Came across this article on Wharton’s website this morning, which is based on a paper about the economic impacts of student loan forgiveness. Everyone just assumes that forgiving student loans — somewhere on the order of $1.6 trillion — would free up a bunch of money and jump the economy, i.e. younger people could buy homes, whatever whatever. Instead, this paper argues that it would increase inequality, and most of the benefit would go to those in upper-income brackets already. I will spare you all the math — if you’re into that stuff, you can read about it at the first link — but here’s some of the essence of it:
Under a universal loan forgiveness policy, in present value terms, the average individual in the top earnings decile would receive $6,021 in forgiveness, compared to $1,085 for those in the bottom earnings decile, the paper stated. In fact, households in the top 30% of the earnings distribution receive almost half of all dollars forgiven. The patterns are similar under policies forgiving debt up to $10,000 or $50,000, with higher-income households seeing significantly more loan forgiveness, the researchers write.
The benefits of student loan forgiveness are unevenly distributed also by race and ethnicity, Catherine and Yannelis found. The average loan balances are the highest among blacks at $10,630, while those for whites are $6,157, and for Hispanics and others they are $3,996. After adjusting for the present value of those loans, universal loan forgiveness would lead to roughly equal average benefits for whites and blacks, but would yield significantly lower average benefits for Hispanics and other groups, the researchers noted.
Right. Makes sense, although again, I don’t always follow all the math. I’m none too bright.
Then you get deeper down the article, and there’s this pull-quote:
While a nice thought, that’s just not true.
Fam, eVeRyThInG is about emotion
I’ve been banging on beliefs vs. data for a few years now, including two semi-recently: the power of belief of narrative, and how funny it is that we claim to be data-driven when we’re clearly belief-driven.
So let’s use student loan forgiveness as an example here. This paper I linked above is claiming it’s a bad idea for data reasons, and arguing we shouldn’t look at the emotion of it. But emotion is commonly how people interact with ideas and decisions and narratives, and here’s the emotion around student loans:
- “Why should these hipster Gen Z assholes get their loans forgiven? I worked to pay mine off.”
- “Why are we handing these younger generations free money? What happened to merit and hard work?”
- “They’d probably just squander it anyway.”
- (partisan one side) “You are animals for not considering this. Think what it would do to spending!”
- (partisan other side) “I’m so tired of these Ds making us into a fucking socialist country.”
That’s all emotion. It’s all belief. An issue like student loans cuts right to the heart of personal accountability, finance, how people spend their time and money, and the impacts (or not) of education. It also has a tangential tie to the housing market. All these things are incredibly powerful belief-drivers for people; a man’s home is his castle, right? Personal accountability is in some ways the primary modern dividing line between left/right. Finance is a massive issue; it determines the quality of life you can have. How people spend their time? How the government deals with money and span of control?
These are huge topics that people have long-held, entrenched beliefs about.
How is data going to change minds on something like that?
It’s the same at work, guys
I had this gig about five years ago. I was on some “digital team” that no one really cared about, and most of them sat in Seattle while I sat in Texas, so they cared about me even less. It was hellacious. Eventually I got fired from it. It sucked at the time, but it led me to bigger and better things, including, somewhat oddly, a new marriage. Long story there.
So let’s talk about belief for a second. The SVP over this “digital” team was this guy David. He was a gay male in his 40s, nice house in Seattle, husband was an accomplished professional, and they had adopted two African kids. The CEO of this company loved David. So right away, there’s a whole belief structure around this guy that he cannot do any wrong — he’s an affluent member of a minority population who adopted children from Africa. Their Christmas cards absolutely slay. They are center on 175 fridges around the world. You gotta love David.
Thing is, David was pulled in 78 different directions on the executive team, so the dude couldn’t manage his way out of a goddamn paper bag. He’d just assign you arbitrary deadlines and tasks, with no rhyme or reason. It was amazing. One time I went to a three-hour dinner in Toronto and all anyone discussed was how off-task he was, but all the while they kept saying “But beautiful family…” as if that made our bitching better. Anyway.
We’d have these meetings all the time where someone would present Google Analytics data from our site, our partners, different destination pages (it was a travel company), etc. Me and this other guy, Matt — who later told a different company he didn’t really know me as part of a reference check, even though we had talked every day for 17 months — we’d sometimes propose ideas based on the data.
99 times out of 100, David would say “Well, I see your point, but I trust my gut. I have access to discussions you guys don’t. I see the data, but we’re going to go a different way.”
Beliefs > data.
Led me to write this post too.
The other funny story about that job and beliefs vs. data was that, every Friday, I used to send out a Google Analytics report to the entire company. The report was inherently dry, so I started putting links and celebrity stuff and photos in there. For a while, people loved it. Eventually people started bitching because, as the reports got read more, people were finding data in them — who visited what, who did what on our site — that contradicted their own control, relevance, and assumptions about the work. So they bitched, my boss shut me down, I made the reports dry again, and then people bitched that the reports weren’t fun anymore. Ha. So I went back to doing “witty” reports. Then I got fired. About a week or two later, I find out that some other kid, a Mormon of 22 years of age, is sending the reports now. I guess on his first report, he just put 750 rows of data together and emailed the entire company. Now everyone is losing it because it’s too much data. “I don’t have time to consume this! I’m too busy!” Etc, etc.
Basically, you can’t win for losing at some joints on this topic.
People frame narratives, and honestly frame themselves, around belief. Student loan forgiveness is but one such issue. You’re not going to change belief with data because people are going to fight tooth and claw against you.
Companies keep trying to shoehorn “data-driven cultures” and send around Slack links about Amazon’s approach, but (a) Amazon is a once-in-three-generations company, and (b) I’d bet you $5000 that, in the early days of Amazon, regularly decisions got made based on Bezos’ gut and nothing else. That’s life.
What else you got on belief vs. data, or student loans?
