The piles of briefing materials that crossed George Bush’s transition desk probably didn’t contain a 1964 Burt Lancaster film called Seven Days in May. That may have been the first major mistake of his presidency. In 118 minutes, the film could have taught him more about government than any volume of Heritage Foundation reports and CIA cables. In fact, it could have taught him the most important presidential lesson he could learn.

The film opens with ominous news: President Jordan Lieman has just discovered that the Joint Chiefs of Staff are plotting a coup. The word comes not through the extensive bureaucratic channels that are supposed to keep the president posted but by way of a lone colonel who happens to stumble across the plot. The president needs to know more, and he needs to know it quickly, but where can he turn? He calls on his oldest (though drunken) friend, Senator Ray Clark, for the treacherous mission. “I don’t like sending you. . .” the president apologizes, “but if there were anyone else I could trust. . . .” Clark has to dash around the Texas desert, sweet talk a hooker, and escape from the plotters’ prison, but in the end he gets the job done. The threat is defused, and the Republic endures. While the odds of a contemporary coup may be slight, the film’s instructive potential remains great. Should it make its way into the White House theater anytime soon, Bush would do well to ask himself these questions: Why wasn’t the president getting a steady stream of information all along, as the plot was developing? And why, once he stumbled upon it, could he trust only a personal friend to tell him the truth?

The absence of a military revolt notwithstanding, you can bet that Bush, like Lieman, presides over a government in which any number of major and minor disasters are ticking along undetected. “It’s incredible that a secret base could have been constructed without our hearing about it sir,” Lieman’s aide murmurs in the film–but, as anyone familiar with the federal government knows, it’s not incredible at all. While Washington has so far escaped a coup, in recent months it has been hit with the equivalent in bureaucratic bombshells: a $ 166 billion explosion called the S&L crisis, a $ 130 billion breakdown in our nuclear weapons plants, and a fairy tale of greed and neglect at HUD, whose bill–$ 10 billion or so–is smaller than its sordidness. The main difference between the film and reality is that reality hasn’t been fortunate enough to have a happy ending.

In film or in fact, the president shouldn’t have to rely on drunken old senators to save the day. Generals, cabinet officers, and agency heads are supposed to keep him in the know. And, within his bureaucratic army, one sentinel in particular should make certain that the president has the information he needs: the Office of Management and Budget. As the political scientist Richard Neustadt told John Kennedy during the 1960 transition, OMB (then called the Bureau of the Budget) is “the nearest thing to institutional eyes and ears and memory. . . which will be available to you.”

Eyes and Ears to a president! It’s quite a responsibility, and it’s no surprise that OMB enjoys a position of almost unparalleled prestige in the government. Two of its recent directors–first David Stockman and now Richard Darman–were said to be the smartest people in Washington. And both, in fact, are supremely smart men. The same goes for most of the organization’s staff. Hardly anyone describes it without resorting to phrases like “cream of the crop,” which in some sense it is.

Why, then, is it that HUD coups, and S&L coups, and weapons-plant coups, and any number of other bureaucratic bungles are catching us unaware? Perhaps the most instructive scene of Seven Days in May comes just after the president learns of the planned mutiny. His aide says, “Yes, sir, I’ll call Bill Condon in the Bureau of Budget–right now,” rightly assuming that if anyone should have the details, he should. But poor Condon may be the most authentic Washington figure Hollywood has produced: he doesn’t have a clue. For those of us stuck not with film presidencies but with real ones, an obvious question arises: If OMB is so smart, why is the government so screwed up?

The problem is that the agency employs the wrong kinds of people and has them do the wrong kinds of things. OMB sports a building full of numbersmen–grand totalers adding up the digits, with adding machines whirring and slide rules flying. But to serve as the president’s eyes and ears, OMB needs to go beyond numbers to program analysis–it needs to know what programs are working, what programs aren’t, and why. When it looks at, say, the Department of Energy, it should be asking questions like these: How many different nuclear weapons plants do we need? Are they working? Or are they sending radioactive waste into the Georgia groundwater? If so, what can we do to fix them?

Answers to these kinds of questions don’t come easily and don’t emerge from the numbers alone. To get them, OMB needs an abundance of investigative talent–an army of Sy Hershes. (And all the better, if the investigators have had some government experience and know the bureaucratic cons–Sy Hersh after he’s worked as a GS-9.) Only after OMB has answered the bigger questions about what’s working and why, can it address the ones that now obsess its numbers-oriented staff, such as “What does it cost?” and “Can we afford it?” Phrased differently, the agency needs to knit its “M” functions (management) with its “B” functions (budget)–for without knowing whether a program is needed or works, how can a budget office determine a proper level of funding?

