James Heckman: Tinder Dates Gone Wrong
Have you ever gone on a Tinder date that was not what you expected? You immediately swipe right at the sight of an attractive person holding an adorable puppy. But, minutes after showing up at the bar, you find your date is dismissive, rude, and has horrible taste in beer. Sometimes what you see isn’t what you get. When you take a step back, you realize that the things you attributed to the person in the profile picture do not reflect the reality of your date.
This disappointing Tinder experience is an example of what economists call heterogeneity. Though you may have a general stereotype of dog owners, there are exceptions to this rule of thumb — your Tinder date, for example. All dog owners may seem identical in their love of man’s best friend, but are they dateable?
Enter James Heckman.
In 2000, along with Daniel McFadden, he won the Nobel Prize for detangling and identifying problematic assumptions economists have historically made about data. One of the main problems with data is what statisticians call selection bias, which is when the data collected does not represent the population accurately, which can lead to incorrect statistical results. Now, due to the “Heckman correction” which allows researchers to check for these biases, the statistics provide a more accurate picture of what’s actually going on.
For example, previous studies showed that women with three or four children made more money than women with fewer children. However, these studies left out unemployed women. By including unemployed women into the data set, Heckman discovered that women with the greater number of children would simply not accept a low salary offer, i.e. working a low-paying job wasn’t worth their time. So, women with more children don’t necessarily make more money; a whole bunch are actually unemployed! These results provided a more nuanced and complete picture of women and their earnings.
Selection bias was (and arguably still is) everywhere, including public policy. In the ’80s, policymakers thought they could increase the success of high school dropouts by having them take the GED — a test that is a rough equivalent of a high school diploma. More credentials, more money, right? According to Heckman: not really. Though people with GEDs should be more successful than those without any high school credentials, they end up making the same amount of money. This is because other factors — like motivation, which economists don’t often look at — was lacking in folks that dropped out of high school and opted for a GED, keeping them from earning higher wages.
Think of the people who show up in your Tinder feed. Though at first you may feel like an entire city’s dateable population is at your fingertips, over time, Tinder narrows your selection based on your preferences. Your swipes tell Tinder what you like (and what you definitely don’t). Maybe, you only like people with puppies, mustaches, and who work on the Hill. This is selection bias in action. You may think that you have a wide range of preferences based on what Tinder shows you, but you forget that Tinder only shows you what you want.
Alternatively, people’s motivations for being on Tinder may be different than yours. Perhaps you’re looking for something to do on Friday night, and they’re looking for a free meal on you. Or, it could be that they’re in a relationship, and on Tinder for Bygone Econ Icons research purposes…cough.
To sum up, let’s go back to the pay and GED examples. Economists gather data, analyze it, and think they’ve come to an unbiased conclusion. They’re running into a problem not unlike your Tinder one. Researchers may have accidentally surveyed people that don’t represent the entirety of the group they’re trying to understand. When researchers tried to learn more about married women’s wages, they didn’t include the population of women who chose not to work. Or for those taking the GED, researchers didn’t understand the motivation of those who dropped out of high school in the first place.
Though we don’t often think about everyday phenomena heterogeneity, we experience it daily. Because we can’t have complete knowledge of anything — from what’s actually in our burger from McDonald’s to the quality of our date on Friday night — we expect that we’ll occasionally be surprised. For researchers, however, this ignorance of the underlying interactions of people and experiences in data can result in misdirected policy efforts or misguided research. Thanks to Heckman, economists have better tools to test for this.
So next time you’re on Tinder and experience date-heterogeneity, think of James Heckman. Or, not. If anything, give that Friday night date a dose of some serious heterogeneity by diving deep into a discussion about heterogeneity. They’ll never see it coming.