How we use economics to get you tipped

A busker who doesn’t like what we do once told me that “economics doesn’t belong in street performance.” I think what he meant was that we—The Busking Project—shouldn’t be the ones making economic decisions that affect how much you get paid.

I get his point. But, when designing our website, we can either create arbitrary designs or we can base our decisions on solid economic foundations, but either way we will affect how much you earn.

It’s a moral question. Should The Busking Project avoid trying to understand and use buskers’ tip data—payments that have been anonymous for thousands of years—or should we try to help buskers earn more money?

I think during a cost-of-living crisis the answer is clear.

However, we should also be utterly transparent about what choices we’ve made and why. So, below is a complete breakdown of the economic principles that we’ve considered regarding how our payments platform works, the tests we’ve done and the results.

[Economic principles are in bold (in case you want to google them.]

It is not true that audiences give cash tips based solely on what they think a show is worth.

First, there are physical limitations. If someone only has a 10p coin and £20 note, they’re either tipping a busker one, or the other, or both. Or, if they only have a bunch of small change in their pocket, they might give all of it without counting.

Tips also depend on something far more basic: which coins are in circulation. The biggest coin in the UK at the moment is the £2 coin — worth is worth $2.40 USD today, and was worth $3.32 when it was launched in 1998. I haven’t seen any studies on this, but I bet you the average tip to buskers in the UK was bigger than the US for that simple fact.

Buskers will probably be happy to hear that the United States will stop minting 1 cent coins in 2026. If you can believe it, the 1 cent coin makes up half of the USA’s coin output every year. I doubt pennies are worth the effort of counting, rolling and depositing individually, and buskers won’t ever have to do it again, soon enough — because people won’t be carrying them any more.

And spare a thought for Lebanese street performers. The highest note in circulation in Lebanon is LL 100,000, which is only worth about $1.12 USD. For reference, average rent for a one-bedroom apartment in the centre of Beirut is around $700 per month.

All that is to say: there are a lot of non-artistic factors that go into determining how much you earn.

Buskers are, of course, well aware that their art isn’t the only important consideration.

Musicians often put ‘seed money’ into their cases. This tells passersby that other people have donated. You or I might call that ‘peer pressure’, but economists call it social proof, which describes how the actions of others impact our behaviour.

When buskers say things like “I think my show is worth at least as much as a beer,” they’re value signalling, suggesting their show has market value, and using reference pricing to say a $10 tip would seem fair.

There are outliers too. I know a Canadian busker who uses the word “twenty” often in his show. He says things like “It must be at least twenty degrees out” and “Now that I’ve got twenty people watching”. So, when he ends his show saying he'd love a ten, twenty or fifty dollar tip, he thinks that he gets far more twenties as a result.

If I remember correctly, he didn’t do great in school, started busking in his mid-teens and continued busking for over three decades. My point is this tactic didn’t come from book learning, it came from clever thinking, without needing to be told that he’s priming his audience by using familiarity bias so that “twenty” feels more normal.

If buskers use economic tactics to get paid more, we should at least try to keep up.

In the hat line I quoted at the top of this post (the $20 sex joke), buskers are anchoring audiences to consider three different values, which act as reference points when tippers get out their wallets.

Our platform is built the same way. When audiences zap your QR code, they’re taken to a page with three options — the default tip in the middle, and then higher/lower buttons on either side:

There is a commonly observed decision bias that comes into play when people are offered three options to choose from. In general, they tend to go for the one in the middle. This is called the compromise effect (although I prefer the less-frequently used goldilocks effect).

There are lots of theories for why.

People generally don’t like to be wasteful with their money, but they also don’t want to be seen as cheap. This is called loss aversion: choosing a high amount would feel like a loss of assets, and a low amount could be a loss of respect. Choosing the option in the middle helps you avoid both types of loss.

Similarly, there’s a trade-off between two ‘attributes’: on the one hand, you want to feel generous by supporting an artist you like; on the other you want to have money. Giving the middle amount means you’ll still get the emotional reward of having given a decent tip to an artist you like, and yet you’ll also save money by not picking the most expensive option. This is called attribute balancing.

