JURISDICTIONAL STATEMENT

The Equal Employment Opportunity Commission (“EEOC” or “Commission”) brought this enforcement action against Defendant Flambeau, Inc., pursuant to the A mericans with Disabilities Act of 1990 (“ADA”), as amended by the Americans with Disabilities Act of 2008, 42 U.S.C. §§ 12101, et seq. R.1 at 1 (Complaint).[1] The district court had jurisdiction under 28 U.S.C. §§ 1331, 1337, 1343, and 1345. Id. The district court entered final judgment as to all claims and parties on December 31, 2015. R.39. The EEOC timely appealed on February 25, 2016. R.46. See Fed. R. App. P. 4(a)(1)(B). This Court has jurisdiction under 28 U.S.C. § 1291.

Whether the district court erred in holding that Flambeau’s wellness program, which required employees to answer disability-related questions and undergo medical exams to enroll in the company’s health insurance plan, fell under the ADA’s “safe harbor” provision for insurance underwriting, 42 U.S.C. § 12201(c), even though the ADA explicitly prohibits employers from requiring employee medical exams or asking disability-related questions as part of an employee health program unless the exams and inquiries are “voluntary,” 42 U.S.C. § 12112(d)(4)?

STATEMENT OF THE CASE

A. Course of Proceedings

This is an employment discrimination case in which the EEOC alleged that Flambeau violated the ADA’s prohibition on involuntary medical exams and disability related inquiries, 42 U.S.C. § 12112(d)(4), by requiring employees to complete health risk assessment questionnaires (“HRAs”) and biometric tests as part of a wellness program. R.1. Flambeau filed a motion for summary judgment arguing that its actions were permissible under the ADA’s insurance “safe harbor” provision, 42 U.S.C. § 12201(c)(2), and that, in any event, the HRAs and biometric tests were voluntary. R.9; R.10, pp.13-18. The EEOC responded, R.24, and Flambeau filed a reply brief. R.34. The EEOC also filed a motion for partial summary judgment, R.15, R.16, which was fully briefed. R.27, R.31. On December 31, 2015, the district court granted Flambeau’s motion for summary judgment, denied the EEOC’s motion for partial summary judgment, and entered final judgment against the EEOC. A-1, R.38 (Order); A-16, R.39 (Judgment).

B. Statement of the Facts

Flambeau manufactures and sells plastic products. A-3, R.38, p.3 (Order). In 1990, Dale Arnold began working at Flambeau’s manufacturing facility in Baraboo, Wisconsin. Id. Arnold worked as a blow-mold operator, a cell leader, and a floor leader, where he assisted the supervisor. R.17-9, p.2 (2/8/12 Arnold letter). He earned uniformly positive evaluations and had excellent attendance. Id.

Flambeau’s health insurance plan

Flambeau offers group health insurance to its employee and their families. R.12, ¶ 3 (Rieland Aff.). Flambeau’s collective bargaining agreement (CBA) states that “[e]mployees will be provide[d] medical benefits as . . . more specifically described in a summary plan description (SPD).” R.17-18, p. 1 (CBA, ¶ 15.1(3)); R.30-9, pp.9-10 (Rieland Depo. 91-92). The CBA sets the employee premium contribution at “25% of [the] COBRA rate.” R.17-18, p.2 (CBA, ¶ 15.4); R.30-9, pp.9-10 (Rieland Depo. 92). Neither the CBA nor the summary plan description mention anything about employees needing to complete an HRA and biometric test in order to participate in the health insurance plan. R.30-9, p.10 (Rieland Depo. 92-93). Flambeau self-insures the health plan, which is administered by United Medical Resources (“UMR”). R.12, ¶¶ 3-4 (Rieland Aff.). Arnold regularly participated in the health insurance plan. A-3, R.38, p.3 (Order).

Flambeau implements a “wellness program”

Concerned about its rising health care costs, Flambeau established a “wellness program” in October 2010. R.12, ¶ 7 (Rieland Aff.); R.12-2 (10/7/10 memo.). The goal of the program was to educate employees about their health in order to encourage healthier habits and thereby reduce health care costs. R.35, ¶¶ 9-10 (reply to statement of facts (SOF)); R.12-2, p.1 (10/7/10 memo.). To encourage employees to participate in the program, Flambeau offered a $600 credit for the health insurance. R.12-2, p.2 (10/7/10 memo.).

The wellness program required employees to complete an HRA, which inquired as to the employee’s medical history, medications, diet, mental and social health, and job satisfaction. A-3; R.138, p.3.[2] Employees also had to undergo a biometric test, which required height and weight measurements, a blood draw (to measure cholesterol, HDL, LDL, glucose triglycerides, and nicotine levels), and blood pressure readings. R.35, ¶ 13 (reply to SOF); A-3, R.38, p.3 (Order). A third-party company, Health Solutions Ltd., administered the program for the 2011 benefits year. R.35, ¶ 17 (reply to SOF). Except as to tobacco use, the information gathered was reported to Flambeau in the aggregate. Id. ¶¶ 14-15. Based on results of the wellness program, Flambeau sponsored weight loss competitions, added healthier foods to the vending machines, charged reduced premiums to employees who tested negative for nicotine or participated in smoking-cessation programs, and offered personalized health and wellness coaching. Id. ¶¶ 29- 31; R.12-7, p.1 (chart) (showing 2011 premium rates, including increased rates for smokers).

In February 2011, Flambeau’s Director of Human Resources, Mark Rieland, learned that sixty percent of the company’s healthcare costs came from just twenty-seven individuals. R.12-6, p.2 (2/15/11 email). Rieland knew this because the company received summaries from its providers as to the claims costs of the plan participants. R.30-9, p.3 (Rieland Depo. 52-53). Rieland sent an email to his healthcare consultant, Bill Siehr, asking, “[W]hat do you recommend we do? Our budget is getting wacked!” R.12-6, p.2 (2/5/11 email). Siehr responded that the Affordable Care Act had “taken some of our ability to reduce costs away by requiring unlimited lifetime maximums and no annual limits on ‘essential benefits.’” Id. at p.1. Accordingly, Siehr said, the company was left with three options to “reduce costs.” Id. It could raise premiums or increase deductibles and co-payments or, Siehr said, the company could “[e]liminate or improve the underlying health conditions which drive the large claims.” Id. Siehr stressed that the greatest savings reduction from the third option would come from requiring participation in the company’s wellness program. Id. Siehr ended his email to Rieland by suggesting they “meet shortly to create a strategy . . . to reduce health costs.” Id. (emphasis added).

Flambeau makes the wellness program mandatory

Flambeau accepted Siehr’s advice and made the wellness program mandatory for those wanting health insurance for the 2012 benefit year. R.12, ¶ 17 (Rieland Aff.). Flambeau switched its wellness program administrator for the 2012 benefit year to Trotter Wellness. R.12, ¶ 14 (Rieland Aff.).

Towards the end of 2011, Flambeau posted notices warning employees that “participation with the biometrics and [health risk assessment] is mandatory to be on Flambeau’s medical insurance. Failure to show up at your scheduled time will result in disciplinary action.” R.28, ¶ 31 (response to SOF); R.17-17, p.1 (notice). The notice reminded employees they were required to “FAST FOR 9 HOURS BEFORE” their appointment R.17-17, p.2 (notice). Another notice stated, “IF YOU DO NOT ATTEND YOUR SCHEDULED BIOMETRICS TIME YOU WILL RECEIVE DISCIPLINARY ACTION” and “PARTICIPATION IN THE BIOMETRICS EXAM AND HRA PACKET ARE MANDATORY REQUIREMENTS TO BE ON THE MEDICAL INSURANCE PLAN.” R.28, ¶ 30 (response to SOF); R.17-17, p.3 (notice). Flambeau set December 14, 2011, as the date for employees to complete the HRA and biometric testing. R.28, ¶ 13 (response to SOF).

Neither spouses nor dependents were required to complete the HRA or the biometric testing. R.11-1, p.17 (Rieland Depo. 54). Figures from the 2012 plan year reveal that spouses comprised about 24.5% of the 1,132 plan participants (the total number of dependents is evidently not in the record), meaning that a sizable number of participants in Flambeau’s health plan were omitted from the HRAs and biometric testing. See R.35, ¶ 19 (reply to SOF); R.30-3, p.5, slides 1 & 3 (9/19/13 plan presentation showing that in benefit year 2012, 854 plan participants were employees and 278 were spouses). Because premiums for the 2012 benefit year had already been set, the results of the HRAs and biometric testing were not used to set premium rates for 2012. R.30-9, p.28 (Rieland Depo. 216-17).

In 2012, Trotter Wellness produced an Executive Report. R.12-4 (report). The report summarized results from the last two benefits years at Flambeau. Id. at pp.1-2. The Executive Summary reiterates that the goal of the wellness program was to improve employee health: “The ultimate goal of the health management program is to produce a net improvement of overall health for the profile population.” Id. at p.11. The report reflects that in 2011, when employees were given a $600 incentive to participate, 97% of all eligible Flambeau employees participated in the wellness program; in 2012, when participation became mandatory to have health insurance, 99% of employees participated. R.12-4, p.2; R.30-9, p.32 (Rieland Depo. 267). In Baraboo, Wisconsin, 78% of employees participated in 2011 while 99% participated in 2012. R.12-4, p.3 (report). The report identifies “risk factors” for the Baraboo employees, and includes a chart listing the percentage of employees at risk for nutrition (eating too few fruits and vegetables), weight, cancer, fitness, coronary, safety, and stress. Id. Nothing in the report concerns actuarial or risk analysis data to be used for underwriting or for setting premium rates. R.12-4 (report); R.30-9, p.28 (Rieland Depo. 219).

