HIPPOS, LLC, Petitioner-Appellant v. MISSOURI DEPARTMENT OF HEALTH AND SENIOR SERVICES, Respondent-Respondent

HIPPOS, LLC, Petitioner-Appellant, v. MISSOURI DEPARTMENT OF HEALTH AND SENIOR SERVICES, Responden t-Respondent. No. SD38218

APP EAL FROM THE CIRCUIT COURT OF MARIES COUNTY Honorable William E. Hickle, Circuit Judge REVERSED AND REMANDED WITH DIRECTIONS Hippos, LLC (Hippos) appeals from the circuit court's judgment affirming three decisions by the Administrative Hearing Commission (Commission). The Commission determined that Hippos was ineligible to receive 13 licenses to operate medical marijuana cultivation, infused products manufacturing, and dispensary facilities. We reverse the circuit court's judgment and remand with directions for the circuit court to

I n Division

2

order the Missouri Department of Health and Senior Services (Department) to grant those 13 licenses to Hippos. Standard of Review We have jurisdiction over this appeal pursuant to § 536.140.6, which permits an appeal from a judgment of the circuit court affirming a decision by the Commission, as in other civil cases. Id. 1 In MO CANN Do, Inc. v. Missouri Dep't of Health & Senior Servs., 686 S.W.3d 642 (Mo. banc 2024), our Supreme Court summarized the applicable standard of review: "Article V, section 18 of the Missouri Constitution articulates the standard of judicial review of administrative actions." Albanna v. State Bd. of Registration for Healing Arts, 293 S.W.3d 423, 428 (Mo. banc 2009). "On appeal, this Court is charged with determining whether the agency actions 'are authorized by law, and in cases in which a hearing is required by law, whether the same are supported by competent and substantial evidence upon the whole record.'" Id. (quoting Mo. Const. art. V, sec. 18).

Moreover, "all final decisions of the administrative hearing commission shall be subject to judicial review as provided in and subject to the provisions of sections 536.100 to 536.140[.]" Section 621.145. "Section 536.140.2 lists several criteria for judicial review of an agency decision." Stone v. Mo. Dep't of Health & Senior Servs., 350 S.W.3d 14, 19 (Mo. banc 2011). This Court reviews whether the agency action:

(1) Is in violation of constitutional provisions;

(2) Is in excess of the statutory authority or jurisdiction of the agency;

(3) Is unsupported by competent and substantial evidence upon the whole record;

1 All statutory references are to RSMo (2016).

3

(4) Is, for any other reason, unauthorized by law;

(5) Is made upon unlawful procedure or without a fair trial;

(6) Is arbitrary, capricious or unreasonable;

(7) Involves an abuse of discretion.

Section 536.140.2.

"Consistent with the constitutional standard, section 536.140.2 provides for appellate review of the administrative ruling, not that of the circuit court[.]" Lagud v. Kan. City Bd. of Police Comm'rs, 136 S.W.3d 786, 791 (Mo. banc 2004); see also Stone, 350 S.W.3d at 19 (citing section 536.140.2) ("An appellate court reviews the decision of the agency rather than the decision of the circuit court."). "For purposes of review, the action of the commission and the order, if any, of the agency shall be treated as one decision." Section 621.145. "Under article V, section 18 and section 536.140.2, the standard of review for administrative decisions is 'whether, considering the whole record, there is sufficient competent and substantial evidence to support the agency's decision.'" Stone, 350 S.W.3d at 20 (alteration omitted) (quoting Albanna, 293 S.W.3d at 428).

MO CANN Do, Inc., 686 S.W.3d at 645-46 (internal footnote omitted). We apply this standard of review here. Procedural Background On November 6, 2018, the citizens of the State of Missouri adopted article XIV, section 1 of the Missouri Constitution, addressing the right to access medical marijuana. Metrc, LLC v. Steelman, 617 S.W.3d 472, 474 (Mo. App. 2021). 2 After the initial adoption of this amendment, Hippos submitted applications to the Department to receive

2 This section of the Missouri Constitution was amended at the general election on November 8, 2022, effective December 8, 2022. Because the events involved in this appeal predate the adoption of this amendment, this change plays no role in our decision here.

