DC Appleseed Center for Law and Justice, Inc. v. DC Department of Insurance, Securities, and Banking

Notice: This opinion is subject to formal revision before publication in the
Atlantic and Maryland Reporters. Users are requested to notify the Clerk of the
Court of any formal errors so that corrections may be made before the bound
volumes go to press.

DISTRICT OF COLUMBIA COURT OF APPEALS

Nos. 16-AA-895, 16-AA-967, and 18-AA-178

DC APPLESEED CENTER FOR LAW AND JUSTICE, INC., ET AL.,
PETITIONERS/INTERVENORS,

v.

DISTRICT OF COLUMBIA DEPARTMENT OF INSURANCE, SECURITIES AND BANKING,
RESPONDENT.

On Petitions for Review of Orders of the
District of Columbia Department of Insurance, Securities and Banking
(MIE-19-14 and MIE-27-14)

(Argued April 17, 2019 Decided August 29, 2019)

Walter Smith, with whom Richard B. Herzog, Marialuisa Gallozzi, Beth
Brinkmann, and Bradley K. Ervin were on the brief, for petitioner-intervenor DC
Appleseed Center for Law and Justice, Inc.

Lisa H. Schertler, with whom David Schertler and Danny C. Onorato were
on the brief, for petitioner-intervenor Group Hospitalization and Medical Services,
Inc.

James C. McKay, Jr., with whom Karl A. Racine, Attorney General for the
District of Columbia, Loren L. AliKhan, Solicitor General, and Caroline Van Zile,
Deputy Solicitor General, were on the brief, for respondent.

Michelle S. Kallen, with whom Toby J. Heytens, Solicitor General for the
Commonwealth of Virginia, and Matthew R. McGuire, Principal Deputy Solicitor
General at the time the brief was filed, were on the brief, for intervenor Mark
Herring, Attorney General of the Commonwealth of Virginia.
 2

Gary Thompson was on the brief for amicus curiae Mary M. Cheh, District of
Columbia Councilmember, in support of petitioner-intervenor DC Appleseed Center
for Law and Justice, Inc.

Before GLICKMAN and MCLEESE, Associate Judges, and WASHINGTON, Senior
Judge.

MCLEESE, Associate Judge: Petitioners-intervenors DC Appleseed Center for

Law & Justice, Inc. (Appleseed) and Group Hospitalization and Medical Services,

Inc. (GHMSI) seek review of orders of respondent, the District of Columbia

Department of Insurance, Securities and Banking (DISB), determining that

GHMSI’s 2011 surplus was excessive, that the excess surplus attributable to the

District was approximately $50 million, and that GHMSI was required to distribute

its excess surplus in the form of rebates to eligible subscribers of GHMSI. We

remand for further proceedings.

I. Factual and Procedural History

This matter has previously been before this court. D.C. Appleseed Ctr. for

Law & Justice, Inc. v. District of Columbia Dep’t of Ins., Sec., & Banking (Appleseed

I), 54 A.3d 1188 (D.C. 2012). The following background material is taken in

significant part from our prior opinion, supplemented and revised as necessary to

reflect subsequent developments.
 3

A. GHMSI

GHMSI is the successor to Group Hospitalization, Inc., a nonprofit

organization created in 1939 by congressional charter to provide health-care services

and medical insurance. Pub. L. No. 76-395, §§ 3, 8, 53 Stat. 1412, 1413-14 (1939);

Pub. L. No. 98-493, § 1, 98 Stat. 2272 (1984). Organized as a “charitable and

benevolent institution,” GHMSI “shall be conducted for the benefit of [its] certificate

holders.” Pub. L. No. 76-395, §§ 3, 8, 53 Stat. 1413-14. GHMSI conducts business

in the District, Maryland, and Virginia.

Although GHMSI initially was not subject to the statutes regulating the

business of insurance in the District, Congress amended GHMSI’s charter in 1993

to domicile GHMSI in the District and place GHMSI under the District’s regulatory

authority. Pub. L. No. 103-127, § 138, 107 Stat. 1336, 1349 (1993). Such regulatory

authority includes review and approval of GHMSI’s proposed health-insurance

premium rates. D.C. Code § 31-3311.01 et seq. (2012 Repl.). D.C. law also requires

GHMSI to maintain certain risk-based capital levels and to report those levels on an

annual basis to the DISB Commissioner. D.C. Code § 31-3451.01 et seq. (2019

Supp.).
 4

In 1998, GHMSI affiliated with CareFirst of Maryland, Inc. GHMSI and

CareFirst of Maryland jointly own CareFirst BlueChoice, Inc. GHMSI is a licensee

of the Blue Cross Blue Shield Association (BCBSA).

B. The Medical Insurance Empowerment Amendment Act

In 2009, the Council of the District of Columbia enacted the Medical

Insurance Empowerment Amendment Act (MIEAA). D.C. Law 17-369, 56 D.C.

Reg. 1346 (Feb. 13, 2009) (codified as amended at D.C. Code § 31-3501 et seq.

(2012 Repl. & 2019 Supp.)). The MIEAA authorizes the Commissioner of DISB to

determine whether a medical-services corporation’s surplus is “excessive” and to

order that any excess surplus be reinvested in “community health.” D.C. Code

§ 31-3506(e), (g)(1) (2019 Supp.). Specifically, the MIEAA and subsequent

amendments added a new subsection to D.C. Code § 31-3506, which now states in

relevant part:

The Commissioner may, on an annual basis, and shall, on
a basis no less frequently than every 3 years, review the
portion of the surplus of the corporation that is attributable
to the District and may issue a determination as to whether
the surplus is excessive. Any such review shall be
undertaken in coordination with the other jurisdictions in
which the corporation conducts business. The surplus may
be considered excessive only if:
 5

(1) The surplus is greater than the appropriate risk-based
capital requirements as determined by the Commissioner
for the immediately preceding calendar year; and

(2) After a hearing, the Commissioner determines that the
surplus is unreasonably large and inconsistent with the
corporation’s obligation under § 31-3505.01.