Dale Arnold’s participation in the wellness program

Arnold participated in the wellness program in 2011. R.35, ¶ 40 (reply to SOF). He was absent on December 14, 2011, however, because he had been hospitalized for congestive heart failure. R.28, ¶ 13 (response to SOF). Arnold returned to work on December 26, 2011. Id. ¶ 14. The next day he contacted the Human Resources department and asked about the procedure for taking the HRA and biometric test. R.17-9, p.3 (2/8/12 letter). He was told the materials had been sent to his home and he should wait for their arrival. Id.

On January 6, 2012, Arnold met with Flambeau’s Benefits and Compensation Specialist, Katie Axelsen. R.28, ¶ 15 (response to SOF). Arnold, who had not received the HRA and biometric test materials, asked Axelsen about obtaining them. R.17-9, p.3 (2/8/12 letter). Axelsen refused to provide the materials. Id. She instead told Arnold, incorrectly, that the materials had been sent to him in early December (they in fact had been sent on December 22, 2011, to his previous address and had not reached him). R.28, ¶ 18 (response to SOF); R.17-9, p.3 (2/8/12 letter). When Arnold tried to explain that he had not received the materials, Axelsen told Arnold he should have been present on December 14, 2011, for the testing. R.17-9, p.3 (2/8/12 letter). She added that “rules are rules and no exceptions will be made,” and she said she had decided to terminate his insurance as of December 31, 2011. Id.

True to her word, on January 9, 2012, Axelsen sent Arnold a letter stating that “effective December 31, 2011, your medical insurance coverage will [sic] be terminated” because of his failure to complete the HRA and biometric testing. R.17-9, p.6 (1/9/12 letter). The letter added that Arnold “will be receiving COBRA information from UMR in the next couple of weeks.” Id. On January 11, 2012, UMR sent Arnold a COBRA notification letter stating that he had been terminated and that if he did not elect to continue coverage at the COBRA rate, his insurance would be terminated as of January 1, 2012 (ten days earlier). R.11-5, p.1 (1/11/12 letter). The letter further stated that the full monthly COBRA payment was $409.30. Id. at p.3. This represented a 334% increase over the $94.44 monthly payment Arnold would have owed, in accordance with the CBA, had he submitted to the HRA and biometric test. R.12-7, p.1 (premium chart). Arnold did not sign up for COBRA coverage because he could not afford it. R.11-3, p.13 (Arnold. Depo. 104). Arnold finally received the HRA materials on January 14, 2012. R.28, ¶ 19 (response to SOF).

Arnold’s complaints to DOL and EEOC

Arnold filed a complaint with the United States Department of Labor (DOL). R.35, ¶ 46 (reply to SOF); R.17-21, p.3. After discussions with DOL, Flambeau agreed to reinstate Arnold’s insurance upon completion of the HRA and biometric test and payment of his back premiums at the regular 25% COBRA rate. R.12, ¶¶ 33-34 (Rieland Aff.). Arnold fulfilled these requirements on approximately May 8, 2012, and Flambeau reinstated his insurance retroactive to January 1, 2012. R.17-22, p.1 (5/8/12 memo); R.12, ¶ 34 (Rieland Aff.). Arnold participated in Flambeau’s health insurance plan through his tenure with the company. R.12, ¶ 37 (Rieland Aff.).

On April 26, 2012, Arnold filed a charge of discrimination with the EEOC. R.17-23, p.3 (charge). He alleged that Flambeau violated the ADA when it required him to participate in an involuntary wellness program with prohibited medical inquiries and exams and terminated his health insurance. Id. The EEOC notified Flambeau of the charge and requested a position statement from the company. R.17-23, p.1 (notice).

On July 3, 2012, Flambeau submitted a position statement denying that it “terminated Mr. Arnold’s medical insurance for not participating in its wellness program.” R.17-25, p.1 (position statement). Rather, Flambeau contended, it offered Arnold the option of continuing his insurance at the full COBRA rate. Id. at p.2. Only new employees, Flambeau stated, were denied health insurance if they chose not to complete the HRA. Id. at p.3. Flambeau further asserted that the “sole purpose” of the testing was to provide “an employee benefit to the employee[s] to learn about their own medical conditions so that they can be fully informed for their own healthcare needs.” Id. at p.2. Flambeau emphasized at the end of the position statement that its goal was to improve employee health. Id. at p.3. Healthier employees, Flambeau said, “are more productive and consume less health care services,” and, Flambeau added, “these efforts benefit the employees” by giving them “the opportunity to lead healthier, happier lives.” Id. The position statement nowhere asserts that Flambeau utilized the results of the HRA and biometric testing to underwrite risks, classify risks, or administer risks in connection with its insurance plan. R.17-25, pp.1-3. To the contrary, Flambeau asserted that it “does not use the results” to determine employee eligibility for insurance. Id. at p.3.

Although Flambeau made the HRA and biometric testing mandatory in an effort to reduce costs, its premiums increased for both 2012 and 2013. R.36, ¶ 36 (response to SOF). Flambeau discontinued the wellness program in 2014 after determining that it did not, in fact, change employee health habits or reduce costs. R.30-9, p.29 (Rieland Depo. 221, 223).

EEOC files suit

The EEOC filed suit under the ADA. R.1 (Complaint). The EEOC alleged that Flambeau’s wellness program violated the prohibition at 42 U.S.C. § 12112(d)(4)(A) against employee medical exams and disability-related inquiries that are not job-related and consistent with business necessity. Id. ¶¶ 9, 17. The EEOC also alleged that the exams and inquiries were not “voluntary” and therefore did not fall under the exception at 42 U.S.C. § 12112(d)(4)(B) for employee health programs. Specifically, the EEOC alleged that exams and inquiries were involuntary because Flambeau terminated Arnold’s health insurance and imposed a financial penalty on him (requiring him to pay the entire premium cost under COBRA to re-enroll) when he failed to complete the exam and answer the inquiries, and because Flambeau threatened employees with disciplinary action for failing to complete the testing. Id. ¶ 18.

Rieland’s statements about the HRAs and biometric testing

During discovery, Rieland testified in more detail about the purpose and use of the mandatory HRAs and biometric testing. He explained that the purpose of making the HRAs and biometric testing mandatory was to “get people to take care of themselves” and “curb diseases before they get beyond control.” R.30-9, p.3 (Rieland Depo. 52). Several times during his deposition, he confirmed that the purpose of the mandatory HRA and biometric testing was to save money by making employees healthier. R.30-9, pp.3-4 (Rieland Depo. 53, 55, 56). But, Rieland admitted, he was never given any data showing the purported cost savings generated by the HRAs and biometric testing. R.30-9, p.27 (Rieland Depo. 213-14). Rieland also conceded that the Trotter Wellness report is devoid of any actuarial or risk analysis data to be used with “setting premiums or underwriting.” R.30-9, p.28 (Rieland Depo. 219).

Significantly, Rieland agreed in his deposition that the data gathered from the HRAs and biometric tests “was not used . . . to decide what benefits individuals should get under the plan” and “was not used . . . to determine what individuals would pay under the plan.” R.30-9, p.5 (Rieland Depo. 71) (emphasis added); see also R.30-9, p.6 (Rieland Depo. 72) (agreeing that the HRA and biometric testing data “was not used at any time to determine either the type of benefits that would be provided under the healthcare plan or . . . premiums”). As to premiums, Rieland conceded that the next year’s premium amounts were the “direct result” of the prior year’s benefits expenses. R.30-9, p.27 (Rieland Depo. 213).

Rieland also confirmed that Flambeau did not need the data from the HRAs or biometric tests to determine what its costs for the health insurance plan were or to learn which participants were suffering from which diseases. Specifically, he stated that even before 2011, Flambeau’s providers gave the company information as to the actual claims incurred by plan participants. R.30-9, p.3 (Rieland Depo. 52-53). Thus, Flambeau already knew—even before requiring the HRA and biometric testing—which people had epilepsy, seizure disorders, etc. R.30-9, p.5 (Rieland Depo. 69). Rieland acknowledged that Flambeau was prohibited by law, however, from using an individual’s health factor to determine whether to insure him or her, to set premium rates, or to determine what benefits to provide. R.30-9, pp.5-6 (Rieland Depo. 71-72).

Only in response to prompting by Flambeau’s counsel did Rieland testify that the HRA and biometric testing data was used to classify or administer risks. When asked by Flambeau’s counsel whether the company used the HRA and biometric test results in “classifying risks or administering risks,” Rieland asked his counsel to “clarify what [he] meant by ‘classify risks.’” R.30-9, p.30 (Rieland Depo. 256). Flambeau’s counsel then referred Rieland to the Trotter Wellness report and asked Rieland whether the report addressed “risk factors for employees at Flambeau?” Id. Rieland responded that it did. R.30-9, p.30 (Rieland Depo. 256-57). After reviewing the report, Rieland also agreed with Flambeau’s counsel that the HRA and biometric testing data was used to classify and administer risks, to manage future costs, and to mitigate risks by encouraging healthy employee behavior. R. 30-9, p.30 (Rieland Depo. 259). When asked by Flambeau’s counsel to provide an example of a program the company implemented to control its costs based on the “risk analysis” contained in the report, Rieland responded that Flambeau added healthier vending food options, subsidized the cost of healthier food items, and “did some weight loss competitions.” R.30-9, p.34 (Rieland Depo. 273-74).

Rieland also agreed that switching from a $600 incentive to participate in the wellness program to being forced to pay 100% of the COBRA rate for health insurance

could be seen as a penalty. R.30-9, p.25 (Rieland Depo. 207).

Hames’ Affidavit

Flambeau submitted an affidavit from Sara Hames, the Vice President for Hays Companies of Wisconsin (“Hays”), to support the company’s litigation position that the aggregated data from the HRA and biometric tests was used to make insurance decisions. R.13, ¶ 1 (Hames Aff.). Hays replaced Bill Siehr & Associates as the company’s healthcare consultant in August 2011. R.13, ¶ 6 (Hames Aff.); R.12, ¶ 13 (Rieland Aff.). Hames’ affidavit explains that she regularly consults with employers regarding the implementation and use of wellness programs. R.13, ¶ 3 (Hames Aff.). She conceded that the “central goal” of wellness programs with biometric screenings and HRAs is “to improve” employee health and thereby “lower the businesses’ benefits expenses.” Id. ¶ 4.

