You said: It is a disturbing fact that police are largely self-selected.
It is a disturbing fact that police are largely self-selected. Society does not go out and find people who would be good cops, but would-be cops choose themselves and apply. Of course some vetting takes place, but it is pretty obvious that many have issues with authority and guns and how they think people should behave and are excessively judgmental about those who don't behave as they think they should.
Claude responded: You're raising a genuine and well-discussed concern in criminal justice reform circles.
You're raising a genuine and well-discussed concern in criminal justice reform circles. There's real substance here, though it's worth thinking through the full picture.
Where you're right:
Self-selection is a real phenomenon in policing. Research does suggest that certain personality types are drawn to law enforcement — those attracted to authority, structure, and the power to enforce rules. Some studies have found elevated levels of authoritarianism among police applicants compared to the general population. The vetting process, while it exists, largely screens for obvious disqualifiers (criminal records, psychological red flags) rather than proactively identifying ideal temperaments.
There's also a well-documented problem with what criminologists call the "warrior mentality" — a us-vs-them framing that some officers bring to the job or develop through training culture, which can lead to excessive force and poor community relations.
Where it's more complicated:
- Many people also choose policing out of genuine community service motivation, family tradition, job stability, or civic duty — the self-selection isn't uniformly negative
- Organizational culture and training arguably shape officer behavior at least as much as initial personality — bad institutions can corrupt good people, and good institutions can channel difficult personalities productively
- The vetting problem is partly a resource problem — thorough psychological screening is expensive and time-consuming, and many departments are understaffed and desperate to fill ranks
- Retention also matters: reform-minded officers often leave frustrated, which can worsen the cultural problem over time
The deeper structural point you're implicitly making — that we should actively recruit for specific temperaments and values rather than passively accept whoever applies — is something reformers genuinely advocate. Countries like Finland and Norway are often cited for their highly selective, professionalized police training models.
It's a legitimate critique of how most Western police forces are built, even if the picture of who becomes a cop is more varied than the worst cases suggest.
You said: It reminds me of a boss I once had who viewed clients as enemies to be defeated rather than partners entitled to mutual profit.
It reminds me of a boss I once had who viewed clients as enemies to be defeated rather than partners entitled to mutual profit.
Claude responded: That's a striking parallel, and it actually maps quite well onto the psychology involved.
That's a striking parallel, and it actually maps quite well onto the psychology involved.
In both cases you have someone in a service role who has fundamentally inverted the relationship. The job exists because of the people being served — clients, or the public — yet the mindset flips them into adversaries. It's a kind of psychological capture by the power dynamics of the role rather than its purpose.
With your boss, that attitude probably showed up in subtle and not-so-subtle ways — negotiations becoming zero-sum contests, client concerns being treated as attacks to be deflected, short-term "wins" that damaged long-term relationships. The person likely didn't even realize how self-defeating it was, because the satisfaction of dominating the interaction replaced the actual goal of mutual success.
Police with that mindset do something structurally identical — the community becomes a population to be controlled rather than people to be protected. Every interaction becomes a contest of authority rather than a problem to be solved together. And similarly, it tends to be self-defeating: it erodes exactly the community trust and cooperation that makes policing actually effective.
There's a concept in organizational psychology sometimes called role corruption — where the tools and authority given to someone to serve a purpose gradually become the point in themselves. The cop stops serving with a gun and badge and starts being someone with a gun and badge. The executive stops serving clients with leverage and starts being someone with leverage.
What's interesting is that in both cases, the people around them — colleagues, subordinates — often normalize it, even adopt it, because the person is in authority. The culture spreads downward more easily than reform spreads upward.
Did your boss's attitude eventually catch up with him, or was he the kind who always seemed to land on his feet?
You said: I can testify that there is a deep effect to putting on a uniform, a badge and a gun.
I can testify that there is a deep effect to putting on a uniform, a badge and a gun.
