FILED
United States Court of Appeals
Tenth Circuit
UNITED STATES COURT OF APPEALS February 9, 2021
Christopher M. Wolpert
TENTH CIRCUIT Clerk of Court
MICHELE KENNETT, individually and
on behalf of the Rule 23 Class,
Plaintiff - Appellee,
No. 19-1004
v. (D.C. No. 1:14-CV-02005-CMA-NRN)
(D. Colo.)
BAYADA HOME HEALTH CARE,
INC.,
Defendant - Appellant.
ORDER AND JUDGMENT *
Before HOLMES, KELLY, and BACHARACH, Circuit Judges.
Defendant-Appellant Bayada Home Health Care, Inc. (“Bayada”) appeals
from the district court’s order denying its motion for summary judgment and
granting Plaintiff-Appellee Michele Kennett’s (“Ms. Kennett”) cross-motion for
summary judgment. The parties’ dispute centers on the scope of Colorado’s
wage-and-hour regulations.
*
This order and judgment is not binding precedent except under the
doctrines of law of the case, res judicata, and collateral estoppel. It may be cited,
however, for its persuasive value consistent with Fed. R. App. P. 32.1 and 10th
Cir. R. 32.1.
Under Colorado’s Minimum Wage Order (the “Wage Order”), employers
must pay their employees time-and-a-half wages for overtime work, with some
job classifications exempted from the overtime requirement. 1 The scope of one of
those exemptions—the “companionship exemption”—is the subject of this case.
Under the companionship exemption, employers need not pay overtime to
“companions, casual babysitters, and domestic employees employed by
households or family members to perform duties in private residences.” 7 C OLO .
C ODE R EGS . § 1103-1:5 (2019). 2 The question before us is whether the phrase
1
The Wage Order—which is promulgated by the Colorado Division of
Labor (the “Division”)—has been amended many times, and different numbers
govern different time periods. The versions of the Wage Orders in effect during
the time period that gives rise to this lawsuit are Nos. 30 and 31. These two
Wage Orders contain identical language with respect to the “companionship
exemption” at issue in this appeal. Accordingly, we refer to them both
collectively as “the Wage Order.”
2
As we have noted, the Wage Order regulation has been amended
frequently. The language at issue in this appeal—found in Wage Orders Nos. 30
and 31—remained in effect through 2019 and the early months of 2020. The most
recent version of the Wage Order containing this language—that is, Wage Order
No. 35—became effective at the beginning of 2019. For convenience, we cite to
this version of the Wage Order. See generally Jordan v. Maxim Healthcare
Servs., Inc., 950 F.3d 724, 727–28 & nn. 3–4 (10th Cir. 2020) (citing and
discussing this same Wage Order). Notably, the Division recently promulgated a
new version of the Wage Order—found in Wage Order No. 36—which supersedes
all prior versions; it became effective on an emergency basis on March 16, 2020,
and, as amended to its current form, became effective July 15, 2020. This new
version of the Wage Order has completely removed the companionship-exemption
language contained in earlier versions of the Wage Order. See 7 C OLO . C ODE
R EGS . § 1103-1:2.2.7 (2020) (providing an “exemption” that “covers the below-
listed in-residence employees” but not including in the following list the
companionship-exemption language).
2
“employed by households or family members to perform duties in private
residences” (referred to herein as the “household modifier”) modifies only
“domestic employees” or all three occupations—that is, also includes
“companions” and “casual babysitters.” If the former (i.e., modifies only
“domestic employees”)—the reading Bayada advances—then all companions,
irrespective of the nature of their employer, are exempt from the overtime
requirement. If the latter (i.e., modifies all three occupations)—the reading Ms.
Kennett advances—then only companions employed directly by households or
family members, as opposed to companions employed by all types of employers,
including third-party employers like Bayada, are exempt.
Unfortunately for Ms. Kennett, we resolved this issue in a recent, published
decision, Jordan v. Maxim Healthcare Services, Inc., wherein we concluded that
“the companionship exemption applies to all companions—including those
employed by third-party employers.” 950 F.3d 724, 731 (10th Cir. 2020). We are
bound by this decision, which governs all salient issues in this case. Accordingly,
exercising jurisdiction under 28 U.S.C. § 1291, we reverse the district court’s
judgment and remand for further proceedings consistent with this order and
judgment.
I
Bayada is a healthcare company that provides in-home nursing services to
its clients. Ms. Kennett worked for Bayada as a Home Health Aide (“HHA”). It
3
is undisputed that Ms. Kennett and her fellow HHAs were “companions” under
Colorado law and that Bayada did not pay them overtime. Ms. Kennett filed a
class action complaint against Bayada on behalf of herself and other HHAs for
allegedly violating the Wage Order by failing to pay overtime. Bayada moved for
summary judgment on the ground that the HHAs fall under the Wage Order’s
companionship exemption for overtime pay. Ms. Kennett cross-moved for
summary judgment on the basis that the HHAs do not fall under the
companionship exemption.
The district court denied Bayada’s motion and granted Ms. Kennett’s
cross-motion. In reaching its decision, the district court held that the “only
grammatically sound reading of the statute . . . dictates that the household
[modifier] is equally applicable to the antecedents ‘companions’ and ‘casual
babysitters’ as it is to ‘domestic employees.’” Aplt.’s App., Vol. III, at 525a
(Order Denying Def.’s Mot. for Summ. J. & Granting Pl.’s Cross Mot. for Summ.
J., entered Sept. 24, 2015). The district court also found that its holding was
supported by the series-qualifier canon, and that the Colorado Division of Labor’s
(the “Division”) contrary interpretation of the regulation was not entitled to any
deference. Id. at 525a–531a. Consequently, the district court held that the Wage
Order’s companionship exemption did not apply to Bayada’s HHAs because they
were not employed directly by a household or family member. Bayada now
appeals from this decision.
4
II
We review the grant or denial of summary judgment de novo, applying the
same legal standard as the district court. See, e.g., Nielson v. Ketchum, 640 F.3d
1117, 1121 (10th Cir. 2011). Summary judgment will be granted if “the movant
shows that there is no genuine dispute as to any material fact and the movant is
entitled to judgment as a matter of law.” F ED . R. C IV . P. 56(a). “Cross motions
for summary judgment are to be treated separately; the denial of one does not
require the grant of another.” Christian Heritage Acad. v. Okla. Secondary Sch.
Activities Ass’n, 483 F.3d 1025, 1030 (10th Cir. 2007) (quoting Buell Cabinet Co.
v. Sudduth, 608 F.2d 431, 433 (10th Cir. 1979)). We also review the district
court’s “conclusions on legal issues de novo . . . and need not defer to its
decisions on questions of law.” Kokins v. Teleflex, Inc., 621 F.3d 1290, 1294–95
(10th Cir. 2010) (quoting City of Wichita v. U.S. Gypsum Co., 72 F.3d 1491, 1495
(10th Cir. 1996)).