D.C. Code § 31-3506(e); see also D.C. Law 18-104, 56 D.C. Reg. 9182 (Dec. 4,

2009); D.C. Law 19-171, 59 D.C. Reg. 6190 (June 1, 2012).

D.C. Code § 31-3505.01 requires GHMSI and similar entities to “engage in

community health reinvestment to the maximum feasible extent consistent with

financial soundness and efficiency.” The MIEAA also provides that “[i]n

implementing the provisions of the [MIEAA], the Commissioner shall consider the

interests and needs of the jurisdictions in the corporation’s service area.” D.C. Code

§ 31-3506.01(b) (2012 Repl.).

C. Surplus Requirements and Excess Surplus Determination

GHMSI is required to maintain a surplus of capital to cover its projected risk,

development costs, and growth. D.C. Code § 31-3451.01 et seq. The National

Association of Insurance Commissioners (NAIC) has developed widely accepted

risk-based capital (RBC) formulae to determine the minimum amount of capital an
 6

insurer should hold to support its business operations. The baseline figure in the

RBC formula is the authorized control level (ACL), which is a reference value that

accounts for the insurer’s size, structure, and volume of risk. The District of

Columbia has adopted statutory minimum requirements for insurance companies’

surplus levels, expressed as an RBC-ACL ratio. D.C. Code §§ 31-2001 to -2013

(2012 Repl. & 2019 Supp.).

If a health insurer’s RBC-ACL ratio falls under certain statutory levels,

company or regulatory action is authorized or required. D.C. Code §§ 31-3451.01

to .06. For example, if an insurer’s surplus falls below 200% RBC-ACL, the insurer

must submit a plan to the Commissioner identifying the conditions that led to that

event and proposing corrective actions to bring the surplus up to a safer level. D.C.

Code §§ 31-3451.01(6), .03. If an insurer’s surplus falls to lower RBC-ACL levels,

the Commissioner is authorized or obligated take increasingly corrective actions.

D.C. Code §§ 31-3451.04 to .06. Additionally, as a licensee of BCBSA, GHMSI is

subject to contractual RBC-ACL standards set by BCBSA. Specifically, early-

warning monitoring is triggered if a BCBSA licensee’s surplus falls below 375%

RBC-ACL, and the licensee must take certain corrective actions. The Commissioner

is required to consider NAIC’s RBC requirements and BCBSA’s capital
 7

requirements in determining whether a medical-services corporation’s surplus is

excessive. 26A DCMR § 4601.4 (2019).

Regulations promulgated by DISB in 2009 establish further procedures for the

Commissioner’s review of medical-services corporations’ surplus. 26A DCMR

§ 4600 et seq. (2019). The regulations require medical-services corporations such

as GHMSI to file an annual financial report with DISB detailing “the company’s

surplus and examin[ing] whether the company’s surplus is considered excessive

under the [MIEAA].” 26A DCMR § 4601.1. If the Commissioner preliminarily

determines that the company’s surplus is excessive, the regulations require a public

hearing “to determine whether the company’s surplus is excessive and unreasonably

large.” 26A DCMR § 4601.5. If the Commissioner makes a final determination that

a company’s surplus that is attributable to the District is excessive, the

Commissioner must order the company to submit a plan for dedication of the excess

surplus to community-health reinvestment. D.C. Code § 31-3506(g)(1); 26A

DCMR § 4603.1.
 8

D. DISB’s 2008 Surplus Determination and Appleseed I Decision

In 2010, the Commissioner concluded that GHMSI’s 2008 year-end surplus

was neither unreasonably large nor excessive. Appleseed, which is a consumer of

health insurance, a subscriber of GHMSI, and an organization with goals including

improving access to healthcare in the District, challenged the Commissioner’s

ruling. Appleseed I, 54 A.3d at 1201, 1210. On review, this court held, in pertinent

part, that the Commissioner’s analysis of GHMSI’s 2008 surplus was incomplete,

because that analysis considered whether GHMSI’s surplus was unreasonably large

without considering whether the surplus was inconsistent with GHMSI’s statutory

obligation to engage in community-health reinvestment. Id. at 1212-15. We

emphasized that the Commissioner, when deciding whether an insurer’s surplus is

excessive, must keep in mind both statutory goals: (A) the insurer’s financial

soundness and (B) maximization of community-health reinvestment. Id. We also

concluded that the Commissioner’s order did not adequately explain the

Commissioner’s findings and conclusions. Id. at 1218-19. We remanded the case

for further proceedings, including a redetermination as to whether GHMSI’s surplus

was excessive, with more complete explanation of the Commissioner’s reasoning.

Id. at 1220-21.
 9

E. Subsequent DISB Orders and Procedural History

On remand, the Commissioner determined that further review of the 2008

surplus would be moot and began the review process for GHMSI’s 2011 year-end

surplus, the most recent year for which surplus information was available at the time.

DISB retained an actuarial firm and a financial-analysis firm to assist in the surplus

review. GHMSI and Appleseed also retained actuarial consultants. As part of the

review process, Appleseed, GHMSI, and their consultants met with DISB’s

consultants and submitted written reports and briefs on whether GHMSI’s 2011

surplus was excessive and what portion of any excess surplus was attributable to the

District. In June 2014, DISB held a public hearing. After the hearing, the

Commissioner received statements from the Maryland Insurance Commissioner and

the Virginia State Corporation Commission’s Bureau of Insurance.

In December 2014, the Commissioner issued a decision and order with respect

to GHMSI’s 2011 surplus. The order concluded that (1) GHMSI’s appropriate level

of surplus as of the end of 2011 was 721% RBC-ACL (approximately $695.9

million); (2) GHMSI’s actual surplus at the end of 2011 was 998% RBC-ACL

(approximately $963.6 million), meaning that GHMSI’s excess surplus was

approximately $267.7 million; and (3) 21% of GHMSI’s excess surplus was
 10

attributable to the District. GHMSI took the position that it had no excess surplus

attributable to the District and had already undertaken community-health

reinvestment beyond what the Commissioner’s order required. GHMSI therefore

did not submit a plan for additional distribution of excess surplus. The

Commissioner disagreed and directed GHMSI to disburse the excess surplus

attributable to the District (approximately $50 million) by issuing rebates to eligible

subscribers.

