MASSACHUSETTS 2020 DEMOCRATIC PARTY PRIMARY Exit Poll Versus Reported Vote Count

Please share this article:

Table updated with latest vote counts on March 8, 2020. Text edited accordingly and for clarity.

By Theodore de Macedo Soares

The 2020 Massachusetts Democratic Party presidential primary was held on March 3, 2020. Election results from the computerized vote counts differed significantly from the results projected by the exit poll conducted by Edison Research and published by CNN at poll’s closing. As in the 2016 Massachusetts primary between candidates Sanders and Clinton, disparities greatly exceed the exit poll’s margin of error. Sanders won Massachusetts in the exit poll and lost it in the computer count.

The discrepancies between the exit poll and the vote count for Sanders and Biden totaled 8.4%— double the 4.0% margin of error (95% CI) for their exit poll differences. Warren’s and Biden’s discrepancies also totaled 8.4%, again double the margin of error (95% CI) for their exit poll differences. . These discrepancies replicate the total discrepancy of 8.0% favoring Clinton in the 2016 Massachusetts Democratic Party primary between her and Sanders. This time two progressive candidates exhibit the same discrepancies now favoring Biden representing the establishment’s choice.

Presidential candidates Biden’s and Bloomberg’s vote counts exhibited the largest disparity from their exit poll projections. Biden’s unobservable computer-generated vote totals represented a 16.2% increase of his projected exit poll share. Given the 1,397,222 voters (97% reporting to date) in this election, he gained approximately 65,200 more votes than projected by the exit poll. Bloomberg increased his vote share by 29% and approximately 36,900 more votes than projected. Their gain came largely at the expense of candidates Sanders and Warren whose combined vote counts were 104,300 less than projected by the exit poll.

Noteworthy is the fact that the 2016 Massachusetts Republican Party exit poll taken at the same time and at the same precincts as the Democratic Party primary, and also with a crowded field of five candidates, was matched almost perfectly by the computer count—varying by less than one percent for each candidate.  

Exit polls are widely recognized—such as by, for example, the United States Agency for International Development  (USAID)—as a means for checking the validity of vote counts. The U.S. has financed exit polls in other countries to “ensure free and fair” elections.

The United States remains one of the few major democracies in the world that continue to allow computerized vote counting—not observable by the public—to determine the results of its elections.[i] Countries such as Germany,[ii] Norway, Netherlands, France,[iii] Canada,[iv] United Kingdom, Ireland, Spain, Portugal, Italy, Denmark, Sweden, Finland and many other countries protect the integrity and trust of their elections with publicly observable hand-counting of paper ballots.[v]

[1] Exit poll (EP) downloaded from CNN’s website by TDMS on election night, March 3, 2020 at 8:00 PM. Candidates’ exit poll percentage/proportion derived from the gender category. Number of EP respondents: 1,394. As this first published exit poll was subsequently adjusted towards conformity with the final computerized vote count, the currently published exit poll differs from the results above.

[2] Candidates’ percentage/proportion of the total computer-generated vote counts derived from reported counts (97% reporting) updated on March 8, 2020 and published by The New York Times. Total number of voters: 1,397,222

[3] The difference between the exit poll proportion and reported vote proportion for each candidate (subtracting values in column two from the values in column three). A positive value indicates the candidate did better and received a greater share of the total reported count than projected by the exit poll. For example, candidate Biden, reported percentage/proportion of the total vote increased by 4.7% compared to his exit poll share.

[4] This column shows the percentage increase or decrease from the candidate’s exit poll projection (difference in column four divided by exit poll proportion in column two).  Shown only for candidates with 4% or more share in the exit poll. 

[5] This column presents a distinct Margin of Error (MOE) of the exit poll (EP) differences between candidate Biden and each of the other candidate’s EP results. The MOE , for example, between Biden and Sanders is 4.0% and the MOE between Biden and Warren is 3.9%.  For simplicity MOE not shown for candidates with less than 4% share in the EP.  MOEs calculated at 95% confidence interval (CI) according to multinomial formula in:  Franklin, C. The ‘Margin of Error’ for Differences in Polls. University of Wisconsin, Madison, Wisconsin. October 2002, revised February 2007. Available at:  https://abcnews.go.com/images/PollingUnit/MOEFranklin.pdf 

[6] The disparities between the exit poll and the reported computer-generated vote counts comparing Biden and each of the other candidates (subtracting each candidate’s difference between exit poll and computer count from Biden’s difference of 4.7%.). Disparities for candidates Sanders and Warren are double their respective MOE. For example, candidate Biden’s unverified computerized vote count exceeded his EP projected vote proportion by 4.7% while Sander’s computerized count understated his EP projected vote proportion by 3.8% for at total discrepancy of 8.4%. This 8.4% disparity, greatly exceeding the 4.0% margin of error based on their exit poll proportions, is statistically significant and it cannot be explained by the MOE.


ENDNOTES

[i] Fittingly, according to a recent Gallup World Poll, only 40% of Americans say they are confident in the honesty of U.S. elections. Finland and Norway with 89% of their citizens expressing confidence in the honesty of their elections along with the citizens of 25 other countries have greater confidence in their elections than do Americans.

[ii] In 2009 the Federal Constitutional Court of Germany ruled that every important aspect of an election must be observable by the public and thus “meet the constitutional requirements of the principle of the public nature of elections.” The Court explicitly ruled that no amount of voting machine testing, security requirements, and licensing procedures can compensate for this constitutional requirement.  With this ruling, Germany abandoned inherently unobservable computerized vote counting and reverted to the hand-counting of every ballot in the precincts in which they were cast and in the plain view of the public.

The court also noted that while vote fraud with hand-counted ballots would be easy to detect, “programming errors in the software or deliberate electoral fraud committed by manipulating the software of electronic voting machines can be recognized only with difficulty.”

[iii] During the 2007 presidential election, eighty-three municipalities (France has 36,569 municipalities) were allowed to use voting machines. Due to security concerns and the inability of voters to determine if their votes are counted correctly a moratorium, that remains today, prevents additional municipalities from introducing voting machines. In the 2012 elections only 64 municipalities continued their use. The French government desires a total ban on their use.

[v] The United States’ long ballots–containing federal, state, and local races–are commonly cited as being unwieldy for hand-counting. The use of Sweden’s method of providing different colored paper ballots for federal, state, and local races that are then sorted prior to hand-counting addresses this objection and allows for at least the hand-counting of federal elections with only three races per ballot.

[iv] In Canada, the results of federal elections are determined exclusively by hand-counted paper ballots.  Some provinces have adopted voting machines for local elections. See here, here and here.



Massachusetts 2020 Democratic Party Primary Exit Poll. Published by CNN at poll’s closing on election night.

Note: The exit poll vote proportions for each candidate was derived from the gender category. Candidate’s share of the male vote was multiplied with the total male proportion and added to the candidate’s proportion of the female vote multiplied with the total female vote to arrive at the candidate’s exit poll share in the state.

Comments below that may be helpful to the reader:

Why the exit polls accessible today differ from the exit poll used here.

Explanation of the method used to calculate the margin of error appropriate for an election with multiple candidates.

On errors in the conduct of an exit poll as the source of the disparities between the exit polls and the unobservable computer counts

Please share this article:

236
Leave a Reply

avatar
77 Comment threads
159 Thread replies
3 Followers
 
Most reacted comment
Hottest comment thread
167 Comment authors
ChuckBobChuckBobChuckBobChuckBobTerrance Newton Recent comment authors
  Subscribe  
Notify of
Lee C
Guest
Lee C

Theodore,

Thank you for this valuable work. Your above report was posted on Twitter. There, @Taniel, a journalist who specializes in elections and political theory, posted the reply below – I think suggesting that the variations were within the M.O.E. Could you possibly comment or clarify?

@Taniel’s reply”

‬⁩ a 4.0% margin of error (going off of this tweet) affects each candidate’s INDIVIDUAL numbers: so Sanders 36 would mean 32-40, and Biden 40 would mean 36-44, for instance. so that covers an 8% *net* difference (=in the margin). see below.

pewresearch.org/fact-tank/2016… pic.twitter.com/uPn45jms2r

Anonymous
Guest
Anonymous

Thanks for your careful analysis and details of your margin of error calculations. What about differential non-response in exit polling? The margin of error is only with respect to the population that would reply to an exit poll– not all voters. How big of an issue do you think that is?

