Other Editorials | CTVotersCount.org

Courant Editorial: Why Is State’s Election Website So Bad?

The Courant has done a service to the voters of Connecticut pointing out a flaw in the Secretary of the State’s web site that does not serve voters. The editorial is not an exaggeration, it might be a bit too kind. UPDATED.

Just a bit of background. What is the information voters most want from an elections web site, state or local? One study says it is not how to register, how to do an absentee vote, nor where to vote. Those are important. Yet, what most voters want to know, most of all, is “What is on the ballot?”.

From the Courant:

Why Is State’s Election Website So Bad?

Want to know which candidates are running in Tuesday’s primary elections? Good luck finding them on the secretary of the state’s website.

This website is an unnavigable embarrassment. Voters should be able, with a click or two, to find out who’s on the ballot. Is that too much to ask?

The secretary’s office issued a news release July 31 saying a list of candidates was posted on http://www.sots.ct.gov/vote. But the list posted midday Friday was in the wrong place and with some of the wrong names.

And just getting to that inaccurate list was a labyrinthian exercise in frustration.

As of lunchtime Friday, the list of contenders for the Aug. 12 primary for governor and lieutenant governor was blank. The candidates were actually listed under the wrong date — “11/04/2014-General.” (Go figure.) Friday’s list included candidates not running in Tueday’s primary.

The secretary said that would be changed Friday afternoon.

This is important information. There’s no excuse for the state not getting it right — and in a place easy to find — just before an election.

We investigated by finding the same information the Courant did. Unlike a most voters, we are familiar with the Secretary’s web site, yet perhaps we are a little less patient. We found the same information (or lack thereof) the Courant found. Then we went back and traced the steps a very patient and persistent voter might traverse:

First we looked under “Voter Information” and found the link for “Town by Town Sample Ballots”. But the latest information there was for last November’s election:

Then after looking for a while, we found the link. Going back, attempting to pretend that we were that patient voter. We got to this screen about six (6) clicks from the home page. Not under “Voter Information”, but under “Election Information and Statistics”:

Note: it has no listing for the Primary 4 days away, only for the general election, and completely missing the race for Governor. The race for Governor might be missing because the candidates are incomplete, due to the Primary. But then why is Lt Governor listed, which is also subject to primary in the same party?

We are still two clicks away from a PDF for each individual office, which must be accessed separately, rather than one complete list, better delivered as a html web page.

Finally, we note that apparently the Secretary’s Office did not do as promised yesterday, and make the link touted in the Press Release have the promised information. In any case, we doubt many voters would expect to need to read press releases to find links to election information.

UPDATE: As David Bidell points out in the 1st comment, as of sometime this morning it was updated. I might have missed it earlier this morning.

NY: don’t follow CA in making “Top Two” error

NY considers “Jungle Primary” we call it another “Centrist Dream”

Political Insanity – doing the same thing that has failed elsewhere, over and over.

NY considers “Jungle Primary” we call it another “Centrist Dream”.

Political Insanity – doing the same thing that has failed elsewhere, over and over.

We have positions on a variety of election issues. The Top Two Primary is one we oppose with a moderate level of intensity. It is a failed idea, not a complete disaster, hopefully an idea that fails after a few tests in States. We and Ralph Nader warned California. Now it is time to warn New York not to make the same mistake..

For once we have progressive opposition from the American Prosepct/Madison Capitol Times editorial: ‘Jungle Primary] Brings Untended Consequences <read>

Would the dysfunction of U.S. politics be dispelled if we got rid of partisan primaries? That’s the contention of Sen. Charles Schumer, D-N.Y. In an op-ed for The New York Times, Schumer argued that the primary system in most states, in which voters choose nominees for their respective parties who then run head to head in November, gives too much weight to the party faithful, who are inclined to select candidates who veer either far right or far left. The cure Schumer proposes for this ill is the “jungle primary,” in which all primary candidates, regardless of party, appear on the same ballot, with the top two finishers, again regardless of party, advancing to the general election. The senator cites the example of California — once the most gridlocked of states, now a place where legislation actually gets enacted — as proof that such primaries work. But Schumer misunderstands what got California working again. In so doing, he also misses the fatal flaws of the jungle primary…

Schumer misunderstands what got California working again. In so doing, he also misses the fatal flaws of the jungle primary…

And what has the jungle primary accomplished? Its adherents had hoped that, in heavily conservative districts where the top two primary finishers were both Republicans, the more centrist of the two would win the November runoff by corralling more Democratic and independent votes. So far, however, that hasn’t happened. Democrats representing more centrist districts, generally in inland California, do tend to be less liberal, but that was the case long before the jungle primary came into effect.

