How is this not immediately obstruction of justice? The article makes it seem like they know who actually wiped it, so just ask them who ordered it and follow the trail, no? It can't even be an accident since not one but two backups were also wiped.
Later in the article, they mention that wiping the server was in response to a security vulnerability leaving 6.7 million sensitive voter records vulnerable. The governor's office claims to not be aware, so it sounds like they're still following the trail, as you put it. Fortunately:
> The FBI is known to have made an exact data image of the server in March when it investigated the security hole
Electronic access, phyiscal access...it sounds more like it was a "routine" disposal of sensitive information 10 months after the election.
United States national ballot records must be kept for 22 months, and then they are likewise disposed. Florida kept (keeps?) its ballots especially long after the 2000 recount.
> Electronic access, phyiscal access...it sounds more like it was a "routine" disposal of sensitive information 10 months after the election.
When a lawsuit is filed (or even when you reasonably know lawsuit will be filed), you can not legally destroy evidence related to it, even in the course of otherwise routine document destruction.
In state agencies and other large orgs, it's common to receive mass email orders from legal to preserve all documents and communications related to matter under dispute, regardless of normal retention policies, with reminders of severe legal consequences to the org and dire personal consequences if the directive is not complied with.
They were informed 3-4 weeks before the lawsuit was filed that they were going to be sued and they should not delete the data. But did it anyway so seems like they might be hiding something
Those couple of paragraphs are confusing to read. They don't mention if the backup servers were also vulnerable to whatever the security problem is, and if they weren't, how they also ended up getting wiped. The article also says the failed attempt also left all that data vulnerable for months -- how is that? I wish the article was more clear and with more info. They also got the name of the coalition wrong. It's the Coalition for Good Governance*, not Governments.
KSU does not have the... "most academically rigorous" reputation. How they ended up handling election records instead of GaTech (literally 30 minutes away, and closer to the Capitol) boggles the mind.
But then, why would you possibly want to house sensitive records at a state university with an actual cybersecurity research focus [1]? That would just be crazy.
Most of the GA state legislature went to or has ties to UGA. Not saying that a school rivalry would affect their decisions, but not saying it wouldn’t either.
I wonder if it was a physical image backup or a logical image backup. If the latter then deleted data on the server would be likely lost if the original disks were reformatted.
Unless you’re aware of a lot more incidents in the past few years, I don’t think twice in 2 years makes for a new norm. The IRS found a backup of that hard drive, according to this article.
> In any case these events are making news because they are not normal.
What’s troubling is that we don’t have a good way to know how often this behavior actually occurs. Would the IRS have found the missing data if there hadn’t been press about it? I suppose that depends on your outlook and level of skepticism.
Or, put another way, I once asked my father if the state of political corruption in the news had always been as dour as what we see today. He said that this was the worst during his lifetime.
I then asked if he thought that the state of world affairs was actually worse, or if technology had simply made it more well known. He was silent for a long time before he answered. He said that he hadn’t considered that.
Finally, when I was a teen I was frustrated when I was lectured after I got caught sneaking out in the night. My mother, with a healthy degree of smug satisfaction, told me that she was always one step ahead of me. I couldn’t throw it in her face that I got away with it on a regular basis, of course.
You must assume that when you catch someone behaving badly once then there may be other incidents that you didn’t recognize. The worst aspect of corruption is not the time that we found out about it. It’s the question of how much we’ve missed.
It would seem a simple matter to charge each individual from the lowest levels with a straightforward crime, and flip them as they ladder up toward the people who gave the order/suggestion/plausibly-deniable-wish.
Simple, that is, if the prosecutor wanted to prosecute.
Or it could be a part of a well-scheduled maintenance schedule that was agreed by IT and the system owners.
I used to have a system like that under my purview, a voice recording system. Unless we were under court order or internal compliance instruction, all back up tapes were to be erased after 1 month of the conversation being recorded. This was a daily job for the VR system admins -- pull tapes from 1 month prior, erase, and recycle back into the recording system to be used again. No special instruction needed to do a BAU task.
If a court order was received a tape was pulled out of the general pool, and sent off to compliance for safeguarding. So if we received a court order, that order would go to CEO / COO, to head of legal &/or compliance, who would then instruct head of IT, who would instruct the VR system admin to pull a tape. If somebody dropped the ball in that chain on day 29 and the order didn't reach the VR system guy, the tape was erased the next day.
