Is there a tendency for female and minority candidates to attract female and minority challengers, respectively?

Question

Not so much this election cycle, but in the past I feel like when there's a female or minority candidate, the challenger is often female or minority, respectively. Is this a known phenomenon?

Answer 1

There have been a number of studies, and the short answer is: "Evidence is inconclusive, but suggestive that the effect is only minor and incidental."

Challenges to Study

There are reasons to believe that ethnicity might effect vote choice, but there are a number of confounding variables which are difficult to control for:

1. covariance with ethnicity/race and political party (i.e. people of the same ethnic or racial groups may join the same political groups for non-ethnic or racial reasons)
2. covariance with ethnicity/race and socio-economic attainment (i.e. people of the same ethnic or racial groups may have similar incomes, and in some cases socio-economic attainment affects racial or ethnic identification as in the U.S. where as income rises, families who historically identified as latino begin to identify as white)
3. covariance with ethnicity/race and education (i.e. people of the same ethnic or racial groups may be of the same educational groups, either due to opportunity or prejudice)
4. geographic distribution (i.e. many ethnic groups group together geographically, in some cases in a 'historic homeland' such as the Kurds in Turkey, or Basque in France/Spain)
5. historic path determinism (i.e. whatever has led to the support of certain candidates affects the future choice of candidates)
6. Idiosyncratic issues for certain ethnic or racial groups which affect outcomes (e.g. Blacks in the U.S. have consistently been at a disadvantage where as other groups who experienced bias eventually wheras other historically biased groups, such as Germans, Irish, Italians and East Europeans and Asians all in turn rose above their disadvantaged status)

Most studies attempt to circumvent this by using randomized experimentation, usually comparing reactions, preferences with identical fictional candidates, varying only race/ethnicity/gender. The advantage of this practice eliminates other factors mentioned above. Unfortunately, it may not be reflected in actual voting behavior. Nonetheless, this is the best method available to Political Science at this time, for this particular question.

Race/Ethnicity

Three recent articles address the issue across different countries (as the question implies it should).

"What Do Voters in Ukraine Want?: A Survey Experiment on Candidate Ethnicity, Language, and Policy Orientation" (Published version gated; Ungated draft version) finds that policy trumps ethnicity and language.

"Candidate Ethnicity and Vote Choice in Britain" argues for a much more nuanced view, where anti-Muslim sentiment hurts Muslim candidates amongst whites, and helps Pakistani Muslims with their co-ethnics.

"Evaluating the Roles of Ethnicity and Performance in African Elections" (gated) contends that co-ethnics support one another, but not in a strictly 'election market' model, but through the formation of voting blocs.

Bottom Line

There is some evidence that ethnicity matters sometimes, but it seems that it matters in specific circumstances and for specific reasons. The study of British elections was certainly affected by the 7/7 attacks which has created an anti-Muslim bias, and a pro-Pakistani backlash within the Pakistani community. Similarly, in Africa many countries have no non-ethnic civil society, so of course ethnicity will matter. How well that generalizes to the other areas of the world, remains to be seen.

Gender

"Choice sets, gender, and candidate choice in Brazil (gated)" contends that, even though women are under represented in elections throughout Latin America--and Latin America has a reputation as being the home of "machismo" there is a universal bias in favor of female candidates in Brazil. The bias is stronger on the political left, and with women, but is present across the political spectrum and in both genders.

"Candidate Gender and the Political Engagement of Women and Men" (gated) argues that in the U.S. women are no more engaged when a woman is involved, but men become less interested when their own party nominates a woman, but it doesn't matter if the other party does.

Bottom Line

With gender, the results are mixed, although there is less evidence of the effect the question asks. There doesn't seem to be much that shows that women are more engaged when a woman is on the ballot.

How This Could Matter

There are two ways in the U.S. right now where this is playing out: The Emerging Democratic Majority and the 2016 election.

The Emerging Democratic Majority

The Emerging Democratic Majority argues that the Democratic Party in the U.S. will have a permanent majority in the near future, because coalitions within the Democrat (e.g. youth, women, ethnic minorities) are growing faster than the coalitions making up the Republican party. A proposed mechanism for this, not by the authors, is that Democrats will be able to put forward more convincing slate by drawing from those coalitions, and that individuals from those groups will support such candidates. However, these studies suggest that this mechanism may not be operative. As social cohesion is implied by the thesis, whether or not the prediction turns out to be true remains to be seen.