Instead, OMB has segregated its M and B functions and set each side to work furiously at tasks that bear little relation to the real needs of government. On the M-side, it has spent part of its time futzing with minor reforms like paperwork reduction and part of its time waging an ideological war on regulation. But virtually no one on the M-side has been charged with getting out of the Executive Office Building to discover what needs less regulation and what needs more.

While the M-side has ideologues deregulating in the dark, the B-side has technocrats playing with make-believe numbers. OMB sets them to work with endless computations about the federal budget. Rather than perform the program analysis that asks, “Is this job-training program needed, and does it work?” OMB budget examiners spend most of their time answering questions like these: “What would be the potential savings if we changed the means test from $ 10,000 to $ 8,000? Estimate the savings three different times, assuming inflation at 4 percent, 6 percent, and 8 percent. Now let’s run those figures a few more times, assuming unemployment at 5 percent, 5.5 percent, and 6 percent.” Finding the answer may require quite a sophisticated use of mathematical models. And, at some stage in policy planning, the answers are important to know. The problem is that this has become OMB’s main function–and it tells you nothing about whether anyone’s actually being trained for a job.

Don’t assume someone else knows the answer; OMB isn’t the only government watchdog asleep on the job. The failure to detect and prevent billion-dollar screw-ups is shared by the agencies themselves, inspectors general, congressional oversight committees, and the General Accounting Office (see John Heilemann, page 38)–not to mention the press, which perversely continues to dispatch its brightest stars to glamor beats like the White House, where they spend their days shouting questions at the president’s helicopter instead of digging into the realities of the president’s programs. The best we get from watchdogs these days are post-mortem analyses–i.e., the press’s too-little, too-late autopsy of the S&L scandal–rather than up-front reporting that could head the problems off. But while oversight failures are widespread, OMB’s mission is special. It’s the president’s watchdog, after all, and the keeper of his purse. If the rest of the government, including the rest of the government’s watchdogs, truly felt the president’s eyes and ears upon it, there’d be more people too frightened or ashamed to give us the boondoggles we’ve come to expect.

With the appointment of Richard Darman as budget director, OMB has recently had a special chance to transform itself into the kind of agency it should be. The words “independent program analysis” may not be on many Washingtonians’ lips, but if anyone could put them there–and explain how they could help rescue the government–it’s Darman. He’s the “most brilliant intellect” in government, says Newsweek. (Perfect SATs, says Darman, whose virtues don’t include modesty.) Perhaps even more important than his brains is his experience. Darman’s held jobs at the Commerce Department, Treasury, Justice, State, Defense, and what used to be Health, Education, and Welfare. There’s probably no one in Washington, and certainly no one in the government, who should know more about which programs work, and which don’t, and how to find out.

The scene that could have truly shot fear into the hearts of GS-15s, and new life into the government, was this: The new budget director, at his confirmation hearings, declares his intention to become the Oversight Czar. “New Mission Cited for OMB,” the Post headline could have read–maybe even, “In Unusual Hire, Darman taps Hersh as Deputy.” In so doing, Darman could have secured a reputation for himself as not only smart but wise.

Nothing like that happened. After 10 months, there’s no meaningful sign of OMB change, no indication that it will play a more meaningful oversight role in the future than it has in the past. If anything, the agency’s existence as a numbers factory has taken on an extra edge of irony: Darman now presides over an agency obsessed with finding clever (and typically illusory) budget cuts made necessary by the tax cuts that his earlier cleverness as a White House strategist helped sell.

On the few occasions when Darman has had to face criticisms of OMB’s oversight failures, he’s done so in ways that reflect no deep understanding of the problem, and no resolve to change it. One critic is the GAO, which in May issued a report politely titled, “Revised Approach Could Improve OMB’s Effectiveness.” Another critic is Senator John Glenn, whose committee has been holding hearings on the poor performance of government watchdogs. Showing a talent for saying the right things at the right times, Darman recently told Glenn that OMB needs to “help reduce the problems that failures of collective oversight, including its own, may have produced.” He added that he agreed with the call for “better integration of management and budget functions.”

What’s missing is action. Longtime OMB employees say the pressures for numbers production are as intense as ever; in the words of one, there’s “not much that he’s done that would suggest a change.”

Chances are that on June 10, 1921, as he signed the Budget and Accounting Act, Warren Harding had no idea how much good he might be doing the world. The law created not only BoB/OMB but blessed it with an institutional twin–the GAO. Though joined at birth, they were physically separated by Pennsylvania Avenue, with the GAO reporting to Congress and BoB to the executive branch. The agencies had the potential to grow up as Washington’s baddest brothers–the Leon and Michael Spinks of the capital–knocking out programs that were irrelevant or ineffective. But neither ever approximated its potential as a government watchdog. Both were more likely to know the cost of toilet paper in the Pentagon’s men’s room than whether the military’s latest supergadget would fly. While GAO, which is about 10 times larger than OMB, has made slight improvements in recent years, OMB has gone from bad to worse.