Claude responded: That's a fascinating firsthand observation.
That's a fascinating firsthand observation. The psychological literature actually backs this up quite strongly.
Philip Zimbardo's Stanford Prison Experiment — however flawed its methodology — pointed at something real: that situational props and role assignments can dramatically alter behavior and self-concept within hours, not years. People become the role the costume signals, to a disturbing degree.
There's also research on what's sometimes called enclothed cognition — the finding that what we wear literally changes how we think and process the world, not just how others perceive us. A lab coat makes people more careful and attentive. It's not hard to extrapolate what a gun, badge, and uniform might do — they signal power, separateness, and the legitimate use of force, all at once.
And the gun is probably the most psychologically potent element. There's research suggesting that merely holding a weapon increases aggressive cognition in the person holding it — not just in those facing it. You are, quite literally, a different person with a loaded weapon on your hip than without one.
The uniform also does something socially powerful — it erases individuality in both directions. The officer feels less like a private person with doubts and more like an institution. And the public responds to the uniform rather than the person inside it, which can reinforce that depersonalization.
Were you in a role yourself where you experienced this, or did you observe it closely in someone else? It sounds like you have some direct sense of what that transformation feels like.
You said: I will pass on your question.
I will pass on your question. If I learned correctly, Athens recruited slaves for law enforcement, kept them separate from the population, and offered a good deal if they performed.
Claude responded: That's a fascinating historical reference, and it speaks directly to the self-selection problem you opened with — Athens essentially eliminated self-selection …
That's a fascinating historical reference, and it speaks directly to the self-selection problem you opened with — Athens essentially eliminated self-selection almost entirely.
The Scythian archers are who you're likely thinking of. Athens used a corps of publicly-owned slaves, mostly Scythians, as a kind of police force — maintaining order, controlling crowds, enforcing magistrates' decisions. At its peak there may have been a thousand or more of them.
The logic, whether consciously articulated or not, was quite elegant in solving several problems at once:
They had no stake in local power struggles. Being foreigners and slaves, they had no family alliances, no political ambitions, no tribal loyalties within Athenian society. They couldn't be the enforcement arm of any faction.
They couldn't self-select for the wrong reasons. Nobody chose this role out of a desire to dominate fellow citizens, because they weren't fellow citizens and had no path to becoming one through the role.
Their incentives were aligned with performance rather than with accumulating personal power or protecting a brotherhood.
It's a striking contrast to modern policing, where officers are deeply embedded in the communities they police, with all the corruption and favoritism that can enable — but also, to be fair, the community knowledge and investment that can be valuable.
The Athenian solution would be considered deeply problematic today for obvious reasons. But the underlying institutional insight — that who you recruit and what their incentives are matters enormously — remains completely relevant.
Do you know where you encountered that piece of history?
You said: I think I've seen it from several sources.
I think I've seen it from several sources. Clearly slaves are out of the question today. A rotating system as in Vietnam drawn from the military veterans avoids excessive local envolvement. As you point out that has the disadvantage of lacking local insight, but does help with corruption avoidance.
Claude responded: The Vietnam model is interesting and underappreciated in Western reform discussions.
The Vietnam model is interesting and underappreciated in Western reform discussions. The rotation principle addresses something fundamental — that long-term embeddedness is a double-edged sword. Local knowledge comes bundled with local entanglements, obligations, and the slow corruption that familiarity breeds.
The military veteran pipeline has some additional features worth thinking through:
Advantages beyond what you mention:
- Veterans have already been screened and observed under stress over extended periods — you know something real about how they perform when it matters
- Military culture, whatever its flaws, does instill a clearer sense of duty to something larger than personal interest
- The hierarchical accountability structure transfers reasonably well
- They've typically learned to operate under rules of engagement — the idea that force has constraints and triggers
The tensions:
- The warrior mentality problem cuts both ways — military training optimizes for a different threat environment than civilian policing
- Combat veterans can carry trauma that doesn't serve community policing well
- America has experimented informally with veteran-heavy police forces and results are mixed
The rotation element may actually be the most powerful piece independently of the veteran question. Even locally recruited officers who rotated regularly between communities would break up the entrenched relationships that enable corruption and the insularity that enables brutality.
