I think interviewers rely more on their intuition to evaluate candidates for managerial positions. For purely engineering positions, a longer, more systematic evaluation is needed.

Why would this be a problem?

He deliberately got himself into an awkward situation, for nothing more than the pleasure of drinking a couple of beers. No-brainers don't get much simpler, and for him to get this wrong suggests there's something more going on.

BTW, his mother already knows he's been drinking.

Another BTW: I didn't make that up arbitrarily, just reasonable conjecture from the ways of the world, and of mothers.

I didn't make it clear, but in the scenario she doesn't know.

You can add as many hypotheses as you like (as could I: "what if she asks point-blank?"), but as I said in my reply to shminux, it doesn't help. This scenario does not work as an illustration of the ethical problem. To scale the example up, it's like asking if a murderer should confess, when what he should have done is not do the murder.

The child does indeed have limited rationality, and is in the care of the protagonist: the protagonist is right to exercise that duty of care by limiting the child's access to chocolate.

The mother only has limited rationality by the protagonist's self-serving account. He thinks he can drive safely after a couple of beers; she thinks it too great a risk, did she know of it. His internal monologue --- under the influence of those same two beers --- triumphantly proves her irrationality by the fact that her assessment differs from his. Pah! she has even let herself be irrationally influenced by one of the family dying in a drunken crash! How irrational she is! She has non-transitive preferences, hahaha! Poor old dear, she's not really a PC, not like us, eh? Of course I can drive her safely, are you calling me a drunk? Yes, officer, this is my car, and we've got a plane to catch, so if you don't mind, no I HAVEN'T been drinking--- And so on. That is the general picture I have in my mind of the person you put in that scenario who thinks he's contemplating "the ethicality of denying her agency".

Or dressed up in jargon, it's my posterior on seeing the evidence of the story, given my prior knowledge of the ways of the world.

ETA: A real answer to what the of course not at all drunk driver could do would be to handle the immediate situation by paying a taxi driver whatever it takes for a two-hour journey. He might then profitably spend those two hours examining the underlying problem: why he chose to have those beers.

BTW, his mother already knows he's been drinking.

I wonder if you can steelman the second setup to make it smell-proof, and then answer the OP's question?

No. For any action X it is easy to dream up a hypothetical situation Y in which X is the right thing to do. In the limit, it reduces to letting Y be "Suppose X was the right thing to do?" This is not a useful exercise. The pattern is an anti-pattern.

Besides which, someone has already commented that the concept of agency used in the OP is unclear. The 5-year-old and the mother still have agency; they are being prevented from exercising it, the one by superior force, the other by concealment of knowledge.

Substituting the word "control" does not change things. The 5-year-old and the mother are still trying to exercise control, that is, they are both trying to achieve purposes; they are being prevented from achieving those purposes, the one by superior force, the other by concealment of knowledge. The protagonist is doing this because his purposes conflict with theirs.

So the question is, when your goals conflict with another's, when is it right to use force or subterfuge to get your way? Suddenly it sounds a lot more commonplace a matter than the distant phrase, "the ethicality of denying agency", and needs no hypothetical steelmanned scenarios. A glance at the real world provides limitless raw material, which can come from as close at hand as one's own everyday life. The question is about the entire subject of how people can live together, the totality of ethics.

I think I'm actually more comfortable with the scenario where you are the primary/sole beneficiary.

Denying someone's agency to benefit them is really treating them like a child, and is only appropriate in a case where they really don't have the capacity to exercise it (besides children, e.g. adults with significant dementia or cognitive impairment.)

By contrast, if someone's irrational behavior is going to negatively affect you, I see more room for mitigating it by denying them information (i.e. lying to them, whether by commission or omission.) In this case I don't see it as being quite the same thing as denying agency, somehow -- you're treating them as an agent, but an adversarial one. Whereas in the other case you're trying to protect them from themselves.

So, you (Swimmer963) think of agenty people as being those who:

1. Are reliable
2. Are skilled (in areas you are less familiar with)
3. Act deliberately, especially for their own interest

It is interesting that all three of these behaviors seem to be high status behaviors. So, my question is this: does high status make someone seem more agenty to you? Could sufficiently high status be a sufficient condition for someone being "agenty"?

The degree to which I feel blame or judgement towards people for not doing things they said they would do is almost directly proportional to how much I model them as agents.

I've noticed that people are angrier at behaviors they can't explain. The anger subdues when they learn about the motives and circumstances that led to the behavior. If non-agents are suppose to be less predictable, I'd guess we're more inclined to judge/blame them.

