Nearly two years ago, the nation's governors began working on a welfare reform plan that would change the system from one that too often traps people in dependency. Benefits are 33 percent lower than they were in 1970, but they can be obtained and kept solely by remaining poor. The governors want a system that promotes independence through education, training and work, that takes care of poor children and that recovers as much of its costs as possible through strengthened child support enforcement.

We are now at a critical point in the welfare debate that has occupied the governors, human service commissioners, advocates, local governments, the administration and Congress. The House of Representatives is expected to vote on two bills that represent two very different approaches to the reform of a welfare system everyone agrees needs to be changed.

I believe that H.R. 1720, the Family Welfare Reform Act of 1987, goes a long way toward achieving the objectives outlined in the governors' policy. Reducing welfare dependency by creating significant work, education and training opportunities for Aid for Families with Dependent Children recipients is the cornerstone of H.R. 1720. Resources are targeted on families most in need of assistance -- those likely to become long-term dependents of the system unless something is done.

To ease the transition to work, the bill creates modest financial incentives so that families who work are better off than those who do not. Day-care subsidies and Medicaid would be available temporarily so that parents who go to work in jobs without benefits won't be asked to put their children' s health or safety in jeopardy.

Further, the legislation is designed to improve efforts to establish paternity of children, ensure that child-support orders are fairly set and regularly updated and give states incentives to use effective enforcement techniques.

The governors recognize the legitimate concerns that have been raised over the cost of H.R. 1720 -- $1.7 billion over the next three years. That seems like a lot of money, unless you compare it with what we have been spending for similar purposes. Over the past three years, before 1987 when appropriations were cut in anticipation of a new program, the Work Incentive Program was funded at more than $600 million. Over three years, then, the cost of H.R. 1720 would exceed that funding by about only $1 billion. This additional money buys the additional training and educational experiences this population needs, as well as the key support services that we know make the programs successful. Like all investment strategies, H.R. 1720 requires some up-front funds. However, the governors believe that there will be quick and substantial savings to both the states and the federal government as people who would otherwise stay on welfare become trained and leave the welfare rolls.

H.R. 3200, the alternative bill, seems at first blush to accomplish the same goals for less money and to give the states more flexibility in designing their own approaches to welfare reform. However, the governors believe it falls far short of the policies we need. H.R. 3200 mandates participation levels that, coupled with limited funding for the welfare-to-work program, would have the practical effect of forcing states to run large numbers of welfare recipients through inexpensive job search programs of limited effectiveness, rather than through the comprehensive education, training and employment programs that are needed to move the hard-core welfare recipient to self-sufficiency. Moreover, according to the Congressional Budget Office, H.R. 3200 will actually cost the states more money than H.R. 1720.

H.R. 1720 is pro-work. It promotes employment opportunities for poor families with children, especially those who remain on welfare for long periods. This group is responsible for the majority of welfare costs. More than 65 percent of the new three-year expenditures for H.R. 1720 are specifically allocated for work, education and training programs and related expenditures. Programs will be targeted to recipients who are most likely to become dependent on welfare: teen-age parents, families who have been on AFDC continuously for at least two years and recipients who do not have high school diplomas. Women with children over the age of 3 (as contrasted with the age of 6 in current law) would be required to participate in education, training and employment programs. Mothers with even younger children would be actively encouraged to enroll and could be required to do so if appropriate infant care were provided. Child care is guaranteed for parents who are enrolled in the work program.

The White House and the supporters of H.R. 3200 object to H.R. 1720 because 1) it contains incentives for states with low AFDC payments to raise them; 2) it requires AFDC coverage to be extended to two-parent households (today about half the states provide welfare only when one parent has left the home); and 3) beause H.R. 1720 provides no "waiver" system by which states can reallocate federal funds targeted for specific purposes to new state-designed welfare initiatives.

The governors do not oppose the incentive plan because benefits are often lowest in the poorest states with the most limited resources. We support coverage of two-parent families by AFDC as long as one parent is required to participate in the education and work programs, no matter how young the child is. We support increased waiver authority for the states, but do not believe H.R. 1720 should be opposed for that reason alone. If the House will put aside partisan considerations and pass H.R. 1720, the Senate then will have the responsibility to move on Sen. Daniel Moynihan's bill, which is less costly than H.R. 1720 and which does have more waiver authority for the states. Then, in conference, the concerns of cost and waivers can be addressed. However, if the House does not act, the Senate will have no reason to move welfare reform up on its crowded agenda, and a historic opportunity to improve a welfare system that nobody likes will have been lost.

The writer is governor of Arkansas and past chairman of the National Governors' Association.