“In a diversity case like this one, [where] the substance . . . is a matter of
state law . . . . our task is ‘not to reach [our] own judgment regarding the
substance of the common law, but simply to ascertain and apply the state law.’”
Id. at 1295 (second alteration in original) (quoting Wankier v. Crown Equip.
Corp., 353 F.3d 862, 866 (10th Cir. 2003)). Critically for this appeal, “we are
bound by ‘our own precedent interpreting a state’s law.’” Jordan, 950 F.3d at
731 (quoting United States v. Badger, 818 F.3d 563, 569 (10th Cir. 2016)); see
5
also Kokins, 621 F.3d at 1295 (“[W]hen a panel of this Court has rendered a
decision interpreting state law, that interpretation is binding on district courts in
this circuit, and on subsequent panels of this Court, unless an intervening decision
of the state’s highest court has resolved the issue.” (alteration in original)
(emphasis omitted) (quoting Wankier, 353 F.3d at 866)).
III
At issue in this appeal is whether “companions,” like Ms. Kennett, who are
employed by third-party employers, fall within the Wage Order’s companionship
exemption. Bayada contends that they do because the household modifier applies
only to “domestic employees”—therefore, all companions are exempt, including
those employed by third-party employers. Bayada reaches this conclusion by
arguing that (1) the Wage Order is ambiguous; (2) the district court should have
applied the last-antecedent rule instead of the series-qualifier canon of statutory
construction in construing the Wage Order; and (3) the Division’s interpretation
of the companionship exemption as applying to all companions, including those
employed by third-party employers, is entitled to deference.
Ms. Kennett contends the opposite. That is, she contends that companions
employed by third-party employers do not fall within the Wage Order’s
companionship exemption because the household modifier applies to “domestic
employees,” “babysitters,” and “companions.” More specifically, Ms. Kennett
argues the district court correctly concluded that (1) the Wage Order is
6
unambiguous; (2) the Wage Order’s plain meaning, as elucidated by applying the
series-qualifier canon, compels a finding that only companions employed by
households or family members are exempt from overtime payments; and (3) the
Division’s interpretation of the Wage Order and the companionship exemption is
not entitled to deference.
We are constrained to conclude that, generally speaking, Bayada has the
better of the argument. As noted supra, we recently addressed the issue of the
companionship exemption’s breadth in Jordan v. Maxim Healthcare Services,
Inc. 3 In Jordan, we held that the Wage Order’s companionship exemption covers
all companions, including those employed by third-party employers. See 950
F.3d at 731. That holding governs this case. We therefore reverse the district
court’s contrary judgment.
We explain our holding in three steps, relying on our analysis in Jordan
along the way. First, we conclude that the Wage Order is ambiguous. Second,
we hold that the district court erred by applying the series-qualifier canon. Third,
we conclude that the Division’s interpretation, even if not entitled to deference, is
at least persuasive authority.
3
In its 28(j) letter, Bayada “submits that both the rationale and
holding in the Jordan v. Maxim Healthcare Services decision are directly
applicable to and controlling in the instant appeal.” See Aplt.’s 28(j) Letter at 1–2
(filed Feb. 25, 2020). Bayada similarly characterized the Jordan case as a
“related appeal” that is “directly aligned” with and “present[s] the exact same
issue” as the instant appeal. Oral Arg. at 0:23–0:45. Having now issued our
published decision in Jordan, we agree with Bayada.
7
A
“Colorado law governs ‘[o]ur interpretation and application of the Wage
Order’ and, more specifically, of the companionship exemption.” Jordan, 950
F.3d at 731 (alteration in original) (quoting Deherrera v. Decker Truck Line, Inc.,
820 F.3d 1147, 1160 (10th Cir. 2016)). Under Colorado law, “we interpret the
companionship exemption according to Colorado’s ordinary rules of statutory
interpretation.” Id. “The primary goal of interpretation is to ‘give effect to the
intent of the enacting body.’” Id. at 732 (quoting United States v. Richter, 796
F.3d 1173, 1185 (10th Cir. 2015)). To accomplish our goal, we “look to the plain
meaning of the [regulatory] language and consider it within the context of the
[regulation] as a whole.” Id. (alterations in original) (quoting Denver Post Corp.
v. Ritter, 255 P.3d 1083, 1088 (Colo. 2011)). But “to the extent that the language
of a [regulation] is susceptible of more than one reasonable interpretation, and is
therefore considered ambiguous, a substantial body of interpretive aids . . . is
available to assist in determining which of these reasonable understandings
embodies the [enacting body’s] intent.” People v. Opana, 395 P.3d 757, 760
(Colo. 2017) (citations omitted).
At the times of the events at issue here, Colorado’s Wage Order exempted
from its overtime provisions “companions, casual babysitters, and domestic
employees employed by households or family members to perform duties in
private residences.” 7 C OLO . C ODE R EGS . § 1103-1:5. Bayada argues that the
8
district court erred when it held that the “only grammatically sound reading of
th[is language] . . . dictates that the household [modifier] is equally applicable to
the antecedents ‘companions’ and ‘casual babysitters’ as it is to ‘domestic
employees.’” Aplt.’s App., Vol. III, at 525a; see Aplt.’s Opening Br. at 18–20.
According to Bayada, the companionship exemption is ambiguous because it is
“susceptible to more than one reasonable interpretation or understanding, leading
to different results.” Aplt.’s Opening Br. at 23. In particular, Bayada contends
the exemption is ambiguous because “the pertinent phrase ‘employees employed
by households or family members to perform duties in private residences’ can
reasonably be read as applying either to all of the enumerated occupations (i.e.,
companions, causal babysitters and domestic employees) or to only the
occupation immediately preceding it (i.e., domestic employees).” Id.; see People
v. Weeks, 369 P.3d 699, 711 (Colo. App. 2015) (holding that a statute was
ambiguous where “the last phrase could be read as applying either to all of the
enumerated [categories] or only to the last one”).
We agree with Bayada. See Jordan, 950 F.3d at 733–34 (finding the
companionship exemption ambiguous because it could plausibly be read as
applying to all companions or only to companions “employed by households or
family members to perform duties in private residences,” and we lacked a
“reasoned way to pick between [these two readings], absent an examination of the
language’s context”); see also id. at 735–39.
9
Ms. Kennett responds by arguing that Bayada’s reading is “contrary to the
unambiguous plain meaning of the text” because it would “introduce a
superfluous ‘and’ in the middle of a list”—a result she describes as
“grammatically grotesque.” Aplee.’s Resp. Br. at 13. However, we rejected an
identical argument in Jordan. See 950 F.3d at 734; see also id. at 735 (noting
that “more often [than not], commas and conjunctions do not imbue a sentence
with only one meaning,” and that the “and” before domestic employees was “one
such case”). We therefore reject Ms. Kennett’s argument as well.
