Here’s a paper from Ryan Enos:
The eff ect of group threat on voter mobilization has been tested using observational data across a number of di fferent geographies and units of analysis. Previous studies have yielded inconsistent findings. To date, no study of voter mobilization has directly manipulated group threat using a controlled experiment. I take advantage of the unique racial geography of Los Angeles County, California, which brings di fferent racial/ethnic groups into close, yet spatially separated, proximity. This geography allows for a randomized, controlled experiment to directly test the eff ects of stimulating racial threat on voter turnout. A test of 3,666 African American and Hispanic voters shows an average treatment e ffect of 2.3 percentage points. The e ect is 50% larger for African Americans than Hispanics. These results suggest that even low propensity voters are aware of the geographic proximity of other groups and can be motivated to participate by this awareness.
See page 21 of the article for an example of the treatment, which includes a map and this bit of text:
But what really interested me about the article was that he imputed ethnicity using available information on last names. (Go to the article and search on “surname.”)
P.S. But, boy, does this paper need some good graphs! I like the paper and want to plug it here, but there’s no grabby graph. I’d like to see that scatterplot of raw data with fitted lines, showing what the researcher found and how these findings came from the data. Regression tables are fine (well, not really; they should be graphs, but that’s another story), but I wanna see what’s happening. I wanna see what’s happening.
Ethnicity based upon last names works fairly well. It was a task I had with a previous employer, in addition to substrings of names being reasonably unique to parts of the world. One issue is that the person's paternal line is going to be the ethnicity/race that is imputed and so half of the interesting information is lost. First names can also help as they often reflect ethnicity/race as well.
One great public data set (not free, but very cheap) if you want to look at country of origin (useful for hispanic/non-hispanic and other questions) is Clark County marriage records. 7% of US marriages happen in the Las Vegas area so if you only take one location it is a decent place. On the marriage record you get country of birth and country of birth of parents.
What are the ethical research guidelines for experimentation (i.e., meddling) during elections? I'm assuming from the start that treatments thought to suppress the vote are right out, but there must be guidelines about selective application of treatments that increase voter turnout…?
Corey,
research like this one, of course, had to pass through an Institutional Review Board for approval. The guidelines that seem to apply are first, as you indicated, you can't suppress turnout. For reasons that are not probably very well thought out, there always seems to be no objection to increasing turnout. Second, it usually has to be in a situation where it is difficult to affect the election outcome. For example, this was done in LA, which is overwhelmingly Democratic, so on a national scale, it does not matter much. The IRB board did not seem to care that it might affect more local elections.
On a more philosophical note, it is interesting to think about how much experimentation we should be comfortable with in elections and other such phenomenon. We accept potentially harmful experimentation in medical science because of the importance of the consequences. One could argue that elections are of equal gravity, so it is interesting to think about whether we should be willing to tolerate more experimentation.
That's an interesting line of thought! Thanks for the reply.
The experiment was great especially all the ground work. I still have reservations about the ethics of the experiment and I am glad to see there was some ethical review.
We accept potentially harmful experimentation in medical science because of the importance of the consequences.
It’s more like – a person, with their *informed consent*, may be allowed to undertake a potentially harmful treatment in a randomised clinical trial as the new treatment is thought to be as good as current best practise.
The thing about electoral trials is that it may influence the outcome in the election which effects everyone in the electorate not just the person whom the treatment is applied to *and* it may have repercussions in future elections.
p24 the author concludes that there is no effect.
For me, he just says that his strange theory has actually no practical basis.
Speaking about me, if I ever received such a letter, I would not even understand it.
In France where I vote, there is a lot of information available (surveys, two-stage elections), so people can very well anticipate what the result will be.