Adding
Recess
to
the
of
Unilateral
Appointments
President's
"Tool
Chest"
Political Research Quarterly
Volume60 Number4
December2007 645-654
? 2007 Universityof Utah
10.1177/1065912907307541
http://prq.sagepub.com
hosted at
http://online.sagepub.com
Powers
Ryan C. Black
AnthonyJ. Madonna
Ryan J. Owens
WashingtonUniversity,Saint Louis, Missouri
Michael S. Lynch
Universityof Kansas, Lawrence
Inthe struggleto controlthefederalbureaucracy,
presidentshaveanoverlookedbutpowerfultool:therecessappointment.By makingrecessappointments,
presidentscanfill vacancieswithouttheadviceandconsentof theSenate.The
unilateral
howrecessappointments
authorsdelineatethreeconditionsthatdefinepresidential
powersanddemonstrate
to imporfit withinthatparadigm.Presidents,the authorsargue,shouldbe morelikelyto makerecessappointments
a
data
of
tantpolicy-making
The
authors
set
everycivilian
agencies.
compile
positions,namely,majorindependent
nomination
andrecessappointment
between1987and2004.Aftercontrollingforotherfactors,theauthorsfindstrong
supportfortheirtheory.
Keywords: recess appointments;presidentialpowers; Congress-presidentialrelations;unilateralpresidency;separation of powers
June 15, 2005, the FederalElections CommisOnsion
a
release
(FEC)published press
announcing
the resignation of Commissioner Bradley Smith.
Smith's departuremeant that four of the six sitting
commissioners had either resigned or were serving
expired terms. President George W. Bush long had
opposed the Bipartisan Campaign Finance Reform
Act (BCFRA), a key statuteenforcedby the FEC, and
now the future of the act depended on how Bush's
nominees would interpretand apply it.' Legislators
who had passed the law and the groupswho had supported it all held their breath, waiting to see whom
the presidentwould nominate-if anyone.
Many supportersof the BCFRA expressedconcern
that PresidentBush would bypass the Senate and use
his constitutionalpowerto recess appointthe new FEC
commissioners.The WashingtonPost suggesteddoing
so would be a "grossmisuse of the recess appointment
power"("No Recess" 2005). Writingto the president,
the sponsorsof the BCFRArequestedthat"younot use
your recess appointmentpowers to fill the current
vacanciesat thatagency"(McCainet al. 2005).2
But the presidentdid exactly that. Ratherthan put
his desired appointees, all of whom expressed some
reservations about the BCFRA, through a grueling
hearingand possible rejectionin the Senate,the president waited for the Senate to recess. Then he
appointedthree candidatesto the commission, using
his constitutionally granted power to fill vacancies
while the Senate is in recess withoutthe advice and
consent of the Senate. The impact of these recess
appointmentswas profound and immediate. As the
president likely anticipated,the new commissioners
continued to interpretfederal campaign finance law
in a mannerthat made it nearly impossible to prove
that groups engaged in illegal, coordinatedcampaign
activity (Schor 2006).
Authors' Note: This articlebenefitedfromcommentsmadeby participantsat the 2007 Midwest Political Science Associationmeetings, the 2007 SouthernPoliticalScienceAssociationmeetings,and
the WashingtonUniversityCenterfor EmpiricalResearchin the
Law workshopseries.The authorsalso thankGaryJ. Miller,James
E Spriggs, Andrew D. Martin,and the anonymousreviewersat
PoliticalResearchQuarterlyfor theirhelpfulcomments.Madonna
thanksthe WeidenbaumCenteron the Economy,Government,and
Public Policy. Owens thanks the George H. W. Bush Library
FoundationandLynchthanksthe EisenhowerInstitutefor generous
financialsupportof this research.
645
646
Political ResearchQuarterly
This FEC episode highlightstwo importantfeatures
of the president'srecessappointmentpower.First,presidents can utilize recess appointmentsto bypass a
Senate or senatorsthat may be hostile to their nominees. Second, recess appointmentsto majorindependent regulatoryboards, like the FEC, can have real
policy impact. Together,these featuresgive the president a powerful tool to affect policy with limited
interferencefrom the Senate.
Recently,presidentialuse of recessappointmentshas
increased.3This increasein recess appointmentscoincides with a growth in contemporarywork by presidency scholars,whichdemonstrateshow presidentscan
and do use "unilateralpowers"to make policy.4We
if it has the
define a presidentialpower as "unilateral"
three
attributes:
the
first,
following
presidentmanipulates ambiguitiesin the Constitution;second,the president must use the power first and alone, puttingthe
legislatureand courtsin a positionsuch thatthey must
reactto the president'saction;andthird,the president's
actionmustaffectpolicy (Howell2003, 2005; Moe and
Howell 1999a, 1999b). Examplesof unilateralpowers
identified by presidentialscholars include executive
orders(Mayer 1999, 2001; Howell 2003), presidential
signing statements(Cooper2002, 2005), and executive
agreements(Moe andHowell 1999a).
In this article,we arguethatscholarsshouldinclude
recess appointments among the unilateral powers
enjoyedby the president.We demonstratethatthe president'suse of recess appointmentsarisesfromambiguities in the Constitution,thatrecess appointmentsallow
the presidentto move first and force Congressand the
courts to react, and that the president uses recess
appointmentsto affect policy. To demonstratethis, we
compiled a data set of every civilian nominationand
recess appointmentmade between 1987 and 2004. We
argue that presidents,as seekers of policy, are more
likely to makerecess appointmentsto positionsthatare
most likely to impact federalpolicy. After controlling
for otherfactors,we find thatpresidentsaremorelikely
to make recess appointmentsto major independent
boardsandagencies.Since suchboardsandagenciesare
we
importantpolicy makersin the federalbureaucracy,
find this to be evidence that supportsour theory.As
such, we arguethatthe power to make recess appointments shouldbe addedto the president's"toolchest"of
unilateralpowers(Howell 2003).
