1996 Data on Atlas for Tyler County, West Virginia cannot not be an error

[America Needs a 13-6 Progressive SCOTUS]

Its simply statistically unbelievable that Clinton did so well in 1996 in this County to the extent that Atlas claims he did. It lists him as having won the County by a nearly 2 on 1 margin with a majority while Dole and Perot were almost tied. This simply makes no sense both when considering historical data and geographical statistics.

If this were not an error, it places the County as the 6th most Dem swinging County from 1992 to 1996 and the 1st most Rep swinging County from 1996 to 2000. In addition, this would make 1996 only 1 of 2 elections since West Virginia statehood where the County voted Democratic since West Virginia statehood, the other being 1912. Nothing in the history of this County or in candidate relations to the County really makes such a shift expected; its not like Russell County, Kansas where such (reverse) shifts can be explained due to it being the home County of Bob Dole.

In a geographical context, Tyler County has historically been part of a geopolitical triplet with neighboring Doddridge and Ritchie County. Neither of these 2 Counties made much of a significant shift at all from 1992 to 1996 (with Doddridge even slightly swinging against Clinton from 1992 to 1996), and neither had such an enormous shift from 1996 to 2000, even if they both shifted massively.

In terms of overall turnout, Atlas lists Dole 1996 votes with a total less than half the total of votes that Bush obtained in 1992. At the same time, votes for Bill Clinton actually decreased slightly from 1992 to 1996 in this County, which would make 1996 an extremely abnormally election for its massively low turnout from this County, as overall turnout would increase back to normal levels in 2000 in this County. Neither of the other 2 Counties listed here earlier had anything in the way of abnormal turnout in 1996.

Overall; I believe there are 2 likely explanations as to the nature of the error here.

1. A leading 1 was accidentally omitted from the total for Dole making it show up as 734 when it should have been 1734.

2. The Dole total was accidentally halved making it show up as 734 when it should have been 1468.

Note: I am using wikipedia totals to write this post; wikipedia uses atlas as a source

[America Needs a 13-6 Progressive SCOTUS]

I just attempted to contact the County Clerk of Tyler County, West Virginia for information about this through the email listed in the County Clerk Sections of the Official Election Results Secretary of State of West Virginia Website. Hopefully, this could help to determine what exactly happened.

[America Needs a 13-6 Progressive SCOTUS]

The data for Pres is definitely inconsistent with the other statewide contested race totals. However, those are the official totals certified by the state, so the error is not with the Atlas. The error appears to be with WV.

What do you think happens if the WV secretary of state agrees?
Will Mac Warner instigate a recount of the votes?
Are the ballots still archived anywhere?

I am hoping somehow that the County Clerk office has the correct totals on hand and that the incorrect totals were generated by a transcription error from the state myself I guess.

[America Needs a 13-6 Progressive SCOTUS]

The data for Pres is definitely inconsistent with the other statewide contested race totals. However, those are the official totals certified by the state, so the error is not with the Atlas. The error appears to be with WV.

What do you think happens if the WV secretary of state agrees?
Will Mac Warner instigate a recount of the votes?
Are the ballots still archived anywhere?

I am hoping somehow that the County Clerk office has the correct totals on hand and that the incorrect totals were generated by a transcription error from the state myself I guess.

That does happen sometimes. The SoS relies on the counties to provide results, and the data can get misread going from the county to the state. But it's also possible that they certified a report that had errors. If so, the erroneous report represents the official totals barring a court challenge to change it. At this point I can't imagine a court that would allow such a challenge since the changing the results has no effect on anything material to the state.

In any case, if there is an error they would probably tell anyone studying the results to use a footnote to explain the discrepancy rather than go through a court proceeding.

I am not sure which of these 2 reasons makes the case for a recount more absurdly stupid:

1. The election happened 23 years ago.
2. There is no chance that such a recount could result in a change of who won anything meaningful.

[America Needs a 13-6 Progressive SCOTUS]

Official results:

Clinton: 1459 (52.5%)
Dole: 734 (26.4%)
Perot: 563 (20.3%)
Browne: 24 (0.9%)

total: 2780

Do you think they merely forgot a "1" for Dole's figure?

Alternative results:

Clinton: 1459 (38.6%)
Dole: 1734 (45.9%)
Perot: 563 (14.9%)
Browne: 24 (0.6%)

total: 3780

1992 results:

Clinton: 1587 (37.8%)
Bush: 1593 (37.9%)
Perot: 1013 (24.1%)
Marrou: 10 (0.2%)

total: 4203

A turnout drop from 4203 to 3780 or even 3514 would not be particularly unusual to be honest. Turnout nationally indeed did drop as a proportion of the total population from 1992 to 1996 and West Virginia was not exactly leading the pack of states that were gaining population fast at the time.

[America Needs a 13-6 Progressive SCOTUS]

There has to have been a digit dropped. Tyler was solid Republican even when West Virginia as a whole was solid Democratic. Going by the official results, 1992 was the only year the Democrats won other than when Teddy split the Republican Party in 1912 letting Wilson get the county with just 37% of the vote. (Teddy was a close second.) The closest the Democrats ever got to a majority was in 1964 when LBJ got 47% to Goldwater's 53%.

If it were an accidental halving of the total, the County would have moved from a 6-vote Bush win in 1992 to a 9-vote Dole win in 1996. Not exactly the most shocking thing ever.

I know that doubling is among the more common types of vote reporting errors; not sure how frequent halving is though, and halving of odd numbers would probably be instantly detected in the next step of vote total processing unless it was somehow rounded up or down in the process. I am not going to take a side right now between whether a dropped digit or a halving was more likely.

[America Needs a 13-6 Progressive SCOTUS]

If it was a halving of the Pub total then that would add 734 to the total cast making it 3514. If 1000 is added then 3780 would have been cast. There were 3828 votes cast for Gov, so 3780 is consistent with that. It would be unusual for Pres to be in the same range as the downballot races, as it would if only 3514 were cast.

It might not make a difference here, but I also just found information while further looking into this issue that the 1996 Republican nominee for West Virginia Governor was from Tyler County. Not sure what the most reasonable thing to expect based on that would be.

[America Needs a 13-6 Progressive SCOTUS]

If you really want to check, go get the precinct results directly from the county itself.

I did attempt to look for that before I sent the email to the County Clerk which I still have not received a reply to for the record. I could not find it anywhere.

[America Needs a 13-6 Progressive SCOTUS]

I just attempted to contact the County Clerk of Tyler County, West Virginia for information about this through the email listed in the County Clerk Sections of the Official Election Results Secretary of State of West Virginia Website. Hopefully, this could help to determine what exactly happened.

For those wondering, I never got a reply on this email. To be honest, although I am still convinced that there is an error here, I will say that I can totally understand how the County Clerk might want to focus their attention to issues actually directly affecting the County today, whatever they are, over fixing an election result error from 23 years ago that has no real chance of actually doing much beyond slightly narrowing the statewide and national popular vote count, and of slightly changing the candidates number of Counties won.