This article is from the source 'nytimes' and was first published or seen on . It last changed over 40 days ago and won't be checked again for changes.
You can find the current article at its original source at http://www.nytimes.com/2016/11/28/upshot/serious-voter-fraud-um-no.html
The article has changed 3 times. There is an RSS feed of changes available.
| Version 1 | Version 2 |
|---|---|
| Trump’s Claim of Millions of Illegal Votes? There’s Simply No Evidence | |
| (2 months later) | |
| Donald J. Trump continues to say that there were millions of illegal votes in the November election, costing him a popular-vote victory. We’re resurfacing an Upshot article from November that debunked the claim. | |
| Usually, the burden of proof for an outlandish claim rests on the person making the assertion. But Donald J. Trump is a master of making big claims, challenging others to disprove them and benefiting from the debate. | Usually, the burden of proof for an outlandish claim rests on the person making the assertion. But Donald J. Trump is a master of making big claims, challenging others to disprove them and benefiting from the debate. |
| On Sunday, he alleged that Hillary Clinton won the popular vote because of millions of illegal votes, presumably many of them cast in California, which is still counting votes and which is providing her with much of her two-million-plus-vote advantage. He said there was “serious voter fraud in Virginia, New Hampshire and California.” | On Sunday, he alleged that Hillary Clinton won the popular vote because of millions of illegal votes, presumably many of them cast in California, which is still counting votes and which is providing her with much of her two-million-plus-vote advantage. He said there was “serious voter fraud in Virginia, New Hampshire and California.” |
| Mr. Trump has offered no evidence for this statement. That’s because there isn’t any. | Mr. Trump has offered no evidence for this statement. That’s because there isn’t any. |
| California’s turnout was low. | California’s turnout was low. |
| If there were a few million fraudulent votes, the turnout rate in California would presumably be quite high. It was not. The turnout rate in California is projected to finish below the national average, according to estimates by Michael McDonald of the United States Elections Project. The turnout rate is based on the state’s voting-eligible population, which excludes noncitizens. | If there were a few million fraudulent votes, the turnout rate in California would presumably be quite high. It was not. The turnout rate in California is projected to finish below the national average, according to estimates by Michael McDonald of the United States Elections Project. The turnout rate is based on the state’s voting-eligible population, which excludes noncitizens. |
| If California had a low turnout despite millions of noncitizen voters, its turnout rate would be among the lowest in the country — edging just above Hawaii and Utah for third lowest. | If California had a low turnout despite millions of noncitizen voters, its turnout rate would be among the lowest in the country — edging just above Hawaii and Utah for third lowest. |
| Other Republicans did quite well. | Other Republicans did quite well. |
| You would expect signs of noncitizen voting in the results. Mr. Trump would have presumably lost the most ground in areas with large noncitizen populations. Instead, his biggest losses came on the affluent coast: in Orange and Marin counties, in San Diego and Silicon Valley. | You would expect signs of noncitizen voting in the results. Mr. Trump would have presumably lost the most ground in areas with large noncitizen populations. Instead, his biggest losses came on the affluent coast: in Orange and Marin counties, in San Diego and Silicon Valley. |
| He underperformed Mitt Romney’s 2012 showing most in Orange County, a redoubt of 20th-century conservatism that voted Republican in every election from 1936 through 2012. Mr. Trump lost the county by a nine-point margin. | He underperformed Mitt Romney’s 2012 showing most in Orange County, a redoubt of 20th-century conservatism that voted Republican in every election from 1936 through 2012. Mr. Trump lost the county by a nine-point margin. |
| These same Orange County residents voted to re-elect all four of their Republican members of Congress — not the result one would expect if Mrs. Clinton were bolstered by a wave of unanimously Democratic illegal votes. | These same Orange County residents voted to re-elect all four of their Republican members of Congress — not the result one would expect if Mrs. Clinton were bolstered by a wave of unanimously Democratic illegal votes. |
| Studies cited don’t support it. | Studies cited don’t support it. |
