Cited: 51 Duke L. J. 901
[*pg 901]
PRESIDENTIAL REVIEW AS CONSTITUTIONAL RESTORATION

JOHN O. MCGINNIS

INTRODUCTION

I. CONSTITUTIONAL DECLINE AND THE RISE OF CENTRALIZED ADMINISTRATION
A. The Decline of Constitutional Federalism
B. The Decline of Tricameralism

II. THE CHANGING WORLD OF POLITICS
A. Political Theory
B. The Executive Orders

III. CONSTITUTIONAL RESTORATION THROUGH PRESIDENTIAL REVIEW
A. Reviving Constitutional Federalism
B. Reviving Tricameralism

IV. RECOMMENDED REVISIONS OF THE FEDERALISM AND REGULATORY REVIEW EXECUTIVE ORDERS
A. Revising the Federalism Executive Order
B. Revising the Regulatory Review Order

CONCLUSION

FOOTNOTES

INTRODUCTION
This Essay mounts a fresh normative defense of the president's exercise of regulatory review authority and his role in enforcing federalism -- responsibilities embodied in executive orders issued by President Ronald Reagan and continued in large measure by President William Jefferson Clinton.1 These executive orders move in some measure toward the restoration of two central principles of the original Constitution -- tricameralism2 (i.e., the combination of bicameralism and the presidential veto) and federalism. The soundness of these orders as a bipartisan, post-New Deal structure of governance derives from their ability to advance the implicit goal of these venerable constitutional principles -- producing government regulation that efficiently promotes the public interest rather than special interests. [*pg 902] These constitutional foundations were in need of repair because the modern administrative state had eroded their effectiveness. This Essay defends presidential review as a step to constitutional restoration and suggests some revisions to these executive orders that will increase their effectiveness in achieving this revival.

Federalism and tricameralism were the cornerstones of the original Constitution. By limiting the federal government to the domestic powers necessary to sustain a free trade zone, the federalism of the original Constitution made states responsible for activities that were best regulated locally, thus encoding a principle of subsidiarity into our polity. Moreover, by forcing states to compete with one another in providing appropriate government regulation, constitutional federalism restrained interest groups and rulers from using state government to serve their interests rather than the public good. Tricameralism -- the requirement that federal legislation be passed by two houses of the legislature and be either approved by the president or passed by a two-thirds vote of both houses over a presidential veto -- also helped assure that the federal government acts for the public good. Tricameralism functioned much like a supermajority rule, which screens public interest legislation from special interest legislation. It also promoted deliberation and consensus, two features that lead to better government regulation overall.

Unfortunately, both of these cornerstones of the Constitution have been gravely weakened in the last century. First, under the modern interpretation of the Commerce Clause, Congress retains essentially plenary power over economic regulation and has used this power to impose regulations on matters that could best be addressed locally. In spite of the recent judicial revival of federalism, the Supreme Court has not restricted the federal government's power in the economic arena.3 Second, tricameralism has declined with the rise of the modern administrative state because, pursuant to congressional delegation, administrative agencies may impose regulations without overcoming the hurdle of legislative passage in two houses and presidential approval. Special interest groups can have substantial influence over administrative regulations, and the regulations may reflect [*pg 903] less broad-based deliberation and consensus than laws and regulations that are passed under the tricameral procedures.

The federalism executive order4 and the regulatory review executive order5 revive the key constitutional principles of federalism and tricameralism and strengthen their objectives for the modern era. Through the federalism order, the president, within the discretion provided by regulatory statutes, can use his authority to revive sensible limitations on the national government by reserving to the states the authority to address regulatory problems within their jurisdictions and the authority to provide distinctive public goods appropriate to their citizens' preferences. In doing so, the president also can revitalize jurisdictional competition among the states to facilitate efficient regulation and restrain special interest groups. Finally, and perhaps most importantly, the order may help revive the decentralized structure that the Framers believed essential to quicken the spirit of self-government within citizens of a continental republic.

It may seem paradoxical that the only official elected by the entire nation can be responsible for enforcing economic federalism. Although this idea may seem superficially surprising, it has deep roots in political economy. The regulatory competition afforded by federalism helps create economic growth, and the president's fortunes (and those of the successor candidate of his party) are more dependent on economic growth than are those of any other federal official.6 In contrast, the regional nature of congressional representation creates a variation of the tragedy of the commons in which congressional representatives seek government action that is advantageous for their state even at the expense of the nation.

For similar reasons, the president has incentives to revive the restraints on lawmaking that the original Constitution imposed through bicameralism and the presidential veto (tricameralism) and that have been eroded by the modern administrative state.7 Regulatory review [*pg 904] by the Office of Information and Regulatory Affairs (OIRA), under presidential direction, compensates in some measure for the decline of tricameralism in the last sixty years. The addition of a layer of review by OIRA compensates for the loss of bicameralism that has occurred through broad delegations of power to administrative agencies. Like bicameralism, the OIRA review process puts an additional screen between government decrees and the citizen by filtering out special interest regulations from those in the public interest. By allowing the president to monitor regulation by the executive branch, the regulatory review order also compensates for the loss of the veto power over government regulation that has resulted from broad delegation. Finally, the substance of the cost-benefit analysis mandated by the regulatory review order8 helps screen public interest regulations from those sought by special interests -- an implicit objective of the original Constitution's requirements for legislative enactments.9

Given this view of the federalism order and the regulatory review order as means toward constitutional restoration, this Essay presents suggestions for strengthening both these orders. For instance, the federalism order should be revised to make more explicit the virtues of regulatory competition and to direct that, within the discretion provided by law,10 the federal government should regulate only when state regulation would be inadequate to address the problem. This perspective necessarily requires the federalism order to be linked to the cost-benefit analysis contemplated by the regulatory review order. This linkage should have institutional advantages as well, because it would provide a more effective enforcement mechanism for the federalism order. Federal agencies would take their federalism responsi- [*pg 905] bilities more seriously if OIRA were directed to reject regulations that fail to comply with the standards of the federalism order.11

Turning to the regulatory review order, this Essay suggests that just as bicameralism and the presidential veto apply to all bills, the regulatory review order should apply to all regulations to screen more comprehensively special interest regulations from public interest regulations. Moreover, because at the margin independent agencies are even more likely to be dominated by special interests than are agencies whose heads are not insulated from presidential removal, the regulatory review order should be extended to independent agencies. Furthermore, although President Clinton substantially restricted the delays that OIRA could impose in the course of its review, the regulatory review order should not impose such restrictions unless the regulatory law at issue can be fairly read to require the implementation of a regulation by a specific date. One of the advantages of bicameralism and the presidential veto lies in delaying legislation until a substantial consensus is reached.

On the substance of cost-benefit analysis, this Essay recommends that the order be revised to take account of the dynamic costs of regulations. Public choice theory has shown that businesses will use regulatory regimes to raise their competitors' costs and that other interest groups will exploit these regimes to obtain benefits at the public's expense.12 Just as the effects of tax revisions should take account of the resulting changes in taxpayer's behavior, so should the costs of regulations take account of changes in interest groups' behavior. If regulations make rent seeking more attractive, as they frequently do, such increased rent seeking decreases the productivity of society and must be reflected in any cost-benefit analysis.

[*pg 906]

Of course, even with these changes, the executive orders will not fully restore the cornerstones of federalism and tricameralism. The tension between one regulatory body -- OIRA -- and substantive regulatory bodies like the Environmental Protection Agency may not provide a deliberative process that is as demanding as the legislative process. Thus, other structures may well be necessary to restore a deliberative process that leads to optimal regulation for the public good. But the federalism and regulatory review executive orders are steps in the right direction.

I. CONSTITUTIONAL DECLINE AND THE RISE OF CENTRALIZED ADMINISTRATION
Individuals need government to create public goods that the market and the family cannot or will not supply. Examples of such public goods include military defense and the rule of law that makes property secure. They also include regulations that address externalities such as pollution. But the same motivations that drive people to create items for exchange in the market will lead them to try to use the government to take resources for themselves.13 Thus, the central dilemma of politics is that when a government is sufficiently powerful to supply public goods, it is also powerful enough to expropriate its citizens' property.14 Based on this understanding, constitutionalism necessarily concentrates on discovering mechanisms that will make government provide a high ratio of public interest goods to expropriation or private interest goods.

One way of understanding the Framers' Constitution is that it offered a particularly sophisticated structure that achieves these ends through the decentralization provided by federalism and the substantial barriers to federal rent-seeking legislation created by bicameralism and presentment. In this Part, the constitutional decline of these structures is briefly traced and the stage is set for understanding the regulatory review and federalism executive orders as movements toward constitutional restoration.

[*pg 907]

A. The Decline of Constitutional Federalism
One advantage of federalism is that a properly designed dual system of government with independent federal and state authority can limit the total amount of rent seeking by interest groups more effectively than a unitary state can. Rent seeking from the national government is limited by giving it only limited powers, which include limited powers of taxation. Rent seeking from state government is limited by putting state governments in competition with one another for capital, including human capital.15 The bridge between the two mechanisms is that the limited powers of the national government sustain the conditions for competition among the state governments.

Recent models of competitive federalism confirm that, as a matter of efficiency, the federal government should regulate only those activities that have spillover effects among the states.16 Where there are no substantial spillover effects (for instance, in the case of environmental regulation when jurisdictions do not pollute beyond their bounds), states are likely to provide appropriate levels of regulation even though they are competing to attract businesses from other jurisdictions. The simplified argument is that such competition allows the cost of regulations to be reflected directly in lost wages -- employees value public goods such as the environment or safety regulation and will accept lower wages to possess them.17 This connection between regulation and wages creates a dynamic that leads to the optimal level of regulation because the level of regulation is set at a point at which the marginal cost of a unit of regulation equals the marginal [*pg 908] benefit for individuals within that jurisdiction.18 As long as the costs and benefits are substantially confined to the state, the state government -- rather than the federal government -- should provide public goods, including appropriate regulatory regimes. This process restrains rent seeking and provides optimal public goods.

Another way of understanding the advantages of jurisdictional competition is that centralized regulation, even when that regulation is the product of a democratic process, has difficulty measuring citizens' preferences because elections are a relatively crude method of assessment.19 In contrast, jurisdictional competition is a discovery mechanism for preferences because it helps employees decide how to value regulations in terms of their preferences. Regulations are available for other states to emulate or reject, which creates, as Justice Brandeis noted, a "laboratory" of democracy in which the successful experiments of yesterday may become the effective public policy of tomorrow.20

Besides the advantages of providing a superior discovery process for regulations, state regulatory authority has two other advantages. First, different states provide different political niches for a diverse people, responding to different preferences with different goods.21 Inhabitants of San Francisco may simply have different preferences and needs from those of Dubuque, and uniform rules that fail to take account of this diversity could leave inhabitants alienated from their government.

Second, federalism increases civic responsibility. Political scientists frequently have noted that in large governments citizens behave strategically, consequently making it harder to reach agreement on [*pg 909] the public goods that will improve the community.22 Federalism and the more encompassing principle of subsidiarity (by which states devolve powers to localities) temper strategic behavior and substitute in its place the genuine concern of one citizen for another.23 As Alexis De Tocqueville said:

It is incontestably true that the love and habits of republican government in the United States were engendered in the townships and in the provincial assemblies. . . . [I]t is this same republican spirit, it is these manners and customs of a free people, which are engendered and nurtured in the different States, to be afterwards applied to the country at large.24
One of federalism's greatest public goods is that it provides a framework for sustaining the virtues necessary for self-government.

There is also historical evidence that the structure of federalism had substantial success in restraining rent seeking and thus in facilitating economic growth. Indeed, some economists suggest that decentralization was responsible for much of the economic progress of the United States as it was transformed from a fledgling republic to an economic superpower during the nineteenth century.25

The decline of federalism has been discussed elsewhere and it is unnecessary to belabor the issue here.26 It will suffice to say that there were two important stages of federalism's decline. The first came in the Progressive Era, when the Seventeenth Amendment's mandate for direct election of senators weakened federalism's restraints on production of private interest goods in two ways. First, because senators became independent of state legislators, they no longer had an incentive to protect state sovereignty.27 To the contrary, after the [*pg 910] Amendment, senators had a natural inclination to encroach on state sovereignty because states were a competing power center for servicing constituents and interest groups. As a result, federalism -- which depended on Congress's adherence to the Constitution's enumerated powers -- lost a crucial institutional defense within the federal government.28 One way of understanding the president's renewed role in the protection of federalism is that, after the Seventeenth Amendment, the president has a greater interest in sustaining state autonomy than do members of the Senate.29

Second, national regulatory regimes of broader scope than ever before were enacted in the New Deal, thus gravely weakening regulatory competition among the states. While the Court struck down some of these initial schemes as beyond Congress's Commerce Clause power,30 public pressure and a series of appointments by President Franklin Delano Roosevelt ultimately ended resistance to the exercise of national authority. By the early 1940s, the Supreme Court had abandoned the constitutional limitations that prevented the federal government from directly regulating manufacturing and the conditions of labor, thereby greatly increasing special interest power to obtain regulatory rents.31 Congress now has gained plenary regulatory power in economic matters,32 essentially ending the wealth-creating jurisdictional competition among the states.

B. The Decline of Tricameralism
Tricameralism consists of the requirement that a bill receive the approval of the House of Representatives, the Senate, and the presi- [*pg 911] dent or, in the event that the president vetoes the bill, obtain two-thirds support in the House of Representatives and the Senate.33 Tricameralism, like federalism, is another structure that helps the government resist rent seeking. Moreover, tricameralism generally contributes to the freedom from regulation that gives primacy to a social order created by voluntary interactions among citizens (such as commercial markets).34 The Framers, of course, did not expressly claim that the goal of bicameralism and presentment was to restrain factions and thereby promote the public good. Nevertheless, some of the Framers, like Alexander Hamilton and Joseph Story, did have an inchoate understanding of matters that public choice theory has since made explicit.35 In any event, the natural result of tricameralism was to advance the public interest by restraining rent seeking.

Bicameralism and the presidential veto will be considered in turn. Like a supermajority rule, bicameralism creates a higher hurdle for legislation because bicameralism requires majority passage in two chambers rather than one.36 Elsewhere Michael Rappaport and I develop the full argument that bicameralism operates as an implicit supermajority rule.37 The electorate often elects individuals with [*pg 912] nonidentical positions on the issues, particularly when the elections take place at different times and in different jurisdictions (as do the elections for the House of Representatives and the Senate).38

A bill that obtains majority support in one house may fail in the other; it will succeed only if it has sufficient support from the voters that it would obtain passage under supermajoritarian rule.39 The presidential veto works much like bicameralism in forcing legislation to obtain more than simple majority support.40 The president is unlikely to share the exact views of the average member of Congress, both because the president is elected from a different jurisdiction (the whole nation) and because the president is elected on the basis of slightly different issues.41 The president, for instance, has more responsibility for foreign policy and for appointments to the Supreme Court.

If the legislature attempts to pass a bill over the president's veto, the Constitution imposes an explicit supermajority rule by requiring a two-thirds vote to override the veto. Upon a veto by the president, a stronger legislative filter is appropriate because vetoed legislation is less likely to be beneficial than legislation approved by the president.42 In this manner, bicameralism and the presidential veto create a [*pg 913] mild supermajority rule that presents a greater hurdle to legislation than simple majority rule in the legislature.

Because a supermajority rule filters legislation in the public interest from special interest legislation, tricameralism may improve the quality of regulatory legislation.43 With certain simplifying assumptions, a supermajority rule is preferable when marginal legislation (legislation that would be enacted under a simple majority rule but would be blocked under a supermajority rule) is undesirable.44 It is certainly plausible that the marginal legislation at issue for the administrative state -- regulatory legislation that would pass in a unicameral legislature but would fail under tricameralism -- is often undesirable.

First, a unicameral legislature under majority rule will tend to produce many undesirable regulatory laws, because, from a global perspective, more interests favor regulation for self-regarding reasons than oppose regulation for self-regarding reasons.45 Perhaps most importantly, federal regulators themselves (including legislators and bureaucrats alike) are in favor of more regulation because it increases their budgets and brings them more status.46 Furthermore, the majority will systematically favor regulation when it can extract resources from the minority. To be sure, Madison's large republic tries to guarantee that no minority is an enduring target of regulatory or other kind of oppression.47 Nevertheless, given the imperfect nature of log- [*pg 914] rolling among coalitions, distinct majorities will use regulation on certain matters to extract resources from minorities. Finally, there are special interests, like those Mancur Olson described,48 that are likely to enjoy greater leverage than the diffuse and inattentive citizenry as a whole and to seek regulation for their own ends.49

Although some special interests will oppose regulation for self-regarding reasons,50 the supermajoritarianism implicit in tricameralism may be justified on the basis that these antiregulatory special interests are unlikely, over the whole range of the legislature's regulatory jurisdiction, to counterbalance the sum of the regulatory, majoritarian, and special interests in favor of regulation.51 Moreover, because the states offer the prospect of regulation to address externalities within their own jurisdictions, the class of special interests that seeks to avoid regulatory competition will also seek an inefficient federal regulatory regime. Thus, inefficient nationalization of regulation may be a hallmark of the modern nation state.52

Second, the mild supermajority rule created by tricameralism tends to filter inefficient regulation from efficient regulation. Supermajority rules make it harder for special interests to succeed because [*pg 915] they must gain more than majority support.53 On the other hand, a supermajority rule -- particularly a mild rule like that implicit in tricameralism -- should not prevent the enactment of efficient regulations because they are more likely to enjoy support from nonspecial interests and thus garner the requisite supermajoritarian support.54 Moreover, desirable legislation generates social benefits, and those who benefit from the legislation can compensate those who lose from it, thereby securing the supermajoritarian support necessary to pass the measure. Accordingly, if self-regarding interests will cause an inefficient mix of regulation under majority rule, a mild supermajority rule will improve the mix through filtering.55

In any event, the Framers' bicameral federal system certainly operates on the assumption that a procedure that requires more than majoritarian support will lead to better legislation. The Framers did not have a high opinion of the beneficence of legislation that could not pass the high hurdle of bicameralism and the presidential veto. Hamilton believed the "inconstancy and mutability in the laws" to be "the greatest blemish in the character and genius of our govern- [*pg 916] ments."56 If greater hurdles of bicameralism and presentment prevented some good laws from being enacted, that disadvantage would be "amply compensated by the advantage of preventing a number of bad ones."57 The purpose of this Essay is less to prove the Framers' principles in these matters than to show that the regulatory review order moves toward a structure that better reflects those principles.

