Two vetoes, one signature and the 1996 welfare law that put marriage in statute

In the fall of 1994, a political party that had spent forty years in the House minority did something that Washington regards as faintly indecent. It wrote down what it would do if elected, put the list on the Capitol steps, and asked the voters to hold it to the list. Then the voters did. Then, to the astonishment of the commentariat, the party tried to do what the list said.

One item on that list was welfare. Not "study welfare," or "convene a commission on welfare," the traditional Washington methods of making a problem last until retirement. The item was to end it as an open-ended federal entitlement and replace it with something built around work, family and the states. It took two presidential vetoes and nearly two years to get there. It ended with a signature on August 22, 1996, and a law whose first finding was about marriage. This is the story in order, with the people involved and the arguments each side still makes.

The system before 1996

To understand the fight, you have to know what was being fought over. Aid to Families with Dependent Children, AFDC, began as Aid to Dependent Children in the Social Security Act of 1935. The original idea was to help widows with young children keep their families together, so a mother would not have to put her kids in an orphanage. It was a modest program for a small population.

Over the decades it grew into something else. It became an entitlement: any family that met the eligibility rules was entitled to cash, and the federal government matched what the states spent without a fixed cap. The rules varied from state to state, and the structure had incentives that even its defenders worried about. In many places, a family's benefit fell or vanished when the parents married, or when a mother took a job and her earnings rose. A program meant to support the family was, in the fine print, penalizing the two things most likely to lift a family out of poverty.

By the early 1990s, both parties agreed it was broken. They did not agree on the repair. When Bill Clinton ran for president in 1992 he promised to "end welfare as we know it," a phrase that gave him room to mean several different things at once. Republicans, in the House especially, meant it more literally.

The Contract with America, 1994

On September 27, 1994, Republican candidates for the House gathered on the steps of the Capitol and signed the Contract with America, a ten-bill list of things they promised to bring to a vote in the first hundred days of a new Congress. Among them was the Personal Responsibility Act, the welfare bill. Another was a $500-per-child tax credit, which we follow in the child tax credit from $400 to $2,000, told bill by bill.

In November 1994 Republicans won control of the House for the first time in forty years, with Newt Gingrich becoming Speaker, and the Senate as well. The contract bills moved fast in the House. The welfare bill's core ideas were simple: end the entitlement, give the states a fixed block of money and wide latitude, require work, set time limits, and discourage out-of-wedlock births.

The Senate was more cautious, and the president was more cautious still. This is where the story gets interesting, because the next two years were less about a right-versus-left fight than about three different centers of power negotiating over what could survive.

The first veto, December 1995

Congress folded welfare overhaul into the large budget bill of 1995, the Balanced Budget Act, which also carried the child credit and a broad package of tax and spending changes. President Clinton vetoed it on December 6, 1995. That was the first veto. His objection was to the overall package, including its cuts to other programs, not only to the welfare provisions.

The months that followed were the long government standoff of the winter of 1995 to 1996, with shutdowns that nobody remembers fondly. Whatever its merits, the standoff shows how hard it is to reform an entitlement when the other side holds the pen that signs.

The second veto, January 1996

Republicans then split welfare off as its own bill, H.R. 4, the Personal Responsibility and Work Opportunity Act of 1995. This time the target was narrower. The president vetoed it on January 9, 1996, objecting, in broad terms, that it did not do enough for children and for the child care and health coverage that working parents would need.

Two vetoes. If you were a Republican member in January 1996, it would have been easy to conclude that the president did not want a bill and that the smart play was to run against him on it in November. There was a real argument in your caucus for doing exactly that. The pressure on the president was equally real: he had campaigned on the promise, and the 1996 election was coming. Neither side could be seen as the one who blocked reform.

The deal in the summer of 1996

By the summer, the parties found a way through. Republican leaders in Congress separated the welfare bill from Medicaid, which had been a major sticking point, and sent the president a bill that kept the structure of the earlier version but removed the Medicaid overhaul that he had opposed. The bill that became the final law was H.R. 3734, and it passed both chambers with votes from members of both parties. I will not give tallies, because the point is the structure, not the scoreboard: this was a bipartisan law, and its supporters and opponents both came from both parties.

President Clinton signed the Personal Responsibility and Work Opportunity Reconciliation Act, known by its initials PRWORA, on August 22, 1996. Several senior officials at the Department of Health and Human Services resigned in protest over the decision, and Senator Daniel Patrick Moynihan, who knew the subject better than almost anyone in the Senate, opposed it. The opposition was principled, and it came from inside the president's own party. The history is not a story of heroes and villains. It is a story of people with strong convictions disagreeing about a hard problem.

What the law actually did

Here is the substance in plain terms.

