VOLUME 119 JUNE 2006 NUMBER 8 HARVARD LAW REVIEW 2006 by The Harvard Law Review Association ARTICLE SEPARATION OF PARTIES,NOT POWERS Daryl I.Levinson&Richard H.Pildes TABLE OF CONTENTS INTRODUCTION....... .2312 I.FROM BRANCHES TO PARTIES........... 2316 A.Madison and the Mechanisms of Political Competition316 B.Presidential,Parliamentary,and Party Government....... .2325 C.Conclsion Separation of parties2329 II.PARTY UNIFICATION AND DIVISION OF GOVERNMENT330 A.The Past,Present,and Future of Unified and Divided Government.. .2330 I.Unified and Divided Governments... .2330 2.Fragmented and Cohesive Parties.... .2332 B.The Functional Differences Parties Make................ .2338 I.Legislative Efficacy. .2338 2.Executive Accountability...... 2342 III.REENVISIONING,AND REFORMING,THE SEPARATION OF POWERS....................2347 A.Separation-of-powers Law2349 I.Rights and Executive Powers During War and Crisis............... .2349 2.The Administrative State............. .2356 3 Judicial Review2364 B.Democratic Institutional Design....... .2368 I.Minority Opposition Rights. 2368 2.Bureaucracy and the Checking Function... 2375 C.Political Parties and the Law of Democracy. 2379 L.Safe Districting… 2380 2.Primary Election Structures. 2381 3.Internal Legislative Rules... .2383 4.Encouraging Divided Government 2384 CONCLUSION........ 2385 231I
2311 VOLUME 119 JUNE 2006 NUMBER 8 © 2006 by The Harvard Law Review Association ARTICLE SEPARATION OF PARTIES, NOT POWERS Daryl J. Levinson & Richard H. Pildes TABLE OF CONTENTS INTRODUCTION ..........................................................................................................................2312 I. FROM BRANCHES TO PARTIES...........................................................................................2316 A. Madison and the Mechanisms of Political Competition..............................................2316 B. Presidential, Parliamentary, and Party Government ...................................................2325 C. Conclusion: Separation of Parties...................................................................................2329 II. PARTY UNIFICATION AND DIVISION OF GOVERNMENT ...........................................2330 A. The Past, Present, and Future of Unified and Divided Government ........................2330 1. Unified and Divided Governments............................................................................2330 2. Fragmented and Cohesive Parties .............................................................................2332 B. The Functional Differences Parties Make .....................................................................2338 1. Legislative Efficacy......................................................................................................2338 2. Executive Accountability ............................................................................................2342 III. REENVISIONING, AND REFORMING, THE SEPARATION OF POWERS ...................2347 A. Separation-of-Powers Law...............................................................................................2349 1. Rights and Executive Powers During War and Crisis............................................2349 2. The Administrative State ............................................................................................2356 3. Judicial Review.............................................................................................................2364 B. Democratic Institutional Design.....................................................................................2368 1. Minority Opposition Rights ........................................................................................2368 2. Bureaucracy and the Checking Function .................................................................2375 C. Political Parties and the Law of Democracy.................................................................2379 1. Safe Districting .............................................................................................................2380 2. Primary Election Structures........................................................................................2381 3. Internal Legislative Rules ...........................................................................................2383 4. Encouraging Divided Government............................................................................2384 CONCLUSION ...............................................................................................................................2385