Unfortunately, as Michael Rappaport and I discuss at length elsewhere, bicameralism and the veto power have been circumvented by government actors' exploitation of an ambiguity in the concept of legislation.58 The problem lies in distinguishing legislation on the one hand and implementation on the other. Because of the imprecision of language and the costs of legislative deliberation, all legislation leaves some discretion to the executive in implementation.59 The exercise of such quasi-legislative discretion, therefore, is not subject to bicameralism and presentment because the executive can exercise it legally. The lack of these supermajoritarian filters in the exercise of executive discretion offers the opportunity to circumvent bicameralism and presentment by recasting matters to be determined by the legislature as matters to be determined by executive discretion.60

Hence, over time, special interests have urged Congress to grant broad delegations of legislative power to agencies whose rulemaking processes can be captured more readily than the complicated and burdensome legislative process.61 Because such delegation can be [*pg 917] made to seem reasonable, an inattentive public cannot easily mobilize to defeat such delegations.62 The citizenry thus cannot police circumvention of the supermajority rules implicit in bicameralism and the presidential veto.63 The rise of the modern administrative state has created the possibility of massive substitution of administrative discretion for the original restraints of bicameralism and presentment.64

It might be argued that excessive delegation circumvents only bicameralism and not presentment because such regulations are made through executive regulation: the president, as head of the executive branch, still must decide whether to issue regulations within the ambit of his discretion under his delegated authority. But this argument depends on the premise that the president himself issues regulations in the same manner that he signs or vetoes bills. This premise is false practically, and at least according to conventional wisdom, false formally as well. As a practical matter, government agency heads issue most regulations under delegations directly from Congress.65 As a formal matter, it is generally thought that the president is not re- [*pg 918] quired to sign off on administrative regulations and that his signature is not needed to make such regulations effective.66

Moreover, the president faces political costs in enforcing his will when he finds that an agency head does not wish to follow his direction.67 While most agency heads serve at the pleasure of the president, they have a degree of autonomy because firing an agency head is politically costly to the president.68 The president will not fire the agency head if the agency head's regulatory agenda imposes fewer political costs than firing him would. Furthermore, the president also faces the typical problems of any principal in monitoring the actions of his agents.69

In particular, a modern agency head's agenda tends to diverge from that of the president because the agency head faces a different incentive structure than the president. We are far from the era in which Cincinnatus, having fulfilled his civic duties, simply retired to [*pg 919] his farm.70 Instead, the modern appointee on retirement from a high government position is likely to look forward to the enjoyment of higher income and status by virtue of his position.71 The extent of this enhanced status and income will depend on his network of friends and supporters and the general reputation in the press that he has acquired by virtue of his government actions.72 Needless to say, the actions that would optimize the agency head's future stream of income and reputational benefits will not necessarily be compatible with the president's agenda at all times.

Moreover, a realist analysis of the agency process would suggest that special interest groups will have substantial success at influencing the particular agencies whose regulatory agenda matter most to them. For instance, special interests can be expected to provide contributions to congressional committee members with influence over these agencies in return for access to and consideration of their concerns by the agencies that most affect them.73 For these reasons as well, the president's authority over regulations does not compensate for the loss of the filter provided by the veto power.74

Therefore, because of monitoring and termination costs -- costs that are built into the constitutional structure of the executive branch -- delegation to agencies does not provide assurance that an official who represents a majority of the nation's citizens will approve [*pg 920] of the regulation. Delegations in the modern administrative state still risk the loss of the constraints on special interest rent seeking that bicameralism and the presidential veto were designed to afford.

II. THE CHANGING WORLD OF POLITICS
A. Political Theory
The federalism and regulatory review executive orders compensate for the decline of federalism and tricameralism. As this Essay shows, the president has an interest in arresting this constitutional decline because his fortunes -- more than those of any other public actor -- are dependent on economic growth. The reason the public looks to him for economic growth is not unrelated to the centralization of power that accompanied constitutional decline: the national leader is now understood as a manager of the economy.75 The timing of the executive orders, however, has been dictated by the growing recognition that the centralization that began in the New Deal has substantially hindered governance in the public interest and tamped down on economic growth.

Today, the political world (and the political theories used to map it) looks very different than it did in the era in which political theorists celebrated the decline of federalism and the rise of the administrative state. Skepticism has grown about the government as a rational planner of social reform because the centralized reforms of the New Deal and the Great Society are widely believed to have had counterproductive consequences and, in some cases, to have been downright failures.76 The failures of these reforms are also congruent with deeper difficulties that fueled the growth of the administrative state. Public choice theory has cast doubt in various ways on the coherence and efficacy of collective decisionmaking.77 In particular, public choice theory has shown that what the Framers called factions -- now known as special interests -- may be able to exercise political power out of proportion to their numbers to obtain resources [*pg 921] and status for themselves.78 In contrast, because the diffuse citizenry has little leverage over centralized democracy, its inattention to government policy is rational and therefore intractable.79 For these reasons, interest groups have additional advantages in gaining resources from the government at the public expense.

Social changes have exacerbated both of these trends. The susceptibility of the national government to special interests has increased. The declining costs of information transmission have increased the ability of interest groups to influence government.80 As a result of these changes, interest groups are better able to organize and monitor the benefits that accrue to their members, thus avoiding the free-rider problems that previously had frustrated their efforts.81 They are also better able to monitor members of Congress, thus eroding legislators' ability to act independently.82 In contrast, the average citizen has become less focused on government activities because the consumption value of politics has declined -- with literally hundreds of channels for entertainment, politics has trouble competing, and thus the average citizen rationally spends less time considering political issues.83

[*pg 922]

The critique of the efficacy of centralized government has powerfully eroded the New Deal paradigm of the administrative state.84 In 1938, James Landis, then dean of the Harvard Law School, argued that, in view of modern problems, the administrative process should replace an outdated "tripartite" system of government, with confidence that unsupervised expert agencies would exercise broad delegations of power for the public good.85 But by 1960 Landis had become substantially disillusioned, suggesting that agencies lacked public spirit and were subject to undue influence by their regulated industries.86 As a consequence, he was an early advocate of greater presidential control.87 By the 1970s, economists had explained more rigorously the reasons that agencies tended to be captured by special interests and thus frustrate the public interest and hamper long-term economic growth.88 The administrative state needed to be restructured to reflect these social truths, a need that ultimately led to the issuance of the regulatory review and federalism executive orders. Moreover, given the perceived advantages of the market and the pathologies of centralized government, regulatory competition among the states naturally appeared to be a more sensible way of organizing society.

As these changes in intellectual outlook are assimilated, it is not surprising that the polity is looking for ways of returning to pre-New Deal limitations on government.89 The president is the appropriate instrument for that restoration because he has more interest in it than [*pg 923] any other national governmental actor. The public may be rationally ignorant of most issues, but they care about the economy and hold the president responsible for it. The challenge for the president is to capitalize on the public's interest in economic growth and use it to recreate structures of good governance -- through executive orders.

B. The Executive Orders
Elected on a platform of restoring the authority of the states,90 President Ronald Reagan promulgated a federalism executive order, Executive Order 12,612.91 The order reasserts the doctrine of enumerated powers and declares that political liberties are best protected by limiting the size and scope of the federal government.92 It mandates that executive branch agencies under the control of the president recognize the distinction between what is best regulated nationally and what is best regulated locally.93

In addition, the federalism order permits federal agencies to preempt state law only when a statute contains language clearly showing that Congress intended to preempt state law.94 It mandates consultation with state and local officials to avoid conflicts between state and federal law.95 In contrast to the operation of the regulatory review order which is monitored by the Office of Management and Budget, an official in each department is made responsible for monitoring compliance with the federalism executive order.96

After promulgating and withdrawing a federalism executive order,97 President Clinton's final federalism order, Executive Order 13,132,98 does not differ substantially from President Reagan's. It too reiterates the importance of federalism and suggests that "national ac- [*pg 924] tion" should be taken in response to problems of "national significance."99 Although not setting forth a test as to what constitutes a national or local problem, Clinton's federalism order directs agencies to "consult with appropriate State and local officials to determine whether Federal objectives can be attained by other means."100 It also requires clear evidence of Congress's intention to preempt before permitting preemption, although the standard is slightly less emphatic than that in President Reagan's order.101 The order continues to require each agency head to designate someone to be responsible for enforcing the order.102 Clinton's federalism order added the additional requirement that agencies provide, to the extent permitted by law, a waiver process to allow states that are accomplishing the same objectives with their own laws to be exempted from those aspects of a federal regulatory regime.103

Elected on a platform of restoring limited government as well as federalism,104 President Reagan issued the regulatory review order, Executive Order 12,291,105 shortly after he took office. The order required the agencies whose heads served at his pleasure to submit their major regulatory proposals to OIRA, a unit of the Office of Management and Budget (OMB), together with a "regulatory impact analysis."106 Only after OIRA signed off on their orders were the agencies permitted to issue the regulations; there were no time limits to the period OIRA could demand for completion of its review.107 The order also established cost-benefit analysis as the overarching principle that would govern regulation, insofar as that principle was consistent with the agency's statutory framework.108 Later, President [*pg 925] Reagan issued another order, Executive Order 12,498,109 that required agencies to provide an annual regulatory agenda, thus giving OIRA more advance warning of what regulations to expect.

These orders continued in force during the presidency of George Bush, President Reagan's successor. In 1992, President Clinton was elected as a "new Democrat," a Democrat who was sensitive to the costs of excessive government.110 President Clinton issued Executive Order 12,866,111 which maintained the principal objectives and structures of the Reagan executive orders while making some important changes. First, President Clinton modified the cost-benefit provisions, making it clear that government would take into account soft as well as hard costs and consider the "equity" and "distributive" consequences of regulation.112 In response to critics of the secrecy of the regulatory review process,113 the new regulatory review order also required very substantial disclosure -- it decreed that in the OIRA review process only the administrator could be lobbied orally by third parties and that the content of these contacts must be disclosed.114 A log of all written and oral communications with OIRA must be kept, and the log must be disclosed to the public.115 Moreover, the order requires that all communication between OIRA and the regulatory agency be disclosed after a final order is issued or the decision is made not to promulgate a final order.116 In response to criticisms of the undue delays that OIRA imposed,117 Clinton's order imposes strict limits on the time OIRA can spend evaluating an agency's regulations.118 The order also expressly contemplates that the president, or [*pg 926] the vice-president acting at the president's direction, can settle disputes between OIRA and a regulatory agency.119

III. CONSTITUTIONAL RESTORATION THROUGH PRESIDENTIAL REVIEW
A. Reviving Constitutional Federalism
The president is the political actor most likely to revive the virtues of the Framers' federalism. To be sure, federal legislators are closer to their individual states than the president. But they are interested in the welfare of people in their individual states, not the welfare of the nation as a whole. Indeed, unlike the contribution that federalism makes to economic growth through jurisdictional competition, the geographic nature of representation at the federal level detracts from economic growth. For instance, each legislator has an incentive to bring back pork barrel legislation for his state despite the economic losses this causes the nation.120 In the case of spending, the representational nature of federalism can cause a geographic tragedy of the commons in which each representative has an incentive to overgraze the federal budget at the expense of the nation's economic prosperity. Exactly the same incentives work to the disadvantage of the nation at the regulatory level: each individual representative wants to obtain a regulatory framework that benefits his state even if it hurts the nation as a whole.

In contrast to their interest in bringing back pork, federal legislators do not have as strong an interest in strengthening state autonomy.121 Protecting autonomy has benefits that are much less visible to voters and thus do not advance the career prospects of politicians.122 Additionally, more powerful states mean more powerful state offi- [*pg 927] cials, and these officials represent an important source of competition for federal legislators for reelection (or, in the case of members of the House of Representatives, for both reelection and promotion to the Senate). Moreover, even if individual members of the legislature want to strengthen state autonomy, they cannot achieve this goal as easily as they can succeed in bringing back pork. Pork barrel legislation creates no free-rider problems among legislators: all legislators are rewarded with pork for their own districts if they support similar legislation for others.123 State autonomy brings no such individualized payoffs because the autonomy of all states is increased by the action of an individual senator protecting state autonomy. Given this free-rider problem, it will be difficult to mobilize the critical mass of legislative support needed to achieve a restoration of state authority through federal legislation.

With the dissolution of competitive federalism permitted by the Supreme Court,124 the principal remnant of the system of constitutional federalism -- the election of national legislators from geographically based districts -- may do more harm than good. It assures a steady stream of pork and regulatory parochialism without sustaining the regulatory competition that might help the nation's economic growth. One way to understand this point is to recognize that federalism in the Constitution has two aspects: governmental federalism and representational federalism. Governmental federalism gave responsibility to the states for areas of governance (largely in areas in which there was little likelihood of interjurisdictional spillovers).125 Representational federalism makes representation in the federal government dependent on support in particular states rather than in national at-large elections.

Perhaps arguments can be made that representational federalism is useful to protect governmental federalism from collapsing from the consolidation of all power in the federal government.126 Nevertheless, [*pg 928] when divorced from the system of governmental federalism, representational federalism has serious defects. More bluntly, constitutional changes since the New Deal have retained the defects of representational federalism -- dominance by parochial factions through local representation -- but added the defects of centralized regulation that comes from dissolving governmental federalism and giving plenary power to the national legislature.

But the pathologies that flow from this system have become so palpable that they set the stage for new ways of reconstituting a federalism that preserves liberty and generates wealth. Here the president is the leading actor on the set. First, the electoral fortunes of the president and his successor are more dependent on economic growth than those of national legislators.127 Moreover, unlike national legislators, the president has no free-rider problems in enforcing a system of competitive federalism. The president also should be less reluctant than national legislators to cede power to state officials because even more substantially empowered state officials are not going to overshadow the president. Finally, the federalist structure of the electoral college also gives the president an interest in appeasing state officials, particularly state governors, because governors and state party committees play an important role in presidential elections.128

The president's role in restoring the framework of restraints that contribute to economic growth has deep historical roots. As Mancur Olson has observed, monarchs who enjoyed stable control over their territory had incentives to create the conditions for economic growth.129 To this end they created rule-of-law institutions, like courts, and tried to keep the peace.130 The president's role is similar, with ad- [*pg 929] ditional benefits provided by modern democracy. Monarchs' incentives lay in the greater taxes and other exactions that they could accrue with greater economic growth. The president's incentives are the prospect of reelection -- his own or that of his party's successor -- and his historical reputation, and therefore he does not take his exactions directly at the expense of the people.131 The voting citizenry provides these incentives, because although citizens may not know much about the details of government policy, they appear to hold the president responsible for economic growth.132

Of course, the president is not going to be a perfect enforcer of federalism. He also has factions that he wants to satisfy at the expense of long-term growth. But he is likely to prove better at preserving federalism because he has a strong countervailing interest in continuing economic growth for the foreseeable future. Moreover, the federalism executive order (and the regulatory review order as well) can act as a precommitment device.133 It thus enables the president to make it more difficult for himself to heed the pleas of these factions at the expense of this overriding interest. More to the point, perhaps, the federalism executive order may redirect these factions to other sources of rents or status that are unlikely to interfere with competitive federalism.134

Although President Reagan issued the federalism executive order, President Clinton continued the order, thus showing that the movement toward presidential enforcement of federalism may be beyond partisan politics.135 Although neither order is explicit about restoring constitutional federalism per se, both resurrect key concepts in distinguishing between what is best regulated locally and what is [*pg 930] best regulated nationally and directing agencies to focus only on national problems.136

The kind of competitive federalism the presidential executive order enforces could enjoy some advantages over the constitutional federalism the Court enforced in the pre-New Deal era. During that time, the Court determined what the federal government could regulate and what was reserved to the states on the basis of rigid categories (e.g., commerce, labor, manufacturing).137 The problem with such categorical distinctions is that they may not permit the federal government to regulate all matters that create interstate externalities (and thus that cannot be addressed optimally by interstate competi- [*pg 931] tion).138 For instance, there is some question whether under the pre-New Deal distinction under the Commerce Clause the federal government could regulate a polluter that was engaged in interstate pollution.139

In contrast, the federalism executive order would allow the president and his advisers the flexibility to make economically sound judgments about what is best regulated locally and what is best regulated nationally, taking into account both the advantages of jurisdictional competition and the advantages of national harmonization.140 For instance, under either President Reagan's or President Clinton's executive orders, the federal government undoubtedly would be able to address interstate environmental problems that manufacturers created. In short, the president is not limited by formalism and could help implement a more effective system of competitive federalism.

But the virtues of informality are also potential vices. A pragmatic federalism will have much greater difficulty resisting the calls for the federal government to do something about the political imperative of the day, even if the problem does not have substantial spillovers. Some of these imperatives are manufactured by interest groups that would benefit from unjustified centralization.141 Others are the result of the passions and rational ignorance of ordinary citizens. Federalism is a structural principle that restrains interest groups, but, for this reason, interest groups are largely hostile to it. Moreover, because it is a structural principle, citizens are largely indifferent to it when it conflicts with issues that stir their passions. Ultimately, a rigid constitutional structure, like that of our original constitutional federalism, may be more impermeable to the acids of interest group rent seeking and political passions.142

[*pg 932]

Nevertheless, much of government regulation concerns matters that, although important, do not stir great political passions. As to these matters, the presidential order can be an effective counterweight to federal bureaucracies' natural tendency to impose federal solutions. Thus, even if the federalism executive order does not restore federalism to its former glory, it may be the most substantial revival possible in the current polity.