RecessAppointments
Article II, section 2 of the Constitution allows
presidents,when the Senate is in recess, to fill vacant
positions that normally require senatorialconfirmation.5 In other words, when a vacancy in a federal
office exists and the Senate is in recess, the president
can appointsomeone to that position without having
to go through the normal senatorial confirmation
process. Recess appointeeshave all the constitutional
powers of a confirmednominee. The primarydistinction between them and confirmed nominees is the
amount of time they can serve. A recess appointee
must leave office at the end the next Senate session or
when the vacancy is filled by a confirmednominee,
making it possible for a recess appointee to serve a
period of up to nearly two years (Hogue 2007).6
Presidentsin recent years have relied increasingly
on the recess appointmentpower.As Figure 1 shows,
since the late 1980s, presidents have made more
recess appointments.Not only has the raw numberof
recess appointmentsincreased in recent years (as of
June 2006, President Bush had already made more
recess appointmentsthan President Clinton), recess
appointmentsrelativeto the total numberof nominations have increased. Additionally, Figure 1 shows
thatthe percentageof time presidentshave chosen to
make recess appointmentshas more than doubled
since the late 1980s.
The nominationprocess discussedin most political
science literaturedoes not consider recess appointments. Rather, it usually takes the following form.
The Senate receives the president's formal nomination. Its presidingofficer then refers that nomination
to the relevantcommittee,where the committee chair
schedules a hearing.For the nominationto proceed, a
majorityof the committeemust then reportthatnomination to the floor. If a majority of senators on the
floor vote to confirm, the nominee waits for a signed
commission from the president.
As such,the nominationprocessis seen as providing
the Senatewith a strongcheckon the president'spower
to nameindividualsto federalvacancies.Both theoretical literature(Chang 2001; Nokken and Sala 2000;
SnyderandWeingast2000; Moraskiand Shipan 1999;
Hammond and Hill 1993; Calvert, McCubbins,and
Weingast1989) and empiricalliterature(Nixon 2004;
Chang2001; SnyderandWeingast2000; McCartyand
Razaghian1999; Krutz,Fleisher,and Bond 1998) generallyfind thatthe presidentandthe Senateshareinfluence overthe nominationprocess.
Corley(2006) providedthe only systematic,empirical analysis of recess appointmentsin the political
scienceliteratureto date.She soughtto explainthe conditions under which presidentsmade recess appointmentsto independentregulatorycommissions.7Counter
Black et al. / Adding Recess Appointmentsto the President's"ToolChest" 647
Figure 1
Trends in the Usage of Recess Appointments
60 -
Number of Recess Appointments
Percent of Recess Appointments
20
18-
5016
S14-
S40 -
120-
o
0
o.C
-210
cc
M
z
a
Z
IV
10
-3
1987 1989 1991 1993 1995 1997 1999 20012003
Year
6-
4
4-
0
1987 1989 1991 1993 1995 1997 1999 2001 2003
Year
Note: Data collected by the authors(see note 15). The data begin with 1987 and conclude with 2004. Figures include a lowess smoothing line with bandwidth0.80.
to traditionalnominationsliteraturethat emphasizes
sharedinfluenceby the presidentandthe Senate,Corley
arguedthat "thepresident'srecess appointmentpower
may tip the balancein his favorwhen it comes to control over the bureaucracy"
(p. 678).
Recess Appointments as
a Unilateral Power
The president's appointmentpower is one of the
strongestformal powers he enjoys. Moe (1987, 489)
described it cogently: "The president's personnel
decisions are strategicallyimportantto the realization
of his interests as a political leader, and the White
House jealously guards its powers and flexibility in
putting them to best use." Since bureaucracieshave
significant powers to interpretand create law, they
play a critical role in furtheringor hampering the
president's interests. Accordingly, whom the president is able to appoint, in large degree, determines
what the presidentaccomplishes.
Presidentscare deeply aboutwhat they accomplish
while in office. To fulfill theirpolicy goals, presidents
are driven to "seek control over the structuresand
processes of the government"(Moe 1985, 238). That
presidents will use whatever legitimate tools are at
their disposal to influence policy, then, is no surprise.
Recent scholarshiphas highlighted the increased
use of presidentialunilateralpowers, or powers that
essentially allow the presidentto make law without
the prior consent of Congress (Moe and Howell
1999a). Examples of unilateralpowers identified by
presidentialscholarsinclude executive orders(Mayer
1999, 2001; Howell 200-3),presidentialsigning statements (Cooper 2002, 2005), and executive agreements (Moe and Howell 1999a). These powers allow
presidentsto accomplish policy goals that would be
impossible in the traditionallegislative process. As
Moe and Howell (1999b, 858) argued, "Unilateral
action can make a big differencein determiningwhat
presidentsare able to achieve-and this is why they
value it and want more of it."
We define threekey criteriathatmust be presentto
name a presidentialaction as a unilateralpower:
The president manipulatesambiguitiesin the
Constitution.
648 Political
Research
Quarterly
*
*
The presidentmustuse the powerfirstand use it
alone,puttingthe legislatureandcourtsin a position such that they must reactto the president's
action.