4

three cultivation licenses, six infused products manufacturing licenses, and five dispensary licenses. On December 26, 2019, the Department awarded Hippos one cultivation license and denied the other two cultivation applications. On January 24, 2020, Hippos timely appealed those denials to the Commission pursuant to § 621.120 and 19 CSR 30- 95.025(6)(B). 3

On January 10, 2020, the Department denied Hippos' six applications for infused products manufacturing. On February 5, 2020, Hippos timely appealed those denials to the Commission pursuant to § 621.120 and 19 CSR 30-95.025(6)(B). On January 23, 2020, the Department denied Hippos' five dispensary applications. On February 21, 2020, Hippos timely appealed those denials to the Commission pursuant to § 621.120 and 19 CSR 30-95.025(6)(B). Hippos' three appeals were each assigned a case number in the Commission: Cultivation (No. 20-0551), Manufacturing (No. 20-1634), and Dispensary (No. 20- 1181). Hippos' appeals were consolidated for the purpose of conducting the hearing, and a three-day, in-person hearing was held on February 17, 18 and 19, 2021. On November 16, 2021, the Commission issued its decision in the Cultivation appeal. On December 8, 2021, Hippos timely filed its "Motion for Reconsideration" of that decision. On December 8, 2021, the Commission granted Hippos' motion for

3 All references in this opinion to the Code of State Regulations are to CSR (12/31/2019). Regulation 19 CSR 30-95.025 was rescinded on July 30, 2023, after the amendment of article XIV, § 1 of the Missouri Constitution in December 2022.

5

reconsideration. On March 15, 2022, the Commission issued an "Amended Decision" in the Cultivation appeal. That same day, the Commission also issued decisions in the Manufacturing and Dispensary appeals. On April 12, 2022, Hippos timely petitioned for judicial review of the Commission's three decisions in the Circuit Court of Maries County pursuant to § 621.145 and § 536.140. On November 7, 2022, the three cases pending in the circuit court were consolidated and designated Case No. 22MS-CC00040 as the surviving case. On August 15, 2023, the court entered its judgment affirming the three decisions previously entered by the Commission. On September 19, 2023, Hippos timely appealed the judgment of the circuit court. Factual Background We must determine whether the Commission's decision is "supported by competent and substantial evidence upon the whole record." M O. CONST. art. V, § 18. "We review the whole record, and we no longer view the evidence in the light most favorable to the agency's decision." Nexgen Silica, LLC v. Mo. Dep't of Nat. Res., 677 S.W.3d 594, 600 (Mo. App. 2023); see Hauck v. Police Ret. Sys. of St. Louis, 721 S.W.3d 150, 157 (Mo. App. 2025). We have applied this principle in summarizing the relevant facts, which are not in dispute. The version of article XIV, § 1 of the Missouri Constitution adopted in 2018 legalized the use and sale of marijuana for medical purposes and gave the Department

6

the responsibility to issue licenses to cultivators, manufacturers, dispensaries, and testing facilities. See id. This article of the Constitution prescribed a minimum number of licenses that the Department was required to grant, but it also permitted the Department to limit the number of licenses. M O. CONST. art. XIV, § 1.3(15)-(17). The Department initially capped the number of licenses at the minimum number required by the Missouri Constitution. The caps for medical marijuana facility licenses were: 60 cultivation licenses; 86 manufacturing licenses; and 192 dispensary licenses (24 per Congressional District). The Department also had the authority to "[e]stablish a system to numerically score competing medical marijuana licensee ... applicants" if "more applicants appl[ied] than the minimum number of licenses[.]" M O. CONST. art. XIV, § 1.3(1)(h). The Department received more applications than the number of available licenses. Thus, the Department was required to "use a system of numerically scoring ten (10) additional evaluation criteria to rank the applications in each such license or certification category against each other." 19 CSR 30-95.025(4). The regulations also gave the Commission or a court of competent jurisdiction the authority to order the Department to grant cultivation, manufacturing, and dispensary facility licenses to appellants timely filing meritorious appeals, notwithstanding the license caps imposed by the Department. See 19 CSR 30-95.025(6)(C). The Department did not score any of the facility license applications. Instead, the scoring was done by Wise Health Solutions, LLC (Wise) pursuant to a contract between

7

Wise and the Department. Consistency and fairness were of utmost importance in the application scoring process. That was explicitly made clear in a training manual provided to scorers, which stated, "consistency across all applications is of the utmost importance!" (Emphasis in original.) Per the contract with Wise, the Department was required to provide Wise with a "Facility Scoring Guide." Scorers were required to score applications "in an objective way using the predetermined point range and instructions outlined in the Facility Scoring Guide[.]" Pursuant to the Facility Scoring Guide: It is imperative that every response is scored in a consistent manner for all applicants. Some of the questions and answers may require a scorer to use his or her own professional knowledge and expertise in scoring the responses. In doing so, a scorer must score every response consistently. For example, if two applicants applying for the same facility type provide identical responses to a question, the score must be the same.

