News: Election 2004: Preliminary Recount Results Suggest That, Conservatively, Statewide Hand Recount Would Result in
50,000+ New Votes Being Counted
The Nashua Advocate
By ADVOCATE STAFF
It seems the mainstream media is not a math-driven medium.
The Advocate has pointed out this phenomenon before, most recently when The Washington Post reported that an estimated
10,000 votes were lost in Columbus alone due to long lines -- and then crowed that such a circumstance likely had no
effect whatsoever on the result of the presidential election.
Well, that's true -- unless, of course, one is on friendly terms with elementary mathematics and grade-school geography,
which informs the literate populace that Columbus inhabits only a small land-area in Ohio, and contains well less than
10% of its population.
And that fractions must be multipled to create whole numbers.
In a story released today, the Associated Press continues the mainstream media's frustrating penchant for paying no
attention whatsoever to the mathematical implications of its own reporting. Specifically, the Associated Press is
reporting that, with 51.1% of Ohio's counties (45 of 88) having completed a 3%-of-ballots manual recount and a
100%-of-ballots machine recount (with a mere handful of these counties having completed also a 100%-of-ballots
hand-recount), 746 new net votes have been uncovered. Virtually all of these new votes are attributable to the hand-,
rather than machine-recounts, according to Pamela Swafford, Deputy Director of Elections in urban Hamilton County. Ms.
Swafford has "attributed the differences [in vote totals] to chads falling out of punch-card ballots during
re-handling." [Emphasis supplied].
Certainly, it's possible that the 45 counties thus far counted represent 51.1% of the state's counties, but much more
than 51.1% of its population; it's possible that some of the "chads falling out" were caused by vote-counting machines,
not hand-tallies; and it's possible that the new net votes are close in number to the new gross vote totals, which have
not yet been released to the public.
But isn't it also possible that the number of new net votes obscures a much larger number of new gross votes? [E.g.,
were Kerry to gain 500 votes then lose 200 votes, his net gain is 300 votes, but 700 gross votes have actually been
"newly" tabulated, or "found," by recounters]. And isn't it possible that 51.1% of the counties means something close to
51.1% of the state population, or, perhaps, even less than that, meaning that the prospective statewide vote gains for
each candidate are underrepresented by the limited recount thus far? And isn't it possible that Ms. Swafford is correct,
and most of the so-called "new" votes were discovered during the 3% hand-recounts? And finally, given that the 3%
recounts were done largely in hand-picked precincts where vote tabulation errors were expected to be few in number,
isn't it possible that the number of "new" ballots found in those precincts represents much less than 3% of the "new"
ballots which would be found in a countywide manual recount of the counties in question?
These questions, at present, are unanswerable.
But The Advocate presses ahead, regardless, with some startling -- yet not at all unreasonable -- speculation.
Specifically, The Advocate notes that if (only quasi-hypothetically) a manual count of 3% of the ballots from 50% of the
state's population has revealed approximately 750 new net votes, then the total number of net votes which could be
revealed in a statewide manual recount exceeds 50,000. And if one presumes that new gross votes in fact vastly exceed
(as would be expected) new net votes, that total could be, in reality, closer to 100,000 new ballots which were not
counted on election day.
When one considers, further, that 92,000 Ohio undervotes remain to be counted; that Dr. Werner Lange's statistical
analysis of Ohio absentee balloting shows that 62,500 illegal absentee ballots might have been cast in Ohio on election
day; and that The Washington Post's "lost vote" analysis suggests that as many as 220,000 votes might have been lost --
in predominantly Democratic wards -- because of long lines caused by negligent or discriminatory machine-placement, it
becomes significantly easier to understand Kerry's Ohio defeat on election day, even nothwithstanding the fact that many
electronic voting machines in Ohio may have been tampered with by agents of private corporations. Indeed, even the exit
polls, which showed Kerry winning the state of Ohio by several points, begin to seem, under this analysis, as accurate
as the nation expected they would be at 6 A.M. on November 2nd.
Of course, even estimating that 100,000 new gross ballots remain to be counted seems a low estimate -- given that the
hand recounts thus far conducted have targeted well-run, efficiently-managed precincts where uncounted votes were few.
The Associated Press's response? "Neither campaign expects the recount to change the outcome. The Republican Bush won
Ohio's 20 electoral votes by a margin of about 119,000."
But from Ms. Swafford, a different accounting of how easily "new" votes are found on hand-recounts of paper ballots (the
ones with the now-infamous "hanging chads"): "Just barely touching them, they pop, they come out. That's not unusual.
When you're talking about that many ballots, that's bound to happen."