| When asked on Monday for evidence to support Mr. Trump’s assertions, Jason Miller, his communications director, cited two studies that he said offered examples of voter fraud: a 2014 Washington Post study on noncitizen voting and a 2012 Pew study on the poor state of voter registration files. | When asked on Monday for evidence to support Mr. Trump’s assertions, Jason Miller, his communications director, cited two studies that he said offered examples of voter fraud: a 2014 Washington Post study on noncitizen voting and a 2012 Pew study on the poor state of voter registration files. |
| Neither study could plausibly be construed to indicate that millions of illegal voters cast ballots in this election. Both were conducted well in advance of the election, and neither supports anything like what Mr. Trump has suggested. | Neither study could plausibly be construed to indicate that millions of illegal voters cast ballots in this election. Both were conducted well in advance of the election, and neither supports anything like what Mr. Trump has suggested. |
| What Mr. Miller referred to as the “Washington Post study” — actually an article by two professors on The Post’s Monkey Cage blog — used a survey to conclude that 6.4 percent of noncitizen adults voted in the 2008 election. | What Mr. Miller referred to as the “Washington Post study” — actually an article by two professors on The Post’s Monkey Cage blog — used a survey to conclude that 6.4 percent of noncitizen adults voted in the 2008 election. |
| This study was widely criticized and later debunked. But even taken at face value, a 6.4 percent noncitizen voting rate would account for only one-fourth of Mrs. Clinton’s margin of victory in the popular vote. | This study was widely criticized and later debunked. But even taken at face value, a 6.4 percent noncitizen voting rate would account for only one-fourth of Mrs. Clinton’s margin of victory in the popular vote. |
| And there is little reason to think that 6.4 percent of noncitizens voted in the 2008 election. The finding is based on the Cooperative Congressional Election Study, a huge study of tens of thousands of respondents. The 6.4 percent is based on just 21 survey respondents who indicated that they were noncitizens and voted. | And there is little reason to think that 6.4 percent of noncitizens voted in the 2008 election. The finding is based on the Cooperative Congressional Election Study, a huge study of tens of thousands of respondents. The 6.4 percent is based on just 21 survey respondents who indicated that they were noncitizens and voted. |
| That tiny number of respondents out of the tens of thousands who took the survey hints at a different explanation: measurement error. | That tiny number of respondents out of the tens of thousands who took the survey hints at a different explanation: measurement error. |
| Sometimes people make mistakes when they take a survey. They might misunderstand the question, accidentally give a wrong response to an interviewer or press the wrong button online. The interviewer can make a mistake, and record the wrong answer. This usually isn’t a significant problem, but it becomes one if you’re trying to find an odd, small finding in a survey of 30,000 people. Even if 99.9 percent of people answered a question correctly in the survey, there would still be 30 erroneous answers out of 30,000. | Sometimes people make mistakes when they take a survey. They might misunderstand the question, accidentally give a wrong response to an interviewer or press the wrong button online. The interviewer can make a mistake, and record the wrong answer. This usually isn’t a significant problem, but it becomes one if you’re trying to find an odd, small finding in a survey of 30,000 people. Even if 99.9 percent of people answered a question correctly in the survey, there would still be 30 erroneous answers out of 30,000. |
| It turns out that number isn’t far off. The study is a panel survey, so many people had been interviewed in previous years. A comparison of their previous answers can help give a sense of just how often people might make a mistake. In 2012, 20 of the self-described noncitizens surveyed reported that they had been citizens when they took the survey in 2010. But that’s unlikely — very few people lose their citizenship. | It turns out that number isn’t far off. The study is a panel survey, so many people had been interviewed in previous years. A comparison of their previous answers can help give a sense of just how often people might make a mistake. In 2012, 20 of the self-described noncitizens surveyed reported that they had been citizens when they took the survey in 2010. But that’s unlikely — very few people lose their citizenship. |
| What’s more, a similar number, 36 people, said they were noncitizens in 2010 but citizens in 2012. That’s possible — since people can obtain citizenship — but the finding is well above the rate of naturalization. | What’s more, a similar number, 36 people, said they were noncitizens in 2010 but citizens in 2012. That’s possible — since people can obtain citizenship — but the finding is well above the rate of naturalization. |