Thepresident'sactionmustaffectpolicy.
In this section, we discuss these three criteria and
apply them to presidentialrecess appointments.By
analyzingall civiliannominationsmadebetween 1987
and 2004, we find evidence that presidents make
recess appointmentsin an effortto affect policy.
ConstitutionalAmbiguities
Presidentsrely on ambiguityin the Constitutionto
act unilaterally. Constitutional ambiguity provides
them with "the opportunity for the exercise of a
residuum of unenumeratedpower" where they can
"lay claim to what is not explicitly grantedto them"
(Pious 1979, 38; Moe and Howell 1999a, 137).
Presidents are in a unique position, among the
threebranchesof government,to exploit constitutional
ambiguities and expand presidential policy-making
powers. In their capacities as chief executives, presidents control the operationsof the government.This
puts the resourcesandpersonnelof the federalgovernment at their disposal, giving presidentsinformation
and expertise that cannot be matched by the other
branches of government (Moe and Howell 1999a,
1999b). Such resources can be used to identify and
exploitconstitutionalambiguities.Presidentsarealso in
a position to act first, therebyenablingthem to claim
powersandforcethe legislativeandjudicialbranchesto
reactto theirmove.8
Recess appointmentsarise from ambiguityin the
Constitution. The Constitution creates the recess
powerwithseeminglyclearlanguage:'"The
appointment
Presidentshall have Power to fill up all Vacanciesthat
mayhappenduringtheRecessof the Senate,by granting
Commissionswhich shallexpireat the Endof theirnext
When probed,however,this seeminglyclear
Session?"9
statementactuallytellsus verylittleabouthow therecess
appointmentpowershouldworkin practice.
For example, determiningwhen a vacancy "happens" is unclear.One interpretationof the clause suggests that a vacancy must originate("happen")when
the Senate is in recess (Mayton 2004; Rappaport
2005). Thus, if the Senate recesses and a vacancy
then arises, the president can fill it with a recess
appointment.A second interpretationof the clause,
and the one that presidentshave pushed successfully,
is that a vacancy must simply "happento exist" when
the Senate goes into recess (Hartnett2005). That is,
even if the position became vacant while the Senate
was in session, if it remainsvacantduringa recess, the
presidentcanmakea recessappointment.The effectsof
on presidentialpowerareeasy to
the two interpretations
first
limits
the powers of the president
discern. The
while the second strengthensthem. That presidents
have arguedthe latter,then, is no surprise.
The clause is likewise unclear as to whether the
Senate must be in "intersession"recess (a recess
occurringbetweensessions) for the presidentto make
a recess appointment,or whetherhe can do so during
"intrasession"recesses (those occurringwithin a session). And if the president can make intrasession
recess appointments, it is unclear how long that
recess must be (Herz 2005).
Some legal scholars suggest that the Constitution
envisions presidents making recess appointments
only during intersession recesses (Rappaport2005).
Early congresses met for relatively short periods of
time and had intersessionrecesses last as long as nine
months.To maintaina functioninggovernment,presidents made recess appointments during such
recesses (Carrier1994; Hogue 2007; Halstead2005).
Despite this, the executive branch has long argued
that recess appointments are permissible during
intrasession recesses. For example, the Harding
administrationasserted that intrasession breaks can
be long enough to cause concern about the proper
functioning of government (Herz 2005). By the
1940s, presidentsbegan to make intrasessionrecess
appointments regularly-Truman alone made 104
such appointments(Hogue 2004).1o
Constitutionalargumentsover intrasessionrecess
appointmentshave recently found their way into the
court system." In February2004, George W. Bush
recess appointedWilliam H. Pryorto the U.S. Court
of Appeals for the Eleventh Circuit during an
intrasessionrecess. Months later,three lawsuits were
filed in federal court by defendants who appeared
before Pryor.They arguedthathis intrasessionrecess
appointmentwas unconstitutional,that Pryor should
have been disqualified from presiding over their
cases, and that they should receive a new hearing.At
issue was the constitutionalityof intrasessionrecess
appointments (Lane 2005). The Eleventh Circuit
upheld Pryor's appointmentin these cases, stating
thatintrasessionrecess appointments,while constitutionally ambiguous, were within the legitimate
purviewof the president.
Since presidents are primarily seekers of policy,
they will push theirinstitutionalpowers as far as they
Blacket al. / AddingRecessAppointments
"ToolChest" 649
to thePresident's
can go, using the ambiguityinherentin constitutional
language to lay claim to strongerpowers. The recess
appointmentpower is no exception. The language in
the constitutionis ambiguous, which opens the door
for unilateralpresidentialaction.
First Mover
As chief executives, presidentsmay use theirpowers to act first on an issue. "If they want to shift the
status quo by taking unilateral action on their own
authority, whether or not that authority is clearly
established in the law, they can simply do itquickly,forcefully, and (if they like) with no advance
notice" (Moe and Howell 1999b, 855). Thus, without
consulting the other branches of government,presidents have the power to change policy.
Once the president changes policy, Congress and
the courts may react. If they do nothing, the president's policy stands. But reacting can be difficult.
When members of Congress choose to fight a president's unilateralaction, they must convince a majority of their colleagues to fight with them, and they
must overcome a cumbersome legislative process
with multiple veto points (Mayer 2001). The courts
must wait for the issue to be broughtto them before
they can react. The likelihood that either branchwill
succeed in alteringpolicy set by the presidentis low.