Then there’s the theory of social image. We want to look fair, generous and reasonable. When presented with three options, it’s easier to rationalise the middle amount as not-the-worst and not-the-best. It’s the safe option, because it’s more defensible than the other two: “I chose the middle option because it wasn’t unreasonably high or insultingly low”.

Finally, the middle option is the default. Not only is it easiest to go with the default, but by having a default at all our platform is giving it an implicit endorsement. We are telling people “this is the right choice”, or the “status quo”. It may seem to tippers that the default is what they are expected to tip.

It’s not just tippers who tend to go with the default option. Even though buskers can change the default settings in their dashboard, they mostly they just stick with the number we give them when they first make an account.

That number is therefore hugely important. If we pick a bad default, buskers will stick with it, and it’ll be our fault that they’ll earn less than they otherwise could.

So, what should we set it to?

Originally, gut feeling meant we decided that “5” was a good default, depending on the currency. New buskers in the UK had their default set to 5 pounds. In the USA it was 5 dollars, in Australia it was 10 dollars and in Japan I think we set it to 500 yen.

But, data from our platform showed that the average tip in most currencies was actually higher than the default that we set — meaning the default number we were endorsing was lower than the amount that people actually wanted to tip.

Obviously not great, but that’s what happens when you trust your gut.

So, we should raise the default, right?

The answer isn’t so simple. A study on charitable donations showed a scale-back effect (or backfire effect) when charities asked for larger donations. As a result, often tippers chose to donate less than the average, or to walk away without donating at all.

Also, a lower bar effect has been observed at cafes and restaurants. Lower suggested tip amounts increased the percentage of people who left a tip at all. This meant that although individual tips ended up being lower on average, the total value of tips increased.

So, we had to run a test. But to do that properly, we’d a) need a lot of street performers to agree to be involved in the study, and b) the ability to factor out things like pitch location and the weather. The perfect location would be at the world’s biggest buskerfest: the Edinburgh Fringe.

So, in 2021 Lily and I applied for a grant to do such a study, which, amazingly, we won. We used those funds to develop a tracking system. Then, in 2022, we travelled to the Fringe to hand out free QR code signs to any busker that wanted to take part in the study.

The test was simple. Every time an audience member zapped a QR code, the tip page would randomly show either a £3, £5, £10, £15 or £20 default in the middle, like this:

As it was random, we could get a pretty good read on the impact of changing the default in terms of the average amount buskers made from tips. We didn’t see a significant scale back effect, but we did see a significant increase in the average tip amount. We showed that at the Edinburgh Fringe, you would earn less with a £3 default value than with a £15 default.

The result is that we changed the platform’s default tip for new UK accounts from £5 GBP to £10 GBP — and expect those buskers to have earned more as a result.

Our Silicon Valley competitors might have billions of dollars, but none of them are specifically trying to maximise how much buskers earn in the street. Nor would they make the kind of effort we did to help buskers out.

This was a multi-year project, and quite personally expensive. Lily and I spent hundreds of pounds of our own money in order to be at the Fringe in person (the grant didn’t cover accommodation).

Of course, it was massively worth it. Aside from everything else, it was hugely rewarding to see so many street performers use our signs at the Fringe—something that has continued to this day.

We’re looking into what other studies we could do. One might be getting buskers all over the world to opt-in to a year-long study randomising their default tip amount. The reason to do such a study is that if we had enough data we’d be able to say things like, “if you’re a musician in Melbourne, you should set your default to $X.”

The bottom line, though, is that we’re constantly working to help buskers earn more money. Yes, we’re doing this by analysing tip data on our platform, and yes, that feels icky. But the result is that you’ll earn more money from our efforts.

And we think that’s worth all the trouble.

Vote to see results!

Again, I’m sorry for our premium members, but no “closing my tabs” this week. I’m still playing catch up after my trip, and have just come down with a fever 🤒

Thanks,

Nick