The importance of this monthly VR tape deletion process was important enough to have some extremely senior compliance person flying out to assure themselves it was actually happening, and happening according to this procedure.
Why 30 days only you ask? Because we were only obligated under law to maintain 30 days worth of recordings. Anything longer than that became a regulatory risk to us.
Under my VR system erasure policy, the guy in question who wiped the election system might not have anybody to point to as ordering him/her to wipe.
In which case he'd have the existing policy to point to as why the wipe was performed when it was, and the legal department to point to for not informing him of the lawsuit and to not delete those tapes. The organization is still liable, it merely shifts which individual within that organization is at fault.
True. But how is that relevant? Nobody can read the law (the annotated copy), so nobody knows that wiping the servers is destroying evidence, and therefore illegal? I doubt that that's how this went down...
The text of Georgias law is publicly available, but there is a version with annotations from legal precendents that is considered necessary to understand the law as it is currently interpreted. LexisNexis collates those annotations for the state of Georgia and maintains a copyright on the annotations, you cannot access them without paying. The government in Georgia feel this is perfectly acceptable as you can tehcnically access the law even though, in practice, you have no idea
The annotated version is freely available: http://www.lexisnexis.com/hottopics/gacode/default.asp. What the kerfuffle is about is that the State of Georgia has (1) Georgia law is highly unusual in giving the annotations authoritative force; (2) the State of Georgia has asserted copyright over the annotated version.
False. You don't need to pay to read the laws, but you do need to pay if you want to read a private groups annotations to the laws, which include things like judicial rulings. The annotations are a value add and not required to understand or practice law in GA. You can also access the judicial rulings and all other sources which the annotations are derived from for free, but you would need to put the work in yourself to compile that data.
It's not quite that clear cut. Georgia places unusual weight to these specific annotations so simply having access to the rulings will not result in the same exact output.
Georgia is somewhat unique in that the annotations actually are part of the codified law. You absolutely do need the annotations to practice, because they are authoritative. So not "false" after all.
The lack of paper ballots or hardcopy proof of voter intent in the GA system is unacceptable. Transparency and verifiability - using means comprehensible to the average voter - are more important to free and fair elections than efficiency.
> The server’s data was destroyed July 7 by technicians at the Center for Elections Systems at Kennesaw State University
I'm willing to bet that there was a whispered conversation between somebody in authority to a subordinate, promising to protect them if they'll take care of it.
Let this be a warning to everybody: you need to cover your butts. The pressure can be really great in situations like this, but if anything hits the fan you better believe the manager will toss you under a boss.
Notice it was wiped on July 7th, over 3 months ago and this information is barely being reported. It's also very peculiar the two backups were subsequently wiped and degaussed in early August. To top it all off, the FBI isn't sure if they still have the image or not? Isn't that highly confidential data that's critical evidence for a federal case?
> It’s not clear who ordered the server’s data irretrievably erased. The Kennesaw election center answers to Georgia’s secretary of state, Brian Kemp, a Republican who is running for governor in 2018 and is the main defendant in the suit. ... The server data could have revealed whether Georgia’s most recent elections were compromised by malicious hackers. The plaintiffs contend that the results of both last November’s election and a special June 20 congressional runoff — won by Kemp’s predecessor, Karen Handel — cannot be trusted.
That's not even the top problem with electronic voting in my book. Counting is.
If there's one central machine crunching all the raw data, you can bet your ass that every APT out there is going to be looking at the possibility of completely changing the outcome of the elections.
The way this problem is tackled with ballots is rather simple: way too many people to bribe to change the outcome.
> The way this problem is tackled with ballots is rather simple: way too many people to bribe to change the outcome
I think we do this correctly in New York City, where I am an election worker from time to time.
Voters mark a paper ballot that is electronically counted. Every counting machine continuously shows two numbers: the number of ballots it has ever received, and the number it has received in that day's election. Multiple people at each poll site sign off on these numbers at the start and end of each day. Lots of people, from parties, interested groups, et cetera, personally verify these numbers throughout the day. (We call these people "observers." Anyone can be one.)
At the end of the day, the poll numbers are printed directly from each machine and posted for observation. They are also transmitted, electronically, to HQ. Poll observers like to observe and independently record these numbers.
Finally, the NYPD carts away the paper ballots for archiving.
