What's Wrong With Mushy?
An issue that I keep coming back to in my exchanges with various of the Bleeding Heart Libertarians is what I see as their unwillingness to offer a clear description of their position. It comes up in details of what they write, such as Jason Brennan’s use of the phrase “minimally decent lives,” a term I have been describing as dishonest mush. It comes up in the overall pattern of the exchange, which started back on Cato Unbound, was continued on libertarianism.org, and revived here when I responded to Jason’s cartoon libertarian post. I have been trying to get the BHL people to tell me what they want to add to libertarianism as other libertarians see it, with a notable lack of success.
This raises the question I have used for my title. Jason wants to argue, in the specific case of “social justice,” that terms do not have to have clear meanings in order to be useful and he claims that lots of terms we routinely use, such as “justice” or “liberalism,” don’t. That is surely true to some extent; the meaning of words is usually at least a little fuzzy at the edges. It is true even for such obvious classifications as male and female, since there are people whose genetics are neither XX or XY, people who are biologically male but (by their report, which I am inclined to believe) psychologically female, people who are genetically of one gender but morphologically of the other, hermaphrodites, and a variety of other sorts of people (not to mention other organisms) that cannot be neatly classified by gender. But there is still a difference between a word that has an adequately clear meaning and one that does not. And there is a further difference between both and a word or phrase which appears to have a clear meaning, is designed to appear to have a clear meaning, but dissolves into mist or turns out to have a very different meaning when you look at it closely.
There are two problems with mush, one having to do with arguments, one with implementing them in institutions. The first may be best described in George Orwell’s classic essay “Politics and the English Language.” If you have never read it I suggest that you now do so, since it is more worth reading than this post.
Clear language promotes clear thought, fuzzy language makes possible unclear thought. The less clearly your ideas are defined, the harder it is to see problems with them, the easier to evade problems when other people point them out. That is part of why, at this point, I prefer the category of left-libertarians of whom Georgists, admirers of the 19th century economist Henry George, are the most familiar example to the ones I have been arguing with lately. The Georgist position starts from the observation that since land is not created by human effort there is a problem justifying anyone's claim to own it and exclude others. It goes on to argue for a land tax used to fund government and, possibly, provide payments to poor people. In an old article I discussed problems with one version of this line of argument and offered an alternative solution to the problem of ownership of land, although not one I find entirely satisfactory. But at least the Georgists are willing to offer an explicit argument, which makes it possible for other people to look for holes in its logic or implications that they find unattractive.
So far the BHL people I have been arguing with are not, which is why I at one point described my exchange with them as trying to nail jelly to the wall. The pattern is illustrated by their attitude to Rawls and his Difference Principle. They speak of Rawls with respect, imply that they, or at least some of them, are in favor of something with a vaguely Rawlsian flavor, but are unwilling to actually defend the argument with which he justified his principle. They think libertarianism should include something beyond natural rights but, aside from making it pretty clear that that something is not utilitarianism, are unwilling or unable to give a clear description of what.
So much for why I find mush irritating. Why do I also find it dangerous?
Suppose one concludes that people with characteristic X deserve special treatment, say a payment of a thousand dollars a year funded by other people’s taxes. Assume, first, that X is clearly defined, someone either has X or doesn’t and it is obvious to everyone which is the case. Let X be blindness, and assume for the moment that it is well defined.
There will still be costs, in addition to the direct cost of the transfers, to implementing the policy. Individuals can and will spend resources supporting or opposing the proposal, and that expenditure is a net cost. If the proposal is implemented, there may be additional costs as some people, most obviously the blind and those who provide goods and services to the blind, try to push the amount up, others to push it down. And there might be a cost due to a reduction in the incentive to avoid blindness or to cure it, although that effect is likely to be small as long as the disadvantages of being blind are much larger than the payment, and disappears if we assume that whether you are blind is entirely outside of your control. That assumption corresponds to the usual assumption in talk about social justice that it involves consequences of characteristics that are not your fault.
