What Al Gore might learn the hard way.

It's early afternoon in the White House Briefing Room. Still fresh to power, the new President and his men stand in front of the famous dark blue curtains to kick off a major new initiative: reinventing the federal government.

"We must give top priority to a drastic and thorough revision of the federal bureaucracy, to its budgeting system, and to the procedures for constantly analyzing the effectiveness of its many varied services," the President had said during last year's campaign "... This is no job for the fainthearted. It will be met with violent opposition from those who now enjoy a special privilege, those who prefer to work in the dark, or those whose private fiefdoms are threatened."

Hopes are high. "Today, we have submitted to the Congress the first in a series of recommendations for reorganizing the executive branch of Government," the President says confidently, promising a 300-person review of the government's operations.

The afternoon? July 15, 1977. The President? Jimmy Carter. Thin the ties, shrink the lapels, trim the sideburns and fast forward to 1993: After a generation of bright federal moments like the $500 billion S & L collapse, this snippet of Carter and Bert Lance has all the trappings of Bill Clinton and Al Gore's September kickoff to ... reinvent the federal government.

We've been here before. Within the last 30 years, LBJ ordered Robert McNamara's Programming-Planning-and-Budgeting Systems instituted government wide. Nixon had Management by Objectives. Reagan brought in the Grace Commission. And Bush encouraged Total Quality Management. The result? Republicans - and Perot - still score points bashing government, and 80 percent of the public believes that the "country needs to make major changes in the way government works," according to a 1993 Washington Post/ABC News poll.

Now comes Gore's National Performance Review (NPR) to remake government. Despite the jargon ("It's time we had a new customer service contract with the American people"), Gore is campaigning against end-of-the-fiscal-year spending sprees, has got religion on cutting the bureaucracy, and is intent on simplifying the Byzantine, 450-level job classification system. But the forces that have killed reform in Washington since the days before Sherman Adams grew fond of vicuna coats are lying in wait. Ah, says the Gore team, this time will be different. Why? Because of "one of the major thrusts of what we're doing is measuring and controlling results, not inputs," says Bob Stone, the Pentagon reformer who's directing the Gore review. "Too much of the time in government, we spend a lot of money on a problem and think we're fixing it."

This is promising talk, but turning talk to action means compelling bureacrats and politicians to do what they have ducked in years past: Report, in a real way, what they're really accomplishing. There's much more at stake here than the fate of another presidential commission. If Washington picks the administration's pockets on reform, then the core Democratic principle - that we, in John Kennedy's phrase, can do better - becomes suspiciously squishy. And Democrats will find it harder and harder to convince voters that they know what they're doing.

When Clinton first announced Gore would lead the NPR in March, the President said, "We|ll challenge the basic assumptions of every program, asking does it work, does it provide quality service, does it encourage innovation and reward hard work?" Consider, however, the Carter experience. In the seventies, he deployed zero-based budgeting (ZBB) to force every agency to reevaluate its reason for being. In Cart er's ZBB, every federal agency sent three plans to the White House every year, one assuming a 20 percent funding cut. (Carter thought bureaucrats would fess up about what didn't work if faced with declining fortunes.) Agencies complied, but overburdened the system with reams of forms: 10,000 annual "decision packages" poured in from every corner of the federal map. Nobody could even read it all, much less act on it.

"The Carter administration, with the best of intentions," says Ronald Johnson, a budgeting scholar at North Carolina's Research Triangle Institute, "got completely bogged down in forms and spent its time on copious details instead of thinking about the point of a program." The clean-up-the-mess-in-Washington rhetoric died from a thousand paper cuts. But the late seventies weren't bucolic: Energy and jobs and the environment were as bad then as they are now. So why weren't the decision packages eagerly read all over Washington and taken to heart, instead of languishing with Carter? Isn't that logical?

