The Trait Advantage (Human Capital, pt. 2)
Skills might be depreciating at a higher rate, but not all human capital is the same
My previous article concluded with several maxims, or first principles, that can be used to understand knowledge, skills, abilities, and other characteristics (KSAOs) as a type of capital. Let’s explore this one:
Like other forms of capital, KSAOs have a degree of durability, but they can appreciate or (more commonly) depreciate in value.
AI is decreasing the “half-life” of skills, meaning they are becoming obsolete more quickly on average. At least, that is a claim being thrown around here and there. I would take these claims with a grain of salt, because most of the supporting data comes from cross-sectional surveys of leaders about their opinions which is a really indirect and potentially fraught method.
That being said, the depreciation of human capital is not a new idea. Higher quality research does show a kind of “use it or lose it” dynamic, which is especially bad for unemployed people, who have their skills depreciate (and wages decrease) even faster. For example, a recent study estimated a 4.3% rate of depreciation for skills among unemployed teachers in Greece. You could argue that in emerging technology fields like AI, we would expect to see a much higher rate of skill depreciation.
Still, the difficulty of measuring skills and skill depreciation (not to mention the conceptual confusion about KSAOs) historically has muddled our thinking about how to measure and manage talent, even relating to our own careers. Thanks to the big-data revolution however, the open-web is now an ocean of jobs data. In this article I unpack some of the recent research on skill depreciation, and its implications. The punchline is that some types of human capital, like traits, are far more durable than others (i.e. technical skills). Of course durability is not the end-all-be-all, but it might affect how we look at, and value human capital into the future.
STEM skills: a fleeting early career advantage?
It’s common knowledge that young professionals who graduate with STEM majors out-earn their counterparts. But how durable is their skills advantage? This question framed an amazing study by David Deming and Kadeem Noray, economists at Harvard, where the researchers used labor market data from Burning Glass to model the process of skill depreciation.
In the study, they characterized the (income) returns to work experience as a “race between on-the-job-learning and obsolescence.” Unsurprisingly, they observed higher rates of skill depreciation (here referred to as ‘turnover of skill requirements’) in technology occupations.
This skill depreciation seems to affect both the earnings and the behavior of workers. Taken from the abstract:
We show that the earnings premium for college graduates majoring in technology-intensive subjects such as Computer Science, Engineering and Business declines rapidly, and that these graduates sort out of faster-changing occupations as they gain experience…
The authors also modeled a fascinating and counterintuitive dynamic, where “higher-ability” workers may actually be incentivized to move out of occupations that have a higher rate of skill turnover. This is because their advantage - the ability to learn new skills quickly - is actually less valuable (on the margin) in fields like computer science, because those skills are going to depreciate at a higher rate.
Psychological traits are pretty durable
Speaking of ability, one really convenient set of data on human capital depreciation comes from the research on cognitive ability and aging. Cognitive ability, which typically rates among the highest predictors of job performance, is remarkably stable across age, at least through traditional working age years (25 - 67). Yes, fluid intelligence (including numeric ability and perceptual speed) peaks in the early 20s and declines thereafter, but cognitive ability is a a foundational aspect of most adults’ human capital that can be relied on to be more or less durable.
Similarly, personality traits, also among the top predictors of job performance, are another foundational aspect of human capital that remain remarkably stable across age. One reason is that personality traits are around 30-60% heritable, as estimated by twin and adoption studies, meaning that somewhere between 30-60% of the variance in personality between people can be attributed to genes.
Of course, personality changes over time for many people, but its “plasticity” seems to be higher for the adolescent and the elderly, according to the research. What’s more, while our personalities can change in response to experiences such as major life events, there is a good deal of evidence that on average, traits actually change towards a more “mature” profile - people tend to get more emotionally stable, more conscientious, and more agreeable as they age. Depending on the job in question, this may be a case of human capital appreciation!
Given the durability of these psychological traits, we could think of them as existing lower in the human capital “stack” - they are the psychological hardware to technical skills’ software. If you want to maximize your human capital in the short-run, you should brush up on your programming ability, but in the long-run, well let’s just hope you are fortunate to be intelligent, conscientious, and emotionally stable!
Trait spillovers / externalities
It’s important to not view human capital in a vacuum though. Borrowing from economics (as I love to do), the idea of externalities applies here: when organizations rent access to your human capital, there are a variety of potential unintended consequences. Research on work-life “spillovers” are an example of externalities at play in the workplace.
Externalities don’t have to be negative however: a positive side effect of hiring someone with the right traits is that their qualities don’t just affect their performance on work tasks, they also impact others around them. Organizational citizenship behavior, a jargony way to say “going above and beyond work tasks to make an organization a better place”, is strongly influenced by personality, with some research finding that up to 17% of the variance in this behavior can be explained by personality traits.
Lastly, a key feature distinguishing human capital from other forms of capital is that it is contained in individuals who can voluntarily terminate the relationship with their employer at any time. In this way, traits can have (often unintended / inadvertent) impacts on human capital durability, viewed from an organization’s perspective. Meta-analyses of turnover predictors show how traits like emotional stability are related to a lower chance of voluntary turnover, for example.
Caveats: The Difference between Traits and Skills
I don’t want any of this to be misinterpreted as a bias towards traits or against skills, so to wrap up I will provide an important point of clarification: While skills may depreciate faster, they are still an essential building block of the human capital “stack”. They may be more pragmatically useful as a means for recommending, matching, and screening job candidates for various roles, particularly at the individual contributor level (which represents the vast majority of human capital).
Instead, the point of this article is to simply highlight a key advantage of traits, which is that they tend to be highly durable. This advantage makes traits ideally suited to weigh heavily in decisions about longer-term talent identification and investment.
But just knowing that traits are durable will not be enough. How should we think about the measurement, valuation, and investment in human capital, given its heterogeneous durability? To find out, stay tuned for the next article in this series: Human Capital, pt. 3!