Cindy
Guest
Cindy

Ted, Thank you so much for doing these! Are you in touch with Bernie Sanders and or someone in his inner circle about these reports? We need him to see the info asap and we need to encourage him to investigate and speak out on this issue. Elizabeth Warren and Tom Steyer as well.

Art
Guest
Art

The CNN exit poll cited by Soares is very different from the table provided (made up?) by Soares. There is something fishy here. See:

https://www.cnn.com/election/2020/state/massachusetts

aiuurgnuyerg
Guest
aiuurgnuyerg

the first footnote addresses this. the exit poll was later adjusted to match the computer count more closely

art
Guest
art

No. The footnote cites a CNN exit poll done in South Carolina in late February. It does not cite the CNN exit poll in Massachusetts in early March. The latter exit poll is pretty much in line with the vote. Check it out. There is something fishy going on with this guy Soares and his outfit. What is the name of his “research company?” Do a google search and you will come up with nothing.

art
Guest
art

There is no evidence at all that this guy Soares and his “research company,” which may not even exist, ever did an exit poll on voters.

Let’s not be gullible.

Dany
Guest
Dany
art
Guest
art

I don’t deny that Soares exists. But does his company? And if so, did that company ever do an exit poll? No evidence for that. If he does exist, who is he, what are his credentials? I asked about the existence of his company, not ab0ut his existence. But this is getting interesting.

Mikhail Bakunin
Guest
Mikhail Bakunin

It is known that CNN and other media, alter exit polling to fit the vote count. https://youtu.be/R6x8Z4NLUAA

art
Guest
art

It is not known at all. Your source, not surprisingly, is RT, the Russian funded tv outfit. There is no evidence that CNN adjusts its exit polls to align with the actual vote and the guest on the utube clip offers no proof of that.

Concerned Berner
Guest
Concerned Berner

This isn’t “altering” data. It’s removing the weighted factors once a sizable and sufficient amount of data have come in. They use weight factors until they have enough data. The data that Ted is using is more altered because he’s using incomplete data sets with high weight factors. Once the final data comes in, they are able to remove these hypothetical weight factors. Ted knows this and is only using the earlier data bc it fits his narrative. He’s a piece of shit

James
Guest
James

It was his data that suggested Russian hackers altered state election results to favor Trump over Clinton. So you can’t have it both ways.
https://www.google.com/amp/s/www.alternet.org/2018/07/reporter-thinks-russian-hackers-changed-votes-trump-and-his-explanation-doesnt/amp/

sherra
Guest
sherra

He’s not claiming to have done an exit poll. He’s using the existing exit polls, the original numbers not the adjusted numbers they do to match results.

Mikhail Bakunin
Guest
Mikhail Bakunin

Art, the links is to South Carolina but there is a drop down box that will take you Massachusetts

Dale Tavris
Guest
Dale Tavris

Soares’ first footnote to the table on the MA exit poll says “As this first published exit poll was subsequently adjusted towards conformity with the final computerized vote count, the currently published exit poll differs from the results above”. CNN always does this — they always adjust exit polls to match the official election results.

art
Guest
art

there is no evidence that “CNN always does this.” Soares says this in his footnote but it is not true.

Curiouser
Guest
Curiouser

Art, how do you know this is not true? Neither of you give any reference for your assertion. I am interested in pursuing this angle. If adjustments ate made to the exit poll data to provide conformity, what then is the point of the poll?

Cindy
Guest
Cindy

Exactly, Curiouser!

Mev
Guest
Todd Vickers
Guest
Todd Vickers

You have to select Massachusetts in the drop-down window.

Just the Facts
Guest
Just the Facts

So he says.

Concerned Citizen #62317
Guest

adjusting the polls to meet the outcome is a known thing.
see here: https://www.reddit.com/r/CAVDEF/comments/5d5f3c/are_the_2016_unadjusted_exit_polls_correct/

Just the Facts
Guest
Just the Facts

…and that should be in the body, not a footnote.

Tom
Guest
Tom

The footnote is embarrassing. The author seems to not know that if the margin of error for Biden’s result is +/-4, and the margin of error for Bernie’s result is +/-4, then the margin of error in a head to head match-up is +/- 8.

Carl Johnson
Guest
Carl Johnson

MoE cannot be reliably calculated. The footnote references a PDF with the formula utilized. BUT, the formula is only accurate if the population was determined with “Simple Random Sampling”. CNN Chooses what locations to poll and how many people to poll at each. Since most polling places are never surveyed, then the entire population doesn’t have equal chance at selection.

There’s good reason CNN doesn’t determine MoE on exit polling, because it’s impossible with this polling technique.

boggis the cat
Guest
boggis the cat

That isn’t how those statistics work. A 4% MOE applies across the board, so if the results between two candidates vary by more than that in total it’s outside the MOE. (The 4% could apply to one or the other, or be split across all of the candidates.)

It’s also odd that Warren would also be down by more than the MOE.

Carl Johnson
Guest
Carl Johnson

Worse, MoE cannot even be calculated because an exit poll doesn’t utilize Simple Random Sampling, where the entire population is given an equal chance to be represented. CNN chooses what sites to poll. Since a significant population is not subject to random sample then MoE isn’t in play. This is an erroneous application of statistics.

Cindy
Guest
Cindy

Actually, CNN does not choose the sites to poll. Edison Research does. Further, CNN is not the only entity that orders this exit poll… there are at least 6 media giants that do including CNN, NBC, ABC, CBS, AP, and FOX.

BERTAL AKTAS
Guest
BERTAL AKTAS

sites are chosen randomly, so your argument is nonsense

Carl Johnson
Guest
Carl Johnson

It’s still not Simple Random Sampling, even if the polling sites are chosen randomly. Every precinct has a lean one side or the other. Selecting certain sites locks your results into a certain lean depending on where they are at.

Real simple random sampling makes the entire population, regardless of precinct, subject to polling. This reduces errors. Again, there’s a reason why CNN doesn’t release an MoE, it’s inappropriate given the sampling methods.

BERTAL AKTAS
Guest
BERTAL AKTAS

You are dumb, margin of errors is for the pool, not candidates.

GregJ
Guest
GregJ

The data cited by Soares is time stamped at 8:00 PM ET and contains 1394 respondents. The data currently available on CNN’s web site is time stamped 10:16 PM ET and contains 1443 respondents. Assuming all 49 additional respondents were for Biden, these numbers align.

BERTAL AKTAS
Guest
BERTAL AKTAS

That means they added those to cook the numbers post-fact

Jon Sneyers
Guest

Not quite. The exit polls initially gave Biden 28.9% with n=1394 (so let’s say 403/1394 voted Biden), and then they processed 49 more exit poll results, and assume by some massive coincidence, ALL of them happened be votes for Biden. That still gives you only 452/1443 = 31.3%. There is no way you can arrive at 34%.

So no, it’s not just “we didn’t process all the exit polls yet”. It is more likely they indeed “adjusted towards conformity”, which makes total sense if you want to answer questions like “what is the influence of gender/age/race on voting behavior” as accurately as possible. But if you want to study election integrity, you should look at the “raw” results, i.e. the actual exit poll results, not the ones that have been adjusted towards the reported actual voting results.

Mike Anderson
Guest
Mike Anderson

Exit polls are not used for election integrity, that is not their purpose. USAID says specifically:

“Verifying official results: Exit polls do not provide sufficient evidence to refute or challenge official results, either at the national level or for individual polling stations.”

ChuckBob
Guest
ChuckBob

Oh, they certainly are, just not in the good ole USA. Look up EVEP exit polls. They’re admissible in court.

Just the Facts
Guest
Just the Facts

No he hasn’t. Because he literally made his numbers up. Check the link he provided to CNN. Note that Biden averaged 34% on the exit polls (NOT 28.9% as TDMS asserts) and Sanders around 28% (NOT 30.4 % as TDMS claims). Same pattern applies in his report on South Carolina. If you misquote sources and make up your own numbers, you can convince gullible folks of just about anything. It’s bad enough when this White House invents its own facts, but I expect better from the rest of America. https://edition.cnn.com/election/2020/entrance-and-exit-polls/massachusetts/democratic

Ian
Guest
Ian

as he explained above: “As this first published exit poll was subsequently adjusted towards conformity with the final computerized vote count, the currently published exit poll differs from the results above.”

art
Guest
art

You are completely correct. The claim that CNN adjusts its exit polls to align with the voting results is right out of the playbook of RT, the Russian funded TV propaganda arm.
See: https://youtu.be/R6x8Z4NLUAA

Sandy
Guest
Sandy

@art – you seem to make major efforts here to discredit the MA comparison to exit polls and now, having received several replies, you pull the RT line. Hopefully you realize this has just tipped your hand, right?