To corrupt this system, you'd have to overwhelm a majority of the poll workers, poll observers, and NYPD officers transmitting the paper ballots at every precinct in your jurisdiction. That's hard.
(The system isn't perfect. The sign offs we're supposed to do at day's end are done at the beginning, because people want to get out early. This weakens, but does not break, the chain of trust and custody I described above. TL; DR If you care about this, please sign up as a poll worker.)
TIL, sounds like a good system. Voting is not only selecting a leader or measure, but it also needs to be validated so that the people believe that the citizenry collectively made the decision. Observers, audit trails, multi-party sign off, etc make the electron results more believable.
If a single server can be wiped, on the other hand, there's no point in holding the election at all -- there's just too much room to quibble over the result.
Isn't this solved by transparency? Each precinct publicly reports their results, then news agencies and data junkies independently verify totals. Combine with a paper trail and random audit on a statistically significant set of precincts.
Having a paper ballot doesn't help if the verification process is broken. In this past presidential primary, a group of election observers signed affidavits that they witnessed counting monkey-business during a "5% audit". The reports assert that Sanders ballots were under-counted 21 and Hillary's were over-counted by 49 so that they matched the previously tallied and reported machine numbers. This issue emerged after Sanders campaign window to challenge the election results had past. What is frustrating is the lack of investigation into the issue; did this event actually occur as described -- were the numbers in a spreadsheet trusted over the actual paper ballots?
I agree, but considering a politician only gets paid while in office, it's in their own interest to make sure they sure they stay in office, regardless of the method they use. Hence why corruption in politics is a thing, why dictatorships are a thing, etc...
Wiping a server while under investigation or at the threat of a potential investigation should automatically be obstruction of justice. This happens far too often in politics and white collar crimes. Disgraceful
I hate to be that guy that brings blockchains into things, but isn't this like the one blockchains are really good at? Why haven't we switched technologies already?
Honest question, I'm all ears if there's a good rebuttal.
It's not enough for elections to be verifiable. They need to be easily verifiable by anybody.
Blockchains may be theoretically secure. But almost nobody is in a position to understand this thoroughly. The debacle around the DAO and frequent hacks of online wallets, even if those events are only tangentially related to blockchain technology, also undermine any potential trust.
It's also unnecessary: I just witnessed the federal election in Germany. In large cities, there's a polling place on every second block or so, each serving maybe 500 voters. The polling place is open to the public. Meaning you can see that the ballot box is empty in the morning, you can watch all day, and after polls close, you are free to watch the poll workers sort and count the ballots, as close as you want. You can then compare the count you witnessed to the official numbers posted online.
The whole process is end-to-end verifiable by anybody, and it requires no special skills other than counting.
A Dutch blog (geenstijl.nl) has used this to organise a initiative to count the votes for the 2014 European parlement elections, because the results only were to be announced days later. They had 1378 people sending in the results they got by waiting for the results in polling stations. And with those results they were able to make a very accurate exit poll.
Blockchain is not necessarily the only answer for everything. The important thing is that the voting scheme is end-to-end verifiable. Even paper would work, although less efficiently. But there are plenty of other technical solutions: https://en.wikipedia.org/wiki/End-to-end_auditable_voting_sy...
Blockchain would definitely solve this: Traceability, verifiability (no double casting). If you dont trust the counters (which no1 in america should) you can also use a form of Secure Multiparty Computation to hide the outcome from the counters until the last ballot is cast. So ya.
As a resident of Georgia, I really wish we had two states - metro Atlanta and the rest of GA. I lived in south GA from the time I was born until graduating from college and then moved to Atlanta. When I go back home it's like another world.
I live in Georgia too, so I'm trying to understand the point of your comment in the context of this server wipe.
Kennesaw is not far outside the Perimeter. I'd guess it counts as "Atlanta".
Why do you wish we had two states? Can diverse cultural elements in a state not come together to act in concert? What does that say about the whole notion of trying to increase "diversity"?
City vs rural, liberal vs conservative... I find that those aren't the important dichotomies that you use to judge the health of a culture or society. It's integrity vs corruption.
The state government is still very conservative because it represent rural Georgia far more than the population would call for. That’s partially because of gerrymandering and partially because of the way that most state governments and the federal government was designed.
