Is there voting fraud in Philadelphia?

Sean Hannity thinks so. The Fox News broadcaster and Donald Trump adviser cites as proof that there were 59 voting divisions in Philadelphia where Republican presidential candidate Mitt Romney got zero votes in the 2012 election.

That could only happen, Hannity argued recently, if the election process in Philadelphia is rigged.

Is Hannity correct?

Yes. And no.

There were, in fact, 59 voting divisions where Romney got zero votes in 2012. There is a simple explanation for how this shutout happened: No one voted for him.

All of those divisions were in wards where nearly all the voters were black. In 2012, President Barack Obama won 99 percent of the black vote in Philadelphia, according to my analysis of key black wards.

It stands to reason that if one candidate averages 99 percent of the vote and another averages 1 percent, there are going to be places where that 1-percenter gets zero votes – just as there were divisions where Romney got two, three or just a half-dozen votes. 

Does it surprise you that black voters would prefer a black president to a white, Mormon conservative multimillionaire? For the record, Obama got 588,000 votes in Philadelphia in 2012; Romney logged 96,000.

Now that you know the facts, will it make any difference?

Not if you are a Republican true believer. As GOP presidential nominee Donald Trump has proven ad nauseum, Republicans are impervious to facts, and this especially holds true when it comes to voter fraud. They believe – no, they know – it is widespread, especially in areas where there are poor, black voters.

Instead of trying to win over those voters, they prefer voter suppression – making it harder for them to cast their ballots.

Preventing voter fraud was the stated purpose of a voter ID law passed by the Republican-controlled legislature before the 2012 election. It never went into effect – the courts declared the law discriminatory. It also noted evidence that no one could find any examples of in-person voter fraud in Pennsylvania.

Does this mean there is no voter fraud in Philadelphia?

No. There is some. The district attorney’s office has prosecuted eight cases in recent years, most of which dealt with ballot tampering – usually by low-level election officials – on voting machines after the polls had closed.

As a rule of thumb, I have observed that fraud cases are limited to minor races where the pool of voters is small – and where a dozen votes can make a difference, like with races for party committee people. Indeed, one case involved someone running for Democratic State Committee.

I have never seen fraud cases involving citywide elections. There’s a reason why: It would take a massive organizing effort, and would likely yield minimal results. 

Suppose, for instance, that you wanted to buy votes and organized a plan to pay select voters $20 a head to vote for Mitt Romney. Your goal would be to get 2,000 extra votes.

A couple of points:

What good would those 2,000 votes do for Romney? Obama won Philadelphia by a margin of 492,000 votes in 2012. Your plan would lower his margin to 490,000.

Also, it is axiomatic that the more people involved in this kind of deal (it’s called a conspiracy in the law), the harder it is to keep it secret. If someone were passing out Jacksons to buy votes, word surely would spread.

During my career, I have spent a lot of time tracking down tips about voter fraud. These days, we have sophisticated tools to sniff it out. Take turnout. Some years ago, we took data on voting in one election for all 1,700 of the city’s divisions and ranked it from highest to lowest turnout.

Our feeling was that if citywide turnout was 60 percent and you had divisions where it was 90 percent, that may indicate fraud. We found a handful of such outliers in our number-crunching and looked into them.

Inevitably, they turned out to be high-rise apartment buildings with a generally elderly population large enough to merit having voting machines in their lobby. The residents were active voters. The machines were nearby. It was convenient to vote – and they did.

Another claim made to prove there is voter fraud is the high number of people registered to vote in the city: We have an over-18 population of 1.16 million and over 1 million registered voters. How can this be?

The answer lies in federal voting laws that make it easier to add people to the rolls and harder to remove them. Under the law, seven years must pass before a person who has not voted can be purged from the rolls.

People who die are purged more quickly, because the state Department of Health forwards data on deaths to election boards each month. People who move in-state and change their driver’s licenses to reflect their new addresses are also purged from the rolls or, more accurately, their files are transferred from their old county to their new residence.

The upshot is that Philadelphia’s voting rolls (and those in other counties) are inflated.

Last year, I did an analysis of voting using data from the City Commissioners. At the time, there were 1,036,943 registered voters, but nearly 300,000 of them had not voted even once in the previous five years. So, the real total of active voters was about 740,000.

If the past is any guide, about 650,000 people will vote in this November’s election.

I can confidently predict that Hillary Clinton will win the city, as every Democratic presidential candidate has done for the last 50 years. 

She will win not because the same people voted five or six times, as Trump alleged in a speech in Altoona. She will win because about 650,000 people will each vote once, and the overwhelming majority will repudiate Trump's candidacy.

That's not cheating. That's democracy in action.