The Department also provided Wise with the "Facility License Application Questions and Scoring Criteria" that was incorporated in the Department's regulations. This included an "Evaluation Criteria Scoring Table" that provided the different scoring methodologies:

8

9

The same methodologies were set forth in the Facility Scoring Guide, which stated that "[s]corers are required to score each response ... using the criteria and points in the Evaluation [Criteria Scoring] Table." One scoring methodology was labeled "0- 10 Point Scale[.]" Pursuant to this scoring methodology, an applicant could receive 0, 4, 7 or 10 "raw" points on a question, based on whether the scorer believed the applicant's answer was Unsatisfactory, Satisfactory, Superior, or Distinctive (terms defined in the Facility Scoring Guide). 4 Another methodology allowed a scorer to award a score of either 0 or 10 raw points for any questions that were reviewed on a Satisfactory/Unsatisfactory or a Yes/No basis. Wise prepared a "Wise Health Solutions Final Work Plan" (Work Plan), which was finalized on September 15, 2019. Wise's Work Plan was approved by the Department. The Work Plan stated: "The most important part of [a scorer's] job is being consistent in grading so that each application is scored in the same manner and compared to other applications received within the specific facility type." Wise also prepared a "Grader Training Manual V3" (Training Manual) for its scorers, which the Department reviewed and approved. The Training Manual stated: The most important part of your job is being consistent in your grading so that each application is scored in the same manner and compared to other applications received within the specific facility type. Conflict, challenges to the process and lawsuits arise when inconsistent scoring results are evident; the companies will likely see their individual scores and will be able to compare across multiple application[s'] results to see for

4 After scoring was completed, the initial "raw" scores were then weighted so that some questions had higher values than others. Weight adjustments were applied to all applications to assign total scores and rankings within each license type.

10

themselves if we are consistent. An example is if a company submits multiple applications with an identical answer to a certain question (on different applications) but receives a different score for each. This should NEVER happen, and would be considered a statistical anomaly. You can and should self-correct this.

(E mphasis in original.) Like the Training Manual, the Work Plan similarly stated that the most important part of a scorer's job was to be consistent in grading so that each application was scored "in the same manner and compared to other applications received within the same facility type." Per Wise's contract with the Department, Wise was required to assign to each question a scorer who possessed the appropriate background and expertise to evaluate that question and response. The Work Plan stated that each scorer was to have the experience and background appropriate to perform the task they were assigned. Wise was required to provide the Department with information concerning each scorer that would allow the Department to determine if the individual was appropriately qualified for the task assigned. Additionally, the Work Plan required Wise to provide the Department, by September 9, 2019, a spreadsheet that included a list of the scorers and their primary subject-matter expertise. Wise did not provide the Department with such a spreadsheet. Wise sent the Department only resumes for each scorer and the questions assigned to each scorer. The Department reviewed the resumes and confirmed its agreement with the assignments. The Department's deputy director and counsel for the section for medical marijuana regulation was Amy Moore (Deputy Director). She testified that notes from

11

the scorers would have helped to show an indication of consistency. However, the Training Manual created by Wise reminded scorers that any emails, notes, or other written materials would be discoverable if any scores were challenged in court. Additionally, the Training Manual stated: "Don't write anything that you don't want everybody to read. Past versions (something deleted) will be discoverable[.] Adhere to this axiom: Say it and forget it; write it and regret it." On behalf of Hippos, Jonathan Andres reviewed and analyzed all the scorers' notes and all of the scoring rubrics produced by the Department in search of any notes concerning Hippos. The only references to Hippos were in the notes of two scorers, Erin Bruckner and Noah Hersh, and, in each instance, the scorer merely listed Hippos' application identification number. Neither scorer provided any description, notes, or narrative regarding their considerations or reasoning about the scores awarded to Hippos. Wise scored all of the license applications, and the Department did not change any of the scores. The scorers were charged with self-correcting their own scores. Pursuant to the Work Plan, if there were any statistically significant variances, Wise was required to re-score the applicable questions to ensure consistency and integrity. Pursuant to the Work Plan, Wise understood that "the application question(s) that were determined to have a variance should be rescored," and "[t]he first score will be discarded and we will record the second score as the final score[.]" Pursuant to 19 CSR 30-95.025(4)(C)(4), "scores that vary significantly from other scores for the same

