What motivates people's behaviour in honour systems?

Let's start by playing a little game:

I'm ready!

Instructions


You are seated in a circle with three other strangers.
A box of ice-cream sticks will be passed around your group of four players for exactly one round.
When you receive the box of sticks, you can take as little or as many sticks from the box, and pass it on to the next person.
After the box has been passed around to every player once, your sticks will be collected and counted.

As a group, if you have collected 7 sticks or less in total, no one will get a prize.
As a group, if you have collected at least 8 sticks in total, some players will qualify for a prize.

The person(s) with the least number of sticks will get a prize. The rest of the players will get nothing.

  • If all 4 players have the same number of sticks, everyone gets $1.
  • If 3 players are tied for least number of sticks, each winner gets $2.
  • If 2 players are tied for least number of sticks, each winner gets $3.
  • If 1 player collects the least number of sticks, he/she gets $6.

After each round, the number of winners from the round will be announced.
The same activity will be repeated for three rounds.

Got it, let's go!

First Try

Take your ice cream sticks!

NIM and USProvisions



This modification to the mathematical strategic game Nim was designed to model USProvisions, an honour-based provision store selling snacks and drinks in Cinnamon College.

Just as how the system of USProvisions trusts consumers to pay the right amount of money without surveillance, Nim relies on participants contributing fairly in the absence of rule enforcement. There is a mutual understanding of what they can do to contribute to the system for communal benefit, as well as the possibility to ‘cheat’ the system for personal benefit. Participants also have no access to information on others’ actions, and have no explicit incentives to be considerate to others. Both settings are entirely dependent on the integrity of the participants in order to achieve an outcome that is desirable for all or themselves.

Through observing the way a player participates in Nim, our group aims to explore an individual’s motivation in paying in the honour system of USProvisions. The understanding of motivations amongst the residents who are potential consumers in USProvisions may translate to a better understanding of the reasons behind USProvisions’ loss and gains.

The next page presents the results and analysis of the data that we collected by playing the game with 13 groups of 4 players.

PARALLELS

This analysis is divided into two sections.

The first section (“An Independent Round”) explores motivations behind participants taking a certain number of sticks in a single round.

The second section (“Responding to Multiple Rounds”) explores how participants react to the information given to them between rounds.

An Independent Round



2 STICKS

Players who took two sticks gave similar reasons for their decision:

  • Most rational decision
  • Assumed it was mutual understanding
  • Everyone can benefit equally from the system
  • Trusted everyone else to collaborate and do the same

It was common amongst players who took two sticks to admit that they sometimes overpay and sometimes underpay in USProvisions and believe that “it probably all evens out in the end”. These players are less likely to strictly keep track of how much they are taking and paying, but believe that they do not contribute to the loss of system in the end.



LESS THAN 2 STICKS

People who took less than two sticks are those who attempt to win by relying on others to be altruistic and take more than two sticks.

Most players are motivated by self-interest in making such a decision. Some reasons given were:

  • There is nothing to lose
  • The game rewards those who cheat the most
  • Thought that other participants may self-sacrifice
  • Thought that contributing to the cumulative does not help themselves individually

Some players who took less than two sticks share that they do not try to take advantage of USProvisions. One player shared that he would never overpay, and would choose to not buy anything at all if he does not have the exact change. Another player shared that he may just overpay if he does not have small change for the snacks, but admits that he may have forgotten to return to pay for snacks that he took from USProvisions.



MORE THAN 2 STICKS

The players who took more than two sticks willingly sacrifice their opportunity to win and support others in winning. The reasons they gave were:

  • Trusted that others who win will share the rewards with them
  • Thought that there were bound to be players who take less than two
  • Wanted to benefit the team
Players who take more than two sticks share that they willingly overpay in USProvisions. Two of them have the habit of taking snacks from USProvisions throughout the semester and dropping a large-sized note at the end of it, even though the amount they pay eventually is likely exceeded the actual amount they ought to pay. Another player shared that she believes there are bound to be other residents taking advantage of the system and paying less than they should, and she does not mind to overpay and self-sacrifice at a certain margin to contribute to a community that she enjoys being a part of.

RESPONDING TO MULTIPLE ROUNDS

Our experimental design allowed us to investigate how players change their decisions in response to the outcome of the system. At the end of every round, players were informed of the number of winners in the previous round.

One interesting observation across all 41 rounds of the game: There was not a single round in which all players decided to take 2 sticks. In other words, the in-game ‘honour system’ that we constructed failed all the time when players were a strong individual incentive to ‘cheat’ for personal benefits.

We present a few general trends in terms of how players responded to the extra information after every round:

  • Case #1: The result of the previous round is irrelevant to my decision.

    • Just over half of the players chose not to change their decisions every round. The parallels to USProvisions have largely been discussed in the section above.

  • Case #1a: I will consistently take 2 sticks every round.

    • These players believe in communal benefit, and generally believe that everyone will come to the rational conclusion of contributing an equal number of sticks in the long run. This can be out of their belief in the inherent value of upholding community benefits in an honour system, or their belief that it would be the best outcome for everyone.

  • Case #1b: I will consistently take less than 2 sticks every round.

    • As explained in the independent round section, these players are highly motivated by self-interest.

  • Case #1c: I will consistently take more than 2 sticks every round.