Although OMB never conducted program evaluation as much as it should have, there was a time when its analysts came closer than they do today. Richard Stubbing’s experience is instructive, both of the type of people OMB should hire and the type of work they should do. He joined the Budget Bureau as an analyst in 1962 after serving as a junior naval officer in the 1950s. Early in his OMB tenure, Stubbing called the Pentagon to ask how the Navy had decided it needed 240 destroyers in the fleet. “Why, come on over, Mr. Stubbing,” an admiral suggested, and Stubbing, elated at his high-level access, paid the Pentagon a visit. The admiral delivered a two-hour lecture, explaining in elaborate detail why each carrier group needed this many destroyers, each submarine group needed that many destroyers, and so on, down the fleet–“and there, Mr. Stubbing, you have it.” Stubbing nodded dutifully and took notes. The only problem was, the admiral had listed the need not for 240 destroyers but for 360, which shows how much can be learned with just one well-placed question. When Stubbing pointed out the discrepancy, the admiral shot back: “Mr. Stubbing, you may discount our estimates by a third!”

During another sixties field trip that shows the value of getting out of the office, Stubbing visited the North American Corporation, a subcontractor for the Navy’s F-111 bomber. Just six months after the contract was awarded, Stubbing discovered the company’s cost overruns had doubled the price. “It blew our minds,” he said. “We went back and broke the word that this was out of control.” The controversial project was eventually killed.

Of course, just because OMB discovers a bad program, there’s no guarantee it will die–Stubbing was an early (and prescient) critic of the B-1 bomber, upon which Congress and several presidents subsequently lavished $ 28 billion, only to see a test flight downed by a flock of birds. Bad programs can, and often do, have strong sponsors behind them. But two lessons emerge from Stubbing’s experience: 1) OMB has the responsibility to single out the bad programs and call them to the president’s attention; and 2) the agency can best perform this role not by just looking at numbers but by looking at programs and questioning the people involved. “You just ask ‘What the hell is going on?’” Stubbing says.

Today, OMB hardly does this kind of asking at all. The journey from bad to worse started in 1981 with the arrival of David Stockman, who not only ditched the agency’s most important function but had the capital applauding his intelligence while he did it. In the pre-Stockman years, OMB spent only part of its time actually working up the president’s budget, which it then presented to Congress and let the legislative committees take their course. In the off-season, budget examiners were expected to keep tabs on the latest developments in their agencies. In truth, more examiners spent the off-season relaxing than investigating; still, at least the expectation of monitoring the program was there, and some examiners took advantage of it. Under Stockman, however, there was no off-season. Stockman threw himself into constant negotiations with Congress, and, in so doing, he kept his examiners furiously busy working up data for the debate.

Stubbing captures the mood of the time in a story he tells partly in admiration for Stockman’s intellect and partly in exasperation at the pointlessness of it all. Boning up for a negotiation on defense, Stockman had his defense analysts work Labor Day weekend–all day Saturday, all day Sunday, all day Monday, and on into the night. Stubbing remembers Stockman asking for the cost of every airplane the military was scheduled to purchase in the next year. And more. Things like the past spare parts costs. Projections of future spare parts costs. With inflation. Without inflation. This required Stubbing’s group of 40 analysts to make agonizingly complex calculations. They delivered one set of figures at 10 a.m. Tuesday, another set at 2 p.m., and another set that evening after Stockman returned from giving a dinner speech in Philadelphia. Stockman bid him good night at 1 a.m. and asked Stubbing to meet him back at the office at 7 a.m. “Jesus, after a while you have no idea what it all meant,” Stubbing says. “The whole division was tied up in mindles numbing detail.”

Stubbing, who left in 1981, wasn’t the only one at the office to question the sense of these name-that-number exercises. That year, a task force of the agency’s civil servants issued an internal report on the effect of the new pressures. It warned, “Examiners feel that they are manipulating numbers in the abstract and progressively losing sight of what lies behind them.”

Reality faded further in 1985, when the Gramm-Rudman-Hollings Act began requiring the budget to meet an annual deficit reduction target. Target is a key work here–the act doesn’t require that the deficit actually be reduced, just that the numbers be manipulated to give that appearance. Even Senator Ernest Hollings, one of the law’s sponsor’s, now calls it “pure sham.” The point is, it’s a time-consuming sham. Now, after working up the numbers to present the president’s budget, and reworking them repeatedly for congressional negotiations, OMB has to play with them some more to meet the deficit target.

This can be done through all kinds of dodges. The administration recently bumped a military pay day from one fiscal year to another, for a “savings” of $ 4 billion. It took the Postal Service “off budget” and saved another $ 1.7 billion. Richard Darman ought to win an award for the trick he unveiled this spring: As Alan Murray reported in The Wall Street Journal, by taking the S&L bailout off-budget, the administration made the $ 166 billion loss look like a $ 14 billion savings in deposit insurance. In actuality, that move cost taxpayers an extra $ 5 billion since it made the bailout bonds more expensive. This has been the contribution of today’s OMB–finding ways to make an S&L debacle seem like a budget-booster, instead of finding ways to prevent it.

