Archive for October, 2010

Why I am not a Libertarian (Part One)

As someone who favours free markets and a generally liberal social policy, I sometimes get tagged as a libertarian. This is to some degree understandable – after all, this is a combination of positions associated with libertarians, and people like to use these ideological markers as reference points. But although I may share some of their conclusions, I absolutely reject their philosophy, and hopefully I’ll go some way towards explaining why.

Firstly, I am not an American. While Libertarianism is in theory a universalist belief system, in practice it is a mostly American movement, and its overseas followers are generally more American than the Americans themselves. This is not intended as a dig at the likes of Daniel Hannan, whom I admire greatly, or indeed the USA, which I think has a great deal to recommend itself as a country. But every broad movement has not just a structure, but a mythology, which you must buy into at least on some level. And the mythology of libertarianism is one of American exceptionalism, and in particular of the infinite wisdom and brilliance of the US constitution – obviously ideas with more appeal in the USA than elsewhere.

My objection isn’t so much that these ideas are false – of course they are, but every mythology is false. Hannan, for example, may claim that the USA has a “sublime set of institutions designed by exceptional men”, but I don’t think he’d defend the proposition too seriously. The electoral college, for example, was clearly broken as early as the 1790s, the setup of Congress leads to endemic dysfunction, and the American political structure he likes best is the open primary, which is a 20th century innovation and nowhere in the constitution. Moreover, although he (rightly) emphasises the conservative aspect of the American Revolution, it was also a radical revolution, as much 1641 as 1689, as much Rousseau as Montesquieu. It was both conservative and radical in the same way as Barack Obama promotes himself – dyed in a long tradition of radicalism that has become the establishment and so conservative, promising both to protect traditional ideas and to make transformative changes.

I don’t know Daniel Hannan personally, but I suspect he would concede all this. “Yes,” he would likely say, “the American constitution has many flaws – it has 27 amendments for a reason, and it could no doubt be amended further to advantage. But no constitution is perfect. It is an amazing document, and worthy of our celebration and respect, and we should use it as a guide.” Well, OK. But the Mappa Mundi is also an amazing document worthy of celebration and respect, but you wouldn’t use it as a guide. It’s an interesting artifact of its time, but it’s just a terrible map, and far from being the basis of serious cartography it’s basically a shambles.

And this is the basis of my disagreement with the mythology – not so much that it’s false, but that it’s dangerous. I side more with Thurgood Marshall, when he said “[I do not] find the wisdom, foresight, and sense of justice exhibited by the Framers particularly profound. To the contrary, the government they devised was defective from the start, requiring several amendments, a civil war, and momentous social transformation to attain the system of constitutional government, and its respect for the individual freedoms and human rights, we hold as fundamental today. When contemporary Americans cite ‘The Constitution,’ they invoke a concept that is vastly different from what the Framers barely began to construct two centuries ago.” Hannan’s “sublime ideals” were far from sublime to women, blacks and Indians. Those institutions were incapable of guiding the country, which descended into civil war because the institutions were inadequate to resolve internal differences. Blacks suffered Jim Crow for generations because the institutions were inadequate to enforce the 14th amendment. The federal government has swollen with huge unchecked powers over the past 75 years, because the institutions are inadequate to enforce the 10th amendment. And so on. The constitution has failed, repeatedly, in its essential functions.

Nor has it been in any way fixed. Marshall said later in that speech that “‘We the People’ no longer enslave, but the credit does not belong to the Framers. It belongs to those who refused to acquiesce in outdated notions of ‘liberty,’ ‘justice,’ and ‘equality,’ and who strived to better them.” But this too is dangerous nonsense. The whole reason that there is such a movement to venerate the original Constitution is that the likes of Marshall drove it into a swamp, resulting in the unchecked and unbalanced leviathan that is USG today. And yet the very fact that the likes of Marshall were able to do so is no credit to the original Constitution – had it functioned properly, they would not have been able. This indeed is the divide between the two wings of American politics – the Tea Party, who worship a failed past, versus the liberal elites, who worship a failing present.

I wish libertarians would think more about how this mythology comes across to others, because I am sympathetic to many of their goals. But when they express their political nostalgia, it’s creepy. At all the times in US history when the government was limited in the ways libertarians like, it was truly terrible in other ways. Taken as a whole, Libertarian mythology communicates that its followers only care about rich white males. Of course, I am not saying that Arnold Kling (for example) is racist, sexist, or anything of the sort. I am saying that he allows himself to seem that way to others. He, and other libertarians, would be better off abandoning that mythology and crafting a new one, based not on a past that never was, but on a future that might yet be, which combined the limited government of the past with the legal equalities of today. I understand why libertarians indulge themselves in the past, but to the extent that they do so, I’m uncomfortable being around them.

