Federal regulatory action–and inaction–often generates top headlines: “Citizen Vigilance Leads to Toy Recalls” “Report Faults FDA on Drug Safety” “EPA Denies California’s Right to Mandate Emissions.” Federal agencies sit uneasily in our system of separated but overlapping branches of government, but they arguably do far more lawmaking than Congress and much more judging than the courts. CQ Weekly and the Congressional Record, for example, have reported that Congress passed nearly two dozen majorstatutes and more than 100 other public laws in 2001, respectively; by contrast, according to the Government Accountability Office (GAO), that year cabinet departments, the Executive Office of the President, and independent agencies promulgated 70 major rules and nearly 3,500 other regulations. Law professor Judith Resnik has calculated that the federal judicial branch conducted approximately 85,000 adversarial proceedings, including trials, in 2001; in the same period, administrative agencies engaged in over 700,000 such proceedings, including immigration and Social Security disputes.

Especially in partisan times, politicians and the press often label
all agency rulemaking and adjudication as either desirable or
troubling. In simple terms, they imply that liberals like regulation,
while conservatives abhor it. But the reality of regulation is much
more complicated. Much depends on who controls regulatory authority,
and how they exercise that power. Not surprisingly, debates over agency
behavior occupy scholars in political science departments, public
policy schools, and law schools. But the regulatory system as a whole,
tied to considerable real-world consequences, rarely draws the
attention of policymakers, including those who would like to see more
stringent federal regulations. As we approach the presidential
election–and a new administration and agency leadership–it’s time to
consider the benefits and problems of different mechanisms of
regulation. Progressives must devote real attention and political
capital to ensuring that agencies employ methods that are as efficient,
effective, and democratically legitimate as possible.

To libertarians and free-market conservatives, the popular way of
viewing the federal administrative state is through public choice
theory, which looks at our vast bureaucratic apparatus and sees “iron triangles” everywhere. To them, agencies, Congress, and organized,
narrow interest groups form powerful bonds, peddling regulatory
initiatives that serve targeted interests at the expense of broader
social well-being. Accountability and the generation of social welfare
are nowhere to be seen.

In Regulation and Public Interests: The Possibility of Good Regulatory Government, Steven Croley, a political scientist and legal scholar at the University of Michigan, offers a more optimistic vision. He conceives of agencies as potentially autonomous institutions that can produce desirable regulatory outcomes, like the ozone and particulate matter rules issued in 1997 under the Clean Air Act. In doing so, he tears down the public choice account of the administrative state and builds an alternative theory of public interested regulation that depends on administrative process. The former effort is formidable, while the latter is admirable, if not completely successful.

In refreshingly readable prose–he is, after all, writing about
regulatory theory–Croley carefully dissects the assumptions and
empirical validity of the public choice worldview. In its most
concentrated form, the theory is an easy target. Drawing on
considerable academic work, Croley hones his attack on public choice’s
premises concerning interest group formation, including that “interest
groups seek regulatory decisions that advance the selfish interests of
their members.” As he demonstrates, while the public choice framework
might explain the power of, say, the oil industry, it has never been
especially good at explaining the creation, maintenance, and regulatory
reach achieved by environmental and consumer groups like the Sierra
Club and Public Citizen for broad social goals.

Croley also questions public choice theory’s predictive power by
looking at several testable propositions: that organized, narrow
interests manipulate agencies and Congress more successfully than other
groups; that the negotiations among these narrow interests, agencies,
and Congress will not be transparent; that agency decisions will
generate minimal public criticism; and that few interests will
participate in regulatory decisionmaking processes. Through a series of
case studies, Croley undermines these propositions, though larger-scale
empirical work might ultimately prove more compelling. His cases
examine, among other topics, the Environmental Protection Agency’s
(EPA) ozone and particulate matter regulations, the Food and Drug
Administration’s (FDA) tobacco restrictions (later struck down by the
Supreme Court), and the Forest Service’s roadless rule for national
forests (subsequently pulled by President George W. Bush’s
administration). In all of these, Croley shows that narrow interest
groups did not, in fact, drive the regulatory outcomes, that
negotiations were remarkably transparent, that the regulatory process
generated considerable debate, and that many interests took part.
Croley is at his best when detailing how a wide range of interests
fights for specific policies in the administrative state. By the end,
we see a very different picture from the one public choice theory
paints. Far from involving a small subset of the polity behind closed
doors, the administrative process is fairly transparent and open to a
diverse set of groups and individuals.

Croley’s book is the latest in a long line of scholarly works
critiquing the theoretical assumptions and methodological commitments
of public choice. Regulation and Public Interests is more than just an
attack, however. The book comprehensively develops an alternative
theory of the administrative state, one in which agencies wield their
authority through various administrative procedures that the public
choice account largely ignores–such as the promulgation of rules after
the public has notice of a regulatory action and an opportunity to
comment (termed “notice and comment rulemaking”), public hearings, and
detailed explanations of decisions–to achieve regulatory outcomes
Croley considers desirable. Whereas public choice theorists see
agencies as dominated by Congress and interest groups, Croley sees them
as much more autonomous and able to push through “public interested
regulation” (a term whose definition shifts somewhat in the book).