SEPARATION OF PARTIES,NOT POWERS Daryl J.Levinson'&Richard H.Pildes** American political institutions were founded upon the Madisonian assumption of vigorous,self-sustaining political competition between the legislative and executive branches.Congress and the President would check and balance each other;officeholders would defend the distinct interests of their different institutions;ambition would counteract ambition.That is not how American democracy turned out.Instead political competition and cooperation along relatively stable lines of policy and ideological disagreement quickly came to be channeled not through the branches of government,but rather through an institution the Framers could imagine only dimly but nonetheless despised:political parties.Few aspects of the founding generation's political theory are now more clearly anachronistic than their vision of legislative-executive separation of powers.Yet few of the Framers'ideas continue to be taken as literally or sanctified as deeply by courts and constitutional scholars as the passages about interbranch relations in Madison's Federalist 51.This Article reenvisions the law and theory of separation of powers by viewing it through the lens of party competition.In particular,it points out that during periods-like the present-of cohesive and polarized political parties,the degree and kind of competition between the legislative and executive branches will vary significantly and may all but disappear,depending on whether party control of the House,Senate,and Presidency is divided or unified.The practical distinction between party-divided and party-unified government thus rivals, and often dominates,the constitutional distinction between the branches in predicting and explaining interbranch political dynamics. INTRODUCTION escribing a set of“wholly new discoveries'”in the“science of poli- tics"that might enable democratic self-government to succeed in the American republic,Alexander Hamilton listed first the "balances and checks"that distinctively characterize the American system of separation of powers.1 In Madison's ingenious scheme of separated powers,“the interior structure of the government'”would be“so con- trivled]""as that its several constituent parts may,by their mutual re- lations,be the means of keeping each other in their proper places."2 By institutionalizing a differentiation between executive and legislative Professor of Law,Harvard Law School. Sudler Family Professor of Constitutional Law,New York University School of Law.Car- negie Scholar 2004. Thanks to Rachel Barkow,Rosalind Dixon,John Ferejohn,Barry Friedman,Jack Gold- smith,David Golove,Don Herzog,Rick Hills,Dan Hulsebosch,Mike Klarman,Lewis Korn- hauser,Liz Magill,Pasquale Pasquino,Jim Ryan.Seana Shiffrin,Matthew Stephenson,Bill Stuntz,Cass Sunstein,Adam Unikowsky,Adrian Vermeule,and participants in workshops and colloquia at Berkeley,Harvard,Minnesota,NYU,and Texas for useful comments on drafts. 1THE FEDERALIST No.9,at 72(Alexander Hamilton)(Clinton Rossiter ed.,1961). 2 THE FEDERALIST No.5I (James Madison),supra note I,at 320. 23I2
2312 SEPARATION OF PARTIES, NOT POWERS Daryl J. Levinson* & Richard H. Pildes** American political institutions were founded upon the Madisonian assumption of vigorous, self-sustaining political competition between the legislative and executive branches. Congress and the President would check and balance each other; officeholders would defend the distinct interests of their different institutions; ambition would counteract ambition. That is not how American democracy turned out. Instead, political competition and cooperation along relatively stable lines of policy and ideological disagreement quickly came to be channeled not through the branches of government, but rather through an institution the Framers could imagine only dimly but nonetheless despised: political parties. Few aspects of the founding generation’s political theory are now more clearly anachronistic than their vision of legislative-executive separation of powers. Yet few of the Framers’ ideas continue to be taken as literally or sanctified as deeply by courts and constitutional scholars as the passages about interbranch relations in Madison’s Federalist 51. This Article reenvisions the law and theory of separation of powers by viewing it through the lens of party competition. In particular, it points out that during periods — like the present — of cohesive and polarized political parties, the degree and kind of competition between the legislative and executive branches will vary significantly and may all but disappear, depending on whether party control of the House, Senate, and Presidency is divided or unified. The practical distinction between party-divided and party-unified government thus rivals, and often dominates, the constitutional distinction between the branches in predicting and explaining interbranch political dynamics. INTRODUCTION escribing a set of “wholly new discoveries” in the “science of politics” that might enable democratic self-government to succeed in the American republic, Alexander Hamilton listed first the “balances and checks” that distinctively characterize the American system of separation of powers.1 In Madison’s ingenious scheme of separated powers, “the interior structure of the government” would be “so contriv[ed]” “as that its several constituent