me Convention: Testing wild ideas using design thinking and the blockchain
Does fairness impact how people consider universal basic income?
Earlier this month Naveen and I went to Stockholm to lead a workshop at the me Convention. The topic was testing crazy ideas in the blockchain space using design thinking. At the conference we wanted to test some ideas around universal basic income. We wanted to see how peoples’ feelings of fairness impact how they used their universal basic income (UBI) and how they felt about other people getting basic income. Here’s how it went down.
The warmup experiment
We started with a short experiment to warm up and get to know one another. We split the workshop participants into groups of three. We asked them to figure out who lives the furthest from Stockholm and that person became the “Decider.” The Decider picked up an envelope from the table. On the envelope was a blue and a red post it note. One of the remaining two people in the group got the blue post it and the other the red post it. Now the Decider opened the envelope and read the instructions.
Hello Decider! Here is some information about your two partners. Read it to yourself and do not share it with them.
Blue Partner — Your blue partner played a game this morning called the “Dictator Game.” In it, your blue partner and another person were paired up. Blue partner was given $20. The rules of the dictator game are simple — the person receiving the money can keep any portion they want for themselves, and give the rest to the other person. Blue partner decided to keep $18 for themselves, and give $2 to their previous partner. So selfish!
Red Partner— Your red partner also played the Dictator game this morning. In it, your red partner and another person were paired up. Red partner was given $20. The rules of the dictator game are simple — the person receiving the money can keep any portion they want for themselves, and give the rest to the other person. Red partner decided to keep $11 for themselves, and give $9 to their previous partner. Quite fair, right?
What you need to do now — You, the Decider, now need to make a decision. You can choose to split 12 gold chocolate coins with blue partner or split 10 gold chocolate coins with red partner (leave the other two in the envelope).
When this game was played in a lab with small amounts of real money, the Decider tended to share with the Red partner. Splitting money that unevenly in the Dictator Game is commonly seen to be outside social norms and the Decider uses her power to punish Blue partner. This is often seen as an irrational economic decision, as splitting $12 with Blue Partner is clearly an economically superior outcome for the Decider. However, as Richard Thaler writes in Misbehaving, “There is clear evidence that people dislike unfair offers and are willing to take a financial hit to punish those who make them.” From Misbehaving, p.142.
During the workshop, we then asked people with a blue post it and a gold candies to stand up, and we counted them. Then we asked people with a red post it and gold coins to stand up and counted them. Of the 15 teams, 11 Deciders chose to split with the red partner and 4 with blue partner. When asked, deciders said they chose red because they liked them more and it felt fairer. They did not express concern about the profit they gave up. The best quote from this experiment came from a Decider who gave to the blue partner, “because everyone deserves a second chance!”
Finally, we revealed the purpose of the experiment. We expected to see many more red partners with gold tokens and that was about how it turned out, and for the reasons described in Misbehaving.
We believe we can run short, simple experiments to see how people will behave in real life.
We also revealed the underlying lesson of the workshop. We believe we can run short, simple experiments to see how people will behave in real life. We don’t have to build anything major or special. We don’t have to “build it and see if they will come.” We can learn how people will really behave and we can do it very quickly and cheaply. For example, we recently ran a full scale test for Token Curated Registries at ConsenSys and learned a lot from that experience.
How does fairness impact how people view UBI
Now we arrived at the heart of our workshop. We wanted to explore how peoples’ feelings of fairness impacted how they felt about universal basic income.
We already knew people care about fairness and that those feelings impact their decisions and actions. We saw it in our first experiment when people who didn’t split a pot evenly were punished by the Decider. We know people care a lot about transactional utility. That’s when people get value from a “deal.” We know people hate mandates but love choices. We also know that poor and rich people have different marginal propensities to consume. (All ideas borrowed liberally from Misbehaving — thank you Richard Thaler!)
Our research question: Could we give UBI to everyone (the universal part) so it feels fair, but then also put it in a specific mental bucket so the poor are likely to spend their UBI while the rich are likely to gift it? Our assumption was people won’t complain about how UBI isn’t fair if everyone gets the same UBI. And since this money isn’t allocated previously in peoples’ budgets, UBI won’t (at the start) change their marginal propensity to consume. So poor people can get a lot of value from the additional income. And richer people who might not get much value from the additional $500 a month MIGHT get a lot of value from donating it to a charity or directly to poorer participants of the UBI.