An aside – it is not even just the exit polls that cast doubt on the final vote counts. There are several additional indicators (which I will not enumerate for you since you are not here to be convinced or to debate). The nice thing about statistics is that often it’s just a way to bolster common sense, but that’d be way too philosophical an argument, so……

James
Guest
James

@art I thought it was just Trump and his supporters that cry fake news when it doesn’t seem to fit their narrative.

Jaro
Guest
Jaro

https://edition.cnn.com/2020/02/03/politics/what-are-exit-polls/index.html

CNN on their exit polling. Last paragraph:

“That’s just because we get more data through the day and as it’s weighted, it gets more precise and refined. We’re just taking the best information we have and refining the results throughout the evening.” After polls close, data is weighted to the official final numbers.

ChuckBob
Guest
ChuckBob

He didn’t make up anything, try reading the man’s explanation.

Alex
Guest
Alex

Can we contact the ACLU or some legal organization about this?

Cindy
Guest
Cindy

There are lawyers who are highly specialized experts in election integrity who also work with other EI experts as well. I think the Bernie Sanders campaign has the resources to hire them and I think he should ASAP. Tom Steyer would too of course and Elizabeth Warren. In fact, maybe they should team up. I do like the idea of contacting the ACLU about this too to represent voters disenfranchised.

Cindy
Guest
Cindy

You know it just occurred to me that Faiz Shakir worked for the ACLU so that might make some coordination there easier. Liking that idea better all the time 🙂

William Rosmer
Guest
William Rosmer

The problem is this article. According to a judge ruling in florida from the 2016 election the dnc is not required to respect the primaries at all. https://observer.com/2017/08/court-admits-dnc-and-debbie-wasserman-schulz-rigged-primaries-against-sanders/

Jimbo Splice
Guest
Jimbo Splice

That court case was in reference to private affairs, such as caucuses. Primary elections are publicly funded and fraud in them constitutes a federal crime.

Cindy
Guest
Cindy

Can others chime in on this distinction…. is it true? Has this been tested in court? If so, maybe that gives us a bit more leverage.

strangequark
Guest
strangequark

The Federal Election Commission

Cindy
Guest
Cindy

John Brakey has suggested we request they save all digital ballots and give us digital copies so we can do a recount. IMO, if enough of us put in these requests it could make a big difference in their willingness to tamper with the results. Even more effective in my opinion, would be for the CAMPAIGNS to write open letters requesting this… and let their supporters make this REALLY PUBLIC.

Noam
Guest
Noam

Problem is, the data has to be reliable, and I’m not willing to trust that they can’t think of creative ways to lie and make it hard to parse anyways.

Cindy
Guest
Cindy

I think Ted’s reports on exit poll data versus reported vote totals become increasingly powerful each time he releases a new one. If they all show the same trend… the establishment candidate(s) benefiting from the “shift” and the progressive candidate(s) losing votes, it is the PATTERN that is the most telling of all. However, I do have to say that the MA shift seems to be massive. It was completely unbelievable to me watching those results come in live AND the polling question answers did not match up at all with the reported results. They made no sense. If you look at how people answered questions and then compare that to the supposed results, you’re left with an obvious head scratcher. It makes no sense unless the reported results are wrong. I think the more they ‘rig” the results, the more easy it will be for people to accept/understand this data.

Just the Facts
Guest
Just the Facts
kathleencampbell67@gmail.com
Guest
kathleencampbell67@gmail.com

In 2016, the exit polls were used to show the theft of the election.
the response of the “authorities” who noticed this, was to simply to do away with exit polls and call California, for instance) without waiting for that election to be over or at least having had the voting stopped. What I did to compensate (however I could — I’m-NOT a statistician) or

lacking that information was to take screen shots every 5 minutes when they were announcing the “results”. When I did that, this method showed many instances of cheating. The media would post each county separately on the TV, so you could easily see if they were calling the race as finished. they very often had a “gaffe,” that required they “disappear” that particular county for a few minutes and then show it back up with the numbers changed. This happened a lot.

I havent done the screenshots this year but I noticed others were using it. I haven’t been well enough to do it. But I would suggest people at this site- should try to learn how to do this, if you don’t already know.

Robin Bergman
Guest
Robin Bergman

We primarily use paper ballots in MA but the scanners are used to count the ballots. We need to hand count the ballots.

Cindy
Guest
Cindy

You can also put in a request for digital ballots (digital copies of the paper ballots stored on the machine). This is something John Brakey has emphasized.

Sandy
Guest
Sandy

So, in your opinion, if one were to rig the vote counts, where would this best be done – IOW – if I had an algorithm ready to go that would redistribute votes according to some preference, where would I input the program? in official scanners? at the tabulation app level?

Second – but related – question: Would this be best done at a precinct or county level? if its the precinct there’ll have to be someone in the know, just in case, which – summed across precincts can be problematic. So may be county or even state level?

Sorry, but I am thinking like a detective – let’s assume the motive is there, and the outcome is as we see it. We still need to have a better idea of the means (electronic, but where?) and opportunity vs potential costs (can’t have too many people in the know or the chances for leaks increase).

Just wondering.

Erelis
Guest
Erelis

A good place to start answering your questions, which is a good one, is to study what happened to Tim Canova in FL against Debbie Wasserman Schultz. I believe the country used paper ballots counted by machines. I would look to a point in the process where a human must be inserted in the process.

Back in the day I worked many a democratic primary in a city dominated by party machine apparatus. The main point of changing results was at a precinct level. Once the voting ended, the person in charge would unseal machine (old flip lever type) and read the results to the rest of us who recorded it on paper forms. If nobody was watching the person in charge, he could easily add, subtract, or transpose the results. I think bad math skills is why some precincts reported more votes for a candidate than registered voters.

(BTW this experience with working inside a polling place convinced me the whole Russiagate thing was a hoax. The only people who can change the results of an election are the people inside the system. That is, it is Americans cheating other Americans.)

Mike Anderson
Guest
Mike Anderson

Yes, this is indeed a bit point we have been making: any vote “hacking” is going to be at the aggregators not the individual machines in the polling booths. There are so many machines, hacking a machine would literally have no effect especially on a state wide election.

But manipulation at an aggregator would, and it would have to be an “inside job” such as 2014 in Georgia.

That said, TDMS here is not detecting fraud. His “gain/loss” math is a bizarre misinterpretation of the difference, and is not the appropriate way to calculate that, Further his use of unweighted data make his effects spurious. Finally in dealing with such small sample sizes relative to the voter population, these polls are not intended for election integrity. Claiming they are is borderline fraudulent in itself.

And to add: the UN 4% is not a “thing” it is not a standard anywhere that we can find, and here it claims it with no reference.

ChuckBob
Guest
ChuckBob

Except that you’re not talking about computerized voting machines. If they’re connected to any type of external network they can be hacked remotely within minutes.
https://www.pbs.org/newshour/nation/an-11-year-old-changed-election-results-on-a-replica-florida-state-website-in-under-10-minutes?fbclid=IwAR3735Bv6CeL6ehei-0kTZ1UkHfUwbGKqAUEXqddBoGzLi9vu9RhoXTsh-Y

Charlottte Burns
Guest
Charlottte Burns

I’ll never forget Elizabeth Warren’s one moment of truth when she told Amy Goodman that she was just a player in the game. Bernie didn’t say word one about the election theft of 2016 or help out those supporters who went to court not once but twice over this. But Bernie does talk a lot about Russian interference in our elections. I often wonder if Bernie’s just a player in the game as well. In our fake wrestling matches known as “elections.”

Cindy
Guest
Cindy

Sometimes you have to kick the door more than once to knock it down. We cannot afford to be defeatists. Our democracy and the future of the world is on the line. Bernie says it’s “Not Me Us.” Well, the “US” part of that formula needs to make the “Me” part understand how important this is and how much we care about it. We also need to get Elizabeth Warren and Tom Steyer involved. In fact, the best case scenario would be to join forces with these three candidates and all their supporters combined. “Shifts” went against them too.

