A Modest Proposal? Regulating Customer Discrimination Through the Firm

Antidiscrimination law appears to be of two minds. The law aggressively polices discrimination in some realms, while leaving other realms to be governed by the preferences of their participants.1 In their essay, Professors Katharine Bartlett and Mitu Gulati suggest recalibrating the boundaries of this area of law so that it covers discrimination in an area that the law now leaves largely up to the realm of personal preference: discrimination by customers.2 

Even if there is agreement that the law should regulate customer discrimination, how should it do so? While Bartlett and Gulati consider a direct ban on discrimination by customers, they reject this approach in favor of regulating firms for two primary reasons: (1) the increased efficacy of firm liability and (2) concern with infringing on customer privacy and autonomy.3  Bartlett and Gulati propose that “entities that already have a legal obligation not to discriminate . . . also should have an explicit obligation to curtail and not to facilitate discrimination by their customers, and to refrain from discrimination when they, themselves, act in the role of a customer.”4 They would allow firms a BFOQ-like exception for “the most compelling instances implicating customer privacy.”5  They describe this proposal as “modest.”6 

This Response considers the claim that such a proposal amounts to a modest change in the law, and then addresses the consequences of regulating discrimination by customers through firms. As a descriptive matter, this Response suggests that Bartlett and Gulati’s negligence-like proposal represents a fairly substantial break with current antidiscrimination laws regulating firms. This fact does not necessarily weigh against their proposal. A substantial break in the law may be required to deal with a sufficiently troubling problem.  In fact, Bartlett and Gulati’s work might lead us to consider anew whether antidiscrimination law should hold firms to a negligence standard in a broader range of circumstances.7 As a normative matter, this Response highlights the downsides of regulating customer discrimination through firms, including some of the same concerns—such as efficacy—that motivate Bartlett and Gulati to reject customer liability. These downsides do not require a rejection of firm liability in favor of either customer liability or no liability.  However, they must be accounted for in considering the optimal approach. Finally, this Response argues that allowing firms too much leeway to take account of discriminatory customer preferences under a BFOQ-type exception fails to hold firms accountable for their role in cultivating these preferences.

I.     A Modest Proposal?

Professors Bartlett and Gulati describe their proposal to place liability on firms to prevent customer discrimination as “modest.”8 In their view, the proposal only “enlarge[s] slightly the obligations that firms already have.”9 This conclusion comes after having rejected a bolder proposal to impose a direct ban on discrimination by customers. In this relative regard, their proposal is modest. But how large of a shift is it away from existing antidiscrimination law? 

In my reading, Bartlett and Gulati’s proposal to hold firms liable when they “facilitate discrimination by their customers” and when they fail “to curtail” discrimination by their customers amounts to negligence liability.10 Bartlett and Gulati suggest that liability could arise not only from affirmative acts by the firm, but also from failures to take due care to prevent harm, a hallmark of negligence liability.11 Bartlett and Gulati also suggest that the touchstone for liability would be the reasonableness of the firm’s actions or failures to act, another hallmark of negligence liability.12   

The greatest impact of this proposal would appear to be as an addendum to employment discrimination law, as workers likely feel most of the harm from discrimination by customers. To be sure, Title VII already imposes some liability on firms for discrimination by customers.13  Nonetheless, imposing liability on a firm for failing to take due care with regard to facilitating or curtailing discrimination seems quite far from the heartland of antidiscrimination law, which generally addresses intentional discrimination.14 Although scholars have acknowledged that aspects of Title VII allow for liability on the basis of employer negligence, these are exceptions rather than the rule.15 

Perhaps most relevant to Bartlett and Gulati’s proposal is work by Professor Noah Zatz recognizing that employers have an affirmative obligation to prevent and redress sexual and racial harassment of their employees by customers.16 Note that this is not a doctrine unique so much to discrimination by customers as it is a doctrine unique to the law of harassment under Title VII, which permits liability for employers in circumstances that look like negligence.17 Liability for third-party harassment fits within Title VII’s intentional discrimination rubric because in these cases some party had discriminatory intent, even if that party wasn’t the employer itself.18 Bartlett and Gulati’s proposal is broader, not requiring any showing of discriminatory intent by anyone, as a firm surely could “facilitate” customer discrimination before any customer ever acts on it. This type of negligence-like liability could, at least in theory, greatly expand the realm of liability for employment discrimination.