B. Reviving Tricameralism
The process of regulatory review can best be understood as an attempt to compensate for the decline of bicameralism and the presidential veto occasioned by delegation. The OIRA review process envisioned in the regulatory review order facilitates this restoration in three ways. First, by requiring OIRA review of major regulatory orders, it reduces the president's monitoring costs and restores in some measure the president's original role in influencing the content of coercive impositions on citizens.143 Second, it introduces another level of bureaucracy that must be navigated before a regulatory proposal becomes binding on citizens.144 This bureaucracy has a different culture from the agencies and is influenced by different factions.145 By introducing another screen, it also recreates some semblance of bicameralism's two hurdles to the enactment of binding laws. Finally, although the procedural requirements of the executive order attempt to mimic the processes of bicameralism and presentment,146 the regulatory review executive order also attempts to achieve the goals of these constitutional cornerstones through a new substantive requirement. The federal government should provide only public goods, including [*pg 933] regulations, that other institutions, like the family and the market, either cannot provide or cannot provide as efficiently.147 The cost-benefit analysis requirement encapsulates this goal, filtering regulations that meet this criterion from those that do not.

As discussed in connection with the federalism executive order, the president is now the logical candidate for reviving normative structures that restrain special interests or factions.148 Because special interest legislation depresses long-term economic growth, the president, as the national leader most closely tied to the prospects of economic growth, is most likely to consider structures to reduce the incidence of special interest legislation.149

Moreover, unlike the Supreme Court, the president is a politically accountable official and has the legitimacy to make policy-based judgments. The Court could attempt to adopt formal rules proscribing excessive delegation by reviving the nondelegation doctrine.150 But it is difficult to draw clear lines to quantify the excesses of delegation, and given this lack of clarity, the Supreme Court could be accused of hiding behind a patina of legal neutrality to advance its political agenda.151 In any event, the Court shows no signs of reviving the nondelegation doctrine, most recently passing up the opportunity to affirm a lower court's use of the doctrine to invalidate an environmental law.152 Just as the Court's failure to revive economic federalism strengthens the case for the president's role in the reconstruction of [*pg 934] federalism, the Court's failure to prevent circumventions of bicameralism through excessive delegation strengthens the case for the president's creation of substitute restraints on special interest legislation.153

For similar reasons, expanded presidential review may be more effective than the hard look doctrine. The hard look doctrine contemplates that to assure regulation in the public interest, courts should require agencies to provide full explanations, examine objectives and relevant factors, and consider alternatives.154 The doctrine provides considerable discretion to courts that may not be well positioned to second-guess agencies, however.155 Additionally, the hard look doctrine may frustrate the rulemaking process by requiring agencies to jump through hoops that are unnecessary in some cases.156 In contrast, the president is able to examine the merits directly through cost-benefit analysis and to eschew costly and unnecessary procedures.

Some might argue that the president's interest in pursuing economic growth might be deleterious because he would tend to avoid regulations that would depress growth, even if that meant discarding regulations that address externalities that have long-term effects. This danger does not seem enormously significant because, when citizens focus on issues in elections, they seem to take into account important public goods like national defense and the environment.157 Moreover, other key actors in the regulatory process, such as agency heads, do not have longer time horizons than the president. To the contrary, the president is more concerned about long-term effects because he wants to be reelected, have a member of his party succeed him, and burnish his historical reputation.

[*pg 935]

1. Restoring Bicameralism Through OIRA. The requirement that a regulation receive the approval of OIRA as well as agency approval introduces, like the addition of another legislative chamber to a previously unicameral legislature, an additional barrier to restricting citizens' freedom. Furthermore, like an additional legislative chamber elected from different jurisdictions, OIRA impedes interest group legislation to some extent. Even assuming that the same interest groups were as capable of influencing OIRA as administrative agencies, the additional layer of review makes it harder for these interest groups to gain rents through regulation just as bicameralism makes it harder for interest groups to gain rent through legislation.

The implicit theory behind bicameralism is that requiring two chambers' approval is likely to improve legislation through the filtering process.158 And this theory, if true for bicameralism in the legislative process, should be true of two layers of administrative review as well. The different interest groups likely to influence OIRA and the regulatory agency may counterbalance one another, creating a regulation more representative of the public interest than any special interest. For instance, management might influence OIRA to counterbalance union influence on agencies. This view of the advantages of a two-step administrative process reflects the more general theory of Federalist No. 10 as well: the more freely counterbalancing factions circulate, the more likely policies will emerge that are closer to the public interest.159

Thus, I disagree with the notion advanced by some commentators that it is imperative for those in the OIRA process to work smoothly with those in the agencies.160 This notion fails sufficiently to take account of the manner in which bicameralism can be seen as an analogue to the regulatory review process. A central advantage of OIRA review is that it offers a distinctive process for regulatory approval that may be influenced by different interest groups at least some of the time. Just as bicameralism is premised on the proposition that creative tension between two different legislative chambers will [*pg 936] lead to better laws, creative tension occurring between OIRA and a regulatory agency should lead to regulations that are more likely to advance the public interest.

It might be argued that if two sets of interest groups are poised to influence regulation, the emerging regulation may be more costly because it must satisfy two sets of interest groups rather than one. This criticism also can be made of legislative bicameralism, and the response at the administrative level is basically the same as the response at the legislative level. Assume that a different set of interest groups attempts to gain rents in the Senate than seeks rents in the House of Representatives. Although bicameralism could abstractly generate worse legislation, the proposed Senate amendments that make the legislation worse will create additional opposition to the legislation. It is not true that legislation can provide limitless rents or we would observe even more rent-seeking legislation.161 Rents are granted until the point at which the marginal political cost outweighs the benefit of granting any more rents. A premise of bicameralism is that, because an interest group generally would have bumped up against the marginal ceiling where additional rents will be politically costly in one chamber, additional interest group activity in the other chamber is more likely to have a counterbalancing effect. The same dynamic should be at work in OIRA. If the interest groups already have come close to exhausting the rent-seeking potential of a particular regulation, it will be hard to add more rents without triggering more substantial opposition.

Moreover, the influential factions at OIRA are likely to be less egregiously rent seeking than the narrow industry- and labor-oriented factions that will influence the administrative agencies.162 First, they are likely to be filtered through political parties because OIRA is closer to the White House, which is traditionally influenced by political parties. Once filtered through a political party, factions tend to [*pg 937] become more broad-based163 -- what Mancur Olson would call a more "encompassing interest."164 These broad-based factions are less likely to take rents for themselves because, as they become a more encompassing interest, it is more likely that the resources used to pay for the rents will be drawn from the encompassing interest itself.165 Second, agencies outside the direct purview of the president, particularly independent agencies, tend to be more easily influenced by congressional power, particularly through congressional committees, than agencies like OIRA that are within the Executive Office of the President.166 Regional factions can be particularly harmful to economic growth when they are given the power to influence regulations that have effects outside their region.167 Because OIRA is less subject to congressional influence, the factional influence on OIRA is likely to be less detrimental than the influence on the regulatory agencies themselves.168

It has been argued that OIRA is not a neutral screen because it is more likely to block regulation or weaken it than insist on more robust regulation.169 But this asymmetry actually brings us closer to the [*pg 938] original design of bicameralism because a premise of bicameralism was the absence of substantial regulation. The Framers did not need to worry that bicameralism would block attempts to repeal burdensome laws because, at the beginning of the Republic, the federal government had passed few such laws.170 In the context of the modern administrative state, however, where there has been substantial delegation already, the Federal Register already contains much special interest regulation. Insofar as OIRA poses less of an additional hurdle for deregulation than it does for regulation, the regulatory review process actually makes the political process more similar to the Founders' design.

2. Restoring the Presidential Veto Through OIRA. Another effect of the regulatory review process is to compensate for the loss of the president's veto power. The review process reduces his transaction costs in monitoring agencies and thus helps him keep agency heads within the margin of autonomy that their independent authority (even in the important but limited sense that they often must be fired to stop their exercise of authority contrary to the president's will) permits.171 It thus strengthens the president's autonomy and makes his regulatory review power similar to his power to veto bills.

It could be argued that the reduction of monitoring costs is illusory because the president still has to monitor the staff of OIRA, including the desk officers at OMB. No structure can eliminate transaction costs in a hierarchy, but OIRA reduces them in two ways. First, for reasons of culture and incentives, OIRA is closer to the president.172 The bureaucracy, and particularly the head of OIRA, largely will be judged a success or failure by its implementation of the president's program: that is its job. Second, conflicts with particular agencies will provide the president and his immediate subordinates with some information about which agency heads warrant close scrutiny. [*pg 939] As congressional committees provide fire alarms for Congress, telling them of agencies that need to be monitored more closely, OIRA provides fire alarms for the president.173

Others have objected to OIRA's micromanagement of the agencies, suggesting that the president would be better off setting clear, broad parameters and letting the agencies follow them.174 Clear policies from the president will doubtless improve implementation of his agenda -- that is one reason that Ronald Reagan was a successful president. But applying broad policies to specific cases requires interpretation because different presidential policy objectives may come into conflict, and it is useful to have OIRA to articulate the nuances of policies centrally. In that way, the tradeoffs among different presidential goals can be refined, particularly when disputes are raised to the president or the vice president under the current executive order. At its best, this process has a dialectical advantage over application of a fixed set of goals; by weighing difficult tradeoffs in different contexts, the OIRA review process actually will improve the articulation of the president's goals, making them more nuanced and transparent.

3. Advancing the Objectives of Tricameralism Through Cost-Benefit Analysis. The final aspect of the regulatory review order that attempts constitutional restoration is the substantive requirement of cost-benefit analysis. The original constitutional filters of bicameralism and presentment offered the promise of distinguishing public goods from special interest legislation overall.175 The cost-benefit analysis accomplishes much the same objective, at least if the baseline for regulation is the absence of government intervention, because cost-benefit analysis should authorize government intervention only if its net benefits are greater than those provided by the market or nonmarket forms of spontaneous order like the family. Such intervention produces public goods -- services that the market or the family either cannot provide or cannot provide as efficiently.176

[*pg 940]

Cost-benefit analysis tries to provide a process that can substitute for the filtering provided by the original Constitution to assure that government action serves the public good rather than special interests.177 Thus cost-benefit analysis revives tricameralism in a manner fundamentally different from the manner in which other provisions of the regulatory review order revive tricameralism. First, cost-benefit analysis is substantive while the other provisions of the regulatory review order impose procedural constraints by adding another layer of review through OIRA. Second, cost-benefit analysis directly aims at the goal of tricameralism -- assuring that government action is in the public interest -- rather than replicating its structure. Because of these features, cost-benefit analysis is perhaps a more radical form of constitutional restoration. It may be justified nonetheless because the other provisions of the regulatory review order cannot fully restore tricameralism. After all, getting a bill through both the mutually jealous Senate and House of Representatives is harder than getting a regulation through a substantive agency and OIRA. Moreover, because the power of special interests has grown since the founding of [*pg 941] the Republic,178 it may well be that the government needs even stronger restraints against special interest-inspired impositions than it did then.

It is beyond the scope of this Essay to respond to the many objections to cost-benefit analysis. I agree with Professors Matthew Adler and Eric Posner that cost-benefit analysis ultimately can be defended (as against other global measures on the efficacy of agency regulation) on the grounds that it is a more feasible metric to apply than others that have been proposed.179 From my analysis of the purpose of the regulatory review order, one other point emerges for this institutional defense. If one of the principal purposes of the regulatory review process is to block special interest regulation, cost-benefit analysis performs a useful function. Special interests want to hide the costs of their regulations and exaggerate the benefits because that makes it harder for political entrepreneurs to rally the diffuse citizenry to oppose these regulations.180 By imposing a single, relatively straightforward symmetric metric on regulations, cost-benefit analysis makes it harder for interest groups to hide costs and exaggerate benefits.181 Many of the other metrics, such as multidimensional assess- [*pg 942] ment, are far more complicated and opaque and they would thus perform this function less effectively.

Direct consideration of costs and benefits by the subordinates of an elected official has some distinctive virtues and vices compared to other structures designed to assure that the federal government produces only public goods. It is useful to compare cost-benefit analysis with a legislative structure that relies on the procedural screens of bicameralism and the veto power to filter out special interest laws and with a judicial process for the invalidation of special interest legislation of the sort Richard Epstein recommends in his book Takings.182

Cost-benefit analysis has some advantages over these structures. Compared to the legislature, the president can more easily employ experts because of the hierarchical nature of the executive branch.183 These experts, in turn, can provide input that lay analysis could not replicate, yet their contributions ultimately would be assessed in political terms because of the president's politically accountable role. In contrast to a judicial scheme, cost-benefit analysis places responsibility in a more politically accountable individual. As a result, cost-benefit analysis can be carried out in a manner that will be more sensitive to political concerns, and it may be ultimately less susceptible to attack and less politically fragile.

To be sure, presidential calculation of costs and benefits has some disadvantages as well. As compared to the legislature, the executive branch is a more centralized decisionmaker. Like all centralized decisionmakers, the president and his subordinates will have difficulty assigning the correct costs and benefits.184 In contrast, the process of bicameralism is somewhat less centralized because it obtains information about costs and benefits from representatives throughout the country.185 Moreover, compared to decisions made [*pg 943] with the insulation of life tenure, cost-benefit analyses executed by a politically accountable agency close to the president are prone to be distorted by political passions in politically sensitive cases.186

For instance, it can be argued that the president will not faithfully carry out serious cost-benefit analyses because he will under- estimate the costs when these costs fall on factions that do not support him and overestimate the benefits when they accrue to factions that support him. In this sense, the fear would be that cost-benefit analysis would simply serve as a cover for rent seeking for a different set of factions rather than that those factions would dominate a pro- cess in which agencies made the final decision.187

There are two responses to this concern. The first is that cost-benefit analysis will discipline some of these political tendencies, which would occur even in the absence of cost-benefit analysis. At least the requirement of the calculus provides a useful framework for comparison.188 Individuals and groups can cite to inconsistencies, if these crop up, and, by applying the same standard across agencies, critics can make interagency comparisons. Second, because of the president's interest in long-term growth, he is likely to be a more faithful arbiter of cost-benefit analysis than other government officials. The defense of cost-benefit analysis undertaken by the president, like all defenses of institutional arrangements, is not that it is perfect but that it is better than the alternatives.

Nevertheless, even if no single aspect of the president's executive order on regulatory review compensates for the loss of tricameralism occasioned by delegation, both the procedural and substantive aspects of the regulatory review order move us back toward the Framers' system designed to filter public good provisions from special interest impositions.

[*pg 944]

IV. RECOMMENDED REVISIONS OF THE FEDERALISM AND REGULATORY REVIEW EXECUTIVE ORDERS
A. Revising the Federalism Executive Order
The normative analysis sketched above leads to the following recommendations for revising the federalism executive order:

1. Strengthening Competitive Federalism. The original Constitution recognized that there were costs to centralized regulation and circumscribed the government's regulatory powers. The federalism order should be amended to strengthen competitive federalism. It should specifically allude to the original design and suggest that federal regulation is warranted, unless otherwise required by law, only when the presence of interstate externalities or spillovers suggests that federal regulation is necessary. In other words, the existence of substantial externalities should be a prerequisite to going forward with the cost-benefit analysis at all.189 The order should specifically note that when an activity within individual states creates no substantial spillovers into neighboring states, and when there is no pressing need for national harmonization of standards, regulatory competition among these states is not in itself an externality but a desirable aspect of federalism.190 Even when there needs to be some federal involvement, the order should direct OIRA to consider whether there remains a role for state regulatory competition within an overall federal framework.

2. Applying Federalism Principles to Agency Recommendations About Legislation. One important limitation on the degree of constitutional restoration that the federalism order affords is that its application is limited to the executive branch. Congress can pass statutes that do not comply with the federalism principles that limit the role of the federal government to solving social problems with substantial externalities. One partial response to this difficulty is to direct agencies to follow the principles of the federalism order in their recommendations as to whether the president should sign such legislation.

