Headline: What Government Agencies Do and Why
They Do It
Byline: James Q. Wilson
By the time the office opens at 8:45 A.M., the line of people waiting to do business at the
Registry of Motor Vehicles in Watertown, Massachusetts, often
will be twenty-five deep. By midday, especially if it is near
the end of the month, the line may extend clear around the building.
Inside, motorists wait in slow-moving rows before poorly marked windows to get a driver's license or to register an automobile. When some one gets to the head of the line, he or she is often told by the clerk that it is the wrong line: "Get an application over there and then come back," or "This is only for people getting a new license; if you want to replace one you lost, you have to go to the next window." The customers grumble impatiently. The clerks act harried and sometimes speak brusquely, even rudely. What seems to be a simple transaction may take 45 minutes or even longer. By the time people are photographed for their driver's licenses, they are often scowling. The photographer valiantly tries to get people to smile, but only occasionally succeeds.
Not far away, people also wait in line at a McDonald's fast-food restaurant. There are several lines; each is short, each moves quickly. The menu is clearly displayed on attractive signs. The workers behind the counter are invariably polite. If someone's order cannot be filled immediately, he or she is asked to step aside for a moment while the food is prepared and then is brought back to the head of the line to receive the order. The atmosphere is friendly and good-natured. The room is immaculately clean.
Many people have noticed the difference between getting a driver's license and ordering a Big Mac. Most will explain it by saying that bureaucracies are different from businesses. "Bureaucracies" behave as they do because they are run by unqualified "bureaucrats" and are enmeshed in "rules" and "red tape."
But business firms are also bureaucracies, and McDonald's is a bureaucracy that regulates virtually every detail of its employees' behavior by a complex and all-encompassing set of rules. Its operations manual is six hundred pages long and weighs four pounds. In it one learns that french fries are to be nine-thirty-seconds of an inch thick and that grill workers are to place hamburger patties on the grill from left to right, six to a row for six rows. They are then to flip the third row first, followed by the fourth, fifth, and sixth rows, and finally the first and second. The amount of sauce placed on each bun is precisely specified. Every window must be washed every day. Workers must get down on their hands and knees and pick up litter as soon as it appears. These and countless other rules designed to reduce the workers to interchangeable automata were inculcated in franchise managers at Hamburger University located in a $40 million facility. There are plenty of rules governing the Registry, but they are only a small fraction of the rules that govern every detail of every operation at McDonald' s. Indeed, if the OMV manager tried to impose on his employees as demanding a set of rules as those that govern the McDonald' s staff, they would probably rebel and he would lose his job.
;
It is just as hard to explain the differences between the two organizations by reference to the quality or compensation of their employees. The Registry workers are all adults, most with at least a high-school education; the McDonald's employees are mostly teenagers, many still in school. The Registry staff is well-paid compared to the McDonald' s workers, most of whom receive only the minimum wage. When labor shortages developed in Massachusetts during the mid-1980s, many McDonald's stores began hiring older people (typically housewives) of the same sort who had long worked for the Registry. They behaved just like the teenagers they replaced.
Not only are the differences between the two organizations not to be explained by reference to "rules" or "red tape" or "incompetent workers," the differences call into question many of the most frequently mentioned complaints about how government agencies are supposed to behave. For example: "Government agencies are big spenders." The Watertown office of the Registry is in a modest building that can barely handle its clientele. The teletype machine used to check information submitted by people requesting a replacement license was antiquated and prone to errors. Three or four clerks often had to wait in line to use equipment described by the office manager as "personally signed by Thomas Edison." No computers or word processors were available to handle the preparation of licenses and registrations; any error made by a clerk while manually typing a form meant starting over again on an other form.
Or: "Government agencies hire people regardless of whether they are really needed." Despite the fact that the citizens of Massachusetts probably have more contact with the Registry than with any other state agency, and despite the fact that these citizens complain more about Registry service than about that of any other bureau, the Watertown branch, like all Registry offices, was seriously understaffed. In 1981, the agency lost 400 workers-about 25 percent of its work force-despite the fact that its workload was rising.
Or: "Government agencies are imperialistic, always grasping for new functions." But there is no record of the Registry doing much grasping, even though one could imagine a case being made that the state government could usefully create at Registry offices "one-stop" multi-service centers where people could not only get drivers' licenses but also pay taxes and parking fines, obtain information, and transact other official business. The Registry seemed content to provide one service.
