Voting and (im)moral behavior: Dr. Kajsa Hansson discusses her new paper

Diffusion of responsibility is a concept from social psychology that is a key factor in what’s known as the ‘bystander effect’. This is a process in which the likelihood of people offering help in an emergency decreases as the number of other people passively observing the situation increases. In essence, diffusion of responsibility occurs when people choose not to help as a result of feeling only a fraction of the responsibility to do so. The thinking goes something like, “Nine other people are here… I’m only 1/10th responsible for doing the right thing.”

There is scientific evidence suggesting that people making market-based decisions involving other people may fall victim to this psychological phenomenon, leading to an increase in immoral behavior in the marketplace. Would this same process be implicated in majority-rule voting?

I recently had the pleasure of talking with Dr. Kajsa Hansson, a behavioral economist and one of the authors of a 2022 paper exploring this exact question. Across three experiments (two of which recruited a total of 1,409 participants on Prolific) the paper looked to see if our behavior changes for the worse when we engage in group voting.

You can access the full paper here.

Could you give me a bit of an introduction to yourself and your area of research?

I have a Ph.D. in economics and I did my doctoral work at Linköping University in Sweden, which is where I was affiliated when the paper we’re going to talk about today got published. I have been working a lot with psychologists, so I'm what’s known as a behavioral economist. I have mainly worked with psychological theories to understand human-decision making and moral behavior.

At the moment I'm a postdoctoral researcher at the Vienna University of Economics and Business, where I work at the Institute for Cognition and Behavior together with economists and psychologists. Together we conduct research that seeks to understand human decision-making and the cognitive processes that underlie those decisions. I also have continuing collaborations with colleagues at the JEDI-lab based at Linköping University.

So, you’re a bit of a hybrid researcher, mixing Psychology with Economics. How has that experience been?

I think more and more it's getting to be kind of a mainstream way to approach behavioral economics. But I do think some of the most traditional economists would look at me and think that I'm a psychologist, and psychologists would look at my work and likely think that I'm an economist. I think the biggest difference between being a very traditional economist would be that I almost exclusively use psychological theories in understanding how cognitive biases can affect human decision-making.

Today we’re discussing your paper “Voting and (im)moral behavior" which explores the social psychological concept known as ‘diffusion of responsibility’ in the context of decisions made either individually or in the context of voting in a group. Can you tell me where the idea for this study came from?

The idea behind this paper came from a line of literature looking at moral behavior in markets. There was a very famous study by Falk and Szech (2013) that showed that people become more immoral when they make decisions in markets. In the economic literature, there’s been a lot of debate about the reason for these results, and many have suggested that it is due to diffusion of responsibility. In markets, you are never solely responsible for the outcome. You share the responsibility with others.

In that Falk and Szech paper, they had participants negotiating for a price to kill a mouse, or to sell and buy a mouse, and then compared those decisions to an individual-choice price list. This was a huge finding, and a lot of papers tried to understand why this immoral behavior happened in that context.

So we took a ‘spin on the wheel’ and we thought that diffusion of responsibility is not only linked to things that we do in markets. In democratic decision-making you usually want people to think about the common good and then provide for the whole society, or to think in terms of what is good for society and not what's good for me. So, we took the concept of diffusion of responsibility and looked at it in the context of voting. It's a kind of contrast, because in markets we expect people to be more selfish and think more of themselves whereas in democratic decision-making we expect something different.

Your paper has three experiments. I was wondering if you could walk me through each one in terms of design and results?

Sure. In the first study, we focused on large groups. We looked at majority-rule voting behavior in a group of 49 people, which is fairly large compared to past research in this line of literature that has typically used a maximum of between 5 and 8 people. We wanted to make a large group and we compared it to an individual decision-making scenario. We asked participants either to donate a sum of money or to keep the money for themselves. That was the moral decision being considered.

