In the months since the Supreme
Court, in INS v. Chadha, struck down the "legislative veto" device in broad terms,
commentary has focused on the resulting crisis in relations between the legislative and executive branches and on various
possible solutions to the crisis. I believe there is no crisis and no need for precipitate solutions. I want to argue instead
that the entire debate concerning legislative oversight of executive functions, both before and after Chadha, really involves a larger and more fundamental problem that cannot be solved through the mechanics of legislative
vetoes or substitute procedures.
The rise of the legislative veto,
at least as applied to domestic spending and regulating, is roughly coincident with the rise of large administrative state.
But this is not just a coincidence. The legislative veto is one of a variety of techniques Congress has used over the past
forty years in an effort to preserve its traditional policymaking authority from erosion—erosion due, ironically, to
the very growth of the size and scope of the federal government.
When Congress considers whether
to expand the federal government's reach into some field previously the domain of the states or private arrangements—highways,
education, medical care, automobile design, and so forth—its approach is to debate the merits of the individual issue
before it, and then, almost always, to resolve the debate in favor of expansion. Congress almost never seriously debates,
much less resolves, the larger problem economists refer to as diseconomies of scale. The term refers to the inherent limits
on the size and range of activities any one organization can undertake efficiently, regardless of the abstract merits of the
activities taken individually, and regardless of the energies and good intentions of the people in charge. Over the past several
decades the number of activities the federal government has taken responsibility for has grown fabulously, while the government
itself is still directed by 537 human beings, and is still the cumbersome and inefficient organization the Founders designed
it to be.
Not only the size of the federal
government but the scope and complexity of its undertakings have increased dramatically. The terms of what may be called our
Grand Political Accommodation in the U.S. are that (a) outright socialization is forbidden; while (b) differential subsidization,
taxation, and regulation are unrestricted, and are in fact encouraged at least some of the time by every influential political
quarter. But partial public management of otherwise private markets requires an endless succession of decisions that are highly
technical and detailed, yet are highly contentious—‘political" rather than "proprietary"—and therefore of
great importance to the representatives who direct the government. Administering a system of compulsory medical care insurance
for the aged and poor, and prescribing the design of new industrial processes around the constraint of "minimum feasible"
air or water pollution, are not only larger but more complex than any of the federal government's peacetime enterprises before
the 1930s.
Now a strong case can be made
that the federal government has passed the threshold of diminishing returns to scale and scope—that it has taken on
more numerous and diverse responsibilities than any organization can manage with tolerable efficiency, even one armed with
the coercive powers of the state. Whether or not this is so, however, it seems quite clear that the growth of government has
worked to the relative disadvantage of the legislative branch and the relative advantage of the executive branch. The legislature
is a collegial body of independent individuals representing a profusion of different views and interests; it is best suited
to making occasional broad decisions requiring the definition of a common social consensus. The executive is a hierarchy of
like‑minded individuals accountable to a single boss; it is best suited to making numerous detailed decisions within
a pre‑established policy framework. The growth in the number and complexity of decisions facing the government has played
to the executive's inherent strengths. While both branches have grown in absolute power as new laws have staked out new domestic
territories for the federal government, the executive branch's superior ability to manage these new territories has given
it a larger share of the division of authority to define and direct national policy.
That this is so can be seen in
the fact that Congress, as a concomitant of its approval of larger and more complex government, has been obliged to delegate
increasing legal and policymaking authority to the executive branch. With the government intervening more and more deeply
into private markets, Congress increasingly has lacked the resources—chiefly time and information—necessary to
enact in law all of the discrete judgments and compromises necessary to guide these interventions. So it has increasingly
fudged—enacting vague and often flatly contradictory statutory standards that have effectively transformed executive
officials (and derivatively judges) into de facto lawmakers. The executive branch, for its part, has responded with brilliant
policymaking innovations that have at once demonstrated its superior versatility in managing the large complex state and encouraged
further legislative delegations. The greatest of these is "informal rulemaking," which subtly combines the efficiency of executive
decision-making with the key legitimating features of judicial and legislative decision-making—due process and public
sanction.
None of Congress's own policymaking
innovations have come close, not even the "legislative veto." It is important to recognize, however, that this is what the
legislative veto was: a means of holding onto a part of the authority delegated to the executive; of avoiding the hard, time‑consuming,
and sometimes impossible compromises of legislating; and of importing some of the efficiencies of executive decision-making
into the legislature. Of course, to describe the legislative veto in this way is not to approve it as a constitutional matter,
as Mr. Justice White appeared to in his dissent in Chadha. It may be that the Constitution
does not, by its terms, forbid the giant modern state—but this hardly means that the terms the Founders did agree on
must give way to the administrative convenience of that state.
