From the perspective of anyone concerned with limiting government and encouraging individual responsibility, the contemporary American welfare state is a disaster. According to a report by the Cato Institute’s Michael Tanner, welfare programs at the federal level alone cost more than $668 billion annually, spread across at least 126 different programs. Add another $284 of welfare spending at the state and local level, and you’ve got almost $1 trillion dollars of government spending on welfare - over $20,000 for every poor person in the United States.

Not only does the U.S. welfare state spend a lot; it spends it badly. Poor Americans receiving assistance face a bewildering variety of phase-outs and benefit cliffs that combine to create extremely high effective marginal tax rates on their labor. As a result, poor families often find that working more (or having a second adult work) simply doesn’t pay. And still, despite massive expenditures by the welfare state, some 16% of Americans are left living in poverty.

Wouldn’t it be better just to scrap the whole system and write the poor a check?

In what follows, I will make the case for a Basic Income Guarantee (BIG) as a replacement for the current welfare state. There are a number of distinct ways of arguing from libertarian premises to a BIG, some of which I have discussed in the past. In this essay, however, I will focus on what I take to be the strongest and most persuasive libertarian argument. I will argue that a BIG, even if it is not ideal from a libertarian perspective, is significantly better on libertarian grounds than our current welfare state, and has a much higher likelihood of being achieved in a world in which most people reject libertarian views.

I begin in the next section by explaining what I mean by a BIG. I then proceed to set out four reasons why libertarians should support a BIG over the current American welfare state. I close with some reflections on libertarian ideals and political compromise.

A Basic Income Guarantee

For purposes of this essay, I will use the phrase “Basic Income Guarantee” quite broadly to refer to a wide range of distinct policy proposals, including Milton Friedman’s Negative Income Tax (NIT), Bruce Ackerman and Anne Alstott’s proposal for a Stakeholder Grant, the Thomas Paine / Henry George inspired idea of a citizen’s dividend, the Alaska Permanent Fund Dividend, and Charles Murray’s 2006 proposal for the government to write a $10,000 each year to every American citizen over the age of twenty-one.[1] There is, of course, quite a bit of variation among these plans in terms of cost, payouts, implementation, and so on. Despite these differences, however, they all have in common two important features.

First, they involve a cash grant with no strings attached. Unlike other welfare programs which encourage or require recipients to consume certain specific kinds of good – such as medical care, housing, or food – a BIG simply gives people cash, and leaves them free to spend it, or save it, in whatever way they choose.

Second, a BIG is an unconditional grant for which every citizen (or at least every adult citizen) is eligible. It is not means-tested; checks are issued to poor and rich alike (though on some proposals payments to the rich will be partially or fully recaptured through the tax system). Beneficiaries do not have to pass a drug test or demonstrate that they are willing to work. If you’re alive, and a citizen, you get a check. Period.

A Pragmatic Libertarian Argument

No libertarian would wish for a BIG as an addition to the currently existing welfare state. But what about as a replacement for it? Such a revolutionary overhaul of the welfare state would almost certainly require a constitutional amendment, both to insulate debate somewhat from the pleas and protests of special interests, and to make it considerably more difficult to renege on the deal afterwards. Charles Murray has given us a rough idea of what such an amendment might look like:

Henceforth, federal, state, and local governments shall make no law nor establish any program that provides benefits to some citizens but not to others. All programs currently providing such benefits are to be terminated. The funds formerly allocated to them are to be used instead to provide every citizen with a cash grant beginning at age twenty-one and continuing until death. The annual value of the cash grant at the program’s outset is to be $10,000.

Suppose, to indulge in a bit of speculative fancy, that this deal was actually on the political table. Should libertarians take it? Given that it is not on the table now, should libertarians make some effort to get it there? I believe the answer to both of those questions is “yes.” A BIG might not be libertarians’ ideal policy – though more on this later – but it is almost certainly a lot better on libertarian grounds than what we have right now. Here are four reasons why.

Less Bureaucracy

Every one of the more than 126 federal welfare programs comes with its own bureaucracy, its own set of arcane rules, regulations, and restrictions, and its own significant (and rising) overhead costs. A BIG, in contrast, requires significantly less in terms of administrative expense. A program in which everyone gets a check for the same amount is simple enough to be administered by a computer program. And even a more complicated proposal, like Murray’s or like Friedman’s NIT, could largely piggyback off of the already existing bureaucracy of the federal tax system.

Eliminating a large chunk of the federal bureaucracy would obviously be good from the perspective of a libertarian concern to reduce the size and scope of government. But it would also be good from the perspective of welfare beneficiaries. Actually getting signed up for all the various welfare benefits to which one is entitled is tremendously costly in terms of time, effort, and skill at bureaucratic navigation. Many people miss out on benefits for which they qualify simply because they don’t know that the program exists, or what they need to do to draw from it. Getting the benefit of a BIG, in contrast, requires just a single signature on the back of a check. If we’re going to spend money on helping the poor, shouldn’t we make sure that they actually get the help we’re paying for?

Cheaper

Second, a BIG could be considerably cheaper than the current welfare state. How much cheaper depends on the details of the particular proposal. Some, like Murray’s, which involve a progressive tax on the BIG once a certain threshold of income is reached, appear to be considerably cheaper. Other analyses, like Ed Dolan’s, suggest only that a moderate BIG would not cost more than what we currently spend.

Part of the explanation of the relatively low cost of a BIG comes from the reduction of bureaucracy, described above. But another reason is to be found in Director’s Law: If you’re like most people, when you hear “welfare” you think about transfers from the rich to the poor. But in reality, most political transfers benefit the middle class at the expense of the poor (and rich). If the BIG is going to replace the welfare state, then transfers to the middle class such as subsidies for higher education, the mortgage interest deduction, and tax benefits for retirement savings ought to be cut right along with (if not before) SNAP, TANF, etc.

Again, how much a BIG would cost relative to the current welfare state depends on the details of the particular BIG proposal. Various proposals need to be evaluated on their own merits, and of course I do not wish to claim that every BIG proposal will be more affordable than our current welfare state. But neither is there any reason to believe that no reasonable proposal could be.

Less Rent-Seeking

Whenever there exists a bureaucracy with the power and discretion to take from some in order to benefit others, there will also exist powerful incentives for individuals to manipulate that bureaucracy in order to better serve their own private interests. Agents of the bureaucracy itself will seek to expand its scope and budget regardless of whether such expansion serves the interests of its clients. And special interest groups will use various political mechanisms to channel the organization’s resources into their own pockets.

In theory, the welfare state doles out money and other resources on the basis of such factors as need and desert. But need and desert are both philosophically contested and impossible to measure objectively. And so, in practice, resources are doled out to those who can make the best political case that they need or deserve it. And this is a contest in which the genuine poor are at a serious disadvantage relative to the better educated, wealthier, and more politically engaged middle class.

A BIG, in contrast, allows virtually no room for bureaucratic discretion, and thus minimizes the opportunities for political rent-seeking and opportunism. It is, as the late James Buchanan once noted, a perfectly general policy that treats all citizens the same. It is thus entirely ill-suited for use as a method of political exploitation. We should therefore expect to see much less rent-seeking and opportunism with a BIG than we do with the present welfare state, and therefore a more effective transfer of resources toward the genuinely needy as opposed to the politically well-connected.

Of course, no policy is perfectly immune to rent-seeking or political manipulation, and others have expressed what seem to me to be some entirely reasonable concerns about a BIG in this respect. But nothing that I have seen has yet convinced me that the problems with a BIG would be worse than those we have now, and there still seems to me to be good reason to think those problems would be considerably diminished.

Less Invasive / Paternalistic

One of the main differences between a BIG and the current welfare state is the unconditionality of the former. Under a BIG, everybody gets a check. Under the current welfare state, only people who meet the various stipulated qualifications are eligible for assistance. The precise nature of those qualifications varies from program to program, but can include not earning too much, not earning too little, not being on drugs, not having won the lottery, making an earnest effort to find work, and so on.

Conditions are put on welfare in order to ensure that assistance goes to the deserving poor, and not to the undeserving. But distinguishing between the deserving and undeserving is difficult business, and requires a variety of invasive, demoralizing, and degrading inspections into the intimate details of applicants’ lives. “Fill out this form, tell us about that man you live with, pee in this cup, and submit to spot inspections of your home by our social workers, or else.”

Maybe the state shouldn’t be in the business of giving out welfare at all. Maybe it shouldn’t be running schools, or highways either. But, as Jacob Levy notes, since it does do these things, libertarians have good reason to demand that it does so in a way that is as “more rather than less compatible with Hayek’s rule of law, with freedom from supervision and surveillance by the bureaucracy, with the ability to get on with living their lives rather than having to waste them proving their innocence.”

The conditional welfare state is not only invasive, it is heavily paternalistic. Restrictions on eligibility are imposed in order to encourage welfare recipients to live their lives in a way that the state thinks is good for them: don’t have kids out of wedlock, don’t do drugs, and get (or stay) married. And benefits are often given in-kind rather than in cash precisely because the state doesn’t trust welfare recipients to make what it regards as wise choices about how to spend their money. This, despite the fact that both economic theory and a growing body of empirical evidence suggest that individuals are better off with the freedom of choice that a cash grant brings. In-kind grant programs like SNAP (food stamps) persist in their present form not because they are effective but because they are the product of a classic Bootleggers-and-Baptists coalition: well-meaning members of the public like the idea that welfare recipients have to use their vouchers on food rather than alcohol and cigarettes, and the farm lobby likes that beneficiaries are forced to buy its own products. Poor people, meanwhile, are deprived of the opportunity to save that a cash grant would give them, and they are forced to waste time and effort trading what SNAP allows them to buy for what they really want.

Utopia is Not an Option

In Libertarian Utopia, we might not have any welfare state all, no matter how limited or efficient. Many libertarians believe that any redistribution of wealth by the state violates individual rights and is therefore morally impermissible. And even those libertarians who do not base their political ideology on a theory of individual rights will worry that welfare states will produce perverse incentives – both on the part of recipients and potential recipients, and in the political processes that sustain and shape government policy.

But we do not live in Libertarian Utopia, nor have any of its prophets yet produced any compelling plan for how to get There from Here. Moreover, most people are not libertarians, and so unless we are willing to impose our views on them by force, we must try to find policy proposals that can command the assent of those who do not share our fundamental moral commitments and empirical beliefs.

From this perspective, the question of social welfare policy becomes less an exercise in ideal theory and more a problem of comparative institutional analysis. The question is not whether a BIG is a perfectly libertarian policy in every way, but whether it is more libertarian than the other realistically available policy alternatives. I believe that the considerations examined above provide us with very strong reason for believing that it is.

But I also believe that a BIG need not be merely a compromise. Even in a Libertarian Utopia, in other words, I think there would be good reasons to provide a social safety net through the mechanism of a BIG. I have written about some of these arguments before, and while constraints of space prevent me from elaborating upon them here, I am happy to do so in the discussion that follows this essay.