[*pg 945]

3. Revising the Preemption Section of the Federalism Order. Executive Order 13,132 imposes a clear statement rule on agencies in preemption matters. Section 4 of the order essentially directs the agency to interpret a statute as requiring preemption only when Congress expressly or clearly intends to preempt state laws and to interpret it as permitting preemption only when that delegation is clearly expressed or intended.191 The standard for preemption in the Clinton order is not very different from that in Reagan's original order, although the Reagan order had a slightly stronger standard for preemption.192 The breadth of the clear statement rule should be slightly narrowed to those areas in which preemption would undercut jurisdictional competition, because such a scope would better comport with a view of the federalism order as constitutional restoration and because it would be easier to defend as a matter of law.193

[*pg 946]

To be sure, the Reagan administration consciously rooted its clear statement rule in the need for a constitutional restoration in the area of preemption.194 Forms of preemption other than express preemption were thought to make it harder for the states to protect their interests in the political process, undermining the procedural protections of federalism. In Garcia v. San Antonio Metropolitan Transit Authority,195 the Court emphasized the importance of the protection that the constitutional legislative process affords to state interests.196 Although Garcia is generally regarded as a case unsympathetic to federalism (because it limited the judicial enforceability of the Tenth Amendment), its reasoning could be seen to provide support for cabining preemption to cases where Congress has expressly displaced state law.197 Regardless of the validity of its refusal to enforce a sphere of sovereign state autonomy through judicial action, Garcia reiterated that one of the most important bulwarks of federalism is the constitutional structure of the legislative process.198 Under this view, the difficulty with forms of preemption other than express preemption is that they deprive the states of the full benefits of this protection by permitting state laws to be preempted indirectly by the inferences and constructions of an unelected federal judiciary rather than directly by the votes of legislators elected from the several states.199 In other [*pg 947] words, when "obstacle" or "occupation of the field" preemption is permitted to displace state law, state autonomy depends not upon the deliberate and clearly enunciated political judgment of those elected from the states but upon the ad hoc determinations of federal judges.200

Fidelity to the separation of powers was also thought to weigh in favor of narrowing the preemption doctrine to favor a clear statement rule because it would force Congress to take responsibility for the hard choice to displace state law rather than sloughing this responsibility off on the courts.201

Finally, an elastic preemption doctrine was thought inconsistent with yet another strand of the separation of powers because it encourages the judiciary to overstep its constitutional bounds.202 Because of the amorphous nature of the tests employed by the courts, "obstacle" and "occupation of the field" preemption tend to entangle the judiciary in an evaluation of the competing policy merits of a solely federal regime and a federal regime supplemented by state law.203

[*pg 948]

Despite the power of these arguments, the jurisprudential justification for the clear statement rule has been challenged in two careful recent articles on preemption. Professor Caleb Nelson suggests that the language in the Supremacy Clause that directs courts to enforce federal law "anything to the contrary notwithstanding" precludes a presumption against preemption, let alone a clear statement approach.204 Nelson asserts that the Supremacy Clause's language was well understood as a "non obstante" clause at the time of the Framing and that such clauses were expressly intended to override the traditional presumption that later statutes should not be read to contradict prior statutes.205 In the case of the Supremacy Clause, it was intended to reject a general presumption that federal law does not contradict prior state law.206 Therefore, even conceding that the political process provides protection for state autonomy, the protection should not generate special canons of statutory interpretation to cut back on preemption.207

Professor Viet Dinh also suggests that a clear statement requirement or even a general presumption against preemption would be anomalous given the other ways in which federal law displaces state law.208 For instance, Professor Dinh notes that under the dormant Commerce Clause or federal common law, state law can be ousted even without clear congressional action.209 He argues that federal interstitial lawmaking cannot be squared with the clear statement rule,210 and thus it would not be consistent to apply the presumption only to preemption.211

Whatever the resolution of this jurisprudential debate on the proper scope of judicial preemption standards, the conception of constitutional restoration advanced in this Essay provides policy reasons [*pg 949] for the president to direct his agencies to follow a clear statement rule within the scope of their statutory discretion.212 Recall that this Essay contends that the federalism order is necessary to restore the jurisdictional competition that has been eroded by Congress's use of plenary powers. Accordingly, at least when Congress is not addressing some interjurisdictional spillover or strengthening the conditions for interjurisdictional competition (such as permitting free movement of people and capital), there are policy reasons for the executive to employ its administrative discretion against interpreting a statute to require preemption or to delegate authority for preemption to the agency. It would follow from this rationale that in those few instances when Congress is addressing spillovers or protecting the free flow of capital among the states, the presumption against preemption should not apply because the model of jurisdictional competition assigns these responsibilities to the federal government.213 As with other issues in federalism, the executive has an advantage over the judiciary in creating a pragmatic doctrine of preemption that actually advances the general values of efficient constitutional design.214

Moreover, this policy rationale for a clear statement rule would provide a firmer legal basis for the federalism order's authority to cut back on agency preemption when agencies are exercising discretion. There is a debate in the literature about whether the policy decisionmaking that an agency can exercise within the scope of its statutory discretion under Chevron U.S.A. Inc. v. Natural Defense Resource Council, Inc. permits the agency to disregard judicial canons of con- [*pg 950] struction.215 The Supreme Court has not consistently adopted a clear statement rule for preemption.216 Therefore it could be argued that if the president directed the agencies to adhere to the clear statement rule, he would be adding a canon of construction based on jurisprudential views contrary to the Court.217

In contrast, if the president based his presumption on his view of jurisdictional competition, he would be injecting the kind of policy consideration with which Chevron undoubtedly permits agencies (and, by extension, the president as their supervisor) to guide policy decisions within statutorily granted discretion.218 Case precedent also suggests that the executive branch should be able to disregard canons of interpretation that simply reflect policy considerations when those canons do not reflect the agency's policy calculus.219 As discussed above, preemption cases that are not based on clear congressional intent necessarily require the judiciary to conduct a policy balance between the relative weight of state and federal interests.220 When a politically accountable agency is available, Chevron suggests that this policy calculus should be made by the agency.221 There is no reason that the president cannot direct the exercise of this discretion on [*pg 951] cross-cutting policy considerations such as the promotion of jurisdictional competition.222

4. Judicial Review of Executive Orders. An issue common to both the regulatory review and federalism executive orders is whether judicial review should be available to litigate compliance with the orders. Currently both orders flatly bar judicial review.223 But if the orders are seen as a means of constitutional restoration, it is not clear that judicial review should be barred, at least for procedural matters. For instance, the Supreme Court surely would hold null and void the attempt to enforce a law that was not passed by the Senate. If the OIRA process is seen as a substitute for bicameralism and presentment, why should the courts be unable to assure compliance with the modern substitute? Similarly, if federalism review is a substitute for the original structure of enumerated powers within the Constitution, it is important enough for the courts to make sure that agencies have jumped through the appropriate procedural hoops. Moreover, permitting judicial review will make it more likely that agencies will scrupulously comply with the procedures set out in these orders.

Strong countervailing considerations, however, suggest that the substantive provisions of the orders, such as cost-benefit analysis and the review of whether sufficient spillovers exist to justify national regulation, should not be subject to judicial review. These provisions involve policy determinations and the weighing of economic costs and benefits -- areas in which the executive has a comparative advantage over the judiciary. Moreover, involvement in these matters might force the judiciary into policy and political thickets that might undermine its reputation for neutrality.

[*pg 952]

B. Revising the Regulatory Review Order
From this normative analysis emerge some recommendations to strengthen the regulatory review executive order as well:

1. Expanding the Scope. First, a revised regulatory review order should reject the decision to restrict the full application of the order to only major rules.224 Just as bicameralism and presentment apply to all proposed laws, so should the OIRA process apply to all regulations. The reasons for a blanket requirement in the regulatory review process are similar to those that underlie the breadth of the requirement of bicameralism. The filtering process afforded by the regulatory review order should improve regulations overall, not only the set of regulations with very substantial regulatory costs. Moreover, without careful review, it is difficult to know whether even a minor regulation will create a dynamic that generates further regulation in the future.225

Professors Posner and Adler suggest that the threshold of $100 million in costs required to trigger OIRA cost-benefit analysis may be sensible because it differentiates "between projects that are large enough to warrant the direct costs of [cost-benefit analysis] and projects that are too small."226 Even setting aside the dynamic of regula- [*pg 953] tions that generate more regulations, it would be better to limit the budget for cost-benefit analysis of a minor rule so that the costs of the analysis itself meet a cost-benefit standard.227 Given the pressure exerted by interest groups to issue regulations, even minor regulations would be improved by the application of a cursory cost-benefit analysis. Senator Everett Dirksen once said of small appropriations: "a billion here and a billion there and pretty soon you're talking big money."228 The same can be said of minor regulations.

Second, if the president and his agency heads have somewhat adverse interests, review of less substantial regulations can create bargaining chips for the president. In the absence of a monitoring problem for insubstantial rules, the agency head can simply exempt them from the calculation of how far he can depart from the president's desires and still keep his job. On the other hand, if these rules are monitored by OIRA, they become part of the agency head's internal calculus and thus the overall bundle of rules should better reflect the president's desires. We can put this in arithmetical terms. As discussed above, an agency head cannot impose more costs on the president than his dismissal would cause.229 If the departures from what the president would want in less significant regulations are taken into account, the agency head's scope for departure in more significant regulations is reduced. Even if OIRA has scarce resources, it should be able to decide which agency regulations to review. Some agency heads warrant a closer watch than others.230

2. Including Independent Agencies in the Regulatory Review Order. The regulatory review order does not include independent agencies -- agencies whose heads do not serve at the pleasure of the [*pg 954] president -- within its full review process.231 Independent agencies' exercise of quasi-legislative authority undermines tricameralism to an even greater extent than the exercise of such discretion by nonindependent agencies because such independence diminishes the president's formal authority to direct agency heads in the exercise of their discretion and makes that authority even less comparable to his veto power. Some scholars argue that the difference is not a purely formal one: interest groups are better able to dominate independent agencies, thus weakening presidential control.232 Therefore, if presidential review is seen as constitutional restoration, the case for including independent agencies under all aspects of the presidential regulatory review order is even stronger than including nonindependent agencies.233 For reasons well articulated by others, I believe such an order would be constitutional.234

3. Deleting Time Limits. Third, the decision in President Clinton's regulatory review order to put limits on the time OIRA can review a prospective regulation is a mistake.235 The requirements of bicameralism and presentment often delay lawmaking because of disagreement between the chambers or because of a presidential veto. The premise behind such potential delay is that it is better to have no [*pg 955] law than law without a consensus sufficient to assure that it is in the public interest. Therefore, the requirements of bicameralism and presentment continue to be applied despite the passage of time without a law being enacted.

To compensate for the loss of bicameralism and presentment, the regulatory review process should similarly not be terminated on an arbitrary date, because the advantages of filtering will be decreased by the agencies' ability to wait out the regulatory review process. It is true that, in an impasse, the agency or OIRA can ask for a determination of the president or, at the president's direction, a determination of the vice president, thus giving OIRA a way of addressing delaying tactics by agencies.236 But it may well be that a decision is not ripe for review by the president (or the vice president) within the ninety-day period, or the president may have other more pressing tasks to which he must attend (including other rules to review).

To be sure, Congress can set a deadline for regulation that the administration must respect.237 In that event, however, Congress has managed to get a specific timing directive approved through bicameralism and presentment and the deadline will have survived the tricameral procedure established by the Constitution. Consequently, there is less need for a regulatory process to substitute for the absence of tricameralism.

4. Revising Cost-Benefit Analysis. The regulatory review pro- cess should be more explicit about cost-benefit analysis. The Constitution established a structure for the production of public goods that the family and the market cannot provide. This view of the Constitution necessarily means that government should not regulate when the [*pg 956] market is superior. The regulatory review order should be more explicit about this value of cost-benefit analysis.

Moreover, compared to President Reagan's order, President Clinton's regulatory review order is not quite as clear that the cost-benefit analysis should be its sole focus.238 This blurring should be cured in a revision. President Clinton's order is not wrong in taking into account soft as well as hard costs, but to be effectively balanced even soft costs should be given as accurate an approximation as possible. An agency must make choices and must tote up benefits and costs as best it can, and a single scale would facilitate this process. Moreover, a single scale is likely to restrain interest groups more effectively because they cannot hide the costs of regulations as easily.239 A clearer focus also will make it easier to compare OIRA's actions across regulations, preventing the president from using differing standards to reward his own factions.

In particular, President Clinton's order's focus on the distributive impacts and equity of regulations,240 if by that it means their redistributive possibilities, is wrongheaded. Insofar as a nation wants to engage in redistribution, the tax system is almost always a more efficient method than regulation. As Professors Steven Shavell and Louis Kaplow demonstrate, an inefficient regulation that redistributes money always can be replaced by an efficient regulation and a change in taxes that delivers the same amount of redistribution.241 The latter combination will be pareto-optimal because the inefficient regulation makes the same distortion in the tradeoff between labor and leisure as the taxes and adds its own inefficiencies as well.242 Regulation for redistribution is rarely, if ever, a public good, and this avenue for spe- [*pg 957] cial interest rent seeking should be blocked when the president has the discretion under statute to block it.

A revised order also should make clear that regulation should consider the costs that regulations themselves bring about.243 Cost-benefit analysis should employ a dynamic scoring of costs that takes account of the more substantial opportunities for rent seeking that result from centralized regulation. There will be cases in which, on a first-order analysis, regulations may be useful to prevent an externality such as pollution. Nevertheless, the regulation itself may cause more externalities than it cures. For instance, it may generate a future opportunity for one company or industry to use the regulatory regime that is created to raise the costs borne by its competitors.244 These public choice problems should be highlighted by the new regulatory review order because they represent some of the costs that the high hurdles of tricameralism in the original Constitution avoided.

5. Revising Disclosure Requirements. The revised order should retain the requirements (initiated in President Clinton's regulatory review order) that OIRA disclose contacts from parties outside the administration about the regulation.245 This disclosure rule helps restrain special interests, thus making the order a more effective substitute for other constitutional structures that have been dissolved. As discussed here, interest groups want to raise the information costs of their opponents.246 One way of doing so, among others, is to keep secret the nature of contacts and arguments for their regulatory rents. If these are disclosed, it is easier for political entrepreneurs to mobilize the diffuse citizenry to oppose rent seeking. Although special interests provide social benefits insofar as they bring information to the attention of the agency that might not otherwise be gathered,247 a disclosure requirement will not substantially decrease the information that [*pg 958] interest groups provide.248 Because the regulatory review order should help assure that the federal government provides public interest goods rather than special interest goods, the disclosure requirement for outside contacts was a salutary revision to President Reagan's regulatory review order.

In contrast, the requirements for internal executive branch disclosure introduced by the Clinton order should be deleted.249 These requirements do not constrain special interests by bringing their activity to the light of day. Instead, they curtail the candor of executive branch deliberations, raising the president's monitoring costs. They may perversely strengthen the hands of special interests because special interests have the resources to mount publicity campaigns that will quote documents out of context and spin the initial deliberations to make OIRA look bad.250 In attempting to avoid this danger, OIRA may pull its punches and reduce the searching nature of its analysis of agency regulations.

Two frequently voiced objections to permitting secret deliberations among OIRA and the agencies seem less powerful once the regulatory review process is seen as a substitute for the original Constitution.251 First, it has been objected that such deliberations undermine review on the record because the judiciary may be unable to determine the real reasons that underlie an agency's decision to enact a particular regulation.252 One response is that the agencies still will have to comply with the Administrative Procedure Act: regardless of their communications with OIRA, their record must support the regulation in order to survive arbitrary and capricious review by the [*pg 959] judiciary.253 But under the conceptualization of the regulatory review process proposed in this Essay, judicial review has proved a poor substitute for constitutional restraints. By filtering regulations in the public interest from special interest regulations, the regulatory review process accomplishes the ends of the original Constitution. Insofar as the disclosure of executive branch deliberations would detract from this filtering, it undermines the essential role of the regulatory review process.

Second, it also is objected that secret deliberations between OMB and the agencies detract from public participation because there is no opportunity to respond in the notice and comment pro- cess.254 But the concept of regulatory review offered in this Essay suggests that regulatory review counteracts the special interest domination of the notice and comment process. While the notice and comment process is styled as public participation, few members of the diffuse public actually contribute to that process relative to special interest participation. In contrast to the notice and comment mechanism, the agents of OIRA better represent the president who, in turn, was elected by the public at large.

CONCLUSION
I do not claim that even revised executive orders can wholly compensate for the decline of the original Constitution. Indeed, it is my view that other institutions, like the World Trade Organization and the Supreme Court, must play important compensatory roles in restraining excessive and special interest government regulation in the modern world.255 Moreover, these executive orders address only the regulatory side of the modern administrative state and do not seek to dissolve the welfare state that also has transformed the polity that the Framers bequeathed to us. In recognizing the distinction between matters of allocation (to be addressed by regulation) and matters of distribution (to be addressed by legislative transfers), the executive orders simply reflect the economic consensus that regulation [*pg 960] is an inefficient way of addressing distribution compared with the tax system.256

The federalism and regulatory review executive orders, even revised according to these suggestions, will not and should not end the debate about the proper scope of government regulation. As Professor Elena Kagan shows, President Clinton deployed the regulatory review process for some regulatory projects that advanced the liberal agenda.257 But this use is not surprising. Even within the confines of the original Constitution's charter for limited federal regulation, Democrats and Whigs battled over the appropriate role for the federal government, in part because they estimated the costs and benefits of state intervention differently.258

But, unlike some other commentators, I do not see the presidential review process simply as a technocratic one designed to create better coordination within the executive branch259 or to advance the undoubted virtue of government accountability.260 Instead, the federalism and regulatory review orders replace New Deal norms concerning the appropriate government structure with norms that are closer to those of the Framers and that more effectively restrain the special interests that seek to live off the modern administrative state.

There are broader lessons from this analysis of the regulatory review and federalism executive orders. If these orders essentially compensate for the decline of certain aspects of the original Constitution, it shows that our constitutional norms can reassert themselves other than through the judicial or amendment process. One of the advantages of a system of separated powers is that, even if some branches fail to honor the constitutional norms that lead to the production of public goods, other actors may have an interest in reformulating these norms. I have argued that the president is most likely to protect the [*pg 961] structures that are aimed at screening out special interest regulations because he benefits most from economic growth.

Moreover, the analysis offered here also suggests that there are routes to the restoration of legal norms that do not simply revive the original Constitution. If we are to restore the Framers' principles of government in a world that they could not have imagined and in which their governmental framework has been distorted substantially, it may not be possible or effective to revive the exact replica of the Framers' design. Thus, new ways must be sought to reconstitute a limited interlocking structure of state and federal government that efficiently produces necessary public goods. The federalism and regulatory review executive orders can be an important part of this process of reconstituting the polity and reclaiming the system of government from the special interests that have been empowered by constitutional decline in the twentieth century.