2016 Election

Both parties' potential nominees include minorities (Cruz, Carson, Rubio as of Feb 3, 2016) and women (Fiorina, Clinton). Observers might hope that selecting an ethnic minority or a woman would shift that voting bloc more towards their party. Extant research suggests that is unlikely to be true, except in idiosyncratic instances--which may exist in the election.

Polls at this point are unreliable, so we cannot say reliably whether that effect is occuring. Anecdotaly, however, there is no evidence that Hillary Clinton is leading against Republican candidates among women, and may even be leading by less than Obama did. At the same time, there is no evidence of a groundswell of Latino or African-American (or Canadian, rimshot) support for Ted Cruz, Marco Rubio and Ben Carson from their respective ethnicities.

To demonstrate the transitivity of ethnic construction, is that as the race stands as of today (Feb. 3, 2016), almost all candidates would have been considered an "ethnic minority" at the turn of the last century. Interestingly, the only non-ethnic in the race who doesn't also have minority children, Jim Gilmore, is unambiguously in last place for the Republican Party, the party with the most 'white' support. Under current conditions, exactly half of the candidates--Clinton, Sanders, Fiorina, Carson, Cruz and Rubio--are viewed as part of an ethnic minority group, now, or a woman.

For further illustration of this point, Eisenhower and Kennedy were both viewed as 'ethnic' candidates in their day. No one thinks of them that way, now. In fact, Hoover was the first ethnically German president, but since his family had immigrated many generations earlier, and changed their name, Eisenhower was the first identified as such, and Kennedy was the first Irish. The second, was Ronald Reagan, but he was not identified as such in the public conscience by the time he ran for President. We still haven't had an Italian-American president, but no one ever talks about Christie or Santorum as ground breaking in that way.

---

Democrats

1. Hillary Clinton (Woman)
2. Bernie Sanders (Jewish)

Republicans

1. Jeb Bush (non-minority)
2. Ben Carson (Black/African-America)
3. Chris Christie (German/Irish/Italian)
4. Ted Cruz (Latino)
5. Carly Fiorina (German, Woman)
6. Jim Gilmore (non-minority)
7. John Kasich (East European)
8. Marco Rubio (Latino)
9. Rick Santorum (Italian)
10. Donald Trump (German)

Bush and Trump are both married to immigrant ethnic minorities, and Bush, Christie, Rubio, and Santorum are Roman Catholic, and Ben Carson is Seventh-Day Adventist.

---

Answer 2

IMHO, it is yet hard to study, so demonstrate, a general trend since the participation of women or minorities at major league elections are still a recent trend in itself.

Besides that, it seems impossible to answer in general objectively, since it was not provided any metrics for what constitutes a tendency. So it can be answered just anecdotally.

But as this question is rooted on a feeling:

I feel like when there's a female or minority candidate, the challenger is often female or minority, respectively. Is this a known phenomenon?

Yes, seeming that it can yield positive tactical results it can form a tendency. There has been at least one remarkable case: Chilean presidential election, 2013. It was the first election on the country with two women as the most voted.

The then ex-president Michelle Bachelet started with a large advantage. Chilean political system can be seen, to easy understanding, as a bipartisanship (similar to US, except that instead of two dominant parties it has two big coalitions-of-parties) At presidential primaries, Bachelet coalition (Nueva Mayoría) more than doubled its challenger coalition (Alianza). Just she herself had more than 150% votes than the sum of all candidates of the adversary coalition.

The primary candidates of Alianza were two men. Then, the winner of Alianza primaries (Pablo Longueira, who had himself less than 30% votes compared to Bachelet) withdrew on 17 July. So, Alianza replaced him by Evelyn Matthei, who had not participated in the primaries.

The news (locally and internationally) stressed similarities between the then two female candidates. Specially the aspect of

the empowerment of female leaders in politics.

It happened way more on news biased to Matthey, what was considered abusive by Bachelet supporters. Specially with theirs historical and political differences regarding Pinochet dictatorship. Fernando Mathei, Evelyn's father, was part of the military junta that ruled Chile from 1973 to 1990, while Alberto Bachelet, Michelle's father, opposed 1973 coup and was imprisoned and tortured for several months until his death in 1974 at the hands of the dictatorship.

Eventually, Michelle Bachelet won the second round on 15 november with 62.16% but some consider that change of the previous male-challenger by a female-challenger (working as a diversionary tactic) can have been one of the reasons that she did not won at the first round (in which Bachelet had 47.74%).