In short, many of the popular stereotypes about government agencies and their members are either questionable or incomplete. To explain why government agencies behave as they do, it is not enough to know that they are "bureaucracies"-that is, it is not enough to know that they are big, or complex, or have rules. What is crucial is that they are government bureaucracies. As the preceding chapters should make clear, not all government bureaucracies behave the same way or suffer from the same problems. There may even be registries of motor vehicles in other states that do a better job than the one in Massachusetts. But all government agencies have in common certain characteristics that tend to make their management far more difficult than managing a McDonald' s. These common characteristics are the constraints of public agencies.
The key constraints are three in number. To a much greater extent than is true of private bureaucracies, government agencies (1) cannot lawfully retain and devote to the private benefit of their members the earnings of the organization, (2) cannot allocate the factors of production in accordance with the preferences of the organization's administrators, and (3) must serve goals not of the organization's own choosing. Control over revenues, productive factors, and agency goals is all vested to an important degree in entities external to the organization-legislatures, courts, politicians, and interest groups. Given this, agency managers must attend to the demands of these external entities. As a result, government management tends to be driven by the constraints on the organization, not the tasks of the organization. To say the same thing in other words, whereas business management focuses on the "bottom line" (that is, profits), government management focuses on the "top line" (that is, constraints). Because government managers are not as strongly motivated as private ones to define the tasks of their subordinates, these tasks are often shaped by [other] factors.
Revenues and Incentives
In the days leading up to September 30, the federal government is Cinderella, courted by legions of individuals and organizations eager to get grants and contracts from the unexpended funds still at the disposal of each agency. At midnight on September 30, the government's coach turns into a pumpkin. That is the moment-the end of the fiscal year at which every agency, with a few exceptions, must return all unexpended funds to the Treasury Department.
Except for certain quasi-independent government corporations, such as the Tennessee Valley Authority, no agency may keep any surplus revenues (that is, the difference between the funds it received from a congressional appropriation and those it needed to operate during the year). By the same token, any agency that runs out of money before the end of the fiscal year may ask Congress for more (a "supplemental appropriation") instead of being forced to deduct the deficit from any accumulated cash reserves. Because of these fiscal rules agencies do not have a material incentive to economize: Why scrimp and save if you cannot keep the results of your frugality?
Nor can individual bureaucrats lawfully capture for their personal use any revenue surpluses. When a private firm has a good year, many of its officers and workers may receive bonuses. Even if no bonus is paid, these employees may buy stock in the firm so that they can profit from any growth in earnings (and, if they sell the stock in a timely manner, profit from a drop in earnings). Should a public bureaucrat be discovered trying to do what private bureaucrats routinely do, he or she would be charged with corruption.
We take it for granted that bureaucrats should not profit from their offices and nod approvingly when a bureaucrat who has so benefited is indicted and put on trial. But why should we take this view? Once a very different view prevailed. In the seventeenth century, a French colonel would buy his commission from the king, take the king's money to run his regiment, and pocket the profit. At one time a European tax collector was paid by keeping a percentage of the taxes he collected. In this country, some prisons were once managed by giving the warden a sum of money based on how many prisoners were under his control and letting him keep the difference between what he received and what it cost him to feed the prisoners. Such behavior today would be grounds for criminal prosecution. Why? What has changed?
Mostly we the citizenry have changed. We are creatures of the Enlightenment: We believe that the nation ought not to be the property of the sovereign; that laws are intended to rationalize society and (if possible) perfect mankind; and that public service ought to be neutral and disinterested. We worry that a prison warden paid in the old way would have a strong incentive to starve his prisoners in order to maximize his income; that a regiment supported by a greedy colonel would not be properly equipped; and that a tax collector paid on a commission basis would extort excessive taxes from us. These changes reflect our desire to eliminate moral hazards-namely, creating incentives for people to act wrongly. But why should this desire rule out more carefully designed compensation plans that would pay government managers for achieving officially approved goals and would allow efficient agencies to keep any unspent part of their budget for use next year?