Now suppose we alter the assumption just a little by replacing “blind” with “legally blind” and giving a very fuzzy definition to the latter, something like “vision bad enough to significantly reduce life opportunities,” which is not too bad a parody of the sort of definitions moral philosophers like to give for disadvantages that they think deserve some sort of compensating special treatment. All of the problems I have described now expand. Anyone with less than perfect vision has an incentive to lobby Congress to broaden the criteria. Anyone with or without perfect vision has an incentive to try to get classified as legally blind, whether by bribing the inspector, getting a well paid physician to testify that he has an obscure optical problem, or merely faking the symptoms—at the cost of the inconvenience of not driving when he might be caught doing so. Make the criteria fuzzy enough—my example might do it—and what you really have is a subsidy to anyone with enough money and/or political influence to get himself qualified as legally blind, whatever the actual state of his vision.
Which is why mush, used in philosophical arguments that are intended to justify legal rules, is dangerous.
Let me end by returning to the specific example of “minimally decent lives” and a slightly more defensible version that came up in the comment thread on the BHL blog, “basic needs.” I begin with the latter.
A reasonably objective definition of “basic needs” might be “enough food and shelter so that their lack would not greatly reduce your life expectancy.” To make it more precise, replace “greatly reduce” with “reduce at least in half relative to those who had such food and shelter.” What would that work out to?
There are parts of the U.S. where housing is pretty cheap, down to about $100/room/month, probably less if I searched further. Assume that people are packed in ten beds to a room, along the lines of housing for tramps in London as described by Orwell in Down and Out in London and Paris. That gets annual housing cost per person down to about a hundred dollars.
On further thought, that’s too high. There are parts of the U.S. where the weather is temperate enough so that living outdoors, perhaps with a roof to shelter you from the rain, is not a serious risk to health. So all you need is some empty land in such an area, enough roofs for everyone to huddle under when it rains, and local authorities willing to put up with the land being used as a refuge for the homeless. Cost per person close to zero. Add a little for porta-potties and a water supply.
I should probably include clothing. There is a place in Boston that my wife, daughter, and some of our friends like to visit that sells second hand clothing at fifty cents a pound. Twenty pounds of clothing should easily last a year in a temperate climate, so call that another ten dollars a year.
What about food? George Stigler, back in 1945, provided an estimate of the lowest cost diet that satisfied what was then the list of nutritional requirements. It cost a little under $40/year. A recent repeat of the calculation, using current prices and current nutritional requirements, produced a figure of about $600/year.
That again is too high, since a sizable fraction of the world’s population survives on considerably less than Stigler’s 3000 calories/day, and probably considerably less than whatever figure the more recent calculation used. Stigler's diet was for someone doing moderately hard labor—again I don’t know the assumptions in the current version. But if all we are asking is what it takes to prevent severe malnutrition for someone who is presumably sitting around unemployed, since if he were employed he would already be making more than the amount we are considering, 2000 calories/day should do it easily. And we can get a better estimate of the cost of a really minimal diet, one designed to keep you from starvation, by simply looking for the foods that have the lowest cost per calorie. The last time I looked at the problem I checked supermarket prices and calorie contents, but now we have the web and someone else has done it. The result was that 200 calories of canola oil or wheat flour cost $.07, of peanut butter $.17.
Those three items, suitably combined, let you choose your mix of carbohydrate, protein, and fat, although I expect you would need a few other things, perhaps some vitamin pills, to avoid serious nutritional deficiencies. Call the total $.10/200 calories, or $1/day. That is not allowing for the cost of food preparation, but that ought to be pretty minimal if we are only trying to provide minimal nutrition, not a tasty meal.
Adding it all up, it looks as though a serious estimate of the cost of “basic needs” in the U.S. at present, taking the term seriously as describing what it takes to stay alive, would come to something around $500/year. Is there any of the BHL authors, any modern American philosopher who uses the term, who would be willing to interpret it that way? Anyone who is not, it seems to me, is pretending to use an objective standard when what he actually intends is something more like “a standard of living I would find tolerable.”
I am not suggesting that people ought to have to live on $400/year, although one risk of making this post is that at some point in the future someone may quote me out of context to claim I am. It is not the life I would choose—although if we assume that the homeless refuge contains some interesting people and not many unpleasant ones and that some generous donor provides a large supply of second hand books I expect I would prefer it by a good deal to not living. I have included no expenditure for medical care in my calculations, which could get unpleasant, but I think the historical evidence suggests that while the lack of medical care reduces life expectancy, it is unlikely to cut it in half.