II. Analysis

The parties raise numerous challenges to the Commissioner’s ruling. Because

we find some of those challenges persuasive, we remand for further proceedings.

A. Standard of Review

On questions of statutory interpretation, we first look to see whether the

statutory language at issue is “plain and admits of no more than one meaning.”

Peoples Drug Stores, Inc. v. District of Columbia, 470 A.2d 751, 753 (D.C. 1983)

(en banc) (internal quotation marks omitted). We will give effect to the plain

meaning of a statute “when the language is unambiguous and does not produce an
 11

absurd result.” McNeely v. United States, 874 A.2d 371, 387 (D.C. 2005) (internal

quotation marks omitted). “We may also look to the legislative history to ensure

that our interpretation is consistent with legislative intent.” Thomas v. Buckley, 176

A.3d 1277, 1281 (D.C. 2017) (internal quotation marks omitted). If a statute is

ambiguous, “we will defer to an agency’s reasonable interpretation of the statute it

administers.” Appleseed I, 54 A.3d at 1211.

Under the D.C. Administrative Procedure Act, this court will uphold an

agency’s decision unless the decision is arbitrary, capricious, an abuse of discretion,

or otherwise not in accordance with the law. D.C. Code § 2-510(a)(3)(A) (2016

Repl.). “The Commissioner must make factual findings on all material contested

issues, the findings must be supported by substantial evidence on the record, and the

conclusions must flow rationally from the findings.” Appleseed I, 54 A.3d at 1216.

The Commissioner’s order must also state the basis of its ruling in sufficient detail

and be fully and clearly explained, so as to allow “for meaningful judicial review of

and deference to the agency’s decision.” Id. Although agency determinations based

upon highly complex and technical matters are entitled to great deference, “[t]he

more technical and complex the subject matter, the more explanation the agency

ought to provide for its decision.” Id. at 1217.
 12

B. 2015 Congressional Amendments

In 2015, Congress amended GHMSI’s charter to provide that GHMSI “shall

not divide, attribute, distribute, or reduce its surplus pursuant to any statute,

regulation, or order of any jurisdiction without the express agreement of the District

of Columbia, Maryland, and Virginia—(1) that the entire surplus of the corporation

is excessive; and (2) to any plan for reduction or distribution of surplus.” Pub. L.

No. 114-113, § 747(a), 129 Stat. 2242, 2486 (2015). Relying on that amendment,

Virginia appears to argue that the Commissioner was flatly precluded from requiring

GHMSI to distribute or reduce its surplus without the express agreement of

Maryland and Virginia. By its plain language, however, the 2015 amendment does

not apply to the present dispute about GHMSI’s 2011 surplus. The 2015 amendment

“shall apply with respect to the surplus of [GHMSI] for any year after 2011.”

§ 747(b), 129 Stat. at 2486 (emphasis added). It is true that requiring GHMSI to

provide a rebate relating to its 2011 surplus would likely have an impact on the size

of GHMSI’s surplus in years after 2011. But Congress was no doubt aware of that

fact and nevertheless made the 2015 amendments inapplicable to distribution of the

2011 surplus. We have no basis to look behind Congress’s clearly expressed

decision. See CareFirst, Inc. v. Taylor, 235 F. Supp. 3d 724, 744-45 (D. Md. 2017)
 13

(plain language of 2015 amendment required conclusion that 2015 congressional

amendment did not apply to determination of GHMSI’s 2011 surplus).

C. Coordination

As previously noted, D.C. Code § 31-3506(e) authorizes the Commissioner to

“review the portion of the surplus of [GHMSI] that is attributable to the District and

[to] issue a determination as to whether the surplus is excessive.” Section 31-

3506(e) further requires that this review “be undertaken in coordination with the

other jurisdictions in which [GHMSI] conducts business.” GHMSI and Virginia

argue that the Commissioner failed to adequately coordinate the review of GHMSI’s

surplus with Virginia and Maryland. We agree. We conclude, however, that in some

respects this issue was not properly presented to the Commissioner, which affects

the scope of relief we grant.

1. Procedural background

The Commissioner’s review of GHMSI’s 2011 surplus began in 2012. During

the period up to the June 2014 hearing, the Commissioner and DISB staff apparently

communicated with Maryland and Virginia insurance commissioners and their staff,
 14

including advising them of the surplus-review hearing and soliciting their

participation. As far as we have been able to determine, at no point during that

period did either Virginia or Maryland object that the Commissioner was failing to

adequately coordinate with them. After the hearing, both Virginia and Maryland

filed brief statements that raised no objection about lack of coordination and that

thanked DISB for the opportunity to comment.

During the period up to the June 2014 hearing, GHMSI did encourage DISB

in general terms to coordinate with Maryland and Virginia to help ensure that

GHMSI would not be subject to conflicting obligations. GHMSI did not, however,

raise any specific objection to the manner in which DISB was coordinating with

Virginia and Maryland. At the hearing, a witness for GHMSI advocated

coordination through “direct communication” but did not assert that such

communication had been lacking.

It appears that the first concrete objection on the ground of inadequate

coordination was not raised until after the December 2014 order, over two years after

the proceeding began. Specifically, GHMSI for the first time argued that the

Commissioner should reopen the proceedings and invite Maryland and Virginia to

participate in a consolidated proceeding. Even after December 2014, the Virginia
 15

Bureau of Insurance did not object to a lack of coordination, instead issuing a report

stating that it recommended that Virginia take a more active role in future

proceedings.

2. Analysis

We hold that any objection to the Commissioner’s failure to adequately

coordinate with Virginia and Maryland before the December 2014 order was not

properly presented. “In the absence of exceptional circumstances, a reviewing court

will refuse to consider contentions not presented before the administrative agency at

the appropriate time.” District of Columbia Hous. Auth. v. District of Columbia

Office of Human Rights, 881 A.2d 600, 611 (D.C. 2005) (internal quotation marks

omitted). “One principal reason for the rule that procedural objections must be

timely made is to give the tribunal and opposing parties the opportunity to correct or

controvert the purported defect when it is still possible to do so.” Id. We see no

justification for requiring the Commissioner to begin this proceeding anew based on

an objection that was not raised until over two years into the proceeding and that

could have been raised far earlier. Cf., e.g., District of Columbia Gen. Hosp. v.

District of Columbia Office of Emp. Appeals, 548 A.2d 70, 75 (D.C. 1988) (refusing

to permit party to belatedly challenge substitution of hearing examiner; “A contrary
 16

rule . . . would only countenance and encourage unacceptable inefficiency in the

administrative process.”) (internal quotation marks omitted).

On the other hand, GHMSI did eventually object on the ground of inadequate

coordination, and that objection sufficed to preserve the issue with respect to

subsequent proceedings. Moreover, because we are remanding on other issues, the

obligation to coordinate is relevant to proceedings on remand. We therefore must

address on the merits the nature of the Commissioner’s obligation to coordinate.

As previously noted, § 31-3506(e) requires that the Commissioner’s review

of GHMSI’s surplus “be undertaken in coordination with the other jurisdictions in

which [GHMSI] conducts business.” The purposes of that provision seem obvious

and do not appear to be disputed: (1) to ensure that the Commissioner gets the

benefit of the knowledge and experience of the other jurisdictions that are regulating

GHMSI; (2) to ensure that the Commissioner takes into account the needs and

interests of subscribers and regulators in the other jurisdictions; and (3) to try to

reduce the extent to which GHMSI is subjected to conflicting regulations in different

jurisdictions. For several reasons, we agree with GHMSI and Virginia that

coordination requires more than simply soliciting and considering input from

Virginia and Maryland.
 17

First, we place substantial weight on the dictionary definitions of

“coordination” and the corresponding verb “coordinate.” Cf. generally, e.g.,

O’Rourke v. District of Columbia Police & Firefighters’ Ret. & Relief Bd., 46 A.3d

378, 383 (D.C. 2012) (“The first step in construing a statute is to read the language

of the statute and construe its words according to their ordinary sense and plain

meaning.”) (internal quotation marks omitted). Those definitions consistently

contemplate more than mere consultation. See, e.g., Webster’s Third New

International Dictionary 501 (2012) (defining “coordinate,” inter alia, as to

“regulate and combine in harmonious action”); Oxford Dictionary of English 384

(3d ed. 2010) (defining “coordinate” as to “bring the different elements of (a

complex activity or organization) into a harmonious or efficient relationship” or to

“negotiate with others in order to work together effectively”).

Second, other courts construing the terms “coordinate” and “coordination”

have understood them to mean more than consultation. See Cal. Native Plant Soc’y

v. City of Rancho Cordova, 91 Cal. Rptr. 3d 571, 602 (Ct. App. 2009)

(“‘[C]oordination’ implies some measure of cooperation that is not achieved by

merely asking for and considering input or trying to work together.”); In re Water

Use Permit Applications, 9 P.3d 409, 501 (Haw. 2000) (requirement of
 18

“coordination” “is a call for cooperation and mutual accommodation”; “The

objectives of the Commission and the counties will not always converge. To the

extent that their respective functions and duties permit, however, the Commission

and counties should be seeking common ground.”). We are aware of one decision

that points in the opposite direction, but that decision does so without analysis and

we do not find that decision persuasive, particularly given the other considerations

present in the context of the MIEAA. Bienz v. City of Dayton, 566 P.2d 904, 916-

917 (Or. Ct. App. 1977) (city complied with coordination requirement “by actively

seeking and considering comments and recommendations from various affected

state and local agencies”).

Third, the coordination requirement was added to the MIEAA in a 2009

amendment. D.C. Law 18-104, § 2(c)(1), 56 D.C. Reg. 9182, 9184 (Dec. 4, 2009).

At the time of that amendment, the MIEAA already required that, “[i]n

implementing the provisions of the [MIEAA], the Commissioner shall consider the

interests and needs of the jurisdictions in the corporation’s service area.” D.C. Code

§ 31-3506.01(b) (2009 Repl.). The coordination requirement thus would ordinarily

be read to impose obligations that run beyond mere consideration of interests. See

generally, e.g., Stevens v. District of Columbia Dep’t of Health, 150 A.3d 307, 315-

16 (D.C. 2016) (“One of the most basic interpretive canons is that a statute should
 19

be construed so that effect is given to all its provisions, so that no part will be

inoperative or superfluous, void or insignificant.”) (brackets and internal quotation

marks omitted).

Fourth, as noted, the coordination requirement is obviously intended to serve

the purpose of minimizing the extent to which GHMSI and similar entities are

subjected to conflicting regulation in different jurisdictions. The importance of

avoiding such conflicting regulation if possible is widely recognized. See, e.g., CTS

Corp. v. Dynamics Corp. of Am., 481 U.S. 69, 88-89 (1987) (noting adverse effects

that may arise from subjecting activities to inconsistent regulation). Any reasonable

interpretation of the coordination requirement must give appropriate weight to that

consideration.

We acknowledge our obligation to defer to the Commissioner’s reasonable

interpretations of statutory provisions that the Commissioner administers.

Appleseed I, 54 A.3d at 1211. For the foregoing reasons, however, we conclude that

the coordination requirement cannot reasonably be understood to require mere

consultation. See Cal. Native Plant Soc’y, 91 Cal. Rptr. 3d at 602-03 (rejecting as

unreasonable interpretation of “coordination” to mean “consultation”). It remains,

though, to determine what more is required.
 20

GHMSI and Virginia initially appeared to contend that the coordination

requirement should be construed to preclude the Commissioner from acting without

the agreement of both Virginia and Maryland. Both GHMSI and Virginia disavowed

that position at oral argument, however. We agree that coordination does not require

unanimous agreement. See Cal. Native Plant Soc’y, 91 Cal. Rptr. 3d at 603 (“[W]e

do not read this ‘coordination’ requirement as requiring the City to subordinate itself

to state and federal agencies by implementing their comments and taking their

direction.”) (brackets and internal quotation marks omitted).

