National | CTVotersCount.org

Two days at the Voting and Elections Summit

Three simple ideas standout among the many things I learned and relearned:

When we are concerned about every cost associated with voting, small and large, compare those costs to what we spend “spreading democracy” elsewhere.
Contemplate what people spend in time and expense for the excitement of the Superbowl. Why are we not similarly engaged in Election Day, where the who wins is much more significant to our lives?
Should we be at least as concerned with protecting and auditing paper ballots, as we are with the footballs used in the semi-finals?

I always get rejuvenated with new ideas and camaraderie of a conference. For the last two days I have participated in the Voting and Elections Summit in Washington, D.C.

If you or anyone you know needs help with registering to vote or absentee voting the source of help is the U.S. Vote Foundation or the Overseas Vote Foundation.

Three simple ideas standout among the many things I learned and relearned:

When we are concerned about every cost associated with voting, small and large, compare those costs to what we spend “spreading democracy” elsewhere.
Contemplate what people spend in time and expense for the excitement of the Superbowl. Why are we not similarly engaged in Election Day, where who wins is actually much more significant to our lives?
Should we be at least as concerned with protecting and auditing paper ballots, as we are with the footballs used in the semi-finals?

Monday and Tuesday, I will be back at the Capitol considering what might be possible in the future, while wondering if we are willing to pay for a voting system worthy of the potential value of trustworthy elections, at the NIST Future of Voting Symposium II. Yes, I went to the 1st Symposium and Connecticut benefited.

Sanity returns (mostly) to Maryland

Paper Ballots Return to Maryland Elections. Once Maryland implements optical scan, there will only be five states left without a voter verified paper record for voting.

Paper Ballots Return to Maryland Elections <read>. Once Maryland implements optical scan, there will only be five states left without a voter verified paper record for voting. Apparently Maryland is very very slow at starting what they mandate.

The Maryland Board of Elections’ state administrator, Linda Lamone, said the coming year will be spent preparing polling center volunteers and educating voters.

Lamone said the General Assembly was given the go-ahead for the paper system in 2007, but only received funding for the switch last year. The cost to lease the machines alone is roughly $25 million, but that does not include additional costs such as storage and transportation of the units before and on Election Day.

She said one of the reasons for the mandate was that lawmakers “felt that it was safer” to have a paper record of every vote cast.

Lamone is know for opposing optical scan. We recall a Connecticut study in 2006 by TrueVoteCT’s Michael Fisher that demonstrated the total costs of acquiring and using optical scan at about half the comparable costs of DRE (Touch Screen) voting systems now in use in Maryland – Optical scanners cost a lot more, but you need only one and a backup to a polling place, you only have to program two, and the last longer.

We point out the obvious that Lamone must have had to transport and store the larger number of DREs as well.

Election Reformers as Entertainers?

Last week Paul Krugman had an interesting Economics editorial that by analogy can apply to some election reformers as well. In general, I agree with Mr. Krugman when it comes to Economics. Beyond that he has a skill making readable opinion pieces that make single points well, even if they are based on very detailed economic theory or analysis. This article is a little different in subject, yet makes an important, useful point that applies widely, including to election reformers

Last week Paul Krugman had an interesting Economics editorial that by analogy can apply to some election reformers as well: Economists as Entertainers <read>

In general, I agree with Mr. Krugman when it comes to Economics. Beyond that he has a skill making readable opinion pieces that make single points well, even if they are based on very detailed economic theory or analysis. This article is a little different in subject, yet makes an important, useful point that applies widely, including to election reformers:

According to researchers in Britain, more than half of the health advice that Dr. Oz gives is either baseless (there’s no evidence for his claims) or wrong (there is evidence, and it contradicts what he says). Julia Belluz at Vox tells us not to be surprised: “He is, after all, in the business of entertainment,” she wrote recently.

But the thing is, there are a lot of Dr. Ozzes out there, including in areas you might not consider the entertainment business…

But I now also suspect that the personality traits you need to be an effective entertainer working with not-so-much-fun subjects like health or monetary policy are inherently at odds with the traits you need to be even halfway competent.

If Dr. Oz were the kind of guy who pores over medical evidence to be sure he knows what he’s talking about, he probably couldn’t project the persona that wins him such a large audience. Similarly, a hired-gun economist who actually knows how to download charts from economic databases probably wouldn’t have the kind of blithe certainty in right-wing dogma that his employers want.

How does this apply to election reformers?

