United States Court of Appeals
For the First Circuit
No. 24-1443
DRAFTKINGS INC.,
Plaintiff, Appellee,
v.
MICHAEL HERMALYN,
Defendant, Appellant.
APPEAL FROM THE UNITED STATES DISTRICT COURT
FOR THE DISTRICT OF MASSACHUSETTS
[Hon. Julia E. Kobick, U.S. District Judge]
Before
Kayatta, Thompson, and Montecalvo,
Circuit Judges.
Christopher G. Michel, with whom Gregg M. Badichek, Aliki
Sofis, Alexander S. del Nido, Issac Saidel-Goley, Quinn Emanuel
Urquhart & Sullivan, LLP, Russell Beck, Stephen D. Riden, and Beck
Reed Riden LLP were on brief, for appellant.
Thomas H. Dupree Jr., with whom Jacob T. Spencer, Orin S.
Snyder, Harris M. Mufson, Justine M. Goeke, Gibson, Dunn & Crutcher
LLP, Mark C. Fleming, William F. Lee, Andrew S. Dulberg, and Wilmer
Cutler Pickering Hale and Dorr LLP were on brief, for appellee.
September 26, 2024
THOMPSON, Circuit Judge.
STAGE-SETTING
Massachusetts and California aren't exactly on the same
page when it comes to noncompete agreements.
Massachusetts generally allows noncompetes if they
comply with certain restrictions (for example, employers can't
require all employees to sign them (excluded employees include
persons aged 18 or younger) and the noncompete period can't last
more than a year (except if the employee breached a fiduciary duty
to the employer or stole the employer's property, in which case a
2-year cap applies)). See, e.g., Mass. Gen. Laws ch. 149,
§ 24L(b)(iv), (c). California generally bans noncompetes unless
they squeeze within one of the few narrow exceptions (like for
noncompetes in a business sale or a partnership breakup). See,
e.g., Cal. Bus. & Prof. Code §§ 16600(a), 16601, 16602. And that's
true "regardless of where and when the contract was signed" and
"whether . . . the employment was maintained outside of
California." See
id. § 16600.5(a), (b).
Today's case plays out against the backdrop of these
different policy choices. We streamline the facts and procedural
history (rather aggressively) to lay bare the nature of the
disputes before us, adding more details later as needed for
specific issues (an example of streamlining is our ignoring a
- 3 -
California state-court lawsuit involving some of the same parties
as here, because that suit doesn't affect our analysis).
As recently as the beginning of this year, then-New
Jersey resident Michael Hermalyn had a plum job with Massachusetts-
headquartered DraftKings. But he quit to take a similar job with
rival Fanatics's California-based subsidiary, a position (he says)
that requires him to live and work in Los Angeles. DraftKings and
Fanatics (for anyone who hasn't seen their ads) are sports betting
and online gaming companies (a description that simplifies things
a bit but is enough for now).
DraftKings thought (among other concerns) that
Hermalyn's new post violated a noncompete he had signed before
quitting — an agreement that had a Massachusetts choice-of-law
proviso and a one-year noncompete clause. So DraftKings sued him
in Massachusetts federal court for breach of the noncompete (along
with other claims not relevant here).
Everyone seems to agree (at least for present purposes)
that if the noncompete is enforceable, Hermalyn breached it by
joining Fanatics. Not surprisingly then, DraftKings asked the
district judge to use Massachusetts law and Hermalyn asked her to
use California law. Siding with DraftKings, the judge — after
using Massachusetts law — ruled the noncompete enforceable and
preliminarily enjoined Hermalyn from competing against DraftKings
- 4 -
in the United States for one year (she did reject DraftKings's
request for a worldwide injunction, however).
Hermalyn then filed this interlocutory appeal, see 28
U.S.C. § 1292(a)(1), making two alternative arguments. The first
is that the judge wrongly held that Massachusetts law governed the
enforceability of the noncompete. The second is that if
Massachusetts law does govern, then the judge should've excluded
California from the preliminary injunction's scope.
