muddled or deliberately obfuscating government often ends up dressing a paternalist policy in tatty liberal clothes.

In On Liberty, John Stuart Mill offered the best-known distinction between liberalism and paternalism: "The only purpose for which power can be rightfully exercised over any member of a civilised community, against his will, is to prevent harm to others. His own good, either physical or moral, is not a sufficient warrant . . . Over himself, over his own body and mind, the individual is sovereign."

This "harm principle" goes to the heart of debates about smoking. The only liberally acceptable grounds on which smoking can be prohibited are to prevent harm to others - which is why the evidence on passive smoking is so important. Though the evidence seems clear to some (including the editors of the British Medical Journal), the argument is not won in scientific circles. Philip Collins, director of the Social Market Foundation and a rare liberal, says: "There is a WMD feel to the passive-smoking debate. The evidence for the health influences of passive smoking is piling up - but it is not yet strong enough to state categorically that harm is occurring." When the evidence is sufficiently strong, says Collins, the liberal argument against a public ban falls away.

But public policy frequently deviates from the harm principle. Take the seat-belt laws. From a liberal perspective, it should be mandatory to wear belts in the back seat - because in the event of a crash you can harm the person in front - but voluntary in the front, where the only harm from being hurled through the windscreen is to yourself. Yet the seat-belt laws were introduced in the reverse order. For front-seat drivers or passengers, they are pure paternalism. However, I have yet to find an MP - even among those most against the "nanny state" - who will call for their removal, perhaps because more than nine in ten of the population now support the law.

And the harm principle - sound though it generally is - does throw up two questions. First, what constitutes "harm"? Second, which "others" qualify for protection?

Mill is pretty clear that harm has to be direct, demonstrable and - almost without exception - physical. Stress, envy, disappointment, disgust or inconvenience do not count. One modern adaptation of the harm principle introduces a category that might be called "fiscal harm". So the grounds for intervening to stop people smoking or getting fat are that they impose a financial burden on the National Health Service, which in turn has a harmful impact on the wallets of non-smokers or slim people. Even if empirically true, this is an entirely bogus argument. If behaviour that costs the NHS money counts as harmful, then we have no option but also to tax skiing, rugby and possibly suicide attempts. Fiscal harm is a false argument used to disguise paternalist intent.

The question of which "others" qualify for protection from harm is more difficult to answer. Historically, members of your own immediate family did not count as "others" - which was why, until very recently, it was legally impossible for husbands to rape or assault their own wives. Now, individuals within the family, and especially children, are accorded similar protection to those outside: the government is limiting physical punishment of children by parents to "reasonable" levels of force.

Indeed, the liberal case for intervention in the privacy of the family may be stronger than that for intervention in public places. There may be some doubt that the stranger next to me in the bar will be harmed by my smoke during the half-hour he sips his beer. There is very little doubt that my children will be harmed by my 60-a-day habit, enthusiastically maintained in a tiny flat or small car. Why is the former to be banned and the latter permitted? If the government is willing to protect children from being struck hard by their parents, on what grounds - given its acceptance of the research on passive smoking - can it withhold from children protection from poisoning by tobacco smoke? As Collins points out, the ban on smoking in public places in Ireland has increased the amount of socialising at home, where the damage from passive smoke may be greater, especially for children.

The editorial writers at the Economist - who should know better - say that "preventing people from puffing away in their living rooms would involve a level of illiberalism that not even this government is prepared to countenance". However, if harm to non-consenting children is clearly established, such a ban could be seen as "illiberal" only if the family, or the physical space of the home, were seen as sacrosanct spaces. And, leaving aside political practicalities, it is not clear why they should be.

One further problem with the current liberal-paternalist debate is that it is solely concerned with the government's role. This misses the role of non-state institutions such as churches or companies as well as friends, family and neighbours. It isn't just the state that can play nanny. Mill identified two methods of policing behaviour: "physical force in the form of legal penalties, or the moral coercion of public opinion". It is a fair reading of Mill that he believed most harmful acts should be prevented by law, while lesser harms were best discouraged using "moral disapprobation in the proper sense of the term".

Yet moral coercion isn't what it used to be. This is the result not only of the decline of non-state moral actors such as churches, but also of our wish to be free of the tyranny of prevailing opinion. There are obvious advantages in no longer having to worry about disapproving neighbours or what you do in front of the vicar. But moral coercion has a positive side, too. It is right that a father who fails to meet obligations to his children be stigmatised as a "deadbeat dad"; that youngsters who fail to yield bus seats to frail old people be condemned by fellow passengers; that a mother who spends her pay in the pub while her children go hungry be shunned by her peers. Stigma is underrated as a positive social dynamic.

