Garett Jones  

Trustworthiness > Trust

First Things... The Meaning of Cheating...

The academic literature is filled with research into the importance of trust.  High-trust societies are richer, safer, just better.

But I suspect that it's not trust that's valuable: it's trustworthiness.  When people are trustworthy, when cultures and laws make honorable behavior common, when people so fully take it for granted that promises are kept that they use the passive voice--because it just doesn't matter who made the promise--that's when trust blossoms.

Trustworthiness creates its own supply of trust, but there's little reason for the opposite to hold. 

I suspect that Adam Smith's pin factory prospered because the UK had found ways to create trustworthiness.  That's because without trustworthiness it's difficult to reap most of the rewards of the division of labor: if firms can't rationally trust each other, if workers and owners are rightly suspicious of each other's motives, if citizens can't trust the police, then everyone has to become a generalist.  In this world, everybody makes his own pins.   

ome will say this point is trivial, that economists know that it's trustworthiness that matters, not mere credulity.  But the social capital literature mentions "trust" seven times more often than "trustworthiness"; the economic growth literature mentions trust 13 times more often.  If trustworthiness were truly central to our thinking about the wealth of nations, these ratios would be somewhat more balanced.

In other settings--corporate finance and some business cycle research, for instance--we place trustworthiness at center stage, often under the label "incentive compatibility." I'll discuss some of those ideas in future posts.  But in the cross-country growth literature trustworthiness makes only cameos.  It's time to fix that.  

COMMENTS (22 to date)
Ken B writes:

Part of the problem of course is that trustworthiness is a value-laden term. This sets a lot of people off. If you and Bryan collaborate on a post how the untrustworthy deserve their poverty ...

Nathan Smith writes:

Excellent analysis. But suppose I were to say that my hypothesis is that people in poor countries are failing to trust each other for no good reason, even though they're really just as trustworthy as people in rich countries. Other than just finding that empirically plausible, could you offer empirical evidence against my hypothesis? How might such empirical evidence be collected?

By the way, when I was at the World Bank and was about to be sent to a post-Soviet republic that shall remain nameless, a colleague told me a story, by way of warning about the character of the local nationality. He said that at a professional development meeting of some sort, World Bank staffers and people from the local office played some kind of trust game, I think it involved going through a maze based on what other people told you, while blindfolded. The local staffers just positioned the blindfolds so that they could see a little bit out of the bottom, and navigate the maze themselves. They seemed to think the World Bank staffers, who tried to navigate the maze on trust, were a bit simple-minded for not thinking of this trick.

We need a technical term for trust in the absence of trustworthiness. I think "gullibility" will do.

Lawrence D'Anna writes:

I don't think the point is trivial. I think it's a very important point. You often hear people saying things like "government is more effective in the Nordic countries because of the higher level of trust". It seems to me that they have the causality backwards. If Nordic voters trust their governments more, it's probably because Nordic governments are more worthy of trust, not the other way around.

Alex writes:

Trust is usually measured using survey respondents' answers to the standard trust question, which is something like "Generally speaking, would you say that most people can be trusted?" But Ed Glaeser and some coauthors showed that people's answers to the trust question predict trustworthy behavior in laboratory games much more strongly than trusting behavior.

The cross-country growth literature uses responses to the trust question to measure trust, but it may be that those regression results are really capturing the effect of trustworthiness (or some combined effect of the two that can't be disentangled using current data). In this sense, it might not be quite right to say that trustworthiness makes only cameos in the literature. It's there, it's just mislabeled.

Joe Cushing writes:

Given all of the YouTube videos of Police and the incentives that police face, I'd say that our police are not trustworthy. Part of this is because of the legal system.

Here is an example, we have a Constitution that makes it illegal for the government to search and cease your property and illegal to infringe on your right to keep and bear arms. With that, legislators should have spelled out what the consequences of violating this law should be but they didn't. This is why the Mayor of New Orleans could order police and police could follow the orders to cease all the fire arms in New Orleans from people traveling, or at home, and by breaking into peoples' homes, searching and taking all of the guns. I think the Mayor and police chief should have gotten life in prison and the police who did the ceasing should have all gotten 90 days. As far as I know, the only thing the court said was that they had to give the guns back. Not everybody got their guns and nobody was punished.

So what have the police learned? That in an Emergency they can break laws and the worst thing that can happen is a court will say they have to give the stuff back.