To understand further the ineffectiveness of today’s OMB, it’s important to understand its structure–the artificial division of the M from the B. The intellectual roots of this separation date back to the first part of the century, with the rise of the scientific management movement. Its theorists argued that there were certain universal management truths that applied no matter what was being managed; hence government managers would need no particular grounding in the substance of actual programs, only an expertise in principles of efficiency. There is something to be said for this view, but not very much. A modern-day parallel is the teacher’s college methodology course devoid of subject content. And just as teachers learn to teach best when teaching a particular subject, the government will manage best when its efficiency experts are actually engaged with a program.

The past decade has seen no shortage of M-side initiatives. Between 1980 and 1987, the flow chart on the M-side was reconfigured six times, and each change brought predictions of an outbreak of government efficiency. The grandest plan was called “Reform ’88,” a long-term project unveiled in 1982 and touted as “one of the most comprehensive and ambitous efforts to improve management ever undertaken…. “It’s now just one more initiative on the trash heap of OMB history.

The project’s first problem was its forest-for-the-trees quality. Reform ’88 vowed to upgrade software. It vowed to modernize cash management. It vowed to reduce paperwork. There is the early days was Deputy Director Joseph Wright, posing for The Washington Post before a garbage pail and announcing that the administration had eliminated 1,998 unnecessary publications. Even taking at face value Wright’s figure of $ 36 million saved, what do you have? About .02168 percent of an S&L crisis.

A second problem with the M-side’s Reform ’88 busywork was that hardly anyone on the B-side–where the sustained contact with government agencies takes place–understood it. A GAO survey of budget examiners found a whopping 10 percent willing to say the support of the M-side was of any use in their actual dealings with agencies.

In his congressional testimony last month. Darman conceded that Reform ’88 had flopped and that the budget examiners weren’t sufficiently involved. But while he offered the right diagnosis, he gave the wrong prescription. Darman wants to bolster the M-side when the right solution is to abolish it–OMB isn’t going to improve the government unless its budget experts and its management experts are the same and they get out of the office to look at what’s really happening.

The same distance from what’s really happening is evident in the M-side’s Office of Information and Regulatory Affairs, the division charged with reviewing the federal government’s regulatory activity. There’s certainly an important watchdog role for OMB to play in regard to regulation. What it ought to be doing is going out into factories and mines, checking to make sure that vital regulations are being enforced. Beyond that, it should be asking: Which regulations are needed and which should be abolished? Are others needed that are not yet on the books? It’s true that some government regulations are dumb and dispensable–OMB has delighted in pointing to examples like an education department rule that prohibits school dress codes from distinguishing between boys and girls–but on the whole the federal government does too little to regulate, not too much. This is particularly true in areas involving safety.

But rather than play discerner and enforcer, OMB has played spoiler. It’s used its powers of regulatory review to create a bureaucratic bottleneck that makes it more difficult than ever to issue safety regulations. Though more than 120 people have died in grain elevator explosions in the past dozen years, the Occupational Safety and Health Administration has been fighting OMB for seven years to get stricter regulations on the elevators. When the Food and Drug Administration tried to place warning labels on aspirin bottles, telling parents that giving aspirin to children with chicken pox or flu can cause a sometimes fatal disease called Reye’s Syndrome, OMB held a private meeting with the aspirin industry–then told the FDA to withdraw the regulation. It took the FDA another three years to get the regulation adopted, and 3,000 more children contracted the disease in the interim. An Environmental Protection Agency regulation on radioactive waste sat on an OMB desk for so long, the EPA staff held a birthday party for it.

The key thing to remember is that with OMB keeping the world safe from aspirin warning labels, there’s no one left in government to ensure that those regulations that do get on the books get enforced. Congress may pass a new mine safety bill and the president may sign it, sending a warm glow through a Washington that feels satisfied it has done its duty toward the miners. But who’s making sure the Mine Safety Administration is enforcing the law? Of course the congressional committees could always do it. Or the press. Or the GAO. But ask a miner how many of these folks he’s seen down there.

The M-side of OMB has traditionally enjoyed a stepchild status in relation to the B-side, and the past decade has been no exception. It’s the budget examiners who are really in the center of the action, and it is through them that the failures of OMB become most apparent.

