The day before beginning this chapter I talked to a woman with little sympathy for those who opt out. When I said that I thought a parent opposed to private schools might be justified in sending her kid to one, she scoffed: 'That won't help solve the problem, will it?' I replied that sending him to the local comprehensive probably wouldn't help much either. The thought that individual sacrifices are futile has been lurking in the background for a while. It's time to put it under the spotlight. Is it true that the decisions of individual parents make no difference? How does the answer affect the issue of hypocrisy? If it is true, does that justify opting out?

Justification 16: Sending him to the local comprehensive won't achieve anything

Let's start with the easy bit. We can see how the insignificance of individual choice might get parents off the charge of hypocrisy by combining two points that have already been discussed and are in any case fairly obvious. First point. There's no way that sending one's kid to the local comp will make as much contribution to the realisation of educational justice as casting the decisive vote. On any view, the justice-gain from sending him there will be less than that achieved by changing the rules. Second point. The downside, for him, of sending him to the local comp given that other kids are opting out may well be greater than it would be if nobody was allowed to opt out. The reason to change the rules is precisely that doing so would make things better - absolutely and positionally - for those who go to comprehensives.

So the disadvantage inflicted on him by being sent to the local comp here and now - a worse educational experience and unfairly poor competitive position - may be worse than the disadvantage (if any) inflicted on him by changing the rules.

Changing the rules and making choices within the rules have different effects - for society as a whole and for the individual child. Change the rules and you get a big pay-back in terms of justice at smaller cost to your child. Send him to the local comp and you get, at best, only a little bit more justice, possibly at much greater cost to him. This is one way to avoid hypocrisy. 'Yes, for lots of justice I'd be willing to make my kid a bit worse off than he could be. But that doesn't mean I'm willing to make him a lot worse off for a tiny increase in justice.' This line of reasoning holds even if sending my kid to the local comp would indeed contribute something to justice. Of course it must also hold if sending him there would have no beneficial impact at all.

But that was the easy bit. There are harder issues. Some are empirical. What are the consequences for educational justice of your individual choice about your child's schooling? Perhaps sending him to the local comp will make no difference either way, perhaps it'll make things better, perhaps it'll make things worse. (Justification 17 appeals to that last suggestion.) The answer to that, of course, is that it depends. Another empirical question. What are the consequences of your indi-vidual choice for your child? Perhaps sending him to the local comp will be the best thing that could happen to him, perhaps it would be a disaster. That too depends. And then there is the moral question, about justification. Suppose you have a best possible estimate of the consequences, for him, for others. What decision are you justified in taking? On that one, I'll suggest that you should be willing to bear your fair share of the cost to bring about your fair share of the gain. But let's begin with the empirical complexities.

The justice-consequences of individual choices

Sending your child to the local comprehensive will not achieve much of the justice that would be achieved by changing the rules so that everybody has to do the same. But the fact that something is small relative to the total picture doesn't mean that it is morally trivial or irrelevant. If I have more money than is fair, and you have less, then I make things better by giving you some of what I've got. The world isn't as fair a place as it would be if everybody did the same, but it's fairer than it was. Still, the case of education is importantly different. There is more of a gap between the individual action and the justice-effect. The causal process by which the action of sending one's child to the local comp contributes to justice is more complicated. And the contribution may be even smaller.

To see why, let's begin by remembering that there are two different ways in which private or selective schools make things unfairly worse for those who don't go to them. These correspond to the two ways in which education is valuable: intrinsically and instrumentally. On the one hand, in intrinsic or absolute terms, creaming off bright and motivated children, siphoning off the energy of their parents, and attracting away the best teachers, results in a poorer education for the rest. Peer group and related effects mean that permitting private and selective schools has a depressing effect on what goes on elsewhere. On the other hand, in instrumental or positional terms, the very fact that some children are acquiring an unusually good education disadvantages the others. Even if one rejected all the evidence about peer group effects on educational experiences, it would still be the case that letting some kids get a better education than others worsens the competitive chances of the latter. In absolute terms, some children get an intrinsically worse education than they would if there were no private or selective schools. In positional terms, the instrumental value of some children's education is reduced by the fact that others are getting a better one. Suppose these various mechanisms do indeed explain how private or selective schools make things worse - and unfairly worse - for those who don't go to them. If Part I is right, that's partly why we should vote to get rid of them. The question is this: do these mechanisms work in such a way that an individual parent's sending her child to the local comp will make things any fairer?