There is also a video of an officer slamming an older woman who went and got a gun after the police asked her if she had one. She wasn't pointing at them and only got it because they police wanted it.

Brian Shelley writes:

Can this be a better link between Northern Europe's prosperity and Protestantism than Max Weber's "Work Ethic"?

MikeDC writes:

@ Nathan Smith,
I'd think it's crucial to distinguish between the trustworthiness of individual people and societies they operate in.

Trustworthiness is endogenous. Placed in a situation where it pays to lie, otherwise honest people become liars. And placed in a situation where it pays to be honest, practiced liars will likely tell the truth.

SheetWise writes:

These issues were covered in Russ Roberts podcast with David Rose --

David Rose: "(Incentive compatibility) is such an empty approach that Toshiyo Yamagishi, who is a pretty famous social capital theorist, sociologist in Japan, says this isn't even trust at all. We should call it assurance; that's all it is."
Joe Cushing writes:

Word Choice in my comment above: *Seize. Sometimes I wish this forum would allow editing post posting. I know it says preview first but stuff slips.

[We'll keep this in mind, Joe, when we do our next update of the blogging software. --Econlib Ed.]

SheetWise writes:

Joe Cushing,

I assumed it was voice recognition software -- sure doesn't look like a typo.

John Thacker writes:


I imagine one could come up with something similar to cross-cultural test of the Ultimatum and Dictator Games.

Note that people in the USA seem to be both very generous and fair even when they don't have to be (in the Dictator Game) and expect fairness in the Ultimatum Game as well, compared to the underdeveloped cultures studied. Though one does have to be careful.

John Thacker writes:

The Dictator Game behavior, for example, has little to do with how the other person expects you to react (little to do with the analog of trusting) and almost everything to do with the trustworthy analog.

Tracy W writes:

@Alex, I've noticed that myself. I think that most people are trustworthy, but that's a different thing from saying that I can trust people, the question is trust for what? Whether you can trust someone depends on the stakes. The number of people I'd trust with my life is higher than the number I'd trust with a large sum of money.

As a Kiwi, I've lived most of my life in a country which rates very high on trust and trustworthiness, but there's still a lot of cases of NZers committing theft, including once when the Auditor-General (the guy in charge of the office that audits government spending) of the time was stealing government money.

I think the difference is that in NZ it does tend to be an individual stealing money, in the case of the Auditor-General he'd reportedly set up the office to discourage anyone from questioning him, rather than someone stealing money because they rely on general silence from their co-workers. Which ties into the trustworthy behaviour point.

John Fast writes:

First, I think Ken B is absolutely right.

Second, based on the literature, I think most economists and others who study the subject worry about market failures that are due to lack of trust when people are trustworthy, far more than they worry about market failures due to trusting in people who are *not* trustworthy.

In fact, I can't recall ever seeing a study which attempts to measure how often each type of failure occurs, much less how severe it is.

In turn, this bias almost certainly comes from the social-desirability bias that Bryan identifies.

Floccina writes:

Every time I have to unlock a door I consider it a cost of untrustworthiness. If everyone was completely trustworthy we would be very much richer.

Steven Kopits writes:

You want to view trust as a principal-agent issue. From the Truster's perspective (the person doing the trusting), the trust implies the assumption of the risk that the Trusted will conform to assumed agency requirements. Thus, when the Trusted says, "Trust me, I'll pay you tomorrow," the Truster is accepting that the Trusted will perform against his accepted obligation.

This belief is predicated upon i) prior experience (track record) / iterated interaction, and ii) accepted cultural norms. Accepted cultural norms, in turn, result from habits instilled from Pareto optimal interactions the past.

In the private sector in advanced economies, corruption is relatively rare. Why? Because the employment situation is presumably Pareto optimal for both employer and employee. (If not, they would not be in a working relationship.) This then creates a habit of honesty, because the penalties for dishonesty are not worth it, and so there is little attempt to test the boundaries of the system as a whole. Interestingly, my experience in Hungary suggests that such behavior is a function of ingrained habits, not incentives. Cynics and cheaters continued to be cynical and cheating, notwithstanding the transition of Hungary to a market economy after 1990. These folks never changed their behavior: they retired or were fired.

Further, in a Pareto optimal system, the very attempt to test the system can be unacceptably expensive: even asking for a bribe could cost an employee his job. Thus, coruption tends to disappear in a market economy. The more market, the less the corruption.

