A Note on the Separation of Powers

The Heritage Guide to the Constitution

A Note on the Separation of Powers

Article I

Separation of Powers is a fixture of the federal constitutional system even as its contours are uncertain and contested. The theory envisions three principal powers of government—legislative, executive, and judicial—and contemplates three distinct branches (the Founders often used the term “departments”) to wield those powers. At the core, the legislative power makes laws, the executive power executes them, and the judicial power resolves disputes between parties.

While the theory does not dictate the ideal structure of the three branches, there were (and are) some patterns in governments that embodied the system. A multimember body typically wielded the legislative power, while numerous judges exercised the judicial power individually or in smaller units. The structure of the executive might vary in size from a small assembly to a single man.

The separation concept developed over time in England as Parliament acquired exclusive authority to make law (by limiting the king’s “prerogative” to do the same), as the judicial power became the exclusive province of the courts (by gradually separating itself from the executive and by virtue of statutes ensuring a measure of independence), and as the executive became wedded to executing and enforcing the laws of Parliament and the judgments of courts. Writing in the mid-eighteenth century, Baron de Montesquieu argued that keeping the three core powers in separate hands helped prevent tyranny. If one branch acted tyrannically, the others might check it, something impossible when all three powers of government were in the hands of one entity. In other words, Montesquieu believed that the diffusion of power helped to secure liberty. Yet he did not believe in a strict separation, for he asserted that the executive ought to have a check on the legislative and that the latter ought to be able to disband the executive’s army.

Shortly after the break with England, early American state constitutions sometimes paid lip service to the idea of separation of powers, containing provisions calling for separation. But quite often, state legislatures were too strong and their executives too weak. For instance, confronting a plural executive council that lacked a veto, the Pennsylvania unicameral assembly repeatedly assumed the powers of the other two branches. Some constitutions (New York and Massachusetts) created executives strong enough to withstand their legislatures and their vortex-like tendency to seize powers vested elsewhere. Meanwhile, the Articles of Confederation, crafted in 1777 and ratified in 1781, paid no heed to the idea of separation at all. It vested the Continental Congress with executive power as well as some legislative and judicial functions.

Most delegates to the Philadelphia convention were determined to avoid the mistakes that plagued the state constitutions. Others were content to have the legislature dominate the executive, conceiving of the latter as but the agent of the former. Even so, there always was a consensus that there would be three branches (legislative, executive, and judicial) vested with the three principal powers.

The introductory sections of the first three articles, shaped by the drafting skills of Gouverneur Morris and the Committee of Style, reflect that consensus. The implication is that only Congress can exercise legislative power (Article I). Only the president can exercise executive power (Article II). And only the federal courts can exercise judicial power (Article III). To make that separation more meaningful, the Incompatibility Clause (Article I, Section 6, Clause 2) bars legislators from simultaneously serving in the executive and judicial branch. The Constitution’s salary and tenure protections for the president and the judges also help ensure that both have some measure of independence from a potentially overbearing Congress.

But the Framers also decided to grant to Congress and the president what James Madison termed “partial agency” (The Federalist No. 47) over certain acts of the other branches. There would, in other words, be some cross-branch checks. The Senate inf luences the exercise of two powers traditionally deemed executive: appointments and treaty-making. The legislature has a judicial power over executive and judicial officers through impeachment. And the executive may check legislation, using his veto, as well as recommend measures to be law and (typically) enjoy some discretion in enforcing the law. During the ratification process, these cross-branch checks led some to complain that the Philadelphia convention had not fashioned a proper separation of powers. For instance, some Anti-Federalists protested that the Senate should not have a treaty or appointment role and that the Constitution ought to have relied upon a separate executive council instead. James Madison retorted that a well-functioning separation of powers required some cross-branch checks and that Montesquieu himself never favored a hermetic separation. Checks and balances had a role in ensuring a more meaningful separation of powers for they helped ensure that no one branch would dominate.

Once one recognizes that the Constitution does not neatly separate the three powers, some difficult questions arise. First, do the functions of the three powers overlap, meaning that more than one branch may exercise the same function? Second, are the Constitution’s checks meant to be the only legal constraints on the exercise of federal powers or does the Constitution permit the imposition through legislation or custom of additional checks?