You might say that I am not making serious arguments against libertarianism as a political philosophy, merely social and aesthetic arguments against the libertarian movement. But I think this is a false distinction. Political positions and the movements behind them are essentially the same thing as far as I am concerned – politics is not a theoretical exercise in an ivory tower, it is the practical business of forming coalitions and choosing our leaders. Social and aesthetic arguments affect that practical business, hence they are relevant – and the mythology of a movement is particularly relevant, since that mythos is frequently more important to the followers than many concrete policy positions.

But if you find this kind of argument unsatisfactory, next time I will be addressing more functional disagreements.

Advertisements

Why I am not a Multiplier

One of the catchphrases of the Less Wrong blog is “Shut up and multiply.” In other words, calculating the expected utility of a situation is all there is to say about its moral status, and although that may result in taking positions that seem horrific to you and I, we must trust expected utility calculations above our own intuitions.

Now, frankly this is an unwitting reductio ad absurdum to me, but this is not another post about utilitarianism, I promise. Instead, it’s about how we approach ethical situations more generally.

In hard science, like physics, the source of truth is experimental data. Theory is good to the extent that it agrees with the known results. In mathematics, it’s the reverse – the source of truth is the axioms of mathematics. Theory is good to the extent that it agrees with the given axioms. But in ethics, neither the axioms nor the results are given. People argue about whether we should be deontologists or consequentialists, but they also argue about whether euthanasia is immoral. So the question is, how do we proceed to build a system of ethics?

So according to Less Wrong, the answer is you start by using intuition to work out what the axioms should be, and then you can use them to decide all the results. But why should your intuition to be a consequentialist take priority over your intuition that torturing one person is worse than a lot of people getting a speck of dust in their eye? When Yudkowski says forget your intuitions, shut up and multiply, we have to ask “Is this one of those intuitions that can be wrong, or one of those that can’t?”

The other way around is to work out the results first, and then build axioms around them. This has certain advantages, in that thinking about concrete cases will result in better and clearer thinking, and you can use multiple reinforcing arguments to get to the same result. So, for example, whether you should be a deontologist or a consequentialist is a vague, waffly question. It’s much easier to argue sensibly if you ask what were the rights and wrongs of Pinnel’s Case. Unfortunately this has problems too. As we haven’t established axioms, we might be using completely unsound arguments, and so get the result wrong. Worse, we won’t be able to use our axioms to solve new problems reliably, because the truth of the axiom is only determined by the truth of the result. And of course, we still have the problem of clashing intuitions, if we don’t like the sound of the axioms we come up with. I’d say this is a better way of going about things, but still unsatisfactory.

The best solution is the concept of reflective equilibrium. In other words, consider your ethical axioms in the light not merely of your intuitions about axioms, but also about results. And do the same with your ethical results, while always keeping axioms and results consistent. And so consider the whole system at once. And so you may have to discard some intuitions here and there as inconsistent with the rest, but overall you have a system the totality of which you are happy with. And in so doing, you have no bullets to bite. You don’t have to “shut up and multiply” because you genuinely embrace both the process and the end result.

An analogy I like is that physics is like cartography – your map must accord with what is on the ground. Maths is like Lego – you have some plans and you fit the pieces together accordingly. Reflective equilibrium is like architecture – it’s no good having a great-looking design if it keeps falling down when you try and build it, and it’s no good building a very sturdy skyscraper, if you were trying to set up a bungalow. You have to change your ideas to work in terms of the physical world, but you also have to shape the physical world to your ideas. And hopefully you reach that happy equilibrium where the building is structurally sound, and also looks the way you want it to. If not, you can always plant vines. 🙂

Unfortunately, this is not how many people approach moral questions. Consider this solution to the trolley problem, which I must have heard a million times:

The best result is the one that maximises utility. If I do nothing, five people will die. If I flip the switch, I will save five people and kill one. 1 < 5. So flip the switch.

The argument makes no sense on its own terms. The premise does not lead to the conclusion. It started by saying we need to maximise utility, but it did not consider utility at any point. Instead, it maximised the number of people alive. And yet this is essentially the standard progressive answer to the question. What on earth is going on?