Hames reviewed aggregate data from the HRAs and biometric tests conducted in the fall of 2010 (when the HRAs and biometric tests were not mandatory). Id. ¶ 7. This data showed that 30% of participating employees had low wellness scores and that a significant percentage of employees were obese or overweight. Id. ¶ 7. She and her colleagues “used this data along with other information, including additional statistics derived from the aggregate HRA and biometric testing . . . as a basis to recommend that Flambeau purchase stop-loss insurance to protect its self-funded health plan from . . . catastrophic claims (of which poor health or excessive weight are often contributing factors).” Id. ¶ 8. Hays then incorporated the cost of the stop-loss insurance into Flambeau’s health plan. Id. “Based upon this cost,” and information as to “the insurance claims faced by the Company and other data,” Hays made premium recommendations “for the 2012 benefits year.” Id.

In early 2012, Hays reviewed the aggregate data from the fall 2011 HRAs and biometric testing. Id. ¶ 9. That data revealed that the health risk results of Flambeau’s employees “were mostly worse than the previous year,” not better. Id. Accordingly, Hays recommended that Flambeau add an onsite clinic to the Baraboo facility, pay 100% of most preventative exams, and lower co-pays for maintenance drugs. Id. Hames hoped these changes would reduce “costs and premiums.” Id.

In 2013, Hames and her colleagues again made recommendations to Flambeau for the 2014 benefit year. Id. ¶ 11. She explained that they went through their “standard process of classifying and assessing risks and projecting plan costs” for Flambeau’s plan. Id. “Then [they] built in cost/savings of plan design changes made the previous year, some of which were influenced by the aggregate results” of the HRA and biometric testing, and they estimated costs for the 2014 benefit year. Id. (emphasis added). Using the cost data, Hays made “premium/premium equivalent rate recommendations” to Flambeau. Id.

C. District Court Decision

The district court granted Flambeau’s motion for summary judgment on the ground that the HRAs and biometric testing fell under the ADA’s “safe harbor” provision, 42 U.S.C. § 12201(c). A-1, R.138 (Order). After noting that the applicability of the safe harbor provision to wellness programs is a matter of first impression in this Circuit, the court turned to Seff v. Broward County, 778 F. Supp. 2d 1370 (S.D. Fla. 2011), aff’d 691 F. 3d 1221 (11th Cir. 2012). Id. at 6. In Seff, the court noted, the Eleventh Circuit affirmed the district court’s holding that the safe harbor provision permitted the employer to charge a $20/week surcharge to employees who participated in the health insurance plan but refused to complete an HRA and biometric test. Id. at pp.6-7. The district court found Seff persuasive, notwithstanding the EEOC’s arguments that the court’s broad reading of the safe harbor provision nullifies the voluntary employee health program exception of § 12112(d)(4)(B). Id. at 8.

The district court acknowledged the EEOC’s proposed wellness regulation, wherein the EEOC stated in the preamble to the proposed regulation that Seff was wrongly decided. Id. at 8 (citing 80 Fed. Reg. 21659, 21622 n.24 (April 20, 2015)). But, the court said, it was not bound by a proposed regulation and, in any event, the regulation did not speak to the applicability of the safe harbor provision to wellness programs used for underwriting. Id. The court also rejected the EEOC’s argument that traditional principles of statutory interpretation required a narrow construction of § 12201(c)(2) as an exception to the prohibitions of § 12112(d)(4). Id.

Having concluded that the safe harbor provision can apply to wellness programs, the court next held that it applied to Flambeau’s wellness program. The court first ruled that the wellness program was a “term” of Flambeau’s insurance plan because Flambeau made it a condition of enrolling in the health insurance plan. Id. at 10. Next, the court held that Flambeau used the wellness program for “underwriting risks, classifying risks, or administering risks,” reasoning that Flambeau’s consultants used the data to classify participants’ health risks, calculate Flambeau’s projected insurance costs, recommend prices for maintenance medications and preventative care, recommend premium rates, and recommend purchase of stop-loss insurance. Id. at 12. Finally, the court held that Flambeau had not used the safe harbor as a “subterfuge,” as Flambeau had not used the wellness program to make disability-related distinctions with respect to employee benefits. Id. at 14-15. The court stated that its ruling made it unnecessary for it to determine whether the HRA and biometric testing were “actually ‘required’ in the manner prohibited by § 12112(d)(4)(A).” Id. at 1.

SUMMARY OF ARGUMENT

The EEOC alleged that Flambeau’s decision to make health risk assessments and biometric tests mandatory for employees wanting health insurance violated 42 U.S.C. § 12112(d)(4), which prohibits disability-related inquiries and medical exams of employees unless they are job-related and consistent with business necessity or are a voluntary part of an employee health program. The district court, however, never reached the critical issue of whether Flambeau’s mandatory HRAs and biometric tests were “voluntary” components of its wellness program. Instead, the district court ruled that the ADA’s “safe harbor” provision, 42 U.S.C. § 12201(c), protected Flambeau’s actions. The court erred. The safe harbor provision permits insurance companies or organizations to administer the “terms” of a bona fide benefit plan that are based on “underwriting risks, classifying risks, or administering such risks” without running afoul of the statute’s prohibitions, unless the provision is used as a subterfuge. The EEOC’s long-standing position, which is consistent with the ADA’s text and legislative history, is that the insurance safe harbor provision does not apply to § 12112(d)(4)(B), which permits disability-related inquiries and medical exams only as part of a voluntary employee health program.

But even if § 12201(c) could provide safe harbor to some employer wellness programs that would otherwise violate § 12112(d)(4)(A), there is no safe harbor in this case for Flambeau’s mandatory HRAs and biometric tests. Contrary to the district court’s conclusion, Flambeau failed to establish on this record that it used the HRA and biometric test data for “underwriting risks, classifying risks, or administering such risks.” Rather, Flambeau adopted the HRA and biometric test requirement to make its employees healthier with the goal of reducing health care costs. That is not underwriting.

The court also erred in holding that the mandatory HRAs and biometric tests were “terms” of Flambeau’s insurance plan. Neither the collective bargaining agreement nor the summary plan description made eligibility for the insurance plan contingent upon completion of an HRA and biometric test. Finally, even if the HRAs and biometric tests constituted “terms” of the plan used for “underwriting,” the safe harbor provision is inapplicable because the record makes clear that Flambeau invoked it as a subterfuge to avoid the prohibition at § 12112(d)(4) on involuntary medical exams and disability-related inquiries.

STANDARD OF REVIEW

This Court reviews a grant of summary judgment de novo, construing all facts and drawing all reasonable inferences in the nonmoving party’s favor. Burnell v. Gates Rubber Co., 647 F.3d 704, 707 (7th Cir. 2011). “The party seeking summary judgment has the burden of establishing that no genuine dispute exists as to any material fact.” Cung Hnin v. TOA (USA), LLC, 751 F.3d 499, 504 (7th Cir. 2014) (citing Celotex Corp. v. Catrett, 477 U.S. 317, 323 (1986)). Summary judgment is appropriate only “if the movant shows that there is no genuine dispute as to any material fact and the movant is entitled to judgment as a matter of law.” Spurling v. C & M Fine Pack, Inc., 739 F.3d 1055, 1060 (7th Cir. 2014) (quoting Fed. R. Civ. P. 56(a)).

ARGUMENT

The court erred in holding that the ADA’s “safe harbor” provision

permits Flambeau to require its employees to complete health risk

assessments and biometric tests as part of a wellness program.

A. The insurance safe harbor provision does not apply to wellness programs.

The district court’s holding that the ADA’s safe harbor provision applies to wellness programs contravenes the text of the statute, the legislative history, and the EEOC’s long-standing interpretation of the ADA. Contrary to the district court’s ruling, one of the core purposes of the ADA was to prohibit involuntary medical exams and disability-related inquiries. The court’s ruling that the safe harbor provision permits employers to evade this prohibition is incorrect and threatens to eviscerate one of the core protections of the statute. This Court should therefore hold that the safe harbor provision of § 12201(c) does not allow employers to evade the statute’s prohibition at § 12112(d)(4)(A) on employee medical exams and disability-related inquiries.

1. The ADA’s text is plain.

The first step in interpreting a statute is to determine whether the language has a “plain and unambiguous” meaning. Robinson v. Shell Oil Co., 519 U.S. 337, 340 (1997). This determination is made with reference “to the language itself, the specific context in which that language is used, and the broader context of the statute as a whole.” Id. at 341. Congress addressed explicitly the question of whether employers can subject employees (and applicants) to medical exams and disability-related inquiries at § 12112(d). Section 12112(d)(1) states that the statute’s prohibition on disability discrimination at § 12112(a) includes “medical examinations and inquiries.” Section 12112(d)(4)(A) repeats the prohibition, stating that “[a] covered entity shall not require a medical examination and shall not make inquiries of an employee as to whether such employee” has a disability “or as to the nature or severity of the disability, unless such examination or inquiry is shown to be job-related and consistent with business necessity.” 42 U.S.C. § 12112(d)(4)(A).

Section 12112(d)(4)(B) contains a second exception, as it permits employers to “conduct voluntary medical examinations, including voluntary medical histories, which are part of an employee health program.” 42 U.S.C. § 12112(d)(4)(B) (adding that an employer may inquire as to an employee’s ability to perform job-related functions). Thus, Congress addressed comprehensively at § 12112(d)(4) the circumstances under which employers can require medical exams and ask disability-related questions of an employee, opting to include only two “safe harbors” to the prohibition against employee exams and inquiries: job-related and consistent with business necessity, or part of a voluntary employee health program. Significantly, Congress did not allude in § 12112(d)(4)(A) to a third “safe harbor” for employee health programs used for insurance underwriting.

