Apparently Charles Murray's new book talks up a fairly old idea: a universal guaranteed income. Supposedly, providing everyone with a government salary of a few thousand dollars a year would solve the problem of destitution (i.e., provide a "safety net"), while not giving people enough comfort to discourage them from improving their standard of living through hard work.

The underlying assumption--which seems to be almost universally taken for granted--is that guaranteed income programs only risk creating a dependent underclass when the guaranteed income is too generous. The American experience with AFDC ought to have put paid to that assumption. Contrary to popular belief, the American welfare system was never all that extravagant, and the American urban underclass during (especially) the 1980s lived very meagerly indeed.

It's true that AFDC also added an extra disincentive to work, in the form of reduced benefits for the legally employed. However, illicit, unreported employment was widely available during the welfare era, and most welfare recipients supplemented their incomes with it on a casual basis. The result was not a thriving underground economy of industrious low-wage workers--as we see today among illegal immigrants, for instance. Rather, welfare cultivated a pathology-rich urban underclass ridden with crime, violence, family breakdown, substance abuse and (of course) grinding, seemingly inescapable poverty. And the same type of social safety net, writes British doctor and social critic Theodore Dalrymple, has now generated an underclass with comparable pathologies in Great Britain.

How, though, could a miserly welfare program, with paltry benefits, have such a profound negative psychological effect on its recipients? I would argue that in fact, the key pernicious attribute of guaranteed income programs is not their generosity, but rather their stability. A guaranteed government income is even more stable than most jobs (which, after all, can disappear overnight). And for a not-insignificant fraction of the population, the stress of trying to hold down a job with a respectable salary is less attractive than the comfort of guaranteed work offering only a subsistence income--let alone a subsistence income that requires no work.

If I were to design a welfare program, I'd take it in exactly the opposite direction from Murray. I'd make it at least as generous as all current programs combined, and focus it exclusively on the needy. However, I would treat it as a government-operated charity program, rather than as an entitlement program. An all-encompassing "charity budget" would be set each year through legislation, and its funds would be doled out in the form of matching grants to registered charitable organizations helping the needy. The usual oversight would be necessary, of course, to ensure that the charitable funds are disbursed honestly. But no charity--and hence no recipient--could ever be sure of receiving as much money in any given month or year as in the previous month or year. A local natural disaster, for instance, might divert much charitable giving towards its victims, thus reducing the amount available for the chronically poor.

Such a system, I claim, would make welfare dependency much less attractive than an entitlement system would--even if it provides the same overall amount of money to the needy. Unlike the recipient of a guaranteed income, a "charity case" under this system would have to endure not only a lower income, but also more instability, than a gainfully employed citizen. This instability would be a powerful incentive to work even for those who are comfortably able to tolerate a financially constrained lifestyle.

As I've pointed out before, the whole concept of entitlement is based on a flawed conception of human need. We cannot hope to end human want--as entitlement programs implicitly aim to do--because human want is infinite and eternal. On the other hand, we are (I believe) morally compelled to contribute to the well-being of our fellow humans, through charitable giving. Government entitlement programs, by straying from the donor-centered ethos of charitable generosity, have managed to end up hurting both the donors and their beneficiaries. The solution, then, is to abandon the false hope of "entitlement", and return once again to the ideal of charity.