DISB argues that the Commissioner’s obligation is to coordinate only with

respect to the process of review, not with respect to any of the actual determinations

as to whether GHMSI’s surplus is excessive or what portion of the excess surplus is

attributable to the District. We disagree. As a textual matter, it seems strained at

best to treat the Commissioner’s determinations as distinct from, rather than the

culmination of, the Commissioner’s review. As a practical matter, reading the

Commissioner’s coordination obligation so narrowly would frustrate the obvious

purpose of requiring coordination. See generally, e.g., Rouse v. United States, 391

A.2d 790, 791 (D.C. 1978) (“As between an interpretation that will effectuate the
 21

obvious intent of our statute and one that will largely frustrate that intent, we

unhesitating[ly] adopt the former.”) (brackets and internal quotation marks omitted).

On a related topic, DISB argues that the Commissioner’s obligation to

“consider the interests and needs of the jurisdictions in [GHMSI]’s service area”

applies only to determining how to reinvest excess surplus. The plain language of

the applicable provision makes clear, however, that this requirement applies to all of

the Commissioner’s decisions “[i]n implementing the provisions of the [MIEAA].”

D.C. Code § 31-3506.01(b).

We turn to what the coordination requirement more concretely demands in the

current setting. In our view, the coordination requirement obliges the Commissioner

to try to work together with Maryland and Virginia with an eye towards agreement

if that is feasible and permissible under applicable law. That may include, but is not

necessarily limited to: (a) inviting the corresponding regulators in Maryland and

Virginia to participate in a joint proceeding; (b) soliciting on-the-record input from

the corresponding regulators, including where appropriate by directing specific

questions to them; (c) giving appropriate weight to the important interest in, to the

extent reasonably possible, avoiding inconsistent regulation of GHMSI; and (d) if

the Commissioner reaches conclusions that differ from those reached by
 22

corresponding regulators, explaining why (i) principles of District law require that

result or (ii) other important considerations outweigh the interest in uniform

regulation.

We note, however, that it takes at least two to coordinate. Neither Virginia

nor Maryland imposes a statutory obligation on its regulators to coordinate (or even

consult) with the District in the regulation of GHMSI. To the contrary, Virginia and

Maryland have each enacted a statute prohibiting GHMSI from complying with

DISB orders requiring reduction or distribution of its surplus without the respective

approval of Virginia or Maryland. Va. Code Ann. § 38.2-4229.2(D) (West 2019);

Md. Code Ann., Ins. § 14-124(a)(6) (West 2019). Each also issued an administrative

order to the same effect. Relatedly, with respect to GHMSI’s stated interest in

avoiding inconsistent regulation, we note that GHMSI specifically asked Virginia

and Maryland to issue those administrative orders, which directly conflict with the

Commissioner’s orders. The Commissioner may appropriately take these

circumstances into account in determining the weight to be given to the interest in

uniform regulation of GHMSI.

We flag a remaining issue for the Commissioner to consider on remand. DISB

suggests in its briefs and suggested at oral argument that some of the
 23

Commissioner’s consultations with Maryland and Virginia were not placed on the

record. In general, formal agency adjudications such as this must be based on

matters of record. See, e.g., Fair Care Found. v. District of Columbia Dep’t of Ins.

& Sec. Regulation, 716 A.2d 987, 996 (D.C. 1998) (“An agency is required to

maintain an official record in every contested case, and is prohibited from issuing

any decision or order in such a case except upon consideration of such exclusive

record.”) (footnote and internal quotation marks omitted). On remand, the

Commissioner should consider this principle in determining the manner of its

coordination with Maryland and Virginia.

D. Commerce Clause

Virginia briefly argues that the Commissioner’s order violates the Commerce

Clause, because the order reduces GHMSI’s overall surplus, thereby burdening

GHMSI subscribers in Virginia and Maryland. Framed at that level of generality,

Virginia’s argument is unpersuasive, particularly given that Congress domiciled

GHMSI in the District and provided the District with regulatory authority over

GHMSI’s activities. Pub. L. No. 103-127, § 138, 107 Stat. 1349; cf. generally, e.g.,

Sprint Commc’ns Co. v. Kelly, 642 A.2d 106, 114 (D.C. 1994) (whether one state’s

tax on company doing business in multiple states violates Commerce Clause
 24

depends on whether tax “(1) applies to an activity with a substantial nexus to the

taxing state; (2) is fairly apportioned; (3) discriminates against interstate commerce;

and (4) is fairly related to services or benefits provided by the states”).

Virginia has not in this case presented a more specific challenge to the

Commissioner’s orders, based on a concrete analysis of those orders and other

particular circumstances that might be relevant to whether the orders impermissibly

discriminate against residents of Maryland or Virginia or otherwise are

impermissible under the Commerce Clause. We therefore do not address that issue.

See, e.g., Graham v. United States, 12 A.3d 1159, 1167 n.10 (D.C. 2011) (“Where a

party generally raises an issue on appeal without supporting argument, we deem [the

issue] abandoned.”). We do note, however, that GHMSI has raised a Commerce

Clause claim in a pending federal case that GHMSI brought against DISB and the

corresponding Virginia and Maryland regulators. CareFirst, Inc. v. Taylor, No.

1:16-cv-02656 (D. Md. filed July 22, 2016).

E. Confidence Level

In determining whether GHMSI’s 2011 surplus was excessive, the

Commissioner tried to assess the future risks that GHMSI faced. The Commissioner
 25

analyzed those risks in terms of a specific benchmark: the likelihood that GHMSI

would fall below 200% of RBC-ACL in the next three years. The Commissioner

tried to determine what level of surplus would provide a sufficient assurance that

GHMSI would not fall below that benchmark. The parties do not dispute that general

approach. The parties do disagree about the appropriate level of assurance, also

called a confidence level. Appleseed argues that the appropriate three-year

confidence level is not higher than 90%, i.e., that GHMSI’s 2011 surplus needed to

be no larger than would create a 90% probability that GHMSI would remain above

the 200% RBC-ACL benchmark in the following three years. GHMSI argues that

the appropriate three-year confidence level is 98%. DISB defends the

Commissioner’s choice of a 95% three-year confidence level. We uphold the

Commissioner’s choice of a 95% confidence level.

In selecting a 95% confidence level, the Commissioner emphasized the

adverse consequences of falling below the 200% benchmark. Specifically, the

Commissioner found that dropping below that benchmark: (1) is generally

understood as a sign of serious financial problems; (2) would permit the Blue Cross

Blue Shield Association to revoke GHMSI’s licensing agreement, causing some

GHMSI subscribers to be reassigned to other providers and all GHMSI subscribers

to lose the benefits of full access to the Blue Cross Blue Shield network; (3) would
 26

likely cause significant concerns among GHMSI subscribers and policyholders,

possibly leading subscribers to leave GHMSI; and (4) given GHMSI’s dominance

in the District’s market, would cause extreme distress in the District’s health-

insurance market. The Commissioner also found, with adequate support in the

record, that it would be difficult for GHMSI to rebuild its surplus.

The Commissioner concluded, however, that a 95% confidence level was high

enough to be “consistent with financial soundness and efficiency.” The

Commissioner explained that falling below the 200% benchmark would not itself

cause GHMSI to be insolvent and would trigger regulatory oversight and other

ameliorative responses by GHMSI. The Commissioner further reasoned that picking

a confidence level higher than 95% would not give effect to the statutory obligation

to maximize community-health reinvestment, because a higher confidence level

would cause GHMSI to “accumulate surplus at a level that [would be] inefficient or

unnecessary for financial soundness.” Conversely, the Commissioner concluded

that picking a confidence level as low as 90%, the figure Appleseed advocated,

“could jeopardize [GHMSI’s] financial soundness.”

In Appleseed I, we explained that “a proper surplus determination under the

MIEAA requires simultaneous consideration of the requirement to engage in
 27

community reinvestment to the maximum feasible extent consistent with financial

soundness and efficiency.” 54 A.3d at 1218-19 (internal quotation marks omitted).

We further explained that, “[a]s to the specification of how surplus and community

reinvestment are to be calculated and balanced, we defer to the agency’s reasonable

discretion in light of its expertise in this subject matter.” Id. at 1215. We conclude

that the Commissioner’s selection of a 95% confidence level in this case reasonably

balanced the competing considerations of financial soundness and community-

health reinvestment. We also conclude that the Commissioner adequately explained

his decision. We are not persuaded by the arguments to the contrary.

First, GHMSI argues that a 95% confidence level was not supported by the

evidence. To the contrary, there was evidence in the record that supported use of a

95% confidence level. In any event, under Appleseed I, the determination of a

confidence level is not a purely factual decision to be based exclusively on actuarial

opinions or evidence of industry practice, though such information is certainly

relevant. Rather, determination of a confidence level requires a policy judgment,

based on a balancing of the pertinent risks and benefits.

Second, Appleseed argues that 2% or 5% risks are necessarily too attenuated

to justify GHMSI’s holding of surplus. We do not agree. A relatively small risk can
 28

reasonably be viewed as substantial if the gravity of the harm is sufficiently great.

See, e.g., Onishea v. Hopper, 171 F.3d 1289, 1297 (11th Cir. 1999) (“[W]hen the

adverse event is the contraction of a fatal disease, the risk of transmission can be

significant even if the probability of transmission is low: death itself makes the risk

significant.”) (internal quotation marks omitted). We thus see no categorical bar to

adopting a confidence level of 95%.

Third, the parties dispute the proper mathematical calculation of the long-term

risks associated with picking a particular three-year confidence level. An expert

retained by DISB stated that under a 90% three-year confidence level GHMSI’s

surplus falling below the 200% benchmark “would be expected, statistically, to

occur once every 10 years.” According to Appleseed’s expert, making certain

simplifying assumptions, a 90% three-year confidence level would mean that

GHMSI’s surplus would be expected to fall below the 200% benchmark once every

thirty years. We need not resolve that dispute. The Commissioner’s December 2014

order does not explicitly adopt either expert’s view, stating only that a 90%

confidence level would mean “a one-in-ten chance of surplus falling below 200%

RBC-ACL.” That statement is indisputably accurate as it relates to the single three-

year period directly at issue.
 29

F. Range v. Fixed Number

The Commissioner concluded that, in order to ensure that GHMSI’s surplus

would not fall below the 200% benchmark at a 95% three-year confidence level,

GHMSI needed a 2011 surplus of 721% RBC-ACL, i.e., approximately $700

million. The Commissioner thus determined that surplus above that amount was

excessive. GHMSI argues that the Commissioner erred in picking a fixed number

rather than a range, because a range is necessary to account for the “imprecision of

surplus analysis and the variability of surplus.” GHMSI also argues that the

Commissioner did not adequately explain the reasoning behind its decision to pick

a fixed number rather than a range. We uphold the Commissioner’s ruling on this

point.

As we have explained, the Commissioner is charged under the MIEAA with

determining whether GHMSI held excess surplus in 2011 that should be reinvested

in community health. The Commissioner acted reasonably in using a fixed number

rather than a range in determining the amount of surplus that was appropriate for

GHMSI in 2011. Like other decisionmakers, administrative agencies are often

required to make specific decisions in complicated contexts based on predictions or

other estimates. Cf., e.g., Florida v. Georgia, 138 S. Ct. 2502, 2514 (2018)
 30

(“Reliance on reasonable predictions of future conditions is necessary.”) (internal

quotation marks omitted). Although agencies may in some circumstances respond

to such uncertainty by picking a range rather than a fixed number, the predominant

approach is “the traditional ‘pick-a-point’ methodology.” Ill. Bell Tel. Co. v. Ill.

