PART-ing Shots

This February in North Carolina, George W. Bush told a giddy crowd, “I'm here to talk about an issue that is going to be an interesting experience in dealing with the Congress [laughter]. And that is Social Security -- formerly known as the third rail of American politics [laughter]. That meant, if you touched it, there would be certain political death.”

He may not have been electrocuted as the result of the felony he tried to commit against Social Security, but he seems to have learned what the rest of the conservative movement has known for years: Don't attack popular social programs directly. Instead, come at them from the side. While the indirect strategy may lack the glamour of a full-frontal counterrevolution, at least it's not suicidal.

The “starve the beast” approach -- the idea that slashing government revenue through perpetual tax cuts eventually creates a fiscal situation that leaves eliminating popular programs as the only responsible solution -- is perhaps the best known of these indirect attacks on government. But the right has another -- and very little known -- club it uses to starve (and bleed) the beast: the Program Assessment Rating Tool, or PART.

Run out of the Office of Management and Budget (OMB), PART was launched in 2003 -- with no individual budget, executive order, or act of Congress -- with goals that sound unobjectionable on their surface: to measure the efficiency and effectiveness of government programs. Cloaked in good-government rhetoric, the program won an Innovations in American Government Award in 2005, jointly administered by Harvard's Kennedy School of Government and the Council for Excellence in Government, perhaps the most zealous backer of the performance-review model.

But underneath the rhetoric, PART is inherently designed to undercut progressive programs -- and, according to research by two political scientists, it has done just that. And when the blade accidentally wounds a program the administration likes, the White House tends to ignore the unfavorable rating and asks for a funding increase anyway.

Run by Clay Johnson, deputy director of management at the OMB, PART takes up countless hours of program and agency time to comply with the reporting requirements. The PART questionnaire asks a series of “yes-no” questions of varying weight, leading to a raw score between zero and 100 based on four measured areas: purpose and design, strategic planning, management, and results. The score translates to “effective” (85 to 100 score range), “moderately effective” (70 to 84), “adequate” (50 to 69), “ineffective” (under 50), or “results not demonstrated” -- which is extremely subjective and often taken to mean ineffective by the administration.

At times, the ratings given to programs or agencies are so biased as to be laugh-out-loud funny. The Social Security Administration, for example, famously efficient and accurate -- and surely serving a good purpose -- is given perfect 100-out-of-100 scores in both the planning and performance for its supplemental security income program in 2005. The score for purpose: 60. It was rated “moderately effective,” and a $462 million cut was requested.

In his 2005 State of the Union address, Bush bragged of reducing or eliminating “more than 150 government programs that are not getting results, or duplicate current efforts, or do not fulfill essential priorities.” In fact, of the 154 programs to which Bush was referring, fully two-thirds were not even rated, according to fact-checking done by OMB Watch, a budget-policy watchdog group. Yet Bush attempted to use good-government rhetoric to eliminate them anyway. Of the programs that were reviewed, 20 percent received one of the two highest ratings -- and were still slated for elimination. And if a program was housed in either the Department of Education or the Department of Housing and Urban Development -- both of which the right wants to see disappear -- and was rated “ineffective,” its chances of survival were terrible: Seventy-eight percent of them were targeted for termination.

A better way for a program to stay alive, though, was to have a purpose the administration supported. Several conservative programs received ratings of “ineffective” but were slated for either level funding or an increase. For example, the Earned Income Tax Credit Compliance Program -- which double-checks the eligibility of low-income workers who have claimed the progressive credit -- was rated ineffective (despite its generous 100 score in the purpose category) but saw its funding request increase $4 million from fiscal year 2005 to 2006.

John Gilmour of William & Mary's Department of Government and David Lewis of Princeton's Woodrow Wilson School have documented this bias. Having done the most in-depth academic review of PART, they concluded that in 2004, “PART scores influence[d] budget allocations for programs housed in Democratic departments but not for others.” Gilmour and Lewis say they defined Democratic programs as those created as part of the Great Society push, when Democrats controlled Congress, or when it was otherwise obvious that the program was favored by Democrats and not Republicans. The pair revisited the budget tool the next year and found that PART's influence had become more consistent across the board. According to Gilmour, though, they never tested the notion that PART's very design leads to lower ratings for progressive programs.

Other experts contend that any evaluation of PART must take into account the conservative assumptions that bias its rating system. “I think one way of looking at it is, what are the underlying values?” says Beryl Radin, a professor at American University who specializes in federal management reform. “And they all seem to be related to efficiency. It's very hard to find anything that deals with equity. If you did a word search in the PART documents you wouldn't find ‘African American' or ‘Hispanic' once, and you'd only find ‘women' in the titles of some programs.”

Beyond its single-minded focus on efficiency, PART also places surprisingly little weight on the results category. “In fact, there was no correlation between results and budgets, but there was with purpose and budget requests,” says Gilmour. “Purpose,” he says, “is the most subjective and easiest to manipulate.”

* * *

Perhaps most insidious, though, is the way in which PART punishes block grants. Because block grants are given directly to states, measuring results and evaluating management at the federal level is extremely difficult. In 2005, PART rated no block-grant programs effective, while finding 43 percent outright ineffective, compared with 5 percent of programs government-wide. Just by coincidence, it is Republicans who insist that regular government programs be transformed into block grants in the name of local control. It's politically easier for Congress to cut a block grant than it is to cut a program the federal government actually manages by the same amount. It's especially easier with PART rating half of the grants ineffective. Voilà! Local control plus performance-reviewed government equals massive cuts in popular programs -- all done without leaving a fingerprint.

Further, progressive programs or agencies are often punished for following the orders of the statute that created them, or a Supreme Court ruling. The Occupational Safety and Health Administration (OSHA) is forbidden by Supreme Court rulings -- notably the “Cotton Dust” ruling of 1981, among others -- from basing decisions on cost-benefit analyses. But PART officials overrode the Supreme Court's mandates.

In this way, the administration gets a twofer: OSHA, a bane of big business, is penalized, and it agrees to change its behavior. PART has done the same to other industry enemies, such as the Consumer Product Safety Commission and the Mine Safety and Health Administration. Francis Meilinger, an OSHA spokesman, says that while it's true OSHA has begun to do cost-benefit analyses, it ignores the results when it makes decisions. An analysis, Meilinger says, is done only to provide “the public with information on the benefits and costs of new regulations.”

It is instructive to apply PART's logic to well-known programs. Take the levee system in New Orleans; clearly the program was ineffective. PART's answer would be to cut funding. The occupation of Iraq is another program that the administration might not want rated, given its poor performance in the areas of planning, accountability, and lack of appropriate measures of results. But then again, PART is most concerned with purpose, so the occupation's funding would likely be safe from its scythe. And that's one more indication that PART itself is ineffective.

Ryan Grim is a writer for