Bartlett and Gulati don’t discuss how their proposed liability would be enforced. They suggest that liability would be imposed “when customer bias causes harm.”19 Because the persons most often harmed by customer discrimination are likely to be workers, limits on standing would probably play some role in restricting the expanded scope of liability. 

Aside from expanding the scope of liability, a shift to negligence liability for discrimination by customers would mean that employers are held to a higher standard when addressing discrimination by customers than they are for other forms of discrimination, including discrimination that originates within the firm itself. Consider, for example, unconscious coworker biases, which can lead some coworkers to not provide support to their minority and female colleagues in ways that make it harder for them to advance, or can infect their evaluations of their minority and female colleagues. Under current law, so long as the employee cannot show that discriminatory intent caused an adverse employment action, which they would be hard-pressed to do in the circumstances described, such claims are not actionable.20 Under Bartlett and Gulati’s proposal, an employer who failed to take reasonable steps to address such discrimination by customers could be held liable.   

Of course, Bartlett and Gulati are only tackling the problem of discrimination by customers. It is fair that their proposal only addresses this problem.  But why should workers who interact with customers be protected under the far more expansive standard of employer negligence? Perhaps these workers have made themselves vulnerable to additional sources of discrimination, and thus they need additional protection.21  Nonetheless, negligence liability seems like more than a modest expansion given the scope of employers’ obligations under current antidiscrimination law.22  Recognizing this does not necessarily weigh against the proposal.  Rather, it might prompt us to consider whether the problem that Bartlett and Gulati describe—the employer’s role in cultivating and reinforcing discrimination against workers by those with whom workers interact—is a problem broader than discrimination by customers and includes discrimination by coworkers, who, like customers, are also essential to workers’ employment outcomes.23  If employers engage in similar conduct vis-à-vis coworkers and current law fails to address it, perhaps Bartlett and Gulati’s proposal should be extended to cover discrimination by coworkers as well.

II.     The Downsides of Regulating Customers through the Firm

Bartlett and Gulati effectively highlight the benefits of firm rather than customer liability, but they do less to highlight the costs.24 This Part takes on that task, not to argue for a rejection of firm liability in favor of customer liability or no liability, but to present a fuller picture of the tradeoffs associated with assigning responsibility to firms rather than customers, and to recognize challenges to be addressed if Bartlett and Gulati’s proposal is adopted.  I write from a sympathetic perspective, as I have previously written about employers’ role in cultivating discriminatory preferences.25The following consequences are ones that might be concerning even for those who agree with Bartlett and Gulati that the law should do more to regulate discrimination by customers, and even for those who agree that firms should bear responsibility.  

Bartlett and Gulati reject customer liability due in part to concerns of efficacy. But firm liability raises its own concerns of efficacy. There is a real question as to how such claims—even if recognized in theory—will fare. Scholars have long lamented the dismal success rates of employment discrimination plaintiffs.26 One explanation for why discrimination claims are so often unsuccessful is that the general conception of what constitutes wrongful discrimination is quite limited: “[M]ost people do not ‘see’ discrimination, except where there is effectively no plausible alternative.”27 Thus, expanding the scope of liability for discrimination far outside of what most Americans consider wrongful might not translate to much success in courts, and may render these claims even less successful than typical discrimination claims. In this regard, direct customer liability might be more effective than firm liability because judges and juries are probably more inclined to fault the individual customer harboring discriminatory bias than the firm that failed to curtail the bias.28

Difficult line-drawing is another reason that Bartlett and Gulati cite for rejecting customer liability.29 But shifting from individual to firm liability does not avoid this problem. Bartlett and Gulati rightly highlight the relationship between customer discrimination and firm discrimination, explaining how customer biases lead to discrimination by the firm, and how firm biases lead to discrimination by customers.30  But because the interaction between customer discrimination and firm behavior is so deep, so many things that an employer does can either “curtail” or “facilitate discrimination by their customers.”31 To give just a few examples from Bartlett and Gulati, a firm could be liable under their proposal for failing to limit customers’ access to information about workers that could allow them to act on discriminatory preferences;32 for failing to change their branding and advertising strategies that promote discriminatory customer preferences;33 or for failing to modify their reliance on evaluation mechanisms through which biased customers affect workers’ employment prospects.34 By contrast, in employment discrimination cases, employer liability is typically limited to circumstances where an employee suffers an adverse employment action due to discrimination—termination, demotion, failure to hire or promote, or the like.35 Questions of employer liability outside of these circumstances have raised difficult questions.36 Deciding whether and when the employer should properly be liable in the types of circumstances catalogued above would be similarly challenging. 