12

question may be rescored," and if re-scored, "the first score will be discarded, and the second score will stand[.]" Pursuant to the Training Manual: "Statistical Anomalies will be reviewed. These may be outliers, unusual patterns, an unusual occurrence in the pool of '0' or '10' scores, and 'same answer receiving the same score' reviews where possible (because the 'same answer' should always receive the 'same score.'" The Training Manual directed scorers: "You can, and should, self-review your grading habits and scoring, regularly looking for anomalies that you may self-correct." It further instructed that "consistency across all applications is of the utmost importance!" After the scores for all applications in each license type were submitted by the individual scorers, Wise informed the Department of completion as to each license type, asserting "no statistically significant variances were identified." The Department performed no review of Wise's scoring to determine if scores varied significantly. All of the application scores the Department received from Wise were accepted and used as valid by the Department. All facility license applications were scored separately. Hippos submitted three cultivation applications (HIPP9706, 9707 and 9708); six manufacturing applications (HIPP9709-9714); and five dispensary applications (HIPP9701-9705). 5

5 We adopt these identification numbers, as provided in demonstrative exhibits before the Commission, for ease of reference, noting that they vary from the actual application identification numbers.

13

On December 26, 2019, the Department released the results of its scoring and ranking of cultivation applications. The lowest-scoring cultivation applicant to receive a license, GF Saint Mary, LLC, received a total score of 1,479.41 points. Hippos' cultivation applications received total scores of: 1,481.33 (HIPP9708); 1,440.13 (HIPP9706); and 1,439.73 (HIPP9707). The Department determined that Hippos was qualified to receive one license for a cultivation facility in Vienna, Missouri, and it granted a license for application HIPP9708. Hippos' cultivation applications HIPP9706 and 9707 met the minimum standards for licensure, but they were denied based on the results of the scoring and ranking. If either of these applications had received a score greater than 1, 479.41, it would have placed within the top 60 cultivation applications and qualified for the award of a cultivation license. On January 10, 2020, the Department released the results of its scoring and ranking of manufacturing applications. The lowest-scoring manufacturing applicant to receive a license, Bloom Medicinals of MO, LLC, received a total score of 1,526.16 points. Hippos' manufacturing applications received total scores of: 1,391.39 (HIPP9709); 1,391.99 (HIPP9710, 9711, 9712 and 9713); and 1,383.19 (HIPP9714). If any of those applications had received a score greater than 1, 526.16, that manufacturing application would have placed within the top 86 manufacturing applications and would have qualified for the award of a manufacturing facility license. Hippos' manufacturing applications met the minimum standards for licensure and were denied based on the results of the scoring and ranking.

14

On January 23, 2020, the Department released the results of its scoring and ranking of dispensary applications. The lowest-scoring dispensary applicants to receive a license in each of the Congressional Districts in which Hippos submitted an application were: (1) Growing Jobs Missouri, LLC, with a total score of 1,496.71 in Congressional District 1; (2) LSL Management, LLC, with a total score of 1,487.44 in Congressional District 2; and (3) TC AppliCO, LLC, with a total score of 1,513.83 in Congressional District 5. Hippos' dispensary applications received total scores of: 1,393.78 (HIPP9703 and 9705 in Congressional District 1); 1,393.78 (HIPP9704 in Congressional District 2); and 1,393.78 (HIPP9701 and 9702 in Congressional District 5). Hippos' dispensary applications met the minimum standards for licensure and were denied based on the results of the scoring and ranking. Hippos' Expert Testimony Concerning Its Cultivation and Dispensary Applications Hippos presented expert testimony from Ashley Picillo (Picillo) concerning its cultivation and dispensary appeals. In 2014, Picillo founded Point 7, LLC, a management consulting firm specializing in cannabis compliance and cannabis facility design, staffing, and community engagement. Point 7 provided services to clients in the U.S. and international markets (Canada, Mexico and Australia), and it worked with existing cannabis operators and those seeking to become cannabis operators. Picillo provided cannabis licensing assistance in 21 different states and also had worked on a

15

committee with the Dean of Excelsior College in New York to assist in developing cannabis-specific college coursework. Picillo and Point 7 wrote approximately 50 applications for Missouri's medical marijuana program, and their clients were awarded 42 licenses. Picillo's firm was named Consultant of the Year in Greenway Magazine for its work in drafting applications for Missouri clients. Picillo utilized the Department's "Medical Marijuana Application Scorer's Guide General Instructions" when evaluating Hippos' responses. Picillo testified that, based on her knowledge and experience, she was capable of evaluating each component of the Evaluation Criteria Scoring Table. Picillo testified there were significant differences between the possible scores of 0, 4, 7 and 10. The methodology Picillo used included reliance upon: (1) Hippos' application answers; (2) the Department's scoring criteria; (3) Point 7's responses prepared for its Missouri clients; (4) scores received on the responses prepared by Point 7; and (5) documentation issued by the Department. Hippos also presented expert testimony from Joe Patterson (Patterson) in connection with certain security questions in the cultivation and dispensary appeals. Patterson formed Ceres Management Group, LLC, a company staffed primarily by former law enforcement officers, former military personnel from special operations, and a former White House staff member with extensive experience in writing government proposals. Patterson and Ceres Management also wrote the security responses, or the entire applications, for 33 medical marijuana applications submitted by their Missouri

