BETA Podcast: Interview with Michael Luca

[music]

Elaine Ung:

Hello and welcome to another episode in BETA's podcast series. BETA is the Behavioural Economics Team of the Australian Government and the team sits in the Federal Department of the Prime Minister and Cabinet.

BETA's mission is to advance the wellbeing of Australians by applying behavioural insights to public policy and administration. Hi, my name's Elaine and I'm part of the team. In this podcast series, we interview a whole range of behavioural economics and behaviour insights academics, practitioners, and policy makers.

This episode is actually part of our mini BX series where we interviewed some of the speakers at the 2018 Behavioural Exchange Conference or BX 2018. There, speakers travelled from all around the world to share the work they've been doing in the behavioural economics and insights field.

In this episode, I interview Michael Luca who is the Lee J. Styslinger III Associate Professor of Business Administration at the Harvard Business School. Michael's research includes collaborating with organisations ranging from Yelp, the crowd sourced review forum of businesses like restaurants, to the city of Boston. His work focuses on designing market places, disclosure policies, and exploring ways to use data to improve decision making within organisations. Hope you enjoy!

Could you please give us a bit of background, who you are and where you come from?

Michael Luca:

So I'm an economist. I went to grad school at Boston University. When I was a grad student, I was studying the economics of information and something that I noticed at the time is that there was a lot of theory around information, but when I look at the world around me, there were a lot of examples of different types of information sources people are using.

I started thinking about things like college rankings and what role that played in people's decision making about where to go and what role that should play in different situations. I started doing this research on the impact of rankings on college decisions and looking at the role of salience of different pieces of the information. After I had written this first paper, I started thinking about all the different types of information that's out there. I said, "Oh, rankings are pretty interesting," but there was this whole new type of information at the time, so crowd sourced reviews. Think about Trip Advisor or Yelp.

I started thinking about how that affects decisions and ended up getting into work around online platforms and the types of information that are available there and how people are using it.

Elaine Ung:

Great. So I guess that does touch on my next question which is how did you become interested in behavioural economics? I understand you've worked, as you mentioned, with Yelp but also with Facebook looking at the data side of things. Could you expand a bit more on that?

Michael Luca:

Sure. I've always been interested in things that had large practical impact. I'm not an ivory tower type. I'm more of a type of trying to be working with organisations and thinking about how they can help to improve the processes that they're doing. How you can improve people's lives through better disclosure, or better designed online platforms. With that in mind, I'm interested in behavioural economics to the extent that behavioural economics can inform the best way to design things.

Elaine Ung:

Great. Could you share with us what you're working on right now? You're talking about the academic side but taking a more practical approach to that. Anything you can share?

Michael Luca:

Sure. One example of a recent paper that we had done was looking at online platforms and the choices that platforms make and how they're affecting decisions. One thing that I've been interested in that domain is the changing role of anonymity over time. I was specifically thinking about Airbnb. I started thinking about this as when people,—I looked at the early platforms so thinking about Priceline, Expedia, Amazon, eBay, all these platforms are basically arm's length transactions. And one way to think about those is they help to make markets more efficient by allowing you to buy stuff that's elsewhere.

But there was another unintended consequence that people viewed as a benefit of the internet age. What people had noticed is there is less discrimination in these markets by the fact that you're taking away some of the information that somebody might use to discriminate. When I looked at that and looked at the way that the internet is evolving over time, I noticed that there's a lot more information that people are looking at now and platforms are designed quite differently. If you think about Airbnb relative to Priceline, Airbnb allows you to see who you are going to stay with and allows the host, importantly, to see who's going to stay with them and then decide whether this is somebody who looks like they should accept or not.

Thinking about this design feature, we had a paper where we showed that there was widespread discrimination on Airbnb resulting from the design choices that they had made.

Elaine Ung:

Could you please give us an example of some of the work that you've done that demonstrates the potential impact of behavioural economics beyond decision making or other examples as well?

Michael Luca:

I think in the context of Airbnb, the fact that they made this design choice changed the ability of somebody to discriminate. Since then Ray Fisman and I put together a guide that we viewed as a managerial guide for how you might use, largely, behavioural economics but other strategies as well to reduce the amount of discrimination on the platform. For example, trying to further automate the transactions or reducing the salience of pictures. So since then Airbnb has implemented a bunch of these strategies and some others in an attempt to reduce the discrimination on the platform.

Elaine Ung:

So that's come out of your work?

Michael Luca:

Yeah. Our work had prompted Airbnb to start thinking about this and they ended up creating a taskforce and now they've actually created a full-time team dedicated to identifying and reducing discrimination.

Elaine Ung:

Wow, that absolutely demonstrates how a lot of the research side can be applied in the practical sense. And it’s great to see that bridge. Hopefully that's what we saw come out of BX as well.

Do you have a favourite behavioural economics concept or bias?

Michael Luca:

I don't have a favourite concept or bias per se. I do have a domain where I think it is very interesting to apply behavioural insights which is online platforms.