For now I will say only that if the idea of a social safety net strikes many readers as obviously incompatible with libertarianism, this is testament to the way in which an excessively narrow understanding of libertarianism has come to dominate our political discourse. For many people, it seems, libertarian thought begins and ends with the ideas of Ayn Rand and Murray Rothbard. And, of course, both Rand and Rothbard were indeed important libertarians, and libertarians from whose ideas I and many others have profited immensely. But while those ideas have played and continue to play an important role in the libertarian intellectual tradition, they do not exhaust that tradition. Once we adjust our eyes to see past the giants of Rand and Rothbard, it is clear that the libertarian intellectual landscape is far more diverse than it first appeared, and far less hostile to the idea of a social safety net.

We can, of course, define libertarianism however we wish, and it is possible to conceive it in a narrow enough way so as to rule out all support for income redistribution by definitional fiat. But any definition of libertarianism that is so narrow as to rule out the likes of John Locke, Thomas Paine, Adam Smith, Milton Friedman, Friedrich Hayek, Robert Nozick, Loren Lomasky, and Eric Mack, to name just a few, seems both historically distortive and pragmatically unhelpful. The arguments these thinkers have advanced on behalf of a (limited) social safety net might be mistaken. But that is something to be established by a careful examination of the substance of the arguments themselves. Arguments about what counts as a “real” libertarian position, especially arguments poorly informed by the writings of seminal historical and contemporary libertarian thinkers, do little to advance the debate.

A Few Words About Work Disincentives

So far in this essay, I have said virtually nothing to say about the many possible objections to a BIG. As a philosopher, this makes me profoundly uncomfortable. I am comforted somewhat by the knowledge that there will be plenty of time to explore these objections in the discussion that follows – and almost certainly plenty of prompting to do so by my fellow discussants! Nevertheless, before I bring this essay to a close, I want to say just a few words about what I take to be the most common, and also the most overrated, objection to a BIG.

Many people argue that a BIG will create a strong disincentive to work. From a theoretical perspective, this makes sense. If you lower the cost of unemployment relative to employment, you’re going to get more unemployment. The famous Negative Income Tax experiments of the 1970s seem to lend some empirical support to this hypothesis.

I find this argument unimpressive for two reasons. First, it is not at all clear that a BIG really would lead to a significant increase in unemployment. The actual findings of the NIT experiments were much more ambiguous than they have generally been represented to be in the nonacademic press. And insofar as a BIG allows welfare recipients who start working to keep more of their money than they would under a conditional welfare system, we should expect at least some reduction of work disincentives relative to the current system.

But suppose that a BIG actually would, on net, increase unemployment somewhat. The second response is: so what? Is it so obviously a flaw in the system if it leads more parents to take time off work to stay home with their children? Or college graduates to take a year off before beginning to work? Or if, among the population as a whole, the balance between work and leisure is slightly shifted toward the latter? My point is not that there isn’t any story that could be told about why work disincentives might be a problem. My point is simply that, even if they were guaranteed to occur, they wouldn’t obviously be a problem. Explaining why these somewhat increased disincentives are a problem requires something more substantial in the way of economic, sociological, and philosophical analysis than often seems to have been assumed.

To my mind, there are other, much better objections to a BIG of the sort I have discussed in this essay. How, for instance, will it handle the issue of children? Would a BIG increase native resistance to increased immigration and thereby hurt the truly needy global poor for the benefit of the (relatively) wealthy American poor? And how could a BIG be politically feasible, given the strong investment various interest groups have in maintaining the current system? I look forward to exploring these questions, and no doubt many more that I have not anticipated, in the discussion that follows.

The central point for which Matt Zwolinski’s lead essay argues is that a basic income guarantee would be less bad than the status quo. Having been recruited for a critical comment, I hope my role is not to dispute this thesis, because I can’t think of how one could defend the status quo from any philosophical standpoint.

I therefore hope I may be excused for focusing on something that Zwolinski did not argue in this essay but that he has argued elsewhere: Is a government-provided basic income guarantee ethically permissible? The answer to this is non-obvious, but I think it is probably no. (Note that this is compatible with the fact that the status quo is even worse.) I will summarize my main reason for this, then address Zwolinski’s arguments from other postings.

I. The Basic Argument Against a Basic Income

My basic argument is this:

1.    A basic income guarantee is permissible only if the state has political authority.

2.    No one has political authority.

3.    Therefore, a basic income guarantee is impermissible.

What is meant by “political authority” here? Roughly, political authority involves a kind of exemption from the usual norms of non-aggression in interpersonal ethics – an exemption that entitles the agent who has authority to coercively impose rules of conduct on the rest of society, in conditions or for reasons that would not justify coercion on the part of ordinary private agents. (For a better explanation, see my The Problem of Political Authority, chapter 1).

Why is premise (1) true? Suppose I decided to provide a basic income for my neighborhood. I don’t have enough justly acquired money to do this, so I extract the needed funds from my neighbors by threatening them with kidnaping and long-term imprisonment if they fail to hand over the funds I require. Sometimes a neighbor evades my efforts, usually by lying to me about his income. I kidnap these neighbors and hold them prisoner in small cells for years at a time. This part is crucial – carrying out the coercive threats on recalcitrant citizens is practically necessary to maintaining a tax system in any realistic society.

This behavior seems impermissible, to put it mildly. Even registered Democrats would agree with this. Now, that does not yet establish that a government-guaranteed basic income is impermissible. Perhaps the government is different from me and other private agents in some ethically relevant way, which exempts them from the moral constraint that enjoins me from behaving in this way. What the example shows is that basic-income advocates need some account of how the government is thus exempt. That is, they need a theory of political authority.

I hold, however, that no agent has ever possessed political authority. I can’t give you the argument for that here. The argument consists in reviewing the most plausible and influential accounts of the basis of the state’s authority – for example, theories based on the social contract, the democratic process, fairness, or utilitarian considerations – and showing, in each case, how the theories fail. If you want to know more about that, I’m afraid you have to read my book (sorry!).

II. The Rectification Argument for a Basic Income 

History is full of injustice, much of it affecting the distribution of property. Following a suggestion of Robert Nozick’s, Zwolinski elsewhere suggests that, since the chain of causation is difficult to trace in individual cases, and since those who are doing especially badly now are disproportionately likely to have been harmed by that injustice, the state might be justified in redistributing wealth to the poor in general, as a kind of approximation to the rectification of injustice.

Here are a few brief problems with this line of thought:

A.    We should expect that the effects of injustices tend to wash out over time. Personal choices and innate abilities probably have much more to do with a person’s present-day income level than events occurring a century or more in the past. This is one reason for adopting a sort of statute of limitations on reparations for past injustices. 

B.    A basic income would redistribute money from the rich and the middle class to the poor. But those from whom the money is taken are not guilty of any crimes for which they owe compensation. For example, no living person is responsible for slavery; it would therefore be unjust to force anyone to pay compensation for it.

C.    Zwolinski suggests that despite point (B), the state might owe compensation for past injustices, because the same state continues to exist despite changes in personnel. Analogously, corporations retain debts and other obligations even when the members who incurred those obligations are no longer in the corporation, or even are no longer alive. Shouldn’t a similar principle therefore hold for governments?

Very briefly, I think this continuity of obligations is possible for private corporations because new members of a corporation, in voluntarily joining the organization, thereby undertake whatever obligations go along with their accepted role in the corporation. These may include, for example, seeing to the payment of debts incurred by past managers. The same does not hold for the state, because new taxpayers do not voluntarily undertake the obligations of the government; rather, they have the government’s debts forcibly imposed on them.

Admittedly, my view here will appeal more to anarchists than to anyone else, since the inability to collect funds to pay its debts would create serious practical problems for running a government.

III. The Freedom Argument

In another posting, Zwolinski suggested that a basic income might be justified on the grounds that it protects individuals’ freedom, because having a basic income insulates one from certain kinds of coercion by one’s employer. Zwolinski characterizes this as a libertarian argument for a basic income.

I think this illustrates the problem with characterizing libertarianism merely as a pro-freedom philosophy. As I understand it, the (typical) minimal-state libertarian’s view is not that the job of the state is to act so as to maximize freedom. Rather, the state is charged with protecting the rights of individuals, while avoiding violating rights itself. Typically one is not justified in violating another person’s rights solely to prevent something slightly worse from happening – not even if the slightly worse thing would involve a diminution of someone’s freedom. For example, it is not ethically permissible to murder an innocent person, even if doing so somehow prevents two other innocent people from being murdered. Or so I believe, and so says the usual deontological conception of rights. I thus find the “freedom” argument unpersuasive. Though a basic income might increase people’s freedom, it would require the state to violate the rights of taxpayers. To justify this, one would have to argue, not merely that the rights violation is required to secure a modest net increase in “freedom,” but that the rights violation is required to prevent something many times worse from happening. Such, I take it, is the logic of rights.

Matt Zwolinski presents a thoughtful essay in favor of a Basic Income Guarantee (BIG) that reminds me strongly of the lawyer’s proverbial “in the alternative” argument about a dog biting a neighbor: “Your honor, my client doesn’t own a dog; even if he did own a dog, then it could not have bitten this man; and even if he did own a dog that bit this man, then it must have been provoked.” 

At the highest level, Zwolinski argues that a BIG is consistent with libertarian theory. And in the alternative, argues that in the real world of practical politics a welfare system of some kind will be with us for a long time, and a BIG is better than the dog’s breakfast of social welfare programs that we have today. Nested within this is another narrower argument in the alternative. He claims that social science evidence indicates that it not clear that a BIG would result in a reduction in work effort. But he argues that even if it did, this would not necessarily be a bad thing.

I’ll begin with a perspective on the narrower argument, and then proceed up to the broader and more philosophical topics.

It is fairly extraordinary to claim that the government could guarantee every adult in America an income even if they did zero work of any kind, and that somehow this would not reduce work effort. Zwolinksi should be able to provide strong evidence for such a claim. But we have scientific gold standard evidence that runs exactly the other way. A series of randomized experiments offered a version of Zwolinski’s proposal between 1968 and 1980. These tested a wide variety of program variants among the urban and rural poor, in better and worse macroeconomic periods, and in geographies from New Jersey to Seattle. They consistently found that the tested programs reduce the number of hours worked versus the existing welfare system, and the tested levels of progressivity of implicit tax rates did not get around this problem by encouraging work, as Zwolinski’s theoretical argument asserts they should.

There was a further series of more than 30 randomized experiments conducted around the time of the welfare debates of the 1990s. These tested many ideas for improving welfare. What emerged from them was a clear picture: work requirements, and only work requirements, could be shown experimentally to get people off welfare and into jobs in a humane fashion. These experiments were an important input into the decision to make work requirements a central tenet of the new welfare regime when the welfare system was converted from AFDC to TANF in 1996.

The paper that Zwolinski cites raises three objections to popular interpretations of the first round of experiments, all of which center around the point that we can never know with certainty the impacts any experimental program would have when scaled up. But there is no serious debate that in the dozens of occasions in which it has been put to the test, changing work requirements changes work effort, just as common sense says it should. Zwolinski’s proposal would reduce work requirements down two steps from the current policy – from current TANF-level requirements, down through the prior AFDC-level requirements, down to no requirements at all – when each of these steps has been shown to reduce work effort. Human society is complex, but as much as we can make almost any non-trivial prediction of social welfare policy, we can state with confidence that Zwolinski’s proposal would lead to fewer work hours in America.