16

clients. Approximately 50 percent of those applications were awarded licenses, whereas the success rate based on the total number of applications was 15.8 percent. Cultivation Applications Hippos received different scores for identical answers to a number of questions in cultivation applications HIPP9706, 9707 and 9708, as illustrated in the following table. Hippos' expert Picillo provided detailed testimony about the scores Hippos had received on Questions 3, 27, 30, 41, 44, 50, 51, 52, 66 and 67. After Picillo's review, she opined that 10 out of the 22 cultivation application responses should have received an increased score, as shown in the following table: Question HIPP 9706 HIPP 9707 HIPP 9708 Picillo's Score 3 7 7 7 10 26 4 7 4 27 10 4 4 10 30 4 4 4 7 41 7 0 7 7 44 0 0 0 7 48 4 4 7 49 7 7 10 50 4 7 7 7 51 10 7 10 10 52 7 10 10 10 66 7 7 10 10 67 7 7 7 10

If Hippos' answers to these questions had received Picillo's increased raw scores, the total weighted score for application HIPP9706 would have been 1,526.52. For

17

application HIPP9707, the total weighted score would have been 1,528.32. These scores would have placed the applications within the top 60 cultivation applications. Dispensary Applications Hippos similarly received different scores for identical answers for a number of questions in dispensary applications HIPP9701-9705, as illustrated in the following table. Picillo testified in detail about why Hippos should have received increased scores on 10 of its responses, also shown in the following table: Question HIPP9701 HIPP9702 HIPP9703 HIPP9704 HIPP9705 Picillo's Score 26 7 7 10 31 4 7 35 10 7 36 10 7 39 4 7 41 0 0 4 43 4 7 7 44 4 7 7 51 10 4 7 52 10 7 10 66 7 7 10 67 4 4 10 D4 4 4 7 D6 4 4 7

If Hippos' answers to these questions had received Picillo's increased raw scores, dispensary applications HIPP9701 and 9702 each would have received a total score of 1,495.76, and applications HIPP9703-9705 each would have received a total score of 1,498.16.

18

Hippos also presented expert testimony from Patterson about Hippos' answers in the dispensary applications to Questions 55-60, which related to security. For Patterson's review of these security questions, he interpreted "minimum expectations" as the term is set forth in the Regulations. 19 CSR 30-95.040(4)(H). The Department agreed that security questions in the applications specifically identified the minimum expectations for each question. The scoring methodology for answers to Questions 55- 60 also used a "0-10 Point Scale[.]" An applicant could receive a score of 0, 4, 7 or 10, based on whether the applicant's answer was Unsatisfactory, Satisfactory, Superior or Distinctive, respectively. Patterson explained how Hippos' identical answers in dispensary applications to Questions 55-60 offered not just benefits beyond the minimum expectations, but significant benefits. Based on Patterson's knowledge and experience, he testified that Hippos should have received a score of at least 7, rather than 4, on its responses to Questions 55-60 in each of its dispensary applications. If Hippos' answers to these questions had received Patterson's increased raw scores, dispensary applications HIPP9701 and 9702 each would have received a total score of 1,555.48, and applications HIPP9703-9705 each would have received a total score of 1,557.88. A total score of 1,555.48 would have placed both applications HIPP9701 and 9702 within the top 24 dispensary applications in Congressional District

1. A total score of 1,557.88 would have placed applications HIPP9703 and 9705 within

the top 24 dispensary applications in Congressional District 1, and application HIPP9704 within the top 24 applications in Congressional District 2.

19

The Department did not present any expert testimony of its own on any issue. It presented no evidence explaining why the expert opinions on scoring introduced by Hippos were unsubstantiated or incorrect. The Department introduced no evidence to explain why the higher scores received by Hippos for identical answers were incorrect, while lesser scores for the same answer were correct. Manufacturing Applications The correct scoring of Hippos' manufacturing applications involves Questions 64 and 68. Tina Szopinski (Szopinski) was assigned to score these questions in the manufacturing license applications and awarded Hippos a score of 4 on each question. Regarding Question 64, the Department's contract with Wise required it to assign application questions "to scoring team members with background and expertise necessary" to evaluate the question and response. Wise provided the Department with copies of the scorers' resumes. Szopinski's 13-page resume is the only evidence of her subject-matter expertise and qualifications to be a scorer. Szopinski's resume does not list or describe any experience or background in the cannabis industry. Her resume also does not list or describe any experience or background in business evaluation or analysis. The Training Manual instructed Szopinski and the other scorers that "[i]n the notes, you should explain how your own expertise and personal experience influenced how you arrived at the assigned score." Szopinski did not explain her expertise or anything else in the notes section of the scoring rubric because that section was left

20

blank. There was only one line in Szopinski's scorer's notes, which the Department chose to redact in the copy of the exhibit admitted in evidence. Jacob Cohen (Cohen) was assigned to score the answers to the same Question 64 on the cultivation and dispensary applications. He had experience as a business analyst, including analyzing companies in the cannabis industry. Based on that experience, Cohen awarded Hippos a score of 10 on its answers to this same question on all of its cultivation and dispensary applications. With respect to Question 68, the scoring methodology only permitted one of two scores. If the scorer found the answer satisfactory, a score of 10 was required. If the scorer found the answer unsatisfactory, a score of zero was required. Szopinski gave all of Hippos' manufacturing applications a score of 4 for this question, which was inappropriate. If Cohen's score on Question 64 had been used on Hippos' manufacturing applications instead of Szopinski's score, and Szopinski's score for Question 68 had been corrected to a 10, each of Hippos' manufacturing applications would have received a total score of at least 1,528.78, which is higher than the lowest-scoring manufacturing application to receive a license. 6

6 In the respondent's brief, the Department did not dispute the accuracy of this statement that Hippos' manufacturing applications would have been eligible to receive licenses.

21

Commission Decisions In denying Hippos' license applications, the Commission's three decisions used the same analysis. In the cultivation, manufacturing and dispensary applications, the undisputed evidence showed Hippos' identical answers to the same questions received different scores, which the Commission acknowledged. The Commission decided that these inconsistencies required that Hippos' answers be re-scored. The Commission also acknowledged that the scoring criteria and the scorers' determinations "hinged on the subjective valuations of the scorers." The Commission adopted the following approach to rescoring: Because the questions were scored subjectively, we look for evidence reflecting the scorer of that question's subjective valuation. The actual scores assigned to identical answers by these scorers represents the most concrete indication of these subjective preferences in the record. Expert testimony and other documentary evidence pale in comparison to the weight these scores carry for identical responses to the same question.

The Commission then re-scored the inconsistent answers by simply choosing the most common score among the inconsistent results, as illustrated below: Cultivation Applications Question HIPP9706 HIPP9707 HIPP9708 Commission Re-scores 26 4 7 4 4 27 10 4 4 4 41 7 0 7 7 48 4 4 7 4 49 7 7 10 7 50 4 7 7 7 51 10 7 10 10 52 7 10 10 10 66 7 7 10 7

22

Using this table, the Commission reasoned that, "[b]ecause Hippos received more 4s than 7s for question 26 [on its cultivation applications], the preponderance of the evidence is that 4 is the appropriate score for this question." After rescoring the inconsistent answers for all applications using this methodology, the Commission concluded that Hippos' corrected scores were not high enough to obtain the requested licenses. Discussion and Decision Hippos' brief contains five points for us to decide. Due to the intertwined nature of the points, we will consider them together. Points 1, 2 and 3 challenge the Commission's decision upholding the Department's denial of Hippos' two applications for cultivation licenses. Point 4 challenges the Commission's decision upholding the Department's denial of Hippos' six applications for manufacturing licenses. Point 5 challenges the Commission's decision upholding the Department's denial of Hippos' five applications for dispensary licenses. Collectively, these points contend the Commission's decisions affirming the Department's denial of Hippos' applications were arbitrary, capricious and unreasonable, and these decisions were not supported by competent and substantial evidence on the whole record. We agree. As the Commission acknowledged, the Department was required to "use a system of numerically scoring ten (10) additional evaluation criteria to rank the applications in each such license or certification category against each other." 19 CSR 30-95.025(4).

23

For most questions, based on the Department's scoring criteria, an applicant could receive 0, 4, 7 or 10 "raw" points on a question, based on whether the scorer believed the applicant's answer was Unsatisfactory, Satisfactory, Superior or Distinctive. The Department did none of the actual scoring. Instead, the scoring was done by Wise as the Department's contractual agent. Therefore, the Department is bound by Wise's actions. See Blanks v. Fluor Corp., 450 S.W.3d 308, 378 (Mo. App. 2014) (it is a fundamental rule of agency law that a principal is bound by the acts which his agent does with actual authority from the principal); see also 3 Am. Jur. 2d, Agency § 2. Pursuant to the Facility Scoring Guide, Wise's scorers were required to score applications "in an objective way using the predetermined point range and instructions outlined in the Facility Scoring Guide[.]" This document explicitly stated: It is imperative that every response is scored in a consistent manner for all applicants. Some of the questions and answers may require a scorer to use his or her own professional knowledge and expertise in scoring the responses. In doing so, a scorer must score every response consistently. For example, if two applicants applying for the same facility type provide identical responses to a question, the score must be the same.

