Lundergan, Carter, Nunn: What's in a Name?

In the South, 2014 is a "name" election. The trick to telling is by counting how many candidates are introduced Viking-style, as "[name], son/daughter of famous state politician". And especially in the South, there are a quite a few candidates who, at least early in their candidacies, got this treatment:

• "Alison Lundergan Grimes, daughter of former Kentucky Democratic Party chairman Jerry Lundergan...."
• "Michelle Nunn, daughter of former Senator Sam Nunn...."
• "Jason Carter, grandson of the former president and Georgia governor...."

Don't be surprised, by the way, if the candidates' actual qualifications are tacked on as an afterthought: "Oh, by the way, Grimes is also Secretary of State of Kentucky. Shoulda mentioned that."

So there are quite a few famous sons and daughters in the running this year. What we wanted to know, though, was this: what does a surname do for a candidacy? We had a few theories: a well-known political relative bolsters name recognition and gives a candidate a boost in the primaries as well as in the general election. A well-liked name can also attract donors who remember the name from the first time around. For example, former Florida Gov. Jeb Bush, as Harry Enten at FiveThirtyEight writes, "is affiliated with a brand Republicans have voted for and still like," and is thus still a pretty good candidate for 2016. We suspect that brand was helpful for him back when he was first elected governor in 1998, as well.

What we decided to do was simply determine whether candidates with famous names perform better in statewide elections. To do this, we drew a random sample of 200 elected senators present in the Senate any time from 1947 to the present. We separated this sample into two groups, which we labeled "Dynastic" and "Nondynastic". As you can probably guess, group Dynastic included all of the senators in the sample who had famous relatives in politics. We defined "famous relatives" as being grandparents, direct descendants of grandparents (including parents, siblings, cousins, aunts, and uncles), or spouses who had previously been one of the following:

• Governor of the candidate's state
• Senator from the candidate's state
• At-large U.S. representative from the candidate's state
• State party chair
• Leader of a house of the candidate's state legislature
• A major-party nominee for President or Vice President
• Cabinet secretary
• Ron Paul

This definition is based on our theory that the major benefit of having a famous relative is name recognition for a statewide election. We've excluded from the definition state legislators, senators and governors from other states, and U.S. Representatives if the state has multiple districts because they haven't necessarily made themselves well-known throughout the state. We've also excluded anyone more distantly related than grandparents because if you're a candidate running on the name of your great-grandfather, most of the people who were around to know your great-grandfather are too dead to vote for you, and it's a bit of a stretch to run as "the fifth cousin of this other guy you like" (even though it worked for FDR). We found lists of family members on this cool Wikipedia page listing all of the political families in the U.S.

After categorizing each senator as either "Dynastic" or "Nondynastic", we determined the margin of victory in each of their first statewide election victories. For most this was the election that put them in the Senate, but for some, like Sen. Jim Jeffords (I-VT), that first election was to an at-large House district, which covers the whole state. There were also a bunch of governors and former governors who were elected to the Senate after they were term-limited: their margin of victory in their first gubernatorial election is used instead of their Senate margin of victory.

We then calculated the mean margin of victory for senators in Dynastic and senators in Nondynastic and calculated a t-score and a P-value for the difference of the two means. Here are the results, for the statheads out there:

The probability of getting these results from a random sample if Nondynastic senators perform as well as Dynastic senators is a minuscule 0.00127. We conclude that, at least historically, senators with famous family names performed better in their first state-wide elections than candidates who didn't have this advantage.

The bottom line

We're not suggesting that Grimes, Carter, and Nunn are going to win by large margins in November. After all, the sample we used has drawbacks--by its nature it only includes people who have won Senate elections already, while this year's crop of Southern Democrats obviously have not. It also takes the average of all the senators' margins of victory without regard to changing conditions from year to year or which state each senator represented. In the end, however, we hoped that most of the errors would just balance out--there have been great years and terrible years for both parties, for example, and a random sample should in theory draw from all of them equally. Most importantly, however, it doesn't mean that a good name can overcome poor campaigning or lackluster polling.

What matters, however, is that this does support our theory that a famous name helps a candidacy. We don't know exactly where it comes in--it could be that it establishes viability in the primary, makes fundraising easier, or gives a candidate broader appeal in the general election--but it seems to help. For Grimes, Carter, and Nunn, their last names seem to have coalesced party support behind them very quickly. The advantage of this, of course, is that in a race like Nunn's in Georgia, she can raise more money and advertise at her leisure, while sniping at a Republican pack still busy fighting with itself. This was an advantage which President Obama had in 2012 and which Sen. John McCain had hoped to exploit in 2008.

But perhaps the best thing about having a name like Carter or Nunn is just the appearance of viability it generates. Whether or not you believe that famous names just perform better in elections, as our data suggest, they do inspire confidence in donors and enthusiasm in the base. And those, as much as anything, are key to winning close elections.