This raises the question I have used for my title. Jason wants to argue, in the specific case of “social justice,” that terms do not have to have clear meanings in order to be useful and he claims that lots of terms we routinely use, such as “justice” or “liberalism,” don’t. That is surely true to some extent; the meaning of words is usually at least a little fuzzy at the edges. It is true even for such obvious classifications as male and female, since there are people whose genetics are neither XX or XY, people who are biologically male but (by their report, which I am inclined to believe) psychologically female, people who are genetically of one gender but morphologically of the other, hermaphrodites, and a variety of other sorts of people (not to mention other organisms) that cannot be neatly classified by gender. But there is still a difference between a word that has an adequately clear meaning and one that does not. And there is a further difference between both and a word or phrase which appears to have a clear meaning, is designed to appear to have a clear meaning, but dissolves into mist or turns out to have a very different meaning when you look at it closely.
There are two problems with mush, one having to do with arguments, one with implementing them in institutions. The first may be best described in George Orwell’s classic essay “Politics and the English Language.” If you have never read it I suggest that you now do so, since it is more worth reading than this post.
Clear language promotes clear thought, fuzzy language makes possible unclear thought. The less clearly your ideas are defined, the harder it is to see problems with them, the easier to evade problems when other people point them out. That is part of why, at this point, I prefer the category of left-libertarians of whom Georgists, admirers of the 19th century economist Henry George, are the most familiar example to the ones I have been arguing with lately. The Georgist position starts from the observation that since land is not created by human effort there is a problem justifying anyone's claim to own it and exclude others. It goes on to argue for a land tax used to fund government and, possibly, provide payments to poor people. In an old article I discussed problems with one version of this line of argument and offered an alternative solution to the problem of ownership of land, although not one I find entirely satisfactory. But at least the Georgists are willing to offer an explicit argument, which makes it possible for other people to look for holes in its logic or implications that they find unattractive.
So far the BHL people I have been arguing with are not, which is why I at one point described my exchange with them as trying to nail jelly to the wall. The pattern is illustrated by their attitude to Rawls and his Difference Principle. They speak of Rawls with respect, imply that they, or at least some of them, are in favor of something with a vaguely Rawlsian flavor, but are unwilling to actually defend the argument with which he justified his principle. They think libertarianism should include something beyond natural rights but, aside from making it pretty clear that that something is not utilitarianism, are unwilling or unable to give a clear description of what.
So much for why I find mush irritating. Why do I also find it dangerous?
Suppose one concludes that people with characteristic X deserve special treatment, say a payment of a thousand dollars a year funded by other people’s taxes. Assume, first, that X is clearly defined, someone either has X or doesn’t and it is obvious to everyone which is the case. Let X be blindness, and assume for the moment that it is well defined.
There will still be costs, in addition to the direct cost of the transfers, to implementing the policy. Individuals can and will spend resources supporting or opposing the proposal, and that expenditure is a net cost. If the proposal is implemented, there may be additional costs as some people, most obviously the blind and those who provide goods and services to the blind, try to push the amount up, others to push it down. And there might be a cost due to a reduction in the incentive to avoid blindness or to cure it, although that effect is likely to be small as long as the disadvantages of being blind are much larger than the payment, and disappears if we assume that whether you are blind is entirely outside of your control. That assumption corresponds to the usual assumption in talk about social justice that it involves consequences of characteristics that are not your fault.
Now suppose we alter the assumption just a little by replacing “blind” with “legally blind” and giving a very fuzzy definition to the latter, something like “vision bad enough to significantly reduce life opportunities,” which is not too bad a parody of the sort of definitions moral philosophers like to give for disadvantages that they think deserve some sort of compensating special treatment. All of the problems I have described now expand. Anyone with less than perfect vision has an incentive to lobby Congress to broaden the criteria. Anyone with or without perfect vision has an incentive to try to get classified as legally blind, whether by bribing the inspector, getting a well paid physician to testify that he has an obscure optical problem, or merely faking the symptoms—at the cost of the inconvenience of not driving when he might be caught doing so. Make the criteria fuzzy enough—my example might do it—and what you really have is a subsidy to anyone with enough money and/or political influence to get himself qualified as legally blind, whatever the actual state of his vision.
Which is why mush, used in philosophical arguments that are intended to justify legal rules, is dangerous.
Let me end by returning to the specific example of “minimally decent lives” and a slightly more defensible version that came up in the comment thread on the BHL blog, “basic needs.” I begin with the latter.