Logic, meet Congress. Congress has locked in three-fourths of the $1.5 trillion budget and oversees line-item spending for everything from the Alaska Natural Gas Transportation System to the Panama Canal Commission. That leaves agencies little power to decide what to spend on their own operations, even on the off-chance bureaucrats admit a program's wasting money.

"Remember, I was sitting at HEW with what was then a $180 billion operation," recalls Joseph Califano, who was Carter's secretary of the old Health, Education and Welfare department." And I had very little control over where 75 percent of the money went. Education, for instance, was run by a computer, doling out money by congressional formula, and Congress wasn't interested in results as much as they were in getting money appropriated."

Carter failed where Johnson had also failed - and for largely the same reasons. Beginning at the Pentagon in 1961, McNamara deployed a system called PPBS to link dollars to a program's performance. Analysis of cost-effectiveness would referee where federal money would go. LBJ liked the idea and applied it across the board in 1965; by 1971, it was dead, the victim of agencies who treated the analysis as a paperwork burden and a Congress that ignored the analysis and continued to appropriate money for political, not rational, reasons.

To improve their odds, Clinton and Gore speak the language recently repopularized by David Osborne and Ted Gaebler in Reinventing Government: Focus on outcomes, not on process. NPR's ultimate credibility hangs on just that: How much can the government deliver? And the only way to find out whether the government is delivering is to evaluate what programs do. Two laws - the Chief Financial Officers Act of 1990 and the Governmental Performance and Results Act of 1993 - now authorize agencies to set goals and measure progress. But neither mandate that budgets be tied to reaching those goals. Here's an old con to watch out for: Bureacrats love forms (remember ZBB) and counting stuff, because both take time. There's a world of difference, though, between counting how many grants the Department of Transportation processes, for example, and figuring out whether the roads have fewer potholes. And unless the bureaucrats and the Congress buy in (remember PPBS), this could all be a waste of time.

After all, we already count more than the Romans did at the pinnacle of the empire. Today at the Bureau of Labor Statistics, for instance, analysts draw on a 328-page book filled with 2,500 "output" indicators. Two million federal workers in 304 organizations tucked away in 62 agencies undergo the productivity review every year. From the number of cases a Justice Department lawyer handles to the "acres of fine lawn maintained" at American military cemeteries in Italy and Tunisia - and around one monument at Gibraltar - we spend hundreds of millions of dollars a year collecting these minutiae.

Form aid

What use is it all put to? "Damn little," says Allen Schick, professor of public policy at the University of Maryland. "They've got so much data on performance already that they don't know what to do with it."

Consider this: When the General Accounting Office surveyed the 103 big agencies that spent 92 percent of the $1.3 trillion in federal outlays in 1991, 76 of them assured auditors that they gather information on their own programs. Press a little harder: How many of these agencies had clearly defined goals? Sixty-seven. And press again: Thirty-three admitted their information-gathering was not related to achieving those goals. Finally, only nine could straight facedly say that the information could usefully measure progress. "The federal experience right now is real thin," admits John Mercer, minority counsel to the Senate Governmental Affairs committee. Two examples from the recent past:

* For years at the EPA, inspectors who monitored the injection of hazardous waste into underground water sources reported the number of inspection visits - not the compliance (or non-compliance the inspection produced. In fact, of the 145 measures the the EPA's Strategic Management and Planning System used in its water oversight, 30 percent were unconnected to any environmental goal - like cleaner water, which was the point of the enterprise. One measure demanded: "Report, by Region and nationally, the number of proposed EPA orders, the number of final EPA ... administrative orders, and the number of final State [administrative orders] issued." The objective for gathering this data? "None," the GAO found.

* For five years, Congress didn't require two welfare-to-work programs - HHS' JOBS and Agriculture's Food Stamp Employment and Training Act, which spend a combined $7 00 million a year to train the disadvantaged - to ask what happened to the people they trained. How many gets jobs, how many keep them, how much money are they making? Or, if trainees were on welfare coming in, did they stay off? Nobody asked. Instead, the law required managers to calculate "participation rates" and "expenditures by target group. "This means Clinton and Gore must make sure what agencies call "performance measures" aren't largely irrelevant busywork. Why not ask the essential questions, too, by tracking a percentage of people over time?