Aside from the discrimination works, one area that I've been thinking about is crowd source review platforms. So going back to that earlier Yelp example that I'd given. We recently had a series of collaborations with Yelp where we started thinking about how Yelp could help to improve policy. We were thinking about a few types of things. One was taking Yelp data to try to create algorithms to help cities decide what restaurants they should go be inspecting for health code violations.

On the other side, we've been working with cities to pull in data about health code violations for restaurants and create disclosure policies where the information is disclosed through Yelp. I think when we think about online platforms, it gives you a whole new area where you could start to implement nudges that you wouldn't have been able to in the offline age.

Elaine Ung:

I think that's a really exciting area in especially data, and understanding data better nowadays, is actually using it for different reasons and different areas where originally it might not have been intended for and the by-products of that. What we see coming out of it is so powerful and it's absolutely the new tool we can use as long as we know how to do it.

Given all the work you've done, do you apply behavioural economics to your everyday life? So how do you BE yourself?

Michael Luca:

I guess there are a variety of ways that I use behavioural economics in my own life. Reading about this in the Wall Street Journal about how people could bring out their more charitable selves, thinking about some of the devices that we could use. I've done that personally. One area recently where we used some behavioural economics to improve our lives as a family is my wife and I decided to run a half marathon. The whole decision was based in behavioural economics. When thinking about what a half marathon is, we use it as a commitment device saying, "Okay. You're planning this well in advance." Kind of in the moment you might never want to do it. Three months beforehand you can say, "Alright, I want to do this." And then you could build up to it.

The second thing is there's this research on goal setting. So thinking about not just the 13 miles you'll be running that day but trying to build up to that 13 miles is going to give me some goals to work toward along the way. During planning there are two behavioural strategies that we use to keep ourselves motivated. One is implementation intentions. We said, "Here's what we want to get toward," but we started making these small plans to say, "What are the steps we need to get there?" The second part is the social component where we publicly committed to the fact that we were running this so that people could shame us if we didn't make it across the finish line.

Elaine Ung:

I was going to say, that public commitment is such a big driver for keeping you on track. It's excellent to see.

Michael Luca:

Yeah. I think the night before it's only the public commitment that stopped me from not running it in the end.

Elaine Ung:

And it works. Could you share with us what you think is the next big thing for behavioural economics?

Michael Luca:

I don't know if there's one next big thing, but I do feel a few limitations and areas for further progress in BE. For instance, one project that we worked on recently was looking at hand gun waiting periods. In the U.S. there's still a lot of gun violence. One of the things we were thinking about is what are some politically viable ways where you could take behavioural research to create new gun policies. So this is with my collaborators Deepak Malhotra and Chris Poliquin. What we found is that even short waiting periods, so meaning you want to buy the gun today, you need to wait three days to do it, could have large impacts on the number of gun deaths. That three day waiting period leads to something like a double digit percent drop in gun violence.

Thinking about that is completely consistent with what you’d expect from the behavioural research. It's something where gun violence is, in many cases, a hot state phenomenon. Even having a short cooling off period could have a big impact. The way that I view this as why is this the next step of BE policy is because it's taking behavioural insights and not kind of nudging given existing policies but baking it into policy. So now there's a new Bill that was introduced in Congress based on this research to say that there should be a three day waiting period for hand guns. That's one area.

The second area I think that there's room for improvement in the way that we think about behavioural insights and policy, is just thinking more broadly and more long term about what the goals that we have in mind are. A lot of interventions will have narrow goals, say, increase school attendance. I think that's great but we need to keep our eyes on what are we ultimately trying to do. Maybe you're trying to increase productivity among students, probability that they'll graduate from school. So I think investing more in long term outcomes and a broader set of outcomes could lead to a richer understanding of how interventions are affecting behaviour in the long run.

Elaine Ung:

I think that's absolutely right. I'm seeing this trend towards a lot of, where behavioural economics and behavioural insights can add value to broad social impact. I think that's a really exciting thing to see come out of it.

One last question. What's the next big project for you?

Michael Luca:

We've got a few things going on. One is I have more collaborations with Yelp trying to think about how they could use their data to improve things like,—we're working with the U.S. Census data and we've talking with people at Census about how you can improve their forecasting models using crowd sourced data from online. That's one area.

The other area I've been thinking a lot about is actually related to this outcomes point that I was just making, which is thinking about how can you more systematically use experiments and business and government settings. We're creating a new course that's basically experiments for managers. So trying to understand what are the frameworks that could tell you when you should run an experiment and when you should just implement something. If you're running an experiment how much should you be focusing on a mechanism versus a policy? I think we're starting to get to a point where we're seeing lots of experimentation being used, but we haven't quite yet seen the full frameworks for somebody who is walking into an organisation to think about how exactly you should experiment.

Elaine Ung:

Oh wow. So lots of exciting new things coming out soon. All the best with that.

Michael Luca:

Thank you.

Elaine Ung:

Thank you so much for your time today.

Hi again. Thanks for listening. If you want to learn more about Professor Luca's research you can find more information on the Harvard Business School website. Professor Luca's research has also been written about in media outlets like the Wall Street Journal, The New York Times, The Washington Post, and Harvard Business Review.

If you haven't heard our previous episodes, listen to them at www.pmc.gov.au/beta. Until next time!