The district court, however, created a third safe harbor for employers seeking to avoid the ban on employee medical exams and disability-related inquiries. According to the district court, it did not need to decide whether Flambeau’s wellness program was “voluntary” under § 12112(d)(4)(B) because the wellness program fell under the “safe harbor” provision of § 12201(c). The court erred. This provision applies to insurance underwriting and was never intended to allow employers to circumvent the prohibition of § 12112(d)(4) against employee medical exams and inquiries. Section 12201(c)(2) states that Titles I (employment), II (public services), and III (public accommodation) of the ADA “shall not be construed to prohibit or restrict . . . a person or organization . . . from establishing, sponsoring, observing or administering the terms of a bona fide benefit plan that are based on underwriting risks, classifying risks, or administering such risks . . . .” Section 12201(c) further provides, however, that this provision cannot be used as a “subterfuge to avoid the purposes” of the ADA. See also 29 C.F.R. § 1630.16(f) (EEOC’s regulation mirroring the statute).

Although the ADA does not define “underwriting risks, classifying risks, or administering such risks,” courts have recognized that “underwriting generally refers to the application of the various risk factors or risk classes to a particular individual or group for the purposes of determining whether to provide coverage.” Zamora-Quezada v. HealthTexas Med. Grp. of San Antonio, 34 F. Supp. 2d 433, 443 (W.D. Tex. 1998) (citing EEOC’s Interim Enforcement Guidance on the Application of the Americans with Disabilities Act of 1990 to Disability-Based Distinctions in Employer Provided Health Insurance, 1993 WL 1497027, at *6 n.15 (June 8, 1993) (“Interim Guidance”)). Courts have also recognized that “[r]isk classification refers to the identification of risk factors and the grouping of those factors which pose similar risks.” Id. (relying on the EEOC’s Interim Guidance, 1993 WL 1497027, at *6 n.15). Thus, the purpose of the safe harbor provision was “to permit the development and administration of benefit plans in accordance with accepted principles of risk assessment.” Barnes v. Benham Grp., Inc., 22 F. Supp. 2d 1013, 1020 (D. Minn. 1998) (citing 29 C.F.R. pt. 1630, App. § 1630.16(f)).

2. The legislative history is consistent with the text.

The legislative history confirms that Congress intended § 12201(c) to permit the insurance industry to continue its traditional underwriting practice—based on risk classification—without running afoul of the ADA’s prohibition on disability discrimination, not to create a third safe harbor to the prohibition of § 12112(d)(4) on employee medical exams and disability-related inquiries. The House Committee on Education and Labor Report states that § 12201(c)(2) was added “to make it clear that this legislation will not disrupt the current nature of insurance underwriting or the current regulatory structure for self-insured employers or of the insurance industry in sales, underwriting, pricing, administrative and other services, claims, and similar insurance related activities based on classification of risks as regulated by the States.” H.R. Rep. No. 101-485, pt.2, at 136 (1990), reprinted in 1990 U.S.C.A.A.N. 303, 419 (emphasis added); see also S. Rep. No. 101-116, at 84 (1989) (“The Committee does not intend that any provision of this legislation should affect the way the insurance industry does business [under] State laws.”), reprinted in H.R. Comm. on Education and Labor, 101^st Cong., 2d Sess., Legislative History of Public Law 101-336 at p.137 (Comm. Print 1990). The House Report further states that under the safe harbor provision, a plan could lawfully limit coverage “based on classification of risk” but could not exclude an individual, limit coverage, or charge a different premium rate “solely because of a physical or mental impairment, except where . . . based on sound actuarial principles or [the action] is related to actual or reasonably anticipated experience.” H. Rep. No. 101-485, pt.2, at 136-37, 1990 U.S.C.A.A.N. at 419-20. Thus, the House Report explained, a blind person could not be denied coverage “based on blindness independent o[f] actuarial risk classification.” Id. at 137, 1990 U.S.C.A.A.N. at 420. Likewise, a group health insurance plan could deny coverage for an individual’s pre-existing condition but not for illness or injury unrelated to that condition. Id. In sum, the House Report says, § 12201(c) was “intended to afford to insurers and employers the same opportunities” they enjoyed before the ADA “to design and administer insurance products and benefit plans . . . consistent with basic principles of insurance risk classification.” Id. at 137-38, 1990 U.S.C.A.A.N. at 420-21.

Thus, nothing in the legislative history suggests that Congress intended the safe harbor provision to apply to wellness programs. The evidence is, in fact, to the contrary. The sole reference to wellness programs appears in the House Report’s discussion of the ADA provision governing voluntary health programs. The House Report acknowledges that an increasing number of employers “are offering voluntary wellness programs” that include medical screenings for high blood pressure, weight, and cancer, etc. Id. at 75, 1990 U.S.C.A.A.N. at 357. The House Report states that such programs are lawful “[a]s long as the programs are voluntary and the medical records are maintained in a confidential manner and not used for the purpose of limiting health insurance eligibility or preventing occupational advancement.” Id. This passage demonstrates that Congress’ understanding was that employee medical exams and disability-related inquiries were permissible as part of a wellness program only when “voluntary”; Congress did not intend them to be permissible under the insurance safe harbor provision as well.

3. The EEOC’s long-standing view is entitled to deference.

Consistent with the text and legislative history, the EEOC’s long-standing view is that the insurance safe harbor provision is inapplicable to § 12112(d)(4)’s prohibition on involuntary medical exams and disability-related inquiries that are part of an employee health program. The EEOC has expressed this view both implicitly and explicitly in regulations and guidance documents. Thus, even if this Court disagrees with the EEOC and concludes that the text of the statute is ambiguous as to whether the insurance safe harbor provision applies to § 12112(d)(4), this Court should defer to the EEOC’s consistent and reasonable view that it does not apply.

The EEOC has substantive rule-making authority under Title I the ADA, 42 U.S.C. § 12116. In 1991, the EEOC issued regulations. Consistent with the statute, 29 C.F.R. § 1630.13(b) states that employee exams and inquiries are unlawful, except as permitted by § 1630.14. Section 1630.14 is titled “Medical examinations and inquiries specifically permitted.” 29 C.F.R. § 1630.14. Also in keeping with the statute, the regulation states that exams and inquiries are permitted of employees if job-related and consistent with business necessity or as a voluntary part of an employee health program. 29 C.F.R. § 1630.14(c)-(d). Significantly, nothing in § 1630.14 hints that the agency understood the insurance safe harbor to provide a third route by which employers can require employee exams and inquiries.

The EEOC’s regulation at 29 C.F.R. § 1630.16(f) restates the statute’s insurance safe harbor exception without mentioning wellness programs.[3] That silence, as with the silence of 29 C.F.R. § 1630.14, likewise suggests that the insurance safe harbor provision is inapplicable to wellness programs. The interpretive guidance to § 1630.16(f) further supports this view, as it emphasizes the limited nature of the safe harbor exception. Specifically, the interpretive guidance provides that the safe harbor “is a limited exemption” applicable only to “those who establish, sponsor, observe, or administer benefit plans.” 29 C.F.R. pt. 1630, App. § 1630.16(f). Mirroring the legislative history, the interpretive guidance further states that “[t]he purpose of this provision is to permit the development and administration of benefit plans in accordance with accepted principles of risk assessment. This provision is not intended to disrupt the current regulatory structure for self-insured employers,” who remain free to administer the terms of a bona fide benefit plan. Id. Further, the interpretive guidance confirms that the purpose of the provision was to permit the “current insurance industry practices in . . . underwriting . . . based on classification of risks as regulated by the States.” Id.

To be sure, the interpretive guidance does not state explicitly that the insurance safe harbor provision is inapplicable to § 12112(d)(4)’s prohibition on involuntary medical exams and inquiries that are part of an employee health program. But that is the clear import of the guidance, which emphasizes that the safe harbor provision is a “limited” exemption intended only to allow the continuation of traditional underwriting practices. The interpretive guidance is particularly persuasive because it resulted from notice-and-comment rulemaking. Specifically, on August 1, 1990, the EEOC issued an advance notice of proposed rulemaking stating that the agency had begun the process of developing substantive regulations pursuant to the newly enacted ADA. See Title I of Americans with Disabilities Act; Implementation, 55 Fed. Reg. 31192 (August 1, 1990). After receiving public comments and soliciting input, the EEOC published a notice of proposed rulemaking setting forth the agency’s proposed regulation at 29 C.F.R. § 1630. See Equal Employment Opportunities for Individuals with Disabilities, 56 Fed. Reg. 8578-01 (proposed February 28, 1991) (to be codified at 29 C.F.R. § 1630). The proposed rule included 29 C.F.R. § 1630.16(f), the provision mirroring the ADA’s safe harbor provision of § 12201(c), as well as the interpretive guidance calling the safe harbor a “limited exemption” whose “purpose . . . is to permit the development and administration of benefit plans in accordance with accepted principles of risk assessment.” 56 Fed. Reg. at 8603. After receiving 697 comments on the proposed regulation, the EEOC adopted 29 C.F.R. § 1630 and the interpretive guidance. See Equal Employment Opportunities for Individuals with Disabilities, 56 Fed. Reg. 35726-01 (July 26, 1991) (codified at 29 C.F.R. § 1630).

The EEOC’s interpretive guidance is “entitled to respect” because it has the “power to persuade.” Skidmore v. Swift & Co., 323 U.S. 134, 140 (1944) (stating that agency “interpretations and opinions” “while not controlling . . . do constitute a body of experience and informed judgment to which courts . . . may properly resort for guidance”). Here, the EEOC’s interpretive guidance is persuasive because of “the thoroughness evident in its consideration, the validity of its reasoning, [and] its consistency with earlier and later pronouncements.” Skidmore, 323 U.S. at 140. As discussed, that interpretation is consistent with the statute, was the product of thorough consideration (including notice-and-c0mment rulemaking), and is consistent with the EEOC’s earlier and later pronouncements as to scope of the safe harbor provision.

