Yesterday I posted the amicus brief that I and a group of constitutional law scholars filed in the Court’s recess appointments case, Noel Canning. Taking a page from Eugene, I thought I’d put break our argument out into three blog posts this week, for those who are curious but don’t want to read through the entire PDF.
The first of the three issues we address is the meaning of the word “happen.” The Clause empowers the President to fill up “Vacancies that may happen during the Recess of the Senate,” and we argue that this originally meant that the office has to become vacant during the recess.
The original meaning of the Clause on this point is covered by the respondent and by the excellent “Brief of Originalist Scholars.” Section One of our brief discusses some of the same arguments and adds two other points based on historical practice — 1, that most of the Solicitor General’s attempts to assert counter-examples in early practice turn out to be mistaken, and 2, that for a long time, the Senate expressly resisted the executive branch’s expansive interpretation of the Clause. (These are I.C. and I.D. below.)
I’ve reprinted Section One of our brief (with a few footnotes omitted) below the fold:
I. Under The Text and Original Meaning of the Recess Appointments Clause, The Office Must Fall Vacant During The Recess
The Recess Appointments Clause states that “[t]he President shall have Power to fill up all Vacancies that may happen during the Recess of the Senate, by granting Commissions which shall expire at the End of their next Session.” U.S. Const. art. II, § 2, cl. 3. The most obvious meaning of the word “happen” is to “occur” or “take place,” and the most obvious meaning of the word “vacancy” is the event where an office becomes vacant. Cf. id. art. I. § 3, cl. 2 (“if [Senate] Vacancies happen by Resignation, or otherwise …”). This implies that the President may fill a vacancy by recess appointment only if the office arises during the recess.
The Solicitor General’s brief contends the word “happen” is ambiguous, and could mean that the vacancy merely exists during the recess. Even if there is some linguistic ambiguity, which we doubt, the “arise” interpretation is the best reading of the text, and is confirmed by the original history and primary purposes of the Clause. The “exist” reading rises or falls on an argument from longstanding practice—but this practice is not so longstanding or unchallenged as the Solicitor General’s brief suggests.
A. The Text Requires The Vacancy To Arise During The Recess
The most straightforward reading of the Clause is that the President may fill offices that became vacant during the recess, through death, resignation, or removal of the previous office holder, by conclusion of a fixed term, or (perhaps) by creation of the office. Now as in 1787, this is the natural reading of the word “happen.” The first American dictionary defined “to happen” as “to come by chance,” “to fall out,” “to come,” or “to befall”—all definitions that emphasize the initial occurrence of an event. Noah Webster, An American Dictionary of the English Language (1828). Nearly every founding-era English language dictionary agreed. See Brief of Scholars of the Constitution’s Original Meaning as Amici Curiae, at 5-7.
Even some of the executive opinions that developed the “exist” interpretation acknowledged that the “arise” reading was more consistent with the Constitution’s text. Attorney General Wirt called it “most accordant with the letter of the Constitution,” 1 Op. Att’y Gen. 631, 632 (1823). Similarly, in 1862, Attorney General Bates wrote to President Lincoln that “If the question were new, and now, for the first time, to be considered, I might have serious doubts” about the “exist[]” interpretation. 10 Op. Att’y Gen 356, 356 (1862).
B. The Clause’s Purpose Supports the “Arise” Interpretation
The Clause’s purpose, in the context of the Appointments Clause as a whole, confirms the most natural reading of the text. Outside of recess appointments, the Constitution requires the Senate’s participation—direct or indirect—in all appointments. First, all judges and principal officers must be appointed “by and with the Advice and Consent of the Senate.” U.S. Const. art. II, § 2, cl. 2. For inferior officers “Congress may by law” delegate the appointment power to the President or to others, without Senatorial confirmation. Id. Hence, in all regular appointments, the Senate must either consent to the nomination directly, or consent to a law that delegates the appointment. These provisions are an important check on the President’s ability to staff the executive branch with “unfit characters,” The Federalist No. 76, at 392 (Alexander Hamilton) (George W. Carey & James McClellan eds., 2001). Attorney General Edmund Randolph aptly observed: “The Spirit of the Constitution favors the participation of the Senate in all appointments.” Edmond Randolph’s Opinion on Recess Appointments (July 7, 1792), in 24 Papers of Thomas Jefferson 165 (Oberg & Looney eds., 2008).