It seems to me, in my experience, that many reformers are always rational and scientific in their approach to any issue, even though they may have biases. We try hard at CTVotersCount, in posts, and in testimony, to provide honest, accurate and comprehensive information about election issues even when the evidence hurts our case. (Hopefully, if there is enough evidence against our position we will change our position, if the weight of accumulated evidence points away from our past position.)

Yet, we run into others who frequently tout things that are unproven, untrue, and contrary to evidence. For example:

Legislators and advocates that ignore the evidence of electronic vulnerability to claim that Internet voting is safe. Despite of the evidence of hacking of business and government; despite the opposition of security experts and computer scientists, including those at Homeland Security, National Institute of Standards and Technology, and Department of Defense studies; despite the fact that online banking is safe only because banks pay billions annual to reimburse for theft.
Advocates that claim that early voting will increase turnout, when it actually decreases turnout. (There are reasons to be for and against early voting, but turnout is not a reason to be for it.)
Advocates for Voter ID who claim, despite evidence to the contrary, that there is a lot of votER fraud. Despite several witch hunts after every election, votER fraud is very very low.
Advocates for absentee/mail-in voting that claim that no votER fraud means that such voting is safe. They ignore and deny the actual evidence of frequent multiple-ballot votING fraud, that is accomplished by mail.
Advocates for the National Popular Vote Agreement that claim, for no reason, that fraud and suppression will not increase in all states, just as we have seen in some swing states, if we go to a popular vote without a uniform, enforceable voting system to match.
Advocates for the National Popular Vote Agreement who claim there could be a recount after a popular vote, when half of states have no recount law, all laws are based on close state votes – not close national votes, and that there is no body to declare and manage a nationwide recount.

Just some of the examples analogous to Paul Krugman’s Economic Entertainers. Yet, we have a hard time calling them Entertainers. Some are incompetent, some deluded, and others we might call advocates or lobbyists for hire.

Close vote highlights concerns with mail-in elections

There is a very close vote on a GMO issue in Oregon, creating a statewide recount and lots of controversy. Brad Friedman has an excellent summary of the controversies and also election integrity concerns: <read>

Oregon GMO Ballot Measure ‘Recount’ Highlights Concern About Excluded Vote-by-Mail Ballots

With razor-thin 812 vote margin, proponents charge 4,600 ballots remain illegally uncounted due to ‘problems’ with signatures

Questions about state procedures underscore shortcomings of Vote-by-Mail balloting…

The entire article is worth reading. Without taking sides here (although personally, I strongly favor GMO labeling) from an election integrity standpoint:

It seems likely, if all the ballots were counted, the Yes would win.
Unless there was significant organized votING fraud most of those ballots would in an ideal world deserve to be counted.
There will always be legitimate questions based on the potential for such organized fraud, since it has been proven to occur quite regularly. There will be charges with or without proof, and claims of lack of proof indicates a lack of fraud.
Its very difficult to accurately compare large numbers of signatures without many that should be questioned are not be questioned and with many that are questioned being actually legitimate.
Any signature check should be subjected to peer-reviewed evaluation of its legitimacy to detect fraud, and avoid false negatives as a prerequisite for considering mass absentee or mail-in voting. (We are not aware of any such evaluation of any system in use for elections).
Beyond the risks to actual integrity, credibility is lost on the part of all those who question fraud and all those who question rejections of ballots. Legitimately both are in question here.

In the specific case here:

We agree with Brad than citizens should be able to observe elections, and the requirement of being an elector in Oregon is too restrictive.
On the other hand, there needs to be limits on the number of observers who can observe closely, legitimate representatives of both sides, especially those from Oregon deserve some level of priority.
And there must be limits on behavior of observers, and protocols for the surfacing and the handling of objections.
We believe that it was sufficiently demonstrated that there were significant disruptions in the 2000 Florida recount that should have been controlled. Without more details and perhaps video’s we are not in a position to judge the extent and legitimacy of the charges of disruption in this case, in Oregon.

Will it take a Pearl Harbor or 9/11 for Internet security (and voting integrity)?

To err is human, to react without thinking is to compound the err

The goal should be to solve a problem of huge risk, without requiring a catastrophe, without attacking others, spending what is necessary and moving on.

Let us also not forget the twin risks of doing nothing and doing too much of the wrong thing, apply as Connecticut tackles our voting system which may have had a wake up call this November, but nothing like Pearl Harbor or 9/11.

