(David Goldman/AP)

“When some Republican governors asked if they could have waivers to try new ways to put people on welfare back to work, the Obama administration listened because we all know it’s hard for even people with good work histories to get jobs today. So moving folks from welfare to work is a real challenge. And the administration agreed to give waivers to those governors and others only if they had a credible plan to increase employment by 20 percent, and they could keep the waivers only if they did increase employment. Now did I make myself clear? The requirement was for more work, not less.”

— Former President Bill Clinton, at the Democratic National Convention, Sept. 5, 2012

Readers may recall that in August we gave Four Pinocchios to Mitt Romney for a television advertisement accusing President Obama of gutting Bill Clinton’s welfare overhaul — and also Three Pinocchios for the Obama administration’s counterspin that Romney himself had sought a similar waiver when he was governor of Massachusetts.

What’s the fuss about? Temporary Assistance for Needy Families (TANF), the centerpiece of the 1996 legislation, established work requirements and time-limited benefits for recipients. But in July the Department of Health and Human Services issued a memorandum saying that it was encouraging “states to consider new, more effective ways to meet the goals of TANF, particularly helping parents successfully prepare for, find, and retain employment.” As part of that, the HHS secretary would consider issuing waivers to states concerning worker participation targets.

HHS’s action set off a firestorm of criticism by Republicans, which was echoed in Romney’s ad.

In his high-profile speech at the Democratic convention, Clinton himself came to Obama’s defense, claiming that the change in rules actually would require “more work, not less.” Last week, we said we wanted to spend some time digging into this statement before making a ruling. After talking to many people on all sides of the welfare debate, we can certainly say it is a very complex issue — which makes it ripe for fact-checking.

The Facts

There are three basic rules in Washington: 1) Nothing happens by accident, 2) Personnel determines policy, and 3) No argument is ever settled. That dynamic is central to understanding the controversy surrounding the HHS memo. In this case, conservatives suspected that the administration was trying to achieve through regulatory fiat what liberals had not been able to accomplish through legislation in the past 16 years.

By many accounts, a key player in the development of the memo was Mark Greenberg, the deputy assistant secretary for policy at HHS’s Administration for Children and Families, which oversees TANF. Key aspects of the memo — which was signed by Earl S. Johnson, director of the Office of Family Assistance — turn up in testimony before Congress that Greenberg gave over the years when he was outside of government, arguing for changes in the 1996 law.

On the opposite side of the long-running debate is Robert Rector, a Heritage Foundation scholar who helped craft the law. (Here’s an example of the two men testifying side by side before Congress, as the yin and yang of the welfare reform debate.)

As soon as the HHS memo was issued, Rector raised the clarion call that Obama, through stealth, was gutting the law because, he says, he recognized that ideas — what he labels “loopholes” — that Democrats had been unable to slip into legislation had suddenly been offered to states in the form of waivers.

Boiled down, one key difference between the two sides is whether one should focus on job search (conservative) or job training (liberal). There was little job training in the 1996 law, which put the main focus on getting people back to work.

The HHS memo listed as possible projects for waivers such ideas as “multi-year career pathways models for TANF recipients that combine learning and work.”

Another issue is whether provisions should be made for people who are disabled. Indeed, the HHS memo touts “projects that demonstrate strategies for more effectively serving individuals with disabilities, along with an alternative approach to measuring participation and outcomes for individuals with disabilities.”

Critics suggest that opens the door for a whole class of people being removed from needing to find work, making it easy to boost published employment rates for others receiving assistance. In other words, if a pool of 100,000 people is reduced by 10,000, then the percentage of people finding work will increase even if the number who are employed remains the same.

We had earlier said that the HHS memo certainly appeared to be a process foul, a position recently supported by the Government Accountability Office. (The Congressional Research Service offered its own view on the Secretary’s waiver authority, in a report released this week by Democrats.)

The Romney welfare ad took an extreme interpretation of the memo, claiming that Obama has already taken action to “drop work requirements.” The ad further states that “under Obama’s plan, you wouldn’t have to work and you wouldn’t have to train for a job. They just send you your welfare check.”

That’s not apparent in the memo — at least not yet, since no waivers have been granted. But interestingly the claim made by Clinton — that the “administration agreed to give waivers to those governors and others only if they had a credible plan to increase employment by 20 percent” — is not in the memo either.