The first question arose during the Washington administration. James Madison and Alexander Hamilton debated the legality of President George Washington’s Neutrality Proclamation (1793), which sought to assure non-involvement in the war between Great Britain and France. In a series of essays, writing as “Helvidius,” and contradicting his earlier support of partial agency, Madison advanced the position of clear and defined separation. He insisted that only Congress could declare American neutrality in the war between France and England and that the Constitution would never permit more than one branch to have the same power. “A concurrent authority in two independent departments, to perform the same function . . . would be as awkward in practice, as it is unnatural in theory.” Parallel power to declare neutrality, said Helvidius, would upend the “partition . . . so carefully made among the several branches” and throw the powers into “absolute hotchpot,” exposing them to a “general scramble.” Madison’s scorn was directed at his erstwhile ally.

Alexander Hamilton (writing as “Pacificus”) defended Washington’s Neutrality Proclamation, arguing that in the course of executing the law of nations the executive could declare the nation neutral. The executive had such authority, Pacificus wrote, because it enjoyed a generic foreign affairs power and because the neutrality declaration helped to enforce the laws of neutrality, laws the president was duty-bound to enforce. Hamilton admitted that perhaps Congress also might decide whether the nation could remain neutral in light of its treaty obligations, but denied that this legislative authority excluded a “similar right of Judgment, in the execution of [the executive’s] own functions.”

Whether the Constitution permits the imposition of additional cross-branch constraints on the exercise of federal powers also arose in the early years. In No. 48, Madison noted that legislatures were apt to usurp the powers of the other branches because they could “with the greater facility, mask, under complicated and indirect measures, the encroachments.” Representative James Madison spoke to this matter in 1789, asserting that the president had an absolute power to remove executive officers and denying that Congress could by statute grant the Senate a check on the president’s power. “If the constitution has invested all executive power in the president, I venture to assert, that the legislature has no right to diminish or modify his executive authority.”

The Supreme Court has long recognized the core functions of the three departments and that these functions overlap to some extent. “The difference between the departments undoubtedly is, that the legislature makes, the executive executes, and the judiciary construes the law,” said Chief Justice John Marshall in Wayman v. Southard (1825). The Court also acknowledged that Congress might leave some discretion to the other departments and in so doing suggested that the powers overlap. When Congress makes rules about commerce, it exercises legislative power. But when it conveys limited discretion to the executive, the latter exercises executive power as it makes similar commercial rules.

The Supreme Court’s “nondelegation doctrine” rests on the notion that the three powers overlap to some extent but that at some point, should Congress vest too much discretion with others (when a delegation lacks an “intelligible principle”), Congress unconstitutionally delegates its legislative powers. (See Article I, Section 1, Clause 1, Legislative Vesting Clause). Because the modern Court has been rather lax in policing delegations, it is clear that Congress can convey a great deal of discretion to the executive and judicial branches to fashion rules.

The question whether and how much Congress can constrain the functions of the other two branches is much more disputed. The Supreme Court itself has been inconsistent, particularly with respect to statutes affecting the executive. In Ex Parte Garland (1866), the Court declared that the president’s pardon power was “not subject to legislative control.” Similarly, in Myers v. United States (1926), the Court refused to allow the Senate to participate in the executive power of removal. Yet in a series of cases from Humphreys’ Executor v. United States (1935) to Morrison v. Olson (1988), the Court sanctioned Congress’s imposition of “for cause” removal restraints on the president’s power to remove officials, both “quasi-judicial” and executive.

Some justices have suggested that the Court has strictly limited Congress’s ability to restrain or curb “express” presidential powers but has been more tolerant with respect to implied powers, such as removal. But insofar as removal is part of the Vesting Clause’s grant of executive power (as the Supreme Court held in Myers), it is no less express than the President’s ability to commute sentences, a power that flows from the Pardon Power Clause.

Other justices have asserted that the Court’s jurisprudence has turned on the whether Congress, in its statutes, has sought to aggrandize itself. When Congress has usurped powers, executive or judicial, the Court has struck it down. When Congress merely weakens the other branches without also aggrandizing itself, say by merely limiting when those branches may remove officers or by requiring public access to their records and meetings, the Courts have been far more forgiving. Yet one might suppose that any legislation that weakens or hampers the other two branches advantages Congress in interbranch struggles.

Needless to say, distinguishing what the Court believes to be permissible congressional regulation of executive power from impermissible encroachments remains a perplexing task. Indeed, the best explanation for the Court’s case law might be that Court has been inconsistent, vacillating between a devotion to formalism and an embrace of functionalism.

Formalists tend to find bright-line, determinate rules in the Constitution’s first three articles. They are drawn to the view that when the Constitution grants a power to one branch, no other branch may exercise it. And they have supposed that unless the Constitution expressly provides otherwise, Congress may not interfere with how another branch exercises its powers. For instance, many formalists are drawn to the view that Congress cannot limit the removal power, the power to appoint (Article II, Section 2, Clause 2), or the power to pardon (Article II, Section 2, Clause 1).