To err is human, to react without thinking is to compound the err

A thoughtful post at the NY Times, that deserves a better title: Hacked vs. Hackers: Game On <read>

The problem, Mr. Kocher and security experts reason, is a lack of liability and urgency. The Internet is still largely held together with Band-Aid fixes. Computer security is not well regulated, even as enormous amounts of private, medical and financial data and the nation’s computerized critical infrastructure — oil pipelines, railroad tracks, water treatment facilities and the power grid — move online.

After a year of record-setting hacking incidents, companies and consumers are finally learning how to defend themselves and are altering how they approach computer security.

If a stunning number of airplanes in the United States crashed tomorrow, there would be investigations, lawsuits and a cutback in air travel, and the airlines’ stock prices would most likely plummet. That has not been true for hacking attacks, which surged 62 percent last year, according to the security company Symantec. As for long-term consequences, Home Depot, which suffered the worst security breach of any retailer in history this year, has seen its stock float to a high point.

In a speech two years ago, Leon E. Panetta, the former defense secretary, predicted it would take a “cyber-Pearl Harbor” — a crippling attack that would cause physical destruction and loss of life — to wake up the nation to the vulnerabilities in its computer systems.

No such attack has occurred. Nonetheless, at every level, there has been an awakening that the threats are real and growing worse, and that the prevailing “patch and pray” approach to computer security simply will not do.

I agree that the problem is huge. We should hope that it does not take an attack like Pearl Harbor or 9/11 to change things. How would World War II have gone without Pearl Harbor – I suspect not much different. I am not a historian. I was not alive then, but overall our reaction to Pearl Harbor was on balance justified, appropriate, and successful. I do not think that 9/11 worked out that way, our wars “of choice” in Iraq and Afghanistan have yet to be successful, have been arguably unjustified and inappropriate as well. They certainly have been costly with no end in sight. When it comes to security, again the Patriot Act was a knee-jerk reaction, with every wishlist item of the security state fulfilled. It is questionable that the fortune and liberties we have sacrificed have been worth it or that all in all we are safer.

The goal should be to solve a problem of huge risk, without requiring a catastrophe, without attacking others, spending what is necessary and moving on.

That has happened once that I know of. It was called Y2K, a disaster avoided, a significant yet limited expense. Y2K was real, those warning about it in the late 1980’s were ignored for many years. The ultimate risk was overblown by the media, then when all went well we had years of poopooing the risk as overblown. For the record, I was a Y2K contractor for a bit over two years for three companies – I did small jobs that needed to be accomplished, where I was uniquely qualified. There were excesses. In fact, I helped save a client from a wasteful proposal. Yet overall we solved and prevented a problem that could have been avoided at a lower cost if more leaders had listened to those who warned us early. Even now, occasionally someone in a discussion will complain about “all the money computer programmers took home working on Y2K”, as if that caused our deficit. Yet, it is worth it to me, it to know that a real problem was avoided, despite the occasional uninformed criticism.

Yet as this article points out, we have already paid a huge, largely unrecognized price for Internet vulnerablity:

The Wake-Up Call
A bleak recap: In the last two years, breaches have hit the White House, the State Department, the top federal intelligence agency, the largest American bank, the top hospital operator, energy companies, retailers and even the Postal Service. In nearly every case, by the time the victims noticed that hackers were inside their systems, their most sensitive government secrets, trade secrets and customer data had already left the building. And in just the last week Sony Pictures Entertainment had to take computer systems offline because of an aggressive attack on its network.

The impact on consumers has been vast. Last year, over 552 million people had their identities stolen, according to Symantec, and nearly 25,000 Americans had sensitive health information compromised — every day — according to the Department of Health and Human Services. Over half of Americans, including President Obama, had to have their credit cards replaced at least once because of a breach, according to the Ponemon Group, an independent research organization.

But the value of those stolen credit cards, which trade freely in underground criminal markets, is eclipsed by the value of the intellectual property that has been siphoned out of United States corporations, universities and research groups by hackers in China — so much so that security experts now say there are only two types of companies left in the United States: those that have been hacked and those that do not yet know they have been hacked.

And this year, American companies learned it was not just Beijing they were up against. Thanks to revelations by the former intelligence agency contractor Edward J. Snowden, companies worry about protecting their networks from their own government. If the tech sector cannot persuade foreign customers that their data is safe from the National Security Agency, the tech industry analysis firm Forrester Research predicts that America’s cloud computing industry stands to lose $180 billion — a quarter of its current revenue — over the next two years to competitors abroad.

