As I mentioned last week, ten years ago I wrote a descriptive and analytical law review article called Private Employees’ Speech and Political Activity: Statutory Protection Against Employer Retaliation, which aims to catalog these often-little-known statutes. This year, I’m returning to the subject, trying to analyze the strongest arguments for and against such statutes. The article (Should the Law Limit Private-Employer-Imposed Speech Restrictions?) will be published later this year in a Journal of Free Speech Law symposium issue, together with other articles that stemmed from an Arizona State symposium on Non-Governmental Restrictions on Free Speech; and this week and next I’d like to serialize it here.
Last Tuesday and Wednesday, I blogged the introduction and the beginning of the argument in favor of such statutes, followed by an explanation of why such statutes usually don’t violate employers’ constitutional rights. This week, I discuss some other arguments against such statutes (and you can see the whole article right now, if you’d like, by looking at the PDF). I begin today with arguments that the statutes should be rejected in order to protect employers’ rights, even if not constitutionally protected rights, not to associate.
[* * *]
Even in the absence of a constitutional right not to associate, of course, all of us have an interest in choosing whom we deal with, whether as friends, fellow club members, business partners, or employees. We may not want to deal with Nazis or Communists or Klansmen or supporters of terrorist movements. We may feel this particularly strongly because of our personal experiences: The child of refugees from Cuba may not want to deal with people who praise Che Guevara; the widow of a police officer who was killed in the line of duty may not want to deal with people who praise attacks on police, or call police officers “pigs”; A black or Jewish employer might especially not want to deal with Klansmen. Likewise, a business owner might want to cut off relations with someone because the business’s other employees or customers don’t want to associate with the person.
But when it comes to employment, our legal system has long generally rejected such claims. An employer may not want to employ union leaders, whose organizing, bargaining, and strike-threatening behavior he sees as jeopardizing the business he has spent a lifetime building up. An employer may not want to employ someone who made discrimination complaints against the employer in the past, perhaps on grounds that the employer views as ill-founded.
A Catholic employer may not want to hire Jehovah’s Witnesses, whose beliefs sharply condemn Catholics. A gay employer may not want to hire people who belong to religious groups that condemn homosexuality (especially ones that the employer views as historically responsible for discrimination or even violence against gays). A Palestinian-American employer who views his nation as being at war with Israel may not want to hire Israeli Jews. Yet the employers are barred from refusing to hire employees on these grounds.
One could respond to these analogies this way: When it comes to constitutionally protected discrimination—for instance, a church’s right to choose its clergy, or a person’s right to choose a spouse, or a person’s right to choose a roommate—the Constitution protects the Right not to associate regardless of whether judges or legislators approve of the reason for the discrimination. But when it comes to economic association, the law should be solicitous of reasonable bases for refusals to deal but not of bigoted or unreasonable ones.
Not wanting to deal with people because of their skin color or nationality or sexual orientation or beliefs about the Trinity is unreasonable and hateful, the argument would go. But not wanting to deal with people because of their political speech or beliefs is reasonable often enough (even if not always), and we should accommodate employers’ desire to do that, even if the Constitution would let us ban such discrimination.
But I don’t think that’s quite right. There’s nothing inherently unreasonable or hateful in an employer’s not wanting to employ union leaders, or employees who had made past complaints. There’s nothing inherently unreasonable or hateful in an employer’s not wanting to deal with employees religious beliefs and don’t just turn on theological questions but sharply condemn the employer’s religion or sexual orientation or other actions or beliefs. There’s nothing inherently unreasonable or hateful in an employer’s not wanting to deal with employees who have voted for certain candidates.
The legal system bans such discrimination not just because it’s somehow hateful or irrational (though some religious discrimination is indeed that), but in large part because it’s valuable to promoting employee collective bargaining, or employee whistleblowing about illegal practices, or the freedom to vote or to openly worship. Likewise, for the reasons given in Part I, it’s valuable to promote the freedom to openly speak and participate in politics beyond just voting.
One might also argue that antidiscrimination laws do burden people’s (nonconstitutional) freedom to choose whom to associate with, but we enact them because there’s massive, society-wide discrimination on certain bases, which justifies a rare departure from a norm of unregulated employment decisionmaking. Antidiscrimination laws were needed to break broad patterns of refusing to hire blacks, Hispanics, women, Jews, Catholics, gays, union, whistleblowers, and the like. But for rarer forms of discrimination, however foolish they may be, it’s enough to rely on the employment market, in which most employers will be reluctant to fire good employees, and most employees needn’t fear being fired, especially since they could easily find other jobs.
And perhaps this does help explain why, for instance, many states haven’t banned discrimination based on marital status, height, weight, off-duty smoking, having children, being a crime victim, and the like, though some states have. No need to restrict employer freedom via employment discrimination law—especially given how expensive employment litigation can be, and how it can chill legitimate problems employer action (more on that below)—to deal with that are fairly rare, the argument would go. And maybe this should also apply to political discrimination, if one thinks it’s rare enough.
This, I think, is likely the strongest combination of arguments against the private employee speech protections I discuss here. The key questions, I suppose, are: (1) How often do employers discriminate (or expressly or implicitly threaten to discriminate) based on employee speech and political activity, especially at a time that is widely seen as more politically polarized than before, and a time in which social media has made it easier than ever to call for boycotts triggered by an employee’s unpopular political views? (2) How much does this chill the free speech we want to see for the sake of self-government, the search for truth, and self-expression? And I don’t have a confident answer to those questions.
 Cf. Catnwell v. Connecticut, 310 US 296, 301 (1940).
 Cf. Espinoza v. Farah Mfg., 414 US 86, 88 (1973) (concluding that “national origin” in Title VII “refers to the country where a person was born,” as well as “the country from which his or her ancestors came”); Minn. Stats. § 363A.03 subd. 25 (defining “national origin” as “the place of birth of an individual or any of the individual’s lineal ancestors”), quoted favorably in id. at 88 n.2 (citing earlier version of statute).
 Some judges might, for instance, take the view that refusal to engage in interracial marriage is morally wrong (or at least is highly shallow, if it’s based on preferences as to appearance rather than based on racial hostility), but people must have a categorical right to choose not to marry someone, even for bad reasons.
 See Discrimination—Employment Laws, Nat’l Conf. of State Legislatures (July 27, 2015), https://www.ncsl.org/research/labor-and-employment/discrimination-employment.aspx. This list notes 19 states that have banned marital status discrimination, and only a few that have banned discrimination based on the other factors I note.
 More broadly, of course people who are skeptical of bans on private discrimination generally, and would either reject them all or limit them to just a very few most harmful forms of discrimination (as many libertarians would), might resist any attempt to add to such laws.