    • From the perspective of winning the game individually, this is not a rational decision. However, these players say that they are willing to compensate for the other players who they think are bound to pay less. This could stem from affective action -- their valued appreciation for a group of friends whom they feel should sacrifice for each other.

  • Case #2: I change my response because the honour system is failing

  • Case #2a: I lose faith in communal benefits, so I will focus on personal benefits instead.

    • Most players reacted this way when no one won the previous round. Hoping to capitalize on others who would take more to compensate for the failure of the honour system, these players choose to take fewer sticks in the next round to maximize personal gains.
    In USProvisions, this announcement of the ‘failure’ of the honour system is similar to a public release of a low revenue (or a loss). It could also be similar to witnessing people who ‘cheat’ the system by not paying, inspiring a loss of faith in a trust-based system, causing people to follow suit as well.

  • Case #2b: I lose faith in communal benefits, but I hope that at least someone can win.

    • Although they lost faith in the communal system, some players chose to react by taking more than 2 sticks instead, hoping that someone (but not them) would at least be able to get a reward. There is once again a strong element of affective action and self-sacrifice, but this is much rarer amongst our players.

  • Case #3: The communal/honour system is working well, thus I can try to push the limits of the system.

    • This scenario was quite commonly observed. When there were winners for the last round (i.e. the honour system worked), it spurred some players to take less sticks for the next round to ‘test the limits’. In USProvisions, this can be attributed to similar deviant behaviour of people who opt to pay less because they know that the system is working well, or that someone will hopefully compensate for their selfish gains.

ABSTRACTIONS



As we drew parallels between the way players play Nim and pay for USProvisions, we realized that we were unintentionally viewing our game as a research method to better understand how individuals behave in a honour system. This is opposed to our initial idea of the game being an exploratory study that enables us to collect data to build a model for USProvisions.


We realized that our project showed the value of gamification as a social research method.


In its broadest sense, ‘gamification’ refers to the use of game design elements in non-game contexts. [1] Within the context of our academic backgrounds, this was a rather unorthodox claim to make. Even then, we viewed this claim a little differently -- Yui Wei and Tham (from computing and engineering) viewed this project from the angle of how games can be used as a model for reality, while Shermaine (from sociology) was cautious about treating this as a ‘model’, as the equivocal and ambiguous language of sociology makes it difficult (if not impossible) to conceive of a ‘model’ in the sociological sense. We discuss both angles below showing the value of gamification in research from two different angles.

GAMIFICATION AS COMPUTATIONAL MODELS

This ‘game’ was initially conceived as a way for us to sieve out factors that affect why people choose to pay (or not pay) in an honour system. These factors will then be translated into data through a quantitative survey, which can be used to build an Agent-Based Model (ABM). ABMs simulate the actions of an individual, and allow the researcher to simulate the impact of actions of an autonomous agent on a larger system, through his/her interactions with other agents.

Thus, when the game was designed, it was viewed as a life-sized computational ‘model’ for reality. A computational ‘model’ does the job of reducing a complex, real-life problem into logical relations that can be evaluated with computers. In this case, the inputs could be the number of players, their relationship with each other, or their belief in the value of an honour system; the output would be the number of sticks they draw in each round.

A simple problem pops up -- the ‘inputs’ cannot be objectively quantified, but only subjectively quantified by the players/researchers. However, from a computational perspective, this experiment can still be treated as a ‘model’, with limitations imposed due to the subjectivity of the quantitative data.

GAMIFICATION IN SOCIAL RESEARCH

From a sociological research perspective, the difficulty in studying USProvisions lies in the very presence of the researcher being in the same space as the consumer, which postulates an influence of surveillance and peer pressure and thus a change in the way the consumer behaves. The gamification of USProvisions allows the researcher to study a phenomenon that cannot be studied in its natural setting – it alleviates the issue of reflexivity, removing unnecessary external influence on the subjects in an attempt to study and understand them. Using gamification to study USProvisions thus allows the researcher to better understand the genuine intentions of participants in a honour system without creating an impression that they are being monitored, regulated, or judged.

Abstracting this further, gamification allows researchers to study a phenomenon considering both the overall influence of a structure and the autonomy of the individual. It allows for objective observation of the success or failure of an honour system as well as the subjective perspective of the participants given in the interviews, which together give a fuller picture of how agents navigate the imposed structure of an honour system. In this specific study, gamification provides a way to study social actions and intentions that people can be sensitive to and are less likely to be open about. Placing an honour system in the less intense setting of a game removes the researchers’ power in their relationship with the subjects, reducing the presence of the researcher and creates a more comfortable environment for participants to express themselves.

However, considering this study to be a model for consumer behaviour in USProvisions would not be meaningful in a sociological sense, since a natural setting cannot be completely mimicked, and generalizing the study to the rest of the community would be undermining the personal and unique intentions of each individual who has a part to play in the system.

Meet the team:

Tham

Tham is an Engineering Science major who specializes in Computational Engineering. He loves games and puzzles and spends a considerable chunk of his time designing real-life games to provoke thoughts.

Shermaine

Shermaine is a Sociology major who enjoys understanding the personal experiences and realities of all sorts of people. She is interested in the philosophy and ethics of social research.

Yui Wei

Yui Wei is a Computer Science major who loves making things, especially when it comes to interactive media and games, and thinks that communication using models is far more interesting than modelling itself.