Having considered the matter on an expedited basis, we
let the challenged order stand. Read on to learn why.
CHOICE OF LAW
Standards of Review
We examine the judge's preliminary-injunction grant for
abuse of discretion. See We the People PAC v. Bellows, 40 F.4th
1, 9 (1st Cir. 2022); NuVasive, Inc. v. Day,
954 F.3d 439, 443
(1st Cir. 2020). Within that system, however, we inspect legal
questions (like which state's law applies) de novo and findings of
fact for clear error. See NuVasive,
954 F.3d at 443. On abuse-
of-discretion review, we defer to the judge if she didn't make an
obvious mistake of judgment (a material legal misstep is a per se
abuse of discretion, for example). See We the People PAC,
40 F.4th
at 25. But on de novo review, we give no deference to what the
judge thought, see Berge v. Sch. Comm. of Gloucester,
107 F.4th
33, 39 n.8 (1st Cir. 2024) — not because we're better or smarter
- 5 -
than our district-court colleagues (we aren't!) but because of the
"institutional advantages" we have over them (including sitting on
multi-member panels, a process that "permit[s] reflective dialogue
and collective judgment" and so "promotes decisional accuracy"),
see Salve Regina Coll. v. Russell, 499 U.S. 225, 232, 233 (1991).1
Massachusetts Law
Because — as the parties agree — diversity jurisdiction
exists over the breach-of-noncompete claim, the "forum" of
Massachusetts (where DraftKings sued Hermalyn) sets the rules for
which state's law decides the noncompete's enforceability (even
though the noncompete itself picked Massachusetts). See Reicher
v. Berkshire Life Ins. Co. of Am., 360 F.3d 1, 4 (1st Cir. 2004).
See also generally Smith v. Gen. Motors LLC,
988 F.3d 873, 879 n.
5 (6th Cir. 2021) (explaining that "even contracts that contain
choice-of-law clauses are themselves analyzed under some state's
law should a controversy arise over the validity of that clause
itself"). So on to Bay State law then.
1 Hermalyn seemingly implies that our review of facts
underlying a choice-of-law ruling isn't for clear error (to the
extent anyone challenges the findings, of course), but remains de
novo. He cites no on-point case that says so, however (his reliedon decisions say that we review a choice-of-law ruling de novo,
but don't say that the facts behind the ruling get something other
than clear-error treatment). And that's probably because the
Federal Reporter is filled with opinions explaining our "usual"
tradition of inspecting fact-findings for "clear error." See,
e.g., Fayard v. Ne. Vehicle Servs., LLC, 533 F.3d 42, 45 (1st Cir.
2008).
- 6 -
Massachusetts usually respects the parties' choice of
law. See NuVasive, 954 F.3d at 443; see also Oxford Glob. Res.,
LLC v. Hernandez,
106 N.E.3d 556, 564 (Mass. 2018); Feeney v. Dell
Inc.,
908 N.E.2d 753, 766 (Mass. 2009). But exceptions exist,
each driven by public-policy concerns (the term "usually" in the
last sentence was a tipoff about exceptions). See Oxford,
106
N.E.3d at 564. And the exception Hermalyn invokes requires him to
show that (i) "application of" Massachusetts law "would be
contrary to a fundamental policy of" California; (ii) California
"has a materially greater interest than" Massachusetts "in the
determination of the . . . issue"; and (iii) California is the
state whose law would control "in the absence of an effective
choice of law by the parties" — i.e., that California has "the
most significant relationship to the transaction and the parties."