Commerce Comm’n, 561 N.E.2d 426, 433 (Ill. App. Ct. 1990). We see no error in

the Commissioner’s decision to apply the traditional approach in the current setting.

We are also satisfied with the Commissioner’s explanation that if the Commissioner

established a range rather than a target, GHMSI might end up holding excess surplus.

GHMSI argues, however, that given the uncertainties involved it is

unreasonable to treat GHMSI as having “violat[ed] the MIEAA” simply because its

2011 surplus was above the fixed number calculated by the Commissioner. We

disagree. The Commissioner’s determination that GHMSI held excessive surplus

did not trigger any punitive consequences. Rather, it simply provided a basis for the

Commissioner’s later order requiring that GHMSI reinvest the excess surplus

attributable to the District in community health. We see no unfairness in imposing

such a consequence as a result of such a determination.
 31

G. Equity Portfolio

GHMSI’s earnings from its portfolio of equity investments constitute a

significant part of GHMSI’s overall revenue, and gains and losses on those

investments could substantially affect GHMSI’s future financial situation.

Accordingly, the actuaries attempted to predict the range of possible future gains and

losses on those investments and to assess the impact of such possible gains and losses

on GHMSI’s overall financial condition. Before the Commissioner, Appleseed

repeatedly argued that the experts retained by DISB and GHMSI were committing

an error in their calculations on that issue, which had the effect of exaggerating the

effect of potential equity-portfolio losses on GHMSI’s overall revenue. GHMSI’s

expert denied that such an error had been committed. The Commissioner adopted

the approach taken by DISB’s expert without mentioning Appleseed’s objection.

We therefore remand this issue for the Commissioner to specifically address

Appleseed’s objection. See Appleseed I, 54 A.3d at 1219 (“[T]he technical nature

of the actuarial reports requires a far more detailed discussion of a decision in which

even a small variance can implicate millions of dollars.”).
 32

H. Premium Growth Rate

Another important consideration in projecting GHMSI’s financial future is

estimating how many new subscribers GHMSI might have in the future and what

premiums those subscribers might pay. The Commissioner projected a most likely

premium growth rate of 8% with respect to GHMSI’s non-Federal Employee Plan

subscribers. Appleseed argues that the Commissioner failed to adequately explain

the reasoning behind this decision. We conclude otherwise. The Commissioner

specifically discussed a number of considerations in explaining that decision,

including the possible impact of the Affordable Care Act, rising health-care costs,

the possibility that subscribers might choose plans with fewer benefits and lower

premiums, and differences between Federal Employee Plan subscribers and non-

Federal Employee subscribers. We view the Commissioner’s explanation as

sufficient.

I. Real-World Data & Risks

Appleseed argues that the Commissioner erred by not testing the reliability

and plausibility of actuarial-model results against the historical record, evidence of

actual real-world risks, and other facts outside the model. We disagree. The
 33

actuarial models and the Commissioner’s analysis reflect reasonable efforts to make

future predictions based in appropriate part on past information. Cf., e.g., Am. Pub.

Gas Ass’n v. Fed. Power Comm’n, 567 F.2d 1016, 1037 (D.C. Cir. 1977) (“Reasoned

decisionmaking can use an economic model to provide useful information about

economic realities, provided there is a conscientious effort to take into account what

is known as to past experience and what is reasonably predictable about the future.”).

We agree with GHMSI, moreover, that “[h]olding surplus to guard against the

potential that one or more costly events that have not previously occurred may occur

in the future is how health insurance companies must be responsibly operated.”

The Commissioner was reviewing in 2013 and 2014 what amount of surplus

would have been appropriate for GHMSI to hold at the end of 2011, based on

predictions about GHMSI’s financial situation in the period from 2012 to 2014. This

meant that the actuarial models used to attempt to predict GHMSI’s future financial

health had been outpaced at least in part by what actually happened. Ultimately, the

Commissioner decided to rely on the actuarial models and approach the process as

if calculations were being done at the end of 2011 with the information available at

that time. To the extent that Appleseed contends that this decision was unreasonable,

we conclude otherwise. Appleseed cites nothing in the MIEAA, or from any other

authority, supporting the theory that the reasonableness of GHMSI’s 2011 surplus
 34

must be assessed in hindsight, based on what actually happened in subsequent years.

The Commissioner did not act unreasonably in instead assessing the reasonableness

of the 2011 surplus based on what would reasonably have been predicted at that

time. In many other contexts, it is well settled that reasonableness of an action

should be determined based on the situation at the time of the action rather than

based on hindsight. See, e.g., Kingsley v. Hendrickson, 135 S. Ct. 2466, 2473 (2015)

(reasonableness of use of force should be assessed based on what “officer knew at

the time, not with the 20/20 vision of hindsight”).

J. Attribution of Excess Surplus to the District

After determining that GHMSI’s overall surplus was excessive, the

Commissioner turned to determining the extent to which the excess surplus was

“attributable to the District.” D.C. Code § 31-3506(e). The MIEAA does not

provide guidance about how to allocate excess surplus to the District. The pertinent

regulation directs the Commissioner to

allocate the portion of the surplus of a hospital and medical
services corporation that is derived from the company’s
operations in the District of Columbia based on the
following factors:

(a) The number of policies by geographic area;
 35

(b) The number of health care providers under contract
with the company by geographic area; and

(c) Any other factor that the Commissioner deems to be
relevant based on the record of a public hearing held
pursuant to [26A DMCR] section 4602.

26A DCMR § 4699.2 (2019).

The Commissioner attributed 21% of GHMSI’s excess surplus to the District.

