Sunday, 3 June 2018

What's in a (first) name?

Following on from yesterday's post on middle initials, it is interesting to wonder how important names are more generally. In a 2010 paper published in the journal Economic Inquiry (ungated and much earlier version here), Saku Aura (University of Missouri) and Gregory Hess (Claremont McKenna College) looked at the effect of first names on a number of different outcomes. They used data on over 5500 people from the 1994 and 2002 waves of the U.S. General Social Survey (and as a reminder, apart from the names data, most of the data for the GSS is available for free online).

Unlike the other studies I've been referring to this week (see here and here and here), the Aura and Hess paper wasn't about academic outcomes, but instead about broader life outcomes. It is related to earlier work on name-based discrimination (ungated here) by Marianne Bertrand and Sendhil Mullainathan. Bertrand and Mullainathan used a field experiment (they sent out CVs with African American and non-African-American names and compared the number that were invited to interviews), whereas Aura and Hess's paper is based on survey data. Aura and Hess also look at a broader range of features of names, rather than just how much more likely they are to be names of African Americans.

Aura and Hess find that:
...more popular names are generally associated with better lifetime outcomes: that is, more education, occupational prestige and income, and a reduced likelihood of having a child before 25. Also, broadly speaking, names starting with vowels and ending in either an ‘‘ah’’ or ‘‘oh’’ sound are related to poorer lifetime outcomes.
Interestingly though, when they look at male and female names separately, the effects of names are not apparent for males, only for females (and even then, only for some of the outcome variables). Some of the problem there is the reduced statistical power from splitting the sample into males and females, but I think there's a robustness issue there as well.

Aura and Hess conclude that their research doesn't support the discrimination argument (although, it doesn't not support it either), because the features of names are also correlated with other person-specific variables such as race and age. However, I'd argue that's exactly what we would expect if discrimination did explain the results.

Anyway, as with yesterday's post on middle initials, if we want to look at the effects of names on academic outcomes such as the number of citations or other measures of research quality, we could easily use a similar method to those in the previous papers I blogged about this week (see here and here). That might be an interesting exercise, alongside looking at the effect of middle initials.

So, to summarise: what have we learned this week about getting research published in top journals and cited a lot? It's possibly better for research papers to have a short title, and to use data on the U.S. Maybe middle initials matter (still an open question in terms of research quality), and maybe first name could matter (it seems there is no specific research on this in an academic context, that I can find). And, thinking back to a much earlier post of mine, having a surname early in the alphabet is a good idea if you have academic career aspirations.

Read more:

No comments:

Post a Comment