Zwolinski’s argument in the alternative – So what, why is this necessarily a bad thing? – is subjective and highly values-laden. Ultimately there is no absolute answer to this question. I will simply note that the moral intuitions of the American electorate appear to be very negative about a BIG. A major driver of welfare reform over decades has been political resistance to work disincentives. As a practical matter, creating a political collation built around the idea of free money for life for every citizen with no strings attached seems extremely unlikely.

This turns out to be highly relevant to considering Zwolinski’s broader argument.

Zwolinksi’s argues that as a practical matter we will have a welfare system for the foreseeable future, and that compared to the current welfare system a BIG would be less bureaucratic, cheaper, reduce rent-seeking, and be less paternalistic. The benefits Zwolinski describes derive primarily from the purity of his proposal. But in that case, the real comparison is not between a theoretical BIG and a theoretical means-tested welfare system; it is between an academic idea that has not yet been subjected to lobbying and legislation, on the one hand, and real laws that are the product of a democratic process, on the other. There is nothing inherent about a BIG that will prevent Congress from creating thousands of pages of special rules, exemptions, tax expenditures, and so on, that are collectively just as convoluted as the current welfare system. After all, “tax each person a given fraction of income” is a pretty simple idea too, but look at the 2013 federal income tax code.

And the likely maintenance or reemergence of the functional equivalents of many of these programs isn’t just the result of cynicism, but of healthy intuitions of natural justice that are essential to maintaining a well-functioning political order. As one example, if part of the motivation for giving adults income is that they spend it supporting their children, would we really allow parents receiving taxpayer money to spend it any way they want with no requirements for child welfare beyond child abuse laws? And as another, a huge and growing portion of the cost of the welfare state is health care. Suppose we gave every adult in America an annual grant of $10,000, and some person who did not buy health insurance with it got sick with an acute, easily treatable condition. Would we really bar them from any urgent medical care and just say “Tough luck, but it’s time to die”? Even if you think this would be a desirable public policy, it’s not plausible that the existence of a BIG would somehow change the political calculus enough to make it substantially more of a reality than it is today.

This points to what I think is the most serious flaw in Zwolinski’s argument: He assumes that the ideal of no government welfare is politically unobtainable, but simultaneously assumes that we could successfully pass and enforce a constitutional amendment enshrining a BIG and nothing but a BIG as forever the law of the land, when there is excellent evidence that such a concept has been politically unviable for more than 200 years of American history and has been getting more unpopular over the past several decades. This is wishing for a magic wand to make democratic politics go away. Why not just assume you can pass a constitutional amendment banning any welfare spending?  Oh, and while we are at it, in order to make a BIG anything other than vastly more expensive than the current welfare system, we are also going to eliminate the mortgage interest credit, 401Ks, and student loans, without any countervailing political deals or economic costs. Zwolinski presents no evidence that there is any prospect that such a set of changes would be politically feasible in our lifetimes.

Let me end with a note of agreement on Zwolinski’s highest-level point. I agree that a social welfare system is consistent with libertarian thought, broadly considered. One of the many, many ways to sub-divide libertarian ideology is that one strand takes liberty to be a (or in extreme cases, the) fundamental human good in and of itself; the other takes liberty to be a means to the end of discovery of methods of social organization that create other benefits. I’ve called the first “liberty-as-goal” libertarianism and the second “liberty-as-means” libertarianism.

Both are real parts of the libertarian intellectual legacy. Hayek, for example, cannot reasonably be construed as supporting an unconditional income for all citizens. Yet central to his thinking was the belief that a society must be adaptable and willing to experiment with new approaches. Trial-and-error learning is far more central to this strand of libertarianism than is trying to divine what Murray Rothbard (or F. A. Hayek) said on a given subject. The problem for proponents of a BIG is that this implies actually accepting the verdict of trials, which thus far have shown every tested variant of a BIG to have pernicious effects.

I agree wholeheartedly with Matt Zwolinski’s dim view of the current American social safety net. Its complex components are paternalistic, create powerful work disincentives, and are riddled with bureaucratic inefficiency. I also agree that his proposal to replace the entire edifice with a basic annual cash grant to every citizen would be an enormous improvement over the status quo.

But there are two deep flaws in his proposal. Most important, a basic income grant sufficient to lift urban families from poverty probably could not win political support initially, but even if it somehow did, fierce opposition to it would surely erupt quickly. The second flaw is that there is a much better way to reform the social safety net, one that would be both politically viable and deliver far more value per dollar. I’ll consider these points in turn.

The cornerstone of Zwolinski’s proposal would be a negative income tax along the lines proposed by Milton Friedman more than 60 years ago—basically, a cash grant that every man woman or child would get, irrespective of income or employment status. Friedman advocated a grant large enough to lift everyone above the poverty line, even those with no income from work or other sources. But a moment’s reflection reveals that a payment large enough to sustain an urban family of four at the official government poverty threshold (about $25,000 today) would quickly doom the program politically.

Imagine, for example, that a group of ten families formed a rural commune and supplemented their $250,000 in cash grants with the untaxed fruits of gardening and animal husbandry. If they located in Colorado or Washington, they could also grow marijuana, both for sale and for personal consumption. Their mornings would be free to drink coffee and engage in extended discussions of politics and the arts. They could hone their musical skills. They could read novels, write poetry, play nude volleyball.

Is it far-fetched to imagine that at least some groups would forsake paid employment in favor of leading lives like these at taxpayer expense? Once such groups formed, wouldn’t it be only a matter of time before journalists found them and created an eager audience for reports of their doings? And wouldn’t most voters react angrily once footage of the reveling commune members began running on the nightly news?

Of course they would, and who could blame them? An Indianapolis dentist with varicose veins rises at 6:00 each morning and drives through heavy traffic on a snow-covered freeway to spend the rest of his day treating patients with bad breath who take offense if they’re charged a fee for breaking an appointment without notice. How could such a person not be indignant at the sight of able-bodied people living it up on his tax dollars?

In short, it is a pipe dream to imagine that an income grant large enough to lift an urban family from poverty could win or sustain political support for long. Voters might support a cash grant if it were far too small to support comfortable living in groups. But the proposal would then fail by definition as an effective social safety net. 

A cash grant could still be an important component of the social safety net, but there would also have to some way to supplement it without undercutting the incentive to work. The most direct response to this concern would be to combine a cash grant that is far too small to lift an urban family from poverty with an open offer to pay sub-minimum wages to those willing to perform useful tasks in the public sphere. Such an offer would also address the strongest objection to the Earned Income Tax Credit, today’s leading income supplement program, which is that it is useless for those who cannot find employment.

Experiments have demonstrated the existence of many useful tasks that can be performed by unskilled workers with proper supervision. (Some examples: landscaping and maintenance in parks; transporting the elderly and handicapped; filling potholes in city streets; replacing burned out street lamps; transplanting seedlings in erosion control projects; removing graffiti from public places; painting government buildings; recycling newspapers and aluminum and glass containers; and staffing day care centers.) In combination, the basic income grant and the sub-minimum wage of the public sponsored job would be just enough to clear the poverty threshold. 

Libertarians might object that a public-sponsored jobs program would entail an undesirable expansion of government bureaucracy. But it wouldn’t be necessary for government to manage the program directly. That task could be bid out to the private sector.

Others might object that a program of low-wage public employment would be akin to forced servitude. Yet no one would be required to perform a public task. And to those who object that the public employment offer would not alleviate poverty for those who choose not to work, we could then say that poverty was truly a voluntary choice, not a condition created by the unavailability of jobs.

Still others might object that the offer of public jobs would lure people away from productive employment in the private sector. But pegging the wage for unskilled public employment significantly below the minimum wage would address this concern. Participants would have strong incentives to move into private employment or more skilled government employment as soon as they were able, and their experience in the public program would help them accomplish that transition.

Paying for the two-pronged solution I propose would require that libertarians rethink their oft-stated objection that taxation is theft. But that’s not an onerous requirement, since equating taxation and theft was not a compelling position in the first place. A country without the power to tax couldn’t field an army and would soon be overrun by a country that had one. Its residents would then have to pay taxes to that country.

For present purposes, the interesting questions are these: What activities should we undertake collectively and how should we raise the revenue to pay for them? People from all ends of the political spectrum agree that society should have an effective safety net. My claim is that even though the one I’ve proposed would not be cheap, it would get the job done at the lowest possible total cost.

Fortunately, there are many things we should be taxing, even if we didn’t need more revenue. We should be taxing vehicles by weight, for example, since those who buy heavier vehicles put others at greater risk of injury and death. We should be charging motorists for using roadways during congested hours, since usage at those times impedes others’ progress toward their destinations. And in light of the robust scientific consensus about the causes of climate change, we also should be taxing greenhouse gas emissions. Simply by taxing activities that do more harm than good, we could raise more than enough to pay for my proposal.

For the reasons that Matt Zwolinski has ably articulated, America now lacks an effective social safety net. His alternative would be an improvement, but there is virtually no chance that it could survive politically. The alternative I propose would be expensive, but much of its expense would be in exchange for valued tasks in the public sphere that the private sector has no incentive to attend to. And lest we forget: Our failure to maintain an adequate social safety net has also proved extremely costly.

Before I begin any substantive response to my commentators, I want to take a moment to extend my deep gratitude to Michael Huemer, Jim Manzi, and Robert Frank, for taking the time to engage so thoughtfully with my essay.  And, of course, I’m very thankful to Jason Kuznicki and the Cato Institute for making this conversation possible. Libertarian intellectuals, and especially libertarian intellectuals affiliated with think tanks, sometimes get accused of spending too much of their time in an intellectual echo chamber. The long history of excellent and ideologically diverse discussions at Cato Unbound in general, and this one in particular, seem to demonstrate pretty clearly that Cato has not succumbed to that vice.

Mike, Jim, and Robert have given me quite a bit to chew on – far too much for me to respond to in a single essay of any reasonable length. Rather than trying to respond to everything at once, which would be impossible, and rather than trying to respond to each essay in turn, which would be tedious, my plan is to organize my responses by topic. In this essay, I will address the issue of work disincentives and work requirements, and in so doing respond to some of the points that Jim and Robert have raised about these issues. In subsequent responses, I will address the feasibility of the Basic Income Guarantee (BIG), Mike Huemer’s anarchist challenge, and the relation between a BIG and the libertarian commitment to freedom.

The BIG and Work Disincentives

Jim Manzi argues that a series of experiments conducted between 1968 and 1980 demonstrates convincingly that a Negative Income Tax (or NIT - which I classified in my original essay as one form of BIG, broadly construed) leads to a reduction in the number of hours worked by recipients. Nor, Jim, thinks, should these results be surprising. “It is fairly extraordinary to claim,” he writes, “that the government could guarantee every adult in America an income even if they did zero work of any kind, and that somehow this would not reduce work effort.”