Accordingly, we conclude that the companionship exemption is ambiguous,
and because of this conclusion, “we may ‘look to other aids in construction’ to
pin down the [exemption’s] proper scope.” Id. at 739 (quoting Specialty Rests.
Corp. v. Nelson, 231 P.3d 393, 397 (Colo. 2010)). These aids include canons of
statutory construction and the Division’s interpretation of the companionship
exemption. We address each in turn.
B
According to Bayada, the district court erred by applying the series-
qualifier canon to conclude that the household modifier applied to all three
occupations, including companions. Instead, Bayada argues, the district court
should have interpreted the household modifier as applying only to domestic
employees (the occupation immediately preceding the household modifier
10
language), consistent with the last-antecedent rule. We agree that the district
court erred in applying the series-qualifier canon.
The series-qualifier canon dictates that “[w]hen there is a straightforward,
parallel construction that involves all nouns or verbs in a series, a prepositive or
postpositive modifier normally applies to the entire series.” Potts v. Ctr. for
Excellence in Higher Educ., Inc., 908 F.3d 610, 615 (10th Cir. 2018) (alteration
in original) (quoting A NTONIN S CALIA & B RYAN A. G ARNER , R EADING L AW : T HE
I NTERPRETATION OF L EGAL T EXTS 147 (2012)). Conversely, the last-antecedent
rule “provides that ‘a limiting clause or phrase . . . should ordinarily be read as
modifying only the noun or phrase that it immediately follows.’” Jordan, 950
F.3d at 746 (omission in original) (quoting Lockhart v. United States, 136 S. Ct.
958, 962 (2016)). Thus, if one were to apply the series-qualifier canon to the
companionship exemption, one would conclude that the household modifier
applies to all three occupations, including companions, such that companions
employed by third-party employers are not covered by the exemption. On the
other hand, if one were to apply the last-antecedent rule, one would conclude that
the household modifier applies only to the occupation immediately preceding it
(i.e., domestic employees) such that all companions—including those employed
by third-party employers—are covered.
In Jordan, we noted that the series-qualifier canon is, “[p]erhaps more than
most of the other canons, . . . highly sensitive to context.” Id. at 745 (quoting
11
S CALIA & G ARNER , supra, at 150). “‘Often the sense of the matter prevails’ over
the meaning the series-qualifier canon suggests.” Id. (quoting S CALIA &
G ARNER , supra, at 150). Moreover, “like all canons of construction, [the
series-qualifier canon] is merely an interpretive aid, not an absolute rule.” Id.
(alteration in original) (quoting Benefield v. Colo. Republican Party, 329 P.3d
262, 267 (Colo. 2014)). That is, “[i]ts utility depends on ‘context and
consideration of other, and often conflicting, interpretive aids.’” Id. (quoting
Benefield, 329 P.3d at 267).
Based on this understanding, we held in Jordan that application of the
series-qualifier canon to the companionship exemption was inappropriate, in light
of the exemption’s context. Id. (“In our view, the context is not suitable for
application of the series-qualifier canon and, accordingly, the district court should
have declined to apply it.”). Specifically, we concluded that application of the
series-qualifier canon would “undermine” the Division’s intent. Id. As implied
by the ordinary and particular meanings of the terms, the Division intended that
the companionship exemption “apply to those household workers who have a
close personal connection to the recipients of their services,” including
“companions employed by third-party employers.” Id.
We determined further that “[t]he aptness of the series-qualifier canon in
this context is . . . undermined” by other considerations:
12
Most notably, using the series-qualifier canon here creates
conflict with the enabling statute, which expressly requires that
any overtime requirements for particular categories of
employees “apply equally to all employers in such industry or
occupation.” C OLO . R EV . S TAT . § 8-6-111(4). In particular, as
most relevant here, applying the household modifier to the term
“companions”—in line with the series-qualifier canon—would
allow for the differential application of the companionship
exemption to employers in the same industry or occupation,
contrary to the mandate of the enabling statute. Not only that,
but application of the series-qualifier canon also gives rise to
surplusage. Given these signs that the context is not suitable for
application of the series-qualifier canon, we believe the district
court should have declined to apply it.
Id. Thus, because the series-qualifier canon conflicted with the enabling statute
and the surplusage canon, we held that (1) application of the canon was
inappropriate, and (2) the district court should have held that the household
modifier applied to domestic employees only.
Ms. Kennett points out that the last-antecedent rule “has not been adopted
by the general assembly [of Colorado] and does not create any presumption of
statutory intent.” C OLO . R EV . S TAT . § 2-4-214. However, this simply means that
we “cannot add [the last-antecedent rule] to the already lengthy list of reasons to
part from the series-qualifier canon.” Jordan, 950 F.3d at 747. We are still free
“to eschew the series-qualifier canon and instead follow the context and other
canons,” which, as here, may yield a result that, in substance, is consistent with
the application of the last-antecedent rule. Id. at 748; see also Weeks, 369 P.3d at
711 (holding that while § 2-4-214 precludes a“presum[ption] that the legislature
13
intended that the last phrase apply only to the last [category in a series],” it does
not create a presumption “that the last phrase applies to all of the previously
enumerated [categories in that series]” either).
In sum, these considerations “compel the conclusion that the
companionship exemption applies to all companions—irrespective of whether
their employers are households or family members on the one hand, or third-party
companies, like [Bayada], on the other.” Jordan, 950 F.3d at 750. “[W]e find
[additional] persuasive support for our conclusion in the Division’s longstanding
and consistent interpretation of the companionship exemption,” id., to which we
now turn.
C
1
Bayada argues that the district court erred by refusing to defer to the
Division’s longstanding interpretation of its own regulation. As Bayada points
out, the Division both promulgates and enforces the Wage Order. See C OLO . R EV .
S TAT . § 8-6-111(4); see also Aplt.’s Opening Br. at 30–31. In 2006, it issued an
opinion letter explaining that the companionship exemption was intended to
mirror the federal Fair Labor Standards Act (the “FLSA”). And the U.S.
Department of Labor—the agency responsible for administering the FLSA—had
promulgated regulations that, at the time, explicitly exempted companions
employed by third-party employers from the FLSA’s overtime requirements. See
14
Aplt.’s App., Vol. I, at 138a–139a (Colo. Dep’t of Labor & Emp’t, Div. of Labor,
Op. Letter, dated Aug. 3, 2006); see also Jordan, 950 F.3d at 729, 750–51
(discussing this opinion letter). Thus, the Division had long interpreted the
companionship exemption as extending to companions employed by third-party
employers, just like the FLSA regulations. See Aplt.’s App., Vol. I, at 139a (“In
concordance with applicable federal regulations . . . it is the Division’s
enforcement policy that the practice of applying the companionship exemption in
situations involving third party employers is acceptable under [the] Colorado . . .