FOOTNOTES

Copyright ? 2001 by John O. McGinnis. ? Professor of Law, Benjamin N. Cardozo Law School. This Essay is based on a paper presented at the Duke Law Journal's Administrative Law Conference held at Duke University School of Law on March 5, 2001. The Duke University School of Law's Program in Public Law provided additional financial support for the 2001 conference. The theory of tricameralism as an implicit supermajority rule offered here was conceived in John O. McGinnis & Michael B. Rappaport, Our Supermajoritarian Constitution, 80 TEX. L. REV. (forthcoming 2002) (manuscript on file with the Duke Law Journal). Professor Rappaport thus shares the credit (such as it is) for those ideas, and I am grateful for his collaboration. Thanks to participants at the annual administrative law conference at Duke Law School, at the New York City Bar Association's Administrative Law section, and at the American Bar Association's Annual Symposium on Administrative Law, where earlier drafts of this paper were presented. I am grateful to Neal Devins, John Duffy, Nelson Lund, Michael Herz, John Manning, Mark Movsesian, Paul Verkuil, and Jonathan Weiner for their helpful comments.
1. President Reagan's executive order on regulatory review, Exec. Order No. 12,291, 3 C.F.R. 127 (1981), was revoked by President Clinton's regulatory review order, Exec. Order No. 12,866, 3 C.F.R. 638 (1994), reprinted in 5 U.S.C. ?601 (2000). President Reagan's executive order on federalism, Exec. Order No. 12,612, 3 C.F.R. 252 (1987), was revoked by President Clinton's federalism order, Exec. Order No. 13,132, 3 C.F.R. 206 (1999), reprinted in 5 U.S.C. ?601 (2000).
2. See infra notes 33-34 and accompanying text.
3. For instance, in United States v. Morrison, 529 U.S. 598 (2000), Chief Justice Rehnquist was careful to cite Wickard v. Filburn, 317 U.S. 111 (1942), the case that symbolizes Congress's plenary authority over economic matters, favorably. Morrison, 529 U.S. at 610. But see United States v. Hickman, 179 F.3d 230, 235 (5th Cir. 1999) (Higginbotham, J., dissenting) (contending that noneconomic activity should not be aggregated even if the activity has economic effects), cert. denied, 530 U.S. 1203 (2000).
4. Throughout this Essay, the "federalism executive order" and the "federalism order" refer to President Clinton's federalism order, Exec. Order No. 13,132, 3 C.F.R. 206 (1999), reprinted in 5 U.S.C. ?601 (2000), which replaced President Reagan's federalism order, Exec. Order No. 12,612, 3 C.F.R. 252 (1987).
5. Likewise, throughout this Essay the "regulatory review executive order" and the "regulatory review order" refer to President Clinton's regulatory review order, Exec. Order No. 12,866, 3 C.F.R. 638 (1994), reprinted in 5 U.S.C. ?601 (2000), which replaced President Reagan's regulatory review order, Exec. Order No. 12,291, 3 C.F.R. 127 (1981).
6. See infra notes 75, 127 and accompanying text.
7. See infra notes 148-49 and accompanying text.
8. Exec. Order No. 12,291 ?2, 3 C.F.R. 127, 128; Exec. Order No. 12,866 ?1(b)(6), 3 C.F.R. 638, 638.
9. Michael Rappaport and I discuss these implicit structural objectives at greater length in John O. McGinnis & Michael B. Rappaport, Our Supermajoritarian Constitution, 80 TEX. L. REV. (forthcoming 2002) (manuscript on file with the Duke Law Journal).
10. The president can direct the agency only in the exercise of discretion the law provides it. Thus, Congress could prohibit an agency from taking the benefits of jurisdictional competition into account in framing its regulations. Cf. James T. Hamilton & W. Kip Viscusi, The Benefits and Costs of Regulatory Reforms for Superfund, 16 STAN. ENVTL. L.J. 159, 160 (1997) (discussing the possibility that superfund legislation could forbid cost-benefit analysis and thus make it impossible for the regulatory review process established by the executive order to deploy cost-benefit analysis).
11. The understanding of competitive federalism as the lodestar of the federalism order also provides for a new perspective on the regulatory review order's approach to preemption. See infra notes 191-222 and accompanying text. The past federalism executive orders have required agencies to find a clear statement before they adopt an interpretation that requires or permits them to promulgate a regulation that conflicts with state law. See infra note 191 and accompanying text. This rationale for a clear statement requirement in the preemption arena, however, is based on a jurisprudential view that conflicts with that of the Supreme Court and has been challenged by recent scholarship assessing the original understanding of the Constitution. Regardless of the resolution of this jurisprudential debate, however, the president's interest in promoting competitive federalism provides a policy rationale for a more nuanced but still potent clear statement rule applicable to all regulatory initiatives that would interfere with regulatory competition by displacing the state regulatory role. In contrast, when a federal regulation comports with competitive federalism, as when it addresses spillovers among state jurisdictions, the clear statement rule need not apply.
12. See infra notes 243-44 and accompanying text.
13. This activity is frequently called "rent seeking" in the public choice literature, particularly when the redistribution is in the form of monopoly profits from regulation and tariffs. DENNIS C. MUELLER, PUBLIC CHOICE II, at 229, 235 (1989).
14. Barry Weingast, The Economic Role of Political Institutions: Market-Preserving Federalism and Economic Development, 11 J.L. ECON. & ORG. 1, 24-28 (1995).
15. See Richard A. Epstein, Exit Rights Under Federalism, 55 LAW & CONTEMP. PROBS. 147, 150 (Winter 1992) (stating that federalism introduces intergovernmental competition because of the continuing exit threat that individuals will move to another state).
16. For an excellent discussion of this thesis, see generally Richard L. Revesz, Rehabilitating Interstate Competition: Rethinking the "Race-to-the-Bottom" Rationale for Federal Environmental Regulation, 67 N.Y.U. L. REV. 1210 (1992) (challenging the argument that competitive federalism serves as the basis for federal environmental regulations). See also Henry N. Butler & Jonathan R. Macey, Externalities and the Matching Principle: The Case for Reallocating Environmental Regulatory Authority, 21 YALE J. ON REG. 23, 25 (1996) (arguing that environmental regulatory authority should be allocated to areas affected by pollution rather than to areas that can harmonize standards). For a more complete discussion of the problem of spillovers in a federal regime, see John O. McGinnis & Mark L. Movsesian, The World Trade Constitution, 114 HARV. L. REV. 511, 561-63 (2000).
17. For a full discussion of this argument, see Walter Oates & Robert M. Schwab, Economic Competition Among Jurisdictions: Efficiency Enhancing or Distortion Inducing?, 35 J. PUB. ECON. 333, 336 (1988). Oates and Schwab assume that individuals cannot move among jurisdictions. Id.
18. Revesz, supra note 16, at 1238. This model of jurisdictional competition provides the answer to those who argue that federal regulation is needed to avoid the "race to the bottom," the claim that competition among the states will cause them to choose suboptimal regulations in order to attract business. Of course, even if there were a race-to-the-bottom problem, federal regulations might not be the best way to address it. Because of the public choice problems previously noted, a centralized institution may be captured by interest groups and consequently choose a bad standard.
19. For a discussion of the reasons that the national political process fails to obtain accurate information about individuals' preferences, see infra notes 80-83 and accompanying text.
20. New State Ice Co. v. Liebmann, 285 U.S. 262, 311 (1932) (Brandeis, J., dissenting) ("It is one of the happy incidents of the federal system that a single courageous State may, if its citizens choose, serve as a laboratory; and try novel social and economic experiments without risk to the rest of the country.").
21. For a discussion of the way associations can perform this function, see MUELLER, supra note 13, at 149-73.
22. See id. at 14 (discussing the distinctions between the formal voting procedures in a democracy and the informal decisionmaking processes of the family).
23. See Michael W. McConnell, Federalism: Evaluating the Founders' Design, 54 U. CHI. L. REV. 1484, 1510 (1987) (examining the Founders' belief that "public virtue" could not be cultivated in a centralized government).
24. 1 ALEXIS DE TOCQUEVILLE, DEMOCRACY IN AMERICA 181 (Henry Reeve trans., Schocken Books 1961) (1835).
25. See Weingast, supra note 14, at 8-9 (noting that federalism facilitated the development of America's large common market).
26. See, e.g., John O. McGinnis & Michael B. Rappaport, Supermajority Rules as a Constitutional Solution, 40 WM. & MARY L. REV. 365, 390-91 (1999) (discussing the decline of federalism); John O. McGinnis, The Original Constitution and Its Decline: A Public Choice Perspective, 21 HARV. J.L. & PUB. POL'Y 195, 204 (1997) (same).
27. McGinnis & Rappaport, supra note 26, at 391; see Jay S. Bybee, Ulysses at the Mast: Democracy, Federalism, and the Sirens' Song of the Seventeenth Amendment, 91 NW. U. L. REV. 500, 557 (1997) (noting that the Seventeenth Amendment greatly increased the independence of senators from state legislatures).
28. See Roger G. Brooks, Comment, Garcia, the Seventeenth Amendment, and the Role of the Supreme Court in Defending Federalism, 10 HARV. J.L. & PUB. POL'Y 189, 206-07 (1987) (discussing the Seventeenth Amendment's crippling effect on senators' interest in protecting the states).
29. I discuss this point further infra notes 127-32 and accompanying text.
30. See, e.g., Carter v. Carter Coal Co., 298 U.S. 238, 310 (1936) (holding that the wage provisions of the Bituminous Coal Conservation Act of 1935 exceeded Congress's power under the Commerce Clause).
31. For a description of this process, see Richard A. Epstein, The Proper Scope of the Commerce Power, 73 VA. L. REV. 1387, 1443-54 (1987).
32. See Grant S. Nelson & Robert J. Pushaw, Jr., Rethinking the Commerce Clause: Applying First Principles to Uphold Federal Commercial Regulations but Preserve State Control over Social Issues, 85 IOWA L. REV. 1, 85 (1999) (discussing the plenary power that the New Deal Court established for Congress in economic matters).
33. See Richard A. Posner, Past-Dependency, Pragmatism, and Critique of History in Adjudication and Legal Scholarship, 67 U. CHI. L. REV. 573, 585 (2000) (suggesting that the House of Representatives, the Senate, and the president with his veto power create "an essentially tricameral legislature").
34. See John F. Manning, Textualism and the Equity of the Statute, 101 COLUM. L. REV. 1, 72-74 (2001) (observing that bicameralism and presentment ensure that a self-interested group is unable to capture the legislative process); see also JAMES M. BUCHANAN & GORDON TULLOCK, THE CALCULUS OF CONSENT 235-36 (1962) (theorizing that the president will use his veto power to reject legislative bargains that do not have the approval of the majority of voters).
35. Hamilton, for instance, focused on the self-regarding laws that the veto would prevent:
[I]t furnishes an additional security against the enaction of improper laws. It establishes a salutary check upon the legislative body, calculated to guard the community against the effects of faction, precipitancy, or of any impulse unfriendly to the public good, which may happen to influence a majority of that body.
THE FEDERALIST NO. 73, at 443 (Alexander Hamilton) (Clinton Rossiter ed., 1961) (emphasis added). Joseph Story made similar points about bicameralism. JOSEPH STORY, COMMENTARIES ON THE CONSTITUTION OF THE UNITED STATES ?276 (Carolina Academic Press 1987) (1833) ("[Bicameralism] operates indirectly as a preventive to attempts to carry private, personal, or party objects, not connected with the common good.").
36. See William T. Mayton, The Possibilities of Collective Choice: Arrow's Theorem, Article I, and the Delegation of Legislative Power to Administrative Agencies, 1986 DUKE L.J. 948, 956 (arguing that Congress's bicameral structure induces a supermajoritarian system).
37. See McGinnis & Rappaport, supra note 9 (manuscript at 60) (analyzing the way in which Congress's bicameral structure affects the political cost of legislation and the power of special interest groups).
38. For further discussion, see id. (manuscript at 9-10). The Framers never argued explicitly that bicameralism was a supermajority rule but they certainly recognized and welcomed the higher hurdle it would create for legislation. One of the Framers' reasons for bicameralism was to prevent a one-house majority from making hasty and unwise decisions. Indeed, even before the framing of the Constitution, critics of unicameral chambers complained that a single legislature was likely to produce bad legislation because of passion and haste. See, e.g., Letter from William Hooper to the Congress of the State of North Carolina (Oct. 22, 1776), in 10 COLONIAL RECORDS OF NORTH CAROLINA 862, 867 (William L. Saunders ed., 1890) ("A single branch of Legislation is a many headed Monster which without any check must soon defeat the very purposes for which it was created, and its members become a Tyranny dreadful in proportion to the numbers which compose it."). The Framers also were express in their fears that a single body representing a popular majority would sometimes pass unwise legislation in a moment of political passion and therefore wanted to impose a requirement of two separate processes to slow down the legislation and permit passion to cool. See, e.g., THE FEDERALIST NO. 62, at 379 (James Madison) (Clinton Rossiter ed., 1961) (noting "the propensity of all single and numerous assemblies to yield to the impulse of sudden and violent passions, and to be seduced by factious leaders into intemperate and pernicious resolutions").
39. McGinnis & Rappaport, supra note 9 (manuscript at 9).
40. Id. (manuscript at 10-11).
41. Id.
42. INS v. Chadha, 462 U.S. 919, 951 (1983) ("The President's participation in the legislative process was to protect . . . the whole people from improvident laws."); see also Manning, supra note 34, at 73 n.290 (suggesting that the veto serves as a particularly good check on political factions because the president is elected by the entire nation).
43. See McGinnis & Rappaport, supra note 9 (manuscript at 25) (discussing the ways in which supermajority rule can improve upon majority rule).
44. See id. (discussing the complications inherent in the simplifying assumptions). That, in turn, depends on both the efficiency of legislation under majority rule and the degree to which the supermajority rule will filter bad legislation from good. Id.
45. The discussion that follows offers some reasons to believe that the pressure for special interest regulation is greater than for special interest deregulation. I have not found a suitable economic model of this important issue in the literature. An alternative way of suggesting that the pressure for special interest regulation may be stronger than the special interest pressure against regulation is to observe that, in a competitive economy, all special interests have an interest in regulation that disadvantages their competitors. See William E. Nelson, The Growth of Distrust: The Emergence of Hostility Toward Government Regulation of the Economy, 25 HOFSTRA L. REV. 1, 5 (1996) (noting that regulatory schemes can favor some special interests by burdening competitors with added costs). Only special interests in industries that create externalities have an interest in special interest deregulation, namely in eliminating regulations that are in the public interest.
46. William A. Niskanen, Jr., Bureaucrats and Politicians, 18 J.L. & ECON. 617, 618 n.4 (1975).
47. See THE FEDERALIST NO. 10, at 80-81 (James Madison) (Clinton Rossiter ed., 1961); see also Larry Kramer, Madison's Audience, 112 HARV. L. REV. 611, 612 (1999) (arguing that Madison sought to "[e]xtend the republic's sphere to include a greater number of parties and territory" because he believed that "the very diversity of the society will provide a margin of safety by, in effect, disabling any group from capturing or controlling government").
48. MANCUR OLSON, THE LOGIC OF COLLECTIVE ACTION: PUBLIC GOODS AND THE THEORY OF GROUPS 137-48 (1977) (describing trade and business lobbies).
49. For a more complete discussion of special interest groups, see infra notes 80-83 and accompanying text.
50. See, e.g., Daniel C. Esty, Toward Optimal Environmental Governance, 74 N.Y.U. L. REV. 1495, 1530 (1999) (discussing the asymmetry of the political process in which special interests oppose environmental laws).
51. Of course, this balance may not be accurate for all regulation. Therefore, bicameralism ultimately must be justified because: (1) the hurdles it places in the way of regulation are in the public interest overall; and (2) it is impossible to define a more limited category of legislation to which bicameralism would apply. Defining a more refined category-which would presumably attempt to identify proposed legislation that enjoys disproportionate support from regulatory, majoritarian, and special interests and would subject only that category of legislation to bicameralism-would raise the administrative costs of bicameralism.
52. See Barry Friedman, Valuing Federalism, 82 MINN. L. REV. 317, 373-75 (1997) (discussing the forces motivating national regulation). The interests in favor of inefficient regulation may well have become even stronger than they were in the Framers' time. First, the interest of bureaucrats in favor of excessive regulation may be stronger because the bureaucracy is larger than it was. Second, the interests that want to avoid jurisdictional competition no longer face the obstacle of the federal government's enumerated powers. For more on the growing strength of these interests, see McGinnis & Rappaport, supra note 26, at 394-95. Thus, the case for bicameralism for laws that apply additional regulations may be stronger than it previously was because there are reasons to suspect that the mix of regulations under majority rule may be more inefficient overall than it was at the time of the Framers.
53. For a discussion of the impact of the Constitution's supermajority voting rules on interest groups, see Jonathan R. Macey, Competing Economic Views of the Constitution, 56 GEO. WASH. L. REV. 50, 56, 72-73 (1987); Jonathan R. Macey, Promoting Public-Regarding Legislation Through Statutory Interpretation: An Interest Group Model, 86 COLUM. L. REV. 223, 248 (1986) (contending that the Constitution's supermajority voting rule raises decision costs for special interest groups).
54. Our political system generally operates as if supermajority rules filter out undesirable enactments from desirable enactments. For instance, because constitutional provisions are so important, the Constitution imposes a dual supermajority rule to screen desirable amendments from undesirable amendments:
The Congress, whenever two thirds of both Houses shall deem it necessary, shall propose Amendments to this Constitution, or, on the Application of the Legislatures of two thirds of the several States, shall call a Convention for proposing Amendments, which, in either Case, shall be valid to all Intents and Purposes, as part of this Constitution, when ratified by the Legislatures of three fourths of the several States, or by Conventions in three fourths thereof, as the one or the other Mode of Ratification may be proposed by the Congress . . . .
U.S. CONST. art. V.
55. Certainly, there are costs imposed by supermajority rules. Perhaps the most important are holdout costs-the costs of empowering minorities who may act strategically in opposing efficient legislation. But holdout costs are not likely to be very high under a mild supermajority rule, particularly because most regulatory legislation is unlikely to be very time sensitive. See McGinnis & Rappaport, supra note 26, at 420-22 (discussing holdout costs). Holdouts lose their leverage if there is a substantial possibility that the bill can be passed in a subsequent Congress when membership has changed slightly. Id. at 422. Administrative costs-the costs of interpreting the supermajority rule-are also low because the rule applies to all legislation and does not require much judicial interpretation.
56. THE FEDERALIST NO. 73, supra note 35, at 444 (Alexander Hamilton).
57. Id.
58. The role of delegation in the circumvention of tricameralism is discussed at length in McGinnis & Rappaport, supra note 9, and the description here is an excerpt and summary of that larger analysis. Here, however, the problem that circumvention causes for presidential control is discussed in substantially greater depth, because presidential control is obviously of substantial relevance to the regulatory review and federalism executive orders.
59. See John F. Manning, Textualism as a Nondelegation Doctrine, 97 COLUM. L. REV. 673, 726 (1997) ("Incidental to discharging their respective constitutional functions of law execution and adjudication, agencies . . . must expound the meaning of the texts they implement, leaving more or less room for the exercise of discretion."); Cass R. Sunstein, Nondelegation Canons, 67 U. CHI. L. REV. 315, 326 (2000) ("The distinction between 'executive' and 'legislative' power cannot depend on anything qualitative; the issue is a quantitative one.").
60. I do not claim that the only reason for delegation has been circumvention of bicameralism and the presidential veto. There may well be bona fide reasons for delegation, such as Congress's need to rely on the greater expertise of executive agencies. The only propositions here are that special interests, over time, will seek to circumvent the constitutional provisions designed to constrain them and that delegation has sometimes been used to accomplish this end.
61. The academic literature on the ability of interest groups to capture agencies is vast. See, e.g., Richard B. Stewart, The Reformation of American Administrative Law, 88 HARV. L. REV. 1669, 1684-87 (1975) (providing a summary of the discussion of agency capture). Here I do not make the claim that special interests always capture agencies; I only assert the more modest claim that, as a general matter, special interests are more likely to capture agencies than they are to capture all the entities in the tricameral process. See Jonathan R. Macey, Transaction Costs and the Normative Elements of the Public Choice Model: An Application to Constitutional Theory, 74 VA. L. REV. 471, 513 (1988) (describing the relative difficulties that interest groups face when attempting to capture the legislative process).
62. Rational ignorance describes the systematic tendency of diffuse citizens to pay little attention to political information. The phenomenon occurs because acquiring political information is both costly and unproductive. It is costly because, to acquire such information, individuals must invest time that they could be using for other more lucrative or pleasurable enterprises. Likewise, it is unproductive because the principal use of such information is to guide voting, but the vote of any individual is unlikely to influence the outcome of an election. MUELLER, supra note 13, at 205-06.
63. Members of Congress have no incentives to police such delegations because the delegations frequently are to their advantage. Such delegation allows Congress to satisfy interest groups while shifting responsibility to agencies for rules that may flout the public interest. Peter H. Aranson et al., A Theory of Legislative Delegation, 68 CORNELL L. REV. 1, 57 (1982). It also allows them to avoid conflict about which interests to satisfy. Id. at 61.
64. See Gary Lawson, The Rise and Rise of the Administrative State, 107 HARV. L. REV. 1231, 1248-49 (1994) ("Administrative agencies routinely combine all these governmental functions in the same body, and even in the same people within that body.").
65. See Elena Kagan, Presidential Administration, 114 HARV. L. REV. 2245, 2327-28 (2001) (describing the normal practice of delegation to agencies). Of course, sometimes Congress delegates power to the president himself on the understanding that he may then delegate the power to a subordinate official. Id. at 2329.
66. See Richard H. Pildes & Cass R. Sunstein, Reinventing the Regulatory State, 62 U. CHI. L. REV. 1, 25 (1995) ("[T]he President has no authority to make the decision himself, at least if Congress has conferred the relevant authority on an agency head."); Peter L. Strauss, The Place of Agencies in Government: Separation of Powers and the Fourth Branch, 84 COLUM. L. REV. 573, 665-66 (1984) (suggesting that the authority to issue regulations remains with the agency head, subject to "whatever political discipline the President might bring to bear"). For a challenge to the conventional wisdom, see Kagan, supra note 65, at 2326-27 (arguing that the president has the authority to revise decisions of subordinates without their consent so long as Congress has not legislated clearly to the contrary).
I recognize that some have argued that the president, instead of the agency head to whom Congress regulates authority, has the constitutional authority to issue regulations. E.g., Steven G. Calabresi & Saikrishna Prakash, The President's Power to Execute the Laws, 104 YALE L.J. 541, 594-95 (1994). Although I have some sympathy with this view as a matter of first principles, it has no support in Supreme Court case law and its leading proponents concede that it requires more academic analysis than it has received to be fully persuasive. Id. at 595 n.208. Thus, on the conventional understanding of the president's power over agencies, which relies on his supervisory and removal authority rather than his authority to displace agency regulations directly, I argue that the regulatory review executive order restores presidential power to something more akin to his veto power. If the argument of Professors Calabresi and Prakash is accepted, the president always has had a power over administrative regulation that would, if exercised, compensate him for the loss of the veto under broad delegations that have characterized the modern administrative state.