The answer is, essentially, that results are being prioritised over axioms. The axiom has not been chosen randomly – it’s liked because it denies others autonomy and justifies unlimited power. But it’s also been chosen because it’s infinitely pliable, and can be made to work for just about any result. It’s disturbing that the result progressives prefer is to murder someone, but hey ho. What’s more illuminating is to see the way that axiom, argument are conclusion are detached from one another.

Or consider this hypothetical: I walk up behind a stranger and shoot him in the head, killing him instantly and painlessly. Is this moral or immoral? Immoral, we all agree – I hope! But why? I haven’t caused the man any suffering. I ended his life, yes, but why is that a bad thing? Over their lives, people are at an equilibrium between happiness and sadness – indeed, neither concept makes sense without the other. All those people who were act-utilitarians a minute ago with the trolley problem, do they stick to their guns and say that this is a morally neutral act? Or do they try and argue that there is positive utility in his life? No. They suddenly take up positions about “right to life” and use loaded terms like “murder” – and quite right too. Yet they’d denounce such thinking in a different context!

And then to make it worse, they claim they are principled. Ultimately, “shut up and multiply” means “I know I can’t justify this position, but you should fall in lockstep anyway otherwise you’re a traitor to Reason.” I am not a fan.

Why I am not a Utilitarian

When seemingly intelligent people trip themselves up over easy ethical questions, it’s almost always because they are using a utilitarian calculus. For instance, if a doctor thinks it would benefit a patient, should he rape her? All reasonable people can agree that it’s a clear moral horror. But Robin Hanson thinks it’s OK. Or rather, he doesn’t think it’s OK. He knows, instinctively, that it’s wrong, but he still can’t bring himself to say so, because it would conflict with his utilitarian calculus.

I bring up this example again not to pick on Professor Hanson, but rather to point out the way that a faulty moral calculus can seduce the well-meaning into evil. I am not making a rhetorical point. Maybe you think it unlikely that a doctor could ever think rape could ever benefit his patient. Well, OK. But now consider the vast number of medical treatments available to a doctor, and the huge amount of power the law grants them – including the power to compel you to receive treatment. You can therefore see how important it is that your doctor not think like Robin Hanson.

The reason that utilitarian (or more commonly, pseudo-utilitarian) ideas are popular is that they provide a very simple prism with which to look at morality. All we have to do is maximise utility. Yes, we may wish to complicate things a little to avoid pitfalls such as utility monsters and so on, but at the core is a neat, plausible concept. But in the words of Mencken, “There is always a well-known solution to every human problem – neat, plausible, and wrong.”

What is Utility?

The first problem with trying to maximise “utility” is that no-one knows what utility is. Originally, utilitarianism was about maximising happiness or pleasure. Then, when people pointed out the obvious problems with this, it retreated to vaguer notions like satisfying preferences. Now they just use the word “utility” to gloss over these problems. You would think that a philosophy that urges us to maximise something would pay close attention to what that something is. You would be wrong.

In my view, utilitarianism can be expressed well in childish language – “make things as good as possible for everyone.” The problem is that when you start honing in on that idea, it falls apart. How do I make things “good” for someone? It’s something we’d all very much like to know, regardless of utilitarianism! Trying to reduce the concept to some fictional “utility” doesn’t make it any easier. Utility is not a real thing that we can parcel out, it’s a metaphor at best. There is no way to maximise utility for one person, let alone over the population. Consider the following hypothetical:

You are invited to a friend’s birthday party, and you have £10 to buy him a present. You could buy him a bottle of wine to make the evening more fun, a book to stimulate his mind, or a box of chocolates which will give him the most pleasure, but would also damage his health. While you are thinking about what to buy him, you ask my advice. I tell you “Buy him the present that would maximise his utility.” Do you consider my advice helpful?

Omniscience

The example I gave was the kind of problem we deal with in the real world. Notice that I did not specify that the friend would be equally grateful for each present, whether he would actually read the book, etc. In real life, when you’re making such a decision, you don’t know how it will play out. Perhaps your friend will strain his arm trying to open the bottle of wine, have to go to the hospital, contract MRSA and die. Not very likely, but possible. How possible? One in a million? One in a billion? You have absolutely no way of knowing. You cannot assign realistic probabilities to these kind of future events. I can’t maximise utility if I don’t know the results of my actions.

You may think the MRSA example is forced. OK, consider Gauguin. He abandoned his wife and children to pursue his art in Polynesia. For the sake of argument, let’s say that because his art gave value to so many people, that that outweighs the suffering of his own family. In other words, that the results of this decision had, in some sense, positive utility. So was Gauguin’s decision moral or immoral, in a utilitarian calculus? After all, he didn’t know at the time that he’d produce great art. He was taking an enormous risk.