Wage Order . . . .”); see also 29 C.F.R. § 552.109(a) (2014) (“Employees who are
engaged in providing companionship services . . . and who are employed by an
employer or agency other than the family or household using their services, are
exempt from the [FLSA’s] minimum wage and overtime pay
requirements . . . .”). 4
The Division subsequently issued enforcement letters reiterating its
position that “the treatment and interpretation of the companions[hip] exemption
in the Wage Order is intended to mirror the current definition and associated
regulations in federal law.” Aplt.’s App., Vol. III, at 584a (Division’s Resp. to
Maxim Emp.’s 2014 Claim, dated July 23, 2014); id. at 588a (Division’s Resp. to
Maxim Emp.’s 2012 Claim, dated Aug. 28, 2012) (“The Wage Order provides
4
As will be discussed infra, since the time period relevant to this case,
the FLSA has been amended to exclude companions employed by third-party
employers from its overtime exemption. See also Jordan, 950 F.3d at 727 n.2.
15
exemptions for individuals performing work that is classified as companion care.
Colorado law is intended to mimic federal law . . . .”); see also Jordan, 950 F.3d
at 729 (discussing these enforcement letters).
Bayada and Ms. Kennett raise a variety of arguments about what level of
formal deference we should give, if any, to the Division’s interpretation of the
companionship exemption. We decline to address this issue. As we held in
Jordan, “[w]e need not decide whether any deference attaches to the Division’s
interpretation because under Colorado law, it is undisputed that we may treat the
Division’s interpretation[s] of its Wage Order as persuasive authority”—if those
interpretations have the power to persuade. 950 F.3d at 751–52 (footnote
omitted); see also Banner Advert., Inc. v. City of Boulder, 868 P.2d 1077, 1083
(Colo. 1994) (“While opinion letters from administrative agencies are not binding
authority, they can be used as persuasive authority.”). And in Jordan, we in fact
“f[ound] the Division’s interpretation persuasive.” 950 F.3d at 752; see id.
(noting that the Division’s “unbroken, oft-expressed position [was] potent
evidence that it intended the companionship exemption to apply to companions
employed by third-party employers”); see also id. at 732 (“The primary goal of
interpretation is to ‘give effect to the intent of the enacting body.’” (quoting
Richter, 796 F.3d at 1185)). Jordan’s view of the Division’s interpretation is
controlling here.
16
Ms. Kennett offers several arguments for why the Division’s interpretation
of the companionship exemption is unpersuasive. See Aplee.’s Resp. Br. at
27–37. These arguments are unavailing in light of Jordan. First, Ms. Kennett
argues that the Division’s interpretation is unpersuasive because it conflicts with
the plain language of the Wage Order. See Aplee.’s Resp. Br. at 29; cf. Rags
Over the Ark. River, Inc. v. Colo. Parks & Wildlife Bd., 360 P.3d 186, 192 (Colo.
App. 2015) (“[W]hile an agency’s reasonable interpretation of its own regulations
is ordinarily entitled to deference, an interpretation that is inconsistent with the
plain language of the regulation is not.” (citations omitted)). But, as we discussed
supra, the companionship exemption is ambiguous and could reasonably be
interpreted as covering companions employed by third-party employers. In other
words, the Division’s interpretation does not conflict with the language of the
Wage Order because it embraces one plausible interpretation of that language.
Second, Ms. Kennett argues that the opinion letter is unpersuasive because
it lacks a reasoned analysis. See Aplee.’s Resp. Br. at 35–37; see also Skidmore
v. Swift & Co., 323 U.S. 134, 140 (1944) (holding that the weight afforded to an
agency opinion letter “depend[s] upon the thoroughness evident in its
consideration, the validity of its reasoning, . . . and all those factors which give it
power to persuade”). But we held in Jordan that the opinion letter contains
sufficient analysis to persuade. See 950 F.3d at 753 (“[The 2006 opinion letter]
read[s] like a syllogism: the Division intended the companionship exemption to
17
mirror federal law; federal law expressly included companions employed by
third-party employers within its exemption; therefore, the companionship
exemption applies to companions employed by third-party employers. We are
thus constrained to disagree with the district court’s assessment that the letter[]
contained ‘no analysis’ and ‘no reasoning.’”).
Ms. Kennett also argues that the opinion letter’s typos and numerous
qualifiers undermine its authority. See Aplee.’s Resp. Br. at 36–37; see also
Aplt.’s App., Vol. I, at 139a (“This position of the Division is based upon the
case-specific information provided, and does not necessarily generalize to other
contexts or situations. . . . The position of the Division may also change over
time. This position is not legal advice, and the opinion of attorneys and the
judicial system may differ.”). However, we concluded in Jordan that these
disclaimers simply “clarify that the opinion letter[] [is] informal and nonbinding.”
950 F.3d at 754. “But even informal and nonbinding letters ‘can be used as
persuasive authority.’” Id. (quoting Banner Advert., Inc., 868 P.2d at 1083).
Moreover, Ms. Kennett does not cite to any authority to show that a single typo
(she fails to specifically identify more than one) will render an agency opinion
letter unpersuasive.
Additionally, Ms. Kennett questions the opinion letter’s persuasive power
because it contends that the companionship exemption is intended to mirror the
FLSA, yet ignores other, significant differences between the FLSA and the Wage
18
Order. Aplee.’s Resp. Br. at 31–35. But the Division never claimed that the
Wage Order mirrors the FLSA—more specifically, its applicable regulations—in
every respect. Rather, it merely stated that the companionship exemption was
intended to mirror the FLSA and its applicable regulations. Thus, no extra
analysis was necessary on the Division’s part to explain the differences between
other portions of the Wage Order and these sources of federal law.
Finally, Ms. Kennett argues that the Division’s opinion letter “is also
problematic in light of the new federal regulations.” Id. at 37. Specifically, she
reasons that the Division’s contention, in the 2006 opinion letter, that the
companionship exemption is intended to mirror the FLSA—in particular, the
FLSA regulations—“loses all credibility,” when one considers that in 2015 the
FLSA regulations were changed to exclude third-party employers from the
companionship exemption, but Colorado’s Wage Order “stayed the same.” Id. at
37–38. However, the most straightforward and determinative response to Ms.
Kennett’s line of reasoning is the one that Bayada offers. It points out,
“[b]ecause the prior version of the FLSA [and its regulations] w[ere] in effect
during the relevant time period at issue in this case, including at the time the
Division drafted the ‘companion[ship]’ exemption and various interpretative aids,
the recent changes [i.e., to the FLSA regulations] are simply not at issue for
purposes of this appeal.” Aplt.’s Reply Br. at 14 n.6. In other words, during the
period material and relevant to this case, the pre-2015 FLSA regulations were in
19
effect, and that was the version of federal law that the Division used as a critical
touchstone in crafting the companionship exemption, which is controlling here,
and the Division’s related guidance materials. That the federal government later
changed its regulations in 2015 could not retrospectively alter the operative
touchstone that the Division used prior to that time—i.e., the touchstone of the
pre-2015 FLSA regulations—and that is the touchstone that is material and
relevant to this appeal.