Because the essential questions are risky. Their answers - such as why the Interior Department spends $66 million irrigating lands for corn, barley, rice, and cotton in the Southwest and the Agriculture Department pays farmers $379 million not to grow crops on that same land - are dangerous to the people who make and carry out those orders. So bureaucrat s rightly fear "ending up in the papers or in front of a sub committee," as one mid-level budget analyst puts it.

That remark underscores two cultural obstacles the NPR must overcome. First, there is something that Gore has come to believe in the last six months: That the more time spent hunting "waste, fraud, and abuse," the less constructive evaluation is actually done. In 1978, Carter established Inspectors General (IG) offices in 33 major departments and agencies with, again, the best intentions. However, IGs thrive on cops-and-robbers, seizing on petty fraud instead of reporting on what works and what doesn't. That's because Congress, in its wisdom, tied the IGs' own performance to the number of reports generated and the dollars of waste identified. Of course, somebody needs to keep an eye out for corruption, but an inordinate amount of time goes into quick-hit inspections to see whether, say, computer contracting rules are being followed to the letter. Why not evaluate whether the computers the agency bought can do the job?

"The idea behind the IGs was well-intentioned," says Dick Kuserow, a former Health and Human Services inspector general, "but a lot of us became mindlessly adversarial, intimidating our departments instead of thinking about how we could improve what we were doing." And short of an in-House bank, Congress loves nothing better than a good waste, fraud, and abuse hearing to fulminate about wasting the taxpayers' dollars. Reagan's Grace Commission, a panel of private sector leaders who came in to identify government waste in the eighties, also talked more about slashing spending - to use budgetspeak - than about smartly spending.

The effect of all this? That's the second obstacle. The NPR is confronting an operation where bureaucrats, accustomed to ducking IGs, fear the risk of exposure. In fact, while the Gore plan to empower employees on the front lines by lifting self-imposed regulations sounds wonderful, and the Vice President has run a smart grass-roots campaign to bring bureaucrats into the game, the empowerment plan is based on the shaky assumptions that federal employees who consciously chose security in taking government jobs are chafing to shake up the system. So it is a serious mistake to expect, as most coverage of the NPR has and will, that "Reinvention Labs" and a handful of successful pilots will produce a new government - complete with overhead savings - any time soon.

Take the fate of the great middle management class - 236,500 out of a General Schedule work-force of 1.5 million - which has grown from 8 percent of the government in 1955 to 29 percent today. To save real money with his promised 100,000 (or more) job cuts, Clinton must go after the middle ranks, and the unions (to whom Gore has been unusually courteous) say they're all for that. But in the context of the major corporate restructuring of the last decade, to which both Clinton and Gore approvingly refer, even that's remarkably tame: Since 1985, for example, IBM has cut its payroll by 45 percent. The deepest federal cuts, assuming they come at all, would be 5 percent.

Yet no matter how small, the management pruning can't come soon enough if the administration's reinvention ideas are going to work. Let's suppose - just suppose - that Congress lets agencies make spending decisions on the spot. That has happened before, and there's strong evidence that reform isn't long for this world once middle managers get hold of it. Consider one of the more complete federal experiments with a key plank of the NPR: results-oriented budgeting, where agencies are given flexibility to spend money as long as performance targets are met.

In 1985, the U.S. Forest Service approved a handful of pilot s to test the idea. Out in Oregon's Ochoco National Forest, rangers stream-lined the budget from 27 line items to six, raised productivity by 25 percent, repaired streambeds that had been neglected by Washington budgeters since 1964, and saved $70,000 the first year. The GAO signed off on the experiment, but by 1989, Forest Service managers nervous about losing control of the budgeting - and the reasons for their jobs - withdrew the budgeting power. Ochoco is now back up to 53 line items: "We're worse off than we were before," says Rod Collins, an Ochoco supervisor. "The people who handled the money said we were operating outside the culture, and they couldn't control that we spent money on."

