By Douglas J. Besharov

This article originally appeared in The Washington Post, January 5, 1992.

Across the country, governors and state legislators are pushing to use welfare benefits to encourage recipients to stay in school or to get a job, to send their children to school, to take their children for preventive health care, to have fewer children and even to get or stay married. That the proponents of this "new paternalism," as it is called, are both Democrats and Republicans, liberals and conservatives, attests to its substantive and political appeal. But there's a real danger that the states may go too far, too fast.

As The Post's Dec. 18 editorial suggested, it makes sense to condition welfare payments on appropriate behavior. Growing anxiety about the dependent and self-destructive behavior of the poor -- be it dropping out of school, teen pregnancy, nonwork, or drug addiction -- gives urgency to more decisive attempts to reshape the behavior of welfare recipients. This was the critical underpinning of the 1988 welfare reform act, which sought to get welfare recipients to work by tying benefits to employment or study requirements while providing transitional health benefits and child care to those returning to work.

Unfortunately, many of the proposals now being made could end up hurting the poor, not helping them, and might set back more reasonable efforts to alter dysfunctional behaviors. So before the process goes much further, some guiding principles would be helpful:

First, the new paternalism should not be an excuse for balancing state budgets on the backs of the poor. Welfare has never been a very popular program, and media images of festering inner-city conditions, by reinforcing racial stereotypes, only further undermine public support. But a more immediate impetus comes from rapidly growing welfare rolls combined with increasing pressures to cut state deficits. Most new proposals highlight how much money they are expected to save.

Second, the behavioral change should be within the reach of the recipient. It is one thing to reduce the welfare benefits of teenage parents who refuse to attend school, as Wisconsin and Ohio do. But is it fair to penalize adult parents for their teenagers' refusal to attend school, as has also been proposed? Any parent who has tried to get a teenager to clean up a bedroom knows how difficult it is to get adolescents to do anything they don't want to do.

Third, behavioral expectations should be well rooted in public support. Proposals, like Maryland's, to have parents obtain preventive health care for their children or suffer a 30 percent reduction in welfare seems unambiguously beneficial to the child, and, if reasonably implemented, would likely enjoy wide support.

Fourth, policy makers should beware of unintended consequences. The history of social engineering is strewn with examples of perverse consequences for even the most apparently benign programs. The negative income experiments of the 1970s, for example, resulted in significantly higher rates of nonwork. What if denying increased welfare to young mothers who have additional children, as proposed in California and Wisconsin, leads to more abortions? Many of the people most eager to discourage welfare mothers from having more children are also the ones most likely to be horrified by higher abortion rates.

Fifth, determining compliance should be easy and fair. Subjective, case-by-case determinations would be a nightmare to administer and likely result in recurring news stories about bad decision making. Thus, Maryland officials have abandoned their effort to condition welfare on the payment of rent, deciding that they could not adequately police payments.

Sixth, rewarding positive behavior can be more useful than imposing penalties. Benefits send the same signal as penalties with fewer drawbacks. Tangible rewards for doing the right thing can uplift and encourage; penalties threaten to discourage recipients who may already feel psychologically beaten down. As the aphorism teaches, you can catch more bears with honey than with vinegar.

Seventh, benefits (or penalties) should encourage the internalization of long-term changes in behavior. Large penalties raise the stakes so much that bureaucrats and the public recoil from imposing them. More important, just as behavior is continuing, so should the benefit or penalty be additive. Thus, all the major proposals aimed at improving school attendance raise or lower welfare payments on a monthly basis in response to the recipients attendance record.

Finally, humility and caution should infuse the new paternalism. The problems faced by the poor make action necessary, but too many questions remain unanswered to rush headlong into radically new programs. Tentative as it may seem, states should adopt a step-by-step approach, securing sound success and avoiding over-promising and overreaching. After all, we are tinkering with the lives of the most deprived and the least powerful among us.

The writer is a resident scholar at the American Enterprise Institute. This article was prepared with the assistance of Amy Fowler.