Moreover, though the Division explicitly referenced its longstanding
intention in the 2006 opinion letter that the scope of the companionship
exemption mirror federal law, it did not indicate that this regulatory position was
somehow immutable, nor that the Division was somehow obliged to march in
lockstep with the federal government’s view of the companionship-exemption’s
scope. Cf. Aplt.’s App., Vol. I, at 138a (noting, without suggesting that change
was not possible, that the Division’s “position” that the Wage Order was
“intended to [] mirror . . . federal law . . . has remained unchanged . . . since . . .
1998”). Therefore, as a matter of logic, it would not signify any lack of
credibility in the position that the Division took in its 2006 opinion letter for the
Division subsequently to persist in its understanding of the companionship-
exemption’s scope—as including third-party employers—even though the federal
government elected to alter its position concerning this question in the 2015
FLSA regulations.
20
In sum, controlled by our analysis in Jordan, we determine that the
Division’s “unbroken, oft-expressed position is potent evidence that it intended
the companionship exemption to apply to companions employed by third-party
employers,” 950 F.3d at 752, and Ms. Kennett’s arguments do nothing to
undermine our determination on this matter.
2
In concluding our analysis, we note that, in a second 28(j) letter on the eve
of oral argument, Ms. Kennett informed us that the Division has issued a Public
Rulemaking Notice in which it stated that the current exemption for companions
“reach[es] only those employed directly ‘by households or family members to
perform duties in private residences.’” Aplee.’s Second 28(j) Letter, Ex. C, at 23
(filed on Nov. 18, 2019). In other words, the Division signaled that it planned for
its interpretation of the companionship exemption to mirror the 2015 amendment
to the FLSA regulations.
In her 28(j) letter, Ms. Kennett argues that the Division’s new
interpretation “confirm[s] the district court’s reading of the regulation, namely,
that the companion[ship] exemption does not apply to third-party employers like
[Bayada].” Id. at 2. In another 28(j) letter, she also argues that the Division’s
June 24, 2019, notice rescinding all Division-drafted opinion letters “support[s]
[her] position (and the district court’s conclusion) that the [Division’s 2006]
21
Opinion Letter . . . [is] not entitled to deference.” Aplee.’s First 28(j) Letter at 2
(filed July 3, 2019). We disagree on both counts.
Notably, these 28(j) letters alert the court to actions the Division has taken
long after the time period that forms the basis for Ms. Kennett’s lawsuit. Yet,
Ms. Kennett has neither cited to any authority showing that the Division’s actions
have retroactive effect nor even made such an argument. Cf. De Niz Robles v.
Lynch, 803 F.3d 1165, 1169–71 (10th Cir. 2015) (discussing the presumption
against retroactive application of laws in the agency context); Ficarra v. Dep’t of
Regulatory Agencies, 849 P.2d 6, 11–12 (Colo. 1993) (noting Colorado’s statutory
and common law presumption favoring the prospective application of statutes);
Colo. Office of Consumer Counsel v. Pub. Serv. Co. of Colo., 877 P.2d 867,
870–72 (Colo. 1994) (holding that a certain regulation violated the Colorado
Constitution’s prohibition against retrospective legislation). More to the point,
Ms. Kennett has not carried her burden of explaining at any point—either in her
supplemental authority or in her subsequent (unsuccessful) motion to certify, filed
after we issued our Jordan decision—why the Division’s recent actions should
affect our resolution of this case or, more specifically, cause us to deviate from
Jordan’s reasoning.
Therefore, we have no need to opine here—and thus do not—on the import
of the aforementioned recent actions of the Division. Suffice to say that Ms.
22
Kennett has not given us any basis for departing from our reasoning in Jordan,
and that reasoning dooms her effort to defend the district court’s judgment. 5
IV
In sum, we hold, pursuant to Jordan, that the companionship exemption in
effect during the time period relevant to this lawsuit covered companions
employed by third-party employers, like Bayada. As a result, Ms. Kennett and
her fellow class members are not entitled to overtime wages under Colorado law.
5
We recognize that the Division not only signaled through its notice
of proposed rulemaking that it planned to exclude companions who work for
third-party employers, like Bayada, from the coverage of the companionship
exemption, but—as alluded to above, see supra note 2—in 2020 it actually
completely removed the companionship-exemption language at issue here from
the applicable regulation. For similar reasons to those stated in text supra, based
on the arguments before us, we have no basis to conclude that there is anything
about this recent development that should cause us to depart from Jordan’s
reasoning in resolving this specific dispute.
23
We thus REVERSE the district court’s judgment and REMAND the case for
further proceedings consistent with this order and judgment. 6
ENTERED FOR THE COURT
Jerome A. Holmes
Circuit Judge
6
On May 9, 2019, amicus curiae Towards Justice filed a motion for
leave to file an amicus brief; Bayada opposed the motion. The motion was
“provisionally granted, subject to reconsideration by the panel of judges that will
be assigned later to consider this appeal on the merits.” See Order, No. 19-1004,
at 1, (10th Cir., filed May 21, 2019). Having carefully considered the matter, we
deny Towards Justice’s motion.
24
Michele Kennett v. Bayada Home Health Care, Inc., No. 19-1004,
Bacharach, J., concurring.
I agree with the majority that Jordan v. Maxim Healthcare Servs.,
Inc., 950 F.3d 724 (10th Cir. 2020) controls, 1 so I concur in the judgment.
But if we were interpreting the minimum wage order on a blank slate, I
would have adopted Ms. Kennett’s interpretation of the minimum wage
order.
The minimum wage order contains an exemption referring to three
jobs:
1. Companions,
2. Casual babysitters, and
3. Domestic employees.
7 Colo. Code Regs. § 1103–1:5 (2019).
1
In Jordan, the Court addressed three arguments that Bayada has not
made here:
1. Dictionaries and federal wage laws can shed insight on the
definitions of companions, casual babysitters, and domestic
employees.
2. The minimum wage order’s enabling statute could conflict with
the household modifier if we were to treat the household
modifier as applicable to companions.
3. The household modifier should not apply to companions or
casual babysitters; otherwise, those terms would be subsumed
by the term domestic employees.
950 F.3d at 735–44. I express no opinion on these arguments.
The parties agree that the last exempt job (domestic employees) is
narrowed by a restrictive clause: employed by households or families to
perform duties in private residences. 7 Colo. Code Regs. § 1103–1:5. The
primary issue is whether this restrictive clause also applies to the other two
jobs (companions and casual babysitters). Our opinion in Jordan answers
“no.” Jordan, 950 F.3d at 754. In my view, however, textual and
interpretive clues support application of the restrictive clause to all three
jobs.
Like the majority, I regard the minimum wage order as ambiguous.