Other EEOC guidance documents have also consistently taken the view, either implicitly or explicitly, that the insurance safe harbor provision does not apply to wellness programs. These guidance documents are also entitled to respect to the extent they have the power to persuade. See Skidmore, 323 U.S. at 140 (stating that agency “interpretations and opinions . . . constitute a body of experience and informed judgment to which courts . . . may properly resort for guidance”); Federal Express Corp. v. Holowecki, 552 U.S. 389, 399 (2008) (stating that the EEOC’s “policy statements, embodied in its compliance manual and internal directives” “are entitled to ‘a measure of respect’ under . . . Skidmore”). More than a dozen years ago, the EEOC issued its “Enforcement Guidance: Disability-Related Inquiries and Medical Examinations of Employees Under the Americans With Disabilities Act (ADA)” (“Guidance”). See Guidance, No. 915.002 (July 27, 2000), available at http://www.eeoc.gov/policy/docs /guidance-inquiries. html. This Guidance, which is in the EEOC’s Compliance Manual, states at Question # 22 that an employer may “conduct voluntary medical examinations and activities, including voluntary medical histories” that are not job-related or consistent with business necessity as part of an employee health program. See id. at Q/A # 22 (emphasis added). Significantly, the Guidance makes no mention whatsoever of the insurance safe harbor, implicitly setting forth the Commission’s view that exams and inquiries are permissible as part of an employee health program only if they are “voluntary,” not if they are used for underwriting.

Consistent with the Guidance, the EEOC issued an informal discussion letter in 2009 stating that employee medical exams and disability-related inquiries are permissible if they are job-related and consistent with business necessity or are part of a voluntary employee health program. See Peggy R. Mastroianni, March 6, 2009, ADA: Disability-Related Inquiries and Medical Examinations; Health Risk Assessment (2009), available at https://www.eeoc.gov/eeoc/foia/letters/ 2009/ada_disability_ medexam_ healthrisk.html. The discussion letter responded to an inquiry asking whether an employer may require its employees to participate in an HRA (which included a health-related questionnaire, blood pressure test, and blood draw) as a condition of enrolling in the company’s health insurance plan. The discussion letter acknowledges that the EEOC “has not taken a formal position on the question” but goes on to state that such a requirement “does not appear to be job-related and consistent with business necessity, and therefore would violate the ADA.” Id. While exams and inquiries are permitted “as part of a voluntary wellness program,” the letter notes, the HRA was not voluntary, as individuals who refused to participate were penalized by being denied the opportunity to enroll in the health insurance plan. Id. Thus, although the letter does not state explicitly that the insurance safe harbor provision would not apply, the letter implicitly suggests this by failing to mention the insurance safe harbor as a third avenue by which employers can lawfully require employee medical exams and make disability-related inquiries.

More recently, the EEOC issued a proposed regulation to amend 29 C.F.R. § 1630.14 that states explicitly in the preamble that the insurance safe harbor provision does not apply to wellness programs. See Amendments to Regulations Under the Americans with Disabilities Act, 80 Fed. Reg. 21659 (proposed April 20, 2015) (to be codified at 29 C.F.R. § 1630.14).[4] The focus of the proposed regulation is to “provide guidance on the extent to which employers may use incentives to encourage employees to participate in wellness programs that include disability-related inquiries and/or medical examinations.” 80 Fed. Reg. at 21659. The preamble to the proposed regulation, however, addresses the question of whether the safe harbor provision applies to wellness programs. Specifically, footnote twenty-four of the preamble states that Seff was wrongly decided and that the safe harbor provision is not “the proper basis for finding wellness program incentives permissible.” 80 Fed. Reg. at 21662 n.24. Rather, the “ADA contains a clear ‘safe harbor’ for wellness programs—the ‘voluntary’ provision of § 12112(d)(4)(B).” Id. To apply the insurance safe harbor provision to wellness programs, the footnote states, would “render the ‘voluntary’ provision superfluous.” Id.

The district court acknowledged the proposed regulation but disregarded it. To be sure, as the court noted, “a proposed regulation . . . is not entitled to deference.” Clay v. Johnson, 264 F.3d 744, 750 (7th Cir. 2001) (internal quotation marks and citation omitted). But in this case, the EEOC’s proposed regulation is at least entitled to consideration, as it embodies the EEOC’s long-standing view that the insurance safe harbor does not apply to the prohibition in § 12112(d)(4) on involuntary exams and inquiries that are part of an employee health program.

B. Even if the safe harbor provision applies to some wellness programs, it does not apply to Flambeau’s mandatory health risk assessments and biometric tests.

Even assuming, arguendo, that the insurance safe harbor provision could apply to some wellness programs, the district court erred in applying it here. The safe harbor provision is an affirmative defense, meaning that employers bear the burden of showing it applies. See Doe v. Mutual of Omaha Ins. Co., 179 F.3d 557, 562 (7th Cir. 1999) (calling § 12201(c) “a defense to liability”); Zamora-Quezada, 34 F. Supp. 2d at 442 (calling the safe harbor provision an “affirmative defense”). Contrary to the district court’s holding, Flambeau failed to establish on this record that the mandatory HRAs and biometric tests were used for “underwriting” or constituted a “term” of the health insurance plan, and the record shows that Flambeau impermissibly used the safe harbor provision as a subterfuge to avoid the prohibition in § 12112(d)(4)(A) on disability-related inquiries and medical exams. The district court’s contrary conclusion was based on an unjustifiably broad construction of the safe harbor provision that threatens to render obsolete the prohibition in § 12112(d)(4) against involuntary medical exams and inquiries that are part of an employee health program. This Court should therefore reverse summary judgment, even if the safe harbor could apply to some wellness programs.

1. Flambeau’s mandatory health risk assessments and biometric tests were not used for “underwriting risks, classifying risks, or administering such risks.”

As discussed above, although the ADA does not define “underwriting,” courts have recognized that “underwriting generally refers to the application of the various risk factors or risk classes to a particular individual or group for the purposes of determining whether to provide coverage.” Zamora-Quezada, 34 F. Supp. 2d at 443 (citing EEOC’s Guidance, 1993 WL 1497027, at *6 n.15). The safe harbor thus permits the use of “accepted principles of risk assessment” for the development and administration of benefit plans. Barnes, 22 F. Supp. 2d at 1020 (citing 29 C.F.R. pt. 1630, App. § 1630.16(f)). This Court has also suggested that the term “underwriting” in § 12201(c) refers to the use of “sound actuarial principles or claims experience.” Doe, 179 F.3d at 562 (observing that insurance company could not rely on the safe harbor provision at § 12201(c) to justify AIDS caps because it could not show the limits were “based on sound actuarial principles or claims experience”).

Stated more concretely, insurance underwriting is concerned with “group[ing] risks with similar characteristics when deciding whom to insure, what coverage to offer to potential insureds, and how much” to charge applicants. Kimberly A. Ackourey, Insuring Americans with Disabilities: How Far Can Congress Go to Protect Traditional Practices?, 40 Emory L.J. 1183, 1187 (1991) (emphasis added). An insurer typically “classifies the riskiness of applicants by determining the applicants’ average losses and by grouping them according to these assumptions concerning their riskiness.” Id. After averaging these losses, an insurer applies them to the applicants and then bases their premiums “on their expected loss class.” Id. In other words, insurers differentiate between “high-risk groups,” who pay more, and “low-risk groups, who should pay less.” Id. Significantly, “insurers generally apply the classification process to individual policies.” Id. (emphasis added). Risk classification has less to do with group health insurance policies because an insurer “usually promises to insure all persons in the group at a fixed price.” Id. Insurers nevertheless could—at least when the ADA was enacted, see infra at p.38, “distinguish among group members by restricting coverage of preexisting illness” or excluding certain procedures. Id.

Employer wellness programs—even those that include HRAs and biometric tests—are not aimed at underwriting. Rather, wellness programs are typically intended to improve employee health and thereby reduce overall costs. Thus, health risk assessments and biometric exams aimed at improving employee health as part of a wellness program would not typically fall under the safe harbor provision. In contrast, a health risk assessment asking disability-related questions for the purpose of acquiring or pricing group health insurance could fall under the safe harbor exception for insurance underwriting. See, e.g., Barnes, 22 F. Supp. 2d at 1020 (holding that disability-related inquiries fell under the safe harbor provision where the information was sought “for the purpose of underwriting, classifying, and administering risks in conjunction with defendant’s search for a new group health plan”); Bloch v. Rockwell Lime Co., N0. 07–478, 2007 WL 4287275, at*6 (E.D. Wis. Dec. 4, 2007) (holding that the safe harbor provision applied where the employer asked disability-related questions for the purpose of gathering bids from insurance companies for its group health insurance plan).

In contrast to the employers in Barnes or Bloch, Flambeau did not require its employees to answer disability-related inquiries for the purpose of pricing out group health insurance plans. Rather, Flambeau’s self-professed purpose for requiring the HRAs—and the biometric tests—was to improve employee health and thereby cut costs. That simply is not underwriting. The district court’s contrary conclusion was based on an overly-broad interpretation of “underwriting risks, classifying risks, or administering risks.” According to the district court, underwriting “refers simply to the process of developing an insurance plan.” A-12, R.138, p.12 (Order) (relying on Seff, 778 F. Supp. 2d at 1374). The court seemed to rule, in essence, that anything having to do with an insurance plan constitutes underwriting. As discussed above, this is plainly not so. The court’s broad interpretation of underwriting has no support in the plain language of the statute and is contrary to the legislative history and case law, which make clear that the safe harbor provision refers to underwriting based on risk classification, not to cost-savings plans. And the record in this case establishes that Flambeau’s mandatory HRAs and biometric tests were not used for underwriting under § 12201(c).