Renee M. Marcelle
Guest
Renee M. Marcelle

So it’s time to get #theresistance #democracyforAmerica #working families the nurses all involved. I’ll do my part to get the word out. If we do this, including letters to the editors; the Intercept, Buzzfeed news Cenq, Algezeera.

Cindy
Guest
Cindy

Exactly Renee!!! Can you please help do that? Can you find others to help do that> I did send a message to Democracy For America and some other groups like Our Revolution and Our Revolution in MA. But not sure they got it… we need to resend to these groups until they respond. Getting these large grassroots organizations with their huge mailing lists and megaphones involved is probably the fastest way we can get this information out in time… any help you can offer… and any help anyone reading this can offer… would be very much appreciated!!!

Charlottte Burns
Guest
Charlottte Burns

I should add, when Greg Palast and Jill Stein challenged the 2016 general election results and a judge told Jill Stein that she had such low %’s that she didn’t have a right to bring this into court, the judge asked Hillary’s lawyers who happened to be present if they would like to challenge the results. They said, “No.” We don’t live in a democracy anymore than the Russians did under Stalin.

Dieslaughing
Guest
Dieslaughing

This needs to be shared more.

Cindy
Guest
Cindy

I agree… please send a link to this page with a short explanation for context to all your social media accounts. I suggest to specifically ask people to share/retweet/etc. In my opinion, we need to put the “Establishment” who were behind these shifts in 2016 across many states, and it looks like we’re on track to see it again, that they will not get away this again. We can also point out that the last time around, that resulted in Trump being “elected.” I also want people to know that Trump and his minions will relentlessly show this in the general. This is an open secret already and we need to spread it to the masses so we get the Democratic nominee that the people actually want. It is a DISGRACE for elections to be stolen from the people.

Just the Facts
Guest
Just the Facts
Don Paul
Guest

Inexpressibly valuable. The discrepancies are worse than in Ohio for the Presidential ‘election’ of 2004. Thank you, Mr. Soares. The time is now to let everyone know. Can you find time to do California and Texas too? Or can others help you in your work?

Cindy
Guest
Cindy

I’m extremely interested in NC, VA, and MN too!

JoAnne
Guest
JoAnne

We use paper ballots in MN that are scanned by machine. The DNC could have manipulated it but I think the fact that Bernie did not win here is because so many people had planned to vote for Amy. She still won 8 counties despite having dropped out. Since Amy told her supporters to vote for Bernie, that’s what they did. *sigh*

Pamela
Guest
Pamela

If you are talking about Amy Klobuchar, she endorsed Biden, not Sanders.

Cindy
Guest
Cindy

That’s the narrative they put out to make the heist more believable. A false narrative is always part of the formula. In fact, I’ve come to look at the propaganda as portending where the heist will occur. They have to give people a reason to believe the results, even though it makes no really think about this one. Ted will release his MN report soon I’m sure and we will see if the same trend holds in MN. BTW, “scanned by machine” is the key phrase in your comment. Once those votes are turned into electrons, the results can be manipulated quite easily. There’s also chain of custody issues with those paper ballots and the fact that it’s near impossible to gain access to them for recounting purposes. Even the candidates paying DEARLY for a recall have problems gaining access to the actual ballots to count.

Mike Anderson
Guest
Mike Anderson

Most states (33) use paper ballots, and most of the rest use VVPAT (7) or Mail (3)

That leaves (7) “bad” states: KY TX TN NJ IN MS LA which still have some paperless machines, though all but LA do have paper trails on some to many.

Massachusetts uses paper ballots ONLY. Wisconsin, California, Vermont, Virginia, New Hampshire, Missouri, Michigan, Florida, Alaska, Alabama, Wyoming

THESE ALL USE PAPER BALLOTS. NOT DRE MACHINES.

Washington, Utah, Colorado all also use paper ballots because they are 100% mail in voting. LOL.

Carol Parham
Guest
Carol Parham

Texas especially!

Espen Hovdenak
Guest
Espen Hovdenak

Are mail in ballots included in the results? If so, that would explain some of the discrepancies. Buttigieg would be expected to over perform, but so would Sanders, while Biden in fact should under perform.

April Smith
Guest

This service is invaluable. And gobsmacking. And incredibly depressing. I live in MA and I know, intellectually, that rigging happens in any state (not just obviously corrupt GA for example). But this makes it more real. Time to ramp up my efforts. I suspect E Warren will be spending more time at home soon…

Cindy
Guest
Cindy

Hi April, could you please share the link to this report with a little context as to why on all your social media channels and specifically ask others to share it too? You’re right… we need to get this out there in a big way and as fast as possible.

Michael Cohen
Guest
Michael Cohen

Thank you.

From footnote 1, “this first published exit poll was subsequently adjusted towards conformity with the final computerized vote count.”

Can you explain this, please?

Cindy
Guest
Cindy

This is a great question. Here is the answer straight from Edison Research, “A properly executed exit poll accounts for absentee and early voters who will not be voting at election day voting locations, a practice that is becoming increasingly common and requires a great deal of care on the pollster’s part to properly model.” I’d like more granular info on this and I’d love to see results broken down by say VBM voters and same day voters in California. I’ll provide the link here too as there is other great info on that page. https://www.edisonresearch.com/makes-exit-poll-legitimate/

Cindy
Guest
Cindy

One thing I learned in 2016 when this was happening is that Edison Exit Polls are used all over the world and are considered the gold standard. In no other countries are the exit poll results “adjusted” (what a euphemism, right?) to match the results. Why? I it makes no sense whatsoever to do so. Can you think of ANY legitimate reason this would happen? Keep in mind too… the media in the US that pays for these exit polls (at least in 2016 this was AP, CNN, NBC, CBS, ABC, and FOX) KNOW the exit poll data for “Who did you vote for?” are “adjusted,” i.e. they’re in on it. Keep in mind that AP feeds are used in almost all print media/newspapers in the country. In 2016, when a lawsuit was launched and these media giants were asked to retain all exit poll data (including unadjusted) for California and other states voting near the last part of the primary, they CANCELLED the exit polls. They are part of the establishment and they know the truth but they won’t dare report on it until they are FORCED to do so. This is another reason it is so important that… Read more »

JoAnne
Guest
JoAnne

Richard Charnin explained this very well during the 2016 primaries.

Christ
Guest
Christ

GWB made it law that exit polls needs to be adjusted to fit the results no matter how widely different the poll and the result would be. The US is the only country that has such a law.

Diane
Guest
Diane

The exit poll is based on less than 3k voters and we had almost 1 million in the primary. Exit polls are not used to validate the actual count – think about. Don’t fall for people trying to split up the Democrats. He is probably a Russian troll or a Republican.

Cindy
Guest
Cindy

No, I know for a fact that Ted is a NOT a “troll.” Exit polls done by Edison Research are used in many countries to test the validity of the reported results. They are the gold standard. They are used when the UN goes in to monitor elections and protect vulnerable democracies. Folks, don’t let anyone tell you that what you’re clearly seeing isn’t true! Let’s all do what juries are told to do every day — USE COMMON SENSE! If all of these results, state by state, show the same trend — establishment candidates gaining ground on these “shifts” and non-establishment candidates losing ground on these shifts — what does your common sense tell you? Put another way, if Buttigieg is the big beneficiary in the New Hampshire Primary and Biden is the big beneficiary in all the rest, what does your common sense tell you? Shouldn’t these shifts be “random” across states/candidates if there was some other explanation? Common sense folks — believe the obvious — believe what you can go investigate and see for yourself — these shifts are hiding in plain sight — they were there in 2016 too, favoring Hillary Clinton and we ended up… Read more »

Alyce L Garrity
Guest
Alyce L Garrity

But the results still could be off by a moderate margin. A person could say they voted for X when they really voted for Y. There are a variety of reasons for this, including not wanting to let their spouse or anyone else know how a person voted. We need to go back to the actually voting BOOTHS, where a person is alone in the booth to cast their vote, unless they have a disability and need help, and then only if the person asks for help themselves, and then only if the assistant is unbiased.