Start with the positional aspect. You are a parent deciding whether to send your son to the local comp. To focus on the precise issue we're interested in, imagine for now that sending him there will make no difference either way in absolute terms to the education received by the other children at the school. You want to know whether sending him there, rather than opting out, will do anything to reduce the positional disadvantage suffered by those who go to the local comp. You know it won't do much, but will it do anything? In one sense the answer must be yes. Other things equal, pushing one's own child up the queue means pushing everybody else down it. But things are complicated. If you do go to the local comp, the private or selective school may simply take in another boy instead of yours. In which case, you might think, the net result will be that somebody else's son gets the better education, yours gets the worse one, and everybody else is left exactly where they were before. Their place in the queue is unchanged. All that's happened is that your son has swapped places with someone else.

This suggestion that individual choices to opt out have no net effect on the positional disadvantage of those left behind should be treated with suspicion. The idea that another child will simply substitute for your own might make sense if we think of the supply of places as fixed. But it isn't - at least not in the case of the independent sector, which has grown every year since 1994. Even if it's true of any particular private school that it can take only a certain number of pupils each year, it's certainly not true of the sector as a whole. So you should think of your decision as making a difference to the total number of children opting out, hence making a difference, positionally speaking, to those who don't. With state selective schools, where the total number of places available may indeed be fixed, things look different. There it could make sense to regard your child, and the one who would take his place if he didn't, as equivalents: who takes the place makes no difference to anybody else.

Though buying your son an education better than others are getting will have a negative impact on them, positionally speaking, the impact will be tiny. One place up or down a very long queue is going to be of much less significance than the contribution to justice represented by the millionaire's giving away his unjust excess to the poor. It's the individual child's contribution - or lack of it - to the absolute quality of other children's education that people tend to have in mind when they think about this justification. So let's turn to the two key mechanisms by which private and selective schools make things worse, in absolute terms, for those who don't go to them: peer group effects and parental energy. Thinking about these in turn, we need to consider whether your action of sending your son to the local comp is likely to improve things for the other children attending it. To focus on a particular case, let's assume that your son is bright and well motivated. (If he's a troublemaker then the way to improve things for others may be to take him away!)

The super-optimistic view would be that your son's presence, on its own, will do something to improve the local comprehensive. A more plausible, but still optimistic, view is that your decision to send him there will itself lead to other parents with similar children making a similar decision, which in turn will lead to more others doing so, and so on. You will begin a benign snowball, and a snowball big enough to make a difference. On more pessimistic scenarios, there is no snowball. Your decision results in one miserable child and no impact on others at all. Clearly, very specific empirical circumstances are going to be crucial. The chances of a benign snowball are going to depend on many different factors: how many other parents in the local area have bright and well-motivated children, how many of them are likely to be influenced by your choice. This will in turn depend on the kinds of networks and modes of communication between parents as well as their beliefs about the likely effects of particular size snowballs and their moral beliefs about the desirability of various outcomes. How many children who might otherwise have opted out are needed to make a difference - how big a snowball is required to achieve something - will itself depend on the characteristics of the school. A single bright and well-motivated child surrounded by demoralised teachers and uninterested classmates might simply become very unhappy and not produce any beneficial results. But a single bright and well-motivated child in a class with a wide range of children and a committed teacher could indeed make a difference - perhaps inspiring just one other child to learn to read who would not otherwise have done so, who in turn shows others that it can be done, and so on. In these various ways, then, the total effect of your decision will be highly context-specific and sensitive to empirical circumstance.

One shouldn't be too hopeful about the benign effects of one kid in a hostile classroom. Nor should one expect that others will somehow notice one's private decision and imitate it. But the prospects for starting benign snowballs can be better than many people realise. Communication is crucial. I have a friend who was worried about the prospect of his son going to the local comprehensive. The primary school had been fine but the secondary had a poor reputation. He organised a meeting at which parents of kids in his son's class could get together and talk about the problem. They ended by agreeing to send their kids to that local school. Confident that others would if they did, that there'd be a critical mass of like-minded parents and similar kids, suddenly that looked a much less risky option. Collective action doesn't have to be political in any formal sense, coercively imposed by the state. If conditions are right, individuals can get together and voluntarily agree to act together, thereby realising a desirable goal not available to them if they act merely as individuals. Parents shouldn't be too quick to judge that they cannot make the relevant difference. The man who organised that meeting did.

I said that one shouldn't expect others to notice one's own decision. That's true of the standard case, but the consequences that follow from an individual's choice for her child depend partly on its symbolic significance. As discussed in Chapter 7, Harriet Harman's decision to send her son out of Southwark will have done more damage to the cause of educational justice than would the same decision made by somebody else. 'If she's not willing to give it a go, why should we?' I know that local public figures - such as clergy - are sometimes particularly reluctant to opt out of the local school, just because they - rightly - feel obliged to factor into their decision its exemplary aspect, the way in which their decision may influence other people's. Snowballs can be malign too.