The jungle primary has had one stunningly perverse effect, however. In a new congressional district east of Los Angeles, Democratic voters had a clear majority — so clear that four Democratic candidates and two Republicans sought the seat in the 2012 primary. Democratic votes split four ways, enabling the two Republicans to advance to November’s ballot. The eventual winner, Gary Miller, chose not to run for re-election this year — understandably, since his record in no way reflected the desires of most district voters.

A weird one-off result? This June, three Democrats and two Republicans sought the statewide office of controller. More Democrats than Republicans tend to file for statewide office in California, and for good reason: The GOP is in free-fall in the state; its share of registered voters has dropped beneath 30 percent; just one Republican (Arnold Schwarzenegger) has been elected to any of California’s 10 statewide offices in the past 20 years. But since Democrats split their votes three ways for the controller’s slot and Republicans just two, a shift of less than 2 percent of the vote would have saddled voters with a Republican-vs.-Republican runoff.

And perhaps the most interesting prediction in the piece, based directly on the example above.

Fast-forward to 2018, when Democrat Jerry Brown, almost certain to be re-elected this November, will be term-limited out of the governor’s office. More Democrats than Republicans will surely line up to succeed him. But under the jungle rules, even though it’s all but certain that the Democratic candidates will collectively aggregate more support, it’s a distinct possibility that two Republicans will face off in November.

This is your solution, senator? Think again.

This all goes to show why it is called a Jungle Primary. We prefer to call it a Crap Shoot, because, like many reforms, its intention is to correct a perceived past problem, but just replaces one imperfect system with another – fighting the last war.

(And who said that centrism is a worthy goal – its usually defined only in the vision of a particular pundit or politician, completely in agreement with that pundit’s own views)

Senator Schumer’s Op-Ed in the New York Times: End Partisan Primaries, Save America <read> Anyone should be suspicions of such a dramatic headline “Save America”.

And a couple of informed letters the Op-Ed generated to the New York Times make additional points <read>

Rather than keeping fringe characters out of office, an open primary would allow a pair of well-funded zealots to lock up a “top two” primary by driving turnout of a few passionate voters in what are typically pretty apathetic primary environments, especially in a gerrymandered district…

California’s June primary under the new system resulted in the lowest voter turnout ever for a statewide election, with just 25.2 percent of voters participating.

The “top two” system has also led to a number of legislative and Congressional districts where voters will be able to choose only between candidates of one party in November, which is hardly democratic. Despite the fact that Democrats have an overwhelming registration advantage in the state, the top-two system nearly resulted in no Democratic candidate on the ballot for one office when multiple candidates split up the Democratic vote.

In addition, it has had the effect of eliminating third parties and write-ins from the November contest, giving voters less choice.

And perhaps the worst result of all — the proliferation of independent expenditure committees, which spent more nearly $4 million to try to win a Bay Area Assembly seat for corporate interests.

Worse than Hurricane Sandy? As bad as climate denial?

A new ZD-Net Editorial: Internet voting: A really bad idea whose time has come

Summary: Believe it or not, most states have some provisions for allowing people to vote over the Internet. The pressure is on to expand it, even though a secure online voting system is impossible using today’s technology.

Climate denial might end human life or at least life as we know it. Internet voting denial can only wipe out our democracy.

A new ZD-Net Editorial: Internet voting: A really bad idea whose time has come <read>

Summary: Believe it or not, most states have some provisions for allowing people to vote over the Internet. The pressure is on to expand it, even though a secure online voting system is impossible using today’s technology.

We have been saying it over and over. In Connecticut, only the Secretary of the State, Denise Merrill gets it. The Senate and House have passed Internet voting twice, unanimously. Governor Malloy vetoed it the first time because it is risky and unconstitutional, nothing changed, yet he signed it the next time.

Its similar to ‘climate denial’, in that both ignore the risks, ignore the science. In Connecticut its thus proven that science denial is not just for Republicans – here its bipartisan and almost unanimous. We do not want to compare the two exactly. Climate denial might end human life or at least life as we know it. Internet voting denial can only wipe out our democracy.

