WELFARE USE: Requiring participation in mandatory employment services and time-limiting benefits decreased welfare receipt.

Two of the approaches states used - mandatory employment services and time limits - reduced the amount of cash benefits paid out (usually because welfare receipt declined). Earnings supplement programs, in contrast, typically increased benefit payments more than traditional welfare programs. While the three principal reform approaches that states used produced consistent effects on earnings, their effects on benefit receipt were anything but consistent. Mandates reduced benefit payment amounts, earnings supplements had the opposite result, and time limits first increased benefit payments and then reduced them. Regardless of strategy - job search first, mixed, or education first - mandatory employment services programs that established a participation quid pro quo for receipt of welfare benefits consistently reduced welfare dependency, as evidenced by reductions in the number of months and amounts received of welfare benefits. The first section of Figure 2 shows the average annual impacts (effects) on welfare payments of 12 programs with participation mandates but no earnings supplements or time limits: All 12 produced welfare savings. Driven in large part by these welfare savings, from a government budget perspective, the most effective mandatory programs (for example, the Portland Jobs First program and the Grand Rapids job-search-first program) returned to the government more than $2 in lower welfare costs and higher taxes paid for every $1 in government investment - a stunning achievement for any social program. In other words, it would have cost these states more not to operate these programs. By contrast, mandatory programs that were able to generate only small welfare savings, possibly because they did not enforce the participation mandate, were less likely to break even, returning in government budget savings only about 40 cents to 80 cents per dollar invested. In stark contrast to the impacts of the traditional mandatory welfare-to-work programs, programs that added earnings supplements to mandates increased public transfer payments (welfare in Minnesota's program, earnings supplement payments minus welfare savings in the Canadian and Milwaukee programs), as shown in Figure 2. The explanation is simple: Earnings supplement programs allow recipients who take jobs to continue receiving some form of cash payment. But does this mean that dependency increases? The answer depends on how one defines "dependency." In the SSP program, supplements were paid outside of the welfare system - to receive a supplement, recipients had to leave welfare. Thus, welfare dependency declined dramatically, but because the amount of the earnings supplements paid to these job-takers exceeded the welfare savings, total public transfer payment amounts (welfare plus earnings sup-plements for all program members versus all control group members) rose. Nor is the answer cut-and-dried in the Minnesota program, even though it used the welfare system to pay earnings supplements. In that program, the percentage of people who were still receiving welfare at the three-year follow-up point rose by a small 4 percentage points, while the percentage who were not working at all fell by 8 percentage points, and the percentage who were combining work and welfare rose by 12 percentage points. In short, while the amount of public transfers rose so did the amount of work. Thus, if welfare dependency is defined as someone who is solely dependent on welfare, the higher fraction of welfare recipients deriving more of their income from earnings means that dependency has fallen, even in the Minnesota program. Finally, earnings supplements can be an efficient means of transferring income to low-income families. Because cash benefits paid as a supplement to earnings stimulate more work effort, not less, and thus are not the sole source of income, earnings supplement programs can generate more than a dollar of increased income for every extra dollar of income transferred. For example, at its peak point, the Canadian program increased recipients' income by more than $3 for every $1 in supplement payments, primarily because earnings also rose by $2. Adding time limits to mandates and earnings supplements, as the two final programs shown in Figure 2 did, produced two quite different effects. Before the time limit, the supplement effect dominates and welfare receipt rises or remains unchanged; after the time limit, when recipients can no longer receive benefits, welfare payment amounts decline. While Figure 2 provides three-year average effects on welfare to facilitate overall comparisons with the other program models, the underlying story is best told by Figure 3, which separates out effects for the Connecticut and Florida time-limit programs into the before- and after-time-limit periods (see the four black "welfare payments" bars). The graph clearly shows welfare benefit amounts rising before the time limit in Connecticut and falling thereafter. (In Florida, effects on welfare receipt in the before-time-limit period is insignificant, in part because benefit levels in Florida were so low that, even with a generous earnings disregard, only modest amounts of work led people to leave welfare.) Although the two programs with time limits typically increased employment during the period before anyone reached the time limit, there was only limited evidence that the time limit compelled people to exit welfare more quickly on their own than they would have otherwise in order to save or bank their remaining months of eligibility for benefits. The programs' pre-time-limit incentive provisions may have mitigated any inclination on the part of recipients to leave welfare early. Back to welfare use summary