There are two ways to view this problem. One is the Bernard Williams point of view, which is to say that yes, it was a moral decision, because the results were favourable. Sure, Gauguin was lucky, but it worked out, so all’s well that ends well. This is of course, an absurd position. It’s not a moral view of the decision at all, but a retrospective labelling. If later on Gauguin’s paintings inspire Barack Obama to start a nuclear war, does that then make Gauguin’s decision immoral after all? It is to strip the word morality of all meaning, and make it a synonym for “success.”

Unfortunately, the alternative position is just as absurd – that would imply that the utilitarian command is not “maximise utility” but “make a good faith effort to maximise utility, based on what you know.” But this is equally stupid, because now morality cannot serve as an adjudicator. If two people disagree in good faith, there would be no way of saying who is in the right. This too is not a morality, but an abnegation of it.

Autonomy

And so we see that utilitarianism is essentially a nonsense. It’s a command to maximise something fictional, in an unknowable future. But hang on, didn’t I claim it was evil? Yes I did, as I shall explain. You see real-world utilitarians just ignore these problems and muddle on through. Because what they really like about utilitarianism is it places no bounds on their actions. Anything can be justified, as long as it maximises utility. And this, not issues like utility monsters or the problem of aggregating, is what makes utilitarianism so immoral.

If we are to return to the rape example, we see that Hanson, being a utilitarian, thinks the doctor should rape the patient if it will benefit her. It’s not that he doesn’t care about her consent – he does – but he thinks it has only limited value, that if need be it can be overridden, for her own good. And this makes it an extremely attractive morality for all those people who think they know better than others, and want to meddle in their lives. It’s no coincidence that utilitarianism is embraced by progressives.

But personal autonomy is of course vital, and not only from a point of view of freedom, but also wellbeing. If others have the power to make you obey their orders for your own good, what is to stop them acting for their own good instead? The best way of encouraging good behaviour towards others is to require consent.

Why I am not a fan of Trolley Problems

My first introduction to philosophy was as a ten-year old, being posed the trolley problem. A train hurtles towards five unaware people on a railway line. The only way to save them is to flip a switch, and divert the train down another section of track – which will kill an innocent bystander. So, do you do nothing and allow the five people to die? Or do you flip the switch and save the five, but in so doing kill one person?

Some people think this is an interesting hypothetical – drawing the lines between our actions and inactions, deontology and consequentialism, etc. I think it’s deeply retarded, in three important ways. And I think it’s retarded in ways that go to the heart of much thinking on morality, fairness, and politics.

The Politics of Division

Firstly, the question does not illuminate, it merely divides. Ethics is of course not a science, because true experimentation is impossible. Science tells us how the world is, but cannot tell us how it ought to be. It can inform our moral choices, but it cannot determine them. If we consider the question of abortion, science can, in theory, tell us at what stage a foetus becomes viable. But if you disagree that viability is the relevant threshold for the permissibility of abortion, there is no experiment that can contradict you.

In the absence of true experimentation, one method of ethical enquiry is the thought experiment, of which the trolley problem is one. The purpose of a thought experiment is to clarify our thinking, and to proceed from easier questions to harder one. To return to the abortion example above, we might agree that infanticide is immoral. What then, we might ask, is the relevant difference between a newborn baby, and that same baby one hour before birth? And what difference between that baby and one hour earlier? And hence we might equate late-term abortion to infanticide, and so to be immoral.

Therefore a good thought experiment will be one where biting the bullet causes psychic pain to most people. There will always be someone willing to take the opposite side, no matter how outrageous – for example here Robin Hanson says that a doctor has a duty to rape his patients if he thinks it will help them. But hopefully we can get consensus on the thought experiment, and so proceed to solve harder problems. This is particularly important for applied ethics – i.e. law, which in a democracy must rest on consensus.

Unfortunately, the trolley problem is a terrible thought experiment in these terms. Switch-flippers say that they’re saving five people and killing one, so that’s a surplus of four. Non-flippers say that you can’t murder one person to save five. There are large groups on both sides, both entirely happy on their own terms, and all we have done is create two groups, each condemning the other. From a flipper point of view, non-flippers are “callous.” From a non-flipper point of view, flippers are “murderers.”