It also changes the officer's self-concept — you're less "owner" of a territory and more a professional applying skills across contexts.
The deeper principle seems to be that institutional design matters more than individual character selection — which somewhat complicates your original point, but also deepens it.
You said: I rather doubt that most soldiers, at least in Vietnam, have a "soldier mentality.
I rather doubt that most soldiers, at least in Vietnam, have a "soldier mentality." That would come from a volunteer army as in the US perhaps.
Claude responded: That's a sharp and important distinction.
That's a sharp and important distinction. Vietnam-era conscript armies were cross-sections of society — lawyers, farmers, teachers, mechanics, reluctant draftees who wanted nothing more than to go home. The military experience was something that happened to them rather than something they chose for what it represented.
So the self-selection problem you identified with policing largely disappears with conscription. Nobody self-selected for warrior values — they self-selected to avoid prison, or simply received a letter.
This actually strengthens the Vietnam rotation model considerably. You're drawing from a genuinely representative sample of society, putting them through standardized training and accountability structures, rotating them to avoid entrenchment, and then returning them to civilian life. The uniform goes on and comes off. It never becomes an identity in the dangerous sense.
The volunteer professional military is almost the opposite dynamic — closer to policing in its self-selection problems. People who choose military service, particularly in a society where it's not the norm, are already a skewed sample. The culture that develops in a fully volunteer force can become quite insular and self-referential in ways a conscript army never does.
There's an interesting broader democratic argument lurking here — that institutions wielding force on behalf of society should reflect society, not a self-selected subset of it. Conscription, jury duty, and sortition-based governance all share that underlying logic.
It suggests that how you populate an institution may matter more than how you train it.
You said: There I would differ: training is far more significant.
There I would differ: training is far more significant. Vietnam has a form of organized corruption. Traffic tickets are not issued, one pays a "bribe" that the officer is allowed to pocket according to an established schedule, and if one cannot pay the vehicle is confiscated until one does.
Claude responded: That's a well-taken correction, and a really illuminating example.
That's a well-taken correction, and a really illuminating example.
What you're describing in Vietnam isn't really corruption in the usual sense — it's a formalized alternative system. The schedule, the predictability, the defined consequences — these are actually hallmarks of institutionalization. It functions more like an informal tax with immediate collection than like the arbitrary extortion that characterizes genuine corruption. Everyone knows the rules, which makes it in some ways more transparent than systems that pretend to be clean while operating corruptly beneath the surface.
But your broader point about training is well taken. The Vietnamese system demonstrates that recruitment source alone doesn't determine outcomes — the institutional framework and training around it shapes behavior just as decisively, perhaps more so. Conscript or volunteer, locally embedded or rotating, people behave according to the incentive structures and norms they're trained into.
Which actually pulls together several threads from your original point in an interesting way:
- Self-selection matters because it determines starting material
- Institutional culture and training then shapes that material
- Incentive structures determine what behaviors actually get rewarded
- And rotation or embeddedness determines what corrupting relationships can form
None of these operates independently. The Vietnamese system arguably gets rotation and recruitment reasonably right while building a formalized corruption into the incentive structure through training and policy.
It suggests that reforming policing requires intervening at all these levels simultaneously rather than pulling any single lever.
You said: The US cop has his pension and other benefits to think about and is reasonably paid, so that system works reasonably well, although it involves paper work and …
The US cop has his pension and other benefits to think about and is reasonably paid, so that system works reasonably well, although it involves paper work and appeals and all sorts of added cost.
Claude responded: Yes, and that's a neat illustration of how incentive structure substitutes for or enables different control mechanisms.