A final concern about relying on firm liability is the way that firms might react to such liability. Bartlett and Gulati note that some firms, especially those that connect workers and customers through online platforms, may respond to liability by limiting customers’ access to information about workers that would allow customers to discriminate (e.g., not making workers’ race or sex known to customers).37 In a manuscript currently under development, I describe this type of action by firms, and increasingly by law, as a move towards “ignorance as equality.”38 While the law has long relied on certain forms of ignorance as equality,39 there has been a turn by both law and private actors to rely on ignorance to prevent discrimination. The type of response that Bartlett and Gulati imagine is enabled by technology: As customers increasingly come to begin their interaction with firms and workers through “apps” and other online mechanisms, technology can be used to ban access to information that would allow customers to discriminate on the basis of protected traits. 

While ignorance may provide a short-term prophylactic against customer discrimination, and is a smart move by firms striving to avoid liability for enabling discrimination by customers, it falls short of the traditional goal of antidiscrimination law. Bartlett and Gulati assert that the “[a]im of anti-discrimination law is to change how individuals act, not what they believe.”40 Others would disagree, and would view the role of law in this area as to “change[] hearts and minds.”41 Ignorance as equality fails to achieve this end. It is cynical about the ability to change discriminatory attitudes, and operates by disabling such attitudes rather than challenging them. Firm liability here may lead to efficient mechanisms to reduce discrimination, but it might not achieve the ultimate ends of a greater transformation in societal attitudes. Moreover, firms denying access to information about workers constructs an impoverished view of the worker, which restricts the intimacy that occurs within so many customer–worker interactions and is part of what makes such interactions so valuable.42

III.     Holding Firms Accountable for Privacy-Based Biases

Bartlett and Gulati suggest importing a BFOQ-like exception into their proposal for firm liability.43 Their effort to account for the interests of privacy and autonomy offered through this approach is commendable. They helpfully attempt to unpack the interest in privacy by recognizing that this seemingly singular interest may be motivated by a multiplicity of interests that might each be entitled to a different level of deference. For instance, they note that “[a] massage therapist’s service, arguably, cannot be effectively performed where the client does not feel comfortable, relaxed, and safe from sexual threat. The same reasoning does not apply to preferences that result from stereotypes linking sex to competence, such as hospital patients’ preferences for female nurses or male doctors.”44 

I would urge slicing these interests even more finely to ensure that the law both combats sex stereotypes and polices employers’ role in cultivating discriminatory customer preferences.45 Bartlett and Gulati state that “the goal should be to respect customer choices in especially personal settings while putting continued pressure on the stereotypes that influence those choices in a biased way.”46 I would argue that the goal instead should be to respect customer needs; otherwise we won’t be putting sufficient pressure on stereotypes that influence biased choices. For instance, in the example above, I would distinguish between general comfort-based preferences and specific safety concerns experienced by someone who has been a victim of sexual trauma. As for the latter, these are the types of concerns that can be documented, and can be limited to a certain subset of persons, such that the law can accommodate necessary sex preferences while still strongly pressing against sex stereotypes.

Recognizing the role of the employer in discrimination by customers is especially important in intimate work settings, where discriminatory customer preferences can appear natural rather than cultivated.47 As Bartlett and Gulati acknowledge, employers play a role in cultivating discriminatory customer preferences.48 This is no less true in intimate settings.49 For example, one court has recognized that a spa perpetuated customers’ biases regarding the sex of their massage therapist by asking for their sex preference.50 The court found that the spa could have instead provided customers with a “description of the therapists’ qualifications” and could have quelled privacy concerns by informing customers of draping policies and telling them that they “can instruct therapists about where they may and may not touch.”51 Compounding this issue is that whenever firms accommodate customers’ discriminatory preferences, they reinforce customers’ preexisting view that this is the only acceptable way these services may be delivered. Such privacy-based preferences thus need to be scrutinized carefully to ensure that the firm is not playing a role in cultivating the preference.   