Flambeau has never asserted that it utilized the HRA and biometric testing data to determine which employees to insure or which medical conditions to cover. Cf. Zamora-Quezada, 34 F. Supp. 2d at 442-43 (“underwriting” concerns the use of risk factors to determine whether to provide coverage; “classifying risks” refers to the identification of risk factors and the grouping of factors with similar risks); Ackourey, 40 Emory L.J. at 1188 (stating that insurers use risk classification to determine who and what to cover and to set premium rates). To the contrary, Flambeau stated in its position statement that it “does not use the results of any testing or surveys to determine whether an employee is eligible for medical insurance.” R.17-25, p.2 (position statement). Further, Rieland testified that the HRA and biometric test results were not used to determine which benefits to offer. R.30-9, p.5 (Rieland Depo. 71). Rieland even acknowledged that Flambeau was prohibited from using an individual’s health factor to determine whether to insure him or her, to set premium rates, or to determine which benefits to provide. R.30-9, p.5 (Rieland Depo. 70-71).

Rieland also twice denied in his deposition that the HRA and biometric test data was used to determine premium rates. Cf. Ackourey, 40 Emory L.J. at 1188 (stating that insurers use risk classification to determine premium rates). When asked whether the data was used in “setting up what . . . premiums would be charged,” Rieland responded, unequivocally, “No.” R.30-9, p.5 (Rieland Depo. 71). At the next page of his deposition, he confirmed that the HRA and test data was not used to determine premiums. R.30-9, p.6 (Rieland Depo. 72). Rieland further conceded that the next year’s premium amounts were the “direct result” of the prior year’s benefits expenses, once more confirming that the HRA and biometric test data did not inform the premium rates. R.30-9, p.27 (Rieland Depo. 213).

Rieland also admitted that Trotter Wellness’s report on the wellness program results is devoid of any actuarial or risk analysis data to be used for underwriting or setting premiums. R.30-9, p.28 (Rieland Depo. 217-19). He further testified that he was unaware of any data showing any purported cost savings due to the mandatory HRAs and biometric testing and that Flambeau never commissioned any study to determine what savings were realized as a result of the HRAs and biometric tests. R.30-9, p.27 (Rieland Depo. 213-14). Nor has Flambeau asserted that it used the wellness data to place caps on claims for certain medical conditions. Cf. Doe, 179 F.3d at 562 (suggesting that AIDS cap would be permissible under § 12201(c) if it were “based on sound actuarial principles or claims experience”). Thus, the record shows that the HRAs and biometric testing were not used for underwriting. Rather, the purpose of the HRAs and biometric tests was to encourage healthier employee habits and thereby reduce health care costs. R.30-9, p.3 (Rieland Depo. 52); R.17-25, p.3 (position statement).

Flambeau’s claim that it needed the HRA and biometric testing data to underwrite, classify, or administer risks is further undermined by the fact that spouses and dependents were not required to complete an HRA or biometric test. Had Flambeau actually been relying on the data for risk classification to determine whether to cover certain individuals or conditions or to set premium rates (even if the company could lawfully do so), it would have needed the data for all of its plan participants. Yet nearly 25% of the 1,132 plan participants in 2012 were spouses (and an unknown number were dependents), meaning that Flambeau lacked data for a sizable portion of its plan participants. R.30-3, p.5, slides 1 & 3 (9/19/13 plan presentation showing that in benefit year 2012, 854 plan participants were employees and 278 were spouses).

Flambeau’s assertion that the wellness program data was used for underwriting is likewise undermined by the position statement the company submitted to the EEOC. In its position statement, Flambeau asserted that the “sole purpose” of the testing was to provide “an employee benefit to the employee[s] to learn about their own medical conditions” to allow them to be “fully informed for their own healthcare needs.” R.17-25, p.2. Healthier employees, Flambeau said, “are more productive and consume less health care services.” Id. at 3. Additionally, Flambeau added, the HRAs and biometric tests benefited employees by providing them “the opportunity to lead healthier, happier lives.” Id. Nowhere in the position statement did Flambeau suggest that it adopted the mandatory HRAs and biometric tests for the purpose of underwriting. Thus, Flambeau’s assertion that the mandatory HRA and biometric testing was used for underwriting is clearly an afterthought that the company adopted for the purpose of defending against the EEOC’s enforcement action.

In an attempt to bolster its claim that the HRAs and biometric tests results were used for underwriting, Flambeau relied below on the affidavit of Sara Hames, the Vice President of Hays, which provided healthcare consulting to Flambeau. But Hames’ affidavit fails to make Flambeau’s point. Hames candidly conceded that the “central goal” of a wellness program with biometric screenings and HRAs is to improve employee health and thereby reduce costs; Hames did not say the “central goal” of a wellness program with biometric tests and HRAs was insurance underwriting. R.13, ¶ 4.

It is true, as the district court stated, that Hames stated that the wellness results prompted Hays to recommend that Flambeau purchase “stop-loss” insurance and that Hays incorporated that cost into their recommendations for 2012 premium rates. Id.¶ 8; A-12, R.138, p.12 (Order). But this recommendation does not show the HRA and biometric test data was used for underwriting the health insurance plan. Also, Hays’ recommendation to purchase stop-loss insurance was based on the HRA and test data from the “fall of 2010,” which was when the wellness program was not required. R.13, ¶¶ 7-8 (Hames Aff.) Because the recommendation to purchase stop-loss insurance was not even based on data collected from the mandatory HRAs and biometric tests (as they became mandatory only beginning in the fall of 2011 for the 2012 benefit year), the court erred in relying on the stop-loss insurance recommendation to conclude that Flambeau’s mandatory wellness plan was used for underwriting.

In any event, Flambeau is over-reaching by claiming that the recommendation to purchase stop-loss insurance was based on the results of the HRAs and biometric tests, as “[c]ompanies that self-insure generally buy . . . a stop-loss insurance policy to protect the employer’s assets against losses above a certain threshold.” Consumer Guide to Group Health Insurance, National Association of Health Underwriters, http://www.nahu.org/consumer/GroupInsurance.cfm (last visited April 15, 2016).

Further, Hames stated that she and her colleagues used the data and statistics from the HRAs and biometric tests “along with other information” as the basis of their recommendation that Flambeau purchase stop-loss insurance, obscuring the actual role the HRA and biometric test data played in her recommendation. R.13, ¶ 8 (Hames Aff.) (emphasis added). Moreover, Flambeau already knew of the past medical conditions and claims of its employees and dependents, which would presumably be the best indicator of the need, if any, for stop-loss insurance to protect Flambeau’s self-funded health plan against catastrophic claims. See R.30-9, p.27 (Rieland Depo. 213) (agreeing that premium rates were “direct result” of benefits paid in the prior year and agreeing that “what you’re liable to incur in the future” is determined by “what you incurred in the past”).

Flambeau’s claim that it utilized the HRA and biometric test data for underwriting in connection with its purchase of stop-loss insurance is also undermined by the fact the company did not require spouses and dependents to complete the HRAs or biometric tests; had the company actually been using the data to estimate the need for, and amount of, stop-loss insurance to purchase, this information would have been crucial. Cf. Edstrom Indus., Inc. v. Companion Life Ins., 516 F.3d 546, 548 (7th Cir. 2008) (discussing litigation in which self-insured employer failed to inform insurance company selling stop-loss insurance that a plan participant recently gave birth to a child with a grave, and costly, medical condition), abrogated on other grounds, Hall St. Assoc., LLC v. Mattel, Inc., 552 U.S. 576 (2008). As for distributing the cost of stop-loss coverage among plan participants to make premium recommendations for the 2012 benefit year, R.13, ¶ 8, that is a matter of simple division, not underwriting.

Hames also stated that the results of the 2012 wellness plan prompted Hays’ recommendation that Flambeau cover 100% of preventive exams and lower the co-pay cost of maintenance drugs, as the district court noted. R.13, ¶ 9 (Hames Aff.); A-12, R.138, p.12 (Order). But recommendations to pay for exams or lower drug co-pays is not “underwriting risks, classifying risks, or administering such risks.” Rather than use risk classification to determine insurability or set rates, these recommendations merely provided general advice as to how Flambeau might be able to reduce its health care costs by improving overall health. Hames’ testimony actually underscores that the wellness data was not used to classify and assess risk; she testified that in 2013 she and her colleagues “went through [their] standard process of classifying and assessing risks and projecting plan costs” for Flambeau’s plan and “then [they] built in cost/savings of plan design changes” from the past year, “some of which were influenced by the aggregate results” of the wellness program, to estimate costs for the 2014 plan year and recommend premium rates. R.13, ¶ 11. Rather than establish that the wellness data was used to classify and assess risks, then, Hames’ statement shows that this was done before Hays added in (unidentified) cost/savings changes from the previous year, only “some of which” were influenced by the wellness data.

The court also reasoned that the wellness plan requirement was used for underwriting, classifying, or administering risk because, the court said, Flambeau’s healthcare consultants relied on the data to recommend that Flambeau “charge cigarette smokers higher premiums.” A-12, R.38, p.12 (Order); see R.12, ¶ 21 (Rieland Aff.) (stating that Flambeau implemented incentives for non-smokers and those in smoking cessation programs). But the record shows that Flambeau adopted non-smoking incentives in 2011 and was charging smokers a higher premium in 2011—before the HRAs and biometric tests became mandatory for the 2012 benefit year. See R.11-1 (Rieland Depo. 34) (Flambeau adopted smoking incentives in 2011); R.12-7, p.1 (reflecting that the 2011 premium rate for a smoker was $50/month more than for a non-smoker).

