Botching Bostock

Yesterday, the Supreme Court of the United States, in the case Bostock v. Clayton County, Georgia, ruled in a 6-3 decision that Title VII of the Civil Rights Act of 1964 prohibits employers from discriminating against their employees on the basis of sexual orientation and gender identity.

Neil GorsuchThe Court’s opinion was written by Justice Gorsuch and joined by Justices Roberts, Ginsburg, Breyer, Sotomayor, and Kagan. Justice Alito wrote a dissenting opinion, joined by Justice Thomas. A second dissenting opinion was given by Justice Kavanaugh. All three opinions can be read in full here.

The relevant statute of Title VII reads as follows:

It shall be an unlawful employment practice for an employer … to fail or refuse to hire or to discharge any individual, or otherwise to discriminate against any individual with respect to his compensation, terms, conditions, or privileges of employment, because of such individual’s race, color, religion, sex, or national origin…

The Court argued, in effect, that discrimination based on sexual orientation or gender identity necessarily involves discrimination based on sex (properly understood as biological sex: male or female) and thus is prohibited by Title VII.

I’m neither a lawyer nor the son of a lawyer, but I know a thing or two about logic and argumentation, so I want to explain, as clearly and concisely as I can, why I think the Court’s central argument is horribly confused and specious.

If we’re going to criticize the Court’s opinion, however, it’s important to recognize how the Court argued. Some commentators have objected to the ruling on the basis of the harmful consequences it will have (undermining protections for women using bathrooms and locker rooms, destroying women’s sports, etc.) but that misses the proper role of the Court. The Court’s task is to interpret the law; in this case, the relevant clause of Title VII. If it turns out that the law has unforeseen or unintended consequences — harmful consequences — surely that’s a fault with the law, to be remedied by the legislative branch, not a fault with the judicial ruling.

Other commentators have argued that the original legislators couldn’t plausibly have understood Title VII to prohibit discrimination on the basis of sexual orientation or gender identity, nor could they have foreseen that it would be applied in that way. However, Gorsuch directly addresses that objection in the opinion. His contention is that it’s a logical implication of the text of the statute, regardless of whether anyone at the time recognized it. His argument is simply that the text as it was written, reasonably interpreted according to standard dictionary definitions, protects against SOGI discrimination precisely because it protects against sexual discrimination. The latter logically demands the former, so he maintains. The complaint that no one in 1964 would have acknowledged such an implication is legally irrelevant. What’s relevant is that it is in fact an implication of the statute. (Gorsuch cites various precedents where a statute is later applied beyond its originally intended scope on the basis of its implications.)

So what exactly was the Court’s central argument? After establishing the meaning of the words “sex,” “because of,” and “discriminate” in the original context of the Civil Rights Act, the Court concludes:

From the ordinary public meaning of the statute’s language at the time of the law’s adoption, a straightforward rule emerges: An employer violates Title VII when it intentionally fires an individual employee based in part on sex. (p. 9)

That’s uncontroversial as far as it goes. What’s controversial is the next plank of the argument, on which the entire ruling hangs. The Court proceeds to argue that discrimination based on sexual orientation or gender identity necessarily involves discrimination based on (biological) sex. That this is the lynch-pin of the argument is clear from how often the point is repeated in the opinion (bold added for emphasis):

An individual’s homosexuality or transgender status is not relevant to employment decisions. That’s because it is impossible to discriminate against a person for being homosexual or transgender without discriminating against that individual based on sex. (p. 9)

There is simply no escaping the role intent plays here: Just as sex is necessarily a but-for cause when an employer discriminates against homosexual or transgender employees, an employer who discriminates on these grounds inescapably intends to rely on sex in its decisionmaking. (p. 11)

At bottom, these cases involve no more than the straightforward application of legal terms with plain and settled meanings. For an employer to discriminate against employees for being homosexual or transgender, the employer must intentionally discriminate against individual men and women in part because of sex. (p. 12)

When an employer fires an employee for being homosexual or transgender, it necessarily and intentionally discriminates against that individual in part because of sex. And that is all Title VII has ever demanded to establish liability. (p. 14)

But, as we’ve seen, an employer who discriminates against homosexual or transgender employees necessarily and intentionally applies sex-based rules. (p. 15)

By discriminating against transgender persons, the employer unavoidably discriminates against persons with one sex identified at birth and another today. (p. 19)

We agree that homosexuality and transgender status are distinct concepts from sex. But, as we’ve seen, discrimination based on homosexuality or transgender status necessarily entails discrimination based on sex; the first cannot happen without the second. (p. 19)

As I say, the entire ruling hangs on this alleged logical connection between one kind of discrimination (sex-based) and another kind (SOGI-based).