Cindy
Guest
Cindy

Slight discrepancies, yes. Ones outside the margin of error and trending all in the same way across multiple states, no, unless something is amiss. Edison is the same firm who does exit polling in countries where elections are being monitored by the UN to keep them honest… including US observers ironically… they are considered the gold standard. In other countries, if they saw these same types of shifts as Ted is demonstrating, they be recounting all the paper ballots, and then if there were still these shifts, they’d call for a new election. The US is the only country where we have accepted these types of shifts for decades. It’s high time we do something about this. We should protect everyone’s vote and the people’s will should be what determines who wins. The machines that are used, including the machines used to digitize same day paper ballots and mail in ballots, are highly vulnerable. There are also major chain of custody issues. Lots of ways to cheat. Lots of barriers to verifying the vote. However, what Ted is doing, especially when we get more states in, will show us enough that we can use common sense, like juries do every… Read more »

trackback

[…] 2020 DEMOCRATIC PARTY PRIMARY Exit Poll Versus Reported Vote Count” [Theodore de Macedo Soares, TDMS|Research]. “The 2020 Massachusetts Democratic Party presidential primary was held on March 3, 2020. […]

trackback

[…] 2020 DEMOCRATIC PARTY PRIMARY Exit Poll Versus Reported Vote Count” [Theodore de Macedo Soares, TDMS|Research]. “The 2020 Massachusetts Democratic Party presidential primary was held on March 3, 2020. […]

ROBIN T ESSER
Guest
ROBIN T ESSER

To Theodore: can you please post a screen shot of the 8 pm downloaded exit poll data from the CNN website? or some other definitive proof of it?

Alyce L Garrity
Guest
Alyce L Garrity

I still don’t think you can rely on exit polls. You can say you voted for X whilst you actually voted for Y for a variety of reasons.

Cindy
Guest
Cindy

Studies have shown that Edison’s Exit Polls are quite accurate. Most people will tell you who they voted for or decline to answer. The real key is to look at the trend across states. So far, we have evidence that in three states, NH/SC/MA, the “shifts” are in favor of the establishment candidates and move away from the non-establishment candidates. You would expect that to be random if what you say is true. Once Ted posts more states, we’ll see if the trend stays the same. It certainly did in 2016. The shifts were overwhelmingly in favor of Hillary Clinton and against Bernie Sanders. That is a fact.

trackback

[…]   […]

Chris
Guest
Chris

Do you happen to have a screenshot of the CNN EP results from 8pm Tuesday, or some other record of how the numbers stood at that time, beyond assertions? That would be hugely valuable.

It is striking that while the number of people included in the EP increased from 1394 to 1443 (+49 people), the proportion of those people voting Biden increased from 28.9% to 34%, indicating a gain of +88 people. Similarly, Sanders went from 30.4% to 28.2%, indicating a loss of 17 people.

So unless people are allowed to change their answers in exit polls, something is quite amiss.

Chris
Guest
Chris

Saw that you just added the data – thanks!!

Dale Tavris
Guest
Dale Tavris

Probably in between those two versions of the exit poll, the exit poll was “adjusted” by CNN to match the official vote count. They routinely do that.

Janet Alfieri
Guest
Janet Alfieri

Thank you for doing this. It explains much of my feeling that something strange was happening in the Mass. primary. I canvassed and found no one who supported Biden. There were no sign holders at the polls holding signs for him. Many people said they supported Warren or Sanders. Some said they supported Pete or Amy. Biden didn’t have any public events in Mass. and no ads. He did horribly in New Hampshire, our neighboring state. Something fishy is going on. We got scrod.

Jess
Guest
Jess

Exactly the same experience for me! I haven’t met ANYONE here in MA who was voting for Biden. I haven’t seen a single Biden sign or bumper sticker, it’s mostly Bernie! I even saw a Cory Booker sticker a few weeks back, but no Biden anywhere! So when I saw that Biden won MA, I was like this CAN’T be right!! I strongly suspect something is up with a lot of these results.

AJ Nstr
Guest
AJ Nstr

Amazing how these things keep happening to progressive candidates.. Particularly Sanders. Yes, curious.

Just the Facts
Guest
Just the Facts

Here is the actual poll that you are referring to (Edison Research for CNN). Note that Biden averaged 34% on the exit polls (NOT 28.9% as you assert) and Sanders around 28% (NOT 30.4 % as you claim). Same pattern applies in your report on South Carolina. If you misquote sources and make up your own numbers, you can convince gullible folks of just about anything. https://edition.cnn.com/election/2020/entrance-and-exit-polls/massachusetts/democratic

Tom
Guest
Tom

Even using these numbers, Bernie underperformed relative to the exit polls by 3.7 percentage points– which is within the margin of error. Biden’s performance was 0.5 percentage points outside the margin of error. Easily explained by some people sheepishly voting for the momentum candidate, and then not wanting to admit it. The margin of error exists for a reason.

Tom
Guest
Tom

Your “gain of 15.7% (Biden) and loss of 12.0% (Bernie) from the exit polls” is an odd way of presenting it. Biden outperformed the exit polls by 4.5 percentage points, and Bernie underperformed by 3.7 percentage points. The margin of error of the poll was +/- 4.0%. You then add those differences together to get 8.2%– as if that is meaningful. Fine, if you add the margins of error together you get 8.0% (4+4=8). And a negative error for one candidate is going to by necessity be a positive error for another candidate, because it all has to add up to 100%.The poll has a margin of error for a reason. Not everyone tells the truth in an exit poll. Momentum toward Biden was huge. Some peoples’ vote finger moved, but their heart didn’t yet. They voted in the direction of momentum, but didn’t want to tell that to the pollsters.

Chris
Guest
Chris

Edison ‘adjusting’ their exit poll data to match election results is absurd. It is at the very least extremely disingenuous of both Edison and CNN to report number of participants in the exit polls if they are modifying the responses of those participants.

For example, based on the data from 8am as compared to the currently available exit polling data, ~17 participants who had indicated support for Sanders, and ~49 participants who had indicated their support for Warren had their responses thrown out or otherwise changed. Additionally, while only 49 new participants were added to the exit poll after 8pm, Biden ‘somehow’ received the support of 88 additional people.

Check out the below spreadsheet for comparisons of all of the numbers:
https://docs.google.com/spreadsheets/d/1VpQ8PgV5NQvMRe4Ga3Q7IZTneduoBXZfKfvu9OetUqc/edit#gid=0

Griffin
Guest
Griffin

Do you have an archive.is link to the version of the poll you used?

Paul F.
Guest
Paul F.

TDMS needs to explain his (1) footnote better. I pulled the numbers from gender in Mass based on his link and it was 28.2% for Sanders. If you want us to take your percentage seriously, you need to explain what you mean by adjusted to conformity or whatever you wrote there. Otherwise, this is FUD.

Tom
Guest
Tom

The most glaring error is in the footnote, where you say the discrepancy (8.2%) between Biden’s and Sanders’ results (Biden overperformed by 4.5, Bernie underperformed by 3.7) is twice the margin of error. FALSE! Either you are flat-out lying, or you misunderstand polls and margin of error. The Biden vote percentage has a plus or minus 4 margin of error. The Bernie vote percentage ALSO has a plus or minus 4 margin of error. Those margins of error are for their individual result. The HEAD TO HEAD margin of error therefore is plus or minus EIGHT percentage points. (The reason is that Biden could be plus four, and Bernie could be minus four, and STILL WITHIN MARGIN OF ERROR). So, in reality, the discrepancy was 0.2 percentage points larger than the margin of error, not twice as big. The whole thing is a sloppy mess.

ROBIN
Guest
ROBIN

Ted: can you please post a screen shot of the 8 pm exit poll data that you used to compile your chart, above,in the above article? I need to see the original data in a form that could not have been altered, in order to accept your assertions. Would you please post it? I don’t want an explanation of why it’s different from the 10 pm exit poll data that is still available at the CNN link – i just want to see the original data in a form that could not have been doctored, please. Thank you.

Jess
Guest
Jess

Send this to the United Nations

AmPewter
Guest
AmPewter

Did you not read your source? Margin of error doesn’t mean 4% between candidates, but 4% for each candidate’s numbers:

In horse-race polls, we want to know the difference in proportions supporting the top
two candidates (and perhaps between other pairs) and we need the confidence interval
for this difference to tell if the lead is statistically significant (or “outside the margin
of error”). Often pollsters, journalists and political scientists calculate this as twice the
reported margin of error of the poll. This is done following the logic that if one candidate
2
is at 55% and the other at 45% and the poll has a ±5% margin of error, then the first
candidate could be as low as 55−5 = 50% and the second could be as high as 45+5 = 50%.
In this case reporters would say the race was a statistical dead heat because the gap
between the candidates (55 − 45 = 10%) is not more than two times the margin of error
of the poll (5%).

