Off The Record #12: The truth about lying
A game to measure dishonesty
We have, at some point in time, heard a variation of the phrase ‘always tell the truth’, be it in tales of the boy who cried wolf or as a warning about what happens to those who lie. As we grow older, situations do not always register as black or white, the subtlety of operating in grey areas comes into play. We learn that not all issues are easily categorized into an ‘either or’ context, more often than not, we have to take a number of variables into account. Therein lies the essence of our Off the Record. We replicated an experiment originally conducted in Kenya in our newly launched lab in Lagos to capture the nuances that exist within and across settings, with a game to measure dishonesty.
Dishonesty manifests in different ways across sectors (fraud in finance, cheating in an exam etc) and understanding mechanisms for avoidance can have great implications. Variables such as culture, social norms, religion, climate and political climate — and the complex interactions between them — define not only how a common issue manifests itself, but the solutions that might be most effective in a given context. Understanding the biases and heuristics that exist in a given context is an important precursor to designing products, services or interventions for individuals and communities. This is the basis of our work, to develop locally relevant insights and capacity. By building and operating labs in multiple developing countries, we are able to capture context specific differences.
What did we do?
To start, we invited 250 residents from the Makoko community in Lagos to take part in the coin-flip game. The demographics of the participants were chosen to mirror the sample population in the Kenyan study. Participants were asked to flip a coin 20 times and report their results. Each time they landed on “heads”, they would be rewarded with a small financial incentive (N75 — $0.20 USD). Every participant was faced with a dilemma when they landed on tails: tell the truth, or lie to increase the total monetary reward received at the end of the game.
Similar to the original experiment, we had multiple treatment arms with an additional treatment to understand inter-group dynamics.
Control — We asked participants to flip a coin 20 times and report their results. Each time they landed on “heads”, they would be rewarded with a small financial incentive (N75 $0.20 USD).
Treatment 1 — Monitoring. Respondents were told “Our Lab Managers are watching you execute this task in the room”, with lab staff observing their actions.
Treatment 2 — Framing. Respondents were told “We anticipate most responses to fall between 8–12 in this exercise.”
Treatment 3 — Rotation. Respondents were asked to change seats every five flips of the coin. Each respondent was paid according to the results from the last computer at which they performed, so this change would remove much of their control over their final total.
Treatment 4 — Group Work. This was introduced in the Nigerian study. Respondents were put into groups of five. The groups were given a five minute window to strategize and discuss the order in which each group member would toss the coin. At the end, the total amount of money earned would be divided among all the members in the group equally.
Results
In Kenya, a high proportion of people said they landed all 20 flips on “heads” than what is realistically possible. Whereas the results were similar in Nigeria, there were higher clusters of people saying they obtained more heads towards the right side of the histogram.
Our results show that the seat rotation group was the most effective in comparison to the control. This is in contrast to Kibera where the framing of expected responses between 8 and 12 was the most effective.
Treatment 3 not only had the participants change seats a few times, but paid participants according to the results from the last computer at which they performed. Here, participants were nudged to honesty presumably because it was less profitable to lie as the benefits would accrue to a stranger.
Interestingly, the new group treatment arm increased dishonesty by 7%. Groups were more likely in this experiment to coordinate to deliver higher overall returns. Some potential reasons we feel might have happened from a behavioural standpoint:
- A smaller number of tosses would require each individual participant to ‘lie’ fewer times, thereby increasing the possibility of falsification
- Groups were better able to coordinate and potentially validate a decision to falsify results. This is presumably due to pressure to engage in a perceivably ‘smarter’ behaviour of cheating.
The second interpretation is particularly interesting in the context of group or cooperative lending, where, if unchecked, group solidarity might lead to lower repayment rates.
So what?
Our results show that the most successful treatments for reducing dishonesty differ in each sample studied. Such findings are useful when deciding how to design a financial product or service for a particular audience. For example, take a bank seeking to increase loan repayments rates in Nigeria and Kenya through SMS reminders. In Kenya, such a provider might use social norm framing by communicating general repayment rates and frequencies; in Nigeria a more successful strategy might be to use gain-framed messages, emphasising the personal benefits of paying on time and in full.
In short, the truth about lying is that, though everybody does it, there are different contextually relevant ways to address it and plan for it.
A final note
Our goal is to make research accessible in order to start, continue or inform conversations that help us to better understand human behavior.
With this in mind, each “Off The Record” post provides access to the full findings from our studies, freely available here.
As a commitment to Open Science, we keep this anonymized data live at this page for all our on-going research efforts.
Access the full data relating to this post here
This blog post is part of our larger “Off The Record” initiative where we share findings from small-scale research projects, designed to collect initial data and kick-start a conversation.
If you would like to learn more about a specific human behavior or have a research idea you think we could explore for a future “Off The Record”, please reach out to us on Twitter or via email.
