Caustic but not necessarily wrong

From Does higher education change non-cognitive skills? by Tyler Cowen.

The original research which sparks his post finds that:

We examine the effect of university education on students’ non-cognitive skills (NCS) using high-quality Australian longitudinal data. To isolate the skill-building effects of tertiary education, we follow the education decisions and NCS—proxied by the Big Five personality traits—of 575 adolescents over eight years. Estimating a standard skill production function, we demonstrate a robust positive relationship between university education and extraversion, and agreeableness for students from disadvantaged backgrounds. The effects are likely to operate through exposure to university life rather than through degree-specific curricula or university-specific teaching quality. As extraversion and agreeableness are associated with socially beneficial behaviours, we propose that university education may have important non-market returns.

I am dubious. Any prolonged intensive experience is likely to change you. That much I am inclined to accept.

Four years of college, or three in the case of Britain, might qualify if you come from a significantly variant background, i.e. lower class or impoverished.

Without having read the gated version, this has the feel for one of those studies in which the controls are so flawed that it will never replicate. But it raises interesting considerations.

My very first thought was that they are talking about three quite different phenomena. There is domain knowledge - the facts and figures which you learn through study. There has always been debate about just exactly how much of the future value of a college education is due to the domain knowledge acquired. Certainly some of the future value is due to the increased domain knowledge, but it might not even be the majority of the value.

Domain knowledge is different from personality which is what the Big Five measures. It is an article of some reliability that personality, past a certain age, is relatively stable and homeostatic. The five attributes are openness, conscientiousness, extroversion, agreeableness, and neuroticism. What the researchers appear to claim is that success in the academy for those from disadvantaged backgrounds is associated with agreeableness and openness. It is unclear whether they are claiming that it increases agreeableness and openness.

If they are only claiming that there is an association, then that could easily be simply a product of survivorship bias. If they are claiming that university increases agreeableness and openness among those from disadvantaged backgrounds, then that flies in the face of other research indicating the stability of personality attributes.

Domain knowledge and personality are different from social norms and behavior. I am perfectly prepared to believe that bright students from disadvantaged backgrounds are capable of learning the social norms and accepted behaviors of the educated elite. Use of our cognitive capabilities to read and understand our physical and social environment and adapt our actions and behaviors accordingly is one of the distinctive characteristics of the human animal. We are evolved (or at least most of us) to do that well.

Read from that perspective, the researchers take on a very colonial mien, becoming the modern day Professor Henry Higgins. "See, we can take these barbarians from the lower class and make them like us."

The tell is in this line from the abstract: "Extraversion and agreeableness are associated with socially beneficial behaviours." True to a limited extent. If you are an ideological adherent of social justice, can't-we-all-get-along mushy thinking is acceptable as a proxy for logic, reason and evidence. But it doesn't pass muster for science; hence the replication crisis.

If your goal is to get along with others, then the researchers are to some degree correct, that is facilitated by a personality high in agreeableness. Perhaps openness as well, though that is much more dependent on context.

But if you goal is for the individual to be successful, there is a negative correlation between success and agreeableness. Research on business executives indicates that the most successful are characterized by high levels of conscientiousness; moderate to low levels of agreeableness; and low levels of neuroticism. Openness and extroversion are, I am pretty certain I recall correctly, not correlated with success one way or the other.

Executives work in environments of uncertainty (so the negative correlation with neuroticism makes sense), high expectations (correlating with high conscientiousness), and constraints (correlating with a negative relationship to agreeableness). The latter perhaps requires elaboration.

In an environment of constrained resources, the most powerful weapon of any executive is the decisive "No!" No, we won't make that questionable investment. No, we won't increase our relatively fixed cost structure by granting across the board raises. No, you are not ready for that promotion. The best executives use no as a means of stripping away the infinity of possibilities to get at a few things that they might be able to do well.

Wielding no as the default basis for social interaction does not fit naturally with a personality that is high in agreeableness. "No" comes easiest to those who are less susceptible to pleading and who are not emotionally concerned about whether others are upset. Executives are low in agreeableness because it is a big part of their job - they have to disappoint people. Politicians on the other hand, while generally sociopathic, also tend to be high in agreeableness because "No" is not in their vocabulary. They hate disappointing people. They want people to like them and to never disappoint them.

The researchers in this instance are allowing their world view to inform, and impose, an unstated assumption about success. They are defining success as getting along with others. They assume that someone from a disadvantaged background should want to get along with others. Therefore, if higher education increases agreeableness and openness (which is what they have claimed but not proven) then that is a good thing.

But if I am from a disadvantaged background, it is quite possible that I am not willing to be the Eliza Doolittle to their Professor Higgins. Perhaps I have agency and define success not as social acceptance but as social power. Perhaps I wish to achieve standing and income. In which case I need to increase my conscientiousness, reduce my sensitive agreeableness, and dispense with neuroticism. I can cultivate openness and extroversion to whatever degree I wish, it doesn't make a difference.

What the academics seem to be saying, if you accept their findings which are likely to be unreplicated, is that they think it is feasible and a good thing to colonize the disadvantaged and prepare them for a life in which they are liked but unaccomplished. At least, that is a caustic read between the lines. Caustic but not necessarily wrong.