Signaling theory and portfolios
Education is a tragedy of the commons. What makes a good portfolio
Hi, I’m Russell. This is issue #4 of Russell’s Index, where I write about the lessons I’ve learned—and continue to learn—as a founding employee at SharpestMinds. Subscribe for a new issue every week(ish).
Signaling theory continued
In last week’s Index, I talked about the signaling theory of education. [1] Higher education leads to higher wages, not because it teaches employable skills, but because it verifies employable traits like intelligence, conscientiousness, and conformity. Given the high cost of education (both in money and time), the signaling theory is a real bummer. The folks who can’t access a “good education” are left behind.
But it’s hard to place blame. It’s a tragedy of the commons-type situation. Everyone is behaving rationally, but the net result is still bad for society. Hiring takes a lot of resources—employers can’t interview everyone. Some filtering upfront based on implicit and explicit signals is necessary.
Given a choice between two candidates—one with a university degree and one without—the candidate with the degree is the less risky choice. Even with the knowledge that the degree is just a signal, it’s still rational to rely on it. The candidate with the degree isn’t guaranteed to be more qualified, but they are more likely to be. They are a safer bet. Brian Caplan sums this up well: [1]
You can’t discover a person’s true work ethic with a glance. You certainly can’t ask, “How good is your work ethic?” and expect candor… A signal doesn’t have to be definitive, just better than nothing.
And, because employers still rely on it so much, getting a degree is still a rational decision for job-seekers. Even Marc Andreeson, who is very critical of America’s education system (he calls it an oligopoly and a cartel), advises the younger generation to pursue a degree anyway. [2] It’s still the most reliable way to boost your future earnings.
Is there a better signal?
There is a better signal—actual work experience. If you’ve already worked elsewhere in a similar role, it’s a strong indicator that you know what you are doing. Mentees at SharpestMinds regularly report that the job search becomes significantly easier when they are looking for their second role. It’s the first one that’s the hurdle.
Technology, however, offers a workaround. For certain types of skills, you can gain experience without getting hired. I made this argument recently in a blog post for Towards Data Science:
For many professions, you need to land a job to start accumulating [work experierence]. But there is a wonderful difference when it comes to data science and machine learning—you can get plenty of experience before getting hired. How? By building things.
I was talking about data science, but this applies to any skill where proof of work is easy to show like programming, design, and writing. All you need is access to the internet and time to learn and build. [3] The web is full of accessible educational resources—you can take MIT courses for free from your living room, for example. Much of the challenge of self-learning comes not from a lack of resources but from filtering the signal from the noise (something a mentor can help with 🙂).
There is a shift happening in the tech industry. An impressive portfolio can be just as good as—if not better than—a university degree as a signal for vetting candidates. Part of this is due to an ongoing culture shift. Folks are starting to recognize the diminishing value-add of a degree where most of the knowledge can be found for free on the internet. [4] Part of it is due to the pace of technological change accelerating faster than university curricula can keep up. The supply of people with degrees in machine learning, for example, is less than the demand for machine learning practitioners. Employers are being forced to use alternative signals—like portfolios.
What makes a good portfolio and other signals
I’ve long been an advocate for putting less weight on formal education and more on portfolios. But recently, while hiring for a technical role, I had to evaluate many candidates. It forced me to think a bit deeper about what kind of signals to look for.
I’ll be honest, almost all the candidates with a CS degree got past the resume screen. Despite being critical of the signaling theory of education, I still found it to be a useful signal. But, I tried to put more weight on portfolios—an impressive hobby project would also get a candidate past the resume screen.
Getting accepted to university signals a certain amount of intelligence, but finishing the degree signals an ability to persevere and see things through. A desirable trait for employees. A good portfolio can also signal this. I found that a portfolio with a single completed project to be more impressive than a smorgasbord of small, or half-finished projects. [5] It shows that you can finish the things you start—that you can put sustained effort into a single project and think long-term. [6]
Another good signal, which should be obvious, is a genuine interest in the role. Some folks emailed me their resumes along with a paragraph or two about why they wanted to work for SharpestMinds. Some folks just emailed me their resume, without much more context. The former made a much better impression. Good writing, attention to detail, enthusiasm, and good manners (send a thank you email!) are some other good signals that can add up.
[1] For further reading on the signaling theory of education, check out Bryan Caplan’s The Case Against Education.
[2] For more of Andreeson’s thoughts on education, check out this podcast episode from a16z:
[3] This is still unfair—not everyone has access to a computer with internet access and the time to self-learn. But it’s still much more egalitarian than academic credentialism.
[4] I’m sure the pandemic is accelerating this trend. Why pay tuition for a shoddy education over Zoom if you can get a better virtual education with world-class teachers elsewhere on the internet (for considerably less money).
[5] At SharpestMinds, we encourage our mentees to build end-to-end ML projects. But there are other things that can signal follow-through—a blog that has been maintained for a long time, or a side hustle that has been kept alive for a while.
[6] Patrick Collison touches on this when talking about what Stripe looks for when hiring: “People who really like gettings things finished”.
Thanks for reading! Subscribe for a new post every week(ish). If you have a comment or want to chat, replies to this newsletter will go directly to my inbox. Or follow me on Twitter @russ_poll.