How then does the Court argue the point? First, it articulates a sufficient condition for violations of Title VII:

If the employer intentionally relies in part on an individual employee’s sex when deciding to discharge the employee—put differently, if changing the employee’s sex would have yielded a different choice by the employer—a statutory violation has occurred. (p. 9)

Having established this condition, it proceeds by way of illustrative examples to show that any SOGI discrimination will inevitably meet this condition and thereby violate Title VII. Here are the two paradigmatic cases offered by the Court:

Consider, for example, an employer with two employees, both of whom are attracted to men. The two individuals are, to the employer’s mind, materially identical in all respects, except that one is a man and the other a woman. If the employer fires the male employee for no reason other than the fact he is attracted to men, the employer discriminates against him for traits or actions it tolerates in his female colleague. Put differently, the employer intentionally singles out an employee to fire based in part on the employee’s sex, and the affected employee’s sex is a but-for cause of his discharge. Or take an employer who fires a transgender person who was identified as a male at birth but who now identifies as a female. If the employer retains an otherwise identical employee who was identified as female at birth, the employer intentionally penalizes a person identified as male at birth for traits or actions that it tolerates in an employee identified as female at birth. Again, the individual employee’s sex plays an unmistakable and impermissible role in the discharge decision. (pp. 9-10)

So here’s the reasoning in the first case. Both employees have the trait attracted-to-men. Only one is fired, and the reason he’s fired is because he’s a man. The other employee has exactly the same trait, but she keeps her job because she’s a woman. Ergo, the first employee was discriminated against based on his sex, which Title VII prohibits.

The problem with the example, though, is that it prejudicially describes the situation so as to deliver the conclusion that there was sex-based discrimination. Suppose we say that the relevant trait is not attracted-to-men but rather same-sex-attracted. Under that description, the biological sex of the employee turns out to be irrelevant: “changing the employee’s sex” would not have “yielded a different choice by the employer.” Presumably what the employer objects to is homosexuality as such, regardless of whether it’s male or female homosexuality. (I suppose there could be cases where an employer discriminates against male homosexuals but not female homosexuals, or vice versa, but obviously such cases aren’t in view here.)

Note in particular the reference to “the employer’s mind” in the excerpt above. What is the objectionable trait in the employer’s mind? Is it attraction to men? Or is it same-sex attraction? Clearly the two are not logically or conceptually equivalent. But the entire argument hangs on the first being the relevant trait rather than the second. Yet it’s most plausibly the second that serves as the basis for the discrimination. If that’s the case, the Court’s argument collapses.

The same analysis can be applied to the second example. In the decision of the employer, is the relevant trait identifies-as-male? Or is it identifies-as-other-than-birth-sex? If it’s the second, then there’s no discrimination based on sex, because “changing the employee’s sex” would not yield “a different choice by the employer.” Again we see that the example has been prejudicially constructed so as to ‘trigger’ the Court’s test for sex-based discrimination.

That’s not quite the end of the issue, however. Notice that the Court’s test asks whether “the employer intentionally relies in part on an individual employee’s sex when deciding to discharge the employee” (emphasis added). Gorsuch anticipates the kind of rebuttal I gave above and offers a response to it, namely, that an employer cannot determine whether an employee is homosexual or transgender without reference to the employee’s sex. Thus, for example, Frank can’t tell whether Andy is gay without knowing that Andy is male, and so Frank would have to “rely in part” on Andy’s sex in any decision to hire or fire him on the basis of Andy’s sexual orientation.

Here’s how Gorsuch tries to make the argument, again by way of example:

There is simply no escaping the role intent plays here: Just as sex is necessarily a but-for cause when an employer discriminates against homosexual or transgender employees, an employer who discriminates on these grounds inescapably intends to rely on sex in its decisionmaking. Imagine an employer who has a policy of firing any employee known to be homosexual. The employer hosts an office holiday party and invites employees to bring their spouses. A model employee arrives and introduces a manager to Susan, the employee’s wife. Will that employee be fired? If the policy works as the employer intends, the answer depends entirely on whether the model employee is a man or a woman. To be sure, that employer’s ultimate goal might be to discriminate on the basis of sexual orientation. But to achieve that purpose the employer must, along the way, intentionally treat an employee worse based in part on that individual’s sex. (p. 11; italics original, bold added)

The point is clear: the employer’s decision to fire the “model employee” depends in part on his recognition that the employee is female. Thus, reasons the Court, the discrimination “relies in part” or “is based in part” on the employee’s sex. The employer is quite self-conscious about this. The employer knowingly (and thus intentionally) reaches his decision partly on the basis of the employee’s sex.

