Timothy Judge, Beth Livingston, and Charlice Hurst investigated the factors influencing pay using three large data sets, each containing data on between 500 and 2000 adults. Each collected personality information using slightly different measures; as each study corroborated the other, I treat this difference as a strength of the paper, as convergent evidence from multiple measures precludes the possibility that the instruments used were generating funky results. In each case, men tended to earn more than women (in one data set this was made explicit as approximately $5,000 less per year), and disagreeable people earned more. However, this mean premium was mostly due to the wage benefit that men received; for women, the premium was much slighter. Across studies, agreeableness made a big difference to male incomes, and a minor one to female ones.
So what makes the mean premium possible? It's not entirely clear - the study investigated some possible reasons such as that disagreeable people find their way into higher status or more complex jobs, but the data didn't support that conclusion. They did find that disagreeable people place more importance on pay and less on communal relationships than others, which sheds light on their priorities but not on how these are achieved. Some possibilities include agreeable people being more prepared to compromise and concede, for example on pay negotiation, or for decision-makers to falsely place warmth and competence as two ends of a continuum (rather than independent factors) and conclude that a people person may be less capable. Highly agreeable men would not only lack those edges, but, as Judge and colleagues point out, are doubly disadvantaged, as they are operating against gender stereotypes by being a soft male. Conversely, the edge that a disagreeable woman gains is blunted by their operating in ways that aren't socially sanctioned. The authors reflect that "exhortations for women not to be nice...might be overblown", and the solutions to gender pay inequality lie foremost with decision-makers.
Perhaps the causality is reversed – well-paid jobs make people less agreeable? There are a few points against this: firstly, the failure to find linkages between job type (such as status) and agreeableness. More convincingly, the investigators ran an additional, experimental study, where 480 student participants made choices in an imaginary scenario as to who they would recommend for a management fast-tract. The pair of candidates only differed according to keywords inserted in the text that speak to the quality of agreeableness, such as modest/immodest. The same pattern emerged – more recommendations for disagreeable people, with a much stronger effect for men than for women. Taken on its own, this suggests that disagreeableness is driving job outcomes, rather than the reverse.
So do nice guy finish poorly, and women last? Well, it depends what matters to you. Judge's third study found that agreeableness - and to a lesser extent, being a woman - was positively associated with life satisfaction, stronger social networks, and community involvement, and negatively associated with stress. Essentially, the disagreeable-man priorities are having exactly the impact you would expect; as the authors conclude, "if disagreeable men win the earnings war, it is a victory that may come at some cost."