Jeremy Rabkin has written a fine essay about the continuing relevance of James Q. Wilson’s 1989 book Bureaucracy: What Government Agencies Do and Why They Do It. I have been fortunate enough to benefit from Wilson’s analysis in my own writing on the Justice Department’s Office of the Solicitor General. His framework showed why the craft values of a bureaucracy often resist the President’s jurisprudential positions.

But I want to use this brief essay to discuss a few issues that I do not believe Wilson sufficiently stresses, and to suggest corrective legislation that might be passed to rectify these distinctive problems of bureaucracy. The first is that the federal bureaucracy stands, as a general matter, not only to the left of the Republican Party, but to the left of the Democratic Party. Wilson is aware of the political bias, but he seems to believe that with a few exceptions, it does not have substantial effects. I cannot agree.

To be sure, Wilson is right that organizational interests often predominate over political ones in routine matters: there is no conservative or liberal way of picking up garbage. And he is right as well that in some agencies, bureaucrats are less liberal than others. The Defense Department employees are far less liberal than those at the Environmental Protection Agency. They may even lean right. Indeed, political views may differ as between the divisions within a department. When I was at the Department of Justice, the career attorneys were predominantly liberal but in widely varying degrees. The attorneys in the Civil Rights and Environment and Natural Resources Divisions stood substantially to left of those in the Criminal and Civil Divisions. This difference is not surprising. The very titles of the former can be understood as political causes as well as government responsibilities, and left-liberals self-select to be part of them.

But my experience casts doubt on Wilson’s apparent view that ideological influence is to be feared largely in new agencies where new jobs are defined with little specificity. The Civil Rights Division and Environment and Natural Resources Division have been around for a long time, but new circumstances arise with new statutes, like the Civil Rights Act of 1991 or the American with Disabilities Act, as do new situations, like the focus on carbon emissions. The policy and litigation world is always being reborn, and left-liberal lawyers are delighted to try to bend the open arc of history their way within the ample bounds of their professional obligations. I suspect the same is true in other agencies that can be defined in terms of their contributions to progressive causes. The EPA and National Labor Relations Board come to mind.

And the recent scandals at the IRS that Rabkin mentions may well show that political beliefs even influence agencies that would have been thought apolitical. Again, it was a new situation that prompted political overreach. Democratic Senators wrote a stern letter to the IRS calling for scrutiny of conservative groups that got tax-exempt status as 501(c)(4) organizations. The head of the IRS division charged with policing such groups appears to have been left of center, and the IRS employees broadly speaking contribute almost entirely to Democrats. At the very least, these civil servants’ ideological leanings may explain why the targeting of Tea Party groups for scrutiny did not set off the alarms bells it would have had their leanings been to the right and their partisan affiliations been mostly Republican.

Second, I think Wilson does not focus sufficiently on where the interests of bureaucrats are most likely to influence their approach to regulation. One should not neglect their interest in increasing the value of their outside options—the income they will earn if, as some do, they come to work in the private sector. Wilson rightly downplays the likelihood that bureaucrats will do favors for outside groups in the hopes of increasing their employment opportunities. He notes that private firms value future competence, not past services. But bureaucrats do have an interest in increasing their value in the future by changing the regulatory environment in which the regulated firms work. The more heavily regulated a sector, the more valuable will be their advice as lawyers or as consultants if they should go to work in that sector. Moreover, the less clear the regulations, the more valuable will be their services. They will have the insider’s knowledge of how regulations are likely to be interpreted. The latter source of value has generally grown with the Supreme Court’s adoption of so-called Auer deference, in which the judiciary is supposed to defer to any plausible interpretation by an agency of its own regulations. The result is greater discretion for the bureaucracy and more income for those who can predict how that discretion will be exercised.

In this way, the personal interests of bureaucrats reinforce left-liberal ideology. And left-liberals distrust the market more than do conservatives, see more market failures, and have greater confidence in government’s capacity to fix them. Beliefs and interests alike incline bureaucrats to overregulation. Moreover, an important part of the Progressive paradigm is the need for executive discretion. On this view, legislation is not flexible enough to address the ever-changing need for government regulation. That is the original Progressive rationale for transferring rule-changing discretion from congressional legislation to agency rulemaking. So-called Auer deference is now transferring that discretion to agency interpretation of rules. Bureaucratic beliefs are also aligned with this troubling development.