Recess appointments afford the president the
opportunityto act first. The presidentneed only wait
for the Senate to recess, when he can appoint a candidate of his choosing to a vacant position. It is then
up to the legislative andjudicial branchesto react to
this appointment.While fear of a negative reaction
has likely tempered presidential use of the recess
appointmentpower (Fisher 1997), the power clearly
gives the presidentan advantagein the struggle over
control of the federal apparatus.
This is not to say that the presidentmay do whatever he wants. If the presidentacts in a reckless manner, it will certainly have political ramificationsand
will inspireCongressand the courtsto react.As such,
the presidentwill use his advantageas first mover to
changepolicy, but in a strategicandmoderatemanner
(Moe and Howell 1999b).
There are several examples where Congress
reactedto recess appointments.Congress has passed
statutesforcing the presidentto submit to the Senate
for confirmationthe names of every recess appointment he makes.12 Another statutorysection declares
that if the Senate rejects the recess appointee'snomination, the appointee cannot be paid. Additionally,
thereis a prohibitionagainstpayinga recess appointee
thathas been recess appointedto the samepositionpreviously (Hogue 2007).13 There is some evidence that
these statutorypay restrictionshave limitedpresidents.
PresidentBush, for example, was recently prevented
from reappointingJohn Bolton to the United Nations
becauseBolton could not have been paid.
Individualsenatorshave also reacted to the president's use of recess appointments. Senator James
Inhofe (R-OK)brieflyplaceda hold on all presidential
nominationsto protest a Clinton recess appointment
made during a ten-day intrasession recess (Hogue
2004). SenatorEdwardKennedy(D-MA) filed an amicus curiaebrief supportingthe argumentthatWilliam
Pryor'sintrasessionrecess appointmentwas unconstitutional(Lane 2005). CurrentSenateMajorityLeader
HarryReid (D-NV) announcedthat he would eliminate all recessesexceptfor the traditionalAugustbreak
duringthe 110thCongress(2007-2008).Thereis speculation that this move was designed to make it much
harderfor Bush to use his recess appointmentpower
(Stanton2006). Despite these variousactions,the president'spowerto makerecess appointmentsstill clearly
allows him to act first and force the otherbranchesof
governmentto react.
Effect on Policy
Presidentshave a short period of time in office to
createa legacy for theiradministrations.
To build a list
of accomplishments,presidents seek power-power
they can use to affectpolicy (Neustadt1960).As H. R.
Haldeman once stated, "You've got the power now,
don't listen to anyoneelse. Yourpoweris going to start
erodingfromJanuary20th on"(quotedin Allison 1996,
10). Presidentshave frequentlyexploitedconstitutional
ambiguities, moved first, and enacted major policy
changes.Throughouthistory,presidentshave used unilateralpower to produce such policy changes as the
acquisitionof the LouisianaPurchase;the freeing of
slaves through the EmancipationProclamation;the
desegregationof the military;andthe creationof agencies such as the EPA, the FDA, and the Peace Corps
(Moe and Howell 1999a).
The recess appointmentpower makes it possible
for the president to influence policy. By making
recess appointments,presidents can appoint candidates who sharetheirideological views, possibly candidates whose ideology would make it difficult for
them to be confirmedby the Senate.
Presidents can affect broad policy via recess
appointments to independent boards. Independent
650
Political ResearchQuarterly
boardsrule on the interpretationof laws, giving them
independentrule-makingpower over specific areasof
policy. Most boards operate according to simple
majority rule, and thus, the median board member
dictates boardpolicy. These two featurescombine to
give the presidenta strong opportunityto affect policy. The fact that these boards create policy via
majority-rulevoting means that the appointmentof
just one member of a board may shift the median of
the board and thus change the policy created by the
board.14If the presidentrecess-appointsmembersof a
board, he is able to select members with ideologies
that are most likely to move policy in a desireddirection without needing to account for the ideological
preferencesof the Senate.
For example, duringthe GeorgeW. Bush administrationthe National Labor Relations Board (NLRB)
issued severalmajorrulings.In Novemberof 2004, by
a vote of 3-2, the NLRB barredtemporaryworkers
from bargainingalongside permanentworkers.That
same month, by anothervote of 3-2, the boardoverruled existing precedentand found that employeesratherthan employers--carriedthe burdenof proving
that employers spreadthreatsof reprisalfor unionizing. In June 2004, again by a 3-2 vote, the NLRB
found that nonunionizedworkershad no right to be
accompaniedby colleagues in disciplinaryhearings.
Collectively,these votes representeda majorprobusiness shift in laborpolicy.The swing vote in all of these
cases came from Bush recess appointees. Recess
appointees,then, can exercise stronginfluenceon policy on behalf of theirappointingpresidents.
While anecdotalevidence suggests recess appointmentsprovidethe presidentwith influenceoverpolicy,
rigorous empirical evidence is harder to come by.
Corley's (2006) examinationof 835 recess appointments to independentboardsfrom 1945 to 2000 found
that presidentsuse recess appointmentsmore often
when their supportin the Senate is low. Despite this
importantevidence,Corleywas not ableto demonstrate
thatrecess appointmentsare in fact tied to presidential
effortsto influencepolicy.Thatquestionremainsunanswered.In the following section, we presentan empirical test thatseeks to demonstratethatpresidentsmake
recess appointmentsto influencepolicy.