The district court’s only legal authority for its determination that Flambeau’s mandatory HRAs and biometric tests constituted “underwriting” is Seff v. Broward County, 778 F. Supp. 2d 1370 (S.D. Fla. 2011), aff’d, 691 F.3d 1221 (11th Cir. 2012). A-6-7, R.138, pp.6-7 (Order). In Seff the employer imposed a $20/week surcharge on employees who participated in the company’s health plan but refused to complete an HRA and biometric test. Significantly, the Eleventh Circuit never reached the issue of whether the employer used the health risk assessment and biometric testing data for underwriting. Seff, 691 F.3d at 1223. Rather, the sole issue on appeal was whether the wellness program was a “term” of the employer’s health plan. Id.

As for the district court’s holding in Seff that the HRA and biometric test data was used for underwriting, classifying, or administering risks, the Seff court erred for the reasons discussed above concerning the proper interpretation and application of the safe harbor provision. In short, the HRA and biometric testing data in Seff was not concerned with using risk classification to determine coverage or premiums. Rather, as in this case, the avowed and actual purpose of the wellness program in Seff was cost reduction. See Seff, 778 F. Supp. 2d at 1371. While the Seff district court concluded that because the wellness plan was meant to lower premiums it was therefore used for “underwriting,” the court cited no provision of the statute to support this conclusion. See id. at 1374.

The only case the Seff court relied upon for authority was Barnes, 22 F. Supp. 2d 1013. But Barnes was not a wellness case. Rather, Barnes was an actual underwriting case, as the disability-related inquiries in that case were made for the purpose of pre-enrollment underwriting when the employer sought out a new health insurance plan. See id. at 1017. In contrast to Barnes, nothing in Seff suggests that the employer’s insurer had requested the HRA and biometric screening for the purpose of pre-enrollment underwriting; the employer’s insurance company was already in place, and the plaintiff was already insured. See Seff, 778 F. Supp. 2d at 1372. Moreover, the employer’s claim in Seff that the wellness plan was about underwriting risks was undercut—as it is here—by the fact that not all the plan participants were required to submit to the HRA and biometric testing (employees incurred only a $20 biweekly surcharge for refusing).

The district court’s expansive view of “underwriting” should also be rejected because it undermines the remedial purpose of the ADA. See generally Steger v. Franco, Inc., 228 F.3d 889, 894 (8th Cir. 2000) (recognizing that “the ADA is a remedial statute” that “should be broadly construed to effectuate its purpose”). In enacting the ADA, Congress found that “discrimination against individuals with disabilities continue[s] to be a serious and pervasive social problem.” 42 U.S.C. § 12101(a)(2). Congress’ avowed purpose in enacting the ADA was “to provide a clear and comprehensive national mandate for the elimination of discrimination against individuals with disabilities,” and Congress sought to achieve that goal by providing “strong . . . standards addressing discrimination against individuals.” 42 U.S.C. § 12101(b)(1)-(2). The district court’s interpretation of “underwriting” as including Flambeau’s use of mandatory HRAs and biometric tests to encourage healthier habits in its employees—which allows nearly any wellness program to fall under the safe harbor exception—undermines the strong protections Congress included in the ADA to guard against disability-based discrimination.

The district court’s excessively broad interpretation of “underwriting” also runs afoul of traditional canons of statutory construction. Assuming the safe harbor provision applies to wellness programs, it is an exception to the general prohibition of § 12112(d)(4)(A) against employee medical exams and disability-related inquiries. As such, it should be narrowly construed. See 1A Sutherland Statutory Construction § 20:22 (7^th ed.) (explaining that a “proviso” typically limits the operation of a general rule and stating that if a proviso’s “restrictive scope . . . is in doubt, the proviso is strictly construed”); see also EEOC v. Kamehameha Sch./Bishop Estate, 990 F.2d 458, 460 (9th Cir. 1993) (in Title VII religious discrimination case, stating that “[w]e construe the statutory exemptions [to discrimination] narrowly”). As with the district court in Seff, the district court in this case failed to heed this fundamental canon of statutory construction and instead applied an unjustifiably broad construction of “underwriting.”

Finally, we note that the district court failed to grapple with the narrowed scope of the safe harbor provision. Ironically, the district court’s expansive interpretation of “underwriting” as essentially encompassing anything having to do with the administration of an insurance plan comes in the wake of recent changes in the law that have severely curtailed the kind of underwriting in which insurers may engage. For instance, at the time of the ADA’s enactment, insurers could lawfully exclude individuals with pre-existing health conditions or limit their benefits. See, e.g., Ackourey, 40 Emory L.J. at 1186-88 (noting in 1991 article discussing the ADA that individuals with pre-existing conditions may be denied insurance, charged higher premiums, or have their benefits limited); 29 C.F.R. pt. 1630, App. § 1630.16(f) (stating in interpretive guidance from 1991 that the safe harbor provision permits employers to deny coverage to an individual with a disability, or to offer different terms and conditions of insurance, if the disability poses “increased risks”). Likewise, the EEOC’s Interim Guidance from 1993 lists “medical history” as a risk factor that could be used for underwriting. Interim Guidance, 1993 WL 1497027 at *6 n.15 (emphasis added).

The Patient Protection and Affordable Care Act (ACA), however, prohibits health insurers from excluding individuals with pre-existing conditions or limiting their benefits. 42 U.S.C. § 300gg-3(a) (“A group health plan and a health insurance issuer offering group or individual health insurance coverage may not impose any preexisting condition exclusion with respect to such plan or coverage.”). Further, the Health Insurance Portability and Accountability Act of 1996 (HIPAA), as amended by the ACA, generally prohibits insurers from discriminating against individuals “in premiums, benefits, or eligibility based on a health factor,” which includes “health status, medical condition . . . , claims experience, receipt of health care, medical history, genetic information, evidence of insurability (including conditions arising out of acts of domestic violence), and disability.” 80 Fed. Reg. at 21660 & n.14. It also permits health insurers to vary premiums for coverage offered in the individual or small group market based only on age, tobacco use, geographic area, and family size. 42 U.S.C. §§ 300gg. Thus, while there is still space under the safe harbor provision for legitimate underwriting based on risk classification, Congress narrowed that space considerably.

2. The health risk assessments and biometric testing were not

“terms” of the insurance plan.

Even if the HRA and biometric testing were used for underwriting, the safe harbor provision does not apply because they were not “terms” of Flambeau’s insurance plan. The word “term” is not defined in the ADA. Therefore, it “carries its ordinary meaning.” Crawford v. Metro. Gov. of Nashville & Davidson Cty., Tenn., 555 U.S. 271, 276 (2009). The Merriam-Webster dictionary defines “term” as a “provision[] that determine[s] the nature and scope of an agreement.” Merriam-Webster Dictionary Online, available at http://www.merriam-webster.com/dictionary/term (last visited April 22, 2016). As the district court acknowledged, ERISA requires a summary plan description to identify a “plan’s requirements respecting eligibility for participation and benefits.” 29 U.S.C. § 1022(b). But the summary plan description makes no mention of mandatory HRAs and biometric testing. See Moore v. Metro. Life Ins. Co., 856 F.2d 488, 492 (2d Cir. 1988) (“Congress intended that plan documents and the [summary plan description] exclusively govern an employer’s obligations under ERISA plans.”). It is also undisputed that neither the underlying insurance plan nor the collective bargaining agreement informed employees that the HRA and biometric testing were prerequisites to enrollment in the health insurance plan (or to enrollment at 25% of the COBRA rate). Thus, the HRA and biometric testing were not provisions—and therefore were not “terms”—of the health insurance plan.

The district court acknowledged that the HRA and biometric testing requirements were omitted from the summary description and the collective bargaining agreement. A-10-11, R.138, pp.10-11 (Order). But the court nevertheless concluded that because Flambeau required employees to complete the HRA and biometric testing to enroll in the plan, they were, ipso facto, terms of the plan. Id. The court reasoned that the summary plan description warned participants that they would be required to enroll “‘in the manner and form prescribed by [defendant]’” and that Flambeau distributed handouts stating that the HRA and biometric testing were mandatory. Id. at 10. Citing Mers v. Marriott International Group Accidental Death & Dismemberment Plan, 144 F.3d 1014, 1023 (7th Cir. 1998), the court also added that “it is well-recognized that a summary plan description does not establish the terms of an employee benefit plan.” Id. at 11. Further, the court said, “a benefit plan may be established without any written document whatsoever.” Id. The court’s reasoning was flawed, and its reliance on Mers was misplaced.

ERISA says nothing about permitting handouts to substitute for the requirement that the summary plan description set out the terms of eligibility. Also, the statement in the summary plan description that employees would be required to enroll “in the manner and form prescribed” clearly refers to the mechanics of enrollment, not to the terms of eligibility. As for Mers, this case actually supports the EEOC’s argument that the HRA and biometric testing were not plan “terms.” In Mers, this Court reiterated that ERISA requires summary plan descriptions to “‘reasonably apprise . . . participants and beneficiaries of their rights and obligations under the plan.’ 29 U.S.C. § 1022(a)(1). In particular, a [summary plan description] must list the circumstances in which . . . ineligibility may occur.” Mers, 144 F.3d at 1021 (citing 29 U.S.C. § 1022(b); 29 C.F.R. § 2520.102-3). This Court went on to hold in Mers that where a direct conflict exists between a summary plan description and the underlying policy, then the summary plan description trumps. Id. at 1023. If, however, the summary plan description is silent on an issue and the underlying plan “clarifies and does not contradict the terms of the [summary plan description],” then the terms of the underlying plan control. Id. at 1024. Thus, the district court was incorrect in asserting that Mers held that a summary plan description does not establish the terms of an employee benefit plan; it does, if the summary conflicts with the plan.

But in any event, Mers’ holding on this point is inapposite, as there is no argument here that the summary plan description was silent on an issue that the underlying policy addressed. Rather, it is undisputed that neither the summary plan description nor the underlying policy stated that an HRA and biometric test were mandatory prerequisites to enrollment. Given Mers’ reiteration of ERISA’s requirement that the summary plan description inform participants of the circumstances of ineligibility, Mers compels the conclusion that the wellness program was not a “term” of the plan.