See
id. at 563-64 (brackets and quotation marks omitted); see also
NuVasive,
954 F.3d at 444.
The word linking the above-listed requisites is "and,"
not "or." Which means Hermalyn must satisfy all of them to get
anywhere. See NuVasive, 954 F.3d at 444; see also Feeney, 908
N.E.2d at 766-67. But he can't satisfy the materially-greater-
interest requisite (as we explain next), thus dashing his hopes of
winning the issue (even assuming without deciding that he could
satisfy the others). See Down-Lite Int'l, Inc. v. Altbaier,
821
F. App'x 553, 555 (6th Cir. 2020) (stressing that that court only
- 7 -
had to "examine the materially-greater-interest prong to show why
[it] will not upset the parties' choice-of-law provision"). See
also generally Commodity Futures Trading Comm'n v. Zelener, 373
F.3d 861, 868 (7th Cir. 2004) (Easterbrook, J., for the court)
(stating that sometimes it's "[b]est to take Occam's Razor and
slice off needless complexity"); PDK Lab'ys. Inc. v. U.S.D.E.A.,
362 F.3d 786, 799 (D.C. Cir. 2004) (Roberts, J., concurring in
part and concurring in the judgment) (noting that "if it is not
necessary to decide more, it is necessary not to decide more").
Arguments and Analysis
The centerpiece of Hermalyn's argument is that Oxford —
an opinion by the Massachusetts Supreme Judicial Court ("SJC") —
makes the materially-greater-interest issue open-and-shut in his
favor. That's because, he continues, Oxford held that a
Massachusetts choice-of-law clause there couldn't "survive" since
California's interest in not enforcing the contract — a
"Confidentiality, Non-Solicitation and Non-Competition
Agreement," see 106 N.E.3d at 561 n.3 — was "materially greater"
than Massachusetts's interest in enforcing it, see
id. at 564.
And applying Oxford evenhandedly, he submits, entitles him to a
victory here.
Unfortunately for Hermalyn, Oxford can't do the work
that he asks of it.
- 8 -
In pressing his Oxford-centric arguments, Hermalyn
mentions how "California" is "where [he] currently resides, works,
and allegedly breached the non-compete covenant." But Oxford held
that "California ha[d] a materially greater interest than
Massachusetts" there "because" the employee Hernandez had
"executed" and "performed" the contract with his Massachusetts-
based former employer Oxford "in California," and had "allegedly
committed a breach of the agreement in California" after he quit
and joined a California competitor, id. at 566 (emphasis added) —
Hernandez (you see) lived and worked in California before and after
he "allegedly violated" the agreement's "nonsolicitation and
confidentiality provisions,"
id. at 560, 569-70. Plus Oxford also
noted that the "subject matter of the contract — Hernandez's
employment with Oxford — [had been] located exclusively in
California."
Id. at 563 (emphases added). Hermalyn's facts are
not like Hernandez's. According to what the judge here found —
and we see no convincingly developed argument contesting her
findings as reversible error — (i) while "Hermalyn primarily worked
for [Massachusetts-headquartered] DraftKings from New Jersey and
New York, he traveled to Massachusetts for work at least 25 times"
during a 2½-year stretch before leaving for Fanatics — which
roughly comes to "once every [6] weeks"; (ii) "Hermalyn does not
contend, nor is there evidence, that he performed any of his work
responsibilities for DraftKings from California"; and (iii) "[a]ny
- 9 -
harms flowing from Hermalyn's likely" noncompete breach "will be
felt by DraftKings in Massachusetts, not California." Given the
dissimilarities between Oxford and our case, Hermalyn's Oxford-
based arguments sputter out.2
Hermalyn is right about one thing, however. Oxford did
say that California's "legislative[ly]" declared "policy" interest
"in favor of open competition and employee mobility" was
"materially greater" than Massachusetts's — a state, Oxford added,
that lacked a "statute akin" to California's. See 106 N.E.3d at
564-65 (quotation marks omitted). But "a paradigm shift" occurred
when Massachusetts passed a law — the Massachusetts Noncompetition
Agreement Act ("MNAA") — that "dramatically reduc[ed] the number
of Massachusetts employees who can be subjected to . . .