The legislative veto is, however,
only one of several devices Congress has fashioned to retain day‑to‑day policymaking authority. Indeed, legislative
veto provisions were in hundreds of statutes for decades, and over consistent presidential opposition, before any were challenged
in definitive cases before the Supreme Court in 1983, suggesting the relative unimportance of the device. Where large stakes
have been involved, Congress has been more likely to rely on other means of influencing or controlling the Administration
in office—appropriations riders requiring or forbidding agencies to undertake certain actions; informal agreements between
committee chairmen and executive officials; and, of course, direct legislative nullification of executive actions, as in cases
of the saccharin ban and the automobile seatbelt‑ignition interlock regulation. The legislative veto stands out from
these and other devices of legislative control only in that it stepped clearly over the constitutional foul line.
Since Chadha, many observers have expressed the hope that Congress will now become "more responsible" and begin making
the tough legislative choices it avoided under cover of the legislative veto. The analysis above suggests that this is a vain
hope. The problem of modern lawmaking has nothing to do with individual legislators avoiding their responsibilities. It is
rather an institutional problem, inherent in the very size and ambitions of modern government and the incorrigible cumbersomeness
of legislative decision-making. In the future, Congress will probably rely more heavily on appropriations riders and similar
techniques, which will probably re‑centralize legislative authority somewhat in the hands of the committee chairmen
and party leaders. (The legislative veto was always a backbencher's idea, widely opposed by influential committee chairmen
and party leaders in both Houses.) But Congress will probably not write "better" laws or take back large chunks of statutory
discretion from the executive branch, so long as it is under such pressure to write and finance so many laws. It will probably
not even write more detailed laws: the Clear Air Act, the reigning paradigm of highly detailed legislative regulation, has
now mired Congress in a thirteen‑year‑long drafting marathon, leaving the drafters in a dispirited state (along
with those who try to interpret, enforce, and obey the results).
Over the past decade, as Congress
was adding legislative veto provisions to more and more regulatory statutes, and considering enacting a generic law to subject
all regulatory decisions to legislative vetoes, the executive and judicial branches were developing their own programs for
increased regulatory oversight. Presidents Ford, Carter, and Reagan issued increasingly rigorous orders requiring the executive
agencies to assess the benefits and costs of their regulations and submit them for review by the Executive Office of the President.
The courts strengthened their own standards of review far beyond the original conception of the "arbitrary or capricious"
standard of the Administrative Procedure Act. In spite of occasional controversies, Congress has generally supported these
efforts, sometimes for the record. The "Regulatory Reform Act of 1982," which passed the Senate unanimously, contained provisions
to strengthen legislative oversight of the regulatory process (a blanket legislative veto) and executive oversight (benefit‑cost
analysis and review by the President's office) and judicial oversight (a "Bumpers Amendment" to the APA). But the remarkable
thing is how far the executive and judicial branches have succeeded unilaterally in tightening management of regulatory decision-making.
The successive executive order
review programs, culminating in President Reagan's Executive Order 12291, are especially notable here. The benefit‑cost
analysis requirements have fallen well short of the regulatory reformers' wild hopes that economics would somehow drive politics
from the regulatory field. But they have imposed a modicum of discipline on the regulatory bureaucracies, narrowing somewhat
the freewheeling discretion afforded by most regulatory statutes. By making regulatory decisions more factual and comprehensible,
the programs have made them more vulnerable to control not only by presidents but by judges and legislators as well. Indeed
the executive order programs, by loading the rulemaking files with detailed empirical analyses and obliging the agencies to
explain their decisions in terms of these analyses, have probably done as much to strengthen judicial review as any of the
judiciary's own doctrinal innovations. Virtually all of the critical court decisions in recent regulatory cases, such as the
Supreme Court's Benzene, Cotton Dust, and Passive Restraints decisions, have turned decisively on economic and statistical
issues that were in the record in the first place because of the executive order programs.
Which is to say that the executive
order programs are one more example of the executive branch's comparative advantage in directing the large administrative
state. The Chadha decision will give somewhat greater impetus to these programs,
although senior congressional committee chairmen will oppose the trend just as they opposed the legislative veto itself. It
is possible, however, that Congress will buck the trend directly. This is suggested by the breathtaking proposal of Congressman
Levitas and others: that Congress require that regulations be affirmatively approved by statute before they can take effect.
This would, of course, avoid the constitutional problems of the legislative veto. It would also oust the courts entirely of
review of regulatory decisions (except on constitutional grounds), and thereby put an end to most of "administrative law"
itself! But it would do so at the cost of swamping Congress with thousands of additional, highly detailed legislative decisions
each year, which would surely be unmanageable and lead to procedures for "approving" agency rules in large batches.
Perhaps a middle ground can be
devised. Congress and the President might be required to affirm only the several dozen most important rules each year (just
as Executive Order 12291 singles out for special treatment "major" rules imposing compliance costs of $100 million or more).
Or statutory approval might be required only of the rules of the so‑called "independent" regulatory agencies, thus giving
the President as well as Congress responsibility for the policies of these agencies. Most presidents would support such procedures
as a general proposition, just as most legislators have supported the executive order programs as a general proposition. In
all events, the Levitas proposal puts the dilemma of congressional control of the large modern state in the most exquisite
possible light: the control is Congress' for the taking, limited not by the Constitution but by the nature of the legislative
process.