Yes, and that's a neat illustration of how incentive structure substitutes for or enables different control mechanisms.
The Vietnamese officer is incentivized through immediate direct reward — the pocket schedule. Simple, low administrative overhead, but it builds the corrupt transaction into the core of the job.
The American officer is incentivized to protect long-term employment and pension — which creates a very different behavioral logic. You don't risk a twenty-year pension for a petty shakedown. The deferred reward system essentially buys future good behavior with present restraint. It's a more sophisticated alignment mechanism.
But as you note it comes with its own overhead — the bureaucratic apparatus required to manage, discipline and adjudicate a professional workforce with employment rights, union protections, and due process claims. The appeals processes and paperwork aren't incidental, they're the necessary counterpart to offering those benefits in the first place. You can't offer job security as an incentive and then remove it arbitrarily.
There's an irony though — the same pension and benefit structure that discourages petty corruption also makes it extremely difficult to remove genuinely bad officers. The protections that align incentives toward good behavior also shield misconduct. Police unions have become perhaps the most powerful institutional obstacle to accountability precisely because they protect the benefit structure that otherwise works reasonably well.
So even this piece of the system has that same double-edged quality — the mechanism that solves one problem tends to create another adjacent one.
You said: I never heard of a Vietnamese officer being removed either, although I doubt I would if it happened.
I never heard of a Vietnamese officer being removed either, although I doubt I would if it happened.
Claude responded: That's a fair epistemic point — absence of evidence and all that.
That's a fair epistemic point — absence of evidence and all that. You'd be unlikely to hear about routine administrative actions in a Vietnamese provincial police station regardless of how frequently they occurred.
But the structural logic does suggest removal is probably rare there too, for different reasons than in America. In a system where the informal payment schedule is officially tolerated or quietly sanctioned, the threshold for what constitutes removable misconduct shifts considerably. The behavior that would end an American officer's career may simply be the Vietnamese system operating as designed.
There's also likely a patronage and hierarchy dimension in Vietnam that substitutes for formal accountability in both directions — protecting officers through personal relationships with superiors while also potentially removing them through those same relationships when they embarrass or displease the wrong person. Less rule-bound, more personal. Which has its own arbitrariness problems.
It points to something interesting — that accountability mechanisms tend to reflect the broader political culture they're embedded in. Vietnam's single party system, its Confucian hierarchical traditions, its relationship between formal and informal rules — these all shape how the police institution actually functions regardless of formal regulations.
America's legalistic, adversarial, rights-based culture produces legalistic adversarial accountability mechanisms with all their overhead and loopholes.
You perhaps can't simply transplant the rotation model without importing the cultural context that makes it function as it does — for better and worse.
You said: I was front seat passenger in car in Cambodia with seatbelt not fastened.
I was front seat passenger in car in Cambodia with seatbelt not fastened. Officer tells driver twenty bucks, no other conversation, then sees I'm foreigner, says, fasten that and get out of here.
Claude responded: That's a wonderful little vignette that captures several things at once.
That's a wonderful little vignette that captures several things at once.
The transaction was completely transparent — no pretense of a fine, no paperwork, no receipt. Just a price, stated flatly. In its way almost refreshingly honest compared to systems that dress the same thing up in bureaucratic clothing.
And then the foreigner calculation is fascinating. His instantaneous reassessment wasn't generosity or respect for your rights — it was pure risk management. Foreigners mean potential complications. An embassy, a journalist, a businessman with connections, a tourist who might complain loudly somewhere that matters. The twenty dollars suddenly wasn't worth whatever uncertainty you represented.
It also shows how discretion is the core of policing in any system. That officer had complete situational authority and exercised it according to his own calculus in both directions — extracting what the traffic would bear from the driver, then making a quick risk-adjusted decision about you. No manual covers that, no training predicted that exact moment. He was improvising within the incentive structure he operated in.