Joseph F Michel
Guest
Joseph F Michel

Simply unbelievable that this story does not get out much further and wider. In the home of democracy?

Cindy
Guest
Cindy

I’m trying 🙂 Can you help? Post to social media, ask your friends to do the same, write articles… whatever it takes… AND can we please get Bernie and the other candidates to speak out on it? Reach out to them?

Adam
Guest
Adam

Just FYI, your data is wrong. Per the very link you supplied to the CNN exit poll data, all of the vote results fall well within the standard deviation for the reported results. Might want to look at your numbers again.

Bojo
Guest
Bojo

Wow! Rigged, eh? That’s all you got? Listen: Sanders demographic did not show up to vote. Check the data. In nearly every single demographic, the turn-out beat previous years, save one: the youngest 18-29 crowd. So sorry, but not sorry. You are cycling BS an you damned well know it.

beejay
Guest
beejay

In the explanations below (footnotes if you will) for [3] shouldn’t it be “(subtracting values in column one from the values in column two)” ? Ex: Biden 33.4% (Column 1) – 28.9% (Column 2) = 4.5% (Column 3)

And for [4] shouldn’t it be (difference in column three divided by exit poll proportion in column one)?? Ex: 4.5%(Column 3) divided by 28.9% (Column 1) = 15.7% (Actually 15.6%)

Dale Tavris
Guest
Dale Tavris

Thank you very much for this info. Very disturbing! Can you tell us if you’ve done similar calculations for the other super Tuesday states, and if so, what you found?

trackback

[…] is awash with conspiracy theories regarding the Democratic primary results for Massachusetts, after a blog from Theodore De Macedo Soares claimed there to be some significant irregularities in the vote count when comparing the exit poll […]

Joseph Couture
Guest
Joseph Couture

So this is absolute nonsense.

First, the source for Exit Poll Proportion supposedly comes from CNN’s exit polling. However, when you click the link to the source material it takes you to South Carolina’s polling.

Assuming that was a typo, the numbers in those column are completely made up. CNN does not provide a top level breakdown of the exit polling by candidate. Each data set is broken down by gender, race, age, education, etc. So if you wanted to get those top line numbers, you need to adjust the subsets by ration. The simplest data set to do that is gender. 44% men, 56% women.

The math for that is as follows:

(.34*.44)+(.34+.56)=.34

This means Biden scored 34% in the exit polling. Biden’s actual result in Massachusetts: 33.6%.

Dan
Guest

Your exit poll numbers are wrong. Also, your link in note [1] is to the SC exit poll. The MA exit poll numbers published by CNN are spot on with the actual results.

Dan
Guest
Dan

Sorry, I missed the note saying that CNN adjusted the exit polls. That’s a big accusation. We need screen shots of what you downloaded on election night.

Cindy
Guest
Cindy

They did these “adjustments” across all states in the 2016 Dem Primary. It should be noted too that we did not see these types of shifts to one particular candidate on the Republican side in 2016 when there were many candidates, only on the Dem side between Hillary Clinton, the establishment candidate, and Bernie Sanders, the non-establishment candidate. In most states, the Dem and the Rep primaries were held at the same time in the same states. I’m almost certain there was no “adjustment” “needed” on the Republican side because the date from the exit polls actually matched the reported vote with only slight (well within the margin of error) — Isn’t this right Ted? I put “adjustments” in quotes because they do NOT do that in other countries where Edison exit polls are used to test the validity of the reported vote nor does it make ANY sense to do so unless they’re trying to hide the fact that the exit poll data does not match the reported votes. In other countries, if the exit poll data does not match reported vote and is outside the margin of error, the take that as a sign there is something WRONG… Read more »

Nikki Alexander
Guest
Nikki Alexander

I am beginning to feel a little less crazy reading this thread. Especially your comments Cindy about common sense. The way Biden swept 10 states just didn’t make sense to me. A dead campaign with no organization on the ground in the states he won, lackluster rallies, donors abandoning him, having lost the first three primaries … It just didn’t make sense. Why would voters flock to him just because he won SC? Because the establishment closed ranks and 3 “team players” dropped out and endorsed him the night before the super Tuesday primaries? Is that really all it takes to change every voter’s mind in 10 states? I live in CA. Biden had only one office, vacant and padlocked, in Compton. How could he possibly have won so many votes here when Bernie’s team was everywhere and active for 8 months? It just doesn’t make sense. Also our voting system which worked perfectly before was completely screwed up by new computer voting machines that “malfunctioned”. The other thing that doesn’t add up for me is that the national polls showed Biden in 3rd or 4th place behind Bloomberg. Nationally. Suddenly he’s in 1st and runs the table in one… Read more »

Cindy
Guest
Cindy

Nikki, you are right on. Thank you for laying that all out so clearly. We need to team up! In fact, several of us on these comments need to team up. We need to get the word out fast and widely. The Corporate Media will not cover this so we have to come up with other methods of distribution. If anyone reading this has a good connection to a large grassroots organization with a large mailing list and big “megaphone,” could you please get this information to them and specifically ask them if they could send it out to their members? Of course, post to social media too…. but we need to get this out in other ways too. If anyone has a connection to the inner circle of any of the candidates affected negatively in any state (Bernie, Warren, Steyer), please reach out to them with this information and do what you can to get these candidates/teams to speak out publicly ASAP. Also, anyone with a podcast or newsletter would be another method of fast distribution. Let’s team up and make getting this information out a priority!

Cindy
Guest
Cindy

Hi Ted, again thank you for doing this. I wanted to ask if you have any idea when you’ll release another one (or more) and which state?

Cindy
Guest
Cindy

By the way, if I were looking for clues as to “who” is/are the controlling force behind these “mysterious” shifts, I would suggest Pete Buttigieg’s phone records (and key staff) between the time he refused to drop out when asked after SC and then 7 hours later he suddenly turns the plane around, goes back to Indiana to drop out, and endorsed Biden before SC. He looked rattled on the plane when telling the reporter pool they were turning the plane around mid-air! NH looked “fixed” for him (the shift went heavily in his favor), then the establishment shift switches to Biden. So who ask Pete to drop out the first time and then who forced him to drop out? I’m not saying this is 100% related but it could provide critical clues…. and I think investigative reporters should be asking these questions and I think any lawsuit to look into this whole mess would likely want that type of info as well. Again… common sense.

Nikki Alexander
Guest
Nikki Alexander

The show “Rising” on Hill TV – YouTube said Obama called Pete. They said it as if it was common knowledge. I didn’t look to see if that was said by anyone else.

Rick
Guest
Rick

I read I believe on VerifiedVoting.org that Mass. has a 1% audit following the electronic tabulation to check the integrity. If it was done or is about to be done: Who are the election integrity people in Mass.? What games were played with them, or will be played, regarding the audit? Some typical games include keeping the citizen observers in a room far from the audit so they can observe through a window, maybe 50 feet away. I’m not saying that is done in MA, but I’ve heard about that in CA during an audit which is *not* done routinely but was done after a suit filed against San Diego County for irregularities, in 2016.

Cindy
Guest
Cindy

I’m not sure about MA but here’s the bottom line in California: the mandatory 1% audit does not RANDOMLY sample ALL of the votes — and to quote John Brakey, this allows them to “steal with impunity! because the so called “random” 1% audit isn’t random at all. The provisionals are routinely excluded, and in 2016, not so surprisingly, very heavily favored Bernie — and by design in many cases. Equally important, a large chunk of the voting is done by mail and good chunk of that is NOT included in the sampling for the 1% audit. Let me explain further. When these ballots come in, the signatures must be verified before the ballots are accepted. In almost all counties, this is done by machine. These machines have a setting that can be changed regarding the sensitivity to accepting a sig or not. So, when the ballots are run through the machine, some are accepted and some are set aside and later looked at by real human eyeballs. In at least most counties, the ones set aside by the machine are not included in the sampling for the mandatory 1% audit. In 2016, the vast majority of the machine rejected… Read more »

Kit Jones
Guest
Kit Jones

The report that I saw about Buttigieg’s turn-about was that when he refused to do as directed, the DNC pulled his financing, devastating his campaign funding and leaving him no place to go but saluting and following orders.