The UBI experiment
We started by showing this video about poor and alcoholic residents of Seattle getting free housing and a place to drink. After watching the video, we asked the participants to rate if they thought it was fair to spend taxpayer dollars to give homeless drunks a free place to live and drink. RESULTS: We assumed the answers would skew further left considering we were in Sweden, with many locals and Europeans among the conference attendees. On a scale of 1–4, 1 being totally unfair and 4 being totally fair, about a third said 4, another third said 3, and the remaining third was split between 1 and 2. Many people thought intervening to help citizens who need help the most was fair. Others (in the minority) thought free housing and booze for alocholics, when the rest of us have to work for our homes and drinks, just feels unfair.
Quick note about econs: In Misbehaving, Thaler introduces us to Econs. Econs are perfectly rational, profit maximizing, fictional beings, created by economists. In a world populated by Econs, how one felt about drinking and public assistance wouldn’t matter. If it was cheaper or better to give people public housing vs citations, or direct cash versus welfare, we’d do it. But we aren’t econs, and there are hesitations about both those things. As we saw in experiment with the Decider, whether people view something as fair plays a strong role in how they will behave next.
After watching the video we ran the experiment. Here are the steps we followed.
First, we distributed envelopes at random to all the attendees that described the job the person has, the amount of money they make, and typical monthly expenses. We included a wide range. We wanted some people to really feel like the UBI they got would make a large impact on their day to day life. These were people with a high marginal propensity to consume. Others were on the other end of the spectrum — the UBI would not impact their day to day much and we expected they would have a lower marginal propensity to consume.
Second, we distributed one “meToken” in the form of a gold chocolate medallion to everyone in the room. The tokens were worth $500. We told everyone they would get one token every month for the duration of the experiment (12 months in this case). The only condition on the meToken was that it had to be spent in the month it was received. It could not be saved for the next month.
Third, we told everyone they had a choice to make each month. They could spend the meToken as cash on anything they wanted, or they could donate the meToken to their preferred charity, or they could donate the meToken to a Holiday Fund. The Holiday Fund would be distributed after all 12 rounds were finished to the 20% lowest incomes in the room and each recipient would get an equal share of the Holiday Fund.
Then we played the rest of the 12 rounds, keeping tracking of what people chose to do with their meTokens.
Finally, we asked people to measure how “fair” the UBI seemed to them, on a scale from 1–4.
What we make of this
First, reactions to the experiment.
As with almost any first time test, there were a lot of questions about the instructions. One big realization was that we should have had the instructions on slides to make it easier for people to follow along. We really wanted to not use slides but it was a mistake in this case. There were also lots of questions about how the meTokens could be used. So overall, while we made it through the experiment, the instructions could be far clearer. Also, it was hard to keep track of how all 45 people were splitting up their meTokens since they did it on paper. I would have liked to know how everyone ended up deciding to spend their meTokens compared to their income and expenses. That means we need to consider a digital way of allocating the meTokens, or perhaps taking a picture of everyone’s sheet at the end of the experiment.
Second, results from the experiment.
The richer people were, the more they donated to charity and the holiday fund. The very rich donated nearly all of their meTokens. In fact, we ended up with 20% of ALL tokens that were distributed being given to the holiday fund (we didn’t capture what percentage was donated to charities unfortunately). Participants reported that they felt good donating to charity or the holiday fund. So on the face of it, we seemed to answer our research question in the positive. However, once we dug a little deeper, we found that some cracks appeared. People didn’t universally rate the system we implemented as fair. In fact it was quite well split up among the participants on whether it was or wasn’t fair.
One person said how can it be fair when rich people get the same as poor people? Another person said it isn’t fair to give the 20% poorest the holiday fund, when it effectively doubled their meToken UBI income, when the person sitting at 21% got nothing from the Holiday Fund at all.
We found this feedback to be really insightful, because in one hour with a simple experiment, we exposed a fundamental fault line in the conversation about UBI. If you follow the universal part, it feels unfair that rich people get money they don’t need. If you don’t give it to everyone, then that provokes a conversation about where exactly to draw the line between those who get and those who don’t. If we had more time, we could have delved deeper into why people felt that way and surfaced some ideas for how to resolve these tensions in a future experiment.
Lots of people to thank
Thank you to the me Convention for having us out. We had a blast! Thanks to all of the participants in our experiment, we couldn’t have done it without you. Hopefully you enjoyed participating and also learned something useful about testing crazy ideas with design thinking. Thanks as always to ConsenSys for being a wonderful place to work and giving us space to explore all sorts of crazy ideas!