What's striking is that his behavior toward you was probably the functionally correct decision by his own institutional logic, which says something uncomfortable — that the system was working as intended in both halves of that interaction.
Did you have a general sense living there of whether the population viewed that kind of arrangement with resignation, or something more like acceptance as just how things work?
You said: I don't think he was worried about complications.
I don't think he was worried about complications. It's policy here to go easy on foreigners, especially Westerners, for tourist and economic reasons. Not that I can get away with anything, but seat belt?
Claude responded: That's a much more straightforward and probably correct explanation.
That's a much more straightforward and probably correct explanation. Occam's razor — explicit policy rather than individual risk calculation. The officer was simply following known rules in both directions simultaneously, which actually makes him look more professional rather than less.
And the seat belt specifically is a nice detail. It's the kind of infraction that sits at the bottom of any reasonable harm hierarchy — you're risking yourself, in a front seat, as a passenger. Even in jurisdictions that enforce it seriously there's something slightly absurd about the priorities it represents when set against actual public safety concerns.
It also suggests the Cambodian system has more layered sophistication than the street interaction implies. There's the informal payment schedule operating at one level, and above it explicit policy distinctions about how different categories of people are treated, presumably handed down from somewhere rather than left to individual officer discretion. That's actually a recognizable bureaucratic structure, just operating with different surface characteristics than a Western one.
It does raise an interesting question about whether the foreigner-friendly policy is genuinely institutionalized through training and instruction, or whether it's the kind of thing that's simply understood through culture and osmosis without ever being written down anywhere — which is itself a very Cambodian and broadly Southeast Asian way of transmitting institutional knowledge.
The unwritten rule that everyone follows can be more powerful and consistent than the written one that requires enforcement.
You said: Most police officers have similar motivations.
Most police officers have similar motivations. In the States a ticket is less likely if from out of town, both for good-reputation reasons and because of inter-state bureaucracy.
Claude responded: That's a perceptive observation that somewhat complicates the starker critiques of policing — the same underlying pragmatic calculus operating across very diff…
That's a perceptive observation that somewhat complicates the starker critiques of policing — the same underlying pragmatic calculus operating across very different systems.
The American out-of-town leniency works through several overlapping logics simultaneously:
The reputational one you mention — a town that develops a reputation for speed traps and aggressive ticketing of travelers suffers economically and politically in ways that locally-focused enforcement doesn't. There's an implicit competition for being seen as reasonable.
The bureaucratic friction is real too — collecting from an out-of-state driver involves interstate compacts, license reciprocity agreements, collection mechanisms that are genuinely more trouble than they're worth for a minor infraction.
And probably a simpler human sympathy element — the out-of-towner who missed a speed limit change gets more benefit of the doubt than the local who should know better.
What's interesting is that this is functionally not that different from the Cambodian foreigner policy in its underlying logic — category of person modifies application of rule — even though the American version would never be written down or officially acknowledged as policy.
It also gently undermines the original concern about self-selection producing uniformly authoritarian officers. Most officers most of the time are making these same ordinary pragmatic calculations that any reasonable person would make.
The problem cases are perhaps the minority who don't make those ordinary human adjustments — for whom the rule is the rule regardless of context or consequence.
You said: Being a traffic cop is one thing, being on a riot squad is another.
Being a traffic cop is one thing, being on a riot squad is another.
Claude responded: Absolutely, and that's probably the sharpest distinction in all of policing.
Absolutely, and that's probably the sharpest distinction in all of policing.
The traffic cop interaction has a human scale — one officer, one driver, a specific minor infraction, a transaction that ends and everyone goes home. All the ordinary social moderating forces operate normally. Eye contact, conversation, reading the other person, mutual face-saving. The humanity of both parties remains present throughout.
The riot squad context systematically strips all of that away. Helmet and visor removes eye contact. Being part of a formation removes individual identity and judgment. Crowd on the other side removes the individual humanity of those being faced. Commands come from behind rather than from personal assessment. And crucially the crowd's own behavior creates a feedback loop of escalation that no individual on either side may actually want.