It is difficult—and perhaps unwise—for the law to attempt to distinguish between customer preferences in these intimate settings that are based in malign “stereotypes linking sex to competence, such as . . . patients’ preferences for female nurses or male doctors,”52 and benign stereotypes linking sex to competence, such as patients’ preferences for female gynecologists.53 Bartlett and Gulati are clear that the former should be rejected, but suggest that the latter might be accepted because “[a]fter a history of male control over women’s reproductive lives, . . . many women prefer female gynecologists, and believe that they receive better care from them.”54 

The law of sex discrimination, however, has urged caution around this line of reasoning, given the history of harmful sex stereotyping based in benign justifications.55 Rejecting such preferences should not raise hackles about customer autonomy, as such preferences are not fixed, but are susceptible to the influence of law. Returning to the example of the gynecologist, while women currently prefer female gynecologists, this preference arose only relatively recently.56 Until just a few decades ago, when gynecology was a male profession, women saw male gynecologists without complaint. This shift in preference was prompted in large part by Title VII, which opened up the medical profession to women.

Moreover, privacy-based sex preferences rely on and reinforce heteronormative assumptions that are increasingly out of sync with Title VII, which has moved toward protecting LGBT employees.57 These preferences are based in the view of same-sex spaces as no-sex spaces, that is, that customers can be comfortable that sex in general and that sexual threats in particular will not be a problem so long as spaces remain sex segregated.58 A law crediting customer preferences premised in a sex binary is increasingly out of place in a world that is coming to accept those who fall outside of it. The need to scrutinize privacy-based preferences closely is thus all the more pressing today.  

IV.     Conclusion

This Response aimed to situate Bartlett and Gulati’s proposal within current antidiscrimination law, identifying how big of a departure such a shift in the law would represent, what challenges it might pose, and how it might not go far enough in challenging the role of employers in cultivating discriminatory customer preferences in the context of intimate work.  While I conclude that their proposal would mark a significant shift in the law, I also conclude that this is no weakness of their proposal.  Rather, their cataloguing of the ways that employers cultivate and reinforce discriminatory preferences should be considered in other contexts, and should be recognized as extending even to the most intimate of work settings.

  1. [1]. Compare 42 U.S.C. § 2000e-2(a)(2) (2012) (banning discrimination in employment), with Elizabeth F. Emens, Intimate Discrimination: The State’s Role in the Accidents of Sex and Love, 122 Harv. L. Rev. 1307, 1309–15 (2009) (explaining how the law does not intervene to prevent discrimination in romantic relationships).

  2. [2]. See generally Katharine T. Bartlett & Mitu Gulati, Discrimination by Customers, 102 Iowa L. Rev. 223 (2016).

  3. [3]. Bartlett & Gulati, supra note 2, at 227.

  4. [4]. Id. at 249.

  5. [5]. Id.

  6. [6]. Id. at 247, 249.

  7. [7]. See David Benjamin Oppenheimer, Negligent Discrimination, 141 U. Pa. L. Rev. 899, 900 (1993) (cataloguing parts of employment discrimination law that already embody a negligence and arguing for a broader embrace of this theory of liability).

  8. [8]. Bartlett & Gulati, supra note 2, at 247, 249.

  9. [9]. Id. at 247.

  10. [10]. Id. at 249, 251.

  11. [11]. See id. at 250 (suggesting that liability should apply when customer discrimination arises in circumstances “that [firms] could prevent”).

  12. [12]. See id. (suggesting that liability should apply “when customer bias causes harm and there are reasonable ways businesses could change their practices to prevent that harm”) (emphasis added); id. at 253 (“The proposed rule would only require firms to take reasonable steps calculated to end the harmful effects of discrimination by its customers.”) (emphasis added)

  13. [13]. Bartlett and Gulati correctly state that “current law generally does not hold firms accountable for customer discrimination that their own practices allow and that they could prevent.” Bartlett & Gulati, supra note 2, at 250. However, they also recognize that employment discrimination law regulates discrimination by customers by banning employers from taking adverse employment actions on the basis of customers’ discriminatory preferences, except when justified by a statutory exception under the bona fide occupational qualification. Id. at 252 & n.120, 253 & nn.125–26. Employment discrimination law also addresses discrimination by customers by requiring employers to take action to prevent and correct sexual or racial harassment of employees by customers. See Noah D. Zatz, Managing the Macaw: Third-Party Harassers, Accommodation, and the Disaggregation of Discriminatory Intent, 109 Colum. L. Rev. 1357, 1359 (2009).

