I agree with Philippe Van Parijs that a Universal Basic Income (UBI) is a morally attractive arrangement, and think he provides a normatively compelling argument for it in terms of real freedom and social justice. But I also believe that it is a mark of a good theory to be able to offer a theory about itself. In this second-order component, the theorist must answer, among others, the question: Why do so many people oppose my theory? Why isn’t it universally shared, given its overwhelmingly evident plausibility? Shared, that is, by a sufficient number of people, both elites and non-elites, to implement its prescriptions.

One answer to this question might be provided by an ad hoc list of propositions: People need to get used to the idea; they have to overcome their moral prejudices and intuitions; they are misled by interested parties into believing that the social costs of a UBI will be unbearable and that the benefits are dubious. These observations are obviously well taken. They suggest some strategies for improving the chances for a UBI to be successful: try to convince people, talk to political elites, demonstrate that the idea has fallen on fertile ground already in some countries, do more realistic econometric analysis on all kinds of second-order consequences, design and conduct large scale experiments, and the like. All of this is actually being done, and with considerable success, most prominently by Van Parijs himself and other people involved in the Basic Income European Network (BIEN), various national research institutes, advocacy groups, and some left-libertarian political parties.

But while interest in and openness towards UBI schemes are generally on the rise, and not only so in the OECD world, nobody would seriously claim that the reality of Basic Income (in the demanding version specified by Van Parijs) is just around the corner anywhere. Why not? I want to suggest an answer and derive a few (second-order) policy implications for proponents of the UBI idea.

As Van Parijs has argued, the ultimate justification for UBI is freedom: the freedom of individuals to say "no" to employers and state agencies (to say nothing about spouses) without being punished through material deprivation. As a general rule, the anticipation of freedom causes fear. As is the case with other instances of achieving freedom, this fear, although it can be passionate and exaggerated, need not be outright paranoiac. It can be based upon reasons. So, who has which reasons to fear what from the freedom that would follow on UBI? In numerous debates and confrontations I had on the desirability and feasibility of Basic Income, I have encountered various kinds of fear.

1. Employers fear that their control over workers will be weakened, as workers would have a livable withdrawal option. A UBI makes it more difficult for employers to recruit workers for "bad" jobs, and requires employers to increase wages if they still want to fill these jobs.

2. Employees fear that a UBI will require a rate of (direct or indirect) taxation that in turn will involve a downward compression of the scale of net income; similarly, the UBI, they fear, will serve as a pretext to replace the wage-graduated "social wage" that employees receive as pensioners, or in the case of unemployment, with a flat-rate transfer. Wage differences will thus no longer, or not to the extent they are used to, translate into differences in income transfers, and the relative loss of income will have to be made up for through savings.

3. Prospective UBI recipients fear that the level of their income, including the rate of increase of their income, will be contingent upon political decisions and fiscal constraints, and thus be determined in the future by majorities who may or may not endorse and remain faithful to the idea of economic citizenship rights.

4. A great variety of individual and corporate actors fear that the moral underpinnings of a social order that is no longer shaped by the "productivist" assumptions that (employed or self-employed, at any rate market-rewarded) work is "normal," free lunches "anomalous," and the demand of "something for nothing" deviant.

It seems to me that proponents of UBI must take these fears seriously. To suggest otherwise would be to ignore the deep traces that more than one hundred years of the hegemony of industrial capitalism have imprinted upon ideas, intuitions, and expectations. In fact, these hegemonic forces have forged an inter-class alliance founded on a work-centered normative belief system that appears to be largely immune to revision, even under the impact of the manifest changes of social and economic realities. Numerous and prominent policy intellectuals advocating "welfare-to-work" schemes believe–or at any rate espouse the belief and encourage people to adopt it–that the onlydevice by which modern societies can both integrate individuals and at the same time grant them a measure of autonomy is the labor contract. Although we can no longer ensure every adult a permanent job that pays a decent wage, this empirically obsolete vision of "normality" is more firmly entrenched than ever at the normative level. Proponents of a UBI have been rightly disgusted by this perversity, but they have yet to find a way of coping with it in politically productive ways.