Testingfor Policy Effects
We arguethatindependentboards,because of their
strong rule-makingfeatures, provide presidentswith
the best chance of affecting policy through recess
appointments.As such, we expect recess appointments to be made at a muchhigherrateto majorindependent boards than to other types of vacancies. To
test this hypothesis, we analyze all civilian nominations made between 1987 and 2004.15We demonstrate that presidents are far more likely to make
recess appointmentsto fill vacancies on major independentboardsthan they are other vacancies.
To test our theory,we presenta multivariateanalysis. We code our dependentvariableas 1 for a recess
appointmentand 0 for all "traditional"nominations.
Ourdatacontains 353 recess appointmentsand 8,612
nominations. We examine whether recess appointments are made to vacancies on independentboards
at higher rates than are made to other vacancies. We
also control for other variables that may impact a
president'sdecision to use a recess appointment.
IndependentBoards
We stresstwo pointsunderscoringthepolicy-making
importanceof independentboards.First, independent
boards are bodies whose policy outputs are conditional upon their votes (Snyder and Weingast 2000).
Thus, the president's ability to sway a board to his
ideological position is contingentupon his ability to
shift the median. The relatively small size of independentboardsallows him ample opportunitiesto do
so. Second, as our examplesregardingthe NLRB and
FEC demonstrate, boards have independent rulemakingpower over their given policy areas.As such,
we expect that presidents are more likely to make
recess appointmentsto major independentagencies
than to other vacancies.
Consistent with Nixon's IndependentRegulatory
CommissionerDatabase(2005),we codemajorindependent agencyas 1 if the agencyto whichthe personwas
nominatedorrecessappointedis one of thefollowing:the
ConsumerProductSafetyCommission,EqualEmployment OpportunityCommission, Federal Communications Commission, Federal Election Commission,
Federal Energy Regulatory Commission, Federal
ReserveBoardof Governors,FederalTradeCommission,
Interstate Commerce Commission, National Labor
RelationsBoard,NationalTransportation
SafetyBoard,
Nuclear RegulatoryCommission, and Securities and
ExchangeCommission.
Controls
We also control for other factors that may influence a president's decision to use recess appointments. The Senate can be a significant constrainton
Black et al. / Adding Recess Appointmentsto the President's"ToolChest" 651
presidentialnominations,and thus the presidentmay
considerthe ideology of that chamberwhen deciding
whetherto make a recess appointment.Nominees are
often blockedin committee,by holds duringthe scheduling process, and occasionally via filibusterson the
floor.'6 Since conflicts over nominations are most
likely to occur when the Senate is ideologically
opposedto the president,we expectpresidentsto make
recess appointmentsmore often as the ideological distancebetweenthemandthe Senateincreases.We measure this ideological difference using the filibuster
pivot. If the presidenthas the supportof the filibuster
pivot, he can be sure that his nominees will not
encountera filibusteron the Senatefloor.As this pivot
becomes ideologicallymoredistantfromthe president,
it will become more difficultfor his nomineesto navigate the confirmationprocess successfully,and recess
appointmentsmay became a more attractiveoption.
This filibusterpivot is coded as the absolutedifference
(using DW-Nominatescores) between the president
and filibusterpivot farthestfrom the president'sideology (Poole and Rosenthal1997).'7
Dilatory tactics in the Senate have been common
in recent decades (Binder and Smith 1997). These
tactics are frequentlyused to delay the confirmation
of presidentialnominees. The presidentmay choose
to make recess appointmentsmore often when he
expects the Senate to delay the confirmationof his
nominees. We controlfor confirmationdelay by measuring the yearly averageof days between a nomination's submissionand a terminalaction by the Senate
for the year in which the nominationis made.
Timing and political context also can be important
factors in a president's decision to make a recess
appointment.During an election year, opponents of
the presidentare more likely to utilize dilatorytactics
to delay nominees until the contest is decided.
Additionally,the Senate is in recess more often during election years, allowing the president to make
greater use of recess appointments. As such, we
anticipatethat in the run-upto an election presidents
will utilize recess appointmentsmore often.
Along similarlines, presidentswho know they will
be replacedby a memberof the oppositepartymay try
to gain last-minutepolicy influence."8We anticipate
thatwhen partisancontrolof the presidencyshifts, the
outgoing presidentwill be more likely to make recess
appointments.This is coded as 1 during the period
betweenanelectionandthe inauguration
if a presidentis
soon to be replacedby a memberof the oppositeparty.
Thishappenstwo timesin ourdata-the transitionfrom
George H. W. Bush to Clinton and from Clinton to
Table1
Probit Regressionof RecessAppointments
Variable
Majorindependentboard
Filibusterpivot
Average yearly delay
Preelectionrun up
Postelection (presidential
partychange)
Postelection (no presidential
partychange)
Post-midtermelection
Monthly presidentialapproval
Intercept
N= 8,965
Log-likelihood= -1,192.811
X2(8,=
Coefficient
StandardError
0.719*
2.396*
0.004*
0.615*
2.324*
0.101
0.410
0.001
0.075
0.113
1.243*
0.110
0.161
-0.007*
-4.127*
0.132
0.003
0.327
589.931
Pseudo-R2 = .198
Note: We reportasymptoticstandarderrors.The statisticalsignificance of all variablesis identicalif we use robuststandarderrors.
*Significantat the .05 level or better(two-tailedtest).
GeorgeW. Bush.We controlfor post-presidentialelection effects when the party of the presidentdoes not
change. We also control for post-midterm election
effects.'19
Finally, Corley (2006, 675) noted that "a recess
appointmentis not cost free." Presidents are more
likely to utilize recess appointmentswhen they know
public backlash will be low. We expect that more
popular presidents will use the recess appointment
power more frequently.We code presidentialpopularity using Gallup Poll average monthly approval
ratings for the month in which the appointmentor
nominationwas made.