Can diverse cultural elements come together and act in concert? Probably not. Why would a former factory worker in south Ga that has seen factories shut down left and right and go overseas and be taken over by automation ever see things the same way as someone in tech in Atlanta who hasn’t been affected? I think the whole 12 years of going to school, I never went to class with a non native English speaker. If your only exposure to any other culture is what you see on TV and never work side by side with people of different cultures, it probably is easier to believe that all Mexicans are rapists, and all Muslims are sitting around at home thinking about how they can kill infidels.
Like minded people tend to cluster together and self segregate.
The state government is still very conservative because it represent rural Georgia far more than the population would call for
You say that like it's a bad thing. Georgia is 47th in the nation in terms of per-capita debt. It boasts the busiest airport in the world, a thriving economy, a thriving tech community, a robust media culture, arguably the best bargain engineering degree in the world at Georgia Tech, a huge influx of outside influences over the past few decades - yielding a great deal of diversity and cosmopolitan attitudes.
Despite some cultural divides between Atlanta downtown and rural Georgia, the state seems to function really well. The state shows that you can have a mix of cultures without perpetual racial clashes, like you see in Missouri, Maryland, etc.
Why would a former factory worker in south Ga that has seen factories shut down left and right and go overseas and be taken over by automation ever see things the same way as someone in tech in Atlanta who hasn’t been affected?
They don't have to see things the exact same way. They certainly don't need a one-size-fits-all heavy-handed government solution to their diverse problem sets. The rural guy in Georgia whose job disappeared needs to find some new work and think about his future. I've moved multiple times across the country for jobs.... he can too, if necessary. Maybe he needs to further his education. Georgia has a fairly good university system. If his kids need education, there's the lottery-funded Hope Scholarship. With just a modest amount of work to keep your grades up, the Hope Scholarship funds 75% to 100% of your tuition!
Like minded people tend to cluster together and self segregate.
And that's a good thing?
Societies that balkanize rip themselves apart. History is replete with examples. Why would we want the same thing here where we have a success story to tell the world about?
Societies that balkanize rip themselves apart. History is replete with examples. Why would we want the same thing here where we have a success story to tell the world about?
I'm not saying it's a good thing. I'm saying it's a thing. As far as "success story". Things aren't too "successful" in places like Albany Ga - where factories have left, was once rated the murder capital of the US, and is consistently rated as one of the poorest cities in the US.
GA has both one of the poorest cities in America (Albany) and one of the wealthiest cities (Johns Creek) and one of the wealthiest counties (Forsyth County). Everything that you tout about GA isn't benefiting areas outside of metro Atlanta.
I've spent half my life in South Ga and the other half in metro Atlanta. It's two completely different worlds as far as opportunity, outlook, and world view.
But if you were gay, Muslim, etc. where would you rather live, Tyty Ga or Atlanta?
I deeply agree with your comment. I think that it can be applied to rural Spain too. It is a symptom of this 2-speed way of living. As I write this I am surrounded at work by a Polish, Romanian, French and Irish. The richness you get from this interactions cannot be surpassed by the TV or the Twitter echo chamber.
Same in NYC, but I don't think splitting states is the answer (or possible). Very few states would be intact if we started that. We should allow more legal autonomy for counties, to reduce intra-state political struggles. And maybe do "foreign" (intra-state) student exchange programs.
It’s not the liberal vs conservative. I am an African American living in a county that made national news 30 years ago for not exactly being welcoming to minorities. If you’re from GA you know where I’m talking about. Just search for “Oprah GA”. It’s definitely still very conservative but it feels more Romney conservative than Trump populist. I’ve never felt any weird vibes here.
Well, what I was getting at is that there are places in Georgia that aren't Atlanta metro that have far more in common with ATL than they do e.g., Collquit County.
It's the same as in most states - the divide is urban/rural and racial diversity. Places like Columbus, Savannah, Macon, or Augusta don't feel anything like the rural counties.
I generally agree with you. Honestly, I'd largely argue that the difference is a population that isn't both insular and predominantly white.
Election results, obviously, are not the full story, but they do provide a glimpse of the divides present within Georgia.
If you look at non-urban counties that voted Clinton in the last election[1], you'll find that the consistent theme is a substantial black population.
But even within cities, I'd argue diversity plays a role. Fulton and Dekalb counties were most heavily in favor of Clinton in the metro ATL area, and they're both majority-minority counties.