The area on the Jersey shore where I grew up was hit very hard by Hurricane Sandy in 2012. It was many weeks before some of the people could even go home. Life was a mess. And then, a little over a week later, was the 2012 election day.

The state made it clear that they would make whatever accommodations it could to help people vote if they were displaced by the storm. So far, so good, but my ears perked up when I heard about “email voting.”

Yes, the state announced that voters could email in a vote. This was part of an effort to make all non-traditional forms of voting, including mail-in and fax, easier. In fact, voters were instructed to ignore the part of the relevant web page where it says “The County Clerk cannot accept faxed or emailed copies of a Application for Vote by Mail Ballot, unless you are a Military or Overseas Voter, since an original signature is required.”

But certainly such circumstances were sui generis, and no sane state authority would contemplate Internet voting in the normal course of things, right? Wrong…

Speaking of around the world, Estonia is the current poster child for electronic voting. Estonians at home and around the world can vote online using a national ID card, a smart card. Clearly a system of digital national IDs has no chance of being adopted in the US, but for all its sophistication, the Estonian system is still vulnerable to tampering according to recent research…

In fact, it’s easy to find research by people who understand computer security pointing out the considerable risks from internet voting. There are other people who would like to increase turnout no matter what and who are happy to declare that all technical problems can be worked out by the experts. Well, the experts have spoken: Internet voting is not and cannot be made secure with current technology.

NPV – A graphic scenario

We are opposed to the National Popular Vote Compact, primarily because it would tend to make the current risky system for determining the President by the Electoral College, much more risky and subject us to open-season for vote suppression, insider fraud, outsider fraud, and legal challenges likely to end with the Supreme Court choosing the President. Paul Choiniere of The Day provides a graphic depiction of a related scenario.

Article published Mar 16, 2014
Connecticut should reject popular vote scheme
Paul Choiniere

The National Popular Vote folks are at it again in Connecticut. They want to end the practice of using the Electoral College system to determine the election of the president. Instead, the candidate who gets the most votes nationally would become president.

They don’t, however, want to amend the Constitution to make this change – the way it should be done – because the movement apparently sees that goal as too difficult.

Instead, its backers seek a clumsy and byzantine approach, one wrought with potential problems. What might happen if the Connecticut General Assembly endorses the plan, enters into the compact and the governor signs the legislation?

Picture this.

Connecticut voters make a clear choice who they want to be president. The state’s support is enough to give that candidate the electoral votes necessary to become president. However, because Connecticut’s choice to lead the country did not win the popular vote, Connecticut has to give its electoral votes to the other guy, electing him (or her).

I expect that would not sit well with many of those who voted in the majority, only to see their choice ignored.

How could such a thing happen? The National Popular Vote legislation requires states that sign on to this movement to agree to cast their electoral votes for the top voter getter nationally, regardless of who that state’s voters want to be president.

So far, nine states and the District of Columbia have passed National Popular Vote laws. They control 136 electoral votes, which means the movement is about halfway to its goal of locking up 270 electoral votes in this manner, the majority necessary to elect someone president.

So, the Electoral College system would remain in place, but all these states would basically be committed to ignoring it and making sure the popular vote, not the Electoral College system, determined the winner.

I can see that plan unraveling quickly. One can easily imagine the anger of voters in, for argument’s sake, a traditional Republican state seeing the state’s electoral votes handed to the Democrat – against the will of the people in that state – giving the Democrat enough electoral votes to win. Can anyone doubt the legislature would be pressed to return in special session, abandon the National Popular Vote compact, and back the voters’ choice?

A rush for the exits would soon follow among other states.

The National Popular Vote organization calls this impossible because the compact prohibits it. They also say such a reversal would face insurmountable constitutional hurdles. That’s debatable. At the very least, it invites potential chaos.

Motivating the popular vote movement are arguments of fairness and making more states relevant. The fairness issue is basic – the candidate with the most votes should win. A strong argument can be made for that position, but is this version of “fair” what is best for the country and Connecticut? The current system has arguably served the nation well for 207 years.

Less credible is the contention that election by popular vote will force the candidates to pay attention to more states. Currently, candidates focus their resources and their visits largely on contested states, particularly those with big electoral numbers. Yet, I suspect, even with a popular vote determining the winner, candidates would continue to give their attention to big population centers, which does not describe Connecticut or its cities.

