A GUIDE TO ELECTION FRAUD: As November 7, 2006 fast approaches, there’s clearly the possibility of the Perfect Storm of mass confusion, controversy, litigation, and just plain ugliness surrounding the results. Millions of us could be riveted to C-Span, well past Thanksgiving, dissecting the latest news from Iowa’s Secretary of State or the 7th Congressional District of Indiana.

So, to help make Election Night (and beyond) experts of all of you, let’s review the specific types of electoral problems we might experience –which fall into four major categories.

Category I: Citizens deprived – by law, mistakes, incompetence, or outright mischief – of being registered as voters.

“Voter suppression” laws and tactics have a long pedigree in American politics. Their purpose: prevent people from even getting inside a voting booth on Election day.

The most infamous, of course, have been explicit prohibitions against blacks, Native Americans, and women. Then there were literacy tests (“How many bubbles in a bar of soap?” was often asked of black would-be voters in the 1950s) and poll taxes. Prisoners serving time have almost always been disenfranchised. (Ex-felons, even off parole, still can’t vote in a half-dozen states.) Dirty tricks? How about blanketing poor neighborhoods with flyers, strongly encouraging people to vote at their local polls – on Wednesday, November 8th?

The most worrisome laws, for 2006, involve requirements in many states to show photo ID, not just to register, but actually vote. While 15 or 30 seconds to check an ID isn’t all that much, multiply it by thousands of voters and it’s not trivial. Not to mention the myriad of judgment calls and (attempted) uniform applications of standards, especially in states that literally say 1 photo ID = two non-photo IDs. (“Uh, will my expired Costco card AND an August cell phone bill work?”)

Then there’s the small-time – but no less infuriating – stuff. Like a voting official telling “Sandra G. Cox” she’s no longer registered, because the “G” on her updated form somehow got entered as a “C” by a county elections official. A registration card lost on a desk somewhere, a voter who mailed one back in but forgot the stamp….

Expect literally hundreds of thousands of citizens showing up at the polls on November 7th, thinking they’re properly registered – and being told they’re not. Some won’t be; others will be; and for many, it won’t be clear. Any voter caught up in such a “Category One” storm should do everything they can to vote – and if still denied, insist on voting a “provisional ballot” while working to straighten out the mess, post election.

Category II: Registered voters, who attempt to vote – but end up discouraged and not even casting votes.

This kind of storm includes the polling stations where voter lines snake outside the buildings, with voters perhaps waiting until midnight or 3 am, in terrible weather.

Some of the reasons this might happen (apart from those mentioned above): Urban, Democratic-leaning poor counties are likely to have fewer voting machines per 1,000 voters, and they’re often improperly maintained. (In fairness, this will also be true in some rural counties, too) They’re also likely to have fewer trained election workers – the biggest single “cost” of holding elections. Add the complexity of DRE voting machines and the chance for breakdowns and confusion, and it’s not hard to see situations where voters simply decide they really can’t be 2 hours late to fix dinner for their families.

As with Category I, the die has pretty much been cast for Category II problems. Polling places by now either have enough machines, or they don’t. All election officials can do is cross their fingers -- and also hope that bad luck isn’t compounded by even worse luck. Like 12 inches of snow falling in the Upper Midwest amidst a howling blizzard; a late-season hurricane barreling up the East coast; an electricity black out; terrible traffic jams in on the expressway – you get the picture.

Category III: Voters whose actual votes are either not recorded at all – or are recorded incorrectly.

This, of course, is the big bugaboo of electronic voting critics -- and understandably. Just what is being recorded when a voter’s finger hits the “Bob Hornswaggle” button on the touch screen? Presumably, an electronic impulse is sent from the screen into the inner workings of a database, where another impulse then records yet another vote for Fighting Bob. But if there’s no paper record of any of this, how can we be truly sure the vote got recorded at all?

In recent decades, this concern has largely been absent from Election Night conspiracy theorists – replaced by other familiar demons, such as “dead voters” listed on registration rolls (whose extra votes are then cast by live impersonators) and “extra ballots” that somehow get filled out and stuffed into ballot boxes by party hacks while election officials like the other way. But it’s actually an “old” scenario; remember the old machine lever voting machines, so common, especially in urban (and largely Democratic) precincts?

Category III problems will be especially watched for on election night by Hacker Conspiracy Theorists – though if you think about it, any “evidence” of such will be very hard to detect. So these storm-watchers will zero in on “suggestive indicators,” such as exit polls that diverge significantly from the released numbers.

The problem is – polling has become an increasingly difficult art in recent years, given the large number of non-responders and “perverse” responders (those that deliberately skew such surveys with bogus responses). Another complication is that more and more voters are casting absentee ballots.