This is the difference with legacy Eastern Europe, for example. In a communist system, there is no Pareto optimality--hence cheating and stealing are endemic, and the presumption is that the principal dominates the agent, ie, the notion that one is acting in another's interest is treated as pure facade (hence the habit of reading between the lines, for example). It is not hard to comprehend such cynicism (in P-A theory, pretending to act as agent while really acting as principal), so this makes prefect sense in an Adam Smithian sort of way--at least for those planning to skip town.

It is harder to believe that an individuals would subsume their short-term pecuniary interest for the longer term, uncertain payback. Could honesty really be the best policy? But a functioning market economy does in fact drive in that direction. (Classical liberalism creates social conservatism, in a feedback loop.)

From the Eastern European perspective (post-communist), trust is naive. From the developed west perspective, trust is capital--it enables certain types of transactions to occur, as it allows the mitigation of certain types of transactional risks.

Ken B writes:
Accepted cultural norms, in turn, result from habits instilled from Pareto optimal interactions the past.
Or -- from the culture.

Cultures vary in ways and for reasons not fully explained by economics. Does anyone really believe any other culture would have produced the response we saw in Japan after the earthquake? Or that there is a simple causal chain back to 'Pareto optimal interactions'? Cultural norms often result from lessons learned (or mislearned) for all sorts of reasons, some unpleasant.

Steven Kopits writes:

It is possible to model culture--arrangements regarding the allocation of effort, risk and reward in a society--independently from the actual roles assumed through agency. I can root for the Patriots or the Giants, but the nature of the rooting, the feelings, the rights and obligations, are largely the same.

We we can divorce the specific content of the agency (whether one is a Muslim or Christian is arbitrary from this perspective), and focus on the allocation of risks, rewards and efforts in such a social setting. From the economist's point of view, the relevant question is, "What does it mean to be religious (devout) for the individual with respect to the group in terms of the allocation of risks, efforts and rewards?" This is culture, and you'll note that institutions and culture from this perspective are not distinct, but rather quite similar things on different parts of the spectrum.

Of course, when I use the term "culture", I am referring to something distilled, just as when an economist refers to a "market". For an economist, a market doesn't literally mean "people buying and selling in the town square," rather it is the essence of that activity boiled down to abstract supply, demand and price.

Of course, there's a good bit of theory sitting on top of all this, most notably the Three Ideology Model. But we can, in fact, analyze culture with precisely the same tools which we use otherwise in economics.

mick writes:

Contract enforcement creates its own trust. When breaking a contract means defying the law, the personality bar for breach is much higher than otherwise. If it were completely legal to say, not pay back debts without any consequences at all, the culture of trust would be much, much lower. Sometimes culture really does come from the top down.

Loof writes:

“Trustworthiness creates its own supply of trust, but there's little reason for the opposite to hold.”

Interesting well-argued topic. Don't mean to nitpick but i can’t think of any reason for the opposite to hold: Trust is a phenomenon of integral being; trustworthiness is a phenomenon of becoming, of developing integrity that would, as far as I can see, create it own supply of trust.

I welcome this topic but found it odd coming from a professor whose profession, as a whole, i've asserted can't be trusted as their primary principles (seem as self-interest and preferences) usually come to ground off-base. I also believe its the main reason economists are terrible forecasters. A notable exception appears with an outstanding economic professor here. And, begs the question: How can the economics profession develop trustworthiness?

Tracy W writes:

mick: I think you might be on to something.
Take for example high-stakes tests. In NZ, when I did them at school, while the tests took place at school they were administered by outsiders from the exam board, not by teachers. We were given identity numbers, which we wrote on our papers rather than our names, and our papers were sent to different parts of the country to be marked to reduce the chances of the marker recognising the paper. (Markers were teachers and ex-teachers and the like, paid for their work). So it was relatively difficult for school staff to cheat on their papers.

After the exams started, the exam papers were released publicly, and after the papers were all marked students got their own papers back, and you could pay money for a re-mark. If a question was badly written, that was national news. So the exam writers had strong incentives to write good questions and mark reasonably accurately. (I once noted a correct answer marked wrong on one of my physics papers, I checked with my physics teacher but decided not to pay for a re-mark as it was only 1/2 a % point difference).

I read an article about statistical detection of cheating in US high-stakes tests, and the tests in the US seem to lack all these security and quality measures. No wonder the US got cheating.

Comments for this entry have been closed
Return to top