“They have I.Q. points on every other person in government,” says a former OMB supervisor. But these best of the best have a few problems. One, there just aren’t enough of them. With only 200 examiners for the entire government, some have up to $ 5 billion in budget responsibility. To really do its job, OMB should be about three times larger; instead, OMB has actually shrunk by about 12 percent in the past eight years. OMB’s second handicap is the average examiner’s background. Most come young, with little, if any, experience in government, meaning they have little sense of the way the bureaucracy actually works. And since many leave after three or four years, by the time they get that sense, they’re gone. Nor do the public policy schools that most young budget examiners attend stress the kind of investigative skills that OMB needs, emphasizing quantitative skills instead. Since even these inexperienced examiners do have the advantage of smarts on their side, with the right kind of supervision, they could still do okay. But they’re typically supervised by others with similar backgrounds.

While there’s nothing wrong with numbers analysis as one tool of program evaluation, there’s plenty wrong when it becomes the main tool, as it has at OMB. One former budget examiner said her proudest moment at OMB came when she ran a regression analysis to disprove a security agency’s claim that excessive overtime was leading to higher attrition–all well and good, but the regression analysis revealed nothing about how well the agency was providing security. Her proudest moment! This isn’t quite as bad as if Sy Hersh said his proudest moment in journalism was catching a typo in Henry Kissinger’s memoirs. But it’s close. Another former budget examiner, Bruce Johnson, said he felt particularly adept at “scrubbing budgets” for excess cost–that is, knowing an agency’s size, he could compute how much money it should spend on salaries, benefits, travel, rent, phone, contracts, and so on. “But the tough call,” he said, “comes, after you’ve done all that, in deciding whether the thing they want to do is worthwile.”

Since OMB spends much more time scrubbing budgets than scouting programs, decisions about “whether the thing they want to do is worthwhile” get made in twilight at best, and often in complete darkness. The average budget examiner will know the conventional wisdom about his program and be up to date with major news accounts or journal articles. He’ll know the administration’s general ideology (i.e., “deregulate”) and his supervisor’s opinions. In other words, taking these factors into account, he guesses.

The most obvious pitfall of this system is the chance of being completely blindsided; how many coup plots get reported through the networks of conventional wisdom? A less obvious, but similarly debilitating, consequence of the examiners’ twilight zone is the process of government by increment. Rather than urging great expansions of the programs that work, and termination of those that don’t, OMB tinkers on the margins. As one former examiner explains, “Instead of getting to the heart of a program, you’re asking, ‘Should we have them take a one-and-one-half percent cut?’” And you’re back to numerical analysis.

Clearly, OMB needs to cast a wider recruiting net. Investigative journalists offer one potential pool of talent; investigative lawyers another–both have experience in peeling back appearances to find out what’s really happening. The ideal OMB examiner would have not only the ability to coax the facts but the mind to understand their meaning, to transform facts into conclusions. Like “Here’s how this program should be changed.”

For a third source of talent, OMB should look to the federal government itself; if some of the bureaucracy’s best staff rotated in, even on a temporary basis, OMB could keep itself stocked with people who are not only smart but also savvy in the tricks of the trade. (Though you wouldn’t want such examiners monitoring the agencies they plan to return to, of course.) As Richard Stubbing’s career shows, it helps to have been around.

Remember, the agencies that deal with OMB are full of bureaucratic ploys, old and new. Picture the typical face-off: a wily Interior Department official with 20 years of bureaucratic survival skills behind him versus a smart but green budget examiner, who is terribly overworked. In describing the problems of inexperience, Ursula Gillis, a candid former examiner, tells a story about a time she was snookered by the IRS, an agency that knew how to get what it wanted. What it wanted was more people to process claims. Knowing that OMB wouldn’t approve the request, the IRS instead asked for $ 400 million to hire not processors but auditors, presenting a study that showed the auditors would collect an extra $ 2.4 billion–for a net savings of $ 2 billion. But the IRS subsequently shifted the money away from enforcement and back into processing. And the $ 2 billion savings disappeared.

Of course, it’s possible that the IRS really needed the processors and was simply playing the necessary game to get them. It’s also possible that it didn’t need them at all but saw a chance for more money and took it. The point is that OMB is supposed to know what the IRS needs and what it’s doing with the money it’s getting, not be the victim of a shell game. The ploy, as it turns out, was an old one. “I was too naive,” Gillis says. “It’s the kind of institutional memory you’d hope would get passed down. But you have lots of turnover. By the time you know what’s going on, you’re out the door.”

Among the more tested bureaucratic tricks is one that budget examiners call The Washington Monument Game: when an agency is threatened with a budget cut, it’ll be sure to propose suspending its most important or popular functions first. That is, when the budget examiner suggests that the Park Service may have to face a 10 percent cut, the Service is likely to respond that this will result in closing down the Washington Monument–the kind of move that will be sure to generate angry letters and congressional intervention. Understanding the principle in theory is easy; having the experience to guard against it in practice is something else. As Gillis says, “If you’ve never worked in government before, it’s hard to sort out the bullshit from what’s real.”