Finally, the Election night judgments of politicians and journalists about an election result, especially in close races, are increasingly suspect. Why? Notwithstanding those “99% of precincts” reporting” squibs on the TV screen at 11 pm EST, seldom do they reflect anything close to that percentage of actual votes.

Large numbers of votes aren’t – and in some case, cannot – be counted until days, even weeks after November 7th. For example, many states process all polling place votes before they count a single absentee ballot. This isn’t bureaucratic laziness; it’s actually a basic “integrity check” so that poll books can be inspected, to see that a voter didn’t sign up to vote at the precinct but also submit an absentee ballot.

Category IV: Real and true votes get cast but are then inaccurately counted, either because they’re missed completely or given to the wrong candidate.

While many confuse Category III issues with Category IV, they’re fundamentally different. This is a counting problem, not a recording problem.

Since DRE machines both record and count, any concern about the integrity of these machines will likely apply to both. But most voters this fall will still use paper ballots. Not chad-prone punchcards, but ballots that might require a bubble to be filled in with a number 2 pencil, or arrows connected with a line, etc.

For decades, computerized counting machines have been used to tally paper ballots. The combination of electricity and software is actually exquisitely suited to this otherwise tedious, labor intensive task. Without knowing as much, impatient journalists and anxious politicos have accelerated the rush to these devices. It lets lots of races to get “called” before bed time, since enough votes get tallied quickly enough to (in most cases) reveal a clear trend that would be hard to reverse.

With the recent hubbub over DREs, the “counting” of votes has re-emerged as a genuine concern. It hasn’t helped that there have been actual situations where county officials have released laughably wrong vote results from these machines – e.g, 16,000 votes cast in a county with just 3,000 voters.

Fortunately, there’s an important “back-up” mechanism for vote counting, though it does require a paper record. It’s the recount. A race that’s close enough will automatically produce a recount. But in almost every state, regardless of the margin, losing candidates have the option to request (and pay for) a partial or full recount, if they suspect any shenanigans. If the recount materially affects the result, the candidate gets his/her money back.

What about states with DREs, that lack Voter Verified Paper Trails? While suspicious losers can demand inspections and software reviews – and doubtless some will be asked for – the lack of a paper trail will be a genuine problem. Again, the die’s been cast on this one, too – we’ll just have to do our best.

In conclusion, two things to keep in mind:

Finally, it’s important to remember two important things about Elections s you watch returns on November 7th:

First the term “2006 Election” is a misnomer. November 7, 2006 is simply a common date, on which there are literally hundreds of thousands of elections, held in tens of thousands of separate political jurisdictions across America.

Partisans and reporters will scrutinize high profile Governor, Senate, and Congressional races for any signs of mischief (and worse) At most, that’s about 100 races. But in reality, each of these contests is the cumulative result of thousands of separately held election events, which occur in individual precincts as well as counties.

Second, no election in America will actually be officially decided on November 7, 2006.

While the time periods vary – some states have 7 days, others as many as 30 – this issue of “election certification” is critical to grasp. Election officials tend to be pretty overwhelmed on Election Day itself -- even if the power doesn’t go out, the doors aren’t accidentally locked, the machines all work flawlessly, the weather is clear and crisp, and voters are polite and orderly. So by having a period of time before final certification, they have a chance to sort through all the inevitable issues that come up – for example, checking voter registration records to ensure a vote can be cast, checking signatures on absentee ballots, double and triple checking the counts, processing provisional ballots

The vast majority of this work is unnoticed and unappreciated. Do journalists care whether Candidate A beat B by a 43,218 to 26,198 margin, as reported on Election night, or by a 43,312 to 26,231 margin, as officially recorded 20 days later? Hardly. But it matters greatly if you’re truly concerned about election integrity.

Votes do get miscounted; voters do get incorrectly un-registered. And while not all mistakes are detected – and some go the other way, allowing legally unqualified people to still cast ballots – the “certification window” is central to ensuring a “fair election.” Yet because the pressure is so overwhelming for “instant gratification,” county election officials have little choice to but to release their “rough draft” totals.

If there ever was a year to “withhold judgment” for a day or a week on a fairly (much less very) close election, 2006 is it. Given the widespread concerns and fears about election mischief and fraud, talk about “apparent winners” should be muted. Voters who see a close election result can usually do the math, and understand who’s “ahead.” But if thousands or even tens of thousands of votes have not yet been counted at all, why not just cool the journalistic jets? Give election workers time to sort out the “last 2%” of votes to decide which should be counted, and which not. In an Election that could go down to the wire, any little bit of calm, patience, and understanding could only do us good.