F I L E D
United States Court of Appeals
Tenth Circuit
JUN 13 2001
PUBLISH
UNITED STATES COURT OF APPEALS PATRICK FISHER
Clerk
TENTH CIRCUIT
DONNIS F. KENT,
Plaintiff - Appellant,
v.
BRUCE MARTIN, in his individual
No. 00-6144
capacity and in his official capacity as
County Clerk of Alfalfa County,
Oklahoma; ALFALFA COUNTY
BOARD OF COUNTY
COMMISSIONERS,
Defendants - Appellees.
Appeal from the United States District Court
for the Western District of Oklahoma
(D.C. No. CIV-99-245-M)
Phyllis L. Walta, Walta & Walta, Hennessey, Oklahoma, for Plaintiff-Appellant.
James L. Gibbs, II (Chris J. Collins and Michael L. Carr on the brief), Collins,
Zorn, Jones & Wagner, P.C., Oklahoma City, Oklahoma, for Defendants-
Appellees.
Before BRISCOE, BALDOCK, and LUCERO, Circuit Judges.
LUCERO, Circuit Judge.
Donnis Kent, a former employee of the Alfalfa County Clerk’s office,
brought this action under 42 U.S.C. § 1983 alleging retaliatory discharge in
violation of the First Amendment. Kent’s employment as a deputy clerk was
terminated six months after her unsuccessful campaign to unseat Bruce Martin as
County Clerk and the publication in a local newspaper of her statements
concerning Martin’s job performance. She appeals the district court’s grant of
summary judgment for defendants Martin and the Board of County
Commissioners. Exercising jurisdiction under 28 U.S.C. § 1291, we reverse.
I
Kent was employed as Second Deputy to the County Clerk in Alfalfa
County, Oklahoma, from 1988 to 1997. In 1993, Bruce Martin replaced Kaye Jay
as County Clerk, and Kent worked under Martin from his election until her
termination in February 1997. During that time, the Clerk’s office contained only
one additional employee, Dorothy Steffey.
When Martin filed for reelection at the end of his term in 1996, Kent ran
against him. Six days after Kent announced her candidacy, Martin began to
document incidents of alleged misconduct on her part; those incidents numbered
seventy-nine after seven months. 1
1
Kent’s alleged performance deficiencies between July 1996 and February
1997 included gossiping, failure to relay phone messages, failure to follow
(continued...)
-2-
Shortly before the August 1996 primary, a local newspaper reported
statements Kent made describing what she believed were Martin’s abuses of his
office, including his absence during approximately five-and-a-half months in a
single year and “double-dipping” when claiming travel reimbursements. Kent
lost in the three-way primary, and six months later, on February 28, 1997, Martin
terminated Kent’s employment. In a letter to the Oklahoma Unemployment
Security Commission, Martin listed three reasons for the termination. Although
two of those reasons concerned incidents occurring on February 20 and 27, 1997,
Martin stated in the letter that he “made up [his] mind on February 14” to release
Kent. (App. at 291.) Martin, however, testified that he decided to fire Kent in
June 1996, before Kent filed to oppose him for County Clerk, but was advised by
the district attorney to wait until “after the campaign was over.” (Id. at 167.)
The parties dispute both the initial tenor of their relationship and the extent
to which it changed as a result of the campaign. It appears from the record that
Martin may have been displeased with Kent’s performance as early as 1993, 2
1
(...continued)
proper bidding procedures, performing outside business in the Clerk’s office, and
general failure to communicate cordially within the office. Kent disputes the
validity of many of the documented incidents.
2
Between January 1993 and January 1996, Martin documented three minor
instances of misconduct involving Kent. On two of those occasions, Kent had
arrived at work three to four minutes late. The third instance concerned
procedure for obtaining bids on fuel and was documented neither in a warning to
(continued...)
-3-
although he claims new and more serious problems arose after the August
primary. On the other hand, Kent alleges that her bid to oust Martin changed
Martin’s attitude toward her but did not affect Kent’s attitude or performance or
the overall efficiency of the workplace. Kent also argues that the reasons Martin
gave for her termination were unrelated to any disruption that her candidacy
might have caused.
In her suit, Kent alleges that defendants violated the First Amendment by
terminating her employment because she opposed Martin for the position of Clerk
and spoke publicly about his absences and other alleged abuses. She appeals the
district court’s grant of summary judgment in defendants’ favor. 3
II
We review a grant of summary judgment de novo, applying the same legal
standard used by the district court. Barker v. City of Del City, 215 F.3d 1134,
1137 (10th Cir. 2000). Summary judgment is appropriate only if the evidence
(...continued)
Kent nor in her personnel file. In his deposition, Martin discussed other instances
of improper conduct by Kent prior to 1996, including “gossiping” and failing to
relay phone messages, but he admitted those had not been documented. (App. at
144.) Martin also testified that Kent was “sneaky,” “rude,” and “talked behind
[his] back” the entire time they worked together (id. at 171), but that he did not
fire her because he “didn’t want to put her out of a job” (id. at 146).
