Why are candidates expected to win their home state?

Question

In coverage of primaries and elections I often hear it stated that candidates are expected to win their home states, or that it is embarrassing not to. I've seen this assertion here, such as in Has a presidential candidate ever lost his home state and went on to win the presidential election?, and from the data I can find it does appear that candidates regularly do far better in their home state than in others. The only explanation I've heard for this effect is that home state voters are 'the people that know the candidate best'. This explanation doesn't sit right with me.

I have a few friends who are very seriously into politics. These friends read political news and do political research daily. They know all of the candidates and their policies and their histories intimately, and have no better understanding of candidates hailing from their own states. I don't see a reason why these people would be more likely to vote for someone just because that someone was geographically close to them. I polled a few of them on their historical votes and they did not seem any more likely to vote for a home state candidate.

I did a little informal poll myself over the past few days, talking to neighbors and cashiers while out. It's not a statistically valid sample size, and it was certainly biased sampling, but only 6 out of 11 knew who our governor was (it's Tom Wolf). Only 1 out of 11 could name a single thing Tom Wolf has done or said in the past 12 months. I personally haven't so much as overheard his name in months*. So it seems like 10 of 11 of my random pollees would have no reason to vote for their governor over another candidate were he running for office.

So what does actually drive the increased performance of a candidate in their own state. Is it simple name recognition? Do candidates simply tend to spend far more time and money advertising and rallying in their home states? Are those I've spoken to an anomaly, and most people are familiar with their home state politicians? Or is there some psychology that I'm not seeing that's at play?

• To be clear, I'm not saying that he hasn't done anything of consequence; just that it seems outside the scope of the average person's knowledge.

Answer 1

In their article, Localism in Presidential Elections: The Home State Advantage [1] published in the American Journal of Political Science, Lewis-Beck & Rice (1983) investigate the home state advantage using quantitative evidence, and try to explain the phenomenon. They discuss the fact that of all the public offices, the presidency is by far the office with the most public exposure. This is even truer currently, with the vast prevalence of TV & online advertising. They go on to discuss why this makes a difference:

Each of us learns the home states of presidential candidates, and we are particularly aware of the fact if the state is ours. This piece of candidate information is repeatedly reinforced through the radio and television network that now embraces our nation. It gives us a chance to show "pride in our own" by voting for a native son. Such local loyalty is not wholly unreasonable. We are offered the psychological satisfaction of identification with a president who is more like our "friends and neighbors." Further, we might hope that as president he would remember "the folks back home" when distributing federal largess.

They identify three variables that affect the level of the home state advantage a presidential candidate can expect, these being the population of the state as a proportion of the national population, the political party, and whether or not the candidate is the incumbent president seeking reelection, and the article even proposes a formula that can be used to determine the level of advantage numerically.

Briefly, however, they expect the level of home-state advantage to decrease as state-population increases, due to the increased strength of local bonds in tighter-knit communities. They find that Democratic candidates can expect a higher degree of home-state loyalty than Republicans, due to lower primary turnout of Democrat voters, which allows a greater opportunity to motivate traditional non-voters to come out to vote for a home-candidate. Finally, due to a "ceiling effect", the incumbent President, who would already expect a high degree of support, dampens the visibility of the home-state effect.

In their conclusions, they found that the home-state advantage could be relied upon to provide the candidate with an increase of the vote share of 4% compared to what they would otherwise expect to obtain, and that this level of support had not changed since 1900, despite increased nationalisation of the electoral process. While these statistics need to be qualified with the age of the article, I expect many of the same overall causes of the home-state effect are still in place today. With all this extra support, then, it seems that any serious candidate for the presidency should be expected to win their home-state, hence why losing it is so embarrassing.

• [1] Lewis-Beck, M. S., & Rice, T. W. (1983). Localism in Presidential Elections: The Home State Advantage. American Journal of Political Science, 27(3), 548. doi:10.2307/2110984

Answer 2

I'd like to add an additional possible factor to CDJB's excellent answer.

Assuming that the candidate actually holds (or held) office in their home state (such as being a current or former representative, senator or governor), they must have already been elected there at least once. They are untested in other electoral markets, with no a priori reason to expect particularly good or poor performance there. But they must be popular enough in their home state to have been elected previously.

Imagine we have four candidates:

• Alice, an Iowan who is considered appealing by the standards of the Iowan electorate.
• Bob, an Iowan who is considered unappealing by the standards of the Iowan electorate.
• Carol, a Texan who is considered appealing by the standards of the Iowan electorate.
• David, a Texan who is considered unappealing by the standards of the Iowan electorate.

Alice, Carol and David all stand possible chances of being elected to lower office in their respective states - Alice is popular in Iowa, and Carol and David may well both be popular in Texas (Texan voters may care about different things than Iowan voters).

Come the Iowa caucuses, Alice, Carol and David might participate. But Bob won't - his political career never made it off the ground. Therefore, the candidates in Iowa will include a single Iowan who is popular in Iowa, and two Texans, only one of whom is popular in Iowa.

I do not know how large this pre-selection effect is compared to the more psychological phenomenon one described by CDJB, but this thought experiment does demonstrate that candidates can be expected to perform above their national average in their home state, even if the voters are perfectly rational and not swayed by home state bias at all.

Answer 3

I have a few friends who are very seriously into politics. ...

These are not the important voters. There is probably only a small minority of people who are really into politics, know what each candidate exactly stands for and how they held up to their promises in the past. This is a utopian view of democracy that does not stand reality.
Most people do not invest much time into politics. They do not read any party's program but simply take what other people (journalists, bloggers) write about it. Sometimes the political opinion may be heavily influenced by what friends think about or by social media.

One important aspect of a politician that is easy to grasp is their origin. This is not about knowing candidates well, it is about having the same background. A politician who grew up in downtown Manhattan will not know first-hand how live in a remote mountain village is (and vice versa). A politician coming from the same background is more likely to have the same views and understand how to solve the problems.
Add this to the self-selection effect mentioned by Drubbels and you can get some more votes than another candidate

Answer 4

So it seems like 10 of 11 of my random pollees would have no reason to vote for their governor over another candidate were he running for office.

You say 10 out of 11 of your random pollees have no reason to vote for your governor, but they don't have any reason to vote for any of the others either. They aren't interested enough in politics to have any political reason to choose anyone in particular. They'll therefore base their decision on non-political factors.

What differentiates the candidates in your pollees eyes? It might be that home-state vs not-home-state is the only differentiator they're aware of. Which makes them far more likely to vote for the home-state candidate.