Mitt Romney - Maine GOP Caucus Fraud

Last Updated : Mar 16, 2012

Summary

On February 11, 2012 the state of Maine held its caucus style Republican party primary election for President. On the morning of the election, Maine GOP officials announced that the elections in Washington County were postponed one week due to possible snow storms.

At 7:40 PM EDT on the night of the election Governor Romney was declared the winner by Maine GOP chaiman Charlie Webster. At the time of the announcement, Mr Webster stated that Governor Romney had beaten Congressman Ron Paul of Texas by less that 200 votes, but there was no doubt that he was the clear winner. When Mr Webster made the annoucement, Governor Romney led by roughly 3% of the vote, and only 84% of the vote had been counted due to the delay in voting in Washington county and delays in counting other areas.

Shortly after this annoucement the Ron Paul presidential campaign issued a statement noting that the weather predictions were nothing unusual for Maine and that Governor Romney should not have been annouced as the winner with so much of the vote left to be taken. The statement also implied that given Governor Romney's recent string of losses, the Maine GOP may have delayed the voting in Washington county so that they could annouce a Romney win that night and then quietly retract that declaration when the total number of votes were counted a week later to reveal Congressman Paul as the winner. Washington county saw only slight amounts of snow that day with activities by groups such as the Girl Scouts going forward as planned. That county was also won by McCain in 2008 and Congressman Paul was polling strongly there before the election.

In the days that followed the election, a number of forums and blogs began to discuss inconsistencies in the voting results put out by the Maine GOP. This culminated in a local news station airing a story in which one caucus chairman flatly asserted that the results were incorrect. The chairman claimed that rules were changed by the Maine GOP to prevent the caucus results from being read aloud. However, that chairman motioned at the onset of the meeting to disregard this rule and all present agreed. He then stated that he not only read the results aloud, but also tweeted them. When he checked the Maine GOP results page, he saw that instead of the votes for his municipality showing a win for Congressman Paul, the results showed a win for Governor Romney. He contacted the Maine GOP and was only able to correct the data after noting the tweet and his position as caucus chairman. While the Results for Governor Romney and Congressman Paul have been corrected, that municipality still has errors.

On February 16, Chairman Webster met with the state GOP Executive Committee. The chairman had previously asserted that the results of Washington county would not be counted as that caucus was held after the state deadline (caucuses had been run as early as Feb 4). This was to be done even though it was the state GOP and not the local caucus boards that moved the elections due to weather. In this meeting, the Executive Committe and the Chairman agreed to count Washington county.

The following day, the state released an updated version of results for the state that showed Governor Romney increasing his lead over Congressman Paul. That same day, Chairman Webster was interviewed by Politico and called supporters of Congressman Paul who suspected fraud "conspiracy theorists." He stated that there were indeed several clerical errors in the initial results and several small towns were left off the results. On February 24, 2012 the Maine GOP released a final version of updated results with Washington County included.

The three sets of results released by the Maine GOP on the 11th, 17th, and 24th were published in PDF form and are available below. An analysis of these results reveals that there were a large number of errors and inconsistencies that remain throughout the data sets. These errors point towards a caucus that can be described under the best terms as negligent in their documentation and book-keeping. Coupling these errors with the actions of state GOP to rush the annoucement of Governor Romney as the winner, the efforts to keep Washington County out of the final total, and the known cases of fraud within the state, there is ample evidence to call the election suspect and possibly even fraudulent.

After the Feb 24th update, the state had counted 87% of its votes and Governor Romney's lead was cut to 115 votes. However, more than a month after the election, no effort has been made to secure the remaining 13% of the votes. This makes Maine the only state to go more than a day or two without completing the count for at least 98% of the state. In fact, the results shown through AP and other sources still reflect the 87% totals as reported in the Feb 17th update and not the final updated results including Washington County. These Feb 17th numbers represent the sole point in the election totals where Romney's lead over Paul was highest and make Maine the only state not to be updated real time by the AP. In short, the AP and the Maine GOP are reporting old results that just happen to be the ones most favorable to Governor Romney. (Google "Maine Election Results" and compare those numbers to the results below to verify).