Finally, let us also not forget the twin risks of doing nothing and doing too much of the wrong thing, apply as Connecticut tackles our voting system which may have had a wake up call this November, but nothing like Pearl Harbor or 9/11. On 9/11, I had a temporary pass to enter the World Trade Center and had friends that worked there – what happened in Hartford on November 4th, and the Courant not getting all the results that night was no 9/11.

Connecticut not alone in election adminstration challenges

Since the election on November 4th we have had all sorts of complaints about Connecticut election administration. Claims that we are the slowest, with the most clueless election officials. And all sorts of cures proposed including more mail-in votes, electronic calculation of results, and reorganization of election administration.

We agree with that their are many problems. We agree with the general outlines of some of the cures. Yet, we caution against knee-jerk reaction, and change without planning and analysis.

We suggest looking at the best practices from other states. Yet, we can also learn from the mistakes and foibles of other states. Often those employing some of those very cures proposed for Connecticut.

We agree with that their are many problems. We agree with the general outlines of some of the cures. Yet, we caution against knee-jerk reaction, and change without planning and analysis.

Lets look at the recent news:

11/17 NJ not so quick in reporting results Using equipment from the same vendor as Connecticut, NJ has problems, delays, and investigations of slow accumulation/reporting of results electronically. Then again, some other states below reported fast, with much less accuracy than Connecticut or New Jersey.

1/25 Mail voting: Not so fast, not so easy, not so simple Take Oregon and their all-mail voting, please. A highly charged ballot question is yet to be decided. In fact they have just counted enough votes to realize they need a recount. Here is the issue, some 13,000 votes were not counted because of possible signature mismatches. So advocates contacted voters after the election to see if they actually voted and requested they come in and sign their ballot or show their signature changed.. We have some of our own issues with all this:

Just how good is their signature matching? Has anyone evaluated their methods. What are the odds they missed more questionable signatures? How many of those 13,000 should not have been questioned?
Does the result depend on which side got more voters to come in and sign (demographics can indicate how a voter might have tended to vote)
And we complain that some results in Connecticut were not available until Nov 5th?
PS: This problem will never happen in Connecticut as we never match signatures. (See no evil…)

11/25 MN lowest turn out since 1986 Many claim, anecdotally and incorrectly, that no-excuse absentee voting is a panacea for increasing turnout. Apparently, anecdotally, it has not helped Minnesota all that much.

11/25 The Maine question: Will 21 mystery ballots change looser into winner? Connecticut has problems with ballot counts not matching check-in list counts, and a greater problem with some officials not checking that those numbers match. At least in Maine there is a recognition that this might be a problem, especially if extra ballots are found after the initial count.

1/26 Electronic result totals not alwasy even close to accurate Here we go again with that electronic tallying of votes. They only missed about one-third of the votes. Fortunately, a news outlet found the error. They say the problem has been fixed, yet sounds more like the error has been corrected in the results of this one election. They are not counting accurately in Kansas any more.

The Perfect Storm meets illegal, disasterous voting

When hurricane Sandy hit the east coast, New Jersey exposed the vote and voters to additional risks.
New report from Rutgers:

The Perfect Storm: Voting In New Jersey In The Wake of Superstorm Sandy

When hurricane Sandy hit the east coast, New Jersey exposed the vote and voters to additional risks.

New report from Rutgers: The Perfect Storm: Voting In New Jersey In The Wake of Superstorm Sandy <read>

CONCLUSION

After Superstorm Sandy, there was no structure in place to make sure that emergency voting directives were followed. There was mass confusion among county officials and voters, alike. Emergency measures such as Internet and fax voting not only violated New Jersey law, but also left votes vulnerable to on-line hacking. Internet voting should never be permitted, especially in emergencies when governmental infrastructure is already compromised.

As the May 2014 National Climate Assessment issued by the U.S. government makes all too clear New Jersey is highly likely to be impacted
negatively by more Superstorm Sandy-like disasters in the near future. This means that it is critical for New Jersey to enact and implement emergency voting procedures that comply with existing election law, and that protect every vote. As such, those emergency measures should
not include Internet and fax voting as an option, under any circumstance.

The report is a quicker read than its size would indicate. It is packed with details that expose the risks of email and fax voting.

How can we vote on Internet that is unsafe for banks, Canada, and alarms the President?

Recent articles highlight the folly and blind faith in technology leading many to trust voting on the Internet.

As Roosevelt said “The only thing we have to fear is fear itself.” seems to apply here.

There are real cyber risks, we need to protect or digital assets. Yet it does not help to jump to the conclusion that every breech is the work of our biggest enemy of the moment.

