Consider the Bathroom Break (Workplace Domination Part Three)

The virtue of the Crooked Timber bloggers’ objections to the Bleeding Heart Libertarians’ line is that it implicitly suggests the difference between liberal and republican conceptions of freedom. Libertarians have usually substituted theories of interference and coercion for a full-blown theory of domination. When Chris Bertram stopped by, he suggested that they wanted to avoid this theoretical debate, but I think it impoverishes the conversation a bit. I’ve written about this at length in the past, but for the purposes of navigating the Bleeding Hearts/Crooked Timber debates on workplace domination, here’s a quick primer on interference, coercion, and domination.

In the comments to my first post, Daniel Levine asked:

“Why, exactly, do you need the distinction between actual productivity-enhancing rules and dominating ones? If your target is domination, it doesn’t seem to matter if the domination enhances productivity, for two reasons.

1. Productivity is, at best, an indirect good for most workers (it may be more directly beneficial for academic workers, since we tend to be abnormally non-alienated from our work).

2. The fact that domination is “good for me” in terms of some of my interests does not generally make it any less domination or much less morally problematic. Part of the core insight, I think, of Pettit, is that domination is about whether my situation is responsive to my own agency and conception of my good, not just whether it maximizes my own interests.”

There’s a debate in the literature about this: one way of defining domination is as arbitrary interference. Arbitrariness generally means “choosing or not choosing at pleasure,” and this is the sense that I believe is correct. Philip Pettit has offered a defense of a stricter, more substantive sense of arbitrariness as “failure to track people’s interests according to their ideas,” but I think this is unworkable at the level of the firm. Certainly, in society as a whole we must avoid arbitrary interference with a person that is not in keeping with their conception of the good, but at the level of the firm it is appropriate to require that all workers set aside their conflicting conceptions of the good and join, for at least as long as they are on shift, with other workers whose efforts are bent toward a common good envisioned by the firm. I would argue that we should also set aside (or translate) our individualistic conceptions of the good when we enter the public sphere and act as citizens, but this is a long debate that is far afield.

Consider the bathroom break: on my view, an “arbitrary” interference would involve some non-principled way of awarding bathroom breaks, like the sweatshop practice of asking permission from pit bosses, who can play favorites or simply deny all requests. In such contexts, it is very demeaning to be denied the right to engage in basic bodily functions. You’ve got to pee when you’ve got to pee.

Yet things are a bit different on an assembly line, where if one person leaves the whole line has to stop. In general, the most non-arbitrary way to deal with this is to have a few extra relief workers available to swap places so that a worker with an urgent need can quickly head to the bathroom. A factory with such a system is simply more productive than one without it, but if everyone needs a break at once, someone may have to wait: this is not domination.

There are definitely domination violations in the sanitation arena, and these are policed by OSHA, which has pretty specific rules about bathrooms and sanitation. Yet they have never enforced a quantitative standard (i.e. four breaks per shift, at least one relief worker per twenty employees) because different industries and different people have different needs.

Instead, OSHA have chosen the famously legalistic “reasonableness” standard. And they actually cite and levy fines very infrequently, either because infractions are few and far between or because their enforcement is quite lax. Evidence suggests that former, but so does basic economics: this kind of domination is simply not productivity-enhancing.

Of course, even bathroom breaks are not always a matter of dignity: a traditional assembly line worker dealing with intestinal issues is going to seriously hamper the functioning of that assembly line, so it’s better for the company to let that worker call in sick (for pay, to avoid people showing up when they shouldn’t) and have replacements available. By way of analogy, consider that long haul buses don’t usually have bathrooms on them, so you have to wait until the next stop, which may be a few hours away. You’re not being “dominated” in that sense, it’s just that onboard bathrooms are too expensive. This may not correspond directly to a persons’ “interests according to their ideas.” Somebody with a chronic condition (IBS, say) might object that this prevents them from taking certain kinds of jobs. The ADA standard of “reasonable accommodation” seems pretty appropriate here: it’s a vague principle, but in practice and through judicial and agency interpretations it gets fleshed out in a pragmatic way that is aligned towards overall productivity.

