While the Rhubarb Upside-Down Cake is baking in the oven, I’m finally writing this long-due post. A good friend of mine was visiting San Francisco back in Feb and I thought about writing this post, but one thing led to another and my well-planned “so-my-friend-visited-from-Israel-last-week” intro was no longer appropriate. That is, until he surprise visited last week again! So here we go again…taking action before it’s too late this time, because its a very fitting introduction to the study I want to talk about which was done in Haifa, the city where he’s from.

I’ve read this research a while back. It’s a great example of how the traditional economic incentives of giving people money or making people pay for something not leading to the action/behavior you want.

So, here’s the problem. Parents are often late to pick up their child or children from daycare in Israel. It’s not really a problem that my friends and I have to think about at our stage in life, but walk with me. Parents are supposed to pick up their kid from daycare at 4pm, but probably because of work and other commitments (or because they’re Israelis?…they can be late sometimes), they’re often running late. They came up with a solution: lets add a small fine to nudge parents to be on time. The fine will be 10 Israeli Shekel, which is around $3 USD, if the parents are more than 10 minutes late.

They ran an experiment with 10 daycare centers around Haifa – 6 of them were randomized to introduce the fine, the rest were not to introduce the fine. And then the researchers observed for 10 weeks – what happened?

There were more parents who came late after the introduction of the fine, than at daycares without the introduction of the fine.

What? The fine was supposed to reduce the number of lateness, but it surprisingly went the other way and increased it.

Alright, well, that’s not great. Lets reverse it then. We don’t want to increase lateness, so lets put it back to normal. On the 17^th week of the experiment, they removed the fines so that all the 10 daycares had no fine again. They found that the effects persisted – meaning that parents who were late with the fines remained late after the fines were removed! There were more parents arriving late to the daycares that previously had the fines even after the fine period was over…ahhhh!

Why did this happen?

The researchers hypothesize that it’s because parents and the daycare had an “incomplete contract” when it came to being late for pick-ups. This means that the exact consequences of arriving late were not specified. Without an explicit rule, some parents might feel that they shouldn’t be late too many times, and they’re not sure how the school will handle these situations. However, with the introduction of the fine, parents are now thinking “I’m paying for it, so its okay I’m a few minutes late! They’re taking care of my child.”

That’s just one possible explanation. In reality, the researchers were so surprised by the results that they wrote, “the possibility of an increase in the behavior being punished was not even considered” in the existing literature. And thus lies the importance of running experiments and testing whether your hypotheses are right or wrong, because we could sometimes surprise ourselves.

Much of this could be culture specific (having worked with an Israeli, I can attest to significant cross-cultural (uh, hiccups?) differences) or it could be because the punishment is not severe enough. Either way, this study humbles us to be fascinated by behavioral research because you can implement some change, hoping for one result and have it go completely the opposite way! Maybe similar to me putting rhubarb, sugar, butter, flour, into the oven hoping for a rhubarb cake and then have it turn out…hm. I guess someone might hope for chocolate cake instead.

See you again soon my friend!

Love,
Steph

follow me on instagram (@brainandthewind) to see the pictures of the rhubarb upside-down cake!

Source for study: Gneezy, Uri, and Aldo Rustichini. “A fine is a price.” The Journal of Legal Studies 29.1 (2000): 1-17.

Here’s the graph with the main result: