There seems to be an upper limit to how much any randomly selected individual would trust another randomly selected individual. Real world observation of society shows that this upper limit varies between different societies and social/environmental conditions.

When the potential pool is narrowed down to the immediate family and closest friends, it becomes normal to claim the capacity for unlimited trust, however this does not seem credible when looking at the revealed preferences of large groups of middle aged individuals.

Therefore it stands to reason that the degree of mutual trust, between individuals, is bounded by some practical upper limit.

Conversely, there appears to be no lower limit, other than zero, for the degree of lack of trust.  In fact this lower limit of zero is practically demonstrated in the real world via prisons that maintain solitary confinement in padded cells for their most troublesome prisoners.

Does this then imply that instead of speaking of a bounded distrust, which appears to not have any meaningful lower bound other than zero, we should speak of bounded trust?


For real world examples, the most useful would be those demonstrated in monetary terms, as other types of expression of trust are difficult to measure or  compare on a YoY or cross-cultural basis.

For example,

There's the widely recognized notion of a 'handshake-deal' where significant transactions, in monetary terms, are settled via simple handshakes and word of mouth, as opposed to contracts. So among certain members of the population, the upper limit of trust can be quite high. The most extreme example would be of one famous individual, Warren Buffett, who regularly boasts of having conducted multi-billion dollar deals on a handshake. 

However there are no cases, that I know of, for deals worth multiple tens-of-billions done on a handshake basis, even Warren Buffet gets lawyers and relies on contracts when his deals get that large. Which implies that the present upper limit of trust in human society, as valued in revealed monetary terms, would be <$10 billion USD.

Prenups are also very common among engaged couples, even of quite modest means, at least in the U.S., which suggests that the practical limit of mutual trust to be in range of a few millions of USD for the majority of the population, at least going by revealed preferences. (As it's unlikely that the average person would trust anyone even more than their soon-to-be spouse, except maybe their parents. Though there is also the similar notion of the 'Bank of Mom-and-Dad', which implies an upper limit to creditworthiness.).


For those situations where trust can't be measured in monetary terms, stated or revealed, the most widely recognized across cultures is likely the bond between soldiers in a platoon on the frontlines of battle. There is quite a high degree of trust for the vast majority of trained soldiers, excluding deserters and such. 

Being willing to risk one's life in pursuit of some highly desirable common objective implies that they would also be willing to risk a smaller amount in pursuit of some less desirable objective in civilian life, though the relationship may not be linear.

However, I do not know of any retired soldiers who feels confident enough in any platoon mate to entrust with billion dollar handshake-deals, which implies that this upper limit is not significantly higher than mentioned above. (Though there might not have been any platoon-mates turned separate large business owners for this to have occurred so far, .)

This phenomena of camaraderie seems broadly similar in every country that has an organized military, and has been the case for most of recorded history, so this type of mutual trust does seem relatively stable across time and across cultures.


Other types, such as the mutual trust between parent and child, in terms of obedience, fidelity, etc., seems to fluctuate according to cultural norms and changing time periods, or at least that's what is commonly alleged. In any case, the upper bound seems to practically exist within any given society. 

For example, no parent, that I know of in the U.S. trusts their teenager with a top-fuel dragster, or stunt plane,  even though there are likely thousands of teenagers who have desire to operate such vehicles and who belong to families capable of affording one. 


In short, there are many plausible reasons to believe in the existence of a practical upper bound to mutual trust between individuals, while real world evidence suggests that the practical lower bound is indeed zero. 

New Comment
12 comments, sorted by Click to highlight new comments since: Today at 3:06 AM

I don't understand the model here.  IMO, trust is multidimensional (prediction robustness to do or not do different things can be different, even for the same trustor and trusteee), and varies over time and with change of circumstance.  It may be useful to talk about the distribution of trust factors or levels, but I don't think this post gives enough basis to start with such.

There seems to be an upper limit to how much any randomly selected individual would trust another randomly selected individual

What makes you say it's an upper limit, rather than a contingent value?  What's the limiting case?  Do you mean randomly selected, or mean or median, or something else?  and are you measuring trust for the distribution (before resolution of the random chance) or for the individuals (after randomly selecting)?

And before you specify any of that, what does this tell us in terms of actions or our own trust behaviors?  Say there's an upper-limit of "would kill or die to prevent a given decision from being made by any given individual".  So what?

IMO, trust is multidimensional (prediction robustness to do or not do different things can be different, even for the same trustor and trusteee), and varies over time and with change of circumstance.  

The expression of trust is reliably measurable along a single dimension, such as monetary units. At least for activities that involve money.

If it is in fact multidimensional this would imply it can be reduced to a single dimension when individuals find it expedient, at least in terms of evaluating potential business partners. Whether or not it actually is multidimensional in practice, or merely claimed to be when convenient, is an issue I will elide.

What makes you say it's an upper limit, rather than a contingent value?

The upper limit isn't set in stone, it varies with time and culture, as I think was pretty clearly stated, so it is also contingent on various factors.

I have a policy of responding to a maximum of two questions per comment, with some exceptions, to keep replies easily readable. If you desire to clarify some additional points you can post it in a separate comment.

It's possible to trust a business partner to deliever on the contract but not trust them to appear on time for meetings. It's also possible to trust them to be on time for meetings but not to honor a contract. 

With business partners it's important to think about the different kinds of expecations you have on your business partner and which of those who can trust them to fulfill. 

I highly doubt there's any question of being on-time for meetings for the principles involved in a billion dollar handshake deal. 

Maybe for small business owners dealing with other small business owners, where the stakes are lower, though even then it seems unlikely unless they're doing it as a favour for family or a friend or something. In that case it wouldn't be reliant on the mutual trust between the business partners but on a trusted intermediary's reputation.

Most business deals are not billion-dollar handshake deals.

For a billion-dollar handshake deal, a person might trust that the deal gets completed but not trust that the other party doesn't leak the sales price. 

When a startup gets acquired the startup founder usually has ideas for how they want the startup to evolve under the new leadership and the trust in the leadership upholding their promises about what they will do after they brought the startup is distinct to the trust in whether they are actually wiring the money. 

When a startup gets acquired the startup founder usually has ideas for how they want the startup to evolve under the new leadership and the trust in the leadership upholding their promises about what they will do after they brought the startup is distinct to the trust in whether they are actually wiring the money. 

This isn't the case in my experience, every successful startup founder I've heard of had, at the very least, ambivalent thoughts about their company's future prospects post-acquisition, in private of course.

Many folks may ostensibly claim that they believe wholeheartedly in the post-acquisition future, but it's likely everyone past the age of 30 or so, startup founder or not, has had some experience with moral mazes. 

So it's unlikely the 'trust' is anything more than nominal for the vast majority of older founders who got acquired.

Maybe someone really young could genuinely believe that all promises will be kept post-acquisition.

Trust in post-acquisition leadership is certainly far from the default. That's what makes it valuable for a funder who actually cares about the post-acquisition future. 

Possibly. How does this relate to the prior comment?

If you know the outcome of how the company will fare post-acquisition you don't need trust. You need trust when it's uncertain.

How do the selling folks acquire knowledge of how their company will fare post-acquisition?

Do you mean ' if they are convinced by the promises of the buying party'?

The whole point of trust, is that you don't have the knowledge. If you would have the knowledge you wouldn't need trust. 

So how do they acquire this knowledge in the first place?