Researchers have found links between leaders’ narcissism, worse reviews of their leadership behaviour and a surprising third factor: how much money their families had when they were growing up.

The study, published online by the Academy of Management Journal and recently highlighted by the Harvard Business Review, claims to be one of the first to explore a tie, if an indirect one, between leaders’ family income during childhood and their leadership behaviours later in life. At a time of increasing income inequality, says Sean Martin, an assistant professor at Boston College, it’s also part of a growing body of research that’s studying the effects of higher incomes on social behaviour.

“We’re trying to understand what it means for how people treat each other,” Martin said in an interview. How does income inequality “relate to the decisions people make in business and in leader-follower relationships? What are the effects of this growing socio-economic trend?”

Martin and his coauthors studied 229 graduates of the US Military Academy who now hold Army leadership positions, which allowed things like college education, current income and the leaders’ hierarchical level to remain standard across the study. They gathered information about the leaders’ parents’ income from their applications to the school. They then asked them to fill out surveys that assessed their narcissism, and a larger group of their subordinates to answer questions about their performance on a variety of leadership traits, such as showing care for others or pushing for innovation and change.

What they found was an indirect relationship between the three. The study showed a strong correlation between the leaders’ parents’ income and narcissism, as well as a significant link between that narcissism and reviews of the leaders’ behaviour. He cautions, though, that what they found “is not causal — it’s a step-by-step chain”, Martin said. And even among the relationships they did see, they’re just saying that higher incomes were associated with more narcissism. “Are there exceptions to the rule? Of course there are.”

It’s not the only study that has attempted to look at leaders and their social class backgrounds. A study published late last year, for instance, found that CEOs from middle-class backgrounds tend to take fewer strategic risks than those who hail from wealthy or poor backgrounds.

Of course, Martin’s study raises many questions, too. Would he see the same results in non-military organisations, which tend to be less hierarchical? How well would their results apply to other workplaces? Martin says there’s been other research that ties income and narcissism, so that relationship would likely hold up. But he admits that the other link, between narcissism and leadership performance, might be different in places that aren’t as team-driven or service-oriented as the military.

In addition, he acknowledges that other factors they couldn’t control for — such as parents’ personality, or the influence of different kinds of lower-level education — could help explain their findings. Martin also says their relatively modest sample size doesn’t include anyone from the super-rich end of the spectrum (their highest group had family incomes of about $275,000), which could effect the results: “It might be that when you get up to several million or more you see that there is more of that noblesse oblige.”

Finally, he’s quick to note that their findings shouldn’t be used as an excuse to start examining family income when it comes to choosing leaders. Rather, he says, leaders should “be careful about rewarding the people who look really flashy,” he says, and “reward people’s tendencies to help and serve and treat other people well. Make that part of your evaluation process”.

Of course, we couldn’t help but ask Martin how he thought his findings might apply to Donald Trump — the GOP presumptive nominee whose father was a wealthy real estate developer himself and who certainly thinks quite highly of himself. In The Atlantic’s recent cover story, “The Mind of Donald Trump,” Northwestern University psychologist Dan McAdams writes that “for psychologists, it is almost impossible to talk about Donald Trump without using the word narcissism”.

But Martin wouldn’t bite. “We absolutely do not intend and do not want to be making any political statement or have any political orientation at all,” he said, noting they started gathering data in 2014, well before Trump announced his candidacy. As they approached publishing, he said, “my colleagues and I said we knew questions about this were going to come. ... But it happened to be a thing of timing more than anything else.”