A final addition I would make to the problem of bureaucracy is the substantial agency costs that the President faces even with his own appointees, costs that are attributable in part to the power of bureaucracy. Wilson emphasizes that agencies are influenced as much or more by Congress as they are by the President. Wilson also recognizes correctly that the President has difficulty making sure that those he appoints to Cabinet and subcabinet posts will loyally carry out his program. This ideological slack is certainly part of the agency costs. But the President faces agency costs even from loyal appointees, because given the power of the career bureaucracy, appointees may often have to compromise to get anything done. An official who is out of step with what his agency wants can often face a stream of unflattering press stories based on sources in the agency. I saw that happen to Attorney General Edwin Meese, an appointee committed entirely to the Reagan program. Given the beliefs and interests of bureaucrats, the compromises that most Republican appointees make are much more damaging to the President’s program than those that Democratic appointees need to make.

I would also emphasize that the President’s formal authority to fire people who do defy his aims either from ideological conviction or because they are unwilling to risk reputational harm is a very weak force to pit against these problems. Firing someone publicly has very large costs: it is treated as an admission that it was a mistake to hire the appointee in the first place. Even getting people to go quietly, if it can be done, obliges the President to find a nominee who can go through the confirmation process again. This process has not insignificant costs, because of the leverage it gives Senators, even if the Senate is under the control of the President’s party. The costs are much higher if the opposition party holds the Senate. These costs endow presidential appointees with a great deal of discretion. They can act contrary to the President’s interests whenever the costs of firing them exceed the costs of letting them have their way.

What can the friends of liberty do to correct this sad problem of the statist influence of bureaucracy? The first order of business is to limit the scope and size government. Bureaucracy will less easily move the world to the Left if there are fewer bureaucrats. But classical liberalism is not a brief for anarchy. Here are four recommendations to circumscribe the untoward influence of bureaucracy within the governance we need.

1. Create more political appointees within departments. Generally, the President nominates and the Senate confirms appointees as far down as the assistant secretary level. But most officials below that level, even deputy assistant secretaries, are usually career bureaucrats who often wield substantial power. A partial solution to this problem would be to double the number of so-called “Schedule C” appointees who are appointed by the Cabinet members but are vetted by the White House. Having more Schedule C appointees has two other advantages for the President’s program: Because Senate confirmation is not involved, firing appointees who have “gone native” and putting in better ones is less complicated. Second, because Schedule C appointees are junior, they look forward to being able to work in future administrations of their party and thus are more likely to hew to the party’s aims. In contrast, Cabinet officials have a “last period” problem. They are unlikely to ever work again in government, and have no particular reason to stay on the reservation. This reform would help Presidents of both parties to press their agendas, even if it would help Republican Presidents more, since they face more obstruction from typically left-leaning career bureaucrats.

2. Create units of deregulation with each department. As my co-blogger Mike Rappaport has suggested, agencies lack any bureaucratic structure that is dedicated to deregulation. For the reasons discussed above, bureaucrats by belief and interest will want to expand not shrink regulation. It is necessary for Congress to establish actual offices to accomplish this task along the lines of the Inspector General. Simply telling bureaucrats to deregulate is unlikely to get the President very far given the culture of bureaucracy. President Obama’s order for agencies to evaluate the effects of previous regulations and eliminate the ineffective ones has brought little action. But there is hope that a dedicated unit for deregulation would have a different culture from the rest of the agency and attract individuals who like the idea of deregulation.
3. Pass some version of the Regulations from the Executive in Need of Scrutiny (REINS) Act. It would require Congress, not the executive, to enact any regulation that would impose a substantial cost on the economy—say more than $1 or $2 billion. Under this regime, the executive branch agency would do the analytical work and then recommend a rule to Congress, which would be required, under its own rules, to take an up-or-down vote. While I do not want to endorse any particular version of such legislation, it is important to force Congress to deliberate on at least the most important rules enacted by the federal government each year.

These debates would bring more open deliberation about matters of fundamental import to society, thereby enhancing the democratic accountability of the administrative state. Even if such roll-call votes on regulations were relatively few in number, they would provide feedback to the executive agencies on how their discretion is best exercised to accommodate the popular will. These votes would also curb excesses in bureaucratic discretion. The most important rules—those that cost society billions of dollars—will have been vetted by Congress. To be sure, even these rules may still be vague or ambiguous enough to be exploited by bureaucrats, but the process would at least give the proponents of limited government the opportunity to persuade their colleagues and the public that this or that regulation was more disadvantageous than advantageous.

4. Eliminate Auer The extreme deference that the judiciary pays to agencies in interpreting of their own rules also maximizes bureaucratic power and encourages vague rules, which, as we have seen, are already in the interest of bureaucrats. Fortunately, some justices on the Court have already been critical of Auer deference. Agencies are given the discretion to translate Congress’s legislation into administrative rules, but this rulemaking entails a notice-and-comment procedure that at least requires them to respond to interested parties and encourages greater care and deliberation. No such notice-and-comment procedure exists for agencies’ interpretations of their own rules, thus unraveling one of the key compromises made in allowing an administrative state to develop. Agencies should not get any deference for interpretations without such safeguards.