Both Jim and Robert think that this kind of reduction of work effort is a serious problem for the BIG, though both are a little unclear on exactly why it is a problem. It might be because allowing some lazy (dope-growing, commune-living) individuals to live off the productive efforts of others violates norms of reciprocity that, they think, we as a society have good reason to endorse. But maybe not. Even if we don’t have good reason to endorse those norms, or even if those norms don’t tell decisively against a BIG, the fact is that most people in our society think they do. And so, as a purely political matter, the fact that a BIG would make people less likely to work is a killer. Even if it’s a genuinely good idea, it’s simply never going to get the votes.

In what follows, I will challenge this argument at two different levels. First, I will argue that neither economic theory nor the available empirical evidence succeeds in demonstrating that a BIG would lead to a significant reduction in work effort. Second, I will argue that even if a BIG does lead to some reduction in work effort, this would not necessarily be a bad thing. It might be a cost, but a cost that is outweighed by other social benefits. I will also argue that in some cases it is not something we should properly characterize as a cost at all.

What Does Theory Predict? What Does the Evidence Show?

In theory, it seems like it should be obvious that a BIG would reduce work effort. As Jim notes, if you guarantee people an income whether they work or not, it would be fairly extraordinary if people didn’t work less.

But this is only half the story. As Ed Dolan has noted elsewhere (and in this very helpful comment on Jim’s post here), a basic income actually has two effects on the labor market, and they push in opposite directions in terms of incentives to work. One effect, the income effect, leads people to work less as their wealth increases. But there is also a substitution effect that leads people to work more as their after-tax hourly earnings increase. Relative to our current welfare system, a BIG would create a larger income effect for most people (depending, of course, on the details of the BIG scheme adopted). But a BIG would also create a stronger substitution effect, by decreasing the very high effective marginal tax rates many welfare recipients face under the current system. Whether a BIG causes a decrease in work effort or not depends, in theory, on which of these two effects is stronger.

So what does the evidence show? Well, as Jim notes, the NIT experiments of ’68-’80 resulted in an overall decrease in work effort. This much is uncontroversial. What’s a little less clear is the magnitude of the decrease, and how we ought to evaluate that decrease from the perspective of public policy. I’ll take up the second of those questions in the next section. For now, here are a few important points to bear in mind when trying to answer the first.

• By most accounts (see this paper by Gary Burtless of the Brookings Institution for a good overview), the reduction in work effort was small – about 120 hours less per year for husbands, about 90 less for wives, and about 130 less for single mothers. That’s not nothing. But it’s about the equivalent of taking a few weeks off from a 40 hour/week job. And that’s far smaller an unemployment effect than most economic models had predicted.
• A “reduction in hours worked” is not the same as “more people spending their lives on the dole.” Most of the reduction of work effort was not caused by people exiting the work force altogether. What happened, instead, was that periods of unemployment just lasted longer. That might still be a problem. But it might not. I’ll have more to say on this in the next section.
• A truly unconditional BIG would reduce work effort less than a NIT. The amount of money you get from an NIT depends on how much you earn by working. The more you earn, the less you get. This keeps the cost of a NIT down, relative to a truly unconditional BIG. But it also means that you lose one of the major benefits of a BIG. Because you keep your entire BIG no matter how much you earn, there’s no extra disincentive to work caused by the threat of losing some of your benefit. We should therefore expect work reductions to be smaller under an unconditional BIG than a NIT. But the experiments to which Jim points only tell us about how people responded to the latter, not the former.
• People worked more than they said they did. In the NIT experiments, the size of recipients’ benefits was determined by the amount of money they reported earning – to a survey company! This, of course, created a fairly strong incentive to underreport one’s income and hours worked. And, indeed, when reported earnings were cross-referenced with more reliable data from the unemployment insurance system, much of the apparent unemployment effect disappeared altogether.

In sum, neither economic theory nor the available empirical evidence make a compelling case for the conclusion that a BIG would cause a significant reduction in work effort.

Is Less Work Necessarily Bad?

Jim Manzi thinks that the work disincentives associated with a BIG are a good reason to reject it – or, at least, a good reason to believe that it is unlikely to gain much favor with the American public. Robert Frank agrees, and advocates a system in which the government serves as an employer of last resort as an alternative to both a pure BIG and the current welfare system.

Are they right? Suppose that a BIG really did cause a significant reduction in work effort. Would this necessarily be a bad thing?

It might. Work is important, in the first instance, because and to the extent that work turns less-useful things into more-useful ones. It is important in a secondary way insofar as meaningful work is one of the central components of human well-being. People feel better about themselves when they spend a good portion of their day doing something useful. And, in a market economy, doing something useful usually means making not only oneself better off, but others better off in the process. Such a fantastic coordination of mutual benefit is something to be celebrated. But it is also something to be protected, for it is neither inevitable nor indestructible.

Still, when it comes to employment disincentives, size matters. And we have already seen that, even in the case of conditional grants like those studied in the NIT experiments, unemployment effects are relatively modest in size. Whether those costs are worth bearing, then, depends on the size and nature of the offsetting benefits.

One of the main benefits, of course, is the freedom to do things other than work without being destitute. For some people that will mean the ability to go back to school and improve their employment prospects. For others, it will mean the ability to spend a few more months with their newborn child. For still others, it might simply mean just a bit less pressure to find another job after being laid off, and a bit more leeway to wait for a job that’s a better match.

Some people, of course, might simply use the freedom that a BIG gives them to “waste their time” sleeping in and watching television. Brink Lindsey cites some data to this effect, presumably to show that people freed from the necessity of work by a BIG would probably just squander it. But even if we grant (and only for the sake of argument!) that people’s lives would not be improved by a little more sleep and relaxation, the data Brink cites don’t show what he thinks they show. That data is about what people who are unemployed now do with their time. But it tells us precious little about what the (presumably different) people would do with the extra time made available through a BIG.

Work is a central component of both personal and social well-being. But it is not the only component. And it may well be a component that we, as individuals and as a culture, tend to err on the side of over­-valuing. John Maynard Keynes thought so when he wrote over 80 years ago that the real, permanent problem of mankind is “how to use his freedom from pressing economic cares, how to occupy the leisure, which science and compound interest have won him, to live wisely and agreeably and well.” And while invoking Keynes might not be the best way to win over the audience of the Cato Institute, a similar sentiment can be found almost fifty years earlier expressed by no less a libertarian hero than Herbert Spencer, who pleaded with his American audience to embrace a new “gospel of relaxation,” and to remember that “life is not for learning, nor is life for working, but learning and working are for life.”

Like most other issues of social policy, then, the issue of work disincentives involves questions of trade-offs. One cannot criticize a BIG merely by showing that it would create some disincentives to work. One must show that it would create too much – that the moral, economic, and cultural costs of those disincentives would not be sufficiently compensated for by the moral, economic, and cultural benefits. Making this determination necessarily involves us, as Jim points out, in questions of value. But unlike Jim, I do not believe that questions of value are necessarily subjective. Some systems of value exhibit a greater degree of internal coherence than others; some better accord with what we know about human nature and the nature of the world we live in; some yield better explanations of our deeply held moral commitments. In short, we can rationally engage with and evaluate systems of values. Without this kind of rational engagement with questions of value, of ends, all the means-end economic analysis in the world will not get us anywhere on the important questions of public policy.

Robert Frank writes,

[E]quating taxation and theft was not a compelling position in the first place. A country without the power to tax couldn’t field an army and would soon be overrun by a country that had one. Its residents would then have to pay taxes to that country.

Presumably, the second and third statements above are supposed to support or explain the first statement … only they don’t. “If I don’t do X, then something bad will happen; therefore, X is not a form of theft” is an invalid argument. So is “If I don’t do X, then someone else will commit a theft; therefore, X isn’t theft.” These are non sequiturs.

At best, the premises might be said to show that taxation is justified even if it is theft. That is a perfectly understandable view. But that won’t do to support the rest of what Frank wants to say: namely, that a basic-income-plus-public-work scheme is ethically permissible. From the premise that taxation is justified to prevent a Russian invasion (or similar event), it does not follow that taxation is justified to implement a social welfare program.

Compare this case: Jean Valjean steals a loaf of bread to feed his sister’s children. Assume that the children would otherwise have starved. It does not follow from this that he didn’t really steal the bread. At most, what follows is that the theft was justified. But one cannot now go on to infer that it would be permissible for me to steal bread for some entirely different purpose – for instance, to feed the ducks at the local pond. Stealing to feed the ducks might be justified nonetheless – but some other argument would have to be given.

Similarly, stealing in order to provide social welfare might be justified, but some argument other than the appeal to national defense would need to be given.

I like Mike Huemer’s book a lot. I think it’s one of the best books of libertarian political philosophy to have been written in the last twenty years. And I like Mike’s response to my essay here too. I especially like the part where he concedes the central thesis of my essay: that a Basic Income Guarantee (BIG) would be better, on libertarian grounds, than the current welfare state.

In a sense, everything after that concession in Mike’s essay is somewhat beside the point. Mike spends most of his essay talking about what a political system ought to look like from the perspective of first-best principles of justice, and argues that such an ideal system would leave no room for a BIG. But the point of my essay was not to argue about what an ideal system would look like. It was to argue about what a better system would look like.

Still, I did open the door to Mike’s line of argument when I suggested in my original essay, without any supporting argument, that a BIG might properly be part of an ideal libertarian system. So let me take this opportunity to say just a little bit more about why I think that is true, and why Mike’s criticisms are not sufficient to convince me otherwise. In this essay, I’ll focus on his first and most fundamental argument – that states lack the legitimacy to impose a BIG. I probably won’t say much more about the reparations argument, since I’ve written about that at length elsewhere. But I hope to say more about the Hayekian freedom-based argument in a future post.

Political Authority and Property Rights

Mike argues that a BIG is morally impermissible because states lack political authority. Enacting a BIG would require taking money by force from some people and giving it to others. And that is something that you and I, as individuals, are generally[1] not morally permitted to do. Since we as individuals are not permitted to engage in coercive redistribution ourselves, neither are states unless there is something special about states that exempt them from the ordinary rules of interpersonal morality. Mike refers to this special something as “political authority,” and the first part of his book is devoted to explaining why he thinks that it doesn’t exist.

I think this is a neat argument, and I find Huemer’s arguments against social contract and consequentialist justifications of political authority very convincing. Notice, though, that Huemer’s argument against the BIG only works if we presuppose the existence and legitimacy of property rights. The reason states need special authority to impose a BIG, on Huemer’s view, is that funding it through coercive taxation would violate the property rights of taxpayers. But nowhere in his book or in his contribution to this debate does Huemer tell us why or how he thinks those property rights are justified.

No philosopher should be expected to argue for everything, especially in as short a piece as he was asked to write for this forum. But property rights are of fundamental importance to this debate because, as other critics have noted, they share a lot of features in common with the kind of political authority Huemer wants to reject. Political authority needs justification because its exercise involves the coercive limitation of individual freedom. But so does the exercise of ordinary rights of property. Political authority imposes unconsented-to moral duties on other persons; so do property rights. When I claim a property right in a piece of land, I claim the right to make the exclusive determination how that land is used, and by whom; I claim that you have a moral duty to act in accordance with this determination; and I claim the right to enforce my determination by the use and threat of physical force.

