Alliterative Affinities
Do Parents Select First Names to Match Surnames?
What’s in a name? For many parents, everything. Consider the widespread popularity of baby name books, with dizzying titles like 100,000+ Baby Names: The most helpful, complete, & up-to-date name book and The Big Book of 60,000 Baby Names. For some parents, the decision is far too important to entrust to anyone besides an expert. In these cases, baby name consultants abound to proffer personalized naming advice, with prices ranging from a few hundred dollars to nearly $30,000 [1]. At the time the linked article was written, the upper reaches of that reported range exceeded 50% of median household income in the United States.
What does all this effort, and in some cases expenditure, yield? The famous pop economics book Freakonomics, by economist-journalist duo Steven Levitt and Stephen Dubner, takes this question to task in its concluding chapter. Levitt and Dubner judge that a child’s name ultimately says far more about that child’s parents than it does about the child itself. A supposedly auspicious name like “Winner” won’t predestine a child for glamour or success, nor will being named “Loser” condemn a child to a life of hardship. This example is drawn from a strange case of a father who literally named one child “Winner” and the other “Loser,” whose life trajectories belied their assigned names.
Levitt and Dubner further argue that, relative to expectations, a parent’s influence on a child is determined largely by the type of person the parent is at the time of the child’s birth, not by how well they master the science of parenting. That is, parents might matter more than parenting (beyond a certain threshold of minimum decency and care).
In the case of names, a name might say less about a person than a person says about a name. In our confusion of the direction of this relationship, we ascribe a sense of status and power to certain names, and parents might be drawn to those names. For example, 2008 and 2009 saw a spike in the number of newborns named Barack [2], plausibly a direct effect of Barack Obama’s ascendancy to the presidency.
Might parents consider alliterative names members of this class of desirable names?
To answer this question, I leverage a 5% sample of the 1930 Census from IPUMS [3]. I know it’s old, but representative, timely sources of personally identifiable information are hard to come by, especially for what many would consider a silly question. I assure you I would be among the first to argue that this is indeed a silly question.
I ask a simple question of these data: if a set of parents has a surname that starts with a given letter (say “Anderson”), are they more likely than other parents to select a first name for their child that begins with the same character (say “Amelia”)?
The answer appears to be yes, unless your surname starts with a J.
We can see this effect in the plot below. The blue bars report the proportion of individuals with a surname that starts with a given character that have a first name that starts with the same character. The bars in gray then report the proportion of individuals with a surname that does not start with a given character that have a first name that starts with the given character.
For example, take a look at the 2 leftmost bars, corresponding to the letter A. Individuals that have surnames that do not start with A have first names that start with A 8.3% of the time, detailed by the gray bar. Comparatively, individuals that have surnames that start with A have first names that start with A 9.0% of the time, reported in the blue bar.
The gray bar can be considered the base rate of names starting with the given first character, and the blue bar is the rate amongst those who could have an alliterative name (because their surname begins with the given character) — I call this the alliteration rate. The alliteration rate exceeds the base rate for each of the letters shown below (I only show the most common 15 characters, comprising 90% of first names), except for J, where the effect is strangely reversed.
Some of the differences between the alliteration and base rates seem trivial. So, how meaningful are these differences? I present these same estimates in a different format below, with 95% confidence intervals for the base and alliteration rates organized by character. Only G and R do not exhibit a significant difference between the alliteration and base rates.
We can make comparisons across characters by normalizing the alliteration rate to the base rate. For example, while the gap between the alliteration rate and the base rate is 0.7 percentage points for A and just 0.3 percentage points for B, both alliteration rates are 8.4% larger than the respective base rates. Thus, an individual with a surname that starts with an A is 8.4% more likely to have a first name that starts with A than is an individual with a surname that does not start with A. If we average these alliterative tendencies across the 15 most common characters, excluding J and weighting by the base rate, we get a mean effect of 7.0%.
The average alliterative effect for individuals of Hispanic origin is 13%, approaching double the population-level average. The average alliterative effect is 4.9% for rural residents and 8.6% for individuals in urban areas. The average effect climbs from roughly 7% for individuals up to 35 years old to over 11% for those 36 and older.
The common practice of a woman taking their husband’s surname in marriages between a man and a woman reduces alliterative tendencies. The alliterative effect for J remains negative and statistically significant for both groups. Only 5 of the remaining 14 alliterative effects are significant for married women, compared to 11 out of 14 for married men.
Alliterative tendencies vary substantially across demographic dimensions. Nevertheless, the tendency to select alliterative names seems rather universal.
That’ll be $30,000, thanks.
References
[1] D. Baer, People are paying up to $29,000 to have a baby name consultant name their child (2016), Business Insider
[2] A. S. Gibbs, Growing Up Barack: Meet Three Boys Named After President Obama (2017), NBC
[3] S. Ruggles, S. Flood, S. Foster, R. Goeken, J. Pacas, M. Schouweiler and M. Sobek, 1930 5% sample (2021), IPUMS
