[JabavuAdams]

Interesting point, but not really relevant to my argument.

What's most interesting about these IDF anecdotes is that you have one ex-soldier completely dismissing behaviour that another ex-soldier admits to. That pattern can generalize to any org, regardless of volunteer / non-volunteer selection. I.e. they can't both be right. One of them is over-generalizing.

EDIT>

I.e. saying "No one in org X is doing Y" is much less likely (given an individual's limited personal experience) than saying "someone in org X is probably doing Y".

[endtime]

The OP said "Org X has property P." You responded with "Org Y does not have property P, and Org Y is like Org X." I claim the second half of that - that things that are true about the IDF are likely to be true about the US military - cannot be assumed. Even ignoring the significant cultural differences between the two countries, there are differences in the demographics the militaries themselves are composed of.

[JabavuAdams]

So at the risk of pedantry ...

I was making a claim about large organizations that depend on loyalty, not even specifically military organizations.

So, the specific difference between military organizations and cultures are neatly excluded.

The pattern (of which I gave a specific example) is:

Person X in large org Y claims something is true for ALL members of org Y, based on X's personal experience.

I submit that humans being what they are and organizations being what they are, it's more likely that there are bad behaviours that are tolerated in sub-organizations of Y even if they violate the officially stated rules.

TL;DR Without perfect knowledge it's better to assume that something somewhere is going wrong than to assume that nothing anywhere is going wrong.

A lot of science, engineering and process control basically boils down to this. Why do we forget it when dealing with human systems?

[endtime]

Behavior/discipline will necessarily be noisier in a conscripted organization than in a self-selected volunteer organization. I don't think you can generalize that away.