As stated, the results themselves are in PDF format and clearly show the election data was contained within a spreadsheet format similar to Excel. There are 435 rows with each row representing a municipality within a county. For each row, there are columns showing the number of votes for Governor Romney, the number of votes for Congressman Paul, the number of votes for Senator Santorum, the number of votes for Congressman Gingrich, and the total number of votes for the municipality. At the bottom of the spreadsheet is a row that shows the totals of each of these columns. In an error free spreadsheet, the total number of votes for all candidates could be added together to obtain the total number of votes for the state. This number would also match the summation of the column denoting the total votes for each municipality.

In all three sets of results issued by the Maine GOP on the 11th, 17th, and 24th, these numbers do not match. Within each spreadsheet, there are roughly 12 municipalities in which the total number of votes in that district does not match the total entered for that row. However, for each set of the results, the incorrect total that would have been calculated at the bottom of the spreadsheet is deleted and the "official" total has been over-written onto its place.

This raises serious questions as to the legitimacy of the results. Pulling the data into our own spreadsheets, we were able to ascertain that the numbers reported for both the individual candidates and the total number of votes matches the official results once the errors in the totals are corrected. However, this does not explain why the person making the spreadsheets was not alarmed that their totals were simply not matching across the board. The only two perceivable conclusions is that the person was not capable of correcting the errors and chose to hide them, or the more sinister reason - that the person making the spreadsheet already had a reason to believe that the totals may not match and acted to cover that fact. Either way, the Maine GOP should explain why they published data that was known to be incorrect.

The results given by the Maine GOP show the results for all municipalities for 15 of Maine's 16 counties. The results for Aroostoock county, the sixth largest county in Maine by population, is given as a single line with no data given for its 67 municipalities. Looking at the population data for each county, voter turnout in that county was 0.19%. This was far below the next lowest county turnout of 0.31% and the average turnout for the state of 0.51%. However, this was consistent with 2008 numbers. The results from this county show that Congressman Paul had more than 3 times the votes of his closest competitor - Governor Romney. So the one county which cannot be verified on a district by district level is the one county that Congressman Paul won in 2008 and the county in which his support was the largest in 2012.

Over the course of the three releases of voter results, a number of municipalities which initially had zero votes were updated to show the true results. These municipalities include ones with later voting dates and ones that voted on the 11th of February, but were mistakenly left off the initial results. Of the 435 municipalities shown, 177 were initially blank with that number being reduced to 156 in the February 17th update and to 104 in the February 24 update. In the end, Congressman Paul won 125 municipalities and Governor Romney won 113.

There were 15 municipalities which were not initially zero but still changed between the first and last release of the data. The end result of these changes was that Congressman Paul lost 36 votes and the remaining candidates gained 13, 12, and 11. This resulted in no change to the overall vote total. In one of those municipalities, the votes or Romney and Paul were flipped. In another, Congressman Paul lost 20 votes and those votes simply disappeared from the total for that district. Note that the flipping of the votes appears to be a legitimate correction, and is used here to show the chaotic nature of the election.

The total evidence of fraud is as follows:

  1. Congressman Paul's support was strongest in two counties: Aroostoock and Washingtonrn
    1. The Maine GOP postponed the election in Washington county under very questionable conditions and then stated that because of that delay those votes would not be counted. They eventually reversed this position.
    2. Aroostoock county is the only one which does not have precent by precent results available. It also had a voter turnout that was far lower than any other district and statistically out of sync with Maine results. However, this was consistent with 2008 results.
  2. The Maine GOP announced Romney as the winner when he was leading by less than 200 votes with entire counties yet to vote and numerous municipalities showing no votes where votes were known to be heldrn
    1. Governor Romney had recently suffered a string of losses and losing a north eastern state to a different rival would have been very detrimental to his campaign.
  3. The Maine GOP established rules to prevent caucus results from being read aloud at the caucus - a determinent to any attempting to verify their cacus results with the official results.
  4. Maine is the only state that went more than a few days without achieving a 100% vote count and sill remains at 87% more than a month later.
  5. The AP and other sources are still reporting the Feb 17 results and not the final Feb 24 results.
  6. There are documented specific cases of incorrect results which favored Governor Romney.
  7. The results presented by the Maine GOP remain incorrectrn
    1. The spreadsheets presented as the results contain errors that any nominal election official would have sought to resolve prior to publishing the data yet no effort was made to do this.
    2. This means that we cannot for sure what data is correct and what is incorrect.
  8. There are unexplained changes in data from one update to the next.rn
    1. Presumably, these changes are the result of the Maine GOP rechecking the results with the caucus chairs.
    2. The results of these changes harmed only Congressman Paul and benefitted every other candidate evenly.

 

Washington County Postponed

On the day of the Caucus, the Republican party announced that the Caucuses for the eastern part of Maine, particularly Washington County and a few other municipalities would be postponed by one week due to a prediction of inclement weather and a possible snow storm. Other municipalities along the east coast of Maine were also closed due to inclement weather. The eastern portion of Maine was the area that Governor Romney did not win in 2008.

 

Romney Declared the Winner

At roughly 7:40pm EDT on the night of the election, Maine GOP Chairman Charlie Webster declared Governor Mitt Romney as the winner of the Maine caucus, stating that the election was decided by less than 200 votes. At the time of the announcement, news agencies were reporting that only 84% of the votes had been counted, meaning that Congressman Paul could have easily made up the 3% difference within the remaining votes to be counted and the votes yet to be held in Washington county. During the announcement, Chairman Webster stated that some counties and municipalities had chosen to hold their elections at a later date and would not be counted. This was the county that the Maine GOP had postponed due to the weather.

 

Response by Congressman Paul

After the announcement that Governor Romney had won the Caucus, the campaign of Congressman Paul issued a statement noting that a large percentage of the state had yet to vote or be counted and the winner should not be annouced until all voting had occurred. The campaign accused the Maine GOP of delaying the votes in Washington County until a week after the election because the GOP was aware that the votes from that county would likely make Congressman Paul the winner and the establishment wanted Governor Romney to be announced as the winner even if this was not the end result. The announcement also noted that the weather that was forecast in Maine was not unusual for the area and that in the end Washington county received only a dusting of snow.

 

Initial Caucus Fraud Accusation

On February 16, 2012 a local news station ran a story on a caucus in Waldo County that was taken on February 4. The news station spoke with the chairman of one of the caucuses in that county, and the chairman confirmed that the state GOP establishment had set rules to prevent the results from being read aloud. The chairman of that particular caucus moved to ignore this new rule at the onset of the meeting and read the results aloud after the vote. He also tweeted those results. Days later, that Caucus chairman found that the votes in his municipality were recorded incorrectly and had Governor Romney as the winner instead of Congressman Paul. He contacted the Maine GOP and was only able to correct the results after noting his public announcement on twitter.

In the most recent version of the results, the votes for Romney and Paul are fixed and reflect the numbers put forth by the chairman. However, even after the news report and the publicity the story generated the end results still reflected an incorrect number of undecided voters and total votes.

 

Maine Executive Committee Meeting

That same day - February 16, 2012 - Charlie Webster met with the Maine Executive Committee and discussed the results of the Presidential Preference Poll. He stated in a press release that the Maine GOP had worked to contact the town chairmen and reconfirm the results of the poll. He noted that in this meeting, it was decided that the later votes from Washington County and other areas would indeed be counted in the total for the state.

 

Politico Interview

On February 17, 2012 Maine GOP Chairman Charlie Webster was interviewed by Politico and discussed the election results. He stated that there were indeed some clerical errors, but that none of the overall results would be changed due to the corrections. He stated that an updated version of the results would be out soon.