Results
Since our dependentvariable is dichotomous, we
estimate our model using probitregression.20Table 1
provides the results of our model.
As expected, presidents were overwhelmingly
more likely to make recess appointmentsto major
independent agencies, key actors in the president's
struggle to control and influence federal policy.
Presidents, as policy-driven actors, seek to shift the
policies of independent agencies toward their own
ideal points. Because most independent agencies
operateaccordingto simple majorityrule, makingthe
median voter important, presidents need to fill as
many vacancies as possible. Recess appointments
allow them to fill vacancies quickly and with the
652
Political ResearchQuarterly
Figure2
PredictedProbabilitiesof RecessAppointmentUsage
0.9II
0.9 -
I
MajorIndependentBoard
0 Other Vacancy
0.8 -
-95% ConfidenceInterval
0.7 -
0 o.6< 0.5
-
0.4 -
S0.30.2 -
0.1 -
0.0
Non-ElectionYear
ElectionYear
Post Midterm
No PresidentialChange
PresidentialChange
Note: Predictedprobabilitieswere estimatedholding all other variablesat their mean. Confidence intervalsfor the predictedprobabilities were estimatedusing the SPost series of commandsimplementedin Stata9 by Long and Freese (2006). The differencebetween recess
appointmentsto majorindependentagencies and other vacancies is significantat the 95 percentlevel for all five panels of the figure.
"right"ideologicalpeople,which,in turn,makesit
morelikelythattheycan influenceagencypolicy.
Moreover,giventhatpresidentsareunableto fire
membersof these agencies,they mustfill the positions withpeoplewho areless likely to "go native."
By proppingup theirown appointees-withouthaving to compromisewith the Senate-presidentscan
morelikely attainalignmentbetweentheirinterests
andthe interestsof majorindependent
agencies.
Filibusterpivotandconfirmation
delayare significantandperformas expected.Interestingly,
theeffectof
is
but
in
the
presidential
popularity significant
opposite
of the expecteddirection.
Thisindicatesthatwhenthe
is low, he is morelikelyto fall
president's
popularity
backon his unilateral
powersin his attemptsto influencepolicy.21
Because interpretingcoefficients in a probit
regressionis difficultanddoes not show substantive
significance,we calculatedpredictedprobabilities,
which are displayedin Figure2. The presidentis
to
alwaysmorelikely to makerecess appointments
boardsthanhe is to othervacanmajorindependent
cies. Timingalso affects the president'schoice to
make recess appointments.The presidentis most
when he knows
likely to makerecess appointments
thatpartisancontrolof theWhiteHousewill change.
Discussion
In the ongoingstruggleto controlpolicy,presidents
have a powerfultool: the recessappointment
power.
allowspresidents
to install
Makingrecessappointments
favoredcandidates
in keypolicy-making
positions,
positionsto whichtheymightnotbe confirmed
in thetraditionalsenatorialadviseandconsentprocess.Welaid
out threeconditionsthatdefinedpresidentialunilateralpowersandthenexplained
howrecessappointments
fit withinthatparadigm.
We theorizedthatpresidents,
as policyseekers,wouldbe morelikelyto makerecess
to majorindependent
appointments
agencies--entities
withthepowerto influencefederalpolicy.Ourfindings
were
supportour theory.Presidentsoverwhelmingly
morelikelyto makerecessappointments
to majorindependentagencieseven when controllingfor other
therecessappointment
factors.
explanatory
Accordingly,
an
is
tool
can use to gain
power
important presidents
over
the
federal
leverage
bureaucracy.
Black et al. / Adding Recess Appointmentsto the President's"ToolChest" 653
The storydoes not end here,though.Ourresults
suggest that presidentsmake recess appointments
will impact
becausetheybelievethoseappointments
publicpolicy.Whetherrecessappointeesactuallydo
impactpolicyas thepresidentwishesis anotherstory,
one thatremainsopenfor empiricalanalysis.
thepresident
is nottheonlyactorthatcares
Moreover,
aboutfederalpolicy.Congresshasa majorstakein the
policyandappointment
processesas well.Ouranalysis
in thisarticlesuggestedthatCongresscan reactto the
decisionandthatits reaction(oranticipated
president's
thechoicespresidents
canconstrain
make.One
reaction)
futureavenueforresearch
wouldexpandon thistheory
and look more closely at how Congress'sreactions
impacttheprocess.As we statedabove,SenateMajority
LeaderHarryReid announcedthatduringthe 110th
he wouldeliminateallrecessesexceptforthe
Congress,
traditional
thatthis
Augustbreak.Thereis speculation
movewas designedto diminishthe president's
recess
If so, therewill be empirical
appointment
capabilities.
theconsequences
of hisactions.
questions
regarding
Notes
1. Indeed, the president's signing statement over the bill
announcedBush's belief that several provisions of the act were
unconstitutional.
2. The sponsorsof the BipartisanCampaignFinanceReform
Act (BCFRA) were Senators John McCain (R-AZ) and Russell
Feingold (D-WI) and RepresentativesChristopherShays (R-CT)
and MartinMeehan (D-MA).
3. The average number of recess appointments made
between 1987 and 2004 has increased.See Figure 1.
4. It is importantto note thatwhile unilateralpowers give the
presidentthe opportunityto act without the consent of the other
branchesof government,the presidentmust consider the potential reactionof the legislative andjudicial brancheswhen choosing to use unilateralpowers.This considerationconstrainshis use
of unilateralpowers.