Croley’s approach, alternately labeled “public interested
regulation” and “administrative process,” rests on the premise that
agency regulators often act to accomplish “some conception of the
public interest.” But it also assumes that laws governing agencies
leave Congress with little influence over the process, and that
presidential oversight and judicial review of actions further diminish
Congress’s role. Administrative procedures also allow smaller groups to
participate along with more powerful groups in the regulatory process.
As a result, this procedural framework permits agencies to enact
“socially desirable regulatory outcomes.” In other words, agencies
dominate, a view in direct opposition to the public choice account.
Croley spends the rest of his book demonstrating how agencies can enact
“good” regulatory outcomes in the face of a “bad” Congress. The
government may not be perfect, but regulatory agencies are the ones
wearing the white hats.

In a matchup between Croley’s public interested regulation theory
and public choice theory, Croley’s story looks exceptionally promising.
But in focusing exclusively on criticism of public choice theory,
Croley ignores a critical question: Who actually controls the
government? Another academic school–the political
institutionalists–argues that agencies are constrained not just by
Congress, but by the White House and possibly the courts. The policy
preferences of the White House and Congress (sometimes unified,
sometimes competing) shape regulatory outcomes, often dramatically.
Croley’s core claims for his alternative theory are compelling, until
it becomes clear that he provides no substantial account for who
actually controls the government. This peculiar silence seems at odds
with recent events and Croley’s own case studies.

Croley relegates the White House to a bit part, one in which the
political leadership of the executive branch mostly weakens any
interference by Congress in agency decision-making. Tellingly, all the
major case studies of socially desirable regulation that Croley
marshals involve executive agencies in President Bill Clinton’s
administration. His smaller examples involve both Democratic and
Republican administrations, but many concern independent regulatory
commissions, such as the Securities and Exchange Commission (SEC),
which arguably have more independence from both Congress and the
President than do executive agencies. No example is provided of
desirable new regulatory activity by a classic executive agency (led by
an appointee who serves entirely at the pleasure of the president and
is not appointed to a fixed term) in a Republican administration.

That is not to say that executive agencies under a Republican
president do not make efficient, fair, and democratically legitimate
regulatory decisions (as they sometimes do). It is merely to say that
political control of the legislature and the White House is a necessary
factor to consider, one that Croley underemphasizes. This is where
political institutions theory, to which I largely subscribe, comes in
handy, insisting that the key to understanding policy outcomes is found
in the interaction among the branches, mainly between Congress and the
White House.

In Croley’s telling, Congress is often detrimental to public
interested regulation, and agencies are often the good guys. But even
for those who share his generally progressive view of what is in the
public interest, things are not so clear. Since January 2007, Congress,
and not federal agencies, arguably has been most concerned with
protecting social interests in our administrative state, struggling
against agencies bent on promoting narrower interests. In recent
months, Congress has, for instance, tried to revamp and strengthen the
Consumer Product Safety Commission (CPSC) in the face of opposition
from the White House and the agency’s own leadership. Even though the
agency admittedly cannot meet growing demands for inspections
(especially of products coming from abroad), the Commission’s acting
chair explicitly asked Congress not to give the agency more money. This
is an example not of an agency promoting the public good, but of one
driven by the political pressures of an anti-regulatory White House to
damage it.

What does this all mean in the real world? Croley’s provocative
account justly focuses attention on two aspects of the administrative
state that deserve wider consideration, especially as we approach the
2008 election: agency leaders and agency procedures. To Croley, many
agency leaders care deeply about social welfare, and agency procedures
provide important opportunities for transparency and public
participation in the regulatory system. He is right that good agency
leaders and agency procedures are necessary for efficient, fair, and
legitimate regulatory outcomes. Croley misses, however, some of the
complex realities of federal institutions, and thus overlooks the ways
in which progressive policymakers can work to make them more
accountable and effective.

For one, as the performance of the Federal Emergency Management
Agency (FEMA) during Hurricane Katrina made clear, we need qualified
and trained agency officials who stick around government long enough to
push through desirable regulatory policies (a concern Croley
essentially dismisses). We know surprisingly little about who serves in
top executive agency positions across a range of recent
administrations, including the current one, other than information such
executives have voluntarily submitted to the Senate and academic
surveys. We need to know if individuals who serve in key positions in
the administrative state are qualified and accountable, and if not, we
need to develop ways to make sure they are.

Congress could enact additional statutory requirements for
particular positions (for instance, experience requirements for a
specific office). In the aftermath of Hurricane Katrina, for example,
Congress mandated that the head of the FEMA have emergency management
and executive leadership experience. The constitutionality of such
requirements, particularly for core executive positions, is not
entirely clear. But it’s worth pushing for them in many circumstances.
The next President can also improve the quality of agency leaders he or
she selects by investing more time during the presidential transition
to considering non-cabinet level positions. In the meantime, we can
press presidential candidates to promise that they will appoint
qualified leaders to agencies and to say now whom they might nominate
so voters and the media can assess them.