It's the Stanford Prison Experiment conditions made real — uniform, role, deindividuation, group dynamics, stress — all simultaneously. The ordinary human moderating forces that make the traffic cop interaction basically functional are almost perfectly disabled.
And self-selection operates differently here too. Officers volunteer for riot duty in most systems, which reintroduces exactly the problem you opened with — who chooses this? The answer is probably not a representative cross-section of the force.
There's also a training issue distinct from general police training — riot control done badly escalates situations that might otherwise have de-escalated naturally. Done well it's almost invisible because nothing happens.
The same person can be a perfectly reasonable traffic cop and something quite different in riot gear.
You said: I told the story of when I was lost on a rural road in Eastern Washington State and did a U-turn in the middle of the road (no traffic).
I told the story of when I was lost on a rural road in Eastern Washington State and did a U-turn in the middle of the road (no traffic). The cop stopped me. I had Missouri plates and was clearly upset and worried. When he asked if I was lost, and I nodded, he said to follow him and led me to the interstate entrance. Now that is helping the public.
Claude responded: That's a perfect example, and a useful antidote to an overly bleak picture.
That's a perfect example, and a useful antidote to an overly bleak picture.
What that officer did required several things working together well. He read the situation accurately — Missouri plates, rural road, visibly distressed driver. The U-turn was obviously not recklessness or disregard for law but simple disorientation. He could have written a ticket and been technically justified. Instead he correctly identified what was actually happening and what the situation actually needed.
Then he went considerably beyond any job requirement. Leading you to the interstate wasn't in any manual. It cost him time and probably paperwork-producing nothing. It was just the right and helpful thing to do.
And notably he did it without making you feel foolish or criminal about being lost, which requires a certain social intelligence and basic decency that no training fully instills — it has to come from somewhere in the person.
It also illustrates your earlier point about local embeddedness having value. A rural Eastern Washington officer probably knows every road in his territory intimately and understands exactly how a stranger could get hopelessly confused there. That local knowledge was directly useful to you.
The interaction had none of the power-assertion dynamics that characterize problematic policing. He had all the authority and used none of it, deploying instead the much more valuable resource of his local knowledge and human judgment.
Those officers exist in large numbers and mostly go unnoticed precisely because nothing newsworthy happens when policing works as it should.
You said: It may have subconsciously helped that we were about the same age and I was driving a Ford Ranger, not some speed mobile.
It may have subconsciously helped that we were about the same age and I was driving a Ford Ranger, not some speed mobile.
Claude responded: Almost certainly, and there's nothing wrong with acknowledging that those social cues operate.
Almost certainly, and there's nothing wrong with acknowledging that those social cues operate.
He was reading the full picture, which is actually what good situational judgment requires. Missouri plates plus rural road plus visible distress plus middle-aged man plus sensible working truck adds up to an extremely legible story — someone's father or uncle who took a wrong turn. No element of that picture suggested anything worth being guarded about.
A young man with local plates in a modified car doing a U-turn on that same road gets a different rapid assessment, fairly or not. Some of that differential is prejudice, but some of it is legitimate pattern recognition that any experienced person develops. The difficult question is distinguishing between the two, which nobody has cleanly solved.
It also speaks to something about mutual recognition — two men of similar age and generation meet on a rural road and each can read the other fairly accurately within seconds. There's a shared cultural vocabulary operating that smooths the interaction before a word is spoken. The officer probably saw someone who reminded him of himself or his neighbors.
Which is humanly understandable but does highlight why policing across cultural or generational distance is so much harder. When that instant mutual legibility isn't available, both parties are working with less information and more anxiety, and the interaction becomes correspondingly more fraught.
The easiest policing interactions are probably those where officer and citizen share the most social context. Which itself says something important about how we staff and deploy police forces.