67. See Kagan, supra note 65, at 2273 (describing the problem of presidential monitoring as a "typical principal-agent dilemma: how to ensure against slippage between the behavior the principal desires from the agent and the behavior the principal actually receives, given the agent's own norms, interests, and informational advantages").
68. See Pildes & Sunstein, supra note 66, at 25 (discussing the political costs to the president of firing agency heads).
69. See Daniel B. Rodriguez, The Positive Political Dimensions of Regulatory Reform, 72 WASH. U. L.Q. 1, 88 (1994) (noting that the president faces an "ubiquitous principal-agent problem").
70. See John O. McGinnis, Principle Versus Politics: The Solicitor General's Office in Constitutional and Bureaucratic Theory, 44 STAN. L. REV. 799, 809 (1992) (discussing the reasons for the decline of Cincinnatus's ideal as the model for government service).
71. See Bruce Ackerman, The New Separation of Powers, 113 HARV. L. REV. 633, 708 (2000) (discussing the power of the "revolving door" in which political appointees stay in government for approximately two years and then move to the private sector).
72. For a discussion of the importance of networks and revolving doors in executive agencies, see Cary Coglianese, Assessing Consensus: The Promise and Performance of Negotiated Rulemaking, 46 DUKE L.J. 1255, 1330 (1997).
73. For this reason, I reject the argument of those who suggest that the oversight process helps create a process that approximates the function of bicameralism. Because committees are much more easily captured by particular special interests than Congress is as a whole, and because the membership of congressional committees is often quite unrepresentative of Congress, committee oversight is less likely than the high hurdles of bicameralism to be a guardian of the public interest. See Einer R. Elhauge, Does Interest Group Theory Justify More Intrusive Political Review?, 101 YALE L.J. 31, 42 (1991) ("[The] committee structure can exacerbate interest group influence.").
74. It might be argued that special interest groups have similar leverage over the president's decision whether to sign or veto a bill. Certainly, special interests have disproportionate leverage at the presidential signing stage of the political process as well, but the disproportion is even greater at the agency level because agency action is less visible and transparent to the public than the president's actions.
75. See, e.g., Kagan, supra note 65, at 2310 ("The American public harbors high and rising expectations about what a President should be able to accomplish.").
76. See generally, e.g., CHARLES MURRAY, LOSING GROUND (1984) (describing the failures of antipoverty programs launched in the 1960s).
77. See MUELLER, supra note 13, at 406 (discussing modern theorems that "raise fundamental questions about the possibility of satisfying collective choice procedures establishing minimally appealing normative properties").
78. OLSON, supra note 48, at 132-34. Special interest groups exert their influence in the political process in a variety of ways. First, they are able to monitor what is actually transpiring in the political process. See, e.g., Michael A. Andrews, Tax Simplification, 47 SMU L. REV. 37, 42 (1993) (discussing how special interest groups have affected the process of tax legislation). Second, special interests often can use experts to persuade legislators of their position. See, e.g., Jonathan R. Macey, Special Interest Groups Legislation and the Judicial Function: The Dilemma of Glass-Steagall, 33 EMORY L.J. 1, 17-21 (1984) (describing the role of investment bankers in persuading Congress to enact legislation prohibiting commercial banks from competing with investment banks). Third, special interests are able to conduct a coordinated and coherent campaign to publicize their position in the media. See, e.g., William Dubinsky, Book Note, 90 MICH. L. REV. 1512, 1514 (1992) (reviewing DANIEL A. FARBER & PHILIP P. FRICKEY, LAW AND PUBLIC CHOICE (1991)) ("Special interest groups can provide legislators . . . with pub- licity . . . ."). Finally, special interests also may exercise greater leverage over legislators through campaign contributions or independent political expenditures. See, e.g., Daniel H. Lowenstein, Political Bribery and the Intermediate Theory of Politics, 32 UCLA L. REV. 784, 826-28 (1985) (evaluating the power of special interest group contributions).
79. See supra note 62.
80. See Peter H. Schuck, Against (and For) Madison: An Essay in Praise of Factions, 15 YALE L. & POL'Y REV. 553, 580 (1997) (explaining how special interest groups inform and represent their members).
81. See Frank H. Easterbrook, The State of Madison's Vision of the State: A Public Choice Perspective, 107 HARV. L. REV. 1328, 1336-37 (1994) (discussing the free-rider problem for factions).
82. Id.
83. John O. McGinnis, The Partial Republican, 35 WM. & MARY L. REV. 1751, 1791 (1994) (reviewing CASS R. SUNSTEIN, THE PARTIAL CONSTITUTION (1993)).
84. See Paul R. Verkuil, Understanding the "Public Interest" Justification for Government Actions, 39 ACTA JURIDICA HUNGARICA 141, 146-47 (1998) (describing the decline in the public's belief that administrative agencies act for the public interest).
85. JAMES LANDIS, THE ADMINISTRATIVE PROCESS 1, 9 (1938). Landis had great confidence in experts to replace the messy system of politics. See id. at 23 (discussing the value of experts). For an excellent discussion of the course of Landis's views, see John F. Duffy, The FCC and the Patent System: Progressive Ideals, Jacksonian Realism, and the Technology of Regulation, 71 U. COLO. L. REV. 1071, 1084-86 (2000).
86. STAFF OF SUBCOMM. ON ADMIN. PRACTICE & PROCEDURE TO THE SENATE COMM. ON THE JUDICIARY, 86TH CONG., REPORT ON REGULATORY AGENCIES TO THE PRESIDENT-ELECT 14 (Comm. Print 1960) (written by James Landis); Duffy, supra note 85, at 1114 (explaining the change in Landis's views).
87. Duffy, supra note 85, at 1115.
88. The classic article is George J. Stigler, The Theory of Economic Regulation, 2 BELL. J. ECON. & MGMT. SCI. 3 (1971) (discussing industrial groups' ability to use the administrative state to their own ends).
89. The most noted Democratic recognition of these changes was former President Bill Clinton's recognition that "the era of big government is over." Alison Mitchell, Clinton Offers Challenge to Nation, Declaring 'Era of Big Government is Over,' N.Y. TIMES, Jan. 24, 1996, at A1 (quoting President Clinton's State of the Union Address).
90. See Ronald Reagan, First Inaugural Address (June 20, 1981), in JAMES M. MCPHERSON, "TO THE BEST OF MY ABILITY": THE AMERICAN PRESIDENTS 442, 443 (2000) ("It is my intention . . . to demand recognition of the distinction between the powers granted to the Federal Government and those reserved to the States . . . .").
91. Exec. Order No. 12,612, 3 C.F.R. 252 (1987), revoked by Exec. Order No. 13,132, 3 C.F.R. 206 (1999), reprinted in 5 U.S.C. ?601 (2000).
92. Id., 3 C.F.R. at 252-53.
93. Id. ?3(b)(1), 3 C.F.R. at 254.
94. Id. ?4(a)-(b), 3 C.F.R. at 255.
95. Id. ?4(d), 3 C.F.R. at 255.
96. Id. ?6(a), 3 C.F.R. at 255.
97. For a description of President Clinton's first federalism order and its suspension, see Jennie Holman Blake, Presidential Power Grab or Pure State Might? A Modern Debate over Executive Interpretations of Federalism, 2000 BYU L. REV. 293, 299-301.
98. Exec. Order No. 13,132, 3 C.F.R. 206 (1999), reprinted in 5 U.S.C. ?601 (2000).
99. Id. ?3(b), 3 C.F.R. at 208.
100. Id.
101. Id. ?4(a)-(b), 3 C.F.R. at 208.
102. Id. ?6(a), 3 C.F.R. at 209.
103. Id. ?7(a)-(d), 3 C.F.R. at 210.
104. See Reagan, supra note 90, at 443 ("It is my intention to curb the size and influence of the Federal establishment . . . .").
105. Exec. Order No. 12,291, 3 C.F.R. 127 (1981), revoked by Exec. Order No. 12,866, 3 C.F.R. 638 (1994), reprinted in 5 U.S.C. ?601 (2000).
106. Id. ?3, 3 C.F.R. at 128-30. Although Executive Order 12,291 requires agency heads to submit major regulatory proposals to the director of OMB, OIRA carries out OMB's functions under the executive order. For a description of OIRA's role in the process that is governed by Executive Order 12,291, see generally Harold H. Bruff, Presidential Management of Agency Rulemaking, 57 GEO. WASH. L. REV. 533 (1989).
107. Exec. Order No. 12,291 ?3, 3 C.F.R. at 128-30.
108. Id. ?3(d), 3 C.F.R. at 129.
109. Exec. Order No. 12,498, 3 C.F.R. 323 (1985).
110. See Richard Benedetto, Dem Group Prods Clinton Toward Center, USA TODAY, Dec. 3, 1993, at 4A (discussing the manner in which President Clinton had campaigned as a supporter of a centrist government opposed to big government solutions to social problems).
111. Exec. Order No. 12,866, 3 C.F.R. 638 (1994), reprinted in 5 U.S.C. ?601 (2000).
112. Id. ?1(a), 3 C.F.R. at 638-39.
113. See, e.g., Michael Fix & George C. Eads, The Prospects for Regulatory Reform: The Legacy of Reagan's First Term, 2 YALE J. ON REG. 293, 315 (1985) (discussing the controversy over the secrecy of the OIRA regulatory review process).
114. Exec. Order No. 12,866 ?6(b)(4)(A), 3 C.F.R. at 647.
115. Id. ?6(b)(4), 3 C.F.R. at 647-48.
116. Id. ?6(b)(4)(D), 3 C.F.R. at 648.
117. See, e.g., Bruff, supra note 106, at 567 (discussing various criticisms of OIRA's ability to delay agency action).
118. Exec. Order No. 12,866 ?6(b)(2), 3 C.F.R. at 646-47 (requiring OIRA to review proposed regulations within ninety days and providing for one thirty-day extension).
119. Id. ?7, 3 C.F.R. at 648.
120. See Steven G. Calabresi, Some Normative Arguments for the Unitary Executive, 48 ARK. L. REV. 23, 35 (1995) (discussing the incentives for legislators to bring home pork).
121. The existence of unfunded mandates-federal orders to state and local governments to provide services without corresponding funds-provides powerful evidence that federal officials have a weak interest in the autonomy of state governments. See generally Edward A. Zelinsky, Unfunded Mandates, Hidden Taxation, and the Tenth Amendment: On Public Choice, Public Interest, and Public Service, 46 VAND. L. REV. 1355 (1993) (offering various reasons for why federal legislators do not resist unfunded mandates).
122. The loss of state autonomy is a rather diffuse phenomenon that is not traceable to federal legislators. See R. DOUGLAS ARNOLD, THE LOGIC OF CONGRESSIONAL ACTION 18-19 (1972) (demonstrating that legislators pay attention to actions that are traceable to them).
123. See Lee Anderson, Oink, Oink, It's Pork Time in Washington, CHATTANOOGA TIMES, Sept. 27, 2000, at B7 (explaining that the operative principle for pork-barrel spending is "[y]ou scratch my back and I'll scratch yours").
124. For a discussion of the dissolution of constitutional federalism, see supra notes 13-32 and accompanying text.
125. The advantages of this system are described supra notes 16-21 and accompanying text.
126. I am a little skeptical of those arguments. The key to representing states' interests in the original Constitution was not simply having the representatives elected from the states but having some of those representatives elected by state legislatures-a constituency powerfully focused on state autonomy. See supra notes 27-29 and accompanying text.
127. See Ray C. Fair, Econometrics and Presidential Elections, J. ECON. PERSP., Summer 1996, at 89-90 (concluding from a study of all American elections since 1916 that the growth of per capita output in the year of the election was the most important factor in determining whether the party of the incumbent president retained the presidency).
128. See JUDITH BEST, THE CASE AGAINST DIRECT ELECTION OF THE PRESIDENT: A DEFENSE OF THE ELECTORAL COLLEGE 156-57 (1971) (describing the importance of state party machinery in presidential elections under the electoral college system).
129. See Mancur Olson, Dictatorship, Democracy, and Development, 87 AM. POL. SCI. REV. 567, 571 (1993) (discussing the advantages that monarchy had over roving bands in generating prosperity). One can think of legislators as behaving a little like roving bands in trying to take back pork (as well as regulations that will help their region) to their constituents at the expense of the commonwealth.
130. Id. Olson was not the first to see the advantages that a monarch had over roving bands or warring city-states in creating the conditions for prosperity. Dante argued that monarchy best led to a polity in which each person could pursue his highest good. See R.W.B. LEWIS, DANTE 128 (2001) (explaining that, in Dante's view, monarchy is a prerequisite for universal peace).
131. See Calabresi, supra note 120, at 67-68 (suggesting that the president is responsive to the majority coalition that placed him in office).
132. See supra note 127.
133. For a discussion of the role of precommitment devices in democratic governance, see JOHN ELSTER, ULYSSES AND THE SIRENS 88-96 (1979).
134. Of course, the president can change his executive order with the stroke of the pen, but this would be embarrassing, and thus an executive order can serve as a precommitment device despite the possibility of waiver. See John O. McGinnis & Michael B. Rappaport, The Constitutionality of Legislative Supermajority Rules: A Defense, 105 YALE L.J. 483, 510-11 (1995) (discussing the value of a supermajority legislative rule applicable to raising taxes in light of the fact that the rule could be waived).
135. It is true that President Clinton's first version of the executive order was criticized particularly by state officials as cutting back on the strong federalism of the Reagan executive order. That order was suspended and the final version was very similar to Reagan's executive order. See Blake, supra note 97, at 299-301; supra notes 97-103 and accompanying text.
136. Exec. Order No. 12,612 ?3(b)(1), 3 C.F.R. 252, 254 (1987), revoked by Exec. Order No. 13,132, 3 C.F.R. 206 (1999), reprinted in 5 U.S.C. ?601 (2000). The order states:
In most areas of governmental concern, the States uniquely possess the constitutional authority, the resources, and the competence to discern the sentiments of the people and to govern accordingly. In Thomas Jefferson's words, the States are "the most competent administrations for our domestic concerns and the surest bulwarks against antirepublican tendencies."
Id. ?2(e), 3 C.F.R. at 253. The order continues, "It is important to recognize the distinction between problems of national scope (which may justify Federal action) and problems that are merely common to the States (which will not justify Federal action because individual States, acting individually or together, can effectively deal with them)." Id. ?3(b)(1), 3 C.F.R. at 254. Exec. Order No. 13,132 ?1(f), 3 C.F.R. 206, 207 (1999) states:
The nature of our constitutional system encourages a healthy diversity in the public policies adopted by the people of the several States according to their own conditions, needs, and desires. In the search for enlightened public policy, individual States and communities are free to experiment with a variety of approaches to public issues. One-size-fits-all approaches to public policy problems can inhibit the creation of effective solutions to those problems.
It continues:
National action limiting the policymaking discretion of the States shall be taken only where there is constitutional and statutory authority for the action and the national activity is appropriate in light of the presence of a problem of national significance. Where there are significant uncertainties as to whether national action is authorized or appropriate, agencies shall consult with appropriate State and local officials to determine whether Federal objectives can be attained by other means.
Id. ?3(b), 3 C.F.R. at 208.
137. See, e.g., United States v. E.C. Knight Co., 156 U.S. 1, 14 (1895):
Manufacture is transformation-the fashioning of raw materials into a change of form for use. The functions of commerce are different . . . . If it be held that the term includes the regulation of all such manufactures as are intended to be the subject of commercial transactions in the future, it is impossible to deny that it would also include all productive industries that contemplate the same thing. The result would be that Congress would be invested, to the exclusion of the States, with the power to regulate, not only manufactures, but also agriculture, horticulture, stock raising, domestic fisheries, mining-in short, every branch of human industry.
138. See Nelson Lund, Federalism and Civil Liberties, 45 U. KAN. L. REV. 1045, 1047 (1997) (describing the difficulties federalism presents in generating an optimal level of efficiency because its costs and benefits are so closely intertwined).
139. See A. Dan Tarlock, Safe Drinking Water: A Federalism Perspective, 21 WM. & MARY ENVTL. L. & POL'Y REV. 233, 251-52 (1997) (questioning whether the Supreme Court's new bright lines in its Commerce Clause decisions will permit federal regulation of the environment).
140. I have previously suggested, however, that when jurisdictional competition is working well, it will be conducive to the optimal amount of national harmonization. McGinnis & Movsesian, supra note 16, at 561.
141. See McGinnis & Movsesian, supra note 16, at 543 n.189 (discussing interest group schemes to obtain more centralized regulatory structures).
142. But such rigid rules may be enforceable only when individuals have strong, elemental attachments to their states-attachments that may have dissipated because of the decrease in information costs and the rise of a mass culture. See John O. McGinnis, Reviving Tocqueville's America: The Rehnquist Court's Jurisprudence of Social Discovery, 90 CAL. L. REV. (forthcoming 2002) (manuscript at 35, on file with the Duke Law Journal) (presenting different explanations for the disappearance of state attachments).
143. Exec. Order No. 12,291 ?3, 3 C.F.R. 127, 128-30 (1981) (providing for centralized review of agency regulations), revoked by Exec. Order No. 12,866, 3 C.F.R. 638 (1994), reprinted in 5 U.S.C. ?601 (2000); Exec. Order No. 12,866 ?6, 3 C.F.R. 638, 644-48 (same).
144. The review is conducted by a separate bureaucracy, OIRA, located in OMB. See Exec. Order No. 12,866 ?6, 3 C.F.R. at 644-48; supra note 106 and accompanying text.
145. As discussed, infra, at notes 162-68 and accompanying text, OIRA is much more directly influenced by the president and is therefore more easily influenced by factions that are influential with presidents.
146. Analogously, the line item veto attempts to increase the power of the presidential veto in appropriation matters because the practice of omnibus bills has reduced the effectiveness of the presidential veto. The version of the line item veto that Congress most recently enacted was invalidated in Clinton v. City of New York, 524 U.S. 417, 448 (1998).