  14. [14]. There is disparate impact liability in theory, but in practice it is quite limited. See Bartlett & Gulati, supra note 3, at 250; Michael Selmi, Was the Disparate Impact Theory a Mistake?, 53 UCLA L. Rev. 701, 738–40 (2006).

  15. [15]. See generally Oppenheimer, supra note 8; Zatz, supra note 13.

  16. [16]. See generally Zatz, supra note 13.

  17. [17]. Title VII places negligence-like liability on employers to prevent harassment by coworkers and not just customers. See Oppenheimer, supra note 7, at 944–67.

  18. [18]. See Zatz, supra note 13, at 1377–80.

  19. [19]. See Bartlett & Gulati, supra note 2, at 250.

  20. [20]. Disparate impact claims on these theories are not likely to prevail. Id. at 250.

  21. [21]. This might be especially true for at least a subset of workers—intimate workers—who interact with customers in an intimate way. See Naomi Schoenbaum, The Law of Intimate Work, 90 Wash. L. Rev. 1167, 1187–89 (2015).

  22. [22]. For more discussion rejecting negligence-like liability here, see id. at 1238.

  23. [23]. See Laura A. Rosenbury, Working Relationships, 35 Wash. U. J.L. & Pol’y 117, 129–34 (2011); Naomi Schoenbaum, Towards a Law of Coworkers, 68 Ala. L. Rev. 605, 613–14 (2017). Generally, discriminatory preferences of both customers and coworkers can lead to discrimination against workers even if the firm itself has no “taste” for discrimination. See Gary S. Becker, The Economics of Discrimination 55, 75 (2d ed. 1971).

  24. [24]. They indicate that the biggest drawback of imposing liability on the firm rather than the customer would be that some forms of discrimination by customers would simply escape liability. See Bartlett & Gulati, supra note 2, at 255 (“Perhaps the biggest gap in the proposed rule is that it will have no effect on businesses that customers avoid because of race or gender bias against the business.”).

  25. [25]. See Schoenbaum, supra note 21, at 1193–96.

  26. [26]. Katie R. Eyer, That’s Not Discrimination: American Beliefs and the Limits of Anti-Discrimination Law, 96 Minn. L. Rev. 1275, 1276 (2012) (noting that “less than 5% of all discrimination plaintiffs will ever achieve any form of litigated relief,” and that “dismissals (on motions to dismiss or at summary judgment) are extremely common in discrimination litigation, accounting for a full 86% of litigated outcomes”).

  27. [27]. Id. at 1278.

  28. [28]. Id. at 1279 (citing studies showing “that most individuals think of discrimination as a phenomenon that is explicit” and “restricted in its manifestations”).

  29. [29]. Bartlett & Gulati, supra note 2, at 248 (“The problem is, rather, determining what constitutes discrimination ... .”).

  30. [30]. Id.

  31. [31]. Id. at 251.

  32. [32]. Id. I return to this point below. See infra notes 50–56 and accompanying text.

  33. [33]. Id. at 252–53. As Bartlett and Gulati acknowledge, this may raise First Amendment concerns. Id. at 235. Although these are not foreign to antidiscrimination law (think of liability for sexual harassment based in discriminatory speech), they are heightened under a legal standard that divorces liability for discriminatory expression from a requirement that such expression result in an adverse employment action or the equivalent.

  34. [34]. Id. at 253.

  35. [35]. See, e.g., Beyer v. Cty. of Nassau, 524 F.3d 160, 166 (2d Cir. 2008) (requiring adverse employment action for Title VII claim to proceed); Jones v. Reliant Energy—ARKLA, 336 F.3d 689, 692 (8th Cir. 2003) (requiring the same).

  36. [36]. See, e.g., Faragher v. City of Boca Raton, 524 U.S. 775, 775 (1998) (addressing when the employer should be liable for sexual harassment when there was no adverse action taken by the employer); Harris v. Forklift, 510 U.S. 17, 21–23 (1993) (addressing when a sexually hostile environment without a tangible adverse action rises to the level of a Title VII violation).

  37. [37]. Bartlett & Gulati, supra note 2, at 251.