5. Article II, section 2 states, "The President shall have
Power to fill up all Vacanciesthat may happenduringthe Recess
of the Senate, by grantingCommissionswhich shall expire at the
End of their next Session."
6. The president is statutorily required to submit to the
Senate for confirmationthe names of all his recess appointees.If
the Senate votes down theirnominations,these recess appointees
cannot be paid from the U.S. Treasury-an action that likely
causes them to resign (Fisher 1997).
7. As we discuss more fully below, legal scholars have
focused on recess appointmentsfor some time, but theirapproach
is normativeratherthan empirical.
8. See Moe and Howell (1999b, 854-56) for a more extensive
discussion of advantages the president enjoys over the other
branchesin the struggle for policy-makingpower.
9. See U.S. Constitution,Article II, section 2, clause 3.
10. These are not the only ambiguitiesin the recess appointment clause. There are others as well. For example, it is not clear
when a vacancy arises. Some positions to independentagencies
have holdover provisions, which allows the officeholder to
remainuntil a successor takes over.Whethera holdoveramounts
to a vacancy is still up for debate.
11. See Evans v. Stephens,387 F.3d 1220 (11th Cir.2004) cert
denied, 533 U.S. 942 (2005). In a rare statementfrom a denial of
cert, Justice Stevens added even more ambiguity to the recess
appointmentprocess when he stated, "It would be a mistake to
assume that our disposition of this petition constitutesa decision
on the merits of whether the President has the constitutional
authorityto fill futureArticle III vacancies, such as vacancies on
this Court,with appointmentsmade absentconsent of the Senate
during short intrasession'recesses' " (533 US at 943).
12. See 5 U.S.C. ? 5503(b).
13. The prohibitionto pay successive recess appointmentsis
based on an Office of Legal Counsel interpretationof 5 U.S.C.
? 5503 as cited in Hogue (2007).
14. It is importantto note that not all vacancies on independent boardswill createan opportunityfor the presidentto change
policy. It will depend on the president'spreferencesand the currentideological makeupof the board(SnyderandWeingast2000;
Moraskiand Shipan 1999).
15. Our nominations data come from the Presidential
Nominationspage of the Libraryof CongressTHOMASWeb site
We downloadedall civil(http://thomas.loc.gov/home/nomis.html).
ian nominationsfor each Congressusing no other searchrestrictions. All data were downloadedfrom THOMAS on January30,
2006, and then cleaned manually.Incompleterecords(e.g., those
lacking nominationdates or position information)were purged
from our data.Nominationssubmittedin groupshad two records:
one for the group nominationand then one for each individual.
These duplicate records were deleted. We used two sources to
obtainourdatafor recess appointments.ForPresidentsReaganand
George H. W. Bush, we examinedthe appendicesof various volumes of thePublicPapersof thePresident,which we retrievedfrom
HEINonline(http://heinonline.org).
Datafor PresidentsClintonand
GeorgeW. Bush were considerablyeasierto obtain,as reportspublishedby the CongressionalResearchServiceprovidedthese datain
tabularform (Garcia2001; Hogue and Bearden2006).
16. Filibustering in the Senate derives its power from the
chamber's lack of a formal, majoritarianmethod for ending
debate.As a consequence,the business of schedulingin the body
is generally a consensual process where the majorityand minority party leaders forge unanimous consent agreements.A hold
occurs when a senatornotifies her or his partyleader that she or
he plans to object to a unanimousconsent agreement.While it is
ultimately still up to the majority leader to determine whether
to honor a hold, objections to unanimous consent agreements
greatly inhibit chamberbusiness.
17. We also controlled for Senate ideology using variables
accountingfor the seat shareof the president'spartyandfor divided
government(partycontrolby the partyopposed to the president).
We estimatedBayesianinformationcriterion(BIC) for threemodels, each includingone of the threecontrols,andfoundthatthe BIC
is the lowest for the model includingfilibusterpivot.
18. This effect has been noted by work on other unilateral
powers, such as executive orders (Mayer 2001; Howell and
Mayer 2005).
19. All of these time effects are comparedto the baseline of
"nonelectionyear."
20. We also estimatedthe model with robust standarderrors.
The statistical significance of our variablesis identical with this
alternativeestimation.
654 Political ResearchQuarterly
21. Despite the statisticalsignificanceof this variable,its substantiveeffect is small. Moving from most popularto least popular presidents in our data, the probability of making a recess
appointmentonly increases from .005 to .014, holding all other
variablesat their mean or modal value.
References
Allison, GarlandW. 1996. Delay in Senate confirmationof federaljudicial nominees. Judicature80:8-15.
Binder, SarahA., and Steven S. Smith. 1997. Politics or principle: Filibustering in the United States Senate. Washington,
DC: Brookings InstitutionPress.
Calvert, Randall L., Mathew D. McCubbins, and Barry R.
Weingast. 1989. A theory of political control and agency discretion.AmericanJournal of Political Science 33:588-611.
Carrier,Michael A. 1994. When is the Senate in recess for purposes of the recess appointments clause? Michigan Law
Review 92:2204-47.
Chang, Kelly H. 2001. The president versus the Senate:
Appointmentsin the American system of separatedpowers
and the Federal Reserve. Journal of Law, Economics and
Organization17:319-55.
Cooper,PhillipJ. 2002. By orderof thepresident:Theuse and abuse
of executivedirectaction.Lawrence:UniversityPressof Kansas.