The flaw in this argument is that it conflates two distinct things:

  1. discrimination on the basis of X
  2. discrimination on the basis of Y, with reliance on X in the process

That one has to take X into account in order to discriminate on the basis of Y simply does not entail that one is thereby discriminating on the basis of X. It’s entirely possible to adopt a normative stance with respect to Y (favoring some Ys over other Ys) without adopting any normative stance with respect to X, even if one has to take X into account when determining Y.

To make things more concrete, consider this scenario. George owns a store that sells shoes for both men and women. He hires two people as store clerks, Andy and Barbara. Over time, George notices that some of his stock is going missing. Based on good evidence, he concludes that one of his two employees has been stealing items. As he further investigates, he discovers that all the stolen items are men’s shoes. Reasoning that a man would be far more likely to take men’s shoes than a woman, he concludes that Andy is the culprit and fires him.

Now clearly George’s decision “relied in part” on Andy’s sex. It was “based in part” on the fact that Andy is a man rather than a woman. Moreover, George’s reliance on that fact was quite intentional. But should we conclude that George is guilty of discrimination based on sex? Did George violate Title VII?

If you think so, I doubt anything else I could say would persuade you otherwise. The “reliance on sex” in George’s decision-making is clearly benign, yet it parallels the “reliance on sex” in Gorsuch’s hypothetical scenario above. What was the relevant trait or action in George’s decision to fire Andy? Was it male-shoe-stealing or was it simply shoe-stealing? What was George’s motivation for firing Andy? Did it involve any prejudice regarding Andy’s sex? The answers to these questions should be obvious.

Enough has been said, I trust, to demonstrate the fallacious nature of the argument at the heart of the Court’s opinion. Even granting what the Court claims about “the ordinary public meaning” of the Title VII statute, the notion that “discrimination based on homosexuality or transgender status necessarily entails discrimination based on sex” is just flat-out confused. It’s a bad ruling that will have very harmful consequences (and not just for religious employers).

If you have the time and patience, I would encourage you to read all three opinions (the majority and the two dissents). Alito’s dissent is devastating; he completely dismantles Gorsuch’s arguments and lays bare the many problematic implications of the Court’s decision.

June 15, 2020, was not a good day for the Supreme Court of the United States.

12 thoughts on “Botching Bostock”

  1. When Gorsuch claims that “sex is *necessarily* a but-for cause when an employer discriminates against homosexual or transgender employees,” the “but-for” standard is essentially a reference to a counterfactual analysis of the decision-making process. (‘If things were otherwise sexually, then a different decision would have been made, and so *sexuality* is the basis for the decision.’)

    As those familiar with the philosophical debate over the analysis of causation know, reducing causation to counterfactual dependence introduces an element of arbitrariness in selecting *which* factor made the difference. Pragmatic or contextual theories of causation highlight this phenomenon. Gorsuch doesn’t seem to recognize that a logical analysis of the “but-for” standard undermines his claim of necessary implication. Given this, his language of ‘impossible,’ ‘necessarily,’ ‘inescapably,’ etc. is wholly inappropriate, and his central argument collapses. Your post just brings this out more clearly.

    On your final illustration, I can imagine some detractors claiming that your shoe-stealing scenario *did* involve sexual discrimination, since the employer made an assumption about which kind of shoes men would be likely to steal. (The discussion has become so polarized nowadays that even *considering* biological sex for any reason whatsoever often counts as sexual discrimination.)

    But there’s an easy way to remove altogether the subjective assumption about ‘what men would be likely to do’. Just tweak the shoe store illustration so that partial DNA left behind in the locked storeroom indicates male DNA. On this basis, they fire the only male employee. Obviously, they had to refer to the sex of the employee to make the decision to fire, but he wasn’t fired *because* he was male (or allegedly did a ‘male kind of thing to do,’ like stealing male shoes). Rather, he was fired for *stealing*, even though this determination was made by way of referencing his sex. (The defendant can’t reasonably claim, “I’m being fired *because of* sex.”) Likewise, having SOGI that differs from one’s biological sex cannot be discerned without referencing biological sex, but one is being fired for SOGI, *not* biological sex. The basic epistemological point seems straightforward and clear.

    Of course, if it was a DNA sample that identified a particular person, then no reference to sex is needed, and the example isn’t parallel in the sense you need. But if it’s a partial DNA sample (which happens a lot), then only some things can be determined (like sex), while not picking out any person in particular. In that case, the employer has to rely on his knowledge of the employee’s sex to make the decision to fire, but he’s being fired because there’s compelling evidence he stole something, and not because of sex. That much seems obvious, as you say.