In any event, if the people of the United States want to abandon the Electoral College system in favor of the direct election of the president, they should do so by formally amending the U.S. Constitution. It may be difficult, but it is preferable to the controversy the National Popular Vote arrangement invites.

Paul Choiniere is editorial page editor.

We agree that “Less credible is the contention that election by popular vote will force the candidates to pay attention to more states”. We would go further and suggest this may well be a reason most of the media supports the Compact. Most voters want more information on candidates, yet more attack ads, more robo calls, and the same soundbites delivered here will do little to quench that thrust.

3rd Harford Elected Registrar maybe eliminated by Council

So may the 1st and 2nd Elected Registrars

The Courant has long been opposed to a third and even a second Registrar in Hartford. The Editorial Board would rather see the Council appoint registrars, as authorized in a Charter Revision last year. As we explained at that time, contrary to claims by the Courant, the proposal could result in an untrained, unqualified, politically appointed registrar or registrars.

Now we learn that the Courant and supporters now believe they got something else wrong in explaining the Charter Revision to the public.

So may the 1st and 2nd Elected Registrars

The Courant combines news reporting with an Editorial: Third Registrar Is Nonessential Employee -Third Registrar Is Nonessential Employee <read>

The Courant has long been opposed to a third and even a second Registrar in Hartford. The Editorial Board would rather see the Council appoint registrars, as authorized in a Charter Revision last yea r. As we explained at that time, contrary to claims by the Courant, the proposal could result in an untrained, unqualified, politically appointed registrar or registrars.

Now we learn that the Courant and supporters now believe they got something else wrong in explaining the Charter Revision to the public, when they said:

The charter changes would allow the city to appoint “one or more” trained, nonpartisan registrars. It would require a small change in state law, which hopefully will be forthcoming.

Now we have the news in the Editorial:

Now another twist. It was widely thought that the city would need a change in state law to
implement the charter change. But this month lawyers for the council and the secretary of the
state reviewed the statutes and think the change can be implemented without a statutory change — in other words, a town can choose to appoint rather than elect registrars, as municipalities can do with town clerks. “The statutes seem to contemplate that registrars can be appointed, if allowed by charter,” said Av Harris, spokesman for the secretary of the state.

But not necessarily only one, if they are now correct the partisan Council may need to appoint at least two:

But there caveats. While not crystal clear, the laws appear to require that there be two registrars
and that they be from the major parties. This is because the registrars have duties involving party primaries, and only the major parties are allowed to hold primaries. Nonetheless, if you are paying the bills, two registrars is better than three.

Speed Up Election Results – Not so fast, with another half-baked solution

UPDATED, With two additional views. And a CORRECTION.
We half agree with the Courant and the Secretary of the State. We have supported the idea, applauded the start that the Secretary took, yet there are problems with the system as proposed, and even more problems with the some of the views and ideas in the Courant’s Editorial. Yet, one half-baked manual system does not deserve a half-baked automated one to solve the problems.

We would like to see the Secretary and the Courant Editorial Board close a polling place and get the data in via smart phone, or close absentee ballots and report via laptop. We will help time them and transparently provide the video on YouTube.

We also remind readers that the Courant is one of the newspapers that led the fight to require expensive paper legal notices instead of allowing for web based notices.

The Courant published an Editorial in today’s print edition: Speed Up Election Results , the online version dated yesterday is titled Half-Baked Reporting System Keeps Election Results A Mystery <read>

We half agree with the Courant and the Secretary of the State. We have supported the idea, applauded the start that the Secretary took, yet there are problems with the system as proposed, and even more problems with the some of the views and ideas in the Courant’s Editorial.

Yet, one half-baked manual system does not deserve a half-baked automated one to solve the problems.

We also remind readers that the Courant is one of the newspapers that led the fight to require expensive paper legal notices instead of allowing for web based notices.