2005. GeorgeW. Bush, EdgarAllan Poe, and the use and
--.
abuse of presidentialsigning statements.Presidential Studies
Quarterly35:515-32.
Corley, Pamela. 2006. Avoiding advice and consent: Recess
appointments and presidential power. Presidential Studies
Quarterly36:670-80.
Fisher,Louis. 1997. ConstitutionalconflictsbetweenCongressand
thepresident.4th ed. Lawrence:UniversityPressof Kansas.
Garcia, Rogelio. 2001. Recess appointmentsmade by President
Clinton.CongressionalResearch Service ReportRL30821.
Halstead, T. J. 2005. Recess appointments:A legal overview.
CongressionalResearch Service ReportRL33009.
Hammond,ThomasH., andJeffreyS. Hill. 1993.Deferenceorpreferof presidential
nomineesto
ence?ExplainingSenateconfirmation
Politics5:23-59.
administrative
agencies.Journalof Theoretical
Hartnett,Edward A. 2005. Recess appointmentsof Article III
judges: Three constitutionalquestions. CardozoLaw Review
26:377-442.
Herz, Michael. 2005. Abandoning recess appointments? A
Comment on Hartnett (and others). Cardozo Law Review
26:443-61.
Hogue, HenryB. 2004. Recess appointmentsto Article III courts.
PresidentialStudies Quarterly34:656-73.
. 2007. Recess appointments:Frequentlyasked questions.
•CongressionalResearch Service Report.RS21308.
Hogue, Henry B., and MaureenBearden. 2006. Recess appointments made by President George W. Bush, January 20,
2001-February 27, 2006. Congressional Research Service
ReportRL33310.
Howell, William G. 2003. Power withoutpersuasion. Princeton,
NJ: PrincetonUniversity Press.
2005. Unilateralpowers:A brief overview.Presidential
-.
Studies Quarterly35:417-39.
Howell, William G., and KennethR. Mayer. 2005. The last one
hundreddays. Presidential Studies Quarterly35:533-53.
Krutz,GlenL., RichardFleisher,andJonR. Bond. 1998. FromAbe
Fortasto Zoe Baird:Why some presidentialnominationsfail in
the Senate.AmericanPoliticalScienceReview92:871-81.
Lane, Charles.2005. Courtdeclines to wade into battleon judges.
WashingtonPost, March22, A15.
Long, J. Scott, and Jeremy Freese. 2006. Regression modelsfor
categorical dependentvariables using Stata. 2nd ed. College
Station,TX: StataPress.
Mayer, Kenneth R. 1999. Executive orders and presidential
power.Journalof Politics 61:455-66.
1. 2001. Withthe stroke of a pen: Executive orders and
presidentialpower. Princeton,NJ:PrincetonUniversityPress.
Mayton,William Ty. 2004. Recess appointmentsand an independentjudiciary.ConstitutionalCommentary20:515-55.
McCain, John, Russell D. Feingold, Christopher Shays, and
MartyMeehan. 2005. McCain, Feingold, Shays, and Meehan
discourage president from FEC recess appointment.http://
www.house.gov/shays/news/2005/november/novfec.htm
(accessed February21, 2007).
McCarty,Nolan, andRose Razaghian.1999. Advice and consent:
Senateresponsesto executive branchnominations1885-1996.
AmericanJournal of Political Science 43:1122-43.
Moe, Terry M. 1985. The politicized presidency. In The new
directionin Americanpolitics, ed. John E. Chubband Paul E.
Peterson235-72. Washington,DC: Brookings Institution.
. 1987. An assessment of the positive theory of congressional dominance.LegislativeStudies Quarterly12:475-520.
Moe, TerryM., and William G. Howell. 1999a. The presidential
power of unilateralaction. Journal of Law, Economics, and
Organization15:132-79.
1999b. Unilateral action and presidential power: A
-~.
theory.PresidentialStudies Quarterly29:850-72.
Moraski,Byron J., and CharlesR. Shipan. 1999. The politics of
SupremeCourtnominations:A theoryof institutionalconstraints
andchoices.AmericanJournalof PoliticalScience43:1069-95.
Neustadt, RichardE. 1960. Presidentialpower and the modem
presidents.New York:Free Press.
Nixon,DavidC. 2004. Separationof powersandappointeeideology.
Journalof Law,Economics,and Organization20:438-57.
. 2005. The Independent Regulatory Commissioner
Database.http://www2.gsu.edu/-wwwirc/.
Nokken, Timothy P., and Brian R. Sala. 2000. Confirmation
dynamics:A model of presidentialappointmentsto independent agencies. Journal of TheoreticalPolitics 12:91-112.
No recess. 2005. WashingtonPost, November 14, p. A20.
Pious, Richard M. 1979. The American presidency. New York:
Basic Books.
Poole, Keith T., and Howard Rosenthal. 1997. Congress: A
political-economic history of roll call voting. Oxford, UK:
OxfordUniversityPress.
Rappaport,Michael B. 2005. The original meaning of the recess
appointmentsclause. UCLALaw Review 52:1487-1578.
Schor,Elana.2006. FEC's coordinationrulingcould shakeup the
midterms.The Hill, March30, p. 13.
Snyder,SusanK., andBarryWeingast.2000. The Americansystem
of shared powers: The president,Congress, and the NLRB.
JournalofLaw, Economics,and Organization16:269-305.
Stanton,John. 2006. Senate plan squeezes recess slots. Roll Call,
December 12.