Who Can You Trust?

Every time you lend money to someone, you do it because you trust that they will pay you back. When you leave your belongings on your train seat and ask strangers to look after them while you use the restroom, you trust that they will not steal your items. When you ask a stranger to take a photo of you next to a monument, you trust they will not run away with your camera. Many activities exist only because people trust others:  self-checkout at the stores, donation-based hobbies, membership fees in organizations, or recycling. Trust in strangers is a social glue that helps society function and thrive. 

So why would people trust strangers in the first place? We know that people trust others in societies where the rules of these societies are clear and well-enforced, and where there is no room for corrupt behavior. We call such rules good-quality institutions. In countries with good-quality institutions, like Sweden or Norway, people leave their phones on their train seats while using the restroom, with the trust that their device will not be stolen. This level of trust is unthinkable in countries with institutions of lower quality, like Romania or Moldova where, if you leave your phone on a train seat, it is likely to get stolen. We wanted to know if good-quality institutions cause social trust to develop.

  In principle, it is very difficult to study the effects of institutions (or, in other words, rules) on anything – institutions develop over the course of history, and once established, they often do not change. On the other hand, it is very important to study the influence of institutions on society as they shape interactions between people in everyday life. Institutional quality seems to be associated with many desirable outcomes. For example, in societies with good-quality institutions, we find greater economic development, longer life expectancy, lower infant mortality, and other elements of wellbeing. On the contrary, in societies with lower-quality institutions, we often find lower economic development, lower general health, and lower happiness . 

  To be able to investigate whether the quality of institutions indeed causes people to trust others more, we conducted an experiment where we exposed participants to different levels of institutional quality. We then checked whether those exposed to lower or higher levels of institutional quality had lower or higher trust in others, respectively.

  Since  “quality of institutions” includes many elements, we had to choose one to focus on. In this particular experiment, we focused on the institutions’ capacity to prevent attempts at stealing public funds. To do this, we used a standard economic game called the public good game. The game divided participants into groups of three people, and each person received some amount of money. 

In our game each participant received 20 euros, which is about 20 US dollars. There is a common pool that all participants share, and each participant can contribute any amount to the common pool. At the end of the game, the common pool doubles all contributions and then the total is distributed equally among the group members. All participants knew this information at the start of the game. This means that the most beneficial strategy for the group is to contribute all of their money to the common pool so that it doubles and the participants receive double the amount at the end. However, nobody is required to contribute their money to the common pool; it largely depends on trust. If the participants trust that the other participants will contribute their money, then they will trust that their reward will be big enough at the end and it will be okay to share their money. If the participants do not trust that the other participants will contribute, then they will not want to put the money in the pool and share that with the other participants. This means that most cautious people will contribute nothing unless they trust others to do it.  

On top of this, we modified this game by making one of the group members an administrator of the public funds. Administrators collect contributions, including their own, send it to the pool, the pool doubles it and redistributes the funds equally among the group members, including the administrator. Crucially, the administrator can withhold any amount of money and keep it for themselves, without transferring to the common pool. Doing so is considered stealing from others.

  Here is where we introduce institutional quality. In “perfect” institutional quality settings, the administrator is checked 100% of the time they send money to the pool. If the administrator makes any attempts to steal others’ contributions, a message will tell them to fix their decision in order to proceed. With slightly imperfect institutional quality, the administrator is checked 99% of the time. This means that there is a 1% chance that their attempt at stealing others’ contributions will succeed and they will be allowed to proceed without being stopped. Conversely, 99% of the time the administrator’s attempt to steal will not succeed and they will be stopped. In low institutional quality settings, the administrator is checked only 50% of the time: attempts at stealing others’ contributions will succeed half of the time.


Summary of Institutional Quality Settings:


  We randomly assigned each group to these different levels of institutional quality. From the beginning of the game, the participants knew which institutional quality they were playing in. There were a few things the participants did not know: 1) how much other group members would contribute, 2) whether the administrator was going to steal, and 3) whether the program actually performed a check. With this in mind, their task was to choose how much to contribute to the common pool.

  The participants played this game only once and then were moved onto a second game, in which we measured the participants’ trust in others. We reshuffled the groups and paired participants in groups of two people. In this second game, group members played only with other group members and not with any administrators, and no one played with the same people as in the first game. Each participant knew this.

  In the second game, we gave every participant 6 euros. They could send any amount to the other group member - the receiver. The receiver receives double the amount sent, and can send any amount back to the sender. As one can imagine, keeping everything is very tempting for the receiver, but it is not very nice. The amount the sender sends is a measure of how much they trust that the receiver will send them back at least as much as originally sent.

Participants did not know the results of the first game, therefore, what happened in the first game could not influence their behavior in the second game. The fact that the participants did not play the second game with the same group members as the first game excludes the possibility that people were taking revenge on what they thought might have occurred in the first game. As the participants did not play the second game with administrators, they could not be taking revenge for the administrators’ potential attempts to steal. 

  We found that participants who were exposed to low institutional quality in the first game sent on average 30% less to the receivers in the second game than participants who were exposed to high institutional quality. As the only difference between people who played the trust game was the exposure to different levels of institutional quality, we conclude that exposure to lower institutional quality caused people to send less money. At the same time, exposure to higher institutional quality caused people to send more. We thus conclude that institutional quality caused higher trusting behavior.

  Our results have implications for the world outside the laboratory. Our work has helped to answer a question social scientists have been asking for many years: do good institutions cause trust to develop in societies? While we cannot exclude that more trusting societies do a better job of building their institutions, we can say that institutions play a role in making societies more trusting. In turn, more trusting societies are better equipped to face global challenges, such as natural and environmental disasters, and more resilient to structural and socio-political changes. Moreover, societies where people trust others more do not need to spend as much on monitoring, crime management, and insurance, and this leaves more money for other public spending, such as healthcare, education, roads, public transport, etc.  

The good news is that, while we cannot directly make people more trusting, we can improve institutions. Therefore, we can help people be more trusting by acting on something that can be improved by policy design. It might be complicated to change institutions overnight (Sweden radically changed its institutions over the span of 25 years!), but showing that efforts are being made in the right direction, or communicating information about existing institutions more effectively, might begin this transition. This will be one of the next chapters of our investigation.



Written By: Dr. Marina Povitkina and Dr. Andrea Martinangeli


Academic Editor: Chemist

Non-Academic Editor: Financial Adviser 



Original Paper

• Title: Institutional Quality Causes Generalized Trust: Experimental Evidence on Trusting under the Shadow of Doubt

• Journal: American Journal of Political Science

• Date Published: 30 April 2023



Please remember that research is done by humans and is always changing. A discovery one day could be proven incorrect the next day. It is important to continue to stay informed and keep up with the latest research. We do our best to present current work in an objective and accurate way, but we know that we might make mistakes. If you feel something has been presented incorrectly or inappropriately, please contact us through our website.