So what might be done? I suggest that efforts to implement a UBI should be governed by principles of gradualism and reversibility. The idea is to provide a context in which people can change their preferences through learning, as in the saying that the appetite comes with the eating (rather than with coercive feeding). Instead of thinking about the UBI in terms of "before" and "after," we need to conceptualize and promote it in the dynamic terms of less and more. This intellectual and political mode of experimental approximation could move along the following pathways.

As is well known, in an eventual steady state of a fully implemented UBI even the "surfer" or the "bohemian" would be entitled to a subsistence level citizen (or even resident) income–a scandalous anomaly by today’s prevailing standards that proponents of UBI are usually quick to mitigate by speculating that nobody is likely to adopt the idle life of a surfer for any length of time. But another strategy of response is possible. Note that in most OECD countries and their social policy systems, numerous types of people in various situations and activities are effectively entitled to tax- financed income transfers, at a subsistence level or even higher. Single parents of infants belong in this category, as do people performing mandatory military service. The same applies to families, university students, and–within the European Union and as long as the Common Agricultural Policy lasts–farmers. Institutionalized exemptions from market labor (or the exclusive reliance of market rewards for labor) are numerous and perfectly legitimate.

One gradualist strategy, then, would be to expand the list of groups, conditions, and activities that are legally eligible for such exemption. Political initiatives to promote such expansion are all the more promising as the "third" or "voluntary" or "self-help" sector of private foundations, cooperatives, and neighborhood organizations begins to play an increasingly visible role both as a social phenomenon and as a policy device to unburden, as well as increase the effectiveness of, state-provided services. To be sure, such "participation income" (as Tony Atkinson has influentially termed it) is still not "unconditional," but rather contingent of non-market services performed. But the more popular, normal, and widespread the sector of such voluntary activities is going to become, the more effectively can the authoritarianism of external bureaucratic control be fought.

As is equally well known, a fully implemented UBI would eventually reach subsistence level (and preferably also both legislative irreversibility and continuous adjustment to current market income); it would also be free of any means testing; and it would be effectively paid to all citizens/residents. But these three features constitute as many axes of gradual approximation. More specifically, one could think of starting with an income supplement that does not cover the subsistence level but would still open up a withdrawal option in terms of hours of work. One could make means-testing less stringent and also invert the means-test from one measuring lack of means to one measuring the presence of (significant) assets, with the implication being that all citizens except those with assets above a specified level receive a basic income.

In his presentation here, Van Parijs does not mention that the UBI is not just universal, unconditional, and subsistence-covering, but also permanent. It is individually paid from adolescence to the end of life. The temporal extension is a further dimension of gradual approximation, and indeed an especially promising one. Elsewhere I have argued for a "sabbatical account" (of, say, ten years) to which every adult person is entitled and upon which she can draw at any time (after the age of, say, 25 years) in the form of chunks of time of at least six months, and use the free time, which is covered by a flat-rate income, for whatever purpose she chooses. This scheme can be understood as a temporary basic income. The freedom of choice as to when, as well as how much of it, to withdraw will help to reduce the labor supply at any given point in time. It will also allow employees to (threaten to) withdraw from particularly undesirable jobs and working conditions, and it will provide opportunities and incentives to restore skills and other aspects of human capital. Instead of "banning" people from the labor market, they are provided with the economically tolerable option of opting out voluntarily and temporarily, thus contributing to the restoration of "full" employment, if at a lower absolute level. Those making use of the option would also indirectly contribute to what I consider to be one of the most attractive feature of UBI (and most of its half-way approximation): the powerful indirect effect it would have upon what we used to call "work humanization" and the gradual elimination of particularly "bad" jobs.

In conclusion, and having in mind the context of the European Union and its integration, let me highlight one dimension in which gradualism is not feasible. A UBI (or whichever of its incomplete approximations) cannot be introduced in one country alone. For such unilateralism is likely to trigger migration effects that are bound to undermine the political and economic viability of any even less-than-complete solution. (Such migration, or emigration-prevention effects are, of course, intended in the very special case of Alaska, as they were intended, before 1989, in the comparable case of West Berlin, with its residence premium paid to citizens as tax credit.) In Europe, however, what is possible in one country is constrained by what is possible in all other countries as well and at the same time. This rule may well be interpreted "Euro-skeptically," as proof that the European Union stands in the way of national policy innovation. But it may also be read, more optimistically, as a design for the implementation of a "social" Europe that might be capable of providing some much-needed meaning and broad popular appeal to the project of European integration.