And this approach is precisely what the administration wants to take national. To counter the inertia that killed the pilots - and could massacre the NPR on a grander scale - we need a wel-staffed Office of Management and Budget (OMB) genuinely analyzing programs. But OMB's auditing staff has remained flat for over a generation. In 1955, for a federal budget of $68.4 billion, OMB had 450 staffers. Forty years, eight presidents, a New Frontier, a Great Society, and $1.5 trillion later, OMB has only 550 staffers, only 140 of whom assess what's working. (In contrast, there are 12,000 staffers in the IG offices.) OMB auditors who visit agencies are like tourists dealing with rug merchants abroad for the first time: Uncertain of the local customs and susceptible to built-in price gougings, they only see the final product - in this case, the agency's budget request - and so bargain from weakness. Who knows which division really needs what they're asking for, or where the fat is?

A quantum leap in making sense of government would mix an OMB warring on mediocrity - not waste, fraud, and abuse, but the pedestrian problems of governing - and generating clear reports to give the press the authority and means to criticize government in a way that's now all too rare. The cost is minimal: IGs cost $1.2 billion a year - one half of one percent of the budget. Even doubling that is nothing compared to the $100 billion in federal mismanagement costs the House Budget committee identified in 1991.

Performance smart

So with vigilance and a touch of honesty, things can change. One of the most ambitious pieces of domestic legislation in the eighties, the Job Training Partnership Act (JTPA), offspring of an odd political one-night stand in 1982 between Senators Ted Kennedy and Dan Quayle, is a case study in how to set outcomes, how agencies will try to fool you, and how reform can win out.

The $2 billion program, which reaches two million adults and teenagers, sets clear-cut goals (60 percent of trainees must be working three months after the program ends). If managers fail for two years running, the feds force the local agency to alter its program. It was not always thus. At first, JTPA required its offices just to count immediate job placements, which encouraged local administrators to exclude the hard-core unemployed or the homeless. Called "creaming," this hustle of only counting the best let the local operators meet their standards without breaking a sweat. Once the Labor Department caught on to the gimmick, is set the bar higher: Sixty-five percent of the people enrolled must be long-term unemployed or dropouts or homeless.

Why isn't everybody doing this? "The problem is, accountability is an all-or-nothing proposition," says one Labor Department evaluator. "Once you agree to it, and tie performance to management and even budgets, then as an agency you can't dodge responsibility. And that's terrifying to the bureaucratic mind." Yet terror can be a force for good. In the fifties and sixties, for example, when the IRS audited more returns and hauled more people in for cheating on their taxes, tax compliance was higher than it is now because people were scared they'd get caught. The same principle holds true here: As long as agencies think there's a possibility of evaluation, they're more likely to do their jobs.

The Democrats' task now is to prove that the feds can accomplish smart things and be held accountable for broad policy outcomes like repairing bridges or housing the poor, even when the federal role is simply writing a check to a state or local government. And while Clinton and Gore are avoiding the Carter/LBJ mistake of instant implementation, the past tells us that an administration's interest in government reform dims as the years wear on. Because private sector renovations of this scale take 10 to 12 years - an eternity in politics, where turnover is the rule, not the exception - what sounds grand today will seem dated before long. Just ask Fred Malek, who was Nixon's OMB point man on Management by Objectives, a plan launched in 1971 to make agencies map out strategic plans and base budget decisions on their progress. (Sound familiar?) "This was a great plan, and a workable one," Malek says now. "But the big push for it had to come in 1974 - and as you know President Nixon and his men had other things on their minds in 1974." If the incumbent administration fails to bring along Congress and defeat the bureaucrats this time, the next guys who try this will have to contend at pulling the plug on the most gee-whiz of reinventions. And the system's got plenty of experience at that already.