The same principle was explicitly included in Wise's Work Plan, which was approved by the Department. The Work Plan stated that "[t]he most important part of [a scorer's] job is being consistent in grading so that each application is scored in the same manner and compared to other applications received within the specific facility type." The Training Manual, also approved by the Department, similarly stated: The most important part of your job is being consistent in your grading so that each application is scored in the same manner and compared to other applications received within the specific facility type. Conflict, challenges

24

to the process and lawsuits arise when inconsistent scoring results are evident; the companies will likely see their individual scores and will be able to compare across multiple application[s'] results to see for themselves if we are consistent. An example is if a company submits multiple applications with an identical answer to a certain question (on different applications) but receives a different score for each. This should NEVER happen, and would be considered a statistical anomaly. You can and should self-correct this.

(E mphasis in original.) In each of Hippos' applications, there were many instances in which identical answers to the same question received inconsistent scores. That should never have occurred if Wise's scorers had followed the instructions that they were given. Neither the Department nor Wise did anything to correct this situation. Hippos introduced expert testimony from two qualified experts, Picillo and Patterson, who had collectively prepared 83 applications for the Missouri program. More than half of their clients' applications successfully received licenses. These experts agreed that: (1) Hippos received inconsistent scores for its answers to the same questions; (2) based on expert analysis, a number of answers should have been given higher scores; and (3) if the higher scores had been assigned, Hippos would have qualified for all the cultivation and dispensary licenses it requested. The Department offered no rebuttal to Hippos' expert testimony and presented no other testimony showing why Hippos should not be given the higher scores about which those experts testified. In the Commission's three decisions denying relief, it acknowledged that Hippos' identical answers to the same questions had received different scores. Therefore, Hippos was entitled to have those answers re-scored. The Commission also acknowledged that

25

the scoring criteria and the scorers' determinations "hinged on the subjective valuations of the scorers." The Commission then re-scored the inconsistent answers by simply choosing the most common score among the inconsistent results. After rescoring the inconsistent answers using this methodology, the Commission concluded that Hippos' corrected scores were not high enough to obtain the requested licenses. The Commission's rescoring methodology was illogical, arbitrary and unreasonable for the following reasons. To avoid being arbitrary, unreasonable or capricious, an agency's decision must be made using some kind of objective data rather than mere surmise, guesswork, or a gut feeling. Bd. of Educ. of City of St. Louis v. Mo. State Bd. of Educ., 271 S.W.3d 1, 11 (Mo. banc 2008); Weisenborn ex rel. Shoemaker v. Missouri Dep't of Mental Health, 332 S.W.3d 288, 293 (Mo. App. 2011); Missouri Nat'l Educ. Ass'n v. Mo. State Bd. of Educ., 34 S.W.3d 266, 281 (Mo. App. 2000). An agency must not act in a totally subjective manner without any guidelines or criteria. Missouri Nat'l Educ. Ass'n, 34 S.W.3d at 281; see Bd. of Educ. of St. Louis, 271 S.W.3d at 11. The Commission correctly concluded that the Department's scoring criteria and the scorers' determinations were based on the subjective valuations of the scorers. This fundamental flaw infected the entire scoring process. After the Commission concluded that it was required to re-score Hippos' applications due to the noted inconsistencies, the Commission stated that it was searching for any "evidence" reflecting the scorers' subjective evaluation. Obviously,

26

any notes from a scorer about his or her reasons for giving a particular score would have been helpful to show consistency, as the Department's Deputy Director conceded. The Training Manual created by Wise, however, reminded scorers that any emails, notes, and other writings would be discoverable if any scores were challenged in court. Additionally, the Training Manual stated: "Don't write anything that you don't want everybody to read. Past versions (something deleted) will be discoverable[.] Adhere to this axiom: Say it and forget it; write it and regret it." Predictably, no notes concerning the scoring of Hippos' application answers were found. The only purpose that could be served by Wise's instruction to scorers was to frustrate later administrative and judicial review. This is yet another fundamental flaw in the Department's scoring process, which was not corrected by the Commission during its administrative review. The Commission compounded the fundamental subjectivity problem by adopting a rescoring methodology that was illogical, arbitrary and unreasonable. The Commission decided that the most important "evidence" to consider were the scores given for an answer. When those scores conflicted, the Commission simply chose the most common answer as the correct one to use for rescoring. In doing so, the Commission concluded these unexplained, inconsistent raw scores were better indicators of a correct score than the well-reasoned rescoring method used by Hippos' experts. The Commission's approach is completely flawed. The raw scores provide no evidence of the scorers' intent because there are no notes explaining why the scores were