Despite this slight departure, Pettit and I agree that “constitutional provision” is generally superior to “reciprocal power,” and indeed the Crooked Timber bloggers depend on this to be the case to defend voice over exit. The easiest way to enact “reciprocal power” is grant both worker and manager equal power to destroy the bond that joins them: managers command, but workers can quit. Thus, the manager can arbitrarily interfere with the worker, but the worker can do likewise by leaving, forcing the employer to search for new employees. No one enforces a “you can’t quit because your manager won’t sleep with you” rule, so (if the power is truly reciprocal) no one would need to enforce a “you can’t fire him because your worker won’t sleep with you” rule.

Of course, the power relations between workers and employers aren’t generally reciprocal! But this is not to say that they never are or cannot be: some high skill, high demand workers can negotiate these issues without help. It’s even possible to imagine some employees who have more power than their employers: consider doctors and lawyers, for instance, who we rarely speak of as “dominated” by their patients and clients.

What makes those professions powerful enough to escape arbitrary interference? One word: exit. Though the Crooked Timber bloggers contest this, I believe that generous unemployment benefits or a basic income guarantee are adequate to supply a kind of reciprocity. Part of this is tied to the “full employment” question: if there is a reserve army of unemployed than the protesting worker does not really have reciprocal power. BUT! The basic income guarantee militates in favor of full employment, which the CT bloggers gloss over in their analysis.

But that doesn’t mean that exit is sufficient. A firm made non-dominating merely by exit options and reciprocal power is one where only constant vigilance and threats deter resurgence of domination. In that sense, this would be a civil “war of all-against-all,” a Rawlsian modus vivendi. Both firms and employees should want more than that, and not just because of domination concerns.

The imposition of procedural rationality in the workplace limits both employer and employee arbitrariness in a way that benefits both. The principled restriction is that both are motivated by the same goal, of engaging in productive work. Most arbitrary interference is at odds with this goal. This is akin to “constitutional provision.”

What I mean by “constitutional provision” is a set of procedures for preventing arbitrary domination rather than for redressing it after the fact. (Rights not revenge.) In the political setting where a constitution limits the activities of the executive and legislative branches, we can say that the legislature is less free to make certain arbitrary interventions into the lives of its citizens. In the same way, constitutional provision in the workplace limits the range of restrictions that employers can place on employees.

Firms are an important part of the republic, and a republic that focuses on non-domination is going to want to prevent domination within the firm. One way to achieve this goal is to couple restrictions on the scope of arbitrary interference within the firm with adequate exit options. But none of this will grant “voice” alone. OSHA doesn’t give an employee “voice,” it gives employees recourse and an opportunity for reprisal after domination has taken place.

Let’s consider two paradigm cases of voice: an employee identifies favoritism, productivity-reducing arbitrary interference from a manager, and then brings attention to the manager’s superiors. This kind of voice benefits both the firm and the employee, but it requires that the employer create a protected channel of communication. If the only person the effected employees can complain to is the manager engaging in favoritism, the employees will be silenced by fear of reprisals. In this case, voice is aligned with the principle of productive collaboration rather than domination, and arbitrary interference results in lost productivity.

Now consider a productivity-reducing use of voice: employees protest automation because fewer workers will be needed to supply the same product. (I’ve recently been having this conversation with fellow academics about Coursera.) In aggregate, the world is better off if the same product (a car or an education) can be produced more cheaply. It doesn’t matter if the mission of a firm is to make widgets or mold young minds: ifa better technique comes along that uses fewer workers, than those workers can do something else. (Belief to the contrary is known as the “lump of labor” fallacy, but in any case these concerns are dispatched by the Basic Income Guarantee and full employment.) An effective exercise of employees’ voice in such a context would make the commonwealth poorer in order to serve the advantage of a few. Employees can certainly exercises voice in such circumstances, but the outcome of that exercise should be in favor of productivity gains. A firm that makes automation decisions does not dominate its employees, because the decision is guided by the principle of productivity, any more than an employee with a better job offer dominates his employer.

The CT bloggers argue that unions are a necessary component of such exercises of voice. Certainly one way to create a protected channel of communication is through a union representative or employee inclusion program. But this is not a necessary consequence of unions, and in fact union members are less likely to report that they have a meaningful role in decision-making than non-union members.