Instead, that assertion appears in a letter written by HHS Secretary Kathleen Sebelius to Sen. Orrin Hatch (R-Utah) after Republicans erupted at HHS’s issuance of the memo. She wrote:

“Specifically, governors must commit that their proposals will move at least 20 percent more people from welfare to work compared to the state’s past performance. States must also demonstrate clear progress toward that goal no later than one year after their programs take effect. If they fail, their waiver will be rescinded.”

The original HHS memo makes no mention of such requirements. Indeed, it is significantly weaker. The memo does not cite a one year time limit to demonstrate progress, for instance.

Instead, the memo suggests that states have at least several opportunities to prove that their plan works: “States that fail to meet interim targets will be required to develop improvement plans. Repeated failure to meet performance benchmarks may lead to the termination of the waiver demonstration pilot.”

In an interview, administration officials said that this is an “iterative process” and Sebelius’ letter represented the most up-to-date version of administration policy. Still, there seems to be some wiggle room, as there is no definition yet of “clear progress.”

“The states have to show they are making progress within a year,” one senior official said. “They need to identify interim targets year after year” if they are not meeting the 20 percent threshold. The original memo also said that “absent special circumstances, the length of an approved project will not exceed five years.”

How many people are we talking about? The latest data (Table 46) shows that in 2010, of the 1.84 million families on welfare, 16.6 percent — or about 300,000 — left the rolls because of a new job.

Looking at the states with Republican governors that the administration said requested waivers, the numbers needed to meet the 20-percent target do not appear to be large. In Nevada, with 0.2 percent (230 families) out 11,520 families getting a job, the target for 20 percent would be an additional 46 families. In Utah, which moved 9.8 percent (735 families) of its 7,496 families to work, the number would be 147 families.

The letters from Nevada and Utah requesting a waiver — which mostly concerned relief from burdensome federal reporting requirements — were sent by the heads of welfare agencies — not by the governors, as Clinton said. The other states that the administration said sought waivers were California, Minnesota and Connecticut, which in theory would need to boost employment by 20,000, 1,573 and 202, respectively.

Some critics of the administration’s approach believe it marks a return to simple “employment exit” statistics that marred the early days of state welfare reform efforts. (Under the updated version of the 1996 law, states must meet “a participation rate” in which at least 50 percent of recipients, before credits, participate in defined work activities.)

In other words, officials would tout that 15 percent of recipients had left the rolls, without acknowledging that the overall welfare population had grown by 20 percent. One former top welfare official said he could easily meet the administration’s requirements by more assiduously tracking people who found jobs but did not inform the welfare agency.

Administration officials insist that will not be the case. For instance, officials said they plan to adjust the targets according to the growth of the economy, so a state could not simply meet its goal because a booming economy has made more jobs available.

The administration also says it will not rely on state-supplied data to measure progress.

“We would not use self-reported data for measuring progress under the waiver,” the administration official said. “Rather, we would match data on individuals leaving TANF to reliable employment data from the National Directory of New Hires (NDNH) and the unemployment system, which tracks earnings. We would do this both prospectively and retrospectively — the latter so that we get a corrected baseline that is accurate.”

The Pinocchio Test

Because welfare funding for the states is delivered as a fixed block grant, there is relatively little incentive for a state to increase its welfare rolls. So even critics of the administration’s approach do not imagine the changes will result in significantly more people on welfare. We stand by our earlier ruling on Romney’s welfare ad.

Still, there is enough uncertainty about how the administration will implement these waivers that it is a stretch for Clinton to declare for certain that a 20 percent threshold must be met — and to claim that more people will end up working under this new system.

The administration’s emerging criteria, which has not been previously reported, certainly sounds reassuring. (Rector says not using state-reported data and relying on NDNH statistics would be “much more reliable.”) But, as with everything in Washington, much will depend on how the baseline number is determined — and then how the process is monitored. In theory, a state could go five years without ever meeting the 20 percent threshold.

The administration, in fact, only revealed the 20-percent figure once it was under fire. Twenty percent sounds like a lot, but there may be less to that number than meets the eye. Given the long history of the two sides battling over the implementation of the 1996 law, critics certainly have a right to be wary about why the administration acted in the way it did.

We wavered on the number of Pinocchios in this instance. But we finally settled on two, mainly because Clinton, in his facile way, made this intense debate appear as if it is mainly a dispute about moving “folks from welfare to work.” It is not quite so simple as that, and neither is it clear yet that the net result is that more people on welfare will end up working.

Two Pinocchios

(About our rating scale)

Check out our candidate Pinocchio Tracker

Follow The Fact Checker on Twitter and friend us on Facebook .

Track each presidential candidate's campaign ads

Read our biggest Pinocchios