Justifying the Authority of Property

None of this would be a problem if we could give an argument for why property rights are obviously justified, but political authority isn’t. But this turns out to be surprisingly difficult to do. The standard Lockean accounts of property to which libertarians so often appeal face a serious challenge – a challenge that found its most influential expression in the writings of Henry George and which libertarians still have yet to answer satisfactorily: why should the fact that individuals own their bodies and their labor give them permanent, bequeathable property rights in external resources that their labor did not create? Why should the mere fact that you exerted labor on a pre-existing object give you the right to coercively prevent others from using it, and why should it give them a duty to refrain from using it without your consent?

Of course, not even Locke held that labor-mixing was sufficient to justify property. In order for an act of labor mixing to generate a legitimate property right, it also had to satisfy the so-called “Lockean Proviso” of leaving “enough and as good for others.” What exactly that means is a matter of some dispute. But one popular way of understanding it is that an act (or system) of appropriation must not make others worse off in order to be legitimate.

And this opens up the door for a very powerful argument on behalf of property rights. Since property rights help us to avoid the tragedy of the commons, and since they facilitate the production and trade that makes economic growth and prosperity possible, perhaps the Lockean Proviso really does get satisfied after all. Perhaps X’s original appropriation of a plot of land not only doesn’t make Y worse off; perhaps it makes Y positively better off.

Those of us born in the last hundred years are Y’s. By the time we arrived on the scene, most of the surface of the earth had already been appropriated by earlier X’s. But how many of us would really want to trade places with an X from 17th century America if we had the chance? Is it really plausible to maintain that we are the ones who got the raw end of the deal?

For most X’s, and most Y’s, I think this is a pretty plausible story. Most of us are far, far better off than the original appropriators who preceded us, in terms of welfare and even in terms of liberty. More importantly, we are better off because of those prior appropriations.

From Lockean Proviso to Social Safety Net

But while this is probably true for most Y’s, it is not obviously true for all. Rights of private property generate increased welfare and freedom for most. But might there not be some who slip through the cracks? The Lockean proviso is an individualistic principle, not a collectivist or utilitarian one. Simply arguing that it benefits most people is not enough. If there are some people for whom the proviso is not satisfied, then those persons have a moral claim against the rest of us. And it is a claim of justice, not of charity.

Here, then, is the sketch of one possible justification for social safety net, which could take the form of a BIG. A safety net is necessary in order to ensure that the Lockean proviso is satisfied for all persons, and hence to ensure that the property rights we take for granted are truly justified. The aim of such a safety net would be modest. It would not be to establish some sort of equality of condition. Indeed, it would not be aimed at any kind of equality at all. Rather, the goal would be to provide individuals with a sufficient level of opportunity – to ensure that they have at least as good a chance to live a free and flourishing life as they would have had if the rest of us had not coercively imposed a system of property rights on them.

This account is merely a sketch. Fleshing it out would require a more detailed account of the Lockean proviso, the precise nature of the counterfactuals involved in assessing questions of “better” and “worse” off, and the exact kind and extent of safety net that could be justified by an argument of this sort. Though I do not agree with everything in them, these papers by Daniel Layman and Peter Vallentyne provide a good start at working out those details.

I’ll close by noting that this sort of argument also points in the direction of a natural answer to the question of how a BIG should be funded. If the point of a BIG is to ensure the satisfaction of the Lockean proviso, and if the Lockean proviso is, after all, a proviso on the appropriation of natural resources, then perhaps a tax on the unimproved value of natural resources is the most morally defensible (not to mention economically efficient) method of raising revenue for a BIG.

In the end, perhaps Henry George had it right.

Note

[1] Mike does concede, in a fascinating discussion in chapter 7 of his book, that individuals are sometimes permitted to engage in coercive redistribution to relieve imminent, dire distress. But he doubts that this concession provides much in the way of justification for the welfare state.

Many thanks to Matt for his thoughtful replies. I want to start by addressing/dodging questions about property rights. The beginning of Matt’s most recent post is reminiscent of some earlier remarks by Kevin Vallier. My response to Matt on the subject of property rights is therefore similar to my earlier response to Kevin: I am not giving a theory of property, mainly because

1. It is reasonable to take property rights for granted here, since there is hardly any thinker who rejects them (and certainly not either Matt or myself);
2. If there were no property rights, taxation would be impermissible anyway – not because taxpayers would have rights to their pretax incomes, but because the state would fail to have any property rights over the taxes that they claim citizens owe them.

But I won’t further explain those points now, since I discussed them in my earlier response to Kevin Vallier.

I suspect that some readers misunderstand my central argument because they confuse me with other libertarians. My central line of thought does not start out like this: “My God, coercion is so awful, how can it ever be justified?” My central line of thought starts out like this: “Wow, we seem to have much more permissive attitudes toward the state than we have toward any other agent. How can that be justified?” To point out that coercion is needed to enforce property rights really does not show an interesting parallel between belief in property rights and belief in authority, because in endorsing property rights, we are not setting up some agent with special entitlements and moral exemptions that apply to no other agent. Those who endorse political authority must answer “what is so special about this one agent?” Those who endorse property rights face no such question.

That being said, Matt has nevertheless articulated what sounds to me like a libertarian moral argument for some sort of wealth redistribution, which is not undermined by anything I have said so far.

Actually, there seem to be two distinct arguments in this vicinity: first, there is the argument that some individuals may have been rendered worse off overall as a result of our system of property rights than they would have been … under some alternative scenario (I’m not sure what the relevant alternative is – the scenario in which those individuals got to appropriate everything? the scenario in which no one ever appropriated anything?) However, it seems that Matt and I agree that only very few individuals would satisfy this description (on any plausible interpretation of the relevant alternative scenario). So I don’t see how this argument could justify anything like Matt’s basic income proposal, which, as I understood it, is supposed to provide a basic income to everyone.

Second, however, there is the argument that I take to be Henry George’s: no one really has any better claim on any bit of natural resources (including especially land) than anyone else. Therefore, everyone is entitled to benefit equally from the value of natural resources (but only of natural resources; value contributed by human activity belongs by default to those who perform the labor). Therefore, it is appropriate to distribute to everyone a sort of basic income: roughly, each person should receive the nth part of the total natural resource rents of the world, where n is the number of people in the world. This is assuming that we can separate the portion of the total economic product that is due to natural resource rents from the portion that is due to other causes.

As a piece of moral reasoning, I have some sympathy with this argument. However, I think it just does not matter very much, because I think the amount of income that a proper Georgist scheme would provide would be minimal – far too small to satisfy the advocates of a social safety net. The reason is that I think the amount of economic value that is attributable to natural resource rents is tiny. Unfortunately, I am not in a position to provide any serious estimate of these rents. Instead, I will just note some ways in which the value of these rents has often been overestimated; I hope that then most readers will agree that a just estimate would be quite small.

Here is a very simple way of overestimating natural resource values: look at the price of some “natural resource” on the commodities exchanges. For example, gold costs $1300 an ounce. But this is the price of the resource after all the work of extracting and purifying it has been done.

Another way of overestimating natural resource values is to consider the value of a parcel of land (with resource-extraction rights included) that is known to have some interesting resource underground, such as oil or gold. Such land can indeed be quite valuable, even before anything has been built on it. However, discovering the location of such interesting resources requires surveying, which is a form of labor (see Caplan and Gochenour’s discussion). To truly capture the unimproved value of land, one must consider the economic value of the land before any surveying has been done, and thus before it is known whether that land has any interesting resources under the ground.

A third way of overestimating natural resource values (specifically, land values) is to consider the value of an empty plot of land in some developed area. For instance, land in some parts of New York City is worth $800 per square foot before anything is built on it. However, almost all of this value is due to human activity, especially the activity of other people who live and have lived in New York. This is shown by the fact that land (with no fewer natural resources) in completely undeveloped areas has much lower value. You might be able to buy a plot of wilderness land in Montana for 8 cents per square foot. If we apply the Georgist rationale strictly, the 8 cents is more relevant than the $800. But even the 8 cents is an overestimate, since some of that value is probably due to human activities (if the rest of the continent were devoid of human occupation, the same plot of land would probably have even lower economic value).

In light of these reflections, I do not think that anyone should expect a livable salary from their fair share of the world’s natural resource rents.

Matt Zwolinksi wrote a brief in favor of the BIG. I replied. Matt responded to objections raised by Robert H. Frank and me.  This is my reply to Matt, with respect to the parts of his response directed to my arguments.

In his current response, Matt now accepts that it is “uncontroversial” that the Negative Income Tax (NIT) has been shown in randomized experiments to cause a reduction in work effort.  But he first raises some technical issues: (1) the decrease was small, (2) it was caused primarily by non-employed people remaining unemployed longer, (3) it would be less under an unconditional BIG, and (4) it was subject to bias because it was survey-based, and that under alternative measurements using administrative data, “much of the apparent unemployment effect disappeared altogether.” [italics in original].  He then proceeds to return to the argument in the alternative from his original essay that a reduction in work is not necessarily a bad thing.  I’ll address the technical issues first, and then move on to the values argument.

Matt’s technical issue (4) seems to undercut his acceptance that a reduction in work effort in these experiments was uncontroversial.  To my knowledge, two of the NIT experiments were subject to post hoc reanalysis in which administrative data was used after the fact to try to create different impact estimates that were then compared to survey-based estimates.  Contrary to my reading of Matt’s statement, in neither case would the administrative data have made the unemployment effect “disappear altogether.”  Even according to the paper that Matt cites as evidence, in one of these experiments “much” of the employment effect would disappear if measured using administrative data, and in the other the change in effect was “somewhat smaller.”  The NIT experiments were designed from the start to measure effects using surveys (and often in fact set up a parallel tracking system), and it turns out that there are general pros and cons to using survey versus administrative data in social experiments.  Further, as I reviewed in Uncontrolled, the later series of more than 30 welfare experiments in the 1990s were designed from the beginning to be read using administrative data, and these experiments consistently found that the absence of work requirements led to statistically significantly lower employment and earnings using this data.  There is no credible interpretation of the NIT experiments and later welfare reform experiments in which we do not observe a causal reduction in work effort in response to the elimination of various kinds of work requirements. 

Matt says, however, that this reduction in work effort was small and would be even smaller under an unconditional BIG.  One problem with this is that the NIT experiments measured the impact of this policy change versus the baseline of the American welfare system circa 1975.  The current welfare system has much more stringent work requirements, which were proved in the1990s welfare experiments to increase work effort. Therefore a hypothetical 2014 BIG would be expected to have a larger reduction in work than the same system would have had in 1975.  I understand (though I do not accept) Matt’s argument from theory that somehow a truly unconditional BIG would partially offset this, because it would have, ceteris paribus, less impact on work effort.  I think any rational observer should remain unconvinced on this theoretical point until it is demonstrated experimentally – unlike similar theories that were disproved in the original NIT experiments.

