RE-Vote, Not Recount, in Ohio

A BUZZFLASH READER CONTRIBUTION
by Anne Pfeiffer   29 November 2004

It appears there will be a recount in Ohio. However, a recount probably won't do much good in three counties that have suspicious numbers Franklin, Mahoning, and Warren. I'll deal with Franklin and Mahoning first, then Warren later. Here are some simple facts about the election in Ohio:

1. Ohio has 88 counties. The three largest counties that went strongly for Kerry are Cuyahoga (which contains Cleveland), Franklin (which contains Columbus the state capital), and Summit (which contains Akron). These counties all went for Gore in 2000. Cuyahoga and Franklin are the only two counties in Ohio with populations topping 1 million.

2. With a population of about a quarter million, Mahoning is not the largest county in Ohio (it's only 10th largest), but it is one of the bluest. Mahoning contains Youngstown, an old steel town that has been hurting economically for decades. Youngstown suffers from chronic high unemployment. Even Bruce Springsteen wrote a song about Youngstown's woes. Gore led Bush there in 2000 by almost 29,000 votes. That was a larger point spread than in Summit County, which has over twice the population of Mahoning. In fact, it was a larger point spread in Gore's favor than any other county except Cuyahoga, which is Ohio's bluest county as well as its largest. Simply put, Mahoning is an important Democratic stronghold in Ohio.

3. Franklin County has two significant errors on its Board of Elections canvass report. On error has already been exposed in the media. That is the 3,893 extra votes that were applied to Bush's total in Gahanna.

4. The second error has been discovered by me and has not yet been reported anywhere, as far as I can tell. The error is that over 20,000 absentee ballots were added to Franklin County's total TWICE. This may seem incredible, but it easy enough to verify. It's right in the canvass report (as Absentee 1 and Absentee 2), and if you sum the totals with a calculator or spreadsheet, you will see that both sets are added in to the county totals. It seems that the single set of absentee ballots may have been counted twice for accuracy. Some candidate totals are identical in the two counts, and some differ by just one vote. Still others are reported as zero in the second count, which implies that only certain races were recounted. This huge error didn't have a significant effect on the presidential race (or on local races I checked), but it did artificially inflate the turnout rate for the county.

5. Cuyahoga and Summit Counties use punch cards, as do most of the counties in Ohio. Franklin and Mahoning use paperless touch screen machines (made by two different manufacturers). Mahoning's is a true computer touch screen system, while Franklin's machines are an older touch pad type that is described in detail on the Franklin County Board of Elections web site. Mahoning was switched over to a touch screen voting system only recently. Both Franklin's and Mahoning's machines are considered to be DRE (Direct Recording Electronic) machines. They are "paperless" because they do not produce any paper trail that can be used for recounts. There is no TANGIBLE way to verify that votes were recorded and counted properly.

6. There was a dramatic increase in voter turnout all over Ohio as compared to 2000. But not in Franklin County, where the reported turnout rate was actually LOWER than it was four years ago! It was also THE lowest reported turnout rate of all Ohio counties (after correcting for Gahanna and absentee errors). Mahoning also had a much lower than expected turnout rate.

7. When fraud or major malfunction is not a factor, DREs have much lower "spoilage" rates than punch card and optical scan voting systems. That alone should have given Mahoning a bump over 2000 since Mahoning was not using DREs at that time. It also makes Franklin's low turnout rate even more suspicious when compared to all the counties that use non-DRE systems.

8. Counties typically had an increase in both registered voters and reported voters. But the normal pattern was for the increase in reported voters to be much higher than the increase in registered voters. In Franklin County, just the opposite was true. There was a much higher increase in registered voters than reported voters. Mahoning also had a higher increase in registered voters than in reported voters.

9. Anecdotal evidence suggests that Franklin and Mahoning Counties had the same long lines and several-hour-long waits to vote as anywhere else in Ohio. In addition, there were "Get Out the Vote" volunteers swarming all over Franklin County in the days leading up to the election. Mahoning had the largest and best-coordinated "Get Out the Vote" effort the county has ever seen.