Id. at 739. Despite this ambiguity, we must decide the better interpretation.
Ms. Kennett’s interpretation is supported by the grammatical structure of
the minimum wage order and the series–qualifier canon. The agency’s
contrary interpretations are nullities and unpersuasive.
1. The text of the minimum wage order suggests that companions are
exempt only when employed by households or families.
To interpret the minimum wage order, the parties agree that the
starting point is the text. In reviewing the text, we consider not only the
words but also their grammatical structure. See Earl T. Crawford, The
Construction of Statutes § 196, at 337 (1940) (“Since one may assume that
the legislature knew and understood the rules of grammar, such rules
should be considered by the courts in their efforts to ascertain the meaning
of a statutory enactment . . . .”).
2
The minimum wage order contains three groupings of exemptions.
The second grouping exempts
companions, casual babysitters, and domestic employees
employed by households or family members to perform duties in
private residences, property managers, interstate drivers, driver
helpers, loaders or mechanics of motor carriers, taxi cab drivers,
and bona fide volunteers.
7 Colo. Code Regs. § 1103–1:5 (emphasis added). We must interpret the
references to companions and casual babysitters. Bayada interprets these
references to include every companion and casual babysitter; Ms. Kennett
would limit these references to individuals “employed by households or
family members to perform duties in private residences.” Id.
Bayada’s interpretation would translate into this list of nine jobs:
1. companions,
2. casual babysitters,
and
3. domestic employees employed by households or family
members to perform duties in private residences,
4. property managers,
5. interstate drivers,
6. driver helpers,
7. loaders or mechanics of motor carriers,
8. taxi cab drivers,
and
3
9. bona fide volunteers.
Ms. Kennett’s interpretation would include these seven jobs:
1. companions, casual babysitters, and domestic employees
employed by households or family members to perform duties
in private residences,
2. property managers,
3. interstate drivers,
4. driver helpers,
5. loaders or mechanics of motor carriers,
6. taxi cab drivers,
and
7. bona fide volunteers.
The primary difference is that with Bayada’s interpretation, the first
and is inserted into the middle of the series. Insertion of the and would be
superfluous, which counsels against Bayada’s interpretation. See People v.
Cross, 127 P.3d 71, 73 (Colo. 2006) (“We . . . reject interpretations [of a
statute] that render words or phrases superfluous . . . .”); Denver Publ’g.
Co. v. Bd. of Cty. Comm'rs of Arapahoe Cty., 121 P.3d 190, 195 (Colo.
2005) (“We interpret every word [of a statute], rendering none superfluous
. . . .”) (quoting Sooper Credit Union v. Sholar Grp. Architects, P.C., 113
P.3d 768, 771 (Colo. 2005)).
We addressed a similar issue in Cruz v. United States, 106 F.2d 828
(10th Cir. 1939). There we interpreted a statutory series based on the
4
superfluity of the conjunction or. The statute criminalized deprivations of
benefits “by means of any fraud, force, threat, intimidation, or boycott, or
discrimination on account of race, religion, or political affiliations . . . .”
Id. at 829 (emphasis added). The defendants argued that the restrictive
clause (on account of race, religion, or political affiliations) referred to all
six of the nouns (fraud, force, threat, intimidation, or boycott, or
discrimination). Id.
The Court rejected this interpretation because the defendants’
interpretation would have injected a superfluous conjunction (or) into the
series:
1. fraud,
2. force,
3. threat,
4. intimidation,
or
5. boycott,
or
6. discrimination
on account of race, religion, or political affiliations. See id. To avoid
making the first or superfluous, the Court concluded that two categories
existed:
5
1. A deprivation of benefits by means of fraud, force, threat,
intimidation, or boycott
2. A deprivation of benefits by means of discrimination on
account of race, religion, or political affiliations
Id. at 829–30. The legislature “clearly indicated” its intent by inserting the
first or between intimidation and boycott. Id. at 830.
The same is true here. In our case, we’re dealing with a different
conjunction (and instead of or). But like the defendants in Cruz, Bayada is
embracing an interpretation that inserts a superfluous conjunction into the
middle of the series. In Ms. Kennett’s interpretation, the conjunction (and)
is not superfluous. It joins the three related jobs (companions, casual
babysitters, and domestic employees), so the modifying restrictive clause
(employed by households or family members to perform duties in private
residences) applies to each of the three jobs.
Bayada points to the other two groupings of exemptions in the
minimum wage order, suggesting that they also include superfluous ands. I
disagree.
The first grouping contains a list of jobs: “administrative,
executive/supervisor, professional, outside sales employees, and elected
officials and members of their staff.” 7 Colo. Code Regs. § 1103–1:5
(emphasis added). Bayada argues that one of the two ands must be
superfluous. I disagree. The first and shows that the whole list is
6
conjunctive, including five jobs; the second and shows that the fifth job
includes staff members for elected officials. So the list reads:
1. administrative,
2. executive/supervisor,
3. professional,
4. outside sales employees,
and
5. elected officials and members of their staff.
See id. Neither and is superfluous in this grouping of five jobs.
Nor is there a superfluous and in the final grouping of exempt
jobs:
Also exempt are: students employed by sororities, fraternities,
college clubs, or dormitories, and students employed in a work
experience study program and employees working in laundries
of charitable institutions which pay no wages to workers and
inmates, or patient workers who work in institutional laundries.
Id. (emphasis added).
The grouping is conjunctive, listing a series of jobs that are exempt.
But within the series, some of the items contain internal punctuation. For
example, the list includes “students employed by sororities, fraternities,
college clubs, or dormitories.” Id. (emphasis added). The or signifies that
students are exempt if they are employed by a sorority, fraternity, college
club, or dormitory. Id. The list also includes three other items: one of these
contains internal punctuation and two don’t. To show that all of these
7
employees are exempt, the final grouping must include at least one and.
But here the conjunction and appears three times.
We typically use only one conjunction to refer to multiple items in a
list. But when some of the items contain internal punctuation, the list can
create confusion. To avoid confusion, we can either use semicolons or
insert an and between each item in the list. See Chicago Manual of Style
6.18–6.19 at 312–13, 6.58 at 326 (16th ed. 2010) (discussing the serial
comma followed by and, and stating that semicolons can be used to aid
clarity when the series contains internal punctuation).
In this grouping of jobs, Ms. Kennett argues, the drafters chose the
second method. They inserted an and between each category to clarify the
existence of four separate categories of jobs (with some of these categories
containing internal series separated by the conjunction or). So Ms. Kennett
regards this grouping as a list of four categories of employees:
1. students employed by sororities, fraternities, college
clubs, or dormitories,
and
2. students employed in a work experience study program
and
3. employees working in laundries of charitable institutions
which pay no wages to workers
and
8
4. inmates, or patient workers who work in institutional
laundries.