147. Exec. Order No. 12,291 ?2(b), 3 C.F.R. at 128 (mandating cost-benefit analysis); Exec. Order No. 12,866 ?1(b)(6), 3 C.F.R. at 639 (same).
148. See supra notes 127-34 and accompanying text.
149. Some kinds of special interest regulation, such as regulations shaped by polluters that fail to make producers pay the full cost of pollution, can help short-term growth. But in the long term, the costs of rectifying the externalities that regulations have failed to correct will hurt long-term growth.
150. The nondelegation doctrine enforces the constitutional provision that "[a]ll legislative Powers herein granted shall be vested in a Congress of the United States, which shall consist of a Senate and House of Representatives." U.S. CONST. art. I, ?1. Therefore, Congress cannot delegate legislative power to one outside the legislative branch, like the president. Delegation also can be conceived as a violation of the separation of powers. J.W. Hampton, Jr., & Co. v. United States, 276 U.S. 394, 406 (1928) ("[I]n carrying out that constitutional division into three branches it is a breach of the National fundamental law if Congress gives up its legislative power and transfers it to the President, or to the Judicial Branch . . . ."). For an excellent discussion of the roots of the nondelegation doctrine, see Bradford R. Clark, Separation of Powers as a Safeguard of Federalism, 79 TEX. L. REV. 1321, 1373-77 (2001).
151. See, e.g., Ernest Gellhorn, Returning to First Principles, 36 AM. U. L. REV. 345, 353 (1987) (discussing the charge that the nondelegation doctrine represents the political agenda of deregulation).
152. Massachusetts v. Am. Trucking Ass'ns, Inc., 121 S. Ct. 1222, 1222 (2001).
153. The Congressional Review Act creates a greater barrier to impositions on the citizen by making it easier for Congress to block regulations. Small Business Regulatory Enforcement Fairness Act of 1996, Pub. L. No. 104-121, ?251, 110 Stat. 857, 868 (codified at 5 U.S.C. Ё 801-808 (2000)). This Act is a congressional move toward constitutional restoration similar to the presidential move advocated in this Essay.
154. Merrick Garland, Deregulation and Judicial Review, 98 HARV. L. REV. 507, 529 (1985).
155. See Richard J. Pierce, Jr., Justice Breyer: Intentionalist, Pragmatist, and Empiricist, 83 ADMIN. L.J. 747, 751 (1995) (noting Justice Breyer's criticism that judges lack the information to second-guess agencies).
156. Id. (noting Justice Breyer's criticism that the hard look doctrine may unduly hamstring the regulatory process).
157. Theodore P. Seto, Drafting a Federal Balanced Budget that Does What It Is Supposed to Do (And No More), 106 YALE L.J. 1449, 1462 n.38 (1997) (suggesting that, because the major political parties represent different views on public spending and public goods, citizens are able to express their preferences about expenditures on public goods by voting).
158. See supra notes 36-57 and accompanying text.
159. THE FEDERALIST NO. 10, supra note 47, at 78-80 (James Madison); see DAVID A. EPSTEIN, THE POLITICAL THEORY OF THE FEDERALIST 99-107 (1984) (arguing that Madison felt that the structure of the Union would prevent control by dueling factions).
160. See, e.g., Pildes & Sunstein, supra note 66, at 17 ("What is needed instead is a definition of shared goals and objectives, accompanied by good and continuing working relations, clear jurisdictional authority, and mutual investment in the basic goals of improving regulatory performance according to agreed-upon criteria.").
161. Cf. McGinnis & Rappaport, supra note 134, at 503-04 (suggesting that legislators cannot simply circumvent supermajority rules applied to tax increases with debt because of limits on rent seeking).
162. One might argue that one advantage of a process without OIRA is that regulations spanning the jurisdiction of several agencies are more difficult for special interests to shape because it is harder for a particular faction to capture multiple agencies than one agency. But even with OIRA, such regulations still will have to receive the approval of multiple agencies. OIRA does not replace the agency process but supplements it.
163. Political parties serve in fact as a counterbalance to narrow special interests. Michael A. Fitts, The Vices of Virtues: A Political Party Perspective on Civic Virtue Reforms of the Legislative Process, 136 U. PA. L. REV. 1567, 1604 (1988); see also Michael H. Schill, Uniformity or Diversity: Residential Real Estate Finance Law in the 1990s and the Implications of Changing Financial Markets, 64 S. CAL. L. REV. 1261, 1320 (1991) (explaining that diversity in state laws ensures that special interests do not prevail on the national level).
164. MANCUR OLSON, POWER AND PROSPERITY: OUTGROWING COMMUNIST AND CAPITALIST DICTATORSHIPS 14 (2000).
165. Id. at 14-17. For instance, a coalition made up of all productive industries is less likely to take rents than a coalition of the auto industry, because the auto industry can hope that the rents will be supported by taxes that other industries and their employees generate whereas the rents for all industries must be supported by taxes from those same industries and their stakeholders.
166. John Ferejohn & Charles Shipan, Congressional Influence on Bureaucracy, 6 J.L. ECON. & ORG. 1, 19 (special issue 1990) (noting that agencies must balance their relationships with both Congress and the president). For a discussion of why committees may be unrepresentative of Congress, see supra note 73 and accompanying text.
167. For a discussion of the tragedy of the commons, see supra note 120 and accompanying text. See also Calabresi, supra note 120, at 84 (discussing the risk of "geographic" capture of agencies).
168. As always, the important judgment in structural matters is the overall assessment compared to the alternatives. Some rent-seeking factions will influence OIRA; the question is whether they are likely to be, on balance, worse than those influencing the agencies.
169. Mark Seidenfeld, A Civic Republican Justification for the Bureaucratic State, 105 HARV. L. REV. 1512, 1552 (1992). For a discussion of such criticisms, see THOMAS O. MCGARITY, REINVENTING RATIONALITY: THE ROLE OF REGULATORY ANALYSIS IN THE FEDERAL BUREAUCRACY 281-88 (1991) (cataloging complaints made against specific executive-controlled agencies).
170. There were many fewer laws at the time of the Framing, and the laws that had been passed prior to the Constitution had faced supermajority rules in the Articles of Confederation. McGinnis & Rappaport, supra note 9 (manuscript at 12).
171. Sidney A. Shapiro, Administrative Law After the Counter-Reformation: Restoring Faith in Pragmatic Government, 48 U. KAN. L. REV. 689, 707-09 (2000) (discussing OIRA's monitoring role). I discuss the argument that the president can simply supplant the discretion of his agency heads and issue regulations himself. See supra note 66 and accompanying text.
172. See Kagan, supra note 65, at 2338 (suggesting that OIRA follows the "policy orientation of the President").
173. For a discussion of the fire-alarm systems in Congress, see Mathew McCubbins et al., Administrative Procedures as Instruments of Political Control, 3 J.L. ECON. & ORG. 243, 250 (1987).
174. Mark Seidenfeld, A Big Picture Approach to Presidential Influence on Agency Policy-Making, 80 IOWA L. REV. 1, 43-44 (1994).
175. See supra notes 43-57 and accompanying text.
176. Cost-benefit analysis attempts to mimic the market in measuring the value of regulation "in terms of willingness to pay." Cass R. Sunstein, Is Cost Benefit Analysis for Everyone?, 53 ADMIN. L. REV. 299, 301 (2001). Cost-benefit analysis takes entitlements as they are and does not seek to change them. But this is a sound limitation as a matter of political economy. Redistribution, if to be undertaken at all, is best done through the tax system, not through regulations. See infra notes 240-42 and accompanying text. Regulation as a general matter should be a matter of allocation. Of course, the president is bound to use regulation for redistributive purposes if Congress so directs. The issue in the regulatory review order is how the president should exercise his discretion in the absence of congressional direction to the contrary.
177. While the Constitution does not mandate cost-benefit analysis, such a mode of thinking was not unknown to the Framers. Benjamin Franklin recommended that individuals consider courses of action by writing down all their advantages and disadvantages:
To get over this, my Way is, to divide half a Sheet of Paper by a Line into two Columns; writing over the one Pro, and over the other Con. Then during three or four Days Consideration, I put down under the different heads short Hints of the different Motives, that at different Times occur to me, for or against the Measure. When I have thus got them all together in one View, I endeavour to estimate their respective Weights; and where I find two, one on each side, that seem equal, I strike them both out. If I find a Reason pro equal to some two Reasons con, I strike out the three. If I judge some two Reasons con, equal to some three Reasons pro, I strike out the five; and thus proceeding I find at length where the Ballance lies; and if after a Day or two of farther Consideration, nothing new that is of Importance occurs on either side, I come to a Determination accordingly. And, tho' the Weight of Reasons cannot be taken with the Precision of Algebraic Quantities, yet, when each is thus considered, separately and comparatively, and the whole lies before me, I think I can judge better, and am less liable to make a rash Step; and in fact I have found great Advantage from this kind of Equation, in what may be called Moral or Prudential Algebra.
BENJAMIN FRANKLIN: REPRESENTATIVE SELECTIONS, WITH INTRODUCTION, BIBLIOGRAPHY AND NOTES 348-49 (Frank Luther Mott & Chester E. Jorgenson eds., 1936); see also Int'l Union, UAW v. OSHA, 938 F.2d 1310, 1319-21 (D.C. Cir. 1991) (suggesting that there exists a continuity between Benjamin Franklin's thought and cost-benefit analysis). I am grateful to Jonathan Weiner for bringing this quotation to my attention.
178. For a discussion of the reasons special interests have grown stronger, see McGinnis & Rappaport, supra note 26, at 394-96. In particular, we note that special interests have multiplied as labor has become more specialized. Id. at 394.
179. Matthew D. Adler & Eric A. Posner, Rethinking Cost-Benefit Analysis, 109 YALE L.J. 165, 225-26 (1999) (arguing that cost-benefit analysis offers a more workable model than alternatives).
180. See Phillip Nelson, Political Information, 19 J.L. & ECON. 315, 323 (1976) (suggesting that rent seekers who lack a political majority have an incentive to take their gain "in a form where the issue can be easily obscured"). Indeed, information costs drive modern interest group politics, in which political success not only depends on satisfying members of a politician's coalition but also on disarming potential opposition. Other ways of facilitating the acquisition of rents include using obscure regulations or legislation. See Zelinsky, supra note 121, at 1374-75 (arguing that "[u]nfunded mandates . . . present legislators with the political temptation to levy hidden municipal taxes" because constituents do not appreciate the link between their municipal tax bills and their legislators' adoption of unfunded mandates); see also John O. McGinnis, The Bar Against Challenges to Employment Consent Decrees: A Public Choice Perspective, 54 LA. L. REV. 1507, 1530-31 (1994) (discussing the collateral attack bar in employment discrimination as an example of legislation that raises the information costs of potential opposition).
181. Cost-benefit analysis has been defended persuasively as a pragmatic tool. Richard Posner, Cost-Benefit Analysis: Definition, Justification, and Comment on Conference Papers, 29 J. LEG. STUD. 1153, 1157 (2000) (endorsing a pragmatic defense of cost-benefit analysis); Cass Sunstein, Cognition and Cost-Benefit Analysis, 29 J. LEG. STUD. 1059, 1077 (2000) (Cost-benefit analysis is "best taken as [a] pragmatic instrument . . . designed to assist people in making complex judgments where multiple goods are involved."). What I try to add is that cost-benefit analysis-particularly through the transparency and discipline it provides-restrains rent seeking and thus restores one of the original functions of the Constitution.
182. RICHARD A. EPSTEIN, TAKINGS: PRIVATE PROPERTY AND THE POWER OF EMINENT DOMAIN 29-31 (1985) (defending judicial activism as a means to ensure that individual rights are not destroyed).
183. For this reason, the practice of delegation grew hand in hand with the acceptance of the "expertise model" of administration. Philip J. Harter, The Political Legitimacy and Judicial Review of Consensual Rules, 32 AM. U. L. REV. 471, 472 (1983) (linking the rise in delegation to the executive branch with the rise of the expertise model).
184. See Roger Clegg, Epilogue: Civil Rights in the Eighties and Nineties, 54 LA. L. REV. 1605, 1612 (1994) (discussing the difficulties that centralized bureaucracies face in economic decisionmaking).
185. Rent seekers in fact provide a service by bringing information to the attention of legislators. See Dwight R. Lee, In Defense of Excessive Government, 65 S. ECON. J. 674, 680-84, 688-90 (1999) (arguing that certain government programs that seem wasteful may truly serve the public interest). Nevertheless, interest groups and constituents will bring concerns to OIRA directly.
186. I previously noted that such passions and interests also could adversely affect the integrity of federalism decisions under the federalism executive order. See supra notes 140-42 and accompanying text.
187. Eric A. Posner, Cost-Benefit Analysis as a Solution to a Principal-Agent Problem, 53 ADMIN. L. REV. 289, 294 (2001) ("Indeed the hard-headed public choice theorist would conjecture that the President is more likely to enforce cost-benefit requirements on agencies that do not generally transfer resources to his constituents, and adopt a lax attitude toward agencies that do.").
188. See supra note 181 (collecting authorities describing the analytical discipline that is imposed by cost-benefit analysis).
189. As discussed above, when the interstate externalities caused by private activity are insubstantial, it is likely that the federal regulatory process will give rise to more substantial costs. For a general discussion of this danger, see BUCHANAN & TULLOCK, supra note 34, at 283-95.
190. When jurisdictional competition is working well, optimal national harmonization may occur as a byproduct. See supra note 140.
191. Exec. Order No. 13,132 ?4(a)-(b), 3 C.F.R. 206, 208 (1999), reprinted in 5 U.S.C. ?601 (2000):
(a) Agencies shall construe, in regulations and otherwise, a Federal statute to preempt State law only where the statute contains an express preemption provision or there is some other clear evidence that the Congress intended preemption of State law, or where the exercise of State authority conflicts with the exercise of Federal authority under the Federal statute.
(b) Where a Federal statute does not preempt State law (as addressed in subsection (a) of this section), agencies shall construe any authorization in the statute for the issuance of regulations as authorizing preemption of State law by rulemaking only when the exercise of State authority directly conflicts with the exercise of Federal authority under the Federal statute or there is clear evidence to conclude that the Congress intended the agency to have the authority to preempt State law.
192. Blake, supra note 97, at 311 (describing the slightly more stringent standard of the Reagan order). The Reagan order, Exec. Order No. 12,612, 3 C.F.R. 252 (1987), revoked by Exec. Order No. 13,132, 3 C.F.R. 206 (1999), reprinted in 5 U.S.C. ?601 (2000), provided:
(a) To the extent permitted by law, Executive departments and agencies shall construe, in regulations and otherwise, a Federal statute to preempt State law only when the statute contains an express preemption provision or there is some other firm and palpable evidence compelling the conclusion that the Congress intended preemption of State law, or when the exercise of State authority conflicts with the exercise of Federal authority under the Federal statute.
(b) Where a Federal statute does not preempt State law (as addressed in subsection (a) of this section), Executive departments and agencies shall construe any authorization in the statute for the issuance of regulations as authorizing preemption of State law by rule-making only when the statute expressly authorizes issuance of preemptive regulations or there is some other firm and palpable evidence compelling the conclusion that the Congress intended to delegate to the department or agency the authority to issue regulations preempting State law.
Exec. Order No. 12,612 ?4(a)-(b), 3 C.F.R. at 255.
193. For instance, it may be that state statutes that restrict advertising retard the process of jurisdictional competition by preventing information from circulating and thus can be preempted. See, e.g., Paul R. Verkuil, Preemption of State Law by the Federal Trade Commission, 1976 DUKE L.J. 225, 247 (advocating FTC preemption of statutes that precluded pharmaceutical advertising).
194. For the reasoning behind this analysis of preemption, see, for example, THE STATUS OF FEDERALISM IN AMERICA: A REPORT OF THE WORKING GROUP ON FEDERALISM OF THE DOMESTIC POLICY COUNCIL 37-42 (1986).
195. 469 U.S. 528 (1985).
196. Id. at 554. In Garcia, the Court refused to invalidate on Tenth Amendment grounds congressional legislation imposing minimum wage requirements on state transit operations. Id. The Court's core holding was that Congress's power to impose its will on the states should be limited, not by ad hoc judicial determinations balancing state and federal interests under the Tenth Amendment, but by the structural arrangements of the federal system, principally the state's participation in the national political process through the voice of its elected representatives. Id.
197. The importance of these protections, at least in areas of traditional state autonomy, is one way to understand Gregory v. Ashcroft, 501 U.S. 452 (1991), in which the Court refused to apply the Age Discrimination in Employment Act of 1967 to state court judges because its application was not clearly indicated. Id. at 470; see also Bus. Roundtable v. SEC, 905 F.2d 406, 413-14 (D.C. Cir. 1990) (citing Garcia to justify the court's refusal to enforce SEC regulations that intruded into traditional state areas without a clear indication of congressional intent).