All of these facts together suggest that Franklin and Mahoning Counties actually had much higher turnout than was reported. I believe that a large number of Franklin and Mahoning County votes were "lost" by the touch screen machines. Look at these figures, which were obtained from the Secretary of State's official website:

Cuyahoga: registered voters in 2000: 1,010,764 reported voters in 2000: 586,914 turnout rate in 2000: 58.07% registered voters in 2004: 1,005,807 reported voters in 2004: 665,334 turnout rate in 2004: 66.15% increase in registered voters 2000-2004: -4,957 increase in reported voters 2000-2004: 78,420 percent increase in turnout rate 2000-2004: 13.92%

Summit: registered voters in 2000: 354,189 reported voters in 2000: 232,252 turnout rate in 2000: 65.57% registered voters in 2004: 368,858 reported voters in 2004: 275,551 turnout rate in 2004: 74.70% increase in registered voters 2000-2004: 14,669 increase in reported voters 2000-2004: 43,299 percent increase in turnout rate 2000-2004: 13.92%

Franklin: registered voters in 2000: 681,949 reported voters in 2000: 417,800 turnout rate in 2000: 61.27% registered voters in 2004: 845,720 reported voters in 2004: 494,792 (after corrections) turnout rate in 2004: 58.51% increase in registered voters 2000-2004: 163,771 increase in reported voters 2000-2004: 76,992 percent increase in turnout rate 2000-2004: -4.51%

Mahoning: registered voters in 2000: 179,546 reported voters in 2000: 116,889 turnout rate in 2000: 65.10% registered voters in 2004: 194,673 reported voters in 2004: 131,938 turnout rate in 2004: 67.77% increase in registered voters 2000-2004: 15,127 increase in reported voters 2000-2004: 15,049 percent increase in turnout rate 2000-2004: 4.10%

You can see how Mahoning and Franklin differ from Cuyahoga and Summit. We can even combine the remaining "blue" counties (counties that went for Gore in 2000 and Kerry in 2004), and see the same pattern as Cuyahoga and Summit. These other counties include Trumbull, Athens, Lucas, Belmont, Lorain, Jefferson, Monroe, Portage, Ashtabula, Erie, and Montgomery. Although I've combined the counties here for efficiency, you can look at these blue counties individually and see the same patterns.

Other "blue" counties: registered voters in 2000: 1,393,942 reported voters in 2000: 878,912 turnout rate in 2000: 63.05% registered voters in 2004: 1,410,764 reported voters in 2004: 1,011,123 turnout rate in 2004: 71.67% increase in registered voters 2000-2004: 16,822 increase in reported voters 2000-2004: 132,211 percent increase in turnout rate 2000-2004: 13.67%

So what can we glean from these facts and figures? Out of all these counties, only Franklin and Mahoning use DRE voting systems. Out of all these counties, only Franklin and Mahoning show the odd pattern of having a greater increase in registered voters than in reported voters. Out of all these counties, Franklin and Mahoning have the lowest increase in turnout rate from 2000; the others average a much higher turnout rate increase of about 13-14%. Does that seem suspicious to anyone else? It is especially fishy when you consider the strategic importance of these two counties and the fact that both were the object of powerful GOTV campaigns Mahoning mainly by Democrats, Franklin by both sides (although the Democrats were reportedly much more successful than the Republicans).

There are a few other lessons to be learned here. First, despite the impression given by the media that conspiracy theorists across America are poring over the election data looking for even the tiniest errors or inconsistencies, it was possible for a mistake involving over 20,000 votes in one of the most critical counties of one of the two most critical states to be overlooked by everyone. Makes you wonder how many other anomalies and errors are out there waiting to be discovered. It is possible that because this mistake was listed in plain sight on the canvass report that people assumed it made sense somehow. They may have thought these were simply two sets of absentee ballots.