parts may, by their mutual relations, be the means of keeping each other in their proper places.”2 By institutionalizing a differentiation between executive and legislative ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– * Professor of Law, Harvard Law School. ** Sudler Family Professor of Constitutional Law, New York University School of Law. Carnegie Scholar 2004. Thanks to Rachel Barkow, Rosalind Dixon, John Ferejohn, Barry Friedman, Jack Goldsmith, David Golove, Don Herzog, Rick Hills, Dan Hulsebosch, Mike Klarman, Lewis Kornhauser, Liz Magill, Pasquale Pasquino, Jim Ryan, Seana Shiffrin, Matthew Stephenson, Bill Stuntz, Cass Sunstein, Adam Unikowsky, Adrian Vermeule, and participants in workshops and colloquia at Berkeley, Harvard, Minnesota, NYU, and Texas for useful comments on drafts. 1 THE FEDERALIST NO. 9, at 72 (Alexander Hamilton) (Clinton Rossiter ed., 1961). 2 THE FEDERALIST NO. 51 (James Madison), supra note 1, at 320. D
2o06] SEPARATION OF PARTIES,NOT POWERS 23I3 powers (as well as by dividing the legislature into two chambers),the separation of powers would harness political competition into a system of government that would effectively organize,check,balance,and dif- fuse power.What is more,the system would be self-enforcing,relying on interbranch competition to police institutional boundaries and pre- vent tyrannical collusion.In the Framers'Newtonian vision,the sepa- ration of powers was to be "a machine that would go of itself."3 To this day,the idea of self-sustaining political competition built into the structure of government is frequently portrayed as the unique genius of the U.S.Constitution,the very basis for the success of American democracy.4 Yet the truth is closer to the opposite.The success of American democracy overwhelmed the Madisonian concep- tion of separation of powers almost from the outset,preempting the political dynamics that were supposed to provide each branch with a “will of its own”that would propel departmental“[a]mbition...to counteract ambition."5 The Framers had not anticipated the nature of the democratic competition that would emerge in government and in the electorate.Political competition and cooperation along relatively stable lines of policy and ideological disagreement quickly came to be channeled not through the branches of government,but rather through an institution the Framers could imagine only dimly but nevertheless despised:political parties.As competition between the legislative and executive branches was displaced by competition between two major parties,the machine that was supposed to go of itself stopped running. Few aspects of the founding generation's political theory are now more clearly anachronistic than their vision of legislative-executive separation of powers.Nevertheless,few of the Framers'ideas con- tinue to be taken as literally or sanctified as deeply by courts and con- stitutional scholars as the passages about interbranch relations in Madison's Federalist 51.To this day,Madison's account of rivalrous, 3 MICHAEL KAMMEN,A MACHINE THAT WOULD GO OF ITSELF 17-18(1986)(summa- rizing the use of the "machine"metaphor in constitutional rhetoric of the late eighteenth century) Led by Woodrow Wilson,Progressives later turned this rhetoric back on itself.See WooDRoW WILSON,CONSTITUTIONAL GOVERNMENT IN THE UNITED STATES 54-55(Transaction Publishers 2002)(19o8)("The government of the United States was constructed upon the Whig theory of political dynamics,which was a sort of unconscious copy of the Newtonian theory of the universe.In our own day,whenever we discuss the structure or development of anything, whether in nature or in society,we consciously or unconsciously follow Mr.Darwin;but before Mr.Darwin,they followed Newton."). 4 See,e.g.,Hugh Heclo,What Has Happened to the Separation of Powers?,in SEPARATION OF POWERS AND GOOD GOVERNMENT 131,134 (Bradford P.Wilson Peter W.Schramm eds.,1994)("[TJhe framework of 1787 has proven durable because the dynamics set up in its allo- cations of power have usually operated more or less as intended.");Arthur M.Schlesinger,Jr.,The Constitution and Presidential Leadership,47 MD.L.REV.54,65 (1987)("The separation of pow- ers is the vital means of self-correction in our system."). 5 THE FEDERALIST No.5I (James Madison),supra note I,at 321-22