  38. [38]. See Naomi Schoenbaum, Ignorance as Equality (unpublished manuscript) (on file with author).

  39. [39]. See Naomi Schoenbaum, It’s Time That You Know: The Shortcomings of Ignorance as Fairness in Employment Law and the Need for an “Information-Shifting” Model, 30 Harv. J.L. & Gender 99, 100 (2007) (discussing limits on preemployment inquiries).

  40. [40]. Bartlett & Gulati, supra note 2, at 238.

  41. [41]. Wendy Brown Scott, Transformative Desegregation: Liberating Hearts and Minds, 2 J. Gender, Race, & Just. 315, 357 (1999).

  42. [42]. Schoenbaum, supra note 21, at 1180–83.

  43. [43]. Bartlett and Gulati state that “there has been no serious attempt to weigh the strength of society’s present interest in preventing customers from discriminating ... against society’s present commitment to personal privacy and autonomy for its citizens.” Bartlett & Gulati, supra note 2, at 241. But we have sometimes done this through Title VII’s BFOQ, which has required courts to weigh the employer’s argument concerning customer preferences against the strictures of antidiscrimination law.

  44. [44]. Id. at 244, 246 (“These are tough cases requiring especially creative strategies—strategies that might both take account of the ‘ordinary’ case in which customer discrimination should be prevented, if possible, and cases in which anti-discrimination goals might be best met by allowing sex- or race-based decisions by customers.”).

  45. [45]. See Schoenbaum, supra note 21, at 1187–89. I am not alone in this argument. See generally Amy Kapczynski, Note, Same-Sex Privacy and the Limits of Antidiscrimination Law, 112 Yale L.J. 1257 (2003) (arguing that customer privacy concerns on which some BFOQ cases rest are just another form of customer preference).

  46. [46]. Bartlett & Gulati, supra note 2, at 244–45.

  47. [47]. See Schoenbaum, supra note 21, at 1193–96.

  48. [48]. Bartlett & Gulati, supra note 2, at 234–38.

  49. [49]. See Schoenbaum, supra note 21, at 1193–96.

  50. [50]. Olsen v. Marriott Int’l, Inc., 75 F. Supp. 2d 1052, 1066–74 (D. Ariz. 1999).

  51. [51]. Id. at 1072–73.

  52. [52]. Bartlett & Gulati, supra note 2, at 244.

  53. [53]. Id. at 246.

  54. [54]. Id.

  55. [55]. See Frontiero v. Richardson, 411 U.S. 677, 684–88 (1973).

  56. [56]. See Tamar Lewin, Women’s Health Is No Longer a Man’s World, N.Y. Times (Feb. 7, 2001),

  57. [57]. See, e.g., Smith v. City of Salem, 378 F.3d 566, 574–75 (6th Cir. 2004) (holding that discrimination on the basis of transgender identity constitutes sex discrimination under Title VII); Rene v. MGM Grand Hotel, Inc., 305 F.3d 1061, 1068–69 (9th Cir. 2002) (Pregerson, J., concurring) (explaining that sexual harassment of a gay male plaintiff is sex discrimination because the man was harassed for failing to meet masculine stereotypes). The EEOC has recently determined that discrimination on the basis of transgender identity and sexual orientation both fall within Title VII’s ban on sex discrimination, and some courts have followed suit. Hively v. Ivy Tech Cmty. Coll. of Ind., 853 F.3d 339, 351–52 (7th Cir. 2017) (“We hold only that a person who alleges that she experienced employment discrimination on the basis of her sexual orientation has put forth a case of sex discrimination for Title VII purposes.”); EEOC v. Scott Med. Health Ctr., P.C., No. 16-225, 2016 WL 6569233 (W.D. Pa. Nov. 4, 2016) (agreeing with the EEOC interpretation), Isaacs v. Felder Servs., LLC, 143 F. Supp. 3d 1190, 1193 (M.D. Ala. 2015) (same).

  58. [58]. Schoenbaum, supra note 21, at 1191–92; see also Naomi Schoenbaum, Heteronormativity in Employment Discrimination Law, Washburn L.J. (forthcoming 2017).


Associate Professor of Law at the George Washington University Law School.

I thank Michael Abramowicz, Katharine Bartlett, and Mitu Gulati for offering generous feedback on a draft of this response.