And of course, this is the state of modern democratic politics, where arbitrary questions are used not to illuminate but divide. The media at the moment is obsessed with the question “Are the coalition cuts going too far?” And so this provides large groups for and against, who can each yell at the other group, and talk among themselves about how the other side just aren’t serious/compassionate. A better question would be “Do you think it is sustainable for the UK to have a budget deficit of 15% of GDP?” Consensus answer: no. So how large a budget deficit can we run in the current climate? And what areas of spending do you want cut to get there? But the media is only interested in an emotional response, because that’s much easier to convey than a reasoned argument, and requires no facts.

The Politics of Ignorance

And this is the second problem with the trolley problem – it is arbitrary, and completely lacking in facts. Who are these people on the tracks, why are they there, and why are you at the controls? What is your relationship with the people? Why is the switch in its current position? What are the railway company’s procedures in such a situation? And so on. These are not irrelevant details around a central issue. These questions are the central issue.

For example, consider that the five people are maintenance workers, and the train was not supposed to go onto their part of the tracks. A saboteur has snuck onto the railway and flipped the switch so as to kill them. He is now making his escape down the other section of the track. Were you to flip the switch, you would be restoring the status quo ante, and only the would-be murderer would suffer.

Alternatively, consider that the five are trespassers – teenagers wandering on the tracks oblivious to the coming train. The other section of track is never used, and the person on it is a railway employee, who has good reason to think he is perfectly safe, and is trying to alert the five to the danger with a loudhailer – without success.

It’s not merely that we might give very different answers in these two cases. It’s that these cases present very different ways of thinking about the problem. Now, the kind of person who would pose such a fatuous hypothetical will no doubt object that you don’t have all this information, that you simply have to go on what you know. But how about the question of why I am at the controls? I must know that. What is my relationship to these people? I must know that. And so on. Suppose I am a friend of the five who has been sent to flip the switch to kill the one. Does that put a colour on my actions?

The Politics of Indifference

And this is the final problem with the trolley problem – implicit is the idea that all six potential victims are strangers, that I am supremely disinterested, and that I am the decider. This is, of course, highly unrealistic – but it’s also the way progressives like to think about politics. To suppose oneself to be eternally wise and powerful, and place oneself in a position of life and death over others, with no need for their consent or approval – what could be more dangerous? But that is not enough for the creator of the trolley problem, because then you as decision maker are stripped of all context and relationship with your victims, and any prior history of decision making.

In the related problem of “Cut-up Chuck,” a doctor has five patients who are near death. One needs a heart transplant, one a liver, and so on. The doctor performs a routine check-up on a new patient, Chuck, who is in perfect health. The doctor notices that Chuck is a donor match for the patients, and so he kills Chuck and transplants the heart to one, the liver to another, and so on, saving five lives by killing one. Is it OK to murder Chuck for the greater good?

Almost everyone will recoil in moral horror from this proposition, yet the logic of it is identical to the utilitarian logic that many are happy to embrace in the trolley problem. So what makes the difference? Firstly, we have named our victim. Rather than our consequences being inflicted on some unknown, nameless other, they are inflicted on a real human being. Secondly, we are not asked to place ourselves in the shoes of the doctor. You are probably not a doctor, but you have been a patient, so you are more likely to identify with Chuck going for his check-up than the doctor killing him. Whereas in the trolley problem it is specifically you in charge of the points. This is dangerous, and unrealistic. The life-and-death decisions in our society are taken by a small group of self-selecting people, not the populace at large. And finally, the perpetrator is given some positive relation to the victim. The doctor is there to help Chuck, not murder him and parcel out his organs. But so too, the points operator is supposed to look after people, not kill them.

A more honest way to pose the trolley problem, or indeed any ethical problem, is to make the audience aware of the costs. Suppose your son was walking down a safe section of track, and a points manager saw him there and diverted a train out of its way without warning. Your son was obliterated. But it’s all for the best, says the points operator, because otherwise the train would have hit five people. Really, I saved four lives. Would you try to have the manager prosecuted for murder?

Some will accuse me of loading the question with emotion. But that’s entirely the point. That person who you would kill when you flip the switch is undoubtedly someone’s son or daughter. Our actions are have real world consequences, and when we empathise with the decision-makers not the victims, it is almost always an excuse to commit horrible crimes. It is listening to Tony Blair whine about “the hardest decision of my life” while paying no attention to the hundreds of thousands of innocent Iraqis dead as a result of his adventurism. It is Sharon Shoesmith, it is John Yoo, it is every crime where the perpetrators are the bureaucrats and the victims are the poor and marginalised. In short, it is modern, “progressive” politics.

And I am not a fan.

Getting started

I don’t know how long I’ll keep this up for, but we’ll see. I intend to start on a variety of subjects I’ve been thinking about for some time.