Of course, unions can enable voice: contract negotiations can become a space where labor and capital constructively engage with each other regarding all aspects of the firm. It’s just that they don’t tend to do that; large union bureaucracies don’t guarantee workplace democracy. They can supply other goods, like solidarity, opportunities for civic engagement, and protection from exploitation in our actually existing world of injustice. But in this sense they serve the “reciprocal power” goal by allowing periodic renegotiation and protection from reprisal. Yet a firm that is restricted by full employment considerations in the labor market and that enacts its own procedural approach to personnel matters will be able to prevent many forms of non-productive arbitrary interference with employees without a union.

Meanwhile, unions do enable one obvious form of interference: they enable the majority of workers in a shop to dictate working conditions and extract dues from the minority of workers. Even though closed shops are unconstitutional, agency shops that require non-union members to work under the collectively bargained contract and to pay dues for that privilege, which dues can then be spent on lobbying and advertising that does not conform to the some members of the minority’s “interests according to their own ideas.”

Is this domination? On my view, it is not, because it fails to be arbitrary: so long as the union does in fact work in a principled (non-arbitrary) way to advance the actual interests of the employee. But of course, there is no guarantee that this will take place. There is even risk of the arbitrary interference being reversed, and a particularly powerful union making unreasonable demands on a firm! Arguably, the public unions that have survived our long period of anti-union sentiment do just that, only the “firm” in this case is the republic itself: just remember that the only union that Scott Walker didn’t attack was the Wisconsin Professional Police Association, for whom citizens’ rights are simply “conditions of employment.”

5 thoughts on “Consider the Bathroom Break (Workplace Domination Part Three)”

  1. I’m still not convinced by your definition of “arbitrary.” If I work at WidgetCo, there’s a very good chance that I don’t give a crap about how many widgets we make, or how high-quality they are. And if you say that I am sort of performatively getting on board with a concern for widget productivity, that’s not much different (as far as I can see) from claiming that, perhaps at least so long as it’s known beforehand, if my firm is all about getting the boss laid, if I take the job, hey, it’s not arbitrary any more, or we’re falling back on exit for everything. It seems like there’s got to be some *genuine* good involved (or, failing that, at least a good whose conception I share for other reasons than that I want the job).

    I’m curious, by the way, about the claim you cite re: non-union workers feeling more empowered (for lack of a better term). Was there any attempt to control for the (it seems quite plausible) possibility that unions may be less likely to get traction in jobs where the workers already feel like they have some control?

    1. If you work for Widget Co, and you don't want to make widgets, then you're presumably there for the money and will prefer workplace decisions that increase salaries, I.e. increasing productivity. Even better if your society is richer and can have a more generous BIG.We can't really control for the “only disempowered workers unionize” hypothesis, but it certainly makes sense. It's also possile that having some control just makes you want more. Their data controlling for actual impact on workplace decision-making is inconclusive: i should have used that, because the point is simply that unionized businesses aren't categorically better at incorporating workers in decisions that effect them, and workers feel that. (There are other benefits, just not “voice” benefits.)

      1. I still feel like more has to be fleshed out here to make something like this view work. Productivity might increase without salaries increasing (there's some connection, but lots of confounding factors).

        I guess you *could* say that, OK, I have an interest in something salary-enhancing, and you're right that if I work there and I'm not motivated by widgets themselves I probably see things that way. So maybe productivity, under the right conditions, but really salary. But now I'm not sure what we buy above and beyond the Pettit definition – it seems like productivity is a non-arbitrary standard only insofar as it gets linked up to my preferences as I see them.

        And it also seems like I can make perverse use of your definition. The CEO of WidgetCo only distributes productivity increases equitably when he's happy, and he's happiest when he's sleeping with employees. So my salary increases the more he gets to sleep with employees – do we count as non-dominated if there's a rule to that effect?

        My gut says that your definition is in the right ballpark, so I'm nitpicking a bit.

        1. If productivity increases, then either salaries, dividends, or investments increase. Investments increase future productivity, and feed back into salaries and dividends. Dividends pay taxes, which fund the welfare state, and salaries go to workers. So, it's not direct, but one way or another productivity gains are good for the republic.

          Your perverse use of my definition is basically a recapitulation of the definition of arbitrary: “choosing or not choosing at pleasure.” So no, under your sleep-with-the-CEO rule, we're dominated in the sense of "having our choices arbitrarily interfered with."

          That said, the firm isn't the only institution in a republic, and there's nothing wrong with, say, the CEO's husband (empowered by divorce law to a position of reciprocal power) sleeping with her to make her happy. Civic republicans can still get laid, on my view.

Second Opinions