Moral disapprobation has diminished for two reasons. First, we feel we should not interfere in the business of others: the modern mantra is "each to his own, I suppose". We lack the personal courage to confront the person dropping litter, leaving dog excrement on the pavement or abusing passers-by. So "antisocial behaviour" has been recast as a responsibility of the state, rather than the community. Antisocial behaviour orders (Asbos), parenting orders and litter police replace the sense of collective responsibility that our insistence on individual freedom has eroded.

Second, anonymity has expanded. Detach people from their families and communities, and the barriers to "bad" behaviour fall away. Infidelity soars at conferences held in large hotels, and internet pornography has surpassed the wildest expectations. (The internet, indeed, provides the ultimate in anonymity: in cyberspace, no one even knows your name.) A man who would be reluctant to buy Hustler in his local shop happily does so on a distant garage forecourt or at an airport news-stand. W H Smith stocks up to seven adult magazines at its railway station and airport branches but, after trials, gave up trying to sell them elsewhere. The official reason is that the "profile" of travel store customers is "more mature, more cosmopolitan". The real reason, however, is the warm cloak of anonymity.

This bears directly on the government's gambling proposals. The problem with mega-casinos is that the punters are strangers: strangers to the croupier, strangers to each other, perhaps even strangers to themselves. As the Casino Operators' Association argues, "small casinos have a hands-on, detailed relationship with their customers" and this could explain "the lower incidence of problem gambling in current UK casinos". People should be free to gamble, but in environments most conducive to responsible gambling. Far-flung, giant casinos, where no one knows anything about you and the booze flows freely, do not meet this requirement. On these grounds, I think, Mill - who had no beef with gambling in itself - would have opposed the current changes.

A liberalism that puts the harm principle at its core, which sees those within family units as being equally deserving of protection to those without, and which places as much emphasis on social as on legal coercion, is a powerful political philosophy, as relevant today as a century and a half ago. However, three weaknesses in classical liberal thought must be addressed.

First, liberalism tends towards a crudely binary view of the world. Either an action is harmful to others, in which case it can be restricted, or not, in which case it cannot. There is no place in Mill for using tax to discourage consumers from harmful actions: "every increase in cost is a prohibition", he writes. Imposing a congestion charge on drivers, or a higher tax on leaded petrol, would not be acceptable to pure liberals, despite the harm caused by pollution. Their view would be that, if harm is proven, a ban should follow. In the real world, however, the government often has to use cost not as a prohibition, but as a prod in one direction or another. Driving is bad, but not bad enough to ban.

The second problem is that pure liberalism struggles to differentiate between freedoms, so that the freedom to vote is given the same weight as the freedom to smoke. A mature politics requires us to recognise that values sometimes conflict - and that no single value holds a trump card.

Collins, for example, supports the law requiring drivers to wear seat belts, even though it transgresses pure liberal principles. "We are dealing here with a relatively trivial freedom, which is outweighed by the clear benefits which flow from its removal," he says. "To accept this is not to give up the central principles of liberalism: it is to recognise that they should not always prevail."

The third and perhaps biggest weakness in liberalism is its failure to recognise the weakness in us. Liberal theorists have an optimistic view of our capacity to resist temptation, seek out worthy goals and make good choices. But most of us are weak-willed, short-termist, prone to prevarication and procrastination, irrational, ignorant, forgetful and foolish. It therefore behoves institutions - including the state - which are in possession of good evidence on the benefits and costs of certain activities to promote, discourage and sanction accordingly. Should the state make me wear a seat belt? Yes, because I am predisposed to think car crashes are less likely to happen to me than they are, because I am too lazy to put on the belt unless there is a potential loss from not doing so, and because it soon starts to feel more like a reflex than a rule. And because it might save my life. Should alcohol, tobacco and junk food be discouraged, using taxation as well as education? Yes. And is all of this paternalism? You bet. So let's be honest about it, rather than find ludicrous alternative arguments such as the white elephant of "cost to the NHS". Better honest paternalism than fake liberalism.

On one day, politicians may need to defend fiercely the rights of individuals to do as they wish, however ill-advisedly; on another, to protect one individual from being harmed by the actions of another; on a third, to protect individuals from themselves. We have to recognise that, even in a truly liberal society, the paternalist argument will sometimes prevail.