Like building new civic centers, baseball stadiums, and bankrolling fishing and hunting retailers there is plenty of real world evidence that Internet voting does not work well, yet we persist despite the evidence. Apparently the technology that actually works to protect Democracy, a technology actually under assault in Connecticut, is Freedom of Information.

Recent articles highlight the folly and blind faith in technology leading many to trust voting on the Internet

From the NYTimes: Obama Had Security Fears on JPMorgan Data Breach <read>

President Obama and his top national security advisers began receiving periodic briefings on the huge cyberattack at JPMorgan Chase and other financial institutions this summer, part of a new effort to keep security officials as up dated on major cyberattacks as they are on Russian incursions into Ukraine or attacks by the Islamic State.

But in the JPMorgan case, according to administration officials familiar with the briefings, who would not speak on the record about intelligence matters, no one could tell the president what he most wanted to know: What was the motive of the attack? “The question kept coming back, ‘Is this plain old theft, or is Putin retaliating?’ ” one senior official said, referring to the American led sanctions on Russia. “And the answer was: ‘We don’t know for sure.’ ”

More than three months after the first attacks were discovered, the source is still unclear and
there is no evidence any money was taken from any institution.

As Roosevelt said “The only thing we have to fear is fear itself.” seems to apply here. There are real cyber risks, we need to protect or digital assets. Yet it does not help to jump to the conclusion that every breech is the work of our biggest enemy of the moment (e.g. China last year, Putin here, and ISIS last week) when we apparently don’t have a clue.

Just as irrational is the fear in then Connecticut Legislature, (and perhaps in the statehouse) where many voted for Internet voting under the threat of being deemed “unpatriotic”.

Of course, Internet voting is not banking. Internet voting is more vulnerable, and more critical to our Democracy. As highlighted by this recent report:

From GMA News: Online voting not ready for worldwide roll-out, study concludes <read>

The research, produced by the Atlantic Council think tank and the online protection firm McAfee, concluded that “security will need to be vastly improved” before it becomes feasible to adopt Internet voting on a large scale.

According to the study, online voting faces more complex obstacles than electronic commerce, where a customer can be reimbursed in the case of fraud or theft.

“Online voting poses a much tougher problem” than e-commerce, the report said.

“Lost votes are unacceptable… and unlike paper ballots, electronic votes cannot be ‘rolled back’ or easily recounted.”

The report said hackers could paralyze an online voting system or, even worse, change the results without being detected.

A major problem of online voting is that any system must verify the identity of the voter, and at the same time guarantee anonymity in the process.

Some experts believe it could be decades before online voting becomes mainstream.
Joseph Hall of the Center for Democracy and Technology said that many security experts believe “the timeline will be 30 to 40 years” before the technological hurdles to online voting are overcome.

One of the problems is the “uncontrolled platform,” in which voting software or computers can be infected, Hall said at a discussion hosted by the Atlantic Council.

Jordi Puiggali of the online voting technology firm Scytl said that while Internet balloting has not been perfect, “we have to consider the risks of voting channels that already exist,” citing practices such as stuffed ballot boxes.

The researchers cited a study released earlier this year by University of Michigan scientists on online voting in Estonia, the first country to hold national elections on the Internet.

That study, which is to be published in a scientific journal next month, revealed vulnerabilities in Estonia’s online voting system.

“Attackers could target the election servers or voters’ clients to alter election results or undermine the legitimacy of the system,” the study said.

Estonian officials have maintained that the system is secure.

Wednesday’s report said that online voting has enormous potential if security can be improved.

“For the digital generation, unsupervised polling via mobile devices may be the ‘killer app’ of e-voting,” the report said, adding that biometric and other security features may need to be perfected.

“Broad adoption of most new technologies generally takes longer than technology optimists hope, but it will happen,” the report added.

“Online voting’s potential benefits in terms of reach, access and participation have the potential to revolutionize the democratic process around the world.”

Count us among the skeptics that Internet voting will be safe in 30-40 years. We say it is a good bet that 20 years from now it it will still be 30-40 years off, and maybe that will be the last we will hear of it. On the other hand it might be possible with a radical redesign of the underlying Internet. (Geeks like myself will remember IP 6, which we were all supposed to be using by about 10 years ago. Great news its up to 4% now.)

From Aljazera: Latest Internet voting reports show failures across the board <read>

Internet voting, a technology often cited as a solution to the United States’ problematic voting machines, received failing security and accessibility grades in the latest in-depth audit conducted by the City of Toronto. Two of the three vendors audited by the city currently have contracts with over a dozen U.S. jurisdictions for similar technologies.