The siphoning off of parental energy and influence is bad for comprehensive schools at two levels. Supporters of state education complained that so many members of the last Conservative cabinet sent their children to private schools. Their concern was at the national level. The worry was that those in power were less likely to act in the best interests of children attending state schools than they would have been if their own children had been doing so, less motivated to care about the education of the mass of the population whom they were elected to serve. Or they would be ignorant about the actual condition of state schools. By changing the rules to require all citizens - including decision-makers and opinion-formers - to participate in the state system, one could expect to harness their self-interest to the interests of the population as a whole, or at least to eliminate their ignorance. Other mechanisms operate locally. The parents of children attend-ing a school make a difference to the quality of the education it provides, through financial contributions to buy new equipment, participation in school activities, membership of the Board of Governors or Home- School Association, or simply moral support for the teachers. Parents lucky enough to have a choice about their children's education are particularly likely to be able to help the school their kids go to. They can contribute money, equipment, non-working time devoted to school-based activities, possession of relevant organisational skills, and perhaps political influence at the local level.

Can the individual make a difference? It depends. It might be that one extra energetic, interested and influential parent makes no difference at all; that a critical mass of such parents is needed to have the beneficial effects. So, again, one has to attend to the fine detail of the concrete situation. How many parents are already contributing in this way? What are the chances of the individual contributing to a snowball? Zealous types who ring round, twist arms and generally apply moral pressure to get other parents involved in activities to support their children's school are a pain. But they play a crucial role in sustaining desirable levels of parental input. Clearly much depends on the amount of the contribution that the individual is prepared to make. A mother who gave the local comp as much as she would pay in school fees might, on her own, without any snowball, make a substantial difference.

Unlike peer group effects, the beneficial effect of your input, as a parent, is not necessarily gained by the school to which you send your child. There's nothing to stop you sending your son to one school and devoting your energy to a different one. I'm not saying that's easily done. It is surely harder to devote one's energies, interest and influence to a school not attended by one's children. But there are thousands of unfilled places as school governors, nearly all in relatively disadvantaged areas. Parents who believe that they can't, as individuals, make a difference to educational justice are deceiving themselves. At the moment, of course, the schools that benefit most from parental input are those that need it least.

There may indeed be circumstances in which sending one's child to the local comp will not contribute anything to the realisation of those values - educational justice, equality of opportunity, and improvement in the education of the worst educated - that would be promoted by changing the rules. If he would be swamped by hostile classmates, if sending him there would lead either to no snowball or to one too small to make a difference to the quality of the education. If one's own energy, interest and influence would be utterly futile on its own and wouldn't produce a snowball (or, more optimistically, if one's input to the local comp did not depend on one's own child going there). But these are atypical situations. In the usual case, an individual's choice could be expected to make some difference. Given not completely inauspicious classmates, he might broaden the horizons of some, yield some crucial job-satisfaction to a teacher who was on the point of giving up. And individual parents can spark collective action. If information flows are good, other parents' knowing that one will send one's kid to the local school if others do the same can make it happen that others do do the same.

There is a lot of uncertainty around. Parents simply have to judge, as best they can, the likely justice-effects of their school choices. The same goes for the effects on their own children. I've already talked a lot about what makes a school 'good enough', and about the different ways - positional and intrinsic - in which education might be judged valuable. The same school that would be disastrous for one child might be perfectly OK, perhaps even ideal, for another. The cost to the child of going to the local comprehensive may be huge - or there may be no cost at all.

It's probably worth emphasising that there are two different 'costs' we might have in mind here. One is the cost of sending him to the local comp compared to opting out. This is a measure of the gap - in overall 'goodness for your son' - between the local comprehensive and the alternative options currently available. (Recall, from discussion of Justification 4, that some parents may think that what kids get from going to a 'normal' school is so valuable that it outweighs the 'narrower' educational advantages enjoyed by those who opt out. For them, the child suffers no cost, all things considered, by going to the local comp. Quite the reverse.) The other is the cost of sending him to the local comp compared to the local comp as it would be if nobody could opt out.

This is a measure of the gap - still in overall 'goodness for your son' - between the local comp under current school rules and how it would be under fairer school rules. (Here it's worth mentioning that some comprehensives under the current rules may be better than they would be under fairer rules. Abolishing private or selective schools could be expected to improve comprehensives overall. But fairer mechanisms allocating kids to comprehensives, or distributing resources between schools in ways that mitigate the unequalising effects of residential segregation, could be expected to make those comprehensives more equal than they are at present. I don't know how these two effects would play out, but it could be that some comprehensives would end up worse, all things considered, than they are now.)

When a parent thinks about the costs and benefits of sending her son to the local comp - to him, to others - she should have in mind the first of these. The issue is what net difference, if any, is made by sending him to one school rather than another. Thinking about that means taking the existing rules as given. But the other cost is relevant too. To see why, we need to move on from these empirical questions to the moral one.