enforceable" noncompetes, while giving those exposed to them
2For similar reasons, Hermalyn gets no help from DCS
Sanitation Mgmt., Inc. v. Castillo, 435 F.3d 892 (8th Cir. 2006),
Cardoni v. Prosperity Bank,
805 F.3d 573 (5th Cir. 2015), and
Waithaka v. Amazon.com, Inc.,
966 F.3d 10 (1st Cir. 2020). DCS
held that Nebraska had a materially greater interest than Ohio
(the contractual choice-of-law state) in the matter because (among
other things) the employees lived and worked in Nebraska before
and after the alleged noncompete breaches. See
435 F.3d at 896-
97. Cardoni ruled that Oklahoma had a materially greater interest
than Texas (the contractual choice-of-law state) in the matter
because (among other things) the employees worked in Oklahoma
before and after the alleged breach. See
805 F.3d at 576-78, 584.
And Waithaka noted that Massachusetts had a materially greater
interest than Washington (the contractual choice-of-law state) in
the matter because the worker "indisputably performed all of his
work pursuant to the contract" in Massachusetts. See
966 F.3d at
34. Compare those cases with Hermalyn's and the differences are
night and day.
- 10 -
"stronger substantive and procedural protections than in the past"
and also "limit[ing]" employers "to substantially reduced post-
employment restrictions." See Jerry Cohen et al., Employee
Noncompetition Laws and Practices: A Massachusetts Paradigm Shift
Goes National, 103 Mass. L. Rev. 31, 31 (2022).
The MNAA was no off-the-cuff, spur-of-the-moment bit of
legislating either. Far from it. The law "gestat[ed]" during a
decade's worth of legislative study and debate. See id.
Interestingly — and we think tellingly — a Massachusetts legislator
actually "propose[d] a bill to completely ban" noncompetes back
"in 2009," see
id. at 53 (emphasis added), a bill "modeled" after
"California" law, see Will Brownsberger, A Study in Persistence
and Compromise (Aug. 13, 2018), https://willbrownsberger.com/a-
study-in-persistence-and-compromise-legislation-regulating-
agreements-not-to-compete/[https://perma.cc./DBV6-VXMH] (emphasis
added) (stating that "the venture capital community" thought that
"California's prohibition of [noncompetes] was a major reason for
California's dominance in technology entrepreneurship").
Concerned Bay State business leaders then stressed how "essential"
noncompetes are "for protecting the ideas that they had developed
through great effort." See
id. And they "promised to leave the
state if [his] legislation passed." See
id. That bill ultimately
went nowhere, however. See
id. But after years and years of
legislative give-and-take, "a hard fought compromise" emerged that
- 11 -
"bann[ed]" noncompetes "for lower level workers, limit[ed] them
for higher level workers[,] and provid[ed] procedural protections
to assure that workers know what they are getting into when they
sign them." See id. (emphasis added); see also Cohen et al.,
supra, at 32-33 (discussing the history of the bill's enactment).
The MNAA became law in August 2018, just days before the
SJC's September 2018 Oxford ruling. See 2018 Mass. Legis. Serv.
ch. 228, § 71. But the MNAA applies only to noncompetes "entered
into on or after October 1, 2018."3 See id. (emphasis added).
Which explains why Oxford never even hinted that the MNAA played
any role there. And which also shows that Oxford had in mind the
state of law before the MNAA when it talked about California's
interest being "materially greater" than Massachusetts's — not the
state of the law after the MNAA.
Perhaps anticipating some of these points, Hermalyn
notes that Oxford "predate[s]" California's 2024 passage of laws
reinforcing its deep-rooted policy against restraint of trade.4
But the fact still remains that Oxford — which (as we said)
acknowledged California's "settled legislative policy" of
promoting "open competition and employee mobility," see 106 N.E.3d
3 Hermalyn's noncompete specifically cites to the MNAA.
4 The drafting committee for these statutes conceded that
while they "express[ed] California's strong desire to enforce its
public policy," they "cannot dictate to courts outside of its
jurisdiction."