In the end, all of these debates are about quantity, not sign.  All reliable evidence seems clear that a BIG would reduce work effort. 

This brings us to the second part of Matt’s argument – less work would not necessarily be a bad thing.  As Matt correctly notes, this is not completely independent of the size of the work reduction.  If a BIG would cause a total reduction of one person-hour per year in all of America, it would technically reduce work, but would presumably not matter to us.  Matt argues, again correctly, that we must compare the costs of work reduction (and presumably other drawbacks) to the benefits created by a BIG.

But Matt’s argument breaks down both in the method of measurement of this trade-off, and in where he places the burden of proof.

With respect to method, I think the key passage form Matt’s response is the following:

Work is a central component of both personal and social well-being. But it is not the only component. And it may well be a component that we, as individuals and as a culture, tend to err on the side of over­-valuing. [Italics in original]

This is closely related to my original critique that Matt wants to assume away democratic governance.  Matt may be right, and the American electorate may be wrong, but it is clear that they disagree on this point.  To grossly oversimplify, the amount of work disincentive is prohibitive versus the benefits of leisure in the view of an electorate that has never supported a BIG in two hundred years, and has been moving away from this concept and toward stricter work requirements for decades.  Matt needs to change a lot of minds before this would become law in a democratic society. 

So what is the basis for Matt’s appeal to the electorate that they should change their views?  Matt writes that:

Like most other issues of social policy, then, the issue of work disincentives involves questions of trade-offs. One cannot criticize a BIG merely by showing that it would create some disincentives to work. One must show that it would create too much – that the moral, economic, and cultural costs of those disincentives would not be sufficiently compensated for by the moral, economic, and cultural benefits. [Italics in original]

Sometimes it is hard to determine the reference case in policy debates, but this is not one of them: Matt is self-consciously proposing a radical departure from current policy and established practice over centuries. He is then asserting that he can demand that those who oppose the change must provide evidence that it would have net negative effects. The burden of proof lies with Matt, and he has yet to make his case.  I will look forward to hearing it from a writer as persuasive and talented as he is.

Most of the economic value of land is due to the benefits of living or working in a high-population-density area. Roughly, the reason why it is beneficial to live in a high-population-density area is that this makes feasible a greater variety of mutually beneficial trades. Call this value “population density value.” Now, there is the question of who is entitled to the population-density value – should landowners capture it, or should it be redistributed?

(A) One view (which I discussed in my last post) is that this value should not be subject to the Georgist land tax, because it is not part of the unimproved value of the land; it is produced value. However, on this view, as I’ve suggested, a proper land tax is only going to collect a pittance.

(B) Another view is that landowners have no special claim to the population density value; they aren’t more responsible for the existence of population density value than the non-landowning members of the same community. Therefore, it is fair to tax away that value. This appears to be the assumption of all those who give large estimates for the proceeds that would be gained from a land tax (including one of the commenters on my previous post).

This latter view, however, still won’t provide much comfort for basic income proponents. Here are three reasons for this:

1. Most current landowners have not really received a large portion of population density value, since the price they initially paid to purchase their land already reflected the population density value as of the time of purchase. It would only be fair to tax them for the increment in population density value since the time of purchase, due to new people moving into the area. Thus, the amount collected would be much smaller than indicated by estimates based purely on the current land value.

2. There is no moral reason (at least not along the lines we’ve been considering) why people in low-population-density areas should receive transfers from people in high-population-density areas. Population-density value is created by people choosing to live near other people. Thus, if we’re following the idea that people should profit from the value that they produce, the people in rural areas should receive little.

3. Most importantly, since population-density value is due to trade opportunities, people who do not contribute to the trading opportunities in their community should presumably not receive a share of the population-density value, and people who make a greater contribution to those opportunities should receive a greater share. This is all based on the assumption, roughly, that individuals ought to capture the value created by themselves.

So for example, unemployed people should receive little or none of the population-density value, since the addition of unemployed residents to a city does little or nothing to increase trade opportunities for others.

Thus, if we rest the case for a basic income on this kind of (libertarian, Georgist or semi-Georgist) rationale, the people whom the basic income proponents are most concerned to help are the ones least entitled to receive it.

Mike Huemer sympathizes ith the core moral claim of Georgism, but he doubts that a Georgist Land Value Tax (LVT) would generate sufficient revenue to fund a BIG. His main concern is that “most of the economic value of land is due to the benefits of living or working in a high-population-density area.” A piece of land in New York City is valuable primarily because there are a lot of other people around with whom the owner can engage in mutually beneficial exchanges. Without those other people around, the land would not be nearly as valuable. But this means that there is a sense in which the main source of land’s value is the productive activity of human beings. And if Georgist moral principles only license the taxation of the unimproved value of land, then most of land’s actual value would appear to be off limits.

I think Mike’s empirical claim about the source of land’s value is correct. And I think it demonstrates an important point about the importance of trade and the division of labor in the development of human prosperity – a point most famously expressed by Adam Smith and beautifully developed by Matt Ridley in his magnificent book The Rational Optimist. However, I don’t think Mike’s point undermines the Georgist defense of a LVT, or my claim that a LVT might suffice to finance a reasonable BIG.

Mike brings up the issue of population density to show that a large part of land’s value depends on the productive activity of other people. But, really, all of land’s economic value depends on the productive activity of other people. Land only has economic value, after all, to the extent that other people are willing and able to pay for it. And the only way other people can put themselves in a position to be willing and able to pay is by producing something themselves.

The same point can be made in a different way with respect to particular aspects of land’s value. Land with subterranean crude oil deposits is valuable primarily because other people have figured out how to put crude oil to productive use. Before that discovery was made, oil was a pollutant, not a resource. The value of a landowner’s oil is thus largely a positive externality created by others’ labor.

So Mike’s point applies more broadly than he gave it credit for. Still, the key question is whether landowners are entitled to the external benefits created by others’ labor, or whether such benefits are the legitimate object of a LVT. The argument I’ve given above shows that the stakes of this question are higher than Mike supposed, but it doesn’t tell us how to answer it.

I think the correct answer must be that, on Georgist principles, the landowner is not entitled to this external benefit. Her ability to capture that benefit depends on putting a fence around a piece of land in order to keep others out. And this, ex hypothesi, is something she has no moral right to do. It’s not benefitting from the positive externalities of others’ labor that’s the problem. It’s the forcibly excluding others from doing so.

This explains why, contra Huemer, people in low-population-density areas are entitled to compensation from people in high-population-density areas. High-population-density land is, ceteris paribus, more valuable than low-population-density land. So when you put a fence around a piece of land in Manhattan, you’re depriving other people of a more valuable resource than you do when you put a fence around a piece of land in West Virginia. Because you have deprived others of a more valuable natural resource, you owe them more in compensation.

But Mike has one more argument up his sleeve. Almost all current landowners purchased their land from someone else. And that purchase price reflected the discounted future value of the then-current land rent. If we instituted a LVT tomorrow, we would wind up taxing landowners on rents that they had already paid for. This seems wrong to Mike. It would only be fair, he suggests, to tax them on the rents that accrue to their property after their purchase. For it is only those rents that constitute a real windfall to the owner.

Mike has identified a real problem here. But it is not by any means a problem that is unique to Georgism. It is a problem that besets any theory that specifies an ideal of justice in conflict with the status quo. Suppose you think, like I do, that the medallion system that cartelizes the taxi industry is unjust. What should be done about it? Taxi drivers did, after all, pay for their medallions. So if we abolish the system, we’ll be depriving them of a resource into which they invested real time and labor. On the other hand, they had no right to that benefit in the first place. So I am more inclined to think that the correct answer is – abolish the cartel as soon as we can. And to those who are harmed by the loss of their monopoly status, we say, “Be grateful you were able to reap monopoly rents for as long as you did. But from now on you can compete for your income like everybody else.”

On Georgist principles, isn’t that the correct response to the landowner, too?

I have described my argument for a Basic Income Guarantee (BIG) as a “pragmatic” one.  But a pragmatic proposal ought to be one that actually has some chance of being carried out successfully. And many doubt that the BIG meets even that bare minimum criterion of pragmatism. By what political process is the wholesale replacement of the welfare state with a BIG supposed to take place? How could such a proposal make it unscathed through the sensationalism, log-rolling, and rent-seeking of real-world political processes? And even if it did emerge unscathed, how long would it stay that way before being mutilated beyond all recognition by the plastic-surgery-by-committee of ordinary politics?

Concerns about the feasibility of a BIG seem to take one of two different forms, both of which find forceful and eloquent expression in these essays by Jim Manzi and Tyler Cowen. The first form I’ll call moral feasibility, and it has to do with whether it would really be possible, once we think through all of the details, to implement a BIG in a morally attractive way. The second is what I’ll call simple feasibility, and it refers to whether a proposal such as the one I have articulated for a BIG could be implemented at all in a way that faithfully executes the original idea. In what follows, I will examine each of these forms of feasibility in turn.

Would a Realistic BIG Be Morally Attractive?

My original essay set out four reasons for thinking that a BIG would be superior to the current welfare state from a libertarian perspective. In response, critics have argued that there are three ways in which the actual effects of a BIG might be morally worse: A basic income might reduce the incentive to work; it might grow and develop in unattractive ways due to various political pressures; and it might fail to provide for the special needs of various segments of the population.

Work Disincentives

I’ve discussed the issue of work disincentives already, and argued that both the empirical and moral premises of the argument are overstated. That said, I do find the three kinds of concerns raised by Michael Tanner extremely persuasive. There does seem something fair about requiring those who benefit from society to contribute their share; work is an effective way for families to escape poverty; and even small changes in work output can have large aggregate effects on economic growth – the force which has done more than any government program ever has or will to eliminate poverty in the past, and which still has much vital work to do in the future.

Still, it’s worth remembering that whatever benefits encouraging work might yield cannot be achieved without incurring serious costs, both moral and economic. If state relief is dependent on willingness to work, then the state needs to decide what counts as “work” and what doesn’t. Is watching your grandchildren while their parents work in exchange for a little bit of money a job or a favor? Making – and enforcing – determinations like that requires a degree of state power and discretion of which libertarians are rightly skeptical. And what about people who genuinely can’t work? If we don’t want such people to be cut off from assistance, then now the state must be in the business of making that determination too: who has a “real” medical problem, and who’s just faking it; whether psychological conditions (and which ones) count as genuine excuses for not working; whether that basketball game that inspectors recorded you playing in your front driveway proves you really could go back to work on the construction site, and so on.

Of course, Jim is right that work requirements are popular with American voters, and they have been for a long time. But the trade-off between work and leisure is, like all trade-offs, one that must be made at the margin. And even if we grant the assumption that the balance struck by our current policies worked well for us in the past, that doesn’t entail that it’s right for our current levels of wealth and economic development.