The constitutional baseline indicates the proper scope of the Recess Appointments Clause. The Clause anticipates circumstances when the Senate cannot consent either directly or prospectively. In this sense the Clause is “to be considered as an exception to the general participation of the Senate. It ought too to be interpreted strictly.” Id.
The “arise” interpretation is consistent with this view. Under the “arise” interpretation, the President may make an immediate recess appointment and forgo Senate confirmations when (but only when) there is no realistic chance for the Senate to participate. Under the “exist” interpretation, by contrast, Presidents can wait until a “recess” in Senate business (however defined) and then appoint officers, including judges, for terms as long as two years, without troubling to obtain advice and consent. Worse yet, he can appoint officers already rejected by the Senate, which some Presidents have done.
The “arise” interpretation is admittedly not perfectly congruent with the evident purposes of the Clause. As a practical matter, the President may not be able to fill a vacancy that occurs very near the end of the session; conversely, there is no real need for recess appointments if the vacancy occurs shortly before the session begins. Attorney General Wirt relied on such pragmatic considerations for his nontextual interpretation in 1823. He worried about what would happen if a vacancy happened during the Senate’s session, but was not discovered until after the Senate recessed, or about “the sudden dissolution of that body by some convulsion of nature; the falling of the building in which they hold their sessions; a sudden and destructive pestilence, disabling or destroying a quorum of that body.” Wirt, supra, 1 Op. Att’y Gen. at 633. Pragmatic considerations of this sort are a potential strike against the “arise” interpretation. But the Solicitor General’s interpretation is far less tethered to the purposes of the Clause: the President can make appointments not just when the Senate is “unavailable” but when it disagrees with him. Moreover, the practical difficulties with the “arise” view can be avoided through the use of acting officers (in many cases) and by the President’s constitutional power “on extraordinary Occasions,” to call the Senate into session when it is in recess. U.S. Const. art. II, § 3.
A further difficulty with following the literal meaning of the Clause is that it might seem to allow Presidents to make recess appointments during the subsequent session so long as they arose during the recess. But this difficulty, if it is one, plagues the “exist” interpretation no less than the “arise” interpretation.
As a matter of practice, it is common ground among the various views that the vacancy must be “fill[ed]”—as well as “happen”—“during the recess of the Senate.” That reading is supported by a practice more longstanding and unbroken than any other discussed in this case. Presidents have always made recess appointments only when they maintained that the Senate was in recess. The Federalist states flatly that under the Clause the President must make the appointment “during the recess.” The Federalist No. 67, at 350 (Alexander Hamilton) (George W. Carey & James McClellan eds., 2001).
C. The Earliest Interpreters Endorsed the “Arise” Interpretation
In the Republic’s early decades, Congress and the executive branch both maintained that an office must become vacant during the recess to trigger the recess appointment power. A 1791 statute established an office and provided that “[i]f the appointment … shall not be made during the present session of Congress, the President may, and he is hereby empowered to make such appointments during the recess of the Senate, by granting commissions which shall expire at the end of their next session.” Act of March 3, 1791, ch. 15, § 4, 1 Stat. 199, 200. This provision would have been wholly redundant under the “exist” view of the Recess Appointments Clause.
The Washington Administration took the same view of the Clause. In the first formal executive opinion on the Clause, Attorney General Edmund Randolph confronted the executive branch’s power to make a recess appointment to the office of “chief Coiner” of the Mint—an office that had been created in April of 1792, a month before the Senate’s recess began. Is it, Randolph asked, a “vacancy which has happened during the recess of the Senate?” Randolph, 24 Papers of Thomas Jefferson, supra, 165. No, he answered: “It is now the same and no other vacancy, than that, which existed on the 2nd. of April 1792. It commenced therefore on that day or may be said to have happened on that day.” Id. Leading commentators on the Constitution, of both Federalist and Jeffersonian persuasions, agreed. See Michael Rappaport, The Original Meaning of the Recess Appointments Clause, 52 UCLA L. Rev. 1482, 1518-1521 (2005) (quoting Hamilton, a Federalist, and Tucker, a Jeffersonian).