198. Garcia, 469 U.S. at 551-54.
199. Clark, supra note 150, at 1429. Federal judges have life tenure, U.S. CONST. art. III, ?1, and thus by constitutional design are not politically accountable in their states. Thus, under the Constitution's structure, there is no substantial political constraint to prevent the federal judiciary from encroaching on state interests. The only constraint is the method of judicial decisionmaking-the constraint of formal rules neutrally applied. Insofar as federal judges instead employ discretionary methods of decision that resemble those of legislators, the Constitution's structure suggests that state interests would be better protected by the political branches.
200. The Reagan administration also found support for this aspect of Garcia's teaching in Gregory v. Ashcroft, in which the Court refused to permit federal age discrimination law to preempt Missouri's mandatory retirement age for its state judges. 501 U.S. at 473. Indeed, in that case, the Court stated that it would require a clear statement from Congress before it would interpret a provision of the law to apply to state judges. Id. at 464. While the Court has not applied the clear statement rule to all preemption cases, the more clearly states are put on notice that their rights of governance are to be displaced by federal law, the more effectively the legislative process will protect state autonomy.
201. It is now widely recognized that Congress delegates power to other entities at least in part to avoid accountability for hard decisions. The question of whether to preempt state law is frequently a peculiarly difficult decision: Congress may be besieged not only by the usual private interest groups who would like centralized regulation but also by elected state officials and state government institutions whose power may be diminished by preemption. The incentives to punt on the preemption issue thus may often be intense. See Clark, supra note 150, at 1427 ("To the extent that proposed legislation (later claimed to regulate the states) is ambiguous, states and their allies in the legislative process lack notice and thus incentive to prevent adoption of [preempting] legislation. . . . A clear statement requirement prevents such evasion.").
202. See S. Candice Hoke, Preemption Pathologies and Civic Republican Values, 71 B.U. L. REV. 685, 694 (1991) (suggesting that preemption can permit the judiciary to make policy decisions and usurp the legislature's role).
203. A doctrinal structure that leads to such a policy-laden role for courts was not thought to comport with the judicial role contemplated under Article III. The decision to locate responsibility for assuring the supremacy of federal law in the federal judiciary rather than in the legislature or in the executive was made by a founding generation well aware that the salient characteristic of the judiciary was that it has, in Hamilton's famous words, no "will, but merely judgment." THE FEDERALIST NO. 78, at 465 (Alexander Hamilton) (Clinton Rossiter ed., 1961). When the judiciary weighs the policy merits (or even measures its view of the extent of the federal interest versus that of the state) in determining the scope of preemption, it is exercising will rather than judgment and consequently its action cannot be distinguished from that of the legislature.
204. Caleb Nelson, Preemption, 86 VA. L. REV. 225, 292-96 (2000).
205. Id. 241-43.
206. Id. at 293.
207. Id. at 302.
208. Viet D. Dinh, Reassessing the Law of Preemption, 88 GEO. L.J. 2085, 2097-2100 (2000).
209. Id. at 2099.
210. Id. at 2098.
211. Professor Bradford Clark criticizes Professor Dinh's analysis, suggesting that "constitutional structure appears to favor a presumption against preemption because the Constitution gives states a role in selecting Congress and the President, but not federal courts." Clark, supra note 150, at 1429.
212. It is clear that the Court believes that at least issues of statutory meaning that result in preemption are subject to the principle announced in Chevron U.S.A. Inc. v. Natural Resource Defense Council, Inc., 467 U.S. 837 (1984). See Smiley v. Citibank (South Dakota), N.A., 517 U.S. 735, 743-44 (1996) (stating that questions of a statute's substantive meaning were subject to Chevron, where the question was whether interpretation of "interest charges" that preempted state law would include late payments). The Court has not decided whether interpretations of jurisdictional provisions (like the scope of express preemption provisions) are subject to Chevron deference. See id. at 744 (assuming without deciding that interpretations of preemption provisions are not subject to Chevron deference). If the Court ultimately determines that such provisions are not subject to deference, this preemption aspect of the federalism order will be less important.
213. The one exception to this analysis may be that even in the area of spillovers, the "obstacle" rationale for preemption should not be permitted. Professor Nelson shows that obstacle preemption is inconsistent with the original understanding. "The mere fact that Congress enacts a statute to serve certain purposes does not automatically imply that Congress wants to displace all state law that will get in the way of those purposes." Nelson, supra note 204, at 287.
214. For a discussion of the virtues of pragmatic federalism in other areas of the federalism executive order, see supra notes 136-40 and accompanying text.
215. See Curtis A. Bradley, Chevron Deference and Foreign Affairs, 86 VA. L. REV. 649, 675-79 (2000) (suggesting that agencies should be able to disregard canons based on policy considerations but not based on jurisprudential considerations). But see Thomas W. Merrill, Judicial Deference to Executive Precedent, 101 YALE L.J. 969, 988 (1992) ("[I]f an agency interpretation is consistent with the language and purpose of a statute, it is hard to see how it could be condemned as unreasonable simply because a judicial canon would suggest a contrary result.").
216. See McGinnis, supra note 142 (manuscript at 29-32) (criticizing recent cases for not adopting a clear statement rule).
217. Basing an interpretation on jurisprudential considerations exacerbates the tension between Chevron and the command of the Administrative Procedure Act for the reviewing Court to "decide all relevant questions of law." See John F. Duffy, Administrative Common Law in Judicial Review, 77 TEX. L. REV. 113, 193-95 (1998) (describing this tension).
218. See John F. Manning, Constitutional Structure and Judicial Deference to Agency Interpretations of Agency Rules, 96 COLUM. L. REV. 612, 625 (1996) (arguing that Chevron is based on the presumption that "interpretive choice entails the exercise of substantial policymaking discretion").
219. Mich. Citizens for an Indep. Press v. Thornburgh, 868 F.2d 1285, 1292-97 (D.C. Cir. 1989) (refusing to limit Chevron deference based on a canon construing exceptions to antitrust statutes narrowly because the Attorney General's interpretation was based on a policy balance), aff'd by an equally divided court, 493 U.S. 38 (1989).
220. See supra note 202 and accompanying text.
221. Chevron U.S.A. Inc. v. Natural Res. Def. Council, Inc., 467 U.S. 837, 865-66 (1984).
222. See Kagan, supra note 65, at 2376 (offering a justification for Chevron deference to presidentially directed statutory construction). The Court, however, may have cut back significantly on the scope of Chevron in general. See United States v. Mead Corp., 533 U.S. 218, 121 S. Ct. 2164, 2171 (2001) (holding that Chevron deference is available when "it appears that Congress delegated authority to the agency generally to make rules carrying the force of law," as by giving the agency notice and comment rulemaking authority, and when "the agency interpretation claiming deference was promulgated in the exercise of that authority").
223. Exec. Order No. 12,612 ?8, 3 C.F.R. 252, 256 (1987), revoked by Exec. Order No. 13,132, 3 C.F.R. 206 (1999), reprinted in 5 U.S.C. ?601 (2000); Exec. Order No. 13,132 ?11, 3 C.F.R. 206, 211 (1999), reprinted in 5 U.S.C. ?601 (2000).
224. The Clinton order restricts OIRA review to "significant regulatory action" and it defines such actions to include only rules that may:
(1) Have an annual effect on the economy of $100 million or more or adversely affect in a material way the economy, a sector of the economy, productivity, competition, jobs, the environment, public health or safety, or State, local, or tribal governments or communities;
(2) Create a serious inconsistency or otherwise interfere with an action taken or planned by another agency;
(3) Materially alter the budgetary impact of entitlements, grants, user fees, or loan programs or the rights and obligations of recipients thereof; or
(4) Raise novel legal or policy issues arising out of legal mandates, the President's priorities, or the principles set forth in this Executive order.
Exec. Order No. 12,866 ?3(f), 3 C.F.R. 638, 641-42 (1994), reprinted in 5 U.S.C. ?601 (2000).
225. See FR蒁蒖IC BASTIAT, SELECTED ESSAYS ON POLITICAL ECONOMY 1 (George de Huszar ed., Seymour Cain trans., Foundation for Economic Education 4th ed. 1995) (1964) ("[A] law produces not only one effect, but a series of effects . . . . [T]he first alone is immediate; it appears simultaneously with its cause; it is seen. The other effects emerge only subsequently; they are not seen; we are fortunate if we foresee them."). See generally Christopher M. Grengs, Note, Making the Unseen Seen: Issues and Options in Small Business Regulatory Reform, 85 MINN. L. REV. 1957 (2001) (discussing the problem of unforeseen effects in the regulation of small business).
226. Adler & Posner, supra note 179, at 240.
227. Another advantage of applying a cost-benefit standard to cost-benefit analysis is that such a limitation could prevent OIRA's bureaucratic aggrandizement.
228. THE NEW INTERNATIONAL DICTIONARY OF QUOTATIONS 184 (Hugh Rawson & Margaret Miner eds., 1986) (quoting Senator Everett Dirksen).
229. See supra notes 67-69 and accompanying text.
230. One argument against expanding the scope of regulatory review to include all regulations is that agencies simply will avoid notice and comment rulemaking and substitute interpretive rules. In my view, the courts should address this kind of substitution by maintaining a strict dichotomy between the deference due regulations and that due interpretive rules. If interpretive rules do not receive substantial deference, agencies still will have incentives to go through the notice and comment process. See Robert A. Anthony, The Supreme Court and the APA: Sometimes They Just Don't Get It, 10 ADMIN. L.J. AM. U. 1, 6-9 (1996) (arguing for a strong distinction in terms of the deference given to interpretive rules and to rules that are subject to notice and comment).
231. See Kagan, supra note 65, at 2277-88 (noting that neither Executive Order 12,291 nor Executive Order 12,866 applied fully to independent agencies). The regulatory planning requirements of Executive Order 12,866 apply to independent agencies. Exec. Order No. 12,866 ?4(c), 3 C.F.R. 638, 642 (1994), reprinted in 5 U.S.C. ?601 (2000).
232. The original case that independent agencies weakened presidential control can be found in MARVER BERNSTEIN, REGULATING BUSINESS BY INDEPENDENT COMMISSION 137-39 (1955). For a more skeptical view, see generally Susan Bartlett Foote, Independent Agencies Under Attack: A Skeptical View of the Importance of the Debate, 1988 DUKE L.J. 223 (suggesting that other complex political forces are more important than independent or dependent status in determining the extent of presidential control). For a modern discussion of agency independence, see Calabresi, supra note 120, at 83-84 (suggesting that geographically based factions and faction members dominate independent agencies).
233. The Clinton regulatory review order already has moved in this direction by applying the regulatory planning process of OMB to independent agencies in a way that allowed the vice president and other participants in the interagency process to request further consideration of rules that conflicted with the president's program. See Exec. Order No. 12,866 ?2(b)-(c), 3 C.F.R. at 640 (stipulating the review roles of the OMB and the vice president in the regulatory planning process). For further discussion of the manner in which President Clinton drew the independent agencies closer to the executive branch, see Kagan, supra note 65, at 2288.
234. See, e.g., PETER M. SHANE & HAROLD M. BRUFF, LAW OF PRESIDENTIAL POWER 355 (1988) (arguing that the presidential authority to issue executive orders is constitutionally derived from Article II, ?3).
235. See Exec. Order No. 12,866 ?6(b)(2), 3 C.F.R. at 647 (setting a strict time limit of ninety days for OIRA review and prescribing one possible extension of thirty days).
236. See id. ?7, 3 C.F.R. at 648:
To the extent permitted by law, disagreements or conflicts between or among agency heads or between OMB and any agency that cannot be resolved by the Administrator of OIRA shall be resolved by the President, or by the Vice President acting at the request of the President, with the relevant agency head (and, as appropriate, other interested government officials). Vice Presidential and Presidential consideration of such disagreements may be initiated only by the Director, by the head of the issuing agency, or by the head of an agency that has a significant interest in the regulatory action at issue. Such review will not be undertaken at the request of other persons, entities, or their agents.
237. Congress did this with the Environmental Protection Agency in the 1980s when it felt that the agency was not issuing regulations fast enough. Sidney A. Shapiro & Robert L. Glicksman, Congress, the Supreme Court, and the Quiet Revolution in Administrative Law, 1988 DUKE L.J. 819, 827. Of course, there are disadvantages to statutorily set regulatory deadlines. Cass Sunstein, Constitutionalism After the New Deal, 101 HARV. L. REV. 421, 481, 509 (1987). However, it is up to Congress to make those tradeoffs.
238. Exec. Order No. 12,866 ?1(a), 3 C.F.R. at 639:
In deciding whether and how to regulate, agencies should assess all costs and benefits of available regulatory alternatives, including the alternative of not regulating. Costs and benefits shall be understood to include both quantifiable measures (to the fullest extent that these can be usefully estimated) and qualitative measures of costs and benefits that are difficult to quantify, but nevertheless essential to consider. Further, in choosing among alternative regulatory approaches, agencies should select those approaches that maximize net benefits (including potential economic, environmental, public health and safety, and other advantages; distributive impacts; and equity), unless a statute requires another regulatory approach.
239. See supra notes 180-81 and accompanying text.
240. Exec. Order No. 12,866 ?1, 3 C.F.R. at 638.
241. Louis Kaplow & Steven Shavell, Why the Legal System Is Less Efficient Than the Income Tax in Redistributing Income, 23 J. LEGAL STUD. 667, 669 (1994).
242. Id. at 677.
243. One must always compare the additional political externalities created by a regulatory regime with the reduction in externalities it effectuates. John O. McGinnis, The Political Economy of Global Multilateralism, 1 CHI. J. INT'L L. 381, 395 (2000).
244. See, e.g., Roger Meiners & Bruce Yandle, Common Law and the Conceit of Modern Environmental Policy, 7 GEO. MASON L. REV. 923, 955 (1999) (discussing how political favors via environmental laws may be used to raise the costs of competitors).
245. Exec. Order No. 12,866 ?6(b)(4), 3 C.F.R. at 644-48.
246. See supra note 180 and accompanying text.
247. See Lee, supra note 185, at 679 (noting that interest groups have incentives to produce information relevant to the political process that individuals do not have).
248. Cf. Richard Nagareda, Note, Ex Parte Contacts and Institutional Roles: Lessons from the OMB Experience, 55 U. CHI. L. REV. 591, 622 (1988) (arguing that interest groups will have no reason to feel embarrassed to provide information about the compliance costs of regulations).
249. Exec. Order No. 12,866 ?6(b)(5), 3 C.F.R. at 648.
250. See J. Skelly Wright, Money and the Pollution of Politics: Is the First Amendment an Obstacle to Political Equality?, 82 COLUM. L. REV. 609, 623 (1982) (discussing the coherent and coordinated campaigns that special interest groups can mount in the media).
251. For general criticism of the miscommunication that exists within the executive oversight process, see Robert V. Percival, Checks Without Balance: Executive Office Oversight of the Environmental Protection Agency, 54 LAW & CONTEMP. PROBS. 127, 185-89 (Autumn 1991).
252. See Steven T. Kargman, OMB Intervention in Agency Rulemaking: The Case for Broadened Record Review, 95 YALE L.J. 1789, 1797 (1986) (acknowledging that an agency may attempt to justify its actions by public record, which may disguise the true intentions behind the decision).
253. See Sierra Club v. Costle, 657 F.2d 298, 407 (D.C. Cir. 1981) (permitting undocketed executive branch communications with agencies but requiring that regulations have support in the record).
254. Kargman, supra note 252, at 1797-98.
255. See generally McGinnis, supra note 142 (discussing the manner in which the Supreme Court is trying to revive the decentralized structure of social norms); McGinnis & Movsesian, supra note 16 (discussing how the World Trade Organization facilitates economic growth and jurisdictional competition).
256. Kaplow & Shavell, supra note 241, at 667.
257. See, e.g., Kagan, supra note 65, at 2282-84 (demonstrating how President Clinton advanced regulatory projects, namely, limiting tobacco advertisements and controlling uses of unemployment insurance, in accordance with the liberal agenda).
258. See G. Edward White, Revisiting James Bradley Thayer, 88 NW. U. L. REV. 48, 57 (1993) (suggesting that Democrats opposed the strong national government favored by Whigs in part because they thought the costs imposed by special interests at the national level would be very high).
259. See COMM'N ON LAW AND THE ECON., FEDERAL REGULATION: ROADS TO REFORM 72-73 (Am. Bar Ass'n ed., 1979) (justifying coordination within the executive branch).
260. See Sunstein, supra note 237, at 460 (discussing the advantages of the executive order in promoting political accountability).