Mr Webster continued the interview by stating that complaints concerning missing votes in Waldo county were understandable, but that most votes within that county went to Santorum and that they wouldn't have any effect on the overall result.

When asked about the accusations by the Ron Paul campaign and his supporters that the results of Washington County were withheld to allow for Governor Romney to be announced as the winner on election night, Webster called such ideas conspiracy theories and stated that he was insulted by the idea that he was not impartial.

 

New Results Issued

On the same day as the Politico interview, the Maine GOP did indeed issue updated results for the Caucus. A week after that release, they issued another set of updated results. So the initial results were issued on February 11 with an updated version released on February 17 and a final set of results issued on February 24.

 

The Results

The Maine GOP provided PDF versions of spreadsheets that held the voting data for each municipality and each county. One PDF was released on February 11, 2012 - the day of the election - and the other two were released on the dates mentioned previously of February 17 and February 24. The total number of votes for each candidate is shown in the table below along with the number of municipalities that still had zero votes counted.

On the election day, 177 of the 435 municipalities had zero votes recorded. As votes were counted in Washington County and other areas that number lessened with the updated versions, going to 156 municipalities on February 17 and 104 on February 24.

As more municipalites were counted, Congressman Paul logged more victories. Although he lost the total number of votes, he did eventually end up winning more municipalities than Governor Romney. The final table below provides links to the PDFs of the results as provided by the Maine GOP.

 

 

Results for February 11, 2012
Results for February 17, 2012
Results for February 24, 2012

 

Problems with Results - Updating

For the 2012 Presidential elections no state other than Maine has gone more than a day without counting at least 98% of the votes. However, more than a month after the votes in Maine the results supplied through Google and AP still show that only 87% of the votes have been counted. Even those results are incorrect as they reflect the results of the February 17 update and not the results of the February 24 update. This version of results, which is shown to anyone who googles something like "Maine GOP results," reflects the data set showing the largest margin of victory for Governor Romney and the results prior to counting Washington County. (Last googled on March 15, 2012).

 

Problems with Results - Totals

An analysis of the PDFs put out by the Maine GOP revealed a number of inconsistencies within the data. The first of these problems was that the totals listed within the documents do not match the true totals. For each municipality, the PDF provides a row that lists the number of votes for each candidate and the total number of votes in the municipality. For a number of municipalities, the total number of votes listed does not match the true total for that area. In the February 11 results there are roughly a dozen municipalities that contain totals that differ from the true value. For example, in York County the municipality of Kennebunk lists 29 votes for Governor Romney, 5 votes for Congressman Paul, 11 votes for Senator Santorum, and 2 votes for Congressman Gingrich. The total number of votes for that municipality was therefore 47, but the total listed is 45. As another example, in the County of Penobscot, the municipality of  Bangor has zero total votes listed even though there are 77 votes recorded in that area.

Of course, an error in the total number of votes for any municipality is not substantial as long as the votes for any given candidate is correct. However, when the data provided by the official source constradicts itself, it indicates a deeper problem. We cannot be sure that the sole problem in the data rests in the totals column.

The results given by the Maine GOP are in PDF forrm and imply that the data was held in a spreadsheet. This terminology of the data being held in the "spreadsheet form" - similar to Excel - was used often in discussing the results. At the bottom of each PDF is a row that shows the total number of votes for each representative, the total number of votes for the state, and the percentage of votes for each candidate. The obvious implication being that those sums were obtained by adding the data in each column from the rows above. However, given the errors noted above in the totals, if a spreadsheet was used to calculate the totals for each candidate and the total number of votes, those two numbers will not match.