That doesn’t tell you the whole story. There were a few groups of Trump voters. The people who actually agreed with his populist, racists rants, those who really lean libertarian or business conservatives but think that Republicans lean more to their liking and held their nose and voted for him. The religious conservatives who may not like him as a person but see him as a means to an end to get their agenda passed. And then you have the single issue voter that doesn’t care about anything except their one pet issue that would vote for anyone that shared their views on that one issue — i.e. anyone against any type of gun control.
I have friends and former coworkers I still hang out with that represent all of those types of Trump voters - except for the nationalist, populist.
I feel that voting data should be made public record after de-identifying individuals. This is one area where stronger standards and norms may be helpful. If there was a market for third-party voter archival services at low costs, this would be far less likely.
The storage and technology should be very cheap, especially with the blockchain hype floating around. As skeptical as I am of new regulation, the requirement that de-identified voting data is made broadly available, and for all intents and purposes, immutable, sounds like a reasonable one worthy of further investigation.
Some governments already use similar services to archive emergency responses from social media, as is required by law in many jurisdictions[1]. That voter data, the foundations of the democratic US society, doesn't receive the same treatment is truly awful.
Thank you for the links, I had no idea about OSET, but find them really interesting. It's always good to look at prior work before rushing in with the new shiny technologies of the month.
> It is naive to trust that US elections are valid demonstrations of democracy
What do "trust", "valid" and "democracy" mean in this context?
Let me re-phrase your question in an attempt to answer it. Strong form: "Do American elections reflect the will of the populace?" Semi-weak form: "Do American elections reflect the votes of the voting populace?" Weak form: "Does the American system of governance work better than others?"
On the strong form, I don't know. Measuring the "will of the populace" and defining a nation's "populace" is a philosophical question.
On the semi-weak form, probably. For every Georgia there are 49 other states. The American electoral system is distributed. Flattening distributions brings buys resilience at the cost of moronic outliers (like this). The same factors that make coördinating systemic changes tough thwart efforts at electoral corruption.
On the weak form, very probably. We just don't have a multi century, multigenerational society that has generated as much wealth for as many people while confronting the number of difficult questions the American system has. It's far from perfect. But its base assumptions about separation of powers, peaceful change of guard and an independent judiciary appear correct. Notably, no competitors to the Washington Consensus boast those three attributes.
On the strong form, certainly and provably not. Studies have been funded to find any correlation between the will of the populace and policy that is enacted. There is zero correlation.
On the semi-weak form, obviously not. Voter suppression is constant and ubiquitous in most states. Gerrymandering has caused international organizations to classify states in the US as non-democracies, and explicitly suppresses the will of the voting population as its stated goal. It is extremely effective. The majority of the population of states vote for one policy position, and the polar opposite party will receive a two-thirds unbeatable veto-proof majority in the legislature. There is incontrovertible evidence that even presidential elections were explicitly and intentionally stolen (Gore) -- we know this happened. There is strong evidence to suggest the subsequent election was stolen as well, with the election in Ohio raising every red flag election monitors know to look for and some they hadn't yet invented because they were so blatant as to be unthinkable (the owner of the company that distributed all voting machines in the state explicitly, publicly promised to deliver the election to the candidate who won). While the system is in principle distributed, in reality it is anything but: a tiny number of very easily hacked, bought, corrupted, and infiltrated swing state electoral systems decide the fate of the entire nation, safe from the scrutiny and security requirements that an actual federal election system would have to undergo.
On the weak form, are you kidding? Nearly every electoral system in any modern democratic state is far superior. Your statements about peaceful change of the guard, separation of powers, and the judiciary (its independence is laughable; judges are explicitly selected partisans who dictate partisan policy positions from the bench instead of voting for it in congress) are all totally irrelevant to the question as it should be formulated in it's weak form: does the American electoral system work better than others? Nothing was asked or stated about the system of governance. You only even answer the question you asked by taking an extreme ideologically charged definition of "work better". The question you implicitly asked (that you answered) is: is the American system of governance effective at preserving a tradition of peaceful transfer of power from oligarch to oligarch? Has the obvious answer, yes. As does the system in North Korea.
>We just don't have a multi century, multigenerational society that has generated as much wealth for as many people while confronting the number of difficult questions the American system has.
We don't have such a society that has generated much wealth for the many; rather, for the few.