A second old-time bureaucratic trick is the fourth-quarter spending spree–the rush of each agency to dump excess funds before the fiscal year’s end, lest it show a surplus and suggest that it’s overfunded. This is the time of year when the Park Service buys electric blankets for the rangers in the Everglades–anything to dump the dough. Here OMB’s failure isn’t so much one of knowledge but one of action. It has groused about this for many years, but when the trade publication Industry Week followed up in 1985, it found OMB wasn’t even tracking the fourth-quarter expenditures, never mind restraining them. The budget bureau’s longtime failure to take effective action on the problem invites the suspicion that there may be bureaucratic pride involved; to the extent that an agency’s bloated budget was one approved by OMB, an examiner’s crackdown reflects badly on himself.

The only way for OMB really to guard against the bureaucracy’s survival schemes is to get out and see programs in action. Had Gillis been quizzing IRS executives the way Stubbing quizzed admirals and defense contractors, chances are she would have picked up on the difference between the agency’s real needs and its old games. Imagine how much sorrow we could save with a small investment in expanding OMB’s staff enough to get the Gillises in the field. Which would you rather pay for–400 more budget examiners, or the S&L bailout?

Of course, even with three times its current staff, OMB couldn’t visit every government program every year. But it doesn’t need to. What it does need to do is to establish a reputation for surprise visits and thorough ones. This was the principle that the IRS itself put to work back in the old days; taxpayers knew the IRS couldn’t audit everyone, but the agency’s random and tough audits were enough to keep most people honest. If government managers knew that OMB could drop in at any time–and that if it did, it would discover the truth about their program–bureaucrats might learn to fear screw-ups the way taxpayers used to fear cheating.

But, like most other examiners, Gillis says she got out to visit programs just a week or two a year. And what kind of visits were they? OMB is not known for arriving unannounced. “If you’re from OMB, they don’t like you to be uncontrolled,” says Gillis. “They’re cautioned about what to say and what not to say.”

For a final sense of how well most of the government controls its watchdogs, consider the principle of Dare to Fail Great–minor screwups may bring you trouble from the budget office, but the agency that founders on a sufficiently large scale gains the advantage. The IRS’s disastrous 1985 tax season is a case in point. This was the year when the IRS, at its Philadelphia office alone, lost track of $ 300 million in withholding tax payments, wiped out the records of 10,000 taxpayers, and dumped other returns in the trash can. As Gillis explains it, the scale of the disaster, which prompted nationwide headlines and a congressional investigation, worked to the agency’s advantage. If the IRS then said it needed more money, no budget examiner would want to risk becoming the next season’s scapegoat by denying the request.

But if you haven’t done the kind of deep investigation that allows you to understand the agency, how do you, eyes and ears to the president, watchdog of watchdogs, happen to know whether the IRS actually needs more employees, or just needs to make better use of the ones it has?

While the underlying weaknesses of OMB remain constant–its lust for numbers, its lack of on-site investigation–the actual circumstances of its failures vary by case and therefore invite individual scrutiny. (For a look at OMB’s role in the S&L crisis, see page 28.)

Incredibly enough, OMB’s failure to head off the defense outrages of the early Reagan administration–of which the $ 433 claw hammer is a minor but apt symbol–weren’t inadvertent but were the product of design. The 1980 Republican platform went so far as to demand the throttling of defense analysts like Stubbing, denouncing the “ill-informed, capricious intrusions of the Office of Management and Budget. . . on defense planning.” The administration made good on its pledge. A conservative defense expert named William Schneider Jr. took charge of OMB’s national security division and put it to sleep. “His mission in life was to cut us off, and he did,” Stubbing says. About 30 percent of OMB’s defense analysts, including Stubbing, left in the first 21 months.

In the meantime the administration just picked a number out of the air–a large one, as it happened: a $ 32.6 billion increase–and told the military to find a way to spend it. To comply, the astonished services had to pull out long-dormant wish lists. Stockman, who in early 1981 had deferred to Defense Secretary Caspar Weinberger on defense spending, later tried to fight him–mobilizing Stubbing & Co. over Labor Day weekend–but by then it was too late. Weinberger had won. The subsequent billions wasted on weapons that don’t work are vivid illustrations indeed of what happens when government keeps its watchdog chained.

(To gauge how well the press generally understands the agency, consider that in 1982, with OMB’s defense analysts under wraps, The New York Times published an OMB primer that explained these are the people “who question the need for new weapons systems. . . .”)

The irony of this surgical strike against OMB’s defense analysts is that they were never an overly vigilant group in the first place. With a few exceptions–like Stubbing and the opposition to the B-1 bomber–the budget examiners had a long history of deference toward the military. As Peter Szanton, an associate OMB director during the Carter years, has written, “OMB imposes far less discipline on the DOD budget. . . than it does on domestic budgets.” Szanton sees this reticence contributing to a list of military ills, including the interservice rivalries that proceed unchecked, constant cost overruns, and the military’s bedazzlement by high tech. OMB’s caution, he says, stems both from the historic clout of the defense secretary and the mystique of military expertise. As a former defense budget analyst explained to Szanton, “When the services say ‘military judgment,’ the curtain just comes down.”