In addition, the canvass report is publicly displayed as a PDF file, although I'm sure Franklin BOE stores the data internally as a spreadsheet. It takes a lot of effort for an individual to get from a PDF to a spreadsheet in order to check all the arithmetic. However, the Franklin Board of Elections really has no excuse in not catching this earlier. It was either gross incompetence or deliberate fraud, and I suspect the latter. I think that this "error" was made deliberately to cover up Franklin's suspiciously low turnout. I believe the Gahanna "error" was also made deliberately, for the same reason. The turnout rate reported for Franklin was 60.95%. After corrections, the turnout rate is actually 58.51%.

If Franklin and Mahoning experienced the same 13-14% increase in turnout rate as the other blue states, how many votes would be added? Let's do some math. The Mahoning turnout rate would be 74.12%. The Franklin turnout rate would be 69.76%. In terms of raw votes, that would be over 12,000 more votes for Mahoning and over 95,000 more votes for Franklin. Together that's about 108,000 votes. That's a large amount of votes that just didn't materialize. I think we need to try to figure out where those votes went.

It is my understanding that candidates asking for recounts are allowed to examine the books that people sign when they come to the polls to vote. It will be time-consuming but important to check the number of people who actually signed in to vote against the number of votes actually cast. What Franklin BOE lists as "public precinct count" on the canvass report is actually a reflection of the number of votes counted, NOT the number of people who signed in to vote. I would suggest they start with the poorest, most African-American precincts and also do some spot checking (contacting registered voters whose names do not appear on the polls and asking them if they voted) to make sure the books weren't altered. Let's just hope the Board of Elections doesn't attempt to drag its feet until the clock runs out.

WARREN COUNTY

Now let's move on to the strange case of Warren County. Warren County made the news twice with regard to the past election. First, it was the last county in Ohio to complete its vote count. Second, the ballots were counted in secret by its Republican election officials. The media were not allowed to observe the counting process and were relegated to an area two floors below where the count was taking place. The justification Warren officials gave was that they had received a report of a terrorism threat, namely a level 10 (out of 10) warning specifically for Warren County. That was a lie. It was denied outright by both Homeland Security and the FBI. So just what were those Republican Warren County officials doing with the ballots they kept to just themselves all night long?

One possibility is that they were destroying ballots.

Let's look at a few facts about Warren County. It is the second-fastest-growing county in the state (Delaware is first). Warren County went from a population of over 158,000 in 2000 to almost 182,000 estimated in 2003. At that rate, the county could have over 200,000 residents by now. These new residents may or may not share the same voting patterns as the residents who were there before 2000. In 2000, almost 70% of the votes cast for president in Warren County went to Bush. Among Ohio counties with populations exceeding 50,000, it is the "reddest" county in the state. Let's look at its election statistics:

Warren: registered voters in 2000: 96,536 reported voters in 2000: 70,109 turnout rate in 2000: 72.62% registered voters in 2004: 125,919 reported voters in 2004: 93,321 turnout rate in 2004: 74.11% increase in registered voters 2000-2004: 29,383 increase in reported voters 2000-2004: 23,212 percent increase in turnout rate 2000-2004: 2.05%

Once again, we see that odd pattern where there is a greater increase in registered voters than there is in reported voters. Let's first compare Warren County to the other red punch card states and then to only those that were the strongest for Bush in 2000. There are 54 counties other than Warren that use punch cards and that went for Bush in both 2000 and 2004. I won't list them individually, but here are their combined statistics:

Other "red" counties: registered voters in 2000: 2,662685 reported voters in 2000: 1,727,051 turnout rate in 2000: 64.86% registered voters in 2004: 2,798,385 reported voters in 2004: 1,989,860 turnout rate in 2004: 71.11% increase in registered voters 2000-2004: 135,700 increase in reported voters 2000-2004: 262,809 percent increase in turnout rate 2000-2004: 9.63%

Unlike Warren County, these counties show the expected pattern: a greater increase in reported voters than in registered voters. Note the turnout rate increase of 9.63% is much higher than Warren County's increase of only 2.05%. But what if we looked at only the "reddest" counties in the state would the numbers look any better? Let's see. There are only seven counties besides Warren that use punch cards and that gave Bush 65% or more of its presidential votes in 2000. They are Clinton, Van Wert, Delaware, Union, Mercer, Holmes, and Putnam.

