[chii]

But the question is whether this sort of hazing produce better outcomes (for the company)?

You point out elite military units, and medicine. Both of these fields tend to have highly competent, highly motivated individuals, and thus their work is quite valuable. If it were better to not have such hazing, then that implies that these units are not operating optimally. But then if they aren't, why hasn't a better optimal system replaced it?

[evgen]

I guess I would assume that it produces better outcomes from the company perspective but honestly can't think of a good counter-example to use as a comparison. After enduring the hazing it is much easier to convince you to do things that might be a bit morally questionable (e.g. shifting your personal ethical Overton window) by reminding you in subtle ways of what you endured to reach this point, what it means to others that have endured with you, etc.

I would assume that in certain adversarial environments it is better to have a small team that will focus on a goal without question than one in which either the goal or the means to accomplish it are ever up for questioning, but someone with any knowledge of academic research on leadership and team dynamics can probably answer better than I.

[goatinaboat]

it is much easier to convince you to do things that might be a bit morally questionable (e.g. shifting your personal ethical Overton window

Same reason that executives like to go to strip clubs together, it creates a sense of “if I go down you go down with me”.

[newsclues]

It creates loyalty and trust. Important when doing dangerous or illegal things.

[TeMPOraL]

I recall this being a topic of interest in social sciences in general. My general impression[0] is that hazing rituals are very important in creating and maintaining group cohesion. The rituals evolve towards being tough enough in terms of pain and self-esteem: enough to filter out people with weak desire to join the group[1], and enough to do some lasting - but not debilitating - damage[2], but not enough to be seen as abusive by the group members[3].

The driver of evolution of these rituals is just group survival: very bad ones will self-destruct a group, good ones let the group be more effective and outcompete less effective rivals[4].

Here are some predictions from this view, which I think all pan out:

- Groups that survived for a long time have such rituals. They evolve towards less tough/abusive if the group isn't in fierce competition. They may ultimately become painless, effortless, make-believe rubber-stamping, but at that point the group is dying.

- When new groups form around some domain, some may go overboard with their rituals - it takes time for groups to self-destruct or be outcompeted.

- Groups overlap, and the demands of supergroup may temper the hazing rituals of subgroups. The most obvious example: the laws of your nation will put a ceiling on what kind of hazing can happen in groups that exist under its jurisdiction.

--

[0] - Not a social scientist, have no sources to cite, going from memory of all the stuff I read in books and blogs over the last decade of procrastination.

[1] - In cases where joining is voluntary - e.g. a fraternity. Contrast that with tribes, where everyone born to in it is expected to go through a coming-of-age ritual.

[2] - The group wants its members to contribute fully to its goals, so rituals will not threaten that - unless there's a steady stream of candidates that needs to be filtered anyway, in which case surviving the ritual unscathed becomes a filtering criterion. On the other hand, scars - be it physical or emotional - are encouraged, if they're not compromising performance. Such scars serve as a reminder of the commitment, in-group identifier, and (particularly with scars gained post-hazing) can confer social status in the group.

[3] - The goal is to boost group cohesion, so having new members harbor resentment towards the group is counterproductive.

[4] - Note that when talking about market players, the "group" here isn't equivalent to a company. A company is a different intersubjective entity, largely independent from a group of people. In a company, people are fungible. A group can form in a company, or across companies, and can easily move from one company to another, retaining its members and rituals. Groups cycle its members too (e.g. people die, or retire from service, or quit the industry), but it's a different lifecycle.

[heavenlyblue]

> But then if they aren't, why hasn't a better optimal system replaced it?

Because we’re not living at the limit of infinity and this discussion is there to change this fact