Szanton argues that the historic failure of OMB to mind the defense store is an illustration of the larger truth that “most White House agencies. . . have little time for implementation. Implementation is merely what happens.”

Perhaps the most instructive look of all at OMB’s lack of interest “in merely what happens” comes from the collapse of HUD, whose cost to the taxpayer is large, but whose perversion of mission and contribution to human suffering is even larger. Remember, among the things that “merely happened” during HUD’s demise is that, as James Watt and Carla Hills grew rich, the homeless population swelled and more children of poverty were consigned to shelters and tenements. If government oversight (or the lack thereof) has a human face, that of children in tenements is it.

Were HUD’s now-familiar story to be entered in some watchdog casebook, it’d be ridiculed as parody. Imagine: A cabinet secretary (Sam Pierce) sleeps through his eight-year tenure, keeping 10-to-4 hours at his multi-billion dollar agency and watching soap operas. Filling the vacuum is the best-connected barmaid in Georgetown (Debbie Dean), who takes control of the department and doles out contracts for questionable projects to her pals. A host of get-the-government-off-the-people’s-back conservatives pocket huge fees for lobbying the agency. In seeking greater efficiency through “privatization,” the agency hands its programs over to private companies who leave the government saddled with $ 5 billion worth of loan liabilities. A southern California country club qualifies for a subsidy. Enter “Robin HUD” (Marilyn Louise Harrell), who steals more than $ 5 million before anyone notices. It’s about as far-fetched as the coup plot in Seven Days in May.

And OMB performed about as well as it did in the movie. As it turns out, the same OMB supervisors who have responsibility for HUD also oversee the S&Ls–a performance that must have set some kind of oversight record for Most Billions Missed.

OMB was busying itself pushing for budget cuts and greater privatization of HUD functions, both of which it achieved. In the meantime, HUD’s greater use of private contractors set the stage for two of the subsequent breakdowns: 1) HUD increased its reliance on private escrow agents like Marilyn Harrell, who handled the sales of foreclosed properties, but frequently failed to forward the funds. The total cost to the government is now estimated at about $ 20 million. 2) HUD turned to private companies to screen applicants for a co-insurance program, in which HUD bore 80 percent of the financial risk. This proved a costlier mistake: the government’s liability for the defaults that resulted now stands at about $ 1 billion.

There was nothing inherently wrong with a smaller HUD, and one that relies more extensively on private contractors. After all, these objectives were part of the philosophy on which Ronald Reagan had campaigned and won. But OMB had an obligation to keep the president posted on the policies’ effects–which it’s now painfully obvious that it failed to do. “They weren’t interested in correcting,” says Charles Dempsey, who served as HUD’s inspector general until 1985. “They were interested in cutting.” Actually, this understates the case: OMB didn’t just ignore the danger signs; in at least one case it worked to suppress them.

To follow the story, first consider the signals ignored and then the one suppressed. As HUD was on the verge of catching fire, a few bureaucratic alarms were sounded. In retrospect, they seem pitifully faint–but OMB should have known that’s the way most inspectors general work and should have been extra vigilant as a result. When most government auditors discover a problem, they couch it in terms that are least likely to arouse the bureaucracy’s enmity. Rather than scream “the building is on fire,” they will remark, alongside some comment about the excessive number of staples purchased last year, that “signs of combustion were evident in several hallways”–and be certain to add that corrective anticombustion action is being taken by the agency. “No one’s going to say ‘the building’s on fire.’ You’ll get in trouble for that,” is how one former budget examiner describes the average inspector general report. “Everyone’s covering their ass.”

Every six months, the inspector general would forward his delicate warnings to OMB, where they were promptly filed and forgotten. For a sampling of the tone, consider the March 31, 1987 report, in which Paul Adams, Dempsey’s successor as inspector general, explained that HUD could save $ 844 million just by refinancing a loan program at more favorable interest rates. Startling as that news was, its effect was blunted by the added soporific, “We have developed recommendations which the department is working on to accomplish this refinancing.” It turns out, as The Wall Street Journal recently reported, that the department was working on no such thing. Top HUD officials resisted the plan, for fear the savings would get plowed back into the mortgage-subsidy program, which it was trying to eliminate. Better just to waste the money.

In the case of the two disasters in question–the escrow agents and the co-insurance program–Adams posted similar clues. Three audits, beginning in August 1987, reported that HUD couldn’t account for millions owed to it by escrow agents and that the delays alone were costing the agency up to $ 16 million a year in interest. (Never mind that the report went on to say that, “HUD has recently implemented measures. . .” to improve the situation.) In the case of the co-insurance program, the inspector general’s warnings dated back to 1985–and with no hedging qualification. All OMB had to do to catch on was read the reports and translate: “This means fire!”

