On December 24,1901, during the administration of Teddy Roosevelt, Attorney General Knox issued an opinion responding to the question whether President Roosevelt could fill a vacant office (the appraiser of merchandise in the district of New York) during the “current holiday adjournment” of the Senate. Both houses had adjourned from December 19, 1901 to January 6, 1902 for the Christmas and New Years holidays.
It is important to appreciate the factual context of this question. The Fifty-seventh Congress had been elected in 1900 but, in accordance with normal practice prior to the 20th Amendment, its first meeting was not until the first Monday of December 1901 (December 2). The taking of a holiday adjournment during the session that began in December had become a matter of routine after the Civil War, with such adjournments typically lasting a maximum of 19 days (there was one exception in 1877, when the adjournment lasted for 26 days). The 17-day adjournment considered by Knox fell squarely within this tradition.
The issue, according to Knox, was “[m]ay the President appoint as in the recess when the Senate had adjourned temporarily to a day certain?” In answering this question in the negative, Knox explained the parliamentary differences between a resolution for “final adjournment of Congress for the session,” on the one hand, and, on the other, “a merely temporary suspension of business from day to day or, when exceeding three days, for such brief periods over holidays as are well recognized and established and as are agreed upon by the joint action of the two Houses.” While Knox acknowledged that the latter type of temporary adjournment might constitute “a recess in the general and ordinary use of that term,” he denied that it could be “the recess” referred to by the Recess Appointments Clause.
Apart from the distinction between “a recess” and “the recess,” Knox grounded his opinion on two structural points. The first was that if temporary adjournments during a session qualified as “recesses” within the meaning of the RAC, then every day-to-day adjournment of the Senate would become a recess during which the president could fill vacant offices. Such a result would be inconsistent with both historic practice and with the purposes of the RAC.
The second point related to the length of the temporary commissions granted under the RAC. Knox saw this as an insuperable barrier to finding the RAC applicable to an “intra-session” recess:
If an adjournment during a session is a recess within the meaning of the clause of the Constitution in question, then the commission of an appointee of the President would extend to the end of the Fifty-seventh Congress, as the Constitution provides that it shall extend to the end of the next session, not the session within which the recess occurs. The only theory to defeat such a conclusion would be that the reassembling of the Senate after each adjournment constitutes a new session, a position wholly untenable in view of the constitutional provision as to adjournments during the session.
23 Op. Atty Gen. at 404 (emphasis in original).
As Professor Hartnett interprets it, Knox’s second point is that “permitting intrasession recess appointments results in some recess appointees holding commissions for substantially longer periods of time than other recess appointees.” Hartnett argues that this point has “little force” because “the commissions of intersession recess appointees could also be of considerably different lengths.”
It seems to me that Hartnett has misunderstood Knox’s primary concern, which was not that there would be differences in tenure among recess appointees. Rather, it was the fact that “intra-session” recess appointees would be permitted to serve over the course of two sessions and, in the case before him, nearly two full sessions.
Perhaps Knox thought it too self-evident to spell out why he considered this unacceptable. The RAC provides for expiration of the temporary commission at the end of the “next session” in order to ensure that advice and consent is utilized at the earliest opportunity. However, if a recess appointment made in December 1901 would not expire until the end of the session that would begin in December 1902, there would be little incentive for the president to seek advice and consent from the Senate when it returned from the holiday adjournment in January 1902. Thus, the “multi-session” interpretation conflicts with one of the purposes of the RAC, without any apparent offsetting benefit.
In addition, Knox must have realized that the Senate would find the “multi-session” result intolerable. To illustrate, consider the events that would transpire later in the Roosevelt administration. From March 4, 1903 to November 9, 1903, the Senate was in recess. During this period, Roosevelt made several recess appointments. On November 9, Congress assembled for an extraordinary session called by the President. This session lasted until December 7, when the regular session began. Although there was no time (or only a few seconds) separating the extraordinary session from the regular session, Roosevelt declared a “constructive recess” between the two sessions and made a number of recess appointments, including re-appointment of some of the individuals appointed in the prior recess.
This did not go over well in the Senate. The matter was referred to the Judiciary Committee, which issued a report in 1905 criticizing Roosevelt’s actions as unconstitutional. The committee rejected the concept of a “constructive recess,” stating that “[t]he word ‘recess’ is one of ordinary, not technical, signification, and it is evidently used in the constitutional provision in its common and popular sense.” 39 Cong. Rec. 3823 (Mar. 2, 1905). It then went on:
It was evidently intended by the framers of the Constitution that it should mean something real, not something imaginary; something actual, not something fictitious. They used the word as the mass of mankind then understood it and now understand it. It means, in our judgment, in this connection the period of time when the Senate is not sitting in regular or extraordinary session as a branch of the Congress, or in extraordinary session for the discharge of executive functions; when its members owe no duty of attendance; when its Chamber is empty; when, because of its absence, it cannot receive communications from the President or participate as a body in making appointments.
Id. (emphasis in original).
It should be noted that the committee report defines “recess” as the antithesis of “sitting in regular or extraordinary session.” The clear import is that the Senate can either be in recess, or in regular or extraordinary session, but not both at the same time. Thus, just as the report dismisses the notions of a “constructive recess” or a “constructive session,” it arguably weighs against that of an “intra-session recess.”
It has been argued that this report, by focusing on whether the Senate is actually in recess or in session, implicitly rejects Knox’s formalistic approach in which the recess/session determination is made solely by looking at the governing parliamentary resolutions. Hartnett goes further and suggests that the committee was likely aware of Knox’s opinion and, by failing to make any distinction between intra-session and inter-session recesses, was effectively registering its disagreement with Knox’s conclusion on that score.
To make the point more concrete, consider that on December 19, 1903, just 12 days after Roosevelt’s controversial appointments, Congress adjourned for the customary holiday period (which in this case lasted until January 4, 1904). If Hartnett is correct, merely waiting for 12 days would have, in the committee’s opinion, transformed Roosevelt’s unconstitutional appointments into lawful actions.
How likely is it that the committee would have taken this position? It seems possible, based on the reasoning of its report, that the committee would have adopted a practical, fact-based approach to determining whether the Senate was in recess or in session during the holiday adjournment, rather than relying on Knox’s formalistic approach. It seems less likely, however, that the committee would have considered the customary 16-day adjournment to be a “recess” within the meaning of the RAC.
What seems inconceivable, though, is that the committee would have agreed that appointments made during the December 1903 holiday adjournment would last for nearly a full year longer than Roosevelt’s actual appointments made earlier that month. It simply would make no sense for the committee to protest Roosevelt’s appointments while condoning a far greater infringement on the Senate’s prerogatives. This point is underscored by the committee’s outrage at the fact that Roosevelt continued in office individuals whose temporary commissions had expired at the end of the extraordinary session, notwithstanding the Senate’s failure to confirm their nominations.
Knox was therefore quite correct to reject the multi-session recess appointment theory. It is less clear that he was correct in concluding that the only alternatives to this theory would be (1) to consider every Senate adjournment as an “inter-session” recess which would be followed by a new “session” within the meaning of the RAC or (2) to defer to the Senate’s judgment in all cases as to whether a particular adjournment was an inter-session recess or merely an intra-session adjournment. There is another possibility, which is that only some adjournments, though formally denominated as intra-session by the Senate, have sufficient indicia of an inter-session recess to be considered as such for purposes of the RAC.
To explore this possibility, though, we will need to visit another administration and another Attorney General.