President Washington repeatedly confirmed Randolph’s view of the Clause. Washington made recess appointments only when the office had become vacant during the recess. On several occasions Washington was confronted with vacancies late in the Senate session, when potential nominees could not be contacted before the end of the session. Id. at 1522. Washington and the Senate found a way to fill these offices consistent with the “arise” interpretation—the President would nominate an officer and the Senate would confirm him, all without the officer’s consent. Id. After the session was over, if the officer refused to serve, Washington treated that as a resignation, which created a new vacancy during the recess and could then be replaced by a recess appointment.
The Solicitor General claims President Washington broke from this practice twice—when he appointed Robert Scot as Engraver of the Mint, and William Clarke as United States Attorney of Kentucky. U.S. Br. 39-40. The history, however, strongly suggests both vacancies arose during a recess.
Washington himself said they arose during a recess. He wrote: “I nominate the following persons [including Scot] to fill the offices annexed to their names respectively; to which, having fallen vacant during the recess of the Senate, they have been appointed.” S. Exec. J., 3d Cong., 1st Sess. 142–43 (1793) (emphasis added). Likewise, “I nominate the following persons [including Clarke] to fill the offices annexed to their names, respectively, which became vacant during the recess of the Senate.” S. Exec. J., 4th Cong., 2d Sess. 216-17 (1796) (emphasis added). Even if Washington misapprehended the facts, he apparently believed he was acting within the recess appointments power because he supposed that the offices “became vacant during” a recess.
Moreover, it is unlikely that Washington was mistaken on the facts. The Act of April 2, 1792, made it “lawful for the functions and duties of Chief Coiner and Engraver to be performed by one person.” Act of April 2, 1792, ch. 16, § 3, 1 Stat. 246, 247. Henry Voigt therefore likely had been acting as Chief Coiner and Engraver. When he decided to extend the Engraver position to Scot, Washington likely created and filled that vacancy at once—during the recess.
As the Solicitor General’s brief suggests, the United States Attorney vacancy filled by Clarke also probably arose during a recess. Clarke’s predecessor, William McClung, “was confirmed on June 2, 1794. S. Exec. Journal, 3d Cong., 1st Sess. 160.” U.S. Br. 40 n.31. The Senate recessed on June 9. S. Exec. J., 3d Cong., 1st Sess. 162. “McClung had declined that appointment by October 1794.” U.S. Br. 40 n.31. The next Senate session began on November 21, 1794. S. Exec. J., 3d Cong., 2d Sess. 163. Unless McClung declined to serve during the seven-day window between his confirmation and the recess—a period shorter than the time it would take for a message to get from New York City to Kentucky—then he must have declined his commission during the recess, rendering the office vacant.
The Clarke appointment thus does not violate the “arise” interpretation or support the “exist” interpretation. But it does present an oddity: because Washington waited until the recess of 1796 to make the Clarke appointment, it appears that Washington filled a vacancy that “happened” during the 1794 recess “during” the 1796 recess (unless there was another nomination and appointment between McClung and Clarke, which the sources do not rule out). See Mary K. Bonsteel Tachau, Federal Courts in the Early Republic: Kentucky, 1789-1816, at 72 n.81 (suggesting that the post may have been filled, if briefly, by George Nicholas, in 1796). In a private letter, Washington indicates that he had simply forgotten about the appointment. Letter from George Washington to Timothy Pickering (Oct. 10, 1796). No one appears to have commented on the aberration. But under no theory does the Clarke appointment support the Solicitor General’s argument.
The Solicitor General makes a similarly unfounded claim that Thomas Jefferson’s early appointments of several district attorneys and marshals filled vacancies that did not arise during the recess. U.S. Br. 41-42. The Solicitor General’s evidence is that the offices were created by a statute on February 13, 1801, during the session. §§ 36-37, 2 Stat. 89, 99-100. But the very sections the Solicitor General cites provide that the previously-appointed marshals and district attorneys who lived in those districts would hold the new offices until the President removed them. Id. So the new offices were likely filled by the old district attorneys and marshal until Jefferson nominated their replacements, which he did during the recess.
Nor is there anything in Jefferson’s letter to Wilson Nicholas, relied upon by the Solicitor General, U.S. Br. 42, suggesting he took a broad view of the recess appointment power. Though Jefferson described the Clause as “susceptible of both constructions,” his chief concern seems to have been that some of the legal fictions used in the previous administrations were too broad. 36 Papers of Thomas Jefferson 433 (2009). He said: “if we find that any of our cases go beyond the limits of such a rule, we must consider what will be the best way of preventing their being considered as authoritative examples.” Id. There is no reason to think that Jefferson supported a broader view of the power than his predecessors had.