27

given. The conflict in these unexplained scores cannot be reconciled by simply assuming the more common score for a particular answer is the correct one. This methodology also ignores the substantial, well-reasoned re-scoring testimony given by Hippos' experts. The Commission did not find that either expert lacked credibility or that the experts' methods were faulty. The Commission's decision is not supported by competent and substantial evidence upon the whole record. When an agency does not find an expert not credible, and the expert is not contradicted or impeached, the agency may not arbitrarily disregard and ignore competent, substantial and undisputed evidence. Williams v. City of Jennings, 605 S.W.3d 152, 160 (Mo. App. 2020). When an agency's decision is not based on substantial evidence, the agency acts unreasonably and arbitrarily. Bd. of Educ. of St. Louis, 271 S.W.3d at 11; Jefferson City Apothecary, LLC v. Mo. Bd. of Pharmacy, 499 S.W.3d 321, 330 (Mo. App. 2016); Stacy v. Dep't Of Soc. Servs., Div. of Med. Servs., 147 S.W.3d 846, 852 (Mo. App. 2004). "Likewise, the agency's failure to consider important aspects or factors of the issue can support a finding that the decision was arbitrary and capricious." Hundley v. Wenzel, 59 S.W.3d 1, 8 (Mo. App. 2001). After considering the entire record, there is not sufficient competent and substantial evidence to support the Department's denial of Hippos' applications or the Commission's decisions to affirm those denials. See Albanna v. State Bd. of Registration for Healing Arts, 293 S.W.3d 423, 428 (Mo. banc 2009).

28

We also note that Wise was contractually obligated to assign to each question a scorer who possessed the appropriate background and expertise to evaluate that question and response. The Work Plan stated that each scorer was to have the experience appropriate to perform their assigned task. Thus, the scores given to Hippos' answers were a form of expert testimony. The inconsistent scores should not have happened, and Hippos' experts provided in-depth testimony about why Hippos' answers should have received scores high enough to qualify its applications for licenses. The Commission's rescoring methodology substituted its own opinion about the proper score in a situation when that was beyond normal lay understanding. The Commission may not substitute its personal opinion for uncontradicted, qualified expert testimony on a topic that is not within the understanding of a lay person. See Lynch v. Treasurer of State, 635 S.W.3d 573, 584 (Mo. App. 2021); Williams, 605 S.W.3d at 159. Finally, with respect to Hippos' manufacturing applications, there were issues with Szopinski's scores on Questions 64 and 68. Szopinski was assigned to score these questions in the manufacturing license applications and awarded Hippos a score of 4 on each question. With respect to Question 64, we again note that Wise was required to make sure each scorer had the requisite experience and background to perform their assigned task. In the case of Szopinski, there is nothing in the record to show that this requirement was met. Szopinski's resume does not list or describe any experience or background in the cannabis industry or in business evaluation or analysis. The Training Manual instructed

29

Szopinski and the other scorers that "[i]n the notes, you should explain how your own expertise and personal experience influenced how you arrived at the assigned score." Szopinski did not explain her expertise or anything else in the notes section of the scoring rubric because that section was left blank. What the record does demonstrate without dispute is that Szopinski was not able to follow the simple grading methodology for a different question. Her score of 4 on Question 68 was inappropriate because this question was to be evaluated on a "Satisfactory(10)/Unsatisfactory(0)" basis. The record is similarly clear that Cohen did have the requisite experience and background to score Question 64. The Commission should have relied on Cohen's score of 10 for Hippos' manufacturing applications. With respect to Question 68, it is undisputed that Szopinski awarded Hippos some points, so her score was not a zero. Since the only other acceptable score was a 10, that score should have been corrected. If the Commission had used Cohen's score for Question 64 and had corrected Szopinski's score on Question 68, Hippos' manufacturing licenses would have been granted. For all the foregoing reasons, Hippos' five points have merit and are granted. Accordingly, we reverse the circuit court's judgment affirming the Commission's decisions denying Hippos' 13 license applications. Pursuant to the authority granted in 19 CSR 30-95.025(6)(C), we remand the case with directions that the circuit court order the Department to grant Hippos the cultivation, manufacturing and dispensary facility licenses for which it applied.

30

JEFFREY W. BATES, J. – OPINION AUTHOR DON E. BURRELL, J. – CONCUR JENNIFER R. GROWCOCK, C.J. – CONCUR