One such study is the famous Wason selection task, and there, evolutionary psychology gives a fundamentally very different sort of answer than what you've given: that we have evolved, innate cognitive modules that solve certain types of problems... but are not used at all when the same abstract form of problem is put in a different context:

Cosmides and Tooby argued that experimenters have ruled out alternative explanations, such as that people learn the rules of social exchange through practice and find it easier to apply these familiar rules than less-familiar rules. [...] They argued that such a distinction [between performance on the problem in a social context and the same problem otherwise], if empirically borne out, would support the contention of evolutionary psychologists that human reasoning is governed by context-sensitive mechanisms that have evolved, through natural selection, to solve specific problems of social interaction, rather than context-free, general-purpose mechanisms.

The explanation on wikipedia is well worth a read.

Hmm, are you sure that they're synonymous? I initially assumed that your post was talking more about holistic vs. analytical reasoning (see e.g. pages 23-27 of The Weirdest People in the World), which seems to have some similarities with System 1/System 2 reasoning, but also differences which don't map so clearly to it:

Holistic thought involves an orientation to the context or field as a whole, including attention to relationships between a focal object and the field, and a preference for explaining and predicting events on the basis of such relationships. Analytic thought involves a detachment of objects from contexts, a tendency to focus on objects’ attributes, and a preference for using categorical rules o explain and predict behavior. This distinction between habits of thought rests on a theoretical partition between two reasoning systems. One system is associative, and its computations reflect similarity and contiguity (i.e., whether two stimuli share perceptual resemblances and co ‐ occur in time); the other system relies on abstract, symbolic representational systems, and its computations reflect a rule ‐ based structure (e.g., Neisser 1963, Sloman 1996).

Although both cognitive systems are available in all normal adults, different environments, experiences, and cultural routines may encourage reliance on one system at the expense of another, giving rise to population ‐ level differences in the use of these different cognitive strategies to solve identical problems. There is growing evidence that a key factor influencing the prominence of analytic vs. holistic cognition is the different self ‐ construals prevalent across populations. First, independent self ‐ construal primes facilitate analytic processing, whereas interdependent primes facilitate holistic processing (Oyserman & Lee 2008). Second, geographic regions with greater prevalence of interdependent self ‐ construals show more holistic processing, as can be seen in comparisons of Northern and Southern Italians, Hokkaido and mainland Japanese, and Western and Eastern Europeans (Varnum et al. 2008).

Furthermore, the analytic approach is culturally more valued in Western contexts, whereas the holistic approach is more valued in East Asian contexts, leading to normative judgments about cognitive strategies that differ across the respective populations (e.g., Buchtel & Norenzayan, in press). Below we highlight some findings from this research showing that, compared to diverse populations of non ‐ westerners, Westerners (1) attend more to objects than fields; (2) explain behavior in more decontextualized terms; (3) rely more on rules over similarity relations to classify objects; (for further discussion of the cross ‐ cultural evidence Nisbett 2003, Norenzayan et al. 2007).

(E.g. this difference wouldn't seem to be something that you'd expect to arise from just System 1/System 2 processing:)

Several classic studies, initially conducted with Western participants, found that people tend to make strong attributions about a person’s disposition, even when there are compelling situational constraints (Jones & Harris 1967, Ross et al. 1977). This tendency to ignore situational information in favor of personality information is so commonly observed—among typical subjects—that it was dubbed the “fundamental attribution error” (Ross et al. 1977). However, consistent with much ethnography in non ‐ Western cultures (e.g., Geertz 1975), comparative experimental work demonstrates differences that, while Americans attend to dispositions at the expense of situations (Gilbert & Malone 1995), East Asians are more likely than Americans to infer that behaviors are strongly controlled by the situation (Miyamoto & Kitayama 2002, Morris & Peng 1994, Norenzayan, Choi et al. 2002, Van Boven et al. 1999), particularly when situational information is made salient (Choi & Nisbett 1998). 6 Grossmann et. al. provide parallel findings with Russians (Grossmann 2008). Likewise, in an investigation of people’s lay beliefs about personality across eight populations, Church et al. (2006) found that people from Western populations (i.e., American and Euro ‐ Australian) strongly endorsed the notions that traits remain stable over time and predict behavior over many situations, while those from non ‐ Western populations (i.e., Asian ‐ Australian, Chinese ‐ Malaysian, Filipino, Japanese, Mexican, and Malay) more strongly endorsed contextual beliefs about personality, such as ideas that traits do not describe a person as well as roles or duties, and that trait ‐ related behavior changes from situation to situation. These patterns are consistent with earlier work on attributions comparing Euro ‐ Americans with Hindu Indians (see Shweder and Bourne (1982) and Miller (1984)). Thus, although dispositional inferences can be found outside the West, the fundamental attribution error seems less fundamental elsewhere (Choi et al. 1999).