- 12 -
at 564 (quotation marks omitted and emphasis added) — ran the
materially-greater-interest analysis without having to factor in
the MNAA (because (again) the MNAA wasn't in play there).
Still looking for a winning argument, Hermalyn writes
that "the MNAA did not purport to abrogate Massachusetts law as
discussed and understood in Oxford," quoting for support this
snippet from the Massachusetts Law Review article mentioned above:
the MNAA "codified aspects of the common law that will continue to
require case-by-case exposition." See Cohen et al., supra, at 31.
Hermalyn's point seems to be that the MNAA didn't change Bay State
"law or policy of limited tolerance towards non-compete
covenants." But the article's animating thesis — revealed in the
following excerpts (we quoted some of them a few paragraphs back)
— couldn't be any clearer: the MNAA "represents a paradigm shift
in favor of employees" that "dramatically reduc[ed] the number of
Massachusetts employees who can be subjected to . . . enforceable"
noncompetes, while offering those covered by them "stronger
substantive and procedural protections than in the past" and
likewise "limit[ing]" employers "to substantially reduced post-
employment restrictions." See id. (emphases added). Which
scotches Hermalyn's attempt to downplay the MNAA's significance.5
Roll Sys., Inc. v. Shupe, No. CIV.A.97-12689, 1998 WL
5
1785455 (D. Mass. Jan. 22, 1998) — cited by Hermalyn — doesn't
compel a different result. Highlighting California's "fundamental
policy against the enforcement of [broad] restrictive covenants,"
- 13 -
Given this tableau — involving (among other features)
two states passing laws reflecting different but careful balances
of conflicting forces in the noncompete area (after the usual push-
and-pull of politics), with Massachusetts opting not to mimic
California's ban and instead generally allowing noncompetes for
higher-level employees like Hermalyn (who unlike lower-level
employees often have business-sensitive info and deep ties with
company customers) — we can't say that Hermalyn has shown (as he
was required to do) that California's "interest" in pursuing its
policy is not just "greater" than Massachusetts's, but is
"materially" so. See Down-Lite, 821 F. App'x at 556 (acknowledging
that "California has a meaningful interest in protecting its
resident from" his Ohio-based former employer's "desire to
restrict competitive conduct," but holding that "that interest is
not materially greater than Ohio's interest in protecting one of
its closely held businesses operating in the global economy").6
see id. at *2, Roll Sys. "disregard[ed]" a Massachusetts "choice-of-law provision in the contract and appl[ied] California law
instead," in a case involving a preliminary-injunction dispute
pitting an employee and his current employer against his former
employer, see id. at *1-3. But Roll Sys. came many years before
the MNAA. Which — whatever else may be said about Roll Sys. — is
enough to distinguish that case from Hermalyn's.
6 DraftKings makes many of the same points about
Massachusetts's public policy on noncompetes. Unimpressed with
DraftKings's effort, Hermalyn blasts the company for "mak[ing] no
headway in attempting to equate Massachusetts'[s] policy interests
with California's." But his protest ignores that he (not
DraftKings) had to show (as the above-text explains) not just that
California's interest equals Massachusetts's, or even that it's
- 14 -
Or to put it differently but with the same result, he hasn't shown
that "California's public policy" eclipses "the parties' clear and
unambiguous agreement to apply [Massachusetts] law." See id.7
Wrap-Up
The short of it is that the judge didn't err by ruling
that Massachusetts law governs Hermalyn's noncompete with
DraftKings.8
SCOPE OF THE PRELIMINARY INJUNCTION
Standard of Review
Judges must closely tailor injunctive relief to the
specific harm alleged. See, e.g., Vaquería Tres Monjitas, Inc. v.
greater than Massachusetts's, but that it's materially greater
than Massachusetts's. Which (again) he hasn't done.
Application Grp., Inc. v. Hunter Grp., Inc., 72 Cal. Rptr.