| This phenomenon could explain all of the noncitizen voting in the congressional election study. Here’s the clincher: There were zero voters among the respondents who indicated that they were noncitizens in both 2010 and 2012. | This phenomenon could explain all of the noncitizen voting in the congressional election study. Here’s the clincher: There were zero voters among the respondents who indicated that they were noncitizens in both 2010 and 2012. |
| Mr. Miller’s citation of a Pew Research study on the quality of voter registration files is far less salient. Yes, there are many millions of people who remain registered to vote after they have moved or died: Very few people remove themselves from the voter rolls when they move, or as they lie on their deathbed. But this has nothing to do with noncitizen voting, and it is not a meaningful contributor to voter fraud. | Mr. Miller’s citation of a Pew Research study on the quality of voter registration files is far less salient. Yes, there are many millions of people who remain registered to vote after they have moved or died: Very few people remove themselves from the voter rolls when they move, or as they lie on their deathbed. But this has nothing to do with noncitizen voting, and it is not a meaningful contributor to voter fraud. |
| Allegations of fraud by the left doesn’t mean fraud has occurred. | Allegations of fraud by the left doesn’t mean fraud has occurred. |
| The political right has not been alone in making unsubstantiated claims about election or voter fraud this year. Some supporters of Bernie Sanders did the same thing in the primaries. The Green Party presidential candidate Jill Stein is pushing for recounts in Pennsylvania, Michigan and Wisconsin, though no recount has ever overturned a lead as large as Mr. Trump’s in these states. | The political right has not been alone in making unsubstantiated claims about election or voter fraud this year. Some supporters of Bernie Sanders did the same thing in the primaries. The Green Party presidential candidate Jill Stein is pushing for recounts in Pennsylvania, Michigan and Wisconsin, though no recount has ever overturned a lead as large as Mr. Trump’s in these states. |
| Last week, a New York Magazine article made a splash by saying that some computer scientists believed there was evidence of voting irregularities in Wisconsin, based on Mr. Trump’s stronger showing in counties with electronic voting. This appears to be a mischaracterization of their views, but the view expressed in the article doesn’t hold up either. Mr. Trump’s edge in electronic-voting counties can be explained by their demographics: The counties are less educated and whiter than the parts of the state with paper ballots. | Last week, a New York Magazine article made a splash by saying that some computer scientists believed there was evidence of voting irregularities in Wisconsin, based on Mr. Trump’s stronger showing in counties with electronic voting. This appears to be a mischaracterization of their views, but the view expressed in the article doesn’t hold up either. Mr. Trump’s edge in electronic-voting counties can be explained by their demographics: The counties are less educated and whiter than the parts of the state with paper ballots. |
| Ultimately, there’s not much that’s odd about the results. Mrs. Clinton underperformed in Wisconsin, but also in Iowa and Minnesota — states with paper ballots and, in Minnesota’s case, a Democratic government. She fared well in California, but gained just as much in Texas, a state with Republican government. | Ultimately, there’s not much that’s odd about the results. Mrs. Clinton underperformed in Wisconsin, but also in Iowa and Minnesota — states with paper ballots and, in Minnesota’s case, a Democratic government. She fared well in California, but gained just as much in Texas, a state with Republican government. |
| There’s a dominant pattern: Mr. Trump’s strength with white working-class voters helped him a lot in crucial battleground states like Michigan and Pennsylvania. Mrs. Clinton had strength with Hispanic voters and well-educated white voters, but it helped her more in noncompetitive states, like Texas and California. These demographic patterns help explain the results without any need to suspect voter fraud. | There’s a dominant pattern: Mr. Trump’s strength with white working-class voters helped him a lot in crucial battleground states like Michigan and Pennsylvania. Mrs. Clinton had strength with Hispanic voters and well-educated white voters, but it helped her more in noncompetitive states, like Texas and California. These demographic patterns help explain the results without any need to suspect voter fraud. |