Arthur
Guest
Arthur

One thing becomes certain in this conversation: paper ballots are more reliable and more trusted.

David Martin
Guest

These results fit perfectly with my observations concerning what I call “The Massachusetts Miracle” of the Democratic primary in 2008 in which the preferred establishment candidate at that time, Hillary Clinton, defeated Barack Obama (we were told) by a resounding margin of 57% to 41% of the vote. This was in spite of the fact that Obama had the endoresement of the black governor or the state, Deval Patrick, and Senators Edward Kennedy and John Kerry. On the same day in the Massachusetts of the Midwest, Minnesota, Obama won the less-easily-rigged caucus by a margin of 67% to 32%. Massachusetts also has a large, ultra-liberal student population and the Boston Globe was perhaps the only major newspaper in the country that was consistently opposed to the Iraq War, a position much closer to Obama’s than to Hillary’s. My larger analysis of the Super Tuesday of 2008 is in “Grand Theft Primary,” http://www.dcdave.com/article5/080209.htm. There is no reason under the sun to trust any vote count overseen by the corrupt-to-the-core DNC.

Cindy
Guest
Cindy

David, thank you for posting this analogy. Yes, this sort of thing has been going on a loooooong time and we need to draw attention to that too. And… we need to finally rise up and do something about it! Please see my comment to Nikki as it applies to you too and everyone reading this and wanting to help.

Don Paul
Guest

I just chipped in $50 toward Ted’s efforts. He’s done good work for years. His tables of improbable differences between exit polls and computerized final results in the 2016 Presidential are worth study–https://tdmsresearch.com/2016/11/10/2016-presidential-election-table/ Four key States evidently went to Trump through vote-fraud then. The differences represented by the Biden and Bloomberg gains in computerized final results for Massachusetts 2020, as presented by Ted, are another order of theft. As Cindy in this discussion says, a ‘pattern; shows in looking at New Hampshire, South Carolina, and Massachusetts. The same ‘pattern’ shows in looking at Clinton versus Sanders in Massachusetts 2016. As Arthur says, further on. ‘One thing becomes certain in this conversation: paper ballots are more reliable and more trusted.’ As many others have said, We need to stop this theft now. We’ll have more bad theater and sickening outcomes, such as the fine documentary “Stealng America, Vote by Vote” recounts about theft of the 2004 Presidential election (https://vimeo.com/2019192), unless and until we raise our voices and increase our numbers into an irrepressible wave. As others suggest, the Sanders and Warren organizations should be engaged among their millions at the grassroots. Many other organizations that care about democracy in this nation… Read more »

Cindy
Guest
Cindy

Don, thank you for supporting Ted. The work he is doing is so incredibly important. Thank you too for providing the link to the documentary. We need a ton more people to see it. Could you please take a look at my comment to Nikki as it applies to you too and everyone reading this who want to help. We need to band together and get the word out fast! We also need to convince the campaigns affected to speak out publicly. They’re scared to do so but we need to give them the courage to do so by having their back and spreading this information. Our democracy is on the line.

Rick
Guest
Rick

Elaborating on what I said above: Since 2014, MA has required, for integrity, a *manual* audit of a certain percentage, I think 1, of randomly selected “paper ballots”, ie what was hand-marked and fed into the optical scanners:

https://www.verifiedvoting.org/state-audit-laws/massachusetts/

Citizens can observe, but probably not conduct, the audit.

I’m looking for the contact info of any election protection organizations in MA. I know of them in other locations like Chicago and California.

Please, if anyone knows such contacts in MA, time may be of the essence (if the audit wasn’t done already). And note my comment above: that the Elections Boards I’ve read about or experienced, really don’t like citizens breathing down their necks, and will do things like keep the observers at a distance in the next room with a window through which to try to guess what’s going on, and (at my local Elections Board for a large county) enough room for about 5 people in the observers’ area.

trackback

[…] ELECTION FRAUD SUSPECTED—EXIT POLL VS. OUCOME DISCREPANCIES: https://tdmsresearch.com/2020/03/04/massachusetts-2020-democratic-party-primary/ […]

trackback

[…] Why the exit polls accessible today differ from the exit poll used here. […]

C. K.
Guest
C. K.

Why isn’t ANYONE doing anything about it? Or did I miss it and steps are being taken to protect our Democracy? This sucks to see this all over again, 2016 and now. What is the point of election monitoring if no one is willing to file a lawsuit and take this further?

Kit Jones
Guest
Kit Jones

I think we all know that Bernie was cheated out of the nomination at the DNC convention in 2016 and is in the process of being cheated out of the 2020 nomination. I use my Fb page*, Twitter, and Instagram to share non-MSM news, and will certainly be sharing this movement’s news on each of those social media sites. See my informal poll on Fb seven hours ago asking for feedback from my Fb friends on this question: “What nefarious methods or misdeeds do you believe are keeping Bernie from receiving fair vote counts?” I invite you to DM me on Messenger with suggestions for updates on this movement’s purpose and actions. Thank you so much for validating my reality of the seriousness and the vital nature of our addressing the hijacking of our electoral system that is currently underway. I think we all knew it as soon as the vote totals began coming in. It was a stunning Super Tuesday here in the Dallas-Fort Worth metroplex where progressives have been running into each other at every turning for the last four years. Btw, for obvious reasons, I voted for Jill in 2016, despite being a Bernie delegate at the… Read more »

Cindy
Guest
Cindy

Hi Kit, I’m wearing two hats here — one is non-partisan election integrity advocate and the other is an ardent Bernie supporter. Really, it was the state by state grand theft of democracy in the 2016 Dem Primary that got me super interested in election integrity so the two are intertwined. One of my goals is to DETER whatever is going on with the machine vote in future states. To that end, I’m trying to convince grassroots groups to send this information out to their email lists and use their megaphone on social media and elsewhere. I’m also trying to urge individuals to do the same. I’m also thinking it might be worth trying to get local media to cover this. And… try try try to get the candidates/campaigns to speak out on what we’re seeing… mostly targeting Bernie but I’ve sent messages to Warren/team (major shift away from her in MA too, although it looks like Bernie would have won comfortably) and Tom Steyer (major shift away from him in SC which looks like a plan to keep him non-viable so Biden could get more delegates and get Tom to drop out). Actually, what I would love to see… Read more »

Cindy
Guest
Cindy

I should add Tulsi Gabbard to that list too and will send her a message about this too. I encourages others to contact all the affected candidates as well. I’m personally spending a lot more time on reaching out to Bernie but I believe it should be called out by all of them!

10:22
Guest
10:22

Thank you for the nice analysis.
Do you mind sharing the confidence intervals and p-values?
Maybe it is worth graphing the distributions to get your point across more easily to the general public.

Michelle Lucier
Guest
Michelle Lucier

When I posted this article on Facebook I was accused of spreading Russian propaganda by supporters of Biden

Linda C
Guest
Linda C

It would seem that someone could request a manual recount of a precinct or town, or a sampling of them to either put this to bed or find something out.

Cindy
Guest
Cindy

I suggest we ask all affected candidates to do this… not just the candidate we’re personally supporting. I’ve contacted Bernie, Steyer, Tulsi, Yang, and Warren. We need to keep pushing these candidates to do so and to get election integrity lawyers/experts to help them so they do it right. I know Tulsi only has a small sliver of the vote and Steyer dropped out after SC, but they both have as much standing I think as Bernie so I think we should really encourage them in this direction. It would be good for Bernie if ANY of these candidates did this… and it would of course be good for our democracy going forward.

Marc Plante
Guest
Marc Plante

Theodore… it would be “interesting” to see the before and after exit poll data, to see how the exit poll data is adjusted for public consumption.

Can you provide this data?

I am working with John Brakey on some of the Mass. data regarding the hand-counted v. machine-counted ballots.

Great work in informing the public. I have been sharing your data as far as I can.