The accessibility report, prepared by researchers at the Inclusive Design Research Centre at OCAD University, and the security report, prepared by researchers at Concordia and Western universities, were obtained by Al Jazeera America through a Freedom of Information Act

request.

Proponents of Internet voting, largely disabilities groups and advocates for military voters overseas, point to the apparent ease-of-use of other Internet-based activities, such as banking, and claim the technology would lead to higher turnout rates.

The reports highlight the difficulty in creating a voting system that isn’t more susceptible to corruption than existing voting technology and that is easy enough to use for voters with a variety of personal computer setups, including those with disabilities who often use alternatives to traditional mice, keyboards and screens.

Got that? Susceptible to corruption. And does not provide expected benefits. Sounds a lot like those civic center, sports stadium, and fishy retail projects! Meanwhile the U.S. Government continues to stonewall:

A nonprofit watchdog group, the Electronic Privacy Information Center, sued FVAP last month to force them to disclose their own audits of Internet voting conducted three years ago. In 2012 the program told Congress it would release the records to the public by the middle of 2013.

“It happens all the time.” All over the place (Part 6)

Here we continue our review of some of our posts of past errors surfaced in Connecticut and around the country, selected from our over 900 posts. Last time we covered CTVotersCount posts from the 1st half of 2011.

Here we continue our review of some of our posts of past errors surfaced in Connecticut and around the country, selected from our over 900 posts. Last time we covered CTVotersCount posts from the 1st half of 2011. Continuing from there:

Smoke, no fire seen. Once again see no problems <Jul 2011>

O Me O My O – Errors in Jersey and Fraud in Ohio <Jul 2011>

Colorado election issues, extreme, yet considered normal, reported in 2011 <July 2011>

State causes confusion in Bridgeport process for ballot access <July 2011>

P.T. Barnum’s city continues three-ring election circus <Aug 2011>

Ballot Skulduggery in Wisconsin? Or Inadvertent Errors? <Aug 2011>

No need to see the ballots, they are “sacred” – Faith based elections <Aug 2011>

Another election in Bridgeport…more absentee ballot allegations <Sept 2011> <Sept 2011>

Organized vote fraud in Florida (Sunshine anyone?) <Sept 2011>

CO: Chain-Of-Custody and Confidence broken <Sept 2011>

Rigged election or three human errors – we may never know for sure. <Sept 2011>

See no problems. Hear no problems. Keep observers out of polling places <Sept 2011>

Bridgeport…a dozen decrees in absentee fraud since 1988 <Oct 2011>

See no problems. Hear no problems. Stop the audit <Oct 2011>

Scanners count same ballots with different results <Nov 2011>

Its an error. No its a typo. Wrong person elected <Nov 2011>

Voting machine investigation leads to serious issues and cover-up <Dec 2011>

Secretary of State removed for vote fraud <Dec 2011>

That brings through the end of 2011. Till next time.

Warning #1: Your absentee or mail-in vote might not count

Voters considering the Constitutional Amendment on the ballot this November and legislators considering what to do if it passes, need to pay heed to the facts and experience of early voting in other state. Common sense is not always a reliable guide.

Did you know that when you vote absentee or mail-in, you might be disenfranchised at a much higher rate than if you voted at the polls?

There are three major reasons that we are aware of:

Your ballot might not be received in time to be counted. It might be delayed or actually lost in the mail. In Connecticut such votes must be received by 8:00pm to be counted.
You might have made a mistake in the somewhat complicated process of placing your ballot inside outer and inner envelopes, or forgetting to properly fill out the envelopes. The law requires that ballots with certain errors be rejected.
You might have overvoted i.e. voted for more candidates than allowed. This can happen when you make an inadvertent mark in a bubble or if you miscount the number of votes in a vote for multiple race. Such errors are caught for you when you vote in a polling, where the error is explained and you are offered a chance to vote again on a new ballot.

Here is a recent article on a study covering some of those problems, in California where mail-in voting is an increasing factor in elections California: Mail-in-ballot rejections analyzed in study | <read>

Voting by mail surpassed 50 percent of votes cast in a general election in California for the first time in 2012. A new study shows that nearly 69,000 mailed ballots, or about 1 percent, were not counted, and why they were rejected. The top three reasons mail-in ballots were rejected: not arriving on time, not being signed or because signatures could not be verified…

Romero said. “People have taken the time to study the issues, fill out the ballot and mail or deliver it. They trust it is going to be counted.”