In Study 1, we found no effect at all, and we really thought we would find something there. We even included a consumption decision to see if there was a difference between a consumption decision where you don't have to think about the morality of it, and where you can think only about yourself. And so we found no diffusion of responsibility effect for their neutral consumption decision, and we found no effect of voting on this moral decision, either. They were practically identical.

In Study 2 we changed the design so that participants had to extract money from a donation. We told them, “Okay, this sum of money is going to be donated to a charity organization. You can choose either to extract some of the money or all of the money, or you can just leave it as it is and all of the money will go to the charity.” Again, we compared individual decision-making to majority-rule voting decisions, but we assigned all participants to groups of five. So participants made decisions either by voting and then we implemented the majority-rule for the voting group, or the people in the individual condition just made their own decisions and had their own decision implemented. In this study, we found a very small but significant effect that was actually opposite to the diffusion of responsibility hypothesis. Here, we found that majority-rule voting led to a slight increase in moral behavior, not immoral behavior.

In Study 3, rather than looking at majority-rule voting, we told participants that the final group outcome would be based on the median vote. They had the same kind of option and set-up as before, where they could extract an amount of money. They were still in groups of five people and they just had to indicate how much they wanted to extract or if they didn't want to extract anything. Just like in Study 1, we found no effect of group voting on moral decision-making, compared to the people making decisions solely as an individual.

In Studies 2 and 3 we also assessed participants’ beliefs about their vote being the decisive vote. If diffusion of responsibility would be something that could affect people in this voting context, we would expect that people who think that their vote is less pivotal to the final outcome would likely act more selfishly. We didn't find support for that either.

Let’s turn back to the broader issue of diffusion of responsibility and its potential role in immoral behavior. What is your takeaway from these three studies?

We find no like clear evidence that diffusion of responsibility is relevant in a voting context. And I think that's important to highlight that… I mean you could still see a diffusion of responsibility in contexts such as teamwork, in that the more people there are on the team, the less you might contribute. Or in markets, for example, our findings do not exclude the possibility that there exists some kind of diffusion of responsibility mentality in markets. But when you make decisions together with others in a voting context, it could also influence what kind of norms there are. So we are making this decision together with others and maybe we’re thinking more in terms of, “Okay, we could actually contribute something to the common good here if I choose not to extract so much from this donation or not to extract from it at all.”

Having explored the potential role of diffusion of responsibility in (im)moral voting behavior, are there any other psychological theories or processes that you're thinking of investigating in this context?

One area of interest is motivated reasoning in a voting context, where individuals lack incentives to be informed about the best policies because their probability of casting a decisive vote is very small. This can lead to detrimental effects as people may simply vote for what they want to be true, protecting their self-image and experiencing cognitive dissonance if their choices deviate. With the low individual cost of uninformed voting, understanding these psychological processes is crucial for improving democratic decision-making and promoting an informed and engaged electorate.

Let's get back to the null effects in your paper. Here at Prolific, we recently hosted a webinar about online data collection platforms like ours and how our samples can help with issues such as reproducibility concerns. It wasn’t that long ago that ‘no findings’ meant ‘no publication’. What does it mean to you as a young scholar knowing that you can design theoretically-sound experiments and know that there's a high-impact avenue for you, even if the effects are null?

I would say it definitely creates a good incentive to do rigorous and good research, with well-derived hypotheses. And this comes from pre-registration where you can show people that you have been thinking a lot and you have these kinds of hypotheses.

If your incentives are to just publish surprising and very novel results, yeah, I think that could create a weird incentive structure and create a lot more stress and produce very weird results as well. So, yeah, I think it's important especially for younger researchers to have this kind of confidence or believe that even though the results are not significant, you don't have to start p-hacking just to find something. You can publish good research either way.

Where can our readers stay up to date with your work?

You can find me on Twitter and my personal website. Also, you can keep up with the CoBe-lab and related work from my co-authors at Jedi-lab.

This research was carried out using the Prolific platform and our reliable, highly engaged participants. Sign up today and conduct your best research with our powerful and flexible tools.

‍