Part of the answer is obvious. Often we do not know whether a man ager or an agency has achieved the goals we want because either the goals are vague or inconsistent, or their attainment cannot be observed, or both. Bureau chiefs in the Department of State would have to go on welfare if their pay depended on their ability to demonstrate convincingly that they had attained their bureaus' objectives.
But many government agencies have reasonably clear goals toward which progress can be measured. The Social Security Administration, the Postal Service, and the General Services Administration all come to mind. Why not let earnings depend importantly on performance? Why not let agencies keep excess revenues?
But in part it is because we know that even government agencies with clear goals and readily observable behavior only can be evaluated by making political (and thus conflict-ridden) judgments. If the Welfare Department delivers every benefit check within 24 hours after the application is received, Senator Smith may be pleased but Senator Jones will be irritated because this speedy delivery almost surely would require that the standards of eligibility be relaxed so that many ineligible clients would get money. There is no objective standard by which the trade-off between speed and accuracy in the Welfare Department can be evalu ated. Thus we have been unwilling to allow welfare employees to earn large bonuses for achieving either speed or accuracy.
The inability of public managers to capture surplus revenues for their own use alters the pattern of incentives at work in government agencies. Beyond a certain point additional effort does not produce additional earnings. (In this country, Congress from time to time has authorized higher salaries for senior bureaucrats but then put a cap on actual payments to them so that the pay increases were never received. This was done to insure that no bureaucrat would earn more than members of
Congress at a time when those members were unwilling to accept the political costs of raising their own salaries. As a result, the pay differential between the top bureaucratic rank and those just below it nearly vanished.) If political constraints reduce the marginal effect of money incentives, then the relative importance of other, nonmonetary incentives will increase. . . .
That bureaucratic performance in most government agencies cannot be linked to monetary benefits is not the whole explanation for the difference between public and private management. There are many examples of private organizations whose members cannot appropriate money surpluses for their own benefit. Private schools ordinarily are run on a nonprofit basis. Neither the headmaster nor the teachers share in the profit of these schools; indeed, most such schools earn no profit at all and instead struggle to keep afloat by soliciting contributions from friends and alumni. Nevertheless, the evidence is quite clear that on the average, private schools, both secular and denominational, do a better job than public ones in educating children. [Professors Note: The evidence for this last statement is a subject of considerable controversy.] Moreover, as political scientists John Chubb and Terry Moe have pointed out, they do a better job while employing fewer managers. Some other factors are at work. One is the freedom an organization has to acquire and use labor and capital.
Acquiring and Using the Factors of Production
A business firm acquires capital by retaining earnings, borrowing money, or selling shares of ownership; a government agency (with some exceptions) acquires capital by persuading a legislature to appropriate it. A business firm hires, promotes, demotes, and fires personnel with considerable though not perfect freedom; a federal government agency is told by Congress how many persons it can hire and at what rate of pay, by the Office of Personnel Management (OPM) what rules it must follow in selecting and assigning personnel, by the Office of Management and Budget (OMB) how many persons of each rank it may employ, by the Merit Systems Protection Board (MSPB) what procedures it must follow in demoting or discharging personnel, and by the courts whether it has faithfully followed the rules of Congress, OPM, OMB, and MSPB. A business firm purchases goods and services by internally defined procedures (including those that allow it to buy from someone other than the lowest bidder if a more expensive vendor seems more reliable), or to skip the bidding procedure altogether in favor of direct negotiations; a government agency must purchase much of what it uses by formally advertising for bids, accepting the lowest, and keeping the vendor at arm's length. [Professor’s note: This last statement about competitive bidding was written before there were no-bid contracts.]
When a business firm develops a good working relationship with a contractor, it often uses that vendor repeatedly without looking for a new one; when a government agency has a satisfactory relationship with a contractor, ordinarily it cannot use the vendor again without putting a new project out for a fresh set of bids. When a business firm finds that certain offices or factories are no longer economical it will close or combine them; when a government agency wishes to shut down a local office or military base often it must get the permission of the legislature (even when formal permission is not necessary, informal consultation is). When a business firm draws up its annual budget each expenditure item can be reviewed as a discretionary amount (except for legally mandated payments of taxes to government and interest to banks and bondholders); when a government agency makes up its budget many of the detailed expenditure items are mandated by the legislature.