1. It ended AFDC and created Temporary Assistance for Needy Families. TANF is a block grant. Instead of matching whatever each state spent, Washington gave each state a fixed annual amount, with a nationwide total of about $16.5 billion a year, set in the statute. States got broad freedom over how to spend it within the law's four stated purposes.
2. It added work requirements. States must have a specified share of their caseload engaged in work or work-related activities, or face penalties. The details and the share of families counted have been argued about ever since.
3. It set a five-year limit. A family may not receive federal TANF assistance for more than sixty months over a lifetime, with a limited exemption that states can use for hardship cases.
4. It required states to keep spending. A maintenance-of-effort rule requires states to keep spending a share of what they had spent before, so the block grant could not simply replace state money.
5. It stated its purposes. The four purposes in the statute are to provide assistance so children can be cared for in their own homes; to end dependence by promoting job preparation, work and marriage; to prevent and reduce out-of-wedlock pregnancies; and to encourage the formation and maintenance of two-parent families.
6. It opened with findings about marriage. The law's first finding, in Section 101, states that marriage is the foundation of a successful society. A federal statute that says so in its text is unusual, and it was a deliberate choice.
7. It strengthened child-support enforcement. The law expanded the tools states use to establish paternity and collect support.

Notice the philosophy. The 1935 program tried to substitute for the family. The 1996 law tried to push the system toward work and toward the family. You can believe that is the right direction, or the wrong one, but there is no mystery about what the authors intended.

The Charitable Choice provision

Buried in the law is a provision that matters to every church reader. Section 104, known as Charitable Choice, says that states using TANF money may contract with religious organizations on the same basis as other nonprofit providers, and that the government may not require such an organization to give up its religious character, such as its religious art, its governing structure or its right to hire on a religious basis, in order to take part. Religious beneficiaries must be offered an alternative provider if they object to receiving services from a religious one.

The point was simple. If a church runs the best job-training program or the best shelter in town, the government should not disqualify it for being a church. The provision was a forerunner of later arrangements, and we trace how it grew in From Charitable Choice to the faith-based office, when churches got a seat. Critics worried about entanglement of church and state, and supporters answered that equal treatment is not an establishment. That argument has not gone away, and it shows up again in disputes over foster care, which we discuss in What the 9-0 Fulton decision means for foster care and faith agencies.

The fair criticism from the left

A fair account has to say what the critics got right, or at least what they argued with force.

First, the block grant was a fixed dollar amount, and its nominal value was not adjusted for inflation. Over time, a fixed number buys less. Critics are right that the real value of the grant has eroded unless Congress acts to change it.

Second, the structure of the block grant gave states wide latitude, and some states spent TANF money on things other than cash for needy families. The statute allowed it; critics say it wandered from the intent.

Third, the time limits and work rules do not bend easily in a recession. When jobs are scarce, a mother who wants to work may not find a job, and the clock keeps running. A law written in a strong economy faces its harshest test in a weak one.

Fourth, the officials who resigned and the senators who opposed it feared that children in the poorest households would be hurt. That worry deserves to be stated plainly, because it is a moral argument and not merely a political one.

The fair criticism from the right

The law also has critics on our side, and we ought to hear them.

First, the work requirement was softer in practice than in theory. The statute allowed states to receive credit toward the requirement for falling caseloads, which meant that a state could reduce its requirement by shrinking its rolls, whether or not those who left were working. Conservatives who wanted a genuine work obligation were disappointed.

Second, the marriage goals did not produce the hoped-for results. Government programs to encourage marriage have a thin record, and honest supporters admit it. The law named the goal, but the means were weak. A culture shift is not easily legislated, which is a lesson conservatives of all people should absorb.

Third, a block grant gives states flexibility, but it also lets them quietly shift the money. Some conservatives would have written tighter rules about how the funds can be used.

The caseload fell sharply in the years after the law, and both sides cite that fact for opposite conclusions. The left says families were pushed off the rolls into hardship; the right says families went to work. The truth probably includes both, and it differs from state to state and year to year. Whoever tells you it was entirely one or the other has not read the research.

What the story teaches

I draw three lessons, and I will be plain about which are facts and which are judgments.

Fact: divided government can produce real laws. A Republican Congress and a Democratic president, each facing an election, produced the most significant welfare change in sixty years. Neither side got everything it wanted. We make that case at length in divided government gave us welfare reform and a balanced budget. Whenever the two parties hold different keys, the country gets gridlock or a deal, and there is no way to know in advance which. The 1996 deal came because each side had something to lose and something to gain.

Judgment: attach aid to work and family, and you shape behavior. The law's authors believed that aid with an expectation of work does more for a family than aid without one. The same principle underlies the debate over the child credit, and we apply it in Should the child credit pay families who don't work? The 2021 argument.

Judgment: a movement must be honest about what it can and cannot do. Congress can write a finding that marriage is the foundation of a successful society. It cannot make people marry. The work after the law is done in churches, neighborhoods and families, and in an economy that rewards work. See why family policy should not punish the parent who stays home for one way the tax code and benefit rules can quietly work against the family.

The bottom line

• The 1996 law ended AFDC and created TANF: a block grant to the states, work requirements, a five-year lifetime limit on federal assistance, and a stated goal of two-parent families.
• It took the Contract with America in 1994, two vetoes, in December 1995 and January 1996, and a signing on August 22, 1996. It was a bipartisan compromise, not a party-line coup.
• The critics from the left and from the right both have real points: a block grant that loses value, soft work rules, and weak results on marriage.
• If you want to judge the policy, read the statute's four purposes on Congress.gov and compare them with what your state does with its TANF money. Your state's own budget is public.
• Next time someone says nothing ever gets done in Washington, remind them of 1996, and then ask what the deal was that got it done.