7
2d 73 (Cal. Ct. App. 1998), isn't a difference-maker for Hermalyn
either — despite what he thinks. Application did hold that
California had a materially greater interest in noncompetes than
Maryland (the contractual choice-of-law state). See
id. at 76,
86. But there — unlike here — "[t]here [was] no showing that [the
employee] was attempting to exploit [the former employer's] trade
secrets or other protected information," see
id. at 86, a
distinguishing point raised in DraftKings's answering brief
without contradiction from Hermalyn in his reply brief. Also
making this case worlds apart from Hermalyn's is that (as we just
said) Application addressed whether California had a materially
greater interest in noncompetes than Maryland, not Massachusetts.
Hermalyn writes that "[s]eparate and apart from the grounds
8
outlined in []his brief, enforcement of" the noncompete "will also
be precluded by the recently promulgated Federal Trade Commission
. . . rule barring most" noncompetes. But he adds that he hasn't
"sought relief based on th[at] rule" since it's not yet in
"effect." See generally Ryan, LLC v. Fed. Trade Comm'n, No. 3:24-
CV-00986-E, 2024 WL 3879954, at *14 (N.D. Tex. Aug. 20, 2024)
(concluding that because the commission lacked the power to issue
- 15 -
Irizarry, 587 F.3d 464, 487 (1st Cir. 2009). And because they're
uniquely placed to design "the scope of injunctive relief to
[their] factual findings," we review only for abuse of discretion.
See, e.g.,
id.
Arguments and Analysis
We can make quick work of Hermalyn's alternative
position: that if Massachusetts law controls DraftKings's bid to
enforce the noncompete, then the judge (under "comity and
federalism" principles) had to — but didn't — exclude California
from the preliminary injunction's scope given California's strong
"public policy barring the enforcement of such [agreements]
against California employees of California companies working in
California." We sidestep DraftKings's claim that Hermalyn waived
or forfeited this argument because as DraftKings also contends,
his argument fails regardless.
Starting with Hermalyn's public-policy point, we've
already explained above why California's policy can't override
Massachusetts's. And the cases he champions — Barnes Grp., Inc.
v. C & C Prods., Inc., 716 F.2d 1023, 1035 (4th Cir. 1983), Keener
v. Convergys Corp.,
342 F.3d 1264, 1268-69 (11th Cir. 2003), and
Herman Miller, Inc. v. Palazzetti Imps. & Exps., Inc., 270 F.3d
that rule, "the [r]ule shall not be enforced or otherwise take
effect" anywhere nationwide "on its effective date of September 4,
2024"). So we needn't say anything more on that subject.
- 16 -
298, 327 (6th Cir. 2001), where reviewing courts narrowed an
injunction's reach — don't move the needle either. Our reasons
for thinking so are simple. (i) California outlaws online sports
betting. See Tak Chun Gaming Promotion Co. v. Long, 314 Cal. Rptr.
3d 890, 897 (Cal. Ct. App. 2023). (ii) But a big part of Hermalyn's
job is creating and keeping relationships with digital-gaming
customers. (iii) So if he can join Fanatics in Los Angeles
straightaway, he'll inevitably interact with clients outside
California where online sports betting is legal. (iv) Clearly his
requested California carveout will give him a way to skirt the
countrywide preliminary injunction's one-year noncompete ban.
(v) And returning to his trio of cases with all this in mind, we
note that none covers a situation like ours — where lopping a state
off a USAwide preliminary injunction on a one-year noncompete would
entirely undercut that injunction's effectiveness. (vi) Which
goes to show that those cases hold no sway here. DraftKings made
each of these romanette-numbered points about the injunction's
breadth in its answering brief, with no pertinently persuasive
push back from Hermalyn in his reply brief.
- 17 -
Wrap-Up
The bottom line is that the judge didn't err by including
California within the preliminary injunction's range.
LAST WORDS
All that's left to say is: affirmed, with appellate
costs to DraftKings.
- 18 -
