Oh, They’re Back

A while ago, Goldy thought Sound Politics might have gone the way of the Buffalo. I guess they did in the fact that it looked like they would be gone, but they’re still around. This is a bad metaphor so I guess instead of continuing it, I should still make fun of them? This Jim Miller piece [h/t to my friend N via email and to Tensor in comments] goes for it, starting with a violation of Betteridge’s Law and then not getting better after that.

Do Illegal Votes By Non-Citizens Sometimes Determine Elections Here In The United States?

No? Does Sound Politics need a new hobby horse? Yes!

I began thinking about that problem more than two decades ago, and concluded, tentatively, that the answer was yes. After the 2000 presidential election, and all the disputes that went with it, I put more time into the problem and, in 2002 gave an example, Maria Cantwell’s defeat of Slade Gorton, where that may have happened.

It sure is convenient to be able to just assume out of thin air that any close election must really belong to you. Because reasons.

The logic of my underlying argument is simple:

As it is written by a simpleton.

1. There are more non-citizens living in the United States than ever before.

Are there? No links or any other evidence was provided. It wouldn’t surprise me if, for example, when we were building the West with a significant amount of Irish, German, and Scandinavian (among others) immigrants that there was a higher number of non-citizens in the US.

2. Some small proportion of them will be tempted to register and vote, illegally.

Well, if some small portion are tempted to do a thing, that thing must happen all the time in fact. QED, I don’t think we need to even go on.

3. Many places have few checks against non-citizens voting, so most of those tempted will succeed.

Some amount of a small number will be tempted could in theory do a thing then that thing is always already happening.

4. The non-citizens here in the United States tend to favor the Democratic Party.

Again, no need to actually prove this.

Therefore, I concluded, some close elections were being tipped to the Democrats by these illegal votes.

So two unproven assertions and two things that are admittedly not very large combine to… something? Here, let me try:

1) Puppies are adorable
2) Most people who hate puppies are Republicans
3) That time Mitt Romney tied his dog to the roof of a car, or whatever
4) A small number of cars have been owned by Mitt Romney

Therefore, I conclude that people who hate puppies are the only reason that Romney won any states.

All through this time, however, I have been unhappy that I could not find any academic studies of the question. And it was obvious to me that it was not a difficult research problem, that you could get a good start on it simply by running a large survey, and asking the right questions, in the right way.

Hey you know how sometimes our trolls will come into the comment threads with large lists that if you’re a dummy — and don’t understand they’ve been obviously ripped out of context — sometimes seem a bit silly? I mean, I don’t know everything this survey does, but I imagine if our trolls actually cared, as opposed to were just copying and pasting from some nonsense list someone compiled, that they would add asking people who can’t vote if they vote. It’s a large survey and it probably does have value. But I wouldn’t posit this piece as the most important.

Now, finally, three researchers, Jesse Richman, Gulshan Chattha, and David Earnest, have done that study, and provided direct evidence for the conclusion I reached more than a decade ago.

I mentioned in the last Open Thread a thing that refutes this. So I’ll quote from that instead of Jim Miller’s quoting of the piece, and then pick it back up with Miller.

The limitations are, in fact, numerous, and not limited to those that Richman and Earnest enumerate. Their estimates rely on a key question from the Cooperative Congressional Election Study: “Are you registered to vote?” Notably, this is not the same question as “Are you registered to vote in the United States?” In principle, non-citizens could be registered to vote only in their home country and respond affirmatively, and truthfully, to the question on the survey.(Respondents are asked for the Zip code at which they are registered to vote, but this could be interpreted as the Zip code at which non-citizens receive absentee ballots from abroad. Mexico, for example, has allowed absentee voting by mail from abroad since 2005.) If this sounds outlandish, consider that 20 percent (15 out of 75) of those non-citizens claiming to be registered in 2008 were in fact verified as not being registered to vote in the United States. Another 61 percent (46 of 75) could not be matched to either a commercial or voter database. That leaves only 14 out of 75 non-citizen respondents claiming to be registered in 2008 who were in fact confirmed as registered to vote in the United States.

Anyway, back to the nonsense.

You may want to apply those numbers to make back-of-the-envelope estimates of the likelihood that illegal votes by non-citizens gave Democratic candidates victories in your favorite close elections. For example, this strengthens my conclusion that Cantwell’s victory in 2000, by just 2,229 votes, was illegitimate. And it makes it nearly certain that Christine Gregoire’s 2004 victory over Dino Rossi, by just 129 votes, was illegitimate. In fact, I will go further and say that, if you could have magically eliminated the non-citizen votes from just Seattle, Dino Rossi would have won the final recount.

Um, OK. They asked 55,400 people about their voting habits and 13 said they voted when they weren’t citizens. Assuming none of them misinterpreted the question and none of their answers were entered wrong, that’s 0.000235% of the population voting despite being unable to vote in the country legally. So yes, if you assume that is a correct number, there were 659.3999 of the 2,810,058 people who voted in the Rossi-Gregoire race that voted who shouldn’t have. That’s higher than the total difference in the vote, but (a) there’s nothing to indicate how they would vote if they exist except for the general results mean they probably would have gone something like half-and-half to each candidate, with a few to the Libertarian and (b) the state GOP and conservative blogs spent months and months looking for illegal votes, but really didn’t turn up anyone in this category. I can’t imagine if there were 659.3999 humans out there who had voted illegally that none of them would have been found. Talk about Sound Politics being a bastion of rank incompetence.

There is nothing difficult in the chain of reasoning that I went through years ago, and I am nearly certain that others came to the same conclusion, independently. I think it likely that unscrupulous Democratic operatives saw that they could gain a few votes by making it easier for non-citizens to vote, and that Republican operatives saw that they could be on the side of truth and justice — and gain a few votes, net — by putting tighter controls on registration and voting. Understanding that non-citizens were sometimes tipping elections to the Democratic Party would explain, for example, why George Soros, and others, put money into the Secretary of State Project.

Not the actual, legit disenfranchisement of voters that happens when you put those things in place. It’s the 0.000235% at most of people.

As has happened far too often in recent years, I wish that research had proved me wrong, wish that our close elections were not sometimes being determined by illegal votes.

Done! Today only I can grant wishes.

Cross posted at Jim Miller on Politics.

I’m totally reading it again!

(For the record: I can think of a few places where illegal votes by non-citizens might help Republican candidates, for example, where there were many immigrants from Russia, or other countries that have suffered from Communism.)

For the record, I’m not sure that immigrants from Russia today are particularly opposed to Communism. Does Jim Miller not know what’s happened in the last couple decades in Russia? Anyway, there can be lots of reasons that people vote.