Madison’s practice is somewhat less clear. His only articulated position on the question, in connection with his delayed appointment of Andrew Jackson to a generalship, see Resp. Br. 39, squarely embraced the “arise” view. But he made five appointments that may or may not have conformed to that view. Compare U.S. Br. 42-43 with Resp. Br. 44. One was the recess appointment of Theodore Gaillard to a judge-ship vacated by the resignation by Dominic Hall. Hall’s date of resignation is uncertain, but it may have been during the session. Compare Edward A. Hartnett, Recess Appointments of Article III Judges: Three Constitutional Questions, 26 Cardozo L. Rev. 377, 400-01 (2005) (arguing it was during the session) with Rappaport, Original Meaning, supra, at 1534 n.142 (arguing that this is not clear). In any case, Gaillard declined the recess appointment.
Two other appointments were to offices created on March 3, 1815, the last day of the Senate’s session. On that day, Madison signed legislation creating two new offices, which he later filled by recess appointment. Looking solely at the Senate Journal, it appears that he signed the bill before the Senate recessed, but a comparison to the House’s records in the Annals of Congress suggests that the Senate may have announced its recess before receiving the message from the Clerk of the House that Madison had signed the legislation. Compare S. Journal, 13th Cong., 3d Sess. 689-90 (Mar. 3, 1815) with 28 Annals of Cong. 1274 (Mar. 3, 1815) (House of Representatives).
Finally, Madison recess appointed the first U.S. Attorney and Marshal for the Territory of Michigan, although those offices were created during the session. U.S. Br. 43. We are not aware of any evidence that would square these two appointments with the “arise” view. Yet we are also less ready than the Solicitor General to infer that Madison supported the “exist” view. In none of these cases is there evidence that Madison actually considered (or was aware of) the dates of the vacancies, so we are inclined to think his explicit embrace of the “arise” view in the Jackson episode constitutes the better evidence of Madison’s view of the Clause’s meaning.
D. Modern Practice Has Diverged From The Original Meaning, But Not To The Extent The Government Claims
The only plausible argument against the original “arise” interpretation of the Clause is the claim that there has been a long contrary practice. Under the usual account, the “arise” interpretation was abandoned in an opinion by Monroe’s Attorney General William Wirt in 1823. See Hartnett, supra, at 401.
But that account downplays a great deal of contrary evidence. First, the executive branch subsequently issued opinions that contradicted Attorney General Wirt’s opinion. One later Attorney General opinion declared, “[i]f vacancies are known to exist during the session of the Senate, and nominations are not then made, they cannot be filled by executive appointments in the recess of the Senate.” 4 Op. Att’y Gen. 361, 363 (1845) (Mason).
Several other Attorneys General expressed similar views, suggesting that the word “happen” meant that even if a vacancy need not arise during the recess, it nonetheless had to continue by chance and not design. 2 Op. Att’y Gen. 525, 528 (1832) (Taney) (“If it falls out that, from death, inadvertence, or mistake, an office … is, in the recess, found to be vacant … the President may fill it . … But vacancies are not designedly to be kept open by the President until the recess, for the purpose of avoiding the control of the Senate.”) (emphasis in original); 3 Op. Att’y Gen 673, 674 (1841) (Legare) (acknowledging possibility that “happen” “impl[ies] something fortuitous and unexpected.”); 4 Op. Att’y Gen. 523 (1846) (Mason); see also 3 Joseph Story, Commentaries on the Constitution of the United States 416-417, § 1553 (1st ed. 1833) (“The word ‘happen’ had relation to some casualty, not provided for by law”); Michael B. Rappaport, Why Nonoriginalism Does Not Justify Departing from the Original Meaning of the Recess Appointments Clause, at 19-20 (draft, Sep. 19, 2013).
On that view, a vacancy happened during the recess only if some unexpected occurrence caused the vacancy to go unfilled during the session. This view is inconsistent with the Solicitor General’s account and its position in this case. There was nothing accidental or unexpected about these vacancies; nominations had been made during the session.