See 7 Colo. Code Regs. § 1103–1:5. Without the three ands, the sentence
would be incomprehensible because of the two internal ors.
Bayada, on the other hand, would connect the job categories this
way:
1. students employed by sororities, fraternities, college clubs, or
dormitories,
and
2. students employed in a work experience study program
and
3. employees working in laundries of charitable institutions which
pay no wages to workers and inmates,
or
4. patient workers who work in institutional laundries.
See id.; Appellant’s Reply Br. at 5. This interpretation inexplicably
switches from and (to separate the first three jobs) to or (to separate only
the final job).
In any event, Bayada’s interpretation would supply a function for
each and in the final grouping. So even under Bayada’s interpretation,
there is no superfluous and in the final grouping.
** *
9
Ms. Kennett’s interpretation is supported by (1) the principles of
grammar and (2) the need to avoid superfluity and inconsistency. The text
itself indicates that the restrictive clause refers equally to companions,
casual babysitters, and domestic workers.
2. Canons of interpretation also suggest that companions are exempt
only when employed by households or families.
In interpreting the minimum wage order, the parties also rely on
canons of interpretation. Though the issue here involves interpretation of
an administrative order rather than a statute, Colorado courts apply the
same rules when interpreting administrative orders and statutes. See
Ferraro v. Frias Drywall, LLC, 451 P.3d 1255, 1260 (Colo. App. 2019).
Two conflicting canons could affect the meaning of restrictive
modifiers that follow the last noun in a series: (1) the series–qualifier
canon and (2) the last–antecedent rule. Ms. Kennett embraces the series–
qualifier canon, and Bayada embraces the last–antecedent rule.
Both canons apply to a series containing multiple parallel nouns.
Under the series–qualifier canon, the restrictive modifier ordinarily applies
to each of the nouns. Antonin Scalia & Bryan A. Garner, Reading Law: The
Interpretation of Legal Texts 147 (2012). Under the last–antecedent rule,
the restrictive clause modifies “only the noun . . . that it immediately
follows.” Barnhart v. Thomas, 540 U.S. 20, 26 (2003). So the two canons,
when applied to the same set of words, lead to two different conclusions.
10
A. The last–antecedent rule is not used in Colorado.
The two canons are treated differently in Colorado. The Colorado
legislature has expressly repudiated the last–antecedent rule. See Colo.
Rev. Stat. Ann. § 2–4–214 (“The general assembly hereby finds and
declares that” the last–antecedent rule “has not been adopted by the
general assembly and does not create any presumption of statutory
intent.”); see also People v. Rieger, 436 P.3d 610, 612 n.3 (Colo. App.
2019) (stating that the last–antecedent rule “was legislatively repudiated in
1981”). Given this statute, the Colorado Supreme Court has stated that the
last–antecedent rule “is no longer applicable in Colorado.” Danielson v.
Castle Meadows, Inc., 791 P.2d 1106, 1113 n.6 (Colo. 1990).
Bayada argues that the last–antecedent canon remains applicable in
Colorado, referring to four opinions: People v. Weeks, 369 P.3d 699 (Colo.
App. 2015); People ex. rel. O.C., 312 P.3d 226 (Colo. App. 2012); People
v. Trujillo, 251 P.3d 477 (Colo. App. 2010); and Chandler-McPhail v.
Duffey, 194 P.3d 434 (Colo. App. 2008). But none support the viability of
the last–antecedent rule to interpret statutes or administrative orders.
In Weeks and O.C., the court relied on legislative and legal history as
grounds for the decisions because the legislature had repudiated the last–
antecedent rule. Weeks, 369 P.3d at 711–12; O.C., 312 P.3d at 230–31. In
Trujillo, the court stated that the statutory repudiation of the last–
antecedent rule “does not create a contrary presumption” favoring the
11
series–qualifier canon. 251 P.3d at 480. The Trujillo court did not apply
the last–antecedent rule or suggest that it continued as a valid interpretive
tool in Colorado. Trujillo instead relied on legislative history and the need
to avoid statutory conflict. Id. at 480–81. And Chandler-McPhail involved
interpretation of a contract, not a statute or administrative order. 194 P.3d
at 437, 440–42.
Bayada also offers numerous examples of opinions in other
jurisdictions applying the last–antecedent canon. None involve the
interpretation of Colorado law.
B. Precedent and context support using the series–qualifier
canon here.
Though Colorado appellate courts have declined to apply the last–
antecedent rule, they have routinely applied the series–qualifier canon.
E.g., In re Estate of David v. Snelson, 776 P.2d 813, 818 (Colo. 1989)
(noting that the General Assembly had “expressly repudiated” the last–
antecedent rule and choosing instead to apply the series–qualifier canon);
People v. Lovato, 357 P.3d 212, 221 (Colo. App. 2014) (applying the
series–qualifier canon).
The majority posits that under Colorado law, a canon of construction
“is merely an interpretive aid, not an absolute rule” and its application
depends on “context.” See Benefield v. Colo. Republican Party, 329 P.3d
262, 267 (Colo. 2014). I agree. For the series–qualifier canon, the context
12
turns on whether the modifying clause would sensibly modify each
preceding noun. The Supreme Court recently illustrated a contextual clue
triggering the series–qualifier canon, using a classic example: “the laws,
the treaties, and the constitution of the United States”:
A reader intuitively applies “of the United States” to “the laws,”
“the treaties” and “the constitution” because (among other
things) laws, treaties, and the constitution are often cited
together, because readers are used to seeing “of the United
States” modify each of them, and because the listed items are
simple and parallel without unexpected internal modifiers or
structure.
Lockhart v. United States, 136 S. Ct. 958, 963 (2016).
The same is true here. The parallel series (companions, casual
babysitters, and domestic employees) contains three categories of caretaker
jobs. Caretakers in each category can be employed by families or
households to work in a home. So all three nouns (companions, casual
babysitters, and domestic employees) can be restricted through the
modifying clause that follows the series: “employed by households or
family members to perform duties in private residences.” 7 Colo. Code
Regs. § 1103–1:5. Given the applicability of the modifying clause to each
noun, the series–qualifier canon would attach the restrictive modifier to
each of the three nouns, applying the exemption to
• companions employed by households or family members to
perform duties in private residences,
• casual babysitters employed by households or family members
to perform duties in private residences, and
13
• domestic employees employed by households or family
members to perform duties in private residences.
See id.
Under the series–qualifier canon, companions hired by a third–party
employer (like Bayada) would not fall within the exemption.
3. The agency’s interpretations were unpersuasive and are now
rescinded.
Bayada also relies on an advisory bulletin and opinion letter. These
documents were unpersuasive and conflicted with each other, and the
agency ultimately rescinded both documents. So we should not base our
decision on these documents.
A. Both the advisory bulletin and the opinion letter are
nullities.