3
Kent does not appeal the grant of defendants’ summary judgment motion
on her additional federal claim of denial of due process and her pendent state
claim for breach of her employment contract.
-4-
shows “there is no genuine issue as to any material fact and that the moving party
is entitled to a judgment as a matter of law.” Fed. R. Civ. P. 56(c).
“In First Amendment cases, an appellate court has an obligation to make an
independent examination of the whole record in order to make sure that the
judgment does not constitute a forbidden intrusion on the field of free
expression.” Barker, 215 F.3d at 1137 (quotation omitted). In a civil rights
action challenging an adverse employment decision allegedly made in reaction to
an employee’s speech, whether the employee’s interest in making the statement
outweighs the state’s interests as employer is treated as a question of law
requiring de novo review. See Cragg v. City of Osawatomie, 143 F.3d 1343,
1346 (10th Cir. 1998).
A
When a government employer has allegedly taken adverse action because
of an employee’s exercise of her right of free speech, we apply the balancing test
derived from Pickering v. Board of Education, 391 U.S. 563 (1968), and Connick
v. Myers, 461 U.S. 138 (1983), (the “Pickering/Connick test”). Barker, 215 F.3d
at 1138; see also Jantzen v. Hawkins, 188 F.3d 1247, 1251 (10th Cir. 1999);
Horstkoetter v. Dep’t of Pub. Safety, 159 F.3d 1265, 1271 (10th Cir. 1998). That
four-part test asks the following questions:
1. Whether the speech in question involves a matter of public
concern.
-5-
2. If so, we must weigh the employee’s interest in the expression
against the government employer’s interest in regulating the speech
of its employees so that it can carry on an efficient and effective
workplace.
3. Employee must show the speech was a substantial factor driving
the challenged governmental action.
4. If so, can the employer show that it would have taken the same
employment action against the employee even in the absence of the
protected speech[?]
Barker, 215 F.3d at 1138–39 (quoting Jantzen, 188 F.3d at 1257 (further citation
omitted)). “The first two questions are ones of law for the court, while the latter
two questions are ones of fact for the jury.” Id. at 1139; see also Horstkoetter,
159 F.3d at 1271.
B
As the district court noted, and defendants conceded at oral argument, an
employee’s candidacy for political office “undoubtedly relates to matters of
public concern,” and the first part of the test is therefore satisfied in this case.
Jantzen, 188 F.3d at 1257.
At issue is the district court’s resolution of the second prong of the
Pickering/Connick test in defendants’ favor. In granting summary judgment for
defendants, the court concluded that their interest in maintaining an effective
workplace outweighed Kent’s interest in her political expression. That
conclusion was based on what the court believed were defendants’ reasonable
-6-
predictions of workplace disruption. The court relied on Jantzen, 188 F.3d at
1257, for the proposition that courts “‘will defer to a public employer’s
reasonable predictions of disruption, but those predications [sic] must be
supported by the presentation of specific evidence’” (quoting Cragg, 143 F.3d at
1347). According to Kent, the district court applied the wrong legal standard
when it accepted evidence of a “prediction of disruption” as evidence of
defendants’ interest in maintaining an efficient workplace by regulating Kent’s
speech. Kent argues that evidence of “actual disruption” is required to justify an
employee’s termination several months after the protected speech occurred.
(Appellant’s Br. at 13.) We agree.
In Jantzen we held, in the context of a deputy sheriff’s termination after
announcing his candidacy for sheriff, that “at the time of [plaintiff] Haugland’s
termination (in contrast to [other deputies’] termination six months thereafter),
there was specific evidence to support [his employer’s] reasonable prediction” of
disruption. 188 F.3d at 1257–58. In that case, plaintiff Haugland was fired
immediately, so that predictions of disruption were the only possible evidence of
the employer’s interest in regulating the expression at the time of the firing. That
is why we drew a distinction between Haugland, who had been fired immediately,
and the fellow deputies who supported his candidacy and were fired six months
-7-
later. 4 Six months after the employees’ expression, a so-called “prediction” of
disruption would be meaningless to justify their termination, and under our case
law evidence of actual disruption would be required to outweigh the employees’
interest in their speech.
In some cases we have deferred to a public employer’s predictions of
disruption—rather than requiring evidence of actual disruption in all cases—only
because we recognize that a public employer does not “have to wait for speech
actually to disrupt core operations before taking action.” Moore v. City of
Wynnewood, 57 F.3d 924, 934 (10th Cir. 1995). In this case, because Martin’s
alleged fear of disruption did not lead him to fire Kent immediately, defendants
must show actual disruption in order to articulate an interest in regulating Kent’s
speech six months after the fact. 5
The factual circumstances we confront in this case are analogous to those
in Prager v. LaFaver, 180 F.3d 1185 (10th Cir. 1999). In Prager, a whistleblower
4
We did not directly address the remaining plaintiffs’ free speech claims
because they had failed to raise the Pickering/Connick issue on appeal. Jantzen,
188 F.3d at 1256 – 57.
5
Defendants’ assertion that “Martin was not required to allow the events
to unfold to the extent they did with Ms. Kent” is irrelevant to our inquiry.