In brief, the Commissioner’s reasoning was as follows: (1) the most relevant

consideration is the geographic location of GHMSI’s health-insurance contracts, “as

measured by the premiums reported and number of policies issued in each

jurisdiction”; (2) the 2011 annual reports of GHMSI and one of its affiliates

indicated that the District accounted for 18% of the premiums from Federal

Employee Program (FEP) policies and 22% of the premiums for non-FEP policies;

(3) because “FEP business is less risky,” greater weight should be given to non-FEP

business than to FEP business, resulting in a weighted premium percentage for the

District of 21%; (4) also using 2011 data, 19% of GHMSI’s policies were held by

D.C. policyholders; (5) again using 2011 data, 15% of GHMSI’s network providers

were located in the District; (6) policyholder data and provider location should be

given less weight than premium data; and (7) weighting these three variables

appropriately resulted in an allocation of 21% of the excess surplus to the District.
 36

GHMSI has repeatedly challenged the reasonableness of allocating a portion

of its surplus, which is unitary, to the District as one of the three jurisdictions in

which GHMSI operates. The MIEAA, however, requires the Commissioner to

undertake such an analysis. D.C. Code § 31-3506(e). We do not understand GHMSI

to be arguing that this requirement of the MIEAA is invalid, and the Commissioner

thus appropriately undertook to allocate a portion of the excess surplus to the

District. GHMSI also at times appears to argue that, in allocating the excess surplus,

the Commissioner was required to create a hypothetical smaller version of GHMSI

that reflected only GHMSI’s activities in the District, and then to conduct the

remainder of the analysis based on that hypothetical entity. We conclude that the

Commissioner reasonably rejected that approach. GHMSI in fact is a bigger

company, and (as GHMSI acknowledges) all of its surplus is available to protect

against future financial problems. Moreover, D.C. Code § 31-3506(f) requires the

Commissioner to consider all of GHMSI’s financial obligations in determining

whether GHMSI’s surplus is excessive. It follows naturally that the full scope of

GHMSI’s activities can appropriately be considered when the Commissioner

determines whether GHMSI’s surplus is excessive and how much of any excess

should be allocated to the District.
 37

GHMSI and Appleseed raise numerous other objections to the

Commissioner’s attribution methods and conclusions. We agree that a remand is

necessary for the Commissioner to more fully address the issues raised by the parties.

One conceivable approach to analyzing what portion of GHMSI’s excess

surplus (if any) should be attributed to the District would be as follows: (1) decide

what parts of GHMSI’s activity should be viewed as attributable to the District; (2)

for any given year, determine the extent to which such D.C. activity added to or

subtracted from the surplus, which in turn might turn on an assessment of (i) how

much profit or loss resulted from the D.C. activity and (ii) to what extent the D.C.

activity required greater or lesser surplus because of differential risk among the

jurisdictions; and (3) because any excess surplus might be built up over years, make

the same determination for other relevant years, rather than simply looking at

information relating to a single year. Actually implementing such an approach

would be extremely daunting from a practical point of view. There may a number

of reasonable approaches to allocating excess surplus among jurisdictions. We

would be inclined to give the Commissioner substantial latitude in determining how

best to approach that issue. We agree with the parties, however, that the

Commissioner in this case has in a number of important respects not adequately

explained the approach the Commissioner took to apportioning surplus.
 38

Specifically, the Commissioner has in our view not provided adequate

explanations in the following respects: (1) why the focus was on a snapshot of 2011

rather than an effort to analyze GHMSI’s surplus history and to determine the

District’s contributions to that surplus over time; (2) why the Commissioner

addressed only to a very limited degree alleged differences among the District,

Virginia, and Maryland with respect to the riskiness and profitability of GHMSI’s

activities; and (3) why, in determining what activity should be attributed to the

District, the Commissioner took as dispositive how GHMSI reported its activity,

even though (a) GHMSI’s manner of reporting its activity has varied over time and

(b) relying on GHMSI’s manner of reporting led the Commissioner to treat FEP

policies differently from non-FEP policies without providing a sufficient rationale

for such differential treatment. We therefore remand for further consideration of

these issues. See generally, e.g., Appleseed I, 54 A.3d at 1216 (“Explanation in

sufficient detail . . . is required for meaningful judicial review and for there to be a

basis for judicial deference to agency determinations.”) (internal quotation marks

omitted).
 39

K. Actuarial Fees & Prejudgment Interest

Citing a letter that it asserts can be found on DISB’s website, Appleseed states

that it asked DISB to reimburse Appleseed’s actuarial fees. Without citation to the

record, Appleseed also states that it asked DISB to award prejudgment interest.

Appleseed further argues that the Commissioner did not address either request.

DISB does not address either issue in its briefs in this court. GHMSI does not dispute

that both requests were made to DISB and that the Commissioner did not address

either request. Rather, GHMSI argues only that both requests should properly be

denied on the merits. In general, however, “an administrative agency’s decision can

be sustained on review only on the grounds on which the agency actually relied.”

Black v. District of Columbia Dep’t of Human Servs., 188 A.3d 840, 850 (D.C.

2018). “When a party asks us to affirm an agency’s decision for a reason not relied

on by the agency, we thus ordinarily remand the case for the agency’s consideration

in the first instance of the reason advanced by the party seeking affirmance.” Id. at

851. We follow that course in the present case. On remand, the Commissioner

should therefore address Appleseed’s requests for reimbursement of actuarial fees

and for prejudgment interest, to the extent those requests remain at issue.
 40

L. Type of Community-Health Reinvestment

The parties have extensively debated the Commissioner’s decision to order a

rebate rather than some other type of community-health reinvestment. We are

remanding the case for further proceedings, however, and therefore it is not at

present clear whether the Commissioner will ultimately determine that it is

appropriate to order community-health reinvestment and if so in what form. We

therefore do not address at this time the disputes about the appropriate type of such

reinvestment. See generally, e.g., Jackson v. Condor Mgmt. Grp., Inc., 587 A.2d

222, 224 (D.C. 1991) (declining to reach issues that “may or may not arise again on

remand”).

For the foregoing reasons, we affirm the Commissioner’s orders in part,

vacate in part, remand for further proceedings.

So ordered.