And if budget examiners had been out talking to HUD officials, the warning could have come even sooner. (After all, it doesn’t take much quantitative analysis to know that a California country club is not a low-income project.) Speaking of the co-insurance program, one HUD official wrote in 1984, “This is the most fraud-prone system ever spawned by HUD. . .” In 1986, another followed: “I am convinced that financial problems of national importance are inevitable unless something is done.”

While the inspector general’s reports gave OMB one set of reasons to start worrying about HUD, a second series of reports provided another. This signal wasn’t just ignored, it was altered. The signal was the product of the 1982 Federal Manager’s Financial Integrity Act, which requires the government’s department heads to report to Congress and the president once a year whether they have “reasonable assurance” that their internal controls and accounting systems are in order.

Pierce, in an unusually responsible move, wrote OMB in 1983 to say that he could not provide reasonable assurance. The fact that he was one of only three of the 17 agency heads to do so might have provided OMB with notice that all wasn’t well at HUD. In 1984 and 1985, Pierce again wrote to say that reasonable assurance was beyond HUD’s grasp. What’s more, guess what problems he said he was having? That’s right: Fraudulent escrow agents. Problems with the co-insurance program. In other words, here was Pierce, breaking from the soap operas to raise his hand and mumble the equivalent of “Well, now that you mention it, we’re not quite sure where all these billions of dollars are going.”

And what happened as a result? Congress yawned and did nothing. OMB went one step further and said, “Hey, shut up over there!” Jack McGrath, the HUD official who filed the reports, says officials on the M-side of OMB pressured the agency to file a rosier scenario. “I got a couple of calls saying, ‘Why are you all still saying you don’t have reasonable assurance?’ There was clearly pressure on the agency to say we did have reasonable assurance.”

As it turned out, the best HUD could muster in 1986, even after the phone calls, was a statement that it had qualified reasonable assurance–within “limits”–which has remained its policy to the present. Beyond asking HUD to change its wording, did OMB ever follow up to make sure the weaknesses described in the reports were being addressed? “No,” says McGrath. “Not at all.” McGrath was later detailed to OMB himself, where he was in charge of receiving the Financial Integrity Act reports from across the government. “If you had a chance to go through them, you’d go through them,” he says. “Basically you just put them in a book.”

But even if the reports were simply filed away, one might think that McGrath, once he arrived at OMB, would have taken the opportunity to share his knowledge of HUD’s problems–to lean over and say, “Pssst. . . the building’s on fire.” But McGrath says he never discussed the agency’s problems. “I guess I assumed they weren’t unknown,” he says. “At that point my focus wasn’t on the department.” And his OMB colleagues didn’t ask.

There’s a kicker to this story of OMB as the watchdog that wasn’t. In 1985, when Paul Adams, the inspector general, conducted his first audit of the co-insurance program, he found $ 36 million in inflated appraisals on $ 120 million worth of property and warned that the program, which, remember, would ultimately cost us $ 1 billion, was in trouble. As he repeated in his semi-annual report: “. . . . appraisals were unsupported, properties were overvalued, and HUD’s insurance risks increased.” Within HUD, he expressed this view to the Assistant Secretary for Housing, Janet Hale. Adams later testified before Congress, in an account that Hale confirmed, that she found his report “premature, unjustified, and unfair,” and she resisted his suggestions for tighter controls over the program.

Hale: “I don’t remember having significant dialogue with the inspector general or his people. I, again, having come from a background of budget, much more than managing co-insurance, but having some very capable people, I know there was extensive conversation, because we developed the response back to the inspector general, agreeing with some of his findings, disagreeing with some of his findings. . . .”

She’s a top official at OMB, with oversight of a quarter of the federal budget–the quarter that includes both HUD and the S&Ls. The woman she replaced, Carol Crawford, who previously oversaw HUD and the S&Ls, now works at the Justice Department as an assistant attorney general for legislative affairs. That is, she keeps Congress posted on the Justice Department’s progress on the HUD investigation.

Before the hearing ended, Lantos asked Hale a revealing, if not terribly difficult, question. Did she, in retrospect, think that Deborah Gore Dean–the HUD scandal’s leading lady–had possessed the “experience, training, judgment” and qualifications to run the agency?

Which richly qualifies her for her current job as a top member of the president’s hear-no-evil, see-no-evil team of watchdogs.

Jason DeParle is a contributing editor of The Washington Monthly.

Our ideas can save democracy... But we need your help! Donate Now!

Follow Jason on Twitter @JasonDeParle. Jason DeParle is a reporter for the New York Times and the author of A Good Provider Is One Who Leaves: One Family and Migration in the 21st Century, from which parts of this essay are adapted. He was an editor at the Washington Monthly from 1987 to 1989.