Could Trump be more right than wrong on voter fraud?

http://www.washingtontimes.com/news/...-than-estimat/

This study and the ODU study that Trump has cited (and the prof. who did it doesn't like it) are based on surveys done by Harvard going back some years. The Harvard guys said it showed "zero" illegal immigrants voting, the ODU guys have estimated 100,000 or so illegals voting, their actual study put it between 38K and 2.8 million. This guy says between almost 600K and 5.7 million.

If you read this article it does a decent job exploring the methodology, at least of the disagreements between these guys and Harvard on the data.

But the Virginia roll purge proves the basic point. they just found about 5.5K illegals registered to vote and about 1,800 of them have cast more than 7,000 votes per their records.

So we KNOW there are illegals voting in the elections. The question is whether it's enough to be statistically significant. Personally I've always felt Trump's estimates were crazy high, but now I'm not so sure.

They found those 5.5K in Virginia b/c in some other process, probably to get a license, they admitted to being illegal. Well I'm betting far more would never own up to it, and millions of illegals have fake IDs and SS#s and such.

So let's be conservative and say that full 50% of the illegals who have registered and voted admitted to it to some other state office, and 50% didn't. that means there are 11K illegal voters in Virginia. Virginia's population is 8.4 million. US population is 321 million, or 2.6% of the overall population.

that means if we just average out Virginia to the nation, which IMO is GROSSLY conservative b/c there are way more illegals in Cali and Texas and the rest of the west than there are in Virginia, the absolute lowest estimate of illegals on the voter roles is 5.5K / 2.6% is 211,000. If we are at all realistic you have to figure with that 50% number above it's more like 425,000, and as I said I think that's low b/c of demographics.

Honestly, it's not far fetched at all that there are a million illegal immigrants on the voter roles nationally. I'm not saying it's the case, but just back of the envelope math says it's not as crazy as I may have thought. Now is it 2 or 3 million? I still think that's a very high estimate, but it's not a stupid, crazy, insane estimate. If we say that only 25% of illegals in Va are admitting to it and being found then it's 22K in Virginia or 850K. Still well shy of 2-3 million, but in the range of that 1 to 1.5 million range as throws at a dartboard go.

And that's in line with the ODU study and this one.

IMO Trump's claim he'd have won the popular vote is still pie in the sky without some much better (and probably impossible to find given how states like cali will never cooperate) proof, but is it possible that votes of illegal immigrants and other forms of voter fraud are having real impacts on elections? yes, absolutely. Lots of elections are decided by 100s of votes at the state level every year, and Trump won those key states by 10K to 20K votes. It doesn't seem insane that a bigger population state like Michigan (about 9 million), also a state with a larger immigrant population, couldn't have 10K or 15K illegal immigrants voting, plus other forms of fraud could get to 20K votes. Not saying it's likely, but it doesn't seem out of the question.

It won't swing any results in a national election in someplace like California, but I could see it being a difference in close House races, even some Senate races, and maybe even be a difference for President in some really close states.

the difference in the Clinton/Trump race in Michigan was 13K votes. Not a whole lot. New Hampshire is a small state, but still the difference was only 2,700 votes.

Anyway, took a break from other math to do this math. Saw that study discussion, thought it was interesting to see if we could get in that range with numbers from Virginia that we do know.

California is going to have a huge number of illegals among their voting rolls. Illegals easily procure driver's licenses and with that license they are eligible to vote.

Deny one persons vote and thet are disenfranchised, huge issue but have thousand of ineligible people vote and its statistically irrelevant and of no concern