(Appellees’ Br. at 14.) The suggestion that Martin deliberately waited to fire
Kent in an attempt to accumulate several documented instances of alleged
misconduct does not relieve defendants of the burden to show actual disruption
caused by Kent’s speech. That documentation goes to the question whether
Martin would have terminated Kent irrespective of her political speech, a question
that is for the trier of fact. Cragg, 143 F.3d at 1347.
-8-
was terminated from the Kansas Department of Revenue several months after he
spoke out regarding abuses on the part of his employer. Id. at 1188–89. This
Court held that a showing of “‘actual disruption of services which results from
the employee[’s] speech’” is required in that situation to justify the employee’s
termination. Id. at 1191 (quoting Ramirez v. Okla. Dep’t of Mental Health, 41
F.3d 584, 594 (10th Cir. 1994) (quoting Schalk v. Gallemore, 906 F.2d 491, 496
(10th Cir. 1990))). As we noted in Jantzen, because of the time lapse, “it was
reasonable [in Prager] to look for proof of actual disruption in order to justify the
subsequent firing.” 188 F.3d at 1258 n.8. 6
In Barker, another factually similar case, an administrative assistant to the
city manager was fired several months after she gave an interview regarding
accusations that the City Council had violated a state statute. 215 F.3d at
1136–37. We held the district court erred in basing summary judgment for the
defendant under the second prong of the Pickering/Connick test on the employer’s
speculative allegations of disruption rather than requiring evidence of actual
6
In Cragg, 143 F.3d at 1347, we stated that “[w]e will defer to a public
employer’s reasonable predictions of disruption [if] . . . supported by the
presentation of specific evidence.” However, we concluded that because “[o]ur
review of the record reveals no evidence of disruption caused by Mr. Cragg’s
political speech,” his employer had “failed to demonstrate by specific evidence”
that his speech “contributed to any disruption.” Id. Thus, as in Prager, because
the employee was fired several months after his initial exercise of speech, the fact
that the employer could not show actual disruption deprived its predictions of
disruption of any probative value.
-9-
disruption. Id. Because the defendant had failed to produce “evidence indicating
that her speech was disruptive,” we held the plaintiff should have been allowed to
proceed with her free speech claim. Id. at 1140 & n.2. In this case, as in Barker,
it was improper to rely on a prediction of disruption stemming from the
employee’s speech when Martin’s decision not to fire Kent for six months raises a
genuine issue of material fact regarding whether disruption actually occurred. 7
Because Kent was not fired until six months after her campaign for office
and public statements concerning Martin’s job performance, evidence of a
prediction of disruption is insufficient to justify summary judgment for
defendants under the second prong of the Pickering/Connick test. Under our
precedent, the district court should have required defendants to make a showing
of actual disruption in order to articulate an efficiency interest in regulating
Kent’s speech. Our cases applying the “reasonable prediction of disruption”
standard have done so in the context of a termination soon after the employee’s
7
Defendants cite Caruso v. DeLuca, 81 F.3d 666 (7th Cir. 1996), in
support of their argument that actual disruption need not be shown under the
second prong of the test. In Caruso, a deputy clerk who opposed the City Clerk in
an election was not reappointed because of her employer’s prediction that
workplace disruption would ensue. There, however, an impending redistribution
of duties within the clerk’s office made it less likely that harmony and efficiency
could be maintained in the wake of election-related antagonism between Caruso
and her boss. Id. at 668, 671. As the redistribution had not yet occurred,
predictions of disruption were the only evidence available. That is not the case
here.
-10-
exercise of speech, when the intent of the termination was to avoid actual
disruption. See Jantzen, 188 F.3d at 1257–58; Moore, 57 F.3d at 927–28, 934.
That legal standard is inapplicable when an employer has allowed an employee to
continue to work after the protected expression.
If there has been no actual disruption justifying termination during the six
months following an employee’s protected speech, it is nonsensical to rely ex post
facto on a “prediction” of disruption to tip the balance in favor of an employer’s
interest in an efficient workplace. We hold that the district court erred when it
granted summary judgment in favor of defendants based on their prediction of
workplace disruption.
C
The district court did not address the question whether defendants can show
actual disruption on summary judgment so as to prevail under the second prong of
the Pickering/Connick test. We therefore remand to the district court for
application of the correct legal standard.
III
The judgment of the district court is REVERSED, and the matter is
REMANDED for consideration of defendants’ summary judgment motion under
the correct legal standard.
-11-