Perhaps libertarians ought to take a lesson from Denmark. Denmark’s generous but relatively efficient social welfare state no doubt plays some role in explaining the fact that Danes work fewer hours per year than workers almost anywhere else in the world. And yet, for all the talk of work’s centrality in a flourishing life, the citizens of Denmark do not seem to be suffering from any existential crisis of meaning. In fact, they’re arguably the happiest people on earth. They are also arguably some of the most economically free people on earth – at least, much freer than us here in the United States, on almost every measure, from respect for property rights to trade and investment. Of course there are complicating factors, and what brief remarks I’ve made here are far from decisive. My point is simply that it’s not as though advocates of a BIG are entirely without evidence for our belief that a little less work, and a little more leisure, would make us better off on net.

Political Pressure

Michael Tanner succinctly expresses a common concern about a BIG:

A UBI would establish as both a legal and a philosophical concept that every American citizen is entitled to a minimum income - exacted from the taxpayers. Once that “right” is established, the political process will inevitably expand it.

This is a legitimate concern. But it is a valid reason to reject a BIG? On this issue, I can do no better than quote the late, great Milton Friedman, who responded to similar objections to his own Negative Income Tax proposal in the following way:

Clearly the dangers exist. But … they must be evaluated in terms of the world as it is, not in terms of a dream world in which there are no governmental welfare measures. The relevant political question is whether the negative income tax is more susceptible or less susceptible to these dangers than alternative programs of the kind we now have or are likely to get. To this question the answer seems to me clear: The negative income tax is less susceptible.

Why has the present grab-bag of programs been adopted? Because each appeals to a special interest that is willing to fight strongly for it, while few are willing to fight strongly against it; because the disinterested who seek to promote the general interest have been persuaded that every measure will contribute to helping a group that is disadvantaged; because, for many of the measures, there was no clear price tag, and the program could be voted without the simultaneous imposition of taxes to pay for it; because, for still others, there are no direct costs at all - minimum wages are perhaps the clearest example; because, finally, no one who opposed the programs had an effective alternative to offer that would meet the real problems.

Politically, the right solution is to have a comprehensive program whose cost is open and clear.[1]

This, incidentally, is the fundamental problem with Pete Boettke and Adam Martin’s critique of the basic income in their otherwise excellent paper. They are quite right to point out that a BIG would be vulnerable to political opportunism and manipulation. But that’s not what matters. What matters is whether a BIG would be more vulnerable to such forces than our current system, or more vulnerable than other proposed reforms. The relevant baseline is the status quo, not libertarian utopia. And all reforms to that status quo, whether they aim at “privatization” or a basic income, will inevitably be filtered through a political process that special interests will try to manipulate to their own advantage. The question is not which reform is perfectly immune to these forces; the question is which is better able to withstand them. And there is good reason to think that a general, simple, and transparent program like a BIG would be considerably less subject to political manipulation than what we have today.

Special Cases

But concerns about special cases are not merely an opportunity for opportunism. They are often, as Jim Manzi nicely puts it, the result of “healthy intuitions of natural justice that are essential to maintaining a well-functioning political order.” It’s not crazy to think that people with severe disabilities ought to get a bigger helping hand than those who are healthy; to think that those who have supported others through their work and taxes have a better claim than those who have not; to think that special arrangements need to be made to ensure the material security of children, or to enable people to cope with skyrocketing medical costs, and so on.

Precisely because of its general nature, a BIG is ill-suited to deal with these special cases. For progressives who favor a BIG in addition to our vast array of current welfare programs, this is not much of a problem. For them, the BIG merely provides an extra degree of freedom and security above and beyond the more particular needs served by other programs. But for libertarians who seek to replace those welfare programs with a BIG, this is a serious challenge. Will a BIG, all by itself, be enough to meet people’s legitimate needs? Can a BIG really be an effective political compromise between libertarians and progressives if it leaves those needs unmet?

Without in any way minimizing the seriousness of this challenge, I think that libertarians have two good responses available to them to meet it. First, the classical liberal critique of “expediency” as expressed by Hayek, Spencer, and others is fundamentally and importantly correct. A public policy guided solely by considerations of expediency is, at the end of the day, simply not expedient. It’s easier to rouse up excitement about the predictable benefits of a policy than it is to induce caution about its unpredictable costs. And exceptions, waivers, and ad-hoc adjustments that look perfectly reasonable when considered in isolation often add up to systems of policy that no one would wish to defend. Paradoxically, the only way to craft an expedient system of policy is to put considerations of expediency to the side and anchor one’s legal system in a set of clear, simple principles.

Second, while it is easy to think up situations of genuine need that a BIG would fail to address, at some point the basic libertarian response just has to be correct. It is not job of the state – it cannot be the job of the state – to guarantee that no one’s genuine need ever goes unmet. The state can promise a guarantee through legislation, but as my old friend and mentor David Schmidtz pointed out, even the promise of a guarantee is not a guarantee itself. And promises have their costs.

Still, the fact that not all needs can be met by the state does not mean that they must go unmet altogether. Indeed, part of the appeal of a BIG for many libertarians and conservatives will be the space it creates for individuals, families, and communities to help themselves. This is not a call for “atomism” or “rugged individualism.” It is, instead, a call to create the conditions necessary for a vibrant and flourishing civil society in which both self-help and mutual aid can play a greater role in meeting those needs that demand more than the basic minimum provided by a BIG. 

Can We Do It?

But all of this is idle talk if a BIG has no realistic prospect of being adopted in any form, no matter how bastardized. And one could certainly be forgiven for thinking that it doesn’t. No major politician in the United States has endorsed the idea. There isn’t any big corporate money behind it. And most ordinary voters still haven’t even heard of the idea. So how could it possibly go anywhere?

It’s a long shot, to be sure. But it’s a long shot that’s made even longer when people who think it’s a genuinely good idea give up on it because they don’t think it has any practical chance of success.

Many libertarians want to abolish the welfare state altogether. Many progressives, on the other hand, would be perfectly happy to see a BIG added on top of our currently existing welfare state. The proposal to replace the welfare state with a BIG might not be anybody’s first-best solution. But that, in a sense, is what makes it promising as a piece of political compromise. A BIG represents a potential overlap between moderate libertarianism and moderate progressivism.

Moreover, from a libertarian perspective, it represents a compromise with the best elements of progressivism. You don’t have to sign on to the problematic belief that inequality per se is a problem to endorse a BIG. A BIG simply tries to ensure that people have enough, not that they have the same. And that, both philosophically and politically, is a far more defensible view. Moreover, a BIG allows libertarians to endorse the admirable progressive goal of a social safety net without the traditional progressive means of achieving that goal through technocratic micromanagement of the economy.

Even still, the wholesale replacement of the welfare state through a constitutional amendment is unlikely to happen anytime soon. And while constitutional decisionmaking behind a “veil of ignorance” has its advantages, there’s something to be said for slow, gradual change informed by experience too. Jim Manzi is on solid ground when he praises the virtues of experimentation.

So if you’re not comfortable replacing the welfare state with a BIG, or if you just don’t think it’s going to happen, what can you do? Well, if we can’t trade the welfare state for a BIG, we can at least make the welfare state – shall we say – BIGger!

Of course, by this I mean “make the welfare state more like a BIG.” As Jesse Walker has recently pointed out, we can do this by cashifying and combining programs. Take programs like food stamps and Section 8 housing vouchers, set common qualifications for them both, and replace the vouchers with cash. That’s not quite the BIG that I proposed. But it’s closer than what we’ve got, and it approximates the pragmatic virtues of a BIG.

Carving out a larger role for cash transfers in the welfare state would shrink the size and administrative cost of government, reduce the intrusiveness and paternalism of the welfare state, and give recipients more dignity and control over their own lives. As an added bonus, it would provide a greater opportunity for us as a society to learn. If you think cash transfers are a bad idea, then let’s start small, see what happens, and go from there. At worst, the experiments fail on a small scale with minimal harm. At best, they work out, and we have a few more data points to support the case for a wider, more comprehensive BIG.

In theory, a universal basic income offers an intriguing alternative to our current dysfunctional welfare state.  But a closer look raises several questions about whether and how a UBI could be implemented in a way that doesn’t create more problems than it solves.

I set aside the question of whether redistribution – for that is what UBI really is – is ever justified.  Matt Zwolinski makes a solid case in favor of such efforts, and certainly a limited amount of redistribution has been supported in the past by prominent libertarians, including Hayek, Nozick, and Friedman among others. On the other hand, as Michael Huemer points out, UBI will, of necessity, violate the Nonaggression Principle at the heart of much of modern libertarianism.  Yet, as interesting as such debates are, there will be some form of government-imposed redistribution for the foreseeable future. The real question therefore is whether UBI offers a better way to fight poverty.

Our current welfare system is clearly a mess. The federal government currently funds 126 separate anti-poverty programs, at least 72 of which provide either cash or in-kind benefits to individuals. For example, there are 33 housing programs, run by four different cabinet departments, including bizarrely the Department of Energy. There are currently 21 different programs providing food or food purchasing assistance. These programs are administered by three different federal departments and one independent agency.  There are eight different health care programs, administered by five separate agencies within the Department of Health and Human Services. And six cabinet departments and five independent agencies oversee 27 cash or general assistance programs. All together, seven different cabinet agencies and six independent agencies administer at least one anti-poverty program. This maze of overlapping bureaucracies is difficult to navigate for those in the system and perhaps even more difficult to supervise and evaluate.

And obviously we should be concerned that the existing welfare system has utterly failed at its primary mission: lifting people out of poverty and enabling them and their children to become independent and self-supporting members of society. Last year alone, the federal government spent nearly $700 billion to fund anti-poverty programs. State and local governments kicked in an additional $300 billion, bringing the total to roughly $1 trillion.  Since the Start of the War on Poverty in 1965, we have spent nearly $19 trillion. Recent studies suggest that welfare programs did help to reduce the worst deprivations of material poverty, especially in their early years. But they have long since reached a point of diminishing returns. We may have reduced the discomfort of poverty, but we have failed to truly lift people out of poverty.

Therefore I am sympathetic to the argument that some form of guaranteed basic income would be an improvement over what we have today. For example, while I am skeptical of some of the predicted administrative savings, there would be clear advantages to a simplified system.  Second, it would treat poor people like adults, expected to save and budget, rather than doling out small allowances for those specific goods and services that the government believes they should have. Third, as Zwolinski notes, it helps break up the entrenched constituencies that support the welfare state.

But despite these potential advantages, I would want to see better answers to several important questions before I could endorse such an approach.

For example, if every American were to receive a flat cash grant that was large enough so as to enable the poor to support themselves in the absence of other welfare programs, the cost would likely be prohibitive.  

Zwolinski does not propose any specific income, but cites Charles Murray’s suggestion of $10,000 per person. Spread over a U.S. citizen population of roughly 296 million, the cost of such a program would be $2.96 trillion, or almost 3 times our current welfare expenditure.  And there is considerable question as to whether $10,000 would be a sufficient grant. Last year, the poverty threshold for a single individual under 65, after all, was $12,119.  