More importantly, the Senate has long contested the “exist” construction. In October of 1862, Attorney General Bates had written that he “might have serious doubts” about the “exist” interpretation as an original matter, but for the executive branch’s contrary practice and the “unbroken acquiescence of the Senate.” 10 Op. Att’y Gen. 356, 356. In December 1862, the Senate took issue with Bates’s claim of ac-quiescence. After noting the increasing number of recess appointments, the Senate instructed the Judiciary Committee to investigate the constitutionality of recess appointments “which have not occurred during the recess of Congress, but which existed at the preceding session of Congress.” Cong. Globe, 37th Cong., 3d Sess. 100 (Dec. 16, 1862).
On January 28, 1863, the Judiciary Committee concluded in strong terms that the practice was unconstitutional. Senator Howard reported: “We think the language too clear to admit of reasonable doubt, and that, upon principles of just construction, this period must have its inceptive point after one session has closed and before another session has begun.” See S. Rep. No. 80, at 3 (37th Cong., 3d Sess.) (Jan. 28, 1863). The Committee argued that a “construction” based on continued practice should guide constitutional interpretation only “in considering the meaning and intent of a doubtful clause.” Id. at 7 (emphasis in original). The Recess Appointments Clause was unambiguous, while Wirt’s interpretation was “only the invention of a phrase not contained in the text, giving it an effect which the text itself, by the ordinary rules of interpretation, forbids. No instrument could long endure such experiments.” Id. at 5.
The same day, the Senate moved to implement the Judiciary Committee’s constitutional objections. It passed a statute that eliminated salaries for all recess appointments “which vacancy existed while the Senate was in session.” Act of Feb. 9, 1863, ch. 25, § 2, 12 Stat. 642, 646. In this statute, signed by President Abraham Lincoln, Congress endorsed the original “arise” view. Cong. Globe, 37th Cong., 3d Sess. 565 (January 28, 1863) (statements of Trumbull, Sherman, Fessenden); Rappaport, Nonoriginalism, supra, at 22-24. Shortly thereafter, two federal courts also rejected the executive branch’s position.
The pay statute’s direct rejection of the “exist” view continued from 1863 to 1940, when it was amended to permit salaries to non-“arise” recess appointees in limited circumstances. Act of July 11, 1940, ch. 580, § 1761, 54 Stat. 751, 751, now codified at 5 U.S.C. § 5503. It is unclear whether the 1940 amendments reflect a twentieth-century change of heart on the constitutional question. The Senate re-port at that time noted only “that the law as it stands may cause serious injustice in preventing the payment of salary to those classes of appointees whose problems will be corrected by this amendment.” S. Rep. No. 1079, at 2 (76th Cong., 1st sess. 1939); see also H. Rep. No. 2646, at 1 (76th Cong., 3d Sess. 1940) (existing statute “from a practical standpoint frequently creates difficulties”). And in requesting the amendments, the executive branch did not even mention the constitutional objections that had originally motivated the statute, nor did it cite the Senate Judiciary report. See Letter from Attorney General Frank Murphy, printed in S. Rep. No. 1079 at 2.
Significantly, even the new statute does not allow recess appointees to be paid in all circumstances, as would be the case if Congress had completely ac-quiesced in the executive practice. This suggests that the Senate has never accepted the “exist” interpretation across the board. Rappaport, Nonoriginalism, supra, at 33-35.
All of that said, it is true that a long string of executive branch practice, and arguably Congress’s role in liberalizing the 1940 pay statute, have reflected a broader construction of the word “happen.” Yet this Court has recognized the authority of modern practice only in cases of ambiguity. See Pocket Veto Cases, 279 U.S. 655, 689-90 (1929); Youngstown Sheet & Tube Co. v. Sawyer, 343 U.S. 579, 610-12 (1952) (Frankfurter, J., concurring). The Court has generally rejected the proposition that practice may displace text and original meaning where it is clear. Myers v. United States, 272 U.S. 52, 174-175 (1956); Walz v. Tax Comm’n, 397 U.S. 664, 678 (1970); INS v. Chadha, 462 U.S. 919, 944 (1983); Bowsher v. Synar, 478 U.S. 714, 723 (1986); Printz v. United States, 521 U.S. 898, 905 (1997).
Acceptance of the Solicitor General’s view on the power of the President to make recess appointments to offices that became vacant while the Senate was in session would greatly increase the importance of modern practice for constitutional interpretation, at the expense of text and early history. As will be seen, even the Solicitor General is unwilling to live with that approach on a consistent basis.