Other "very red" counties: registered voters in 2000: 217,927 reported voters in 2000: 148,612 turnout rate in 2000: 68.19% registered voters in 2004: 251,035 reported voters in 2004: 186,603 turnout rate in 2004: 74.33% increase in registered voters 2000-2004: 33,108 increase in reported voters 2000-2004: 37,991 percent increase in turnout rate 2000-2004: 9.00%

Once again, we see the expected pattern of a greater increase in reported voters than in registered voters. Once again, we see a turnout rate increase much higher than Warren County's. How many votes would Warren County gain if it had enjoyed the same 9% increase in turnout rate as these other very red states? The turnout rate would be 79.16% and the vote count would be about 6,400 votes higher. If those were all Kerry votes, Warren would still be very red, just as it is now.

If 6,400 or so Kerry-voting ballots were destroyed, stolen, or hidden by Warren County officials during their secret count, how would anyone ever know? Of course it's only a theory that they eliminated ballots. If the low turnout is genuine, they still could have replaced or altered ballots. They could have spoiled some Kerry ballots by punching the holes for Bush or another candidate, thus creating overvotes. They also could have "un-spoiled" some undervote ballots by punching the holes for Bush. This would keep the spoilage rate from being too high. It's even possible they used a combination of these methods. With a count conducted in secret, how will we ever know? The "chain of evidence" has been compromised forever.

There is one more point to be examined here. The idea promoted by the media is that Bush won because "red" voters showed up in greater quantities than "blue" voters. But looking at Ohio, we see that the opposite is actually true. Blue counties in Ohio had a larger increase in turnout rate than red counties. Even dividing counties up into smaller and smaller groups does not change this: Strongly blue counties have better turnout rate increases than moderately blue counties. Moderately red counties have better turnout rate increases than strongly red counties. Where did all those votes for Bush come from? The election in Ohio stinks. Badly.

So now we have votes in three important counties that may have been lost forever. (There are several other counties that use DREs that I have not had time to study. It's possible these counties also have problems.) We have numerous reports of civil rights violations, Jim Crow, and other illegal forms of voter suppression throughout the state. With all these anomalies that may be uncorrectable through a recount, the only logical solution is to conduct a re-vote throughout the entire state using paper ballots and pens. Republican and Democratic volunteers in each precinct could be responsible for counting the ballots together. If other large countries such as Canada can conduct pen and paper elections and have the votes tallied within hours, there is no reason Ohio cannot do the same. I'm sure the Ohio Democratic and Republican parties will not have any trouble finding volunteers willing to participate in counting the votes.

There are some who will argue against a re-vote on legal grounds, saying that the US Constitution specifies the day for national elections. But I think when there is strong evidence of massive vote fraud, an arbitrary date becomes less important. You can be certain that if Ohio was won by a third party candidate under circumstances suggesting obvious fraud that both Democrats and Republicans would be crying out for a re-vote. The only question is whether Republicans have enough integrity to do the same when it is their own candidate who "won" under fishy circumstances. I think we already know what the answer is as far as the party's top officials go. But perhaps Republican voters in the state will actually pressure their party to do the right thing. Perhaps the rest of the nation will be vocal enough to demand a re-vote in Ohio. The alternative discarding Ohio's votes is probably not palatable to anyone. Accepting a slate of electors arrived at under dubious circumstances is probably palatable only to Republican leaders.