  2. Ron DiGiacomo

    >>>Under that description, the biological sex of the employee turns out to be irrelevant: “changing the employee’s sex” would not have “yielded a different choice by the employer.”<<<

    James,

    Aren’t you treating the sex of the employee and the employee’s attraction both as variables? I think to be fair to the court’s logic, the attraction to men should be seen as a constant whereas the sex of the employee should be considered a variable.

    If I’m correct, then “changing the employee’s sex” from male to female would indeed have “yielded a different choice by the employer” (if we allow attraction to remain constant). We’d no longer have a male attracted to a male but rather a woman attracted to a male. Under such a scenario the employee would keep the job if his sex were female (all else remaining constant). To be fair to the court’s argument I don’t think we may change both the sex of the employee and the attraction of the employee.

    Even if we change the criterion to SSA, then by manipulating the *only* variable in a SSA scenario (e.g. the sex of the employee), the resultant outcome becomes opposite sex attraction, which again allows the employee to keep the job. I don’t think we deal fairly with the court’s logic if we simply manipulate the sex of the employee and then treat the attraction as a variable that is inexorably linked to the same sex of the employee.

    There are other issues with SCOTUS’ ruling that are problematic.

    1. Ron,

      “Aren’t you treating the sex of the employee and the employee’s attraction both as variables?”

      No, I’m pointing out that you get different results depending on how the objectionable trait is defined.

      “Even if we change the criterion to SSA, then by manipulating the *only* variable in a SSA scenario (e.g. the sex of the employee), the resultant outcome becomes opposite sex attraction, which again allows the employee to keep the job.”

      I don’t understand this. If (1) the employee is a man attracted to a man, (2) the relevant trait is SSA, and (3) the sex of the employee is changed, the outcome is an employee who would be a woman attracted to a woman. The employee would still have been fired (given that the presenting concern is discrimination on the basis of sexual orientation) but there would be no discrimination on the basis of sex (precisely because changing the sex doesn’t change the outcome).

      1. Right. If the relevant trait is SSA, then you’d be fired *whether or not* you’re a man or a woman. How does that entail discrimination *on the basis of* sex? That would be the absence of sex discrimination.

        Yes, singling out “but-for biological sex” gives the appearance that the latter is the factor that ‘makes a difference.’ But singling out “but-for SSA” instead makes the latter the factor that ‘makes a difference’. Unless Gorsuch can show why the former choice of description is required rather than the latter, his conclusion that SOGI discrimination *entails* sex discrimination simply doesn’t follow. But his argument requires the entailment, since he explicitly notes that these categories are conceptually distinct.

  3. Ron DiGiacomo

    Sorry, let me try again.

    Maybe leave (2) out for a moment and then we can reintroduce it. I think we are in agreement.

    Here’s how I’d see the court responding to your rejoinder.

    If “the employee is a man attracted to a man” and *only* “the sex of the employee is changed” (leaving attraction to males *constant*), then from SCOTUS’ perspective we are not left with: “the outcome is an employee who would be a woman attracted to a woman.” Rather, aren’t we left with a female (variable) with male attraction (constant)? In that sense, getting the job is a function of sex; (yes, as it relates to SSA). From there we can drill down even further.

    When dealing with woman vs. man up for the same job, the attraction to males either is acceptable to an employer or not based upon the sex of of employee. Yes, the disqualifying trait in question is SSA. But without a little more work, I don’t see how we are fairly interacting with SCOTUS merely by our flipping the sex of the male without also holding the attraction to males constant. Reason being, I think their position requires holding the (male) attraction constant, which means I would have to deal with that internal constraint in my critique of their position. After all, isn’t their point that attraction to males (being constant) shouldn’t tip a job toward a woman over a man.

    Let me try to approach your point of SAA another way.

    SCOTUS reasons from a man and woman up for the same job, both being attracted to men. Woman gets the job because of attraction to men. Man doesn’t get job because of attraction toward men. The constant is attraction to men. The variable is sex of applicant. If man with attraction toward men was a woman (yet with the same attraction toward men being constant), then she’d get the job.

    My question for SCOTUS.

    Rather than woman vs man for the job, what if Rocco and Skippy are up for the same job? Rocco gets it because of Skippy’s SSA. What can SCOTUS say about sexual discrimination? Nothing. Must’ve been a SSA thing all along. SSA is the only distinguishing factor. Now we apply that to woman vs man scenario.

    1. Ron DiGiacomo

      Hi Greg,

      Yes, the distinguishing variable is SSA. My particular interest was in how best to untangle and ultimately tease out SSA from SCOTUS’s “sexual discrimination” interpretation of the scenarios.