Assuming we can get qualified individuals into important agency
positions, they also need more comprehensive training, not only about
their agencies, administrative law, congressional oversight, and White
House regulatory review procedures, but also about media relations,
management, and other executive skills. Many members of Congress attend
intensive orientation sessions; so too should top-level bureaucrats.

Qualified, trained agency leaders then need to stay put to engage in
meaningful regulatory activity. Their average tenure is far shorter
than the tenure of members of Congress or the federal judiciary,
despite their extensive lawmaking and adjudicatory responsibilities.
Research by political scientists Roger Davidson and Walter Oleszek has
found that, as of the start of the 108th Congress, the average member
had spent almost 11 years in the House of Representatives and the
average senator had provided almost 16 years of service. Federal judges
enjoy life tenure. The average tenures of Supreme Court Justices who
retired between 1970 and 2004 and lower court federal judges who
retired between 1983 and 2003 were, legal studies show, 25.6 years and
20 to 25 years, respectively. On the other hand, agency leaders seem to
stay for “a social season and a half and then leave,” as one Eisenhower
staff member quipped. Looking at all Senate-confirmed agency appointees
from October 1981 to September 1991, the GAO found that the median
appointee tenure was 2.1 years. Researchers at the RAND Corporation
determined that from 1949 to 1999 most top-level appointees in the
Defense Department served only 11 to 20 months. To encourage agency
leaders to remain in government for more than two years, we should be
open to a range of reforms, including increasing salaries and benefits,
streamlining the nomination and confirmation process, and, in some
instances, giving agency heads more policy authority.

Leadership is just one necessary component of a desirable
bureaucracy; agency operating procedures are another. In Croley’s view,
agency procedures such as notice and comment rulemaking are paragons of
democratic legitimacy, particularly when contrasted with the relatively
opaque communications between Congress and interest groups (as he sees
them). But just because such tools exist doesn’t mean they get used. In
fact, in recent administrations some agencies have increasingly enacted
policies without providing prior notice and opportunity to comment at
all. They instead turn to such items as guidance documents and interim
and direct final rulemaking, which skip traditional notice and comment
procedures at the front end, or to informal adjudications. The GAO
estimates that approximately half of the final regulatory actions
listed in the Federal Register in 1997 were completed without prior
notice and comment. Public policy professor Stuart Shapiro has found
that about 40 percent of rules promulgated in November and December
2003 were direct or interim final rules. Statutory deadlines for prompt
agency action seem to encourage agencies to forgo notice and comment
procedures. Sometimes, skipping notice and comment is justified (for
emergency reasons or for routine announcements); sometimes it is not.
But the result is the same: less public input, and an increasingly
opaque agency rulemaking process.

Of course, some agency procedures that increase outside involvement
may even be detrimental to good regulatory governance. The Data Quality
Act, two sentences snuck into an appropriations bill in 2000, allows
interest groups to challenge the reliability of agency information. In
a 2004 series, the Washington Post demonstrated how industry groups
have become skilled at using the Act to introduce “uncertainty” into
scientific decisions by agencies to the detriment of larger social
concerns. For instance, the makers of a popular herbicide called
atrazine–which scientific research has linked to cancer and hormonal
disruption in animals–escaped regulation because the Act allowed them
to introduce doubt about the scientific community’s findings. White
House regulatory review, as recently strengthened by Executive Order
13,422, also provides opportunities for arguably beneficial regulatory
initiatives to be halted. In May, the National Oceanic and Atmospheric
Administration (NOAA) submitted its proposed rule to limit the fishing
of krill, a marine species and important food source for whales and
other animals in the Pacific Ocean. But the Office of Information and
Regulatory Affairs (OIRA), which leads the White House regulatory
review process, rejected the rule in November. The same executive order
also requires the agency regulatory policy officers who coordinate with
OIRA on regulatory review to be political appointees. This new
provision undermines transparency and good regulatory governance, and
it should be repealed. Instead, we need regulatory procedures that
encourage public participation and foster transparency (while balancing
their ability to delay regulatory outcomes), and we must shy away from
policymaking mechanisms that bar prior public comment unless they are
truly needed.

We cannot lose sight of the immense policy repercussions of agency
regulatory activity and inactivity, and the ways in which intellectual
developments can help guide policymaking. Nor can we wait until we find
the perfect theory of regulation to act. Because no matter what happens
in November 2008, the administrative state will shift, perhaps
considerably, opening the possibility for real change in the way the
government constructs, staffs, and monitors regulatory agencies. And it
is then that we can–and must–push for reforms to improve regulatory
governance, the often overlooked, but critical part of creating the
rules that govern our country.