2006] SEPARATION OF PARTIES, NOT POWERS 2313 powers (as well as by dividing the legislature into two chambers), the separation of powers would harness political competition into a system of government that would effectively organize, check, balance, and diffuse power. What is more, the system would be self-enforcing, relying on interbranch competition to police institutional boundaries and prevent tyrannical collusion. In the Framers’ Newtonian vision, the separation of powers was to be “a machine that would go of itself.”3 To this day, the idea of self-sustaining political competition built into the structure of government is frequently portrayed as the unique genius of the U.S. Constitution, the very basis for the success of American democracy.4 Yet the truth is closer to the opposite. The success of American democracy overwhelmed the Madisonian conception of separation of powers almost from the outset, preempting the political dynamics that were supposed to provide each branch with a “will of its own” that would propel departmental “[a]mbition . . . to counteract ambition.”5 The Framers had not anticipated the nature of the democratic competition that would emerge in government and in the electorate. Political competition and cooperation along relatively stable lines of policy and ideological disagreement quickly came to be channeled not through the branches of government, but rather through an institution the Framers could imagine only dimly but nevertheless despised: political parties. As competition between the legislative and executive branches was displaced by competition between two major parties, the machine that was supposed to go of itself stopped running. Few aspects of the founding generation’s political theory are now more clearly anachronistic than their vision of legislative-executive separation of powers. Nevertheless, few of the Framers’ ideas continue to be taken as literally or sanctified as deeply by courts and constitutional scholars as the passages about interbranch relations in Madison’s Federalist 51. To this day, Madison’s account of rivalrous, ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– 3 MICHAEL KAMMEN, A MACHINE THAT WOULD GO OF ITSELF 17–18 (1986) (summarizing the use of the “machine” metaphor in constitutional rhetoric of the late eighteenth century). Led by Woodrow Wilson, Progressives later turned this rhetoric back on itself. See WOODROW WILSON, CONSTITUTIONAL GOVERNMENT IN THE UNITED STATES 54–55 (Transaction Publishers 2002) (1908) (“The government of the United States was constructed upon the Whig theory of political dynamics, which was a sort of unconscious copy of the Newtonian theory of the universe. In our own day, whenever we discuss the structure or development of anything, whether in nature or in society, we consciously or unconsciously follow Mr. Darwin; but before Mr. Darwin, they followed Newton.”). 4 See, e.g., Hugh Heclo, What Has Happened to the Separation of Powers?, in SEPARATION OF POWERS AND GOOD GOVERNMENT 131, 134 (Bradford P. Wilson & Peter W. Schramm eds., 1994) (“[T]he framework of 1787 has proven durable because the dynamics set up in its allocations of power have usually operated more or less as intended.”); Arthur M. Schlesinger, Jr., The Constitution and Presidential Leadership, 47 MD. L. REV. 54, 65 (1987) (“The separation of powers is the vital means of self-correction in our system.”). 5 THE FEDERALIST NO. 51 (James Madison), supra note 1, at 321–22
L&P-BOOKPROOFS 09/13/06-11:39PM 23I4 HARVARD LAW REVIEW [Vol.I19:2311 self-interested branches is embraced as an accurate depiction of politi- cal reality and a firm foundation for the constitutional law of separa- tion of powers.In the Madisonian simulacrum of democratic politics embraced by constitutional doctrine and theory,the branches of gov- ernment are personified as political actors with interests and wills of their own,entirely disconnected from the interests and wills of the of- ficials who populate them or the citizens who elect those officials.Act- ing on these interests,the branches purportedly are locked in a perpet- ual struggle to aggrandize their own power and encroach upon their rivals.The kinds of partisan political competition that structure real- world democracy and dominate political discourse,however,are al- most entirely missing from this picture.6 As this Article describes,the invisibility of political parties has left constitutional discourse about separation of powers with no conceptual resources to understand basic features of the American political sys- tem.It has also generated judicial decisions and theoretical rationali- zations that float entirely free of any functional justification grounded in the actual workings of separation of powers.Ignoring the reality of parties and fixating on the paper partitions between the branches,the law and theory of separation of powers is a perfect fit for the govern- ment the Framers designed.Unfortunately,they miss much of the government we actually have. Ironically,one of the few places in constitutional law where parties do appear is in the most celebrated judicial opinion of the separation- of-powers canon,Justice Jackson's concurrence in the Youngstown case.?After laying out his now-familiar tripartite categorization of ex- ecutive action,Justice Jackson went on to emphasize that modern separation-of-powers analysis must understand the powers and moti- vations of the branches of government in light of their relation to po- litical parties: [The]rise of the party system has made a significant extraconstitutional supplement to real executive power.No appraisal of his necessities is