All these complexities of doing business in or with the government are well-known to citizens and firms. These complexities in hiring, pur chasing, contracting, and budgeting often are said to be the result of the "bureaucracy's love of red tape." But few, if any, of the rules producing this complexity would have been generated by the bureaucracy if left to its own devices, and many are as cordially disliked by the bureaucrats as by their clients. These rules have been imposed on the agencies by external actors, chiefly the legislature. They are not bureaucratic rules but political ones. In principle the legislature could allow the Social Se curity Administration, the Defense Department, or the New York City public school system to follow the same rules as IBM, General Electric, or Harvard University. In practice they could not. The reason is politics, or more precisely, democratic politics.
* * *
Public versus Private Management
What distinguishes public from private organizations is neither their size nor their desire to "plan" (that is, control) their environments but rather the rules under which they acquire and use capital and labor. General Motors acquires capital by selling shares, issuing bonds, or retaining earnings; the Department of Defense acquires it from an annual appropriation by Congress. GM opens and closes plants, subject to certain government regulations, at its own discretion; DOD opens and closes military bases under the watchful guidance of Congress. GM pays its managers with salaries it sets and bonuses tied to its earnings; DOD pays its managers with salaries set by Congress and bonuses (if any) that have no connection with organizational performance. The number of workers in GM is determined by its level of production; the number in DOD by legislation and civil-service rules.
What all this means can be seen by returning to the Registry of Motor Vehicles and McDonald's. Suppose you were just appointed head of the Watertown office of the Registry and you wanted to improve service there so that it more nearly approximated the service at McDonald's. Better service might well require spending more money (on clerks, equipment, and buildings). Why should your political superiors give you that money? It is a cost to them if it requires either higher taxes or taking funds from another agency; offsetting these real and immediate costs are dubious and postponed benefits. If lines become shorter and clients be come happier, no legislator will benefit. There may be fewer complaints, but complaints are episodic and have little effect on the career of any given legislator. By contrast, shorter lines and faster service at Mc Donald's means more customers can be served per hour and thus more money can be earned per hour. A McDonald's manager can estimate the marginal product of the last dollar he or she spends on improving service; the Registry manager can generate no tangible return on any expenditure he or she makes and thus cannot easily justify the expenditure. Improving service at the Registry may require replacing slow or surly workers with quick and pleasant ones. But you, the manager, can neither hire nor fire them at will. You look enviously at the McDonald's manager who regularly and with little notice replaces poor workers with better ones. Alternatively, you may wish to mount an extensive training program (perhaps creating a Registration University to match McDonald' s Hamburger University) that would imbue a culture of service in your employees. But unless the Registry were so large an agency that the legislature would neither notice nor care about funds spent for this purpose-and it is not that large-you would have a tough time convincing anybody that this was not a wasteful expenditure on a frill project.
If somehow your efforts succeed in making Registry clients happier, you can take vicarious pleasure in it; in the unlikely event a client seeks you out to thank you for those efforts, you can bask in a moment's worth of glory. Your colleague at McDonald's who manages to make customers happier may also derive some vicarious satisfaction from the improvement but in addition he or she will earn more money owing to an increase in sales.
In time it will dawn on you that if you improve service too much, clients will start coming to the Watertown office instead of going to the Boston office. As a result, the lines you succeeded in shortening will become longer again. If you wish to keep complaints down, you will have to spend even more on the Watertown office. But if it was hard to persuade the legislature to do that in the past, it is impossible now. Why should the taxpayer be asked to spend more on Watertown when the Boston office, fully staffed (naturally, no one was laid off when the clients disappeared), has no lines at all? From the legislature's point of view the correct level of expenditure is not that which makes one office better than another but that which produces an equal amount of discontent in all offices.
Finally, you remember that your clients have no choice: The Registry offers a monopoly service. It and only it supplies drivers' licenses. In the long run all that matters is that there are not "too many" complaints to the legislature about service. Unlike McDonald's, the Registry need not fear that its clients will take their business to Burger King or to Wendy's. Perhaps you should just relax.
If this were all there is to public management it would be an activity that quickly and inevitably produces cynicism among its practitioners. But this is not the whole story. For one thing, public agencies differ in the kinds of problems they face. For another, many public managers try hard to do a good job even though they face these difficult constraints.
DISCUSSION QUESTION: What are the advantages and disadvantages of trying to run the government more like a business?
--