The image below is a screen capture from the PDF released by the Maine GOP on the day of the election - February 11. The first two rows in the image contain the voting results for the last two entries in the dataset, which includes the county of York and the municipalities of Wells and York. The third row is a summation of all 435 municipalities that held elections in Maine and it shows the proper results for all candidates as well as the correct number of total votes at 5585. The obvious implication being that those totals were obtained by summarizing the data above. The problem with these results is that if a true summary is done of the total number of votes column, then the spreadsheet would have produced a total number of votes of 5493 in the far right column due to the errors listed previously. This means that the creator and/or publisher of the results overwrote the total number of votes to the correct number of 5585 without attempting to address the inconsistency in the spreadsheet and without attempting to confirm that totals for each representative was correct. Given that the 5585 number is also written in the far left column, it seems likely that the creator of the spreadsheet added the total number of votes for all candidates and arrived at the number of 5585 and then changed the sum when it did not match.

The PDFs presented by the Maine GOP for the later results did correct some of the inconsistencies in the totals, but most remained. Both PDFs were still presented as if the total contained within the spreadsheet matched the total shown at the bottom and the total number of votes from each candidate. The screen captures from the later dates are shown below. The true total number of votes within the February 17 results was 5732 and not 5814 as shown while the February 24 results had a total number of votes of 6172 and not the reported 6250. Note that in these later results, the total number of votes is moved out of the column for the totals and only appears in the column which should denote the municipality. Again, we believe that this is because the person creating the sheets added the total for each candidate and then wrote that number in the far left without addressing why the numbers did not match.

Again, the total number of votes for each candidate is consistent with the total presented to the public. This means that you can take the spreadsheet and sum the column for the number of votes for any given representative and come out with the number shown at the bottom of the spreadsheet. However, summing the total number of votes does not match the total shown at the bottom of the spreadsheet.

This error in totals was not consistent from one release to the next as some errors in the total number of votes were corrected between the Feb 11 release and subsequent releases while other errors were not. This means that while the Maine GOP was furiously revalidating the data with Caucus chairmen as Chairman Wilson stated, no one attempted to address the fact that the actual data being presented to the public contradicted itself.

 

Problems with Results - Counties and Numbers

The state of Maine is comprised of 16 counties. The county of Aroostook contains 67 municipalities, has the 6th largest population in the state, and makes up most of the nothern section of the state. In the three sets of data presented by the Maine GOP, the results are given for every municipality within every county with the exception of Aroostook. The results for this county are provided as one line of data.

The results for Aroostook county show that it strongly favored Congressman Paul, and that he had more than three times the number of votes as his nearest competitor. The results also show that this county lies well outside the statistical norm for voter turnout in the state. In Aroostook county only 0.19% of the people cast a vote in the Caucus. The next lowest county had 0.31% of the population cast a vote. With an average turnout of 0.51% across the state, counties won by Congressman Paul were typically won with much higher than this nominal turnout. However, the county in which the Congressman is most statistically ahead provided the lowest statistical turnout. The fact that the results of this county are not shown for each municipality means that this is the only county in which Caucus leaders cannot personally verify their results.

The highest per capita turnout came from Washington County, the same County that the Maine GOP Chair stated would not provide a statistical difference on the day of the election. In that county, almost 1% of the population cast a vote and Congressman Paul almost doubled Governor Romney in votes.

 

Problems with Results - Changes

Between the initial publication of the results on February 11 and the final results on February 24, there were 15 districts in which the results changes with no explanation. The assumption would be that the Maine GOP confirmed the numbers with the local caucus leaders and corrected errors. However, if the Maine GOP did revalidate the caucus numbers, they did not attempt to address the issue of their totals not matching in the spreadsheet. They also did not manage to fix the voting numbers in the municipality shown in the previous news clip. The end result of these changes was no change in the total number of votes, but a reduction in the total number of votes for Congressman Paul by 36 and an equal increase in all three remaining candidates. The two most glaring changes were in the municipality of Portland in Cumberland County where the votes for Romney and Paul were flipped from one set of results to the next, and in Limington municipality in York county where Congressman Paul lost all 20 of his votes and the total number of votes for that distict went from 26 to 6. Numerous discussion boards have maintained that the error in Cumberland county was valid and that it was found shortly after the results were posted.

References

[1] Website: Politico.com Article: Maine GOP: Results went to spam Author: Tim Mak Accessed on: 03/15/2012

User Comments