real- istic which overlooks that he heads a political system as well as a legal system.Party loyalties and interests,sometimes more binding than law, extend his effective control into branches of government other than his 6 There have been a handful of important,recent exceptions.See,e.g.,MARK TUSHNET, THE NEW CONSTITUTIONAL ORDER (2003)(basing analysis of constitutional structures on the assumption that American government will be divided for years to come);Bruce Ackerman,The New Separation of Powers,I13 HARV.L.REV.633 (2000).Note the "new"in these titles.In an insightful essay,Professor James Gardner makes similar points to some we raise here about the effects of party competition on Madison's vision of separation of powers.See James A.Gardner Democracy Without a Net?Separation of Powers and the Idea of Self-Sustaining Constitutional Constraints on Undemocratic Behavior,79ST.JOHN'S L.REV.293,308 (2005). 7 Youngstown Sheet Tube Co.v.Sawyer,343 U.S.579(1952)
L&P - BOOKPROOFS 09/13/06 – 11:39 PM 2314 HARVARD LAW REVIEW [Vol. 119:2311 self-interested branches is embraced as an accurate depiction of political reality and a firm foundation for the constitutional law of separation of powers. In the Madisonian simulacrum of democratic politics embraced by constitutional doctrine and theory, the branches of government are personified as political actors with interests and wills of their own, entirely disconnected from the interests and wills of the officials who populate them or the citizens who elect those officials. Acting on these interests, the branches purportedly are locked in a perpetual struggle to aggrandize their own power and encroach upon their rivals. The kinds of partisan political competition that structure realworld democracy and dominate political discourse, however, are almost entirely missing from this picture.6 As this Article describes, the invisibility of political parties has left constitutional discourse about separation of powers with no conceptual resources to understand basic features of the American political system. It has also generated judicial decisions and theoretical rationalizations that float entirely free of any functional justification grounded in the actual workings of separation of powers. Ignoring the reality of parties and fixating on the paper partitions between the branches, the law and theory of separation of powers is a perfect fit for the government the Framers designed. Unfortunately, they miss much of the government we actually have. Ironically, one of the few places in constitutional law where parties do appear is in the most celebrated judicial opinion of the separationof-powers canon, Justice Jackson’s concurrence in the Youngstown case.7 After laying out his now-familiar tripartite categorization of executive action, Justice Jackson went on to emphasize that modern separation-of-powers analysis must understand the powers and motivations of the branches of government in light of their relation to political parties: [The] rise of the party system has made a significant extraconstitutional supplement to real executive power. No appraisal of his necessities is realistic which overlooks that he heads a political system as well as a legal system. Party loyalties and interests, sometimes more binding than law, extend his effective control into branches of government other than his ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– 6 There have been a handful of important, recent exceptions. See, e.g., MARK TUSHNET, THE NEW CONSTITUTIONAL ORDER (2003) (basing analysis of constitutional structures on the assumption that American government will be divided for years to come); Bruce Ackerman, The New Separation of Powers, 113 HARV. L. REV. 633 (2000). Note the “new” in these titles. In an insightful essay, Professor James Gardner makes similar points to some we raise here about the effects of party competition on Madison’s vision of separation of powers. See James A. Gardner, Democracy Without a Net? Separation of Powers and the Idea of Self-Sustaining Constitutional Constraints on Undemocratic Behavior, 79 ST. JOHN’S L. REV. 293, 308 (2005). 7 Youngstown Sheet & Tube Co. v. Sawyer, 343 U.S. 579 (1952)
2006] SEPARATION OF PARTIES,NOT POWERS 23I5 own and he often may win,as a political leader,what he cannot command under the Constitution.8 Justice Jackson astutely recognized that the separation of powers no longer works as originally envisioned because interbranch dynamics have changed with the rise of political parties,which in Youngstown had diminished the incentives of Congress to monitor and check the President.Yet this part of Justice Jackson's opinion has been ignored entirely.Even after decades of dissecting Justice Jackson's Youngs- town opinion,neither the Supreme Court nor any other federal court has ever quoted this critical insight,nor has it received much notice by legal scholars.Justice Jackson's sophisticated realism about the workings of government is widely admired