Of course, some suggest using the basic income to replace middle-class social welfare programs such as Social Security and Medicare, as well as those targeted to the poor. The idea of abolishing Social Security and Medicare is far more problematic, both politically and practically, than using UBI to replace more conventional welfare programs. Besides, it still wouldn’t raise enough money to fund a truly universal basic income. Using CBO data for 2013, eliminating welfare state programs including Social Security, Medicare, Medicaid, income security and so forth (but excluding tax expenditures) would yield only $2.13 trillion. If we also included, as some have suggested, so-called tax expenditures, such as the mortgage interest deduction and the exclusion of employer contributions, as well as Social Security, EITC and CTC related tax expenditures, we could add an additional $393 billion for a total of $2.5 trillion. That still wouldn’t be enough.

Others would limit grants to adults only.  This would clearly be more affordable, dropping the cost to roughly $2.25 trillion.  However, limiting participation to adults would leave families with several children well below the poverty level. Consider that the poverty threshold for a family of four was $23,624 in 2013, while grants for the two adults in the family would total $20,000. One possibility would be to adjust the benefit downward for each additional person in a household, recognizing that there are some economies of scale as household size increases. This would reduce costs and the incentive to increase household size (potentially by having more children) while also allowing the initial benefit to be set higher (benefiting smaller households], but would introduce another layer of complexity.

Another issue that would arise in any national level implementation of a UBI is how to address the regional variation in the cost of living. The benefit might be more than sufficient in low cost states like South Dakota, but it might not be enough in high cost states like California and New York. A recent study by the Tax Foundation looked at the purchasing power of $100 in each state, with the relative value ranging from $84.60 in Washington D.C. to $115.74 in Mississippi. Our current system addresses this disparity to some extent, although some of the variation may be due to states increasing benefit generosity for reasons other than cost of living differences. In The Work versus Welfare Trade-off 2013, I found that the benefits package from the same seven programs ranged from $25,491 in Arkansas to $49,175 in Hawaii. The impact of the UBI would vary by location, and low-income people in high cost areas could be worse off. It is not hard to imagine a scenario where people advocate for some kind of benefit adjustment based on the cost of living in the area. While this could potentially be a better design, it would again add a layer of complexity to what initially seemed like a very simple program.

Moreover, whatever the initial size of the UBI, there will be enormous political pressure to increase it.  A UBI would establish as both a legal and a philosophical concept that every American citizen is entitled to a minimum income - exacted from the taxpayers. Once that “right” is established, the political process will inevitably expand it. Murray argues that $10,000 is the correct amount. But how long before some politician comes along and says, “No one can live on $10,000. We need to make it $11,000.” Soon another politician, not wanting to be thought less compassionate than the first, will propose $12,000. There would also be pressure to “carve-out” additional payments to certain groups, like families with a person with a disability or some kind of long-term illness, adjusting for age (since the elderly have higher health care expenses), and so on. We could then get to the point that we have moved far away from the unconditional universal cash grant, to some extent sacrificing the very simplicity that makes up part of the UBI’s appeal.

The more affordable alternative would be to make the grant available only to those with incomes below some predefined level. While under this approach the grant itself would not be universal, it would nevertheless establish a universal floor of income below which no American would be allowed to fall. Friedman’s negative income tax takes this approach. Similarly, Murray would begin to tax back a portion of benefits for families with earned income above $25,000.

The design of any negative income tax becomes crucial. If the benefit is phased out fairly quickly for those with incomes above the poverty level, the program may well be affordable, even cheaper than the current welfare system. But too rapid a phase-out would create a “poverty cliff,” where the marginal tax rate for earning additional income would significantly discourage work or other efforts to escape poverty. A more gradual phase-out minimizes this problem, but adds considerably to the expense of the program.

Zwolinski believes that the studies of NIT and work are ambiguous, at least as far as the degree to which the NIT discourages work. I also think it is fair to ask whether an NIT would discourage work more or less than the current system. (My own research has shown that the current welfare system, where benefits exceed minimum wage in 35 states, has serious potential to discourage work. In the NIT experiments the control group was eligible for traditional welfare, so the work disincentives were relative to the welfare system in place at the time. That being said, there have been significant changes to the welfare system since the NIT experiments, introduction of tax credits etc., so to some degree they cannot tell us that much about how the NIT would compare to the welfare system we have today.)

While the NIT experiments have many important insights, they are to some extent limited in how much they can inform this debate. By their nature, the experiments were temporary, which means that they cannot tell us anything about the permanent work-disincentive effects of an NIT. Would a generation that grew up with an NIT be affected differently?

Zwolinski asks “so what?” if the NIT does reduce work effort. But I can think of at least three reasons why we should be concerned. First, if we are to accept some level of redistribution, it seems fair to require that those receiving the benefits of such efforts take steps that would enable them to become self-supporting as soon as possible. Second, if we actually want to help the poor escape poverty, rather than simply make poverty less uncomfortable (which would be the compassionate thing to do), we know that work is one of the keys to achieving that goal. And, third, a reduction in labor force participation lowers GDP growth, making all of us a little bit poorer. For example, CBO estimates that Obamacare subsidies will encourage people to leave the labor force, reducing aggregate labor compensation by one percent.

A better approach might then be to tie any benefit more explicitly to work. For example, Rep. Paul Ryan, Sen. Marco Rubio, and President Obama have all suggested changes to the Earned Income Tax Credit (EITC). This would in essence establish a minimum income for anyone who was willing to work. It would also be both more effective and better targeted than an increase in the minimum wage.

But the benefit would not be truly universal. It therefore raises the question of what to do about those who are disabled or who cannot work for other reasons. For that matter, what about those who cannot find jobs?  Would there be a parallel welfare system to care for those people?  If so, it would appear that all we have done is add a new benefit on top of the current system. Moreover, the EITC is already one of the most fraud-ridden of all federal programs.

The virtues of the universal basic income lie in its universality and simplicity.  But the closer the program hews to those goals the more likely it is to increase the cost of the welfare state.

Until these questions can be answered, it might be worth experimenting with something similar to what the British government recently undertook with some of its major welfare programs. Britain consolidating its six major welfare programs (the jobseeker’s allowance, the income-support allowance, the employment-support allowance, the child tax credit, the working tax credit, and housing benefits) into a single cash grant, payable monthly to recipients.

The United States could follow suit by consolidating our own disparate welfare programs and, instead, pay recipients a direct cash benefit instead. Such a baby step would allow us to realize some, though not all, the upside of a UBI, while giving us time to further investigate the potential problems.

Opponents of the welfare state have long criticized its supporters for believing that good intentions justified even failed programs. In considering some form of a universal basic income, we should avoid falling into the same trap.

This series on the libertarian case (or lack of one) for a basic income has been very useful. I have commented at length on several of the individual contributions. Let me try to sum up my reactions without repeating myself too much.

(1) As Zwolinkski and Huemer seem to agree, the libertarian case for a basic income is largely a pragmatic one. It falls into the same category as marijuana legalization. If we ask, “Does the state have the moral authority to decide whether I can use/grow/sell marijuana?” then the answer is, obviously, “no.” But when Washington, the state where I live, held a referendum on marijuana legalization, that is not the question I asked. I asked, “Will this proposal, flawed though it is, make things better or worse than they are now?” The answer was yes, so I cast my vote in favor.

(2) We need to be careful with terminology. Zwolinski uses the term Basic Income Guarantee (BIG) to refer to any policy that guarantees a minimum income level to everyone, regardless of family structure or labor force participation. The term is a useful but a broad one. It is important to recognize that it covers a wide variety of different policies, including a negative income tax (NIT) and its many variants, as well as the policy I call a universal basic income (UBI). The key difference between the NIT and UBI versions of BIG is that an NIT reduces the basic grant by a set percentage as the beneficiary’s income rises, whereas a UBI applies no benefit reductions. Under a UBI, everyone gets the full grant regardless of their income or wealth.

More formally, let B be the benefits a given person receives, G be the amount of the minimum income guarantee, Y be a person’s income from other sources, and t be the benefit reduction rate (or effective marginal tax rate) that applies as income rises. For all forms of BIG, B=G-tY. For an NIT, t is a positive number greater than 0 and less than 1. For a UBI, t=0.

Another way to put it is that a negative income tax is means tested, as are most existing U.S. welfare policies, such as TANF, SNAP, and Medicaid. A UBI, as I use the term, is not means tested.

(3) Jim Manzi’s response essay shows why it is important to distinguish between NIT and UBI variants of a BIG. Manzi, like many basic income critics, is concerned about work incentives. He argues that “it is fairly extraordinary to claim that the government could guarantee every adult in America an income even if they did zero work of any kind, and that somehow this would not reduce work effort.” In support of this contention, he cites the results of a set of income maintenance experiments conducted in the 1970s and 1980s. Those experiments, he says, “consistently found that the tested programs reduce the number of hours worked versus the existing welfare system.”

Work incentives are an important issue. Prompted by Manzi’s concerns, I have recently posted a long two-part response on my own blog covering theory and evidence related to BIGs and work incentives.

My conclusion is that Manzi is right, on grounds of both theory and evidence, about the work incentive effects of an NIT. However, he is wrong, on both theory and evidence, to say that replacing our current welfare system with a UBI would reduce work incentives. On the contrary, I think there is every reason to believe that replacing TANF, SNAP, and other programs with a UBI would significantly increase both average hours worked and average labor force participation, with the increases concentrated among households that are just below or just above the poverty line.

(4) Robert Frank raises three important issues in his response essay on basic income and public work.

First, he correctly notes that one of main sources of resistance to a UBI/BIG is the fear that people would take the money and use it to finance lives devoted to folk music and nude volleyball. Part of my response, as argued in the two-part post linked earlier, is even though some individuals would no doubt increase their consumption of leisure, there are strong theoretical and empirical reasons to believe that those would be exceptions, and that the average response to a UBI, especially among low-income households, would be toward more work, not less.

I would add that the worry that people might choose to spend their lives in idleness if they could afford to do so is more conservative than libertarian. Libertarians do not subscribe to the “gospel of work,” as Zwolinski calls it. They subscribe to the gospel of freedom of choice. Freeing people from the micromanagement of bureaucrats who are sure they know what everyone should choose is a feature, not a bug, of a basic income policy.

The second issue Frank raises is that of how large the “G” parameter of a BIG/UBI should be. He points out that “a basic income grant sufficient to lift urban families from poverty probably could not win political support initially, but even if it somehow did, fierce opposition to it would surely erupt quickly.” My own view is that to be effective, a UBI need not be sufficient by itself to lift everyone out of poverty. Instead, I view a UBI as a minimal but dependable platform on which people can build a better life for themselves, if they choose to do so.

Third, Frank advocates financing a BIG/UBI “simply by taxing activities that do more harm than good.” I agree that if we are going to tax, it is better to tax “bads” than “goods,” but I don’t think we have to resort to new taxes of any kind in this case. Instead, it makes more sense to finance a basic income by cutting things from the federal budget as it now stands. First, we should eliminate most existing means-tested income support programs—TANF, SNAP, etc. Second, we should scrap “middle-class welfare” in the form of tax loopholes like the mortgage interest deduction and other tax expenditures. Third, we should integrate a UBI/BIG with existing policies toward disability, unemployment insurance, and retirement in a way that eliminates the possibility of double dipping. For example, retirees and disabled persons should be able to choose between Social Security benefits or the UBI, but not both.