The only thing that keeps Americans from publicly admitting these obvious truths about the voting system is a need to believe that the system works; we accept that obviously undemocratic results are democratic because to accept the alternative (which happens to be true) is unthinkable and would cause chaos. It is a group lie we tell ourselves to make society function.
"We don't have such a society that has generated much wealth for the many; rather, for the few."
You really should get out more. Go visit some countries around the world. Study history beyond the 4th grade level. Your claim is so bad, it isn't even wrong.
Indeed; I have seen the "you are dumb and I am smart, you are a fourth grader" 'argument' before, though it's never been applied towards me until this point -- consider yourself a pioneer in that regard. I am honored; right wing idiots who have access to the internet and little else to offer have found my post and think that what I am saying threatens their worldview. There isn't a better sign in the universe that I'm doing something right.
> On the strong form, certainly and provably not. Studies have been funded to find any correlation between the will of the populace and policy that is enacted. There is zero correlation.
This is not the strong form parent was talking about. Here is the strong form:
"Do American elections reflect the will of the populace?"
American elections are not the same as the policy changes made by the people elected; You are arguing against something like "Does American governance reflect the will of the populace?"
Can you link to some of those studies? I'm very interested in seeing the method they used in determining the will of the populace.
Good question, but it seems like you already have an idea or an answer in mind. It might help if provide your opinion or a source and some more thoughts on it.
I think you might thinking they are not but, there are multiple angles to this, do want to talk about gerrymandering, foreign state actors hacking electronic voting machine, corrupt politicians interfering, our own intelligence agencies hacking our elections.
(From wikipedia)[0] Spoliation has three possible consequences:
in jurisdictions where the (intentional) act is criminal by statute, it may result in fines and incarceration (if convicted in a separate criminal proceeding) for the parties who engaged in the spoliation;
in jurisdictions where relevant case law precedent has been established, proceedings possibly altered by spoliation may be interpreted under a spoliation inference, or by other corrective measures, depending on the jurisdiction;
in some jurisdictions the act of spoliation can itself be an actionable tort
Can anyone comment on viability of recovering a multiple degaussing?
> The new e-mails, which were sent by the Coalition for Good Governance to Ars, show that Chris Dehner, one of the Information Security staffers, e-mailed his boss, Stephen Gay, to say that the two backup servers had been "degaussed three times."
I think the email itself is evidence of crime. Why bother with recovering the data when the absence of it allows the prosecution to portray the worst imaginable criminal scenario. They actually shot themselves in the foot because no matter how bad their crime is it cannot be as bad as erasing the data and degaussing 3 times. The prosecution can claim the absolute worst has happened. And they have no way to prove them wrong. And a murder weapon (the email)
Prosecution cannot claim anything without evidence. Now there’s only evidence of obstruction of justice by some IT manager, but no evidence to support charges of election fraud by executive branch of government.
Giving the benefit of the doubt to the prosecutor goes against the “innocent until proven guily” ideology. You’re basically saying, “he unknowingly deleted evidence he didn’t know was evidence. That proves it’s a conspiracy because we now don’t have evidence to prove our case”
It’s the reason cops are seen as immune to the law; The courts give them the benefit of the doubt a lot.
It's clearly evidence of knowingly deleting data. It may not be evidence of knowingly deleting evidence. The distinction matters, because deleting data is not a crime.
Degaussing 3 times. Who does that when deleting data? Unless they don't want investigators to find the data? And isn't the data in this case a public record? Why delete a public record indelibly unless they had manipulated that record and wanted to hide the fact..?
But then that would incriminate them even more! right? because it's not just "We erased it to make room for new data" ... This is in the realm of obstruction of justice.
Yes and no. If the "standard procedure" when they normally wipe the data as part of routine is to overwrite (Which I'd probably hope so) then the fact that they did so a) Makes it impossible to recover, and b) Doesn't make the action any more incriminating.
This is entirely separate from questions of obstruction of justice.
If that were actually true then we would see wild differentials between poles and election results. I admit the possibility that results may be shifted a few percent either way, enough to swing an election, but this will only ever be at the margins. Should the population be truly set on one candidate or the other (more than 5% difference) then rigging would differ from poling very significantly and be detected.
You'd hope so but in practice, this gets reported as the polls being off and there's speculation about ivory tower intellects not asking the right people or using outdated methodologies ... the news media and mathematicians get blamed for being wrong. This is what actually happens.