Bayada argues that the agency’s 2012 advisory bulletin shows the
drafter’s intent underlying the minimum wage order. The advisory bulletin
states that companions and casual babysitters are exempt. Appellant’s
App’x vol. 1, at 158a. Given this statement, Bayada contends that the
advisory bulletin exempts all companions irrespective of whether they were
employed by a family or household.
But in 2019, the agency stated that the advisory bulletin “is no longer
operative as reflecting Division opinions, interpretations, or rulings.”
Notice Regarding Opinion/Exemption Letters at 2 (June 24, 2019). Bayada
argues that this language operates prospectively, allowing the advisory
14
bulletin to continue supplying retroactive guidance for past minimum wage
orders. I disagree. The agency stated that the advisory bulletin had
“properly noted that it was . . . not ‘legal advice,’ ‘and should not be relied
upon as an official record of action or law.’” Id. This statement does not
permit continued use of past advisory bulletins for any purpose. See
MGPC, Inc. v. Dep’t of Energy, 763 F.2d 422, 431 (Temp. Emer. Ct. App.
1985) (interpreting rescission of an order as a nullity prohibiting any effect
even in the past).
Bayada also relies on a 2006 opinion letter from the agency,
supporting its position on the companion exemption. The letter
• stated that the agency intended the companion exemption to
“mirror” federal law, which then exempted companions with
third-party employers (see 29 C.F.R. § 552.109(a) (2006)), and
• approved “the practice of applying the companionship
exemption in situations involving third party employers.”
Appellant’s App’x vol. 1, at 138a–39a. But in 2019, the agency stated that
its previous opinion letters “were never binding” and “are now rescinded.”
Notice Regarding Opinion/Exemption Letters at 1 (June 24, 2019).
Despite the language in the notice, Bayada urges continued
application of the 2006 opinion. I disagree. The agency stated that
• opinion letters had been issued ex parte, with input from only
one party,
• the opinion letters “were not binding determinations with legal
effect,” and
15
• “all such letters are now rescinded.”
Notice Regarding Opinion/Exemption Letters at 1 (June 24, 2019). The
agency made clear that its opinion letters were never binding because they
had been issued ex parte and lacked any legal effect. Because the 2006
opinion letter has been rescinded, it is also a nullity.
B. When the advisory bulletin and opinion letter were in effect,
they were meaningful only to the extent that they were
persuasive.
Colorado law normally affords great deference to an agency’s
reasonable interpretation of its ambiguous order. Sierra Club v.
Billingsley, 166 P.3d 309, 312 (Colo. App. 2007); see also Citizens for
Clean Air & Water v. Colo. Dep’t of Pub. Health & Env’t, 181 P.3d 393,
396–97 (Colo. App. 2008) (“An agency’s interpretation of its own
regulations . . . is entitled to deference and must be accepted if there is a
reasonable basis in the law.”). But Colorado law, like federal law,
recognizes that an interpretation in the form of an advisory bulletin or
opinion letter lacks the force of law, so it is “‘entitled to respect,’ but only
to the extent that [the interpretation has] the ‘power to persuade.’”
Brunson v. Colo. Cab Co., LLC, 433 P.3d 93, 96 (Colo. App. 2018)
(quoting Christensen v. Harris Cty., 529 U.S. 576, 587 (2000)). The
persuasiveness of an agency interpretation “is derived in part from the
‘thoroughness evident in its consideration’ and from its reflection of a
16
body of experience and informed judgment.” Brunson, 433 P.3d at 99
(quoting Skidmore v. Swift & Co., 323 U.S. 134, 140 (1944)).
C. The 2012 advisory bulletin was unpersuasive when it was in
effect.
Even before rescission of the advisory bulletin, it was unpersuasive.
The advisory bulletin discussed companions and casual babysitters but did
not define domestic employees as exempt or non-exempt. Without defining
domestic employees as a comparator, a court could not fairly infer whether
the bulletin referred to companions and babysitters when employed by
third parties.
The agency’s interpretation in the advisory bulletin also conflicted
with a 2006 opinion letter defining companions and companionship
services. For example, the opinion letter stated that companions spent no
more than 20% of their weekly workloads on general household work.
Appellant’s App’x vol. 1, at 138a. By contrast, the advisory bulletin
described companionship services without mentioning any such limitation
on general household work. Appellant’s App’x vol. 1, at 158a. This
inconsistency undercut the persuasive value of both the advisory bulletin
and the opinion letter. See Brunson, 433 P.3d at 99 (“[T]o be persuasive,
an agency interpretation must be consistent and contemporaneous with
other pronouncements of the agency . . . .”).
17
D. The 2006 opinion letter was unpersuasive when it was in
effect.
Even if the 2006 opinion letter were not a nullity, it would also be
unpersuasive. The opinion letter not only conflicted with the advisory
bulletin but also displayed a lack of thoroughness. The opinion letter
didn’t address the language of the minimum wage order, canons of
interpretation, case law, or regulatory history. Instead, the opinion letter
relied solely on an assertion that the companion exemption “mirror[ed]”
federal law. But it’s impossible to read the companion exemption as a
mirror image of contemporaneous federal law.
In 2006, federal law provided that
• companions employed by a family or household were not
exempt and
• companions employed by a third party were exempt. 2
See 29 C.F.R. § 552.109(a) (2006). But the minimum wage order cannot be
read this way. There were only two reasonable interpretations:
1. All companions are exempt (Bayada’s interpretation).
2. Only companions employed by households or families are
exempt (Ms. Kennett’s interpretation).
7 C.C.R. § 1103–1:5 (“Other exemptions are: companions, casual
babysitters, and domestic employees employed by households or family
2
Federal law changed in early 2015. Compare 29 C.F.R. § 552.109
(2014), with 29 C.F.R. § 552.109 (2015).
18
members to perform duties in private residences . . . . (emphasis added)). It
is impossible to read the minimum wage order as limiting the companion
exemption to companions employed by third parties. So the minimum wage
order did not “mirror” federal law. 3
The opinion letter supplied no other basis for its conclusion, and the
reasoning would be wrong even under Bayada’s interpretation of the
minimum wage order. So even if the opinion letter hadn’t been rescinded,
it would have lacked persuasive value.
4. Conclusion
Based on the arguments in this case, I would have avoided
superfluous language, applied the series–qualifier canon, and declined to
defer to the agency’s obsolete and unpersuasive interpretation of the
minimum wage order. To avoid superfluous language and apply a viable
canon of interpretation, Ms. Kennett properly restricts the exemption for
companions to those “employed by households or family members to
perform duties in private residences.” 7 Colo. Code Regs. § 1103–1:5. But
Jordan controls, so I join my colleagues in the disposition.
3
Bayada and the majority also rely on various enforcement letters,
which similarly state that the companion exemption was intended to
“mirror” or “mimic” federal law. See Appellants’ App’x vol. 3, at 584a,
588a. These enforcement letters share the conclusory, unpersuasive
reasoning of the 2006 opinion letter.
19