A sideways look at economics

Ever feel like you’re doing all the work?

At Fathom we have an unwritten code of offering our colleagues a drink whenever we make one for ourselves. A legacy, possibly, of the early years of the company, when the team was a lot smaller. Or testament to what a nice, friendly bunch we are, perhaps. But for all of this niceness, there have been, and may still be, a few issues with this system.

When I joined Fathom, making a drink could turn into an ordeal. I could barely remember everybody’s name never mind how much milk, or how many sugars, they liked in their tea. The prospect of making up to 17 drinks ultimately changed my behaviour and presumably that of my colleagues, too. There were many ways to game the system: get to work before most people arrived and make your drinks then; pick up a latte on the way to work; wait for somebody else to offer you a drink even though you really want one now; offer to make a drink straight after somebody else has made a drink to give the appearance that you offer every now and then; make a scene every time when you offer a drink so people remember it, creating the impression that you offer more than you actually do; eat lunch at your desk, but pop out for a cheeky espresso afterwards; take the passive-aggressive approach – i.e. offer drinks in a really hurried voice in the hope that the person receiving the offer senses your anxiety, feels sorry for you and says no, or simply doesn’t have enough time to decide if they want a drink before you walk off.

Such a system was not ideal. After all, comparative advantage suggests that it was inefficient for senior Fathom employees, such as our CEO, to be making 17 drinks for the team. (When it was our CEO’s turn to make drinks, he would go out to the coffee shop and buy 17 coffees.) It’s not that senior Fathom employees aren’t skilled at making coffees (although I have some doubts), it’s just that senior employees are better off using their time, knowledge and experience on the stuff that other members of the team can’t do. For example, our CEO may be brilliant at everything, but since he has a comparative advantage at being a CEO, he’s better off focusing on that.

We use the theory of comparative advantage in our lives more than we probably realise. For example, we might decide whether we should do our own housework or whether we’re better off working a little harder, earning a little more money and paying somebody to do that task for us. Should I iron my own shirts, or does it make sense to send them to the laundry? They iron them better than I do, and in less than half the time. Surely it’s more efficient for society as a whole that I focus my energies on being an economist and pay for my ironing to be done. Many readers would probably agree with that, unless, of course, you enjoy a bit of extreme ironing, like these gentlemen.

I digress. As our team continued to grow we agreed that it was no longer feasible to offer drinks to the whole office – we narrowed it down to groups of four. Suddenly, making drinks became a lot more manageable. But the euphoria of that change soon wore off and it occurred to me that some members of my four might not be making their fair share. Laura, for example, got called out once and has turned into a drinks-making machine. The same can’t be said for some other members. It got me thinking and I decided to collect data on the drink-offering and drink-accepting habits of my group. I know this might sound like odd behaviour, and may defy normal office etiquette, but there’s an economic argument for such an experiment. Richard Thaler, who won a Nobel Prize this week for his contributions to behavioural economics, would surely agree.
Admittedly, doing this research did satisfy my curiosity and allowed me to properly assess whether it was true that Joanna wasn’t making her fair share of drinks. (It turns out that Kevin doesn’t make his fair share either.) But the intention wasn’t to make this data public and use this blog post as a platform to raise awareness of this issue in order to cajole my colleagues into making their fair share. No. The real reason was to analyse human behaviour in a microeconomic setting and see if I can infer any conclusions that can be applied to the macro economy.
As you can see from the table Laura has offered more drinks (31 offers) than anybody else over the observation period. She’s also made more drinks for others than anyone else (36 to be precise). I come in second, with 28 offers and 29 drinks made although admittedly, it’s possible that I may have modified my own behaviour since I knew would be publishing this information. So, in the interest of neutrality, I will ignore my own statistics. Compared to Laura though, Kevin and Joanna fare miserably; their offer-to-accept ratio – the key statistic from this study – is less than half of Laura’s. Their made-to-accept ratio is even worse. To be fair, Joanna did point out that she has the highest decline-to-accept ratio, which is true. But she also drew 7 blanks in her 14 offers, giving her a blank-to-offer ratio of 50%, which suggests that she has mastered the old trick of offering drinks when nobody else wants one. Note: 19 drinks-making events took place when I was not present – I was able to record their outcomes since they were documented on our group chat where many coffee deals take place.
The bottom line is that Laura makes way more drinks than anybody else, probably because she drinks more than everybody else. She tells me that she is not annoyed by this. In fact, she says that she’s happy to have a break from her desk every couple of hours and make drinks – truly selfless and barely believable. Kevin told me that he had no idea that his D to A ratio was so shoddy and has upped his game considerably since he saw the results last week. Joanna has insisted that she would make more drinks if she wasn’t so busy all of the time (the high B to O ratio raises a red flag though).

Admittedly, these were not the reactions I was expecting. I thought that Laura would demand that her colleagues shoulder more of the drinks-making burden, and that Kevin and Joanna would admit that they are rational agents and that the logical thing for them to do when they want a drink is to simply sit and wait for Laura to make one for them. After all, research has shown that among flatmates, the person who normally bears the burden of cleaning the flat is the person with the lowest tolerance for mess. It’s a similar situation to drinks making really. The one with the lowest tolerance for not having a drink is the one that makes all the drinks: that would be the one who drinks the most – i.e. Laura. A payoff matrix for the situation might look something like this:

In this world the dominant strategy for Laura is to offer to make drinks since she will get a better payoff by doing so. For the rest of us, we would only offer to make a drink if Laura didn’t offer. But we know that will never happen, at least when Laura is in the office, so we never offer to make drinks. The equilibrium outcome is Pareto-efficient and in the bottom left quadrant.

But what if Laura didn’t like making drinks so much? Such situations do not always lead to such optimal outcomes, and we are not all lucky enough to have a Laura sitting next to us. What if Laura didn’t like drinks any more than we did? A drink standoff could ensue. We may even feel guilty and decide to help Laura with the drinks from time to time – how might that change the payoff?

The bottom line is that there are many permutations to situations like this and each will depend on how much the individuals involved like making drinks, how much do they like drinking and how much empathy they have. Macroeconomists would do well to recognise that sometimes even perfectly rational agents make decisions that do not lead to Pareto-efficient outcomes. It is therefore incumbent on us to heed the lessons of Richard Thaler and possibly even the lessons from the drinking and drink-making habits of the employees at Fathom.