The district court’s observation that “a benefit plan may be established without any written document whatsoever,” A-11, R.138, p.11, also fails to support the court’s conclusion that the wellness program was a “term” of the plan. This is not a case involving an unwritten benefit plan; it is a case involving a written benefit plan with a 100+ page summary plan description. The question therefore is not whether a benefit plan may be established in the absence of written documents but whether an employer can change the terms of a written plan by distributing a handout to plan participants. See generally Sandstrom v. Cultor Food Science, Inc., 214 F.3d 795, 797 (7th Cir. 2000) (where there was a written plan, holding that the court need not decide whether ERISA permits unwritten plans). As stated, ERISA’s requirement that the summary plan description set forth the circumstances of ineligibility undercuts the district court’s conclusion that Flambeau’s handouts could add a term to the plan. Moreover, precedent from this Court and others suggests that a handout is insufficient to alter the terms of a written benefit plan. See Sandstrom, 214 F.3d at 797 (“Statements by plan administrators, side agreement and understandings, or even special offers made to many of a firm’s employees, do not change the contents of the plan applicable to other employees.”); Moore, 856 F.2d at 492-93 (stating that “an ERISA welfare plan is not subject to amendment as a result of informal communications between an employer and plan beneficiaries” and holding that booklets, filmstrips, and other presentations explaining the insurance plan did not “override plan documents and [summary plan descriptions] created pursuant to ERISA”).

Although the district court did not explicitly rely upon it, we note that the Eleventh Circuit’s decision in Seff also fails to support the conclusion that the mandatory HRAs and biometric tests were “terms” of Flambeau’s health insurance plan. In Seff, the Eleventh Circuit held that the wellness program there—which also involved an HRA and biometric screening—was a “term” of the employer’s health insurance plan. 691 F.3d at 1223. The Eleventh Circuit reached this conclusion despite acknowledging that the acting benefits manager had testified that the wellness program was not a term of the health plan, and despite acknowledging that employees could enroll in the health insurance plan without completing the HRA and biometric testing (for a $2o/biweekly charge). Id. at 1222-23. The Eleventh Circuit concluded summarily that the manager’s testimony was irrelevant because the plaintiff “presents no substantive argument that the issue of whether the employee wellness program was a written term contained within the physical plan documents for Broward’s group health plan is material to the determination of the safe harbor provision’s applicability.” Id. at 1224. This reasoning does not withstand scrutiny. As discussed above, this Court has recognized that ERISA requires a summary plan description to set out the circumstances of eligibility, which the plan in Seff—as well as in this case—clearly failed to do. Additionally, to the extent Flambeau’s argument is that the HRA and biometric testing were “voluntary” because employees could obtain insurance at 100% of the COBRA rate, this shows that the HRA and testing were not terms of the plan.

3. Flambeau used the safe harbor provision as a “subterfuge.”

Even if the mandatory HRA and biometric testing were terms of Flambeau’s insurance plan and were used for underwriting, summary judgment was still inappropriate because Flambeau used § 12201(c)(2) “as a subterfuge to evade the purposes” of the ADA.

As with “underwriting” and “term,” the ADA does not define “subterfuge.” In construing the term under the ADEA, the Supreme Court held in Public Employees Retirement System of Ohio v. Betts, 492 U.S. 158 (1989), that “subterfuge” should be accorded its ordinary meaning, i.e., “a scheme, plan, stratagem or artifice of evasion” and suggested it means an intent to evade the statute. Id. at 171 (internal quotation marks and citation omitted). Relying on that definition, this Court has said that “subterfuge” occurs under 29 U.S.C. § 623(j) of the ADEA when the employer tries to use the exemption “as a way to evade another substantive provision of the act.” Minch v. City of Chicago, 363 F.3d 615, 629 (7th Cir. 2004) (citing Betts, 492 U.S. at 181). As this Court noted in Minch, other circuits have also “looke[d] to Betts as the relevant precedent on subterfuge” under the ADA’s safe harbor provision. Id. at 627; see, e.g., Ford v. Schering-Plough Corp., 145 F.3d 601, 611 (3d Cir. 1998) (applying Betts’ definition of subterfuge to the ADA’s safe harbor provision). The burden of establishing subterfuge rests on the plaintiff. See Betts, 492 U.S. at 181 (employees bear the burden on subterfuge); Minch, 363 F.3d at 623 (discussing how “[a] plaintiff can establish subterfuge”).

Here, the EEOC showed that Flambeau “took advantage of the [safe harbor] exemption” “in order to evade a different substantive provision of the statute,” i.e., the prohibition in § 12112(d)(4) on involuntary exams and disability-related inquiries. Minch, 363 F.3d at 623. Flambeau’s intent to use the safe harbor provision to evade the prohibition on medical exams and disability-related inquiries except as part of a voluntary employee health program is evidenced by the omission from Flambeau’s position statement of any reference to underwriting or classifying risks, or to the safe harbor provision. R.17-25 (position statement). To the contrary, Flambeau asserted in the position statement that the “sole purpose” of the testing was to enable employees to learn about their own medical needs. R.17-25, p.2. Flambeau also denied unequivocally that it used the results of the wellness program to determine employee eligibility for insurance. R.17-25, p.3. Not until litigation began did Flambeau assert that it used the mandatory HRA and biometric testing data for underwriting. Flambeau’s belated invocation of the safe harbor provision to navigate around the statute’s prohibition on involuntary exams and inquiries therefore shows that Flambeau sought to use the provision as a subterfuge.

Rieland’s deposition testimony and Hames’ affidavit further support the conclusion that Flambeau invoked the safe harbor provision to evade § 12112(d)(4). As discussed, supra, their statements make clear that the goal of the mandatory HRAs and biometric testing was to make employees healthier, not to assess risk to determine insurability or coverage. The notion of making the HRAs and biometric tests mandatory actually originated in the February 15, 2011, email from Bill Siehr. R.12-6. In that email, Siehr suggested that Flambeau combat the “whack[ing]” of Flambeau’s healthcare budget by “[e]liminating or improv[ing] the underlying health conditions” that “drive the large claims.” Thus, this email reveals that Flambeau did not require the HRAs and biometric tests for the purpose of underwriting risk, classifying risk, or administering such risk; rather, Flambeau did so in a (failed) effort to make its employees healthier.

Likewise, the fact Flambeau required only employees—not spouses or dependents—to complete the HRA and biometric testing shows that Flambeau never intended to use the data for underwriting, as the company lacked data from a sizable number of plan participants. This provides additional evidence that Flambeau’s intent in invoking the safe harbor provision was to avoid the prohibition of § 12112(d)(4) on medical exams and inquiries that are neither job-related and consistent with business necessity nor part of a voluntary employee health program.

The district court thus erred in concluding that Flambeau had not used the safe harbor as a subterfuge. The court reasoned that no subterfuge occurred because “the purpose of the ADA is not to prohibit employers from asking for medical and disability-related information” but to eliminate disability-based discrimination. A-14, R.138, p.14. This is incorrect. The plain text of the ADA, § 12112(d)(4), states that the prohibition against “discrimination” “shall include disability-related inquiries and medical exams,” and § 12112(d)(4)(A) explicitly prohibits employee exams and inquiries unless they are job-related and consistent with business necessity or part of a voluntary health program. Contrary to the district court’s conclusion, then, one of the clear purposes of the ADA was to prohibit involuntary medical exams and disability-related inquiries of employees, unless they fall under two narrow exceptions.

The legislative history also demonstrates that Congress was concerned about involuntary medical exams and disability-related inquiries. The Senate Committee on Labor and Human Resources Report explains that such exams and inquiries “serve[] no legitimate employer purpose, but simply serve to stigmatize the person with a disability.” S. Rep. No. 101—116 (1989), reprinted in H.R. Comm. on Education and Labor, 101^st Cong., 2d Sess., Legislative History of Public Law 101-336 at p.137 (Comm. Print 1990). For example, the Senate Report stated, an employer should not be able to require an employee with hair loss to be tested for cancer, unless the testing is job-related. Id. at p.137. The Senate Report recounted that testimony before the Committee indicated that “there still exists widespread irrational prejudice against persons with cancer,” and the Report notes that an individual with cancer “may object merely to being identified, independent of the consequences.” Id. at p.138. Thus, contrary to the district court’s conclusion, one of the purposes of the ADA was to prohibit unlawful medical exams and disability-related inquiries; those results can be used by employers to discriminate based on disability, which the ADA prohibits, but the inquiries and exams can be stigmatizing in their own right. Flambeau’s attempt to invoke the safe harbor provision to evade the prohibition on involuntary exams and disability-related inquiries therefore constitutes a subterfuge.

CONCLUSION

For the foregoing reasons, the judgment of the district court should be reversed and the case remanded for further proceedings.

Respectfully submitted,

P. DAVID LOPEZ

General Counsel

JENNIFER S. GOLDSTEIN

Associate General Counsel

LORRAINE C. DAVIS

Assistant General Counsel

s/Anne Noel Occhialino

ANNE NOEL OCCHIALINO

Attorney

Equal Employment

Opportunity Commission

Office of General Counsel

131 M St. N.E., 5th Fl.

Washington, D.C. 20507

(202) 663-4724 (phone)

(202) 663-7090 (fax)

Annenoel.Occhialino@eeoc.gov

[1] “R.#” refers to the district court docket entry.Where a cited document is included in the Appendix attached to this brief, “A-#” refers to its location in the Appendix.

[2] The HRA for the 2013 benefit year also asked questions regarding family medical history. R.35, ¶ 13 (reply to SOF).

[3] Although the safe harbor is contained within Title V of the Act, it references Title I, under which the EEOC has substantive rulemaking authority. 42 U.S.C. § 12116.

[4] The proposed regulation is expected to become final by May 2016.

STATEMENT CONCERNING APPENDIX

APPENDIX