      It’s hard to believe that all conservative judges didn’t see that the court’s sexual discrimination paradigm glaringly breaks down when both employees or applicants are of the same sex. If liberal judges saw the problem, I wouldn’t expect that to have changed anything,

    2. “But without a little more work, I don’t see how we are fairly interacting with SCOTUS merely by our flipping the sex of the male without also holding the attraction to males constant.”

      But this misses the point entirely, which is that the Court’s selection of the ‘constant’ is arbitrary; it lacks the necessity that Gorsuch’s argument requires. This is precisely the problem Greg emphasized in his first comment. Define the constant one way: voilà, SOGI discrimination is sex discrimination! Define it another (actually more sensible) way: oh, now it isn’t!

      1. Ron DiGiacomo

        James,

        I wasn’t interacting with Greg’s first post but with your original piece.

        Even if I were to grant that the court’s implicit constant was arbitrary, to assert SSA as the relevant variable doesn’t of itself establish biological sex as irrelevant. I suspect that’s why you took time to try to tease out and distinguish the relevant difference between one’s sex and SSA. (You didn’t just dismiss SCOTUS’s variable as arbitrary.) In doing so, you made both attraction and sex variables. I find that problematic for reasons already cited.

        Best wishes,

        Ron

        1. “In doing so, you made both attraction and sex variables.”

          No, I didn’t. I held the attraction constant; I just defined the attraction in different (non-question-begging) terms. That’s the entire point. :)

          1. Ron DiGiacomo

            You manipulated the person’s sex and the attraction was altered with it. That is not to keep attraction “constant.” It’s to keep attraction inexorably tied to a variable. What you’re now defining as “constant” has become technically a function of the variable. But constants don’t work that way. Constants by definition don’t change with the variable. If the person’s sex was treated as a variable, which it was, then so was the attraction if it was altered with the sex variable.

            Aside from what I find to be a somewhat esoteric tagging of terms, I’ll leave you with my other points. I’ll also continue to reflect on what you’ve stated.

  4. Ron DiGiacomo

    James,

    I’ll take one more swipe at resolving our difference.

    SCOTUS: If a man with x talent and attracted to men loses a job to a woman with identical x talent and attracted to men, the variable is not x talent or attraction to men. Both those variables are the same for each person. Therefore, SCOTUS reasons (badly) that the variable is sex, implying sex discrimination. That’s SCOTUS’ position.

    Your point is that sex discrimination is not the discriminating variable but rather SSA. I agree. To make that point, you argue that if the man with SSA was instead a woman with SSA, the woman with SSA would be discriminated against no differently than a man with SSA. You reason that since SSA for men and women is essentially the same trait, the only discriminating variable left is SSA.

    (I agree that SSA for men and women are essentially the same trait. There’s no relevant distinction. AND, it would be incumbent for SCOTUS to show otherwise, lest they appear arbitrary. However, I think it’s also incumbent upon us to show how the relative trait must be SSA as opposed to sex, as obvious as that might appear to us without argumentation.)

    Please consider, SCOTUS would argue back that both the woman with SSA and the man with SSA would be discriminated against based upon sex: The woman with SSA (i.e. attraction to women) would be discriminated against *relative to* a man with the exact same attraction to women. In such a case, the talent would be the same and the attraction to women would be the same. The discriminating factor would be sex. Likewise, the man with SSA (i.e. attraction to men) would be discriminated against relative to a woman with the exact same attraction to men. In such a case, the talent would be the same and the attraction to men would be the same. The discriminating factor would again be sex. By flipping the man with SSA to a woman with SSA we haven’t put the matter to rest.

    You and SCOTUS are at an impasse, perhaps both calling the other position’s variable arbitrary.

    Obviously your instincts are correct and SCOTUS’ are incorrect. One way to show that is by taking sex off the table(!) and making the *only* variable SSA. We do that by comparing a man attracted to men to another man of equal talent who is not attracted to men. The former man loses the job based upon his attraction to men. The *only* variable is SSA. We’ve left SCOTUS without an appeal to sex discrimination. Sex is no longer a variable given two men with one discriminated against. (We can apply the same reasoning to two women, only one with SSA.)

    The point is, as long as we continue to argue man vs woman (both with same talent and same attraction to men), we will be at an impasse with SCOTUS. They’ll attribute the variable to sex whereas we understand it correctly to be SSA. I don’t think that’s adequately addressed until we show that it cannot be be sex.

    Like other disagreements we’ve had, I also chalk this one up not to a different instinct or different conclusion but rather to the way in which we might arrive at that conclusion by refuting the opponent in a way that leaves no imaginable rejoinder.

Comments are closed.