Cindy
Guest
Cindy

Hey Marc, big John Brakey fan here! Wanted to say thank you for what you’re doing! Could you tell us a bit about what you’re finding with hand counts vs machine counts…. with geography included. Are you doing the same in Vermont? I’d be really interested in that.

trackback

[…] that happened in Clinton’s favor throughout the 2016 primaries, in Massachusetts the discrepancies between the vote count and exit poll for Biden and Sanders was 8.2%, which is double the 4% margin […]

trackback

[…] MASSACHUSETTS 2020 DEMOCRATIC PARTY PRIMARY Exit Poll Versus Reported Vote Count Posted on March 4, 2020 by Theodore de Macedo Soares https://tdmsresearch.com/2020/03/04/massachusetts-2020-democratic-party-primary/ […]

trackback

[…] MASSACHUSETTS 2020 DEMOCRATIC PARTY PRIMARY Exit Poll Versus Reported Vote Count ← “They must be doing this on purpose”: Texas (would-be) voter on the 5-HOUR WAIT at his polling place (on the campus of an African-American university) […]

Linda Tolini
Guest
Linda Tolini

Demand a recount ,

trackback

[…] Why the exit polls accessible today differ from the exit poll used here. […]

Don Paul
Guest

It’s great the read the ardor and research among participants in this thread! I just posted a piece on which I’d been working since the Democrats’ Iowa caucus. You may find it useful. We appear to have many allies. I particularly enjoyed remarks by Barbara Simons, Steven Freeman, and Joel Bleifuss, and I loved the photo of Germans with paper ballots! https://donpaulwearerev.com/flipping-the-script/paper-ballots-all-the-way-to-make-sure-votes-of-the-real-america-count-in-2020 You can hear the piece, along with a song, on bandcamp, too.

Cindy
Guest
Cindy

I’ve been thinking about how to get the word out fast and how to have the biggest impact…. I think we need to contact the Michigan offices of Rashida Tlaib immediately, plus her congressional office… and the Washington offices of Pramila Jayapal immediately, plus her congressional office. I also think we need to work on getting one or more of the super celebrities endorsing Bernie Sanders to speak out now….. actors/actresses/musicians/athletes/etc… the bigger the name the better and especially those who have been on the trail with him/performed at his rallies. I’ve been tweeting at Michael Moore for a while now but we need to contact them all and not just by social media. Politicians will be much more reticent to speak out than celebrities I think and celebrities can garner massive attention with one tweet and or one press conference… plus that would FORCE the corporate media to cover it, even if negatively. I’m going to ask here for help…. and lots of help!!! Can you help? Anyone have a connection to any of these people? It would even be good if people could show up in person at these offices and rallies with signs and try to talk… Read more »

Carl Johnson
Guest
Carl Johnson

The formula you referenced to get the MoE is dependent up on a “Simple Random Sample” which is defined as “a subset of a statistical population in which each member of the subset has an equal probability of being chosen.”

Exit polling does not utilize simple random samples. These are people who pick polling sites to visit, and then poll intervals of people. There’s too much bias and not enough coverage for this to be considered statistically random because so much of the sample population is decided by humans.

Therefore extrapolating an MoE from exit polling is a misuse of statistics. There’s a reason why CNN does not publish an MoE on these exit polls and why one had to be extrapolated.

Carl Johnson
Guest
Carl Johnson

Just to follow up. I’ve heard that CNN has polling locations selected randomly by a 3rd party. This still doesn’t make the samples qualified to be considered “Simple Random Sample”. It just means you’re locked into whatever lean each of those precincts have. Simple Random makes the entire population subject to polling, not random locations and a subset of each location.

Michael Q. Rudnin
Guest

Lee Camp @ Redacted Tonight / Moment of Clarity covered this. https://www.youtube.com/watch?v=t78Ff26-V8s

trackback

[…] Why the exit polls accessible today differ from the exit poll used here. […]

Boris Hayete
Guest
Boris Hayete

The Massachusetts numbers are incorrectly computed. For example, the Sanders exit poll average should be (0.31*0.44 + 0.26*0.56)*0.98 = 27.6% (based on CNN polls posted online). This is very close to the actual percentage, 26.7, and quite far from the reported 30.4. There’s a deliberate or accidental error in this table, and a very large one, that ENTIRELY accounts for the discrepancy. Always check the referred-to datasources, people!

James Chapman
Guest
James Chapman

The author has donated to Sanders for years. Bias at play here? And apparently is finding discovering that Bernie is getting screwed in primary after primary. What are the chances? Lol.

James Chapman
Guest
James Chapman

The author has donated to Sanders for years. Bias at play here? And apparently is finding discovering that Bernie is getting screwed in primary after primary. What are the chances? Lol.

Josiah Haswell
Guest
Josiah Haswell

So, the big story isn’t that CNN and NYT and others are fudging their exit-poll data? Do you have the original “uncorrected towards conformity” data? If not, why not? Wouldn’t major pollsters fudging their polls be a much bigger story than a polling anomaly that results in like, at most 1 delegate biased towards Biden?

Max
Guest
Max

See this response

Alex N
Guest
Alex N

Do all the electoral districts in MA use unverifiable computer voting? Or are there some districts that still use paper ballots? It would be interesting to compare the exit poll / final vote count disparities, if any, between districts that use computer voting machines and districts that use other ways of collecting votes, I.e. lever machines, etc.

Jason C
Guest
Jason C

Hi Theodore,

Thanks for doing this really important work. It’s been eye-opening for me, and I’d like to share it widely. Before that, however, I’d like to do my due diligence and ask about a couple of your claims I’ve had a hard time verifying.

1) You note that the Edison poll results published on CNN have been adjusted to conform to the computerized vote counts. Has CNN admitted to doing this somewhere? Is this a normal process for CNN? Where can I find more information about how the results are adjusted?

2) Is there a way that I can verify the numbers you pulled from CNN’s website on election night? I looked at the earliest save I could find on the wayback machine (5 minutes after midnight on March 4), and it showed the same results shown today (though I’m not sure if the wayback machine captures that kind of data…) It goes without saying that all your conclusions on discrepancies depend entirely on the accuracy of the numbers in these great tables you’ve made.

Thanks again.

Jason

Cindy
Guest
Cindy

Just saw this petition (and signed it) calling for the UN to monitor our elections, giving Ted’s exit poll research as a reason: https://sign.moveon.org/petitions/u-n-is-needed-to-oversee-democratic-primaries-due-to-election-fraud

Cindy
Guest
Cindy

Lee Camp certainly has a way of driving the point home! This one is so well done, maybe we put part of it in a commercial!
https://www.youtube.com/watch?v=Rk6iTd18lPI

trackback

[…] as in Texas, election results from the digital vote counts deviated significantly from the exit poll results conducted by Edison Research. The difference in Massachusetts is that […]

Sean Smith
Guest
Sean Smith

You’re a joke. Biden legit crushed Sanders.

Michael Q. Rudnin
Guest
Will
Guest
Will

I read that article by ‘Fact Check’. I cannot weigh-in on the objections they raised about technical issues of methodology used by TDMS. But as long as TDMS applied the same methodology to all his analyses of the various states, it is still anomalies to have outcomes that always hurt Sanders. Even a flawed methodology, consistently applied, should yield a random scatter of deviations favoring and hurting different candidates. The results seem to suggest a concerted manipulation to hurt Sanders.

ThisOldMan
Guest
ThisOldMan

I live & vote in Massachusetts, so I know for a fact that we DO use paper ballots! They are immediately scanned by a machine that each voter has to personally feed the ballot into. So, at the very least, a hand recount is possible here. If this discrepancy is so compelling, why then did not OurRevolutionMA not even try to demand such a recount?

trackback

[…] margin of error. Sanders won Massachusetts in the exit poll and lost it in the computer count. The discrepancies between the exit poll and the vote count for Sanders and Biden totaled 8.4%— dou……Warren’s and Biden’s discrepancies also totaled 8.4%, again double the margin of error. […]

Zeb
Guest
Zeb

Did CNN change their exit poll numbers? This exit poll https://edition.cnn.com/election/2020/primaries-caucuses/entrance-and-exit-polls/massachusetts/democratic seems to indicate Biden won.

Gavin
Guest
Gavin

He explains in the notes that their exit polls are adjusted to conform with incoming computer-counted results. The reason for this is that technically these exit polls weren’t meant to detect discrepancies, but to get demographic data about voters. TDSM’s numbers are from the initial exit polls CNN out, hence why he calls it the “raw data.” These polls haven’t been adjusted to conform with “official” results.

ChuckBob
Guest
ChuckBob

There’s a whole page of discussion on that very subject.

Terrance Newton
Guest
Terrance Newton

That Biden and the DNC cheated is not telling. What’s more telling is both Sanders and Warren didn’t utter a word of descention about the exit polls. The possibility that Warren and Sanders could be in on this con is not farfetched.