Functionalists favor flexibility in the separation of powers and are more comfortable with the idea that a function might simultaneously and constitutionally rest with more than one branch. Their reasoning sometimes rests less upon text, structure, and history and more on a sense that the Constitution fashions a flexible separation of powers. Generally, when functions or powers overlap, functionalists tend to believe that Congress’s statutes should prevail over contrary executive and judicial conclusions, likely because the Supremacy Clause provides that federal statutes are supreme and because Congress is the most representative of the branches. Functionalists also are more apt to believe that Congress can properly regulate the exercise of executive and judicial powers. The source of this power to regulate is often said to be the Necessary and Proper Clause (Article I, Section 8, Clause 18).

When the Court embraces formalism, the majority opinion asks whether the Constitution contains a clear textual commitment of a function to one branch, or alternatively, whether one branch is seeking to perform a function properly the domain of another branch. In INS v. Chadha (1983), for example, the Court struck down the so-called “legislative veto” exercised by one house of the Congress because the Court believed that the veto permitted a chamber to make law outside the confines of the Presentment Clause (Article I, Section 7, Clause 2). And in Bowsher v. Synar (1976), the Court declared unconstitutional a law that gave the comptroller general, part of the legislative branch, the executive power of determining what kind of budget reductions should be made under a statute.

When the Court adopts functionalist standards, it asks whether one branch has unduly interfered with the essential functions of another branch, thus allowing for a wide range of variable overlap. In Morrison v. Olson, the Court declared that when Congress established the office of independent counsel, it did not unduly hamper the president’s essential executive power of enforcing the laws, even though the president had little supervisory or removal power over the independent counsel. In dissent, Justice Antonin Scalia adopted formalist reasoning. He insisted that prosecution was wholly within the executive’s domain and that Congress could not strip away that function and vest it in an independent officer.

Functionalist opinions have implicitly given the Court’s imprimatur to the creation of what is often called a “fourth branch” of government. So-called independent agencies, not part of the executive branch, such as the Securities and Exchange Commission, the Federal Election Commission, etc., execute federal law through civil prosecutions. But because such agencies typically create rules and adjudicate violations, the agencies and their officers are sometimes seen as “quasi-legislative” and “quasi-judicial.” Functionalists in the academy tend to welcome cases that have enlarged congressional flexibility to create new governmental structures. Formalists are more apt to lament the fourth branch as inconsistent with a Constitution that, in the first three articles and the Oaths Clause (Article VI, Clause 2), presupposes three branches and three governmental powers.

As hinted at above, functionalist opinions have permitted the concentration of the three powers within particular governmental subunits, allowing one entity to make binding rules, adjudicate violations of the law, and prosecute legal violations before the courts. This concentration of the three powers, which displeases some formalists, is found in both executive and independent agencies. In contrast, many functionalists see wisdom in vesting the three powers of government in a body of experts able to change law more rapidly. (See essay “A Note on Administrative Agencies”).

Lest too much be made of functionalism and formalism, it is important to recognize that they represent general approaches and do not inevitably yield conf licting answers. In Clinton v. City of New York (1998), the Court struck down the Line Item Veto Act, an act ceding the president the power to cancel certain newly enacted provisions of law. Both the majority and the dissenting groups of justices were composed of interesting amalgams of functionalist and formalist justices.

The Court’s stance toward federal judicial power is equally complicated, with respect to both both what non-judicial authority Congress may convey and Congress’s attempts to exercise judicial authority. In some laws, Congress has assigned rulemaking, law execution, and appointing authority to judges. The Court has approved such statutes when they concern tasks related to the judicial power, while reserving the right to strike down laws that assign tasks wholly alien to the exercise of the judicial power. Hence Congress can delegate to federal courts the powers to fix punishments and appoint prosecutors, as these are linked to the judicial function. Yet it likely cannot grant the courts authority to create rules of military discipline or to appoint a deputy secretary of state.

The Court also has limited Congress’s ability to assume judicial power. The Constitution contains some hints that Congress cannot exercise judicial power itself. The Bill of Attainder Clause (Article I, Section 9, Clause 3) prohibits Congress from declaring that certain individuals are guilty of a crime or shall be punished. The Ex Post Facto Clause (Article I, Section 9, Clause 3) bars Congress from changing criminal law and punishment and applying those laws to prior acts, a prohibition that means that Congress cannot effectively dictate guilty verdicts in pending criminal cases.