A reasonably objective definition of “basic needs” might be “enough food and shelter so that their lack would not greatly reduce your life expectancy.” To make it more precise, replace “greatly reduce” with “reduce at least in half relative to those who had such food and shelter.” What would that work out to?
There are parts of the U.S. where housing is pretty cheap, down to about $100/room/month, probably less if I searched further. Assume that people are packed in ten beds to a room, along the lines of housing for tramps in London as described by Orwell in Down and Out in London and Paris. That gets annual housing cost per person down to about a hundred dollars.
On further thought, that’s too high. There are parts of the U.S. where the weather is temperate enough so that living outdoors, perhaps with a roof to shelter you from the rain, is not a serious risk to health. So all you need is some empty land in such an area, enough roofs for everyone to huddle under when it rains, and local authorities willing to put up with the land being used as a refuge for the homeless. Cost per person close to zero. Add a little for porta-potties and a water supply.
I should probably include clothing. There is a place in Boston that my wife, daughter, and some of our friends like to visit that sells second hand clothing at fifty cents a pound. Twenty pounds of clothing should easily last a year in a temperate climate, so call that another ten dollars a year.
What about food? George Stigler, back in 1945, provided an estimate of the lowest cost diet that satisfied what was then the list of nutritional requirements. It cost a little under $40/year. A recent repeat of the calculation, using current prices and current nutritional requirements, produced a figure of about $600/year.
That again is too high, since a sizable fraction of the world’s population survives on considerably less than Stigler’s 3000 calories/day, and probably considerably less than whatever figure the more recent calculation used. Stigler's diet was for someone doing moderately hard labor—again I don’t know the assumptions in the current version. But if all we are asking is what it takes to prevent severe malnutrition for someone who is presumably sitting around unemployed, since if he were employed he would already be making more than the amount we are considering, 2000 calories/day should do it easily. And we can get a better estimate of the cost of a really minimal diet, one designed to keep you from starvation, by simply looking for the foods that have the lowest cost per calorie. The last time I looked at the problem I checked supermarket prices and calorie contents, but now we have the web and someone else has done it. The result was that 200 calories of canola oil or wheat flour cost $.07, of peanut butter $.17.
Those three items, suitably combined, let you choose your mix of carbohydrate, protein, and fat, although I expect you would need a few other things, perhaps some vitamin pills, to avoid serious nutritional deficiencies. Call the total $.10/200 calories, or $1/day. That is not allowing for the cost of food preparation, but that ought to be pretty minimal if we are only trying to provide minimal nutrition, not a tasty meal.
Adding it all up, it looks as though a serious estimate of the cost of “basic needs” in the U.S. at present, taking the term seriously as describing what it takes to stay alive, would come to something around $500/year. Is there any of the BHL authors, any modern American philosopher who uses the term, who would be willing to interpret it that way? Anyone who is not, it seems to me, is pretending to use an objective standard when what he actually intends is something more like “a standard of living I would find tolerable.”
I am not suggesting that people ought to have to live on $400/year, although one risk of making this post is that at some point in the future someone may quote me out of context to claim I am. It is not the life I would choose—although if we assume that the homeless refuge contains some interesting people and not many unpleasant ones and that some generous donor provides a large supply of second hand books I expect I would prefer it by a good deal to not living. I have included no expenditure for medical care in my calculations, which could get unpleasant, but I think the historical evidence suggests that while the lack of medical care reduces life expectancy, it is unlikely to cut it in half.
So much for "basic needs." "Minimally decent life" is harder to give any content to. For most of us, I suspect it gets interpreted as the least attractive life we would feel reasonably happy about living, which in practice probably means life at something like half whatever our annual expenditure presently is. That is not an objective standard and I think it would be hard to come up with any convincing argument, certainly with any argument convincing to a libertarian, to show that everybody is entitled to achieve it, if necessary paid for by other people.
For anyone who has not been following the argument from the beginning here are are the relevant links:
The Libertarianism.org discussion
More argument can be found on the comment threads to some of those posts.
And may I humbly suggest to the people at libertarianism.org that they ought to provide a page showing, with links, the structure of such exchanges? And, if they already have done so, tell me how to find it.
posted by David Friedman @ 2:37 PM
4 comments