by constitutional lawyers, but his most penetrating realist insight-recognizing party competi- tion as a central mechanism driving the institutional behavior that separation-of-powers law aims to regulate-has been missed. This Article seeks to recover and build upon Justice Jackson's in- sight,reenvisioning the law and theory of separation of powers by viewing it through the lens of party competition.We emphasize that the degree and kind of competition between the legislative and execu- tive branches vary significantly,and may all but disappear,depending on whether the House,Senate,and presidency are divided or unified by political party.The practical distinction between party-divided and party-unified government rivals in significance,and often dominates, the constitutional distinction between the branches in predicting and explaining interbranch political dynamics.Recognizing that these dy- namics shift from competitive when government is divided to coopera- tive when it is unified calls into question many of the foundational as- sumptions of separation-of-powers law and theory.It also allows us to see numerous aspects of legal doctrine,constitutional structure,com- parative constitutionalism,and institutional design in a new and more realistic light. The Article proceeds as follows.Part I lays out the conceptual case for switching the focus of separation of powers from branches to par- ties,arguing that political competition in government often tracks party lines more than branch ones.Recognizing that party competi- tion can either create or dissolve interbranch competition,depending on whether government is unified or divided by party,suggests that the United States has not one system of separation of powers but (at least)two.Part II looks more closely at these dual-party-divided 8 Id.at 654 (Jackson,J.,concurring). 9 It is cited,in passing,in two law review articles.See Norman C.Bay,Executive Power and the War on Terror,83 DENV.U.L.REV.335,345 (2005);Linda S.Greene,After the Imperial Presidency,47 MD.L.REV.99,106 (1987)
2006] SEPARATION OF PARTIES, NOT POWERS 2315 own and he often may win, as a political leader, what he cannot command under the Constitution.8 Justice Jackson astutely recognized that the separation of powers no longer works as originally envisioned because interbranch dynamics have changed with the rise of political parties, which in Youngstown had diminished the incentives of Congress to monitor and check the President. Yet this part of Justice Jackson’s opinion has been ignored entirely. Even after decades of dissecting Justice Jackson’s Youngstown opinion, neither the Supreme Court nor any other federal court has ever quoted this critical insight, nor has it received much notice by legal scholars.9 Justice Jackson’s sophisticated realism about the workings of government is widely admired by constitutional lawyers, but his most penetrating realist insight — recognizing party competition as a central mechanism driving the institutional behavior that separation-of-powers law aims to regulate — has been missed. This Article seeks to recover and build upon Justice Jackson’s insight, reenvisioning the law and theory of separation of powers by viewing it through the lens of party competition. We emphasize that the degree and kind of competition between the legislative and executive branches vary significantly, and may all but disappear, depending on whether the House, Senate, and presidency are divided or unified by political party. The practical distinction between party-divided and party-unified government rivals in significance, and often dominates, the constitutional distinction between the branches in predicting and explaining interbranch political dynamics. Recognizing that these dynamics shift from competitive when government is divided to cooperative when it is unified calls into question many of the foundational assumptions of separation-of-powers law and theory. It also allows us to see numerous aspects of legal doctrine, constitutional structure, comparative constitutionalism, and institutional design in a new and more realistic light. The Article proceeds as follows. Part I lays out the conceptual case for switching the focus of separation of powers from branches to parties, arguing that political competition in government often tracks party lines more than branch ones. Recognizing that party competition can either create or dissolve interbranch competition, depending on whether government is unified or divided by party, suggests that the United States has not one system of separation of powers but (at least) two. Part II looks more closely at these dual — party-divided ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– 8 Id. at 654 (Jackson, J., concurring). 9 It is cited, in passing, in two law review articles. See Norman C. Bay, Executive Power and the War on Terror, 83 DENV. U. L. REV. 335, 345 (2005); Linda S. Greene, After the Imperial Presidency, 47 MD. L. REV. 99, 106 (1987)