Yet those rules apply only to punishment. Civil laws that affect exercises of judicial power face fewer such constraints. While Congress cannot declare who wins in an ongoing civil case nor revive a case in which a court has entered a final judgment, it may change the underlying substantive law applicable to an ongoing civil case and create new causes of action applicable to acts occurring in the past. In a sense there is a formal (if implied) rule against legislative usurpations of judicial power, a rule said to reflect the Founders distaste for the sorts of legislative aggrandizement that was all too common in the states. Nonetheless, if Congress is displeased with the result in a particular civil case, pending or otherwise, it can come close to dictating the outcome if it enacts new legislation changing the substantive law and creating a cause of action that reaches prior acts.

With respect to congressional regulation of exercises of judicial power, Congress has long dictated quorum rules, rules of procedure, and the like. It also has imposed rules about when courts must meet. These powers likely stem from Congress’s power to create necessary and proper laws in order to carry federal powers into execution. The limits of congressional power remain uncertain, and it is unclear whether Congress could, for example, require a supermajority of a multi-member panel in order for a court to strike down laws as unconstitutional or whether it could require judges to write opinions in all cases.

The Supreme Court also has cited the Constitution’s separation of powers as a basis for a number of implied branch privileges. In United States v. Nixon (1974), the Supreme Court said that executive privilege (like judicial privilege) was an implicit feature of the separation of powers, reading the Constitution as ceding to each branch a limited dominion over conversations and documents within it. In Nixon v. Fitzgerald (1982), the Court rooted its finding of absolute presidential immunity for official acts in the separation of powers. Obviously, not all immunity claims are successful, as Clinton v. Jones (1997) rejected the view that the separation of powers cedes the president a temporary immunity from civil suit while in office. Hence while the separation of powers shields the president from suits contesting his official acts, it poses no barrier to suits challenging his private conduct.

In modern times, some scholars have expressed skepticism about the extent to which the Constitution’s system of separation of powers has actually prevented tyrannies and concentrations of federal power. Some scholars have said that for the separated powers system to work effectively, different political parties must control the various branches, for co-partisans are unlikely to check their allies in the other branches. Ambition does not much counteract ambition when one branch generally sees eye-to-eye with another. Others have asserted that courts will do little to check the political branches during crises and only rise to the occasion when an emergency has passed, meaning that temporary tyrannies are possible. Finally, some scholars have suggested that the separation of powers system is in shambles because Congress has delegated so much authority to the president that the courts cannot meaningfully check the latter. These scholars suggest that public opinion acts as the chief, if not the only, restraint on the executive.

Saikrishna B. Prakash

David Lurton Massee, Jr. Professor of Law, Univeristy of Virginia School of Law

Gerhard Casper, Separating Powers: Essays on the Founding Period (1997)

A Federalist Society Symposium, The Presidency and Congress: Constitutionally Separated and Shared Powers, 68 Wash. U.L.!. 485 (1990)

W. B. Gwyn, The Meaning of the Separation of Powers (1965)

Gary Lawson, The Rise and Rise of the Administrative State, 107 Harv. L. Rev. 1231 (1994)

Daryl J. Levinson and Richard H. Pildes, Separation of Parties, Not Powers, 119 Harv. L. Rev. 2311 (2006)

John F. Manning, Separation of Powers as Ordinary Interpretation, 124 Harv. L. Rev. 1939 (2011)

Eric A. Posner & Adrian Vermeule, The Executive Unbound (2011)

Saikrishna Prakash, Regulating Presidential Power, 91 Cornell L. Rev. 215 (2005)

Peter L. Strauss, Formal and Functional Approaches to Separation-of-Powers Questions--A Foolish Inconsistency?, 72 Cornell L. Rev. 488 (1987)

M.J.C. Vile, Constitutionalism and the Separation of Powers (1998)

Wayman v. Southard, 23 U.S. (10 Wheat.) 1 (1825)

Ex parte Garland, 71 U.S. (4 Wall.) 333 (1867)

Myers v. United States, 272 U.S. 52 (1926)

Humphrey's Executor v. United States, 295 U.S. 602 (1935)

Wiener v. United States, 357 U.S. 349 (1958)

United States v. Nixon, 418 U.S. 683 (1974)

Buckley v. Valeo, 424 U.S. 1 (1976)

Nixon v. Fitzgerald, 457 U.S. 731 (1982)

INS v. Chadha, 462 U.S. 919 (1983)

Bowsher v. Synar, 478 U.S. 714 (1986)

Morrison v. Olson, 487 U.S. 654 (1988)

Plaut v. Spendthrift Farm, Inc., 514 U.S. 211, 218 (1995)

Clinton v. Jones, 520 U.S. 681 (1997)

Clinton v. City of New York, 524 U.S. 427 (1998)