We see several problems with the Courant’s expectations and the system tested last year, presumably the same system tested this year, since the Secretary of the State’s web site hosts the same training as before:

The system expects every single Moderator and Head Moderator to input all the results on election night, that is about eight hundred individuals, many in their 70’s and beyond, expected to start election day at the polls at 5:00am, working the polls until 8:00pm and then work to close the polls, close the voting machines, count some ballots by hand, secure ballots, secure materials, and report results.
In a simple Municipal Election in Glastonbury, with 26 candidates, polling place Moderators have approximately 78 numbers to input, while the Absentee Ballot Moderator would have 12 times that number to input or 936.
In Bridgeport those numbers would be 19 candidates, 57 numbers for each polling place, and 2850 for the Absentee Moderator.
In Greenwich it varies by district, in District 1 it is 55 candidates and 165 numbers for the polling place, perhaps 3600 for the Absentee Moderator.
I would challenge the Courant Editorial Board to work for 15 straight hours, service the public, managing a team of individuals that work one day a year, and within one-half hour input those numbers.
To add to the challenge, the Secretary offers input via smart phone, so the entry can be performed at the polling place, if it has cell service, saving the drive to Town Hall. By the way, Moderators keep a log of incidents during the day and that has to be typed in as well.
If you are a central count Absentee Ballot Moderator you do not have that log to put in, you are at Town Hall, so you can surely use a laptop computer, you started the day a bit later, but in addition to entering a few hundred or thousand numbers, you have to print the optical scanner tape, which can easily use up more than the whole half-hour the Courant Editorial Board expects so that they can get the results on their schedule.
That is up to thirty-four individuals in a town all doing that at the same time, hopefully few would have problems with their passwords or need other help from the town or state.

Once again, we are in favor of a fully baked solution:

Allow towns to hire competent data entry help, to arrive fresh at town hall at 8:00pm, and type in the data under the guidance and supervision of the Moderators and Head Moderator. (In one medimu -sized town where I have worked, as Absentee Moderator – I read the numbers, the Head Moderator typed them into a spreadsheet, a Registrar watched him to check his input, he printed the data and the Registrar and I checked it against my hand written records and the machine tapes – we always found a couple of things to correct in the process)
Forget the smart phones, just too slow for this much data.
Test the system in real life and set reasonable expectations for timing. Most towns should be able to get the data in by Midnight, but sometimes there will be good reasons for delays.
The goal should be reasonably accurate data the first time. That means double checking entry. Double checking any transcription and manual addition required (Try as we might, it is not possible to machine count write-ins and other special situations that require hand counting of ballots)

Here is the Editorial with our annotations in []

By Wednesday afternoon, official results for all of Tuesday’s local elections were still not up on the secretary of the state’s website.

This is crazy. [Perhaps, but lets consider what we would say about this editorial after reviewing it]

By contrast, Florida voters knew by 8 p.m. Tuesday all their local election results — because Florida state law says officials have to report them to the state a half-hour after polls close and update them every 45 minutes thereafter. [Florida has improved after 2000, but Connecticut has wisely opted not to have its electronic voting machines connected to phone lines or the Internet. Pretty much impossible to get the job done this quickly without an electronic connection from each scanner to some central location]

Also, Florida has early voting, and officials are required by law to count those votes and absentee ballots ahead of time so that they’re ready for posting as soon as polls close. [Counting Absentee Ballots ahead of time is not such a great idea, since it opens the same issues as reporting Presidential elections from East to West, in this case with days of advanced notice. We are in favor of early voting, yet it would be very expensive in Connecticut with our town by town election management. This is not a simple, nor an inexpensive change. Wisely Connecticut does not allow reporting of any absentee results until 8:00pm. We do not allow counting to start before 10:00am on election day, and we get it done on election day.]
[Consider other states, like California, which counts absentee ballots for weeks after elections. Somehow their voters and media have survived]

But residents of Hartford, West Hartford, Windsor, Waterbury, Tolland and a bunch of other towns and cities that rely on the secretary of the state’s website for election results couldn’t get them the day after the election. [We agree that is too long. It is not the reporting system. The current system, with all its faults works for most of the towns, much faster than that. Something else must be delaying those results. Sometimes it is better to get the right results than pressure overwhelmed officials (see Bridgeport 2010)]

In some towns, voters won’t know Thursday, either. [Once again, they should ask their local officials for an explanation. Apparently the Courant has not considered reporting on the actual reason for such delays, instead assuming its the reporting system]

Instead, curious townsfolk who clicked on those municipalities on the secretary of the state’s website (www.ct.gov/sots) saw the message “Check back later for ‘Official Elections Results’ as submitted by the town.”

This is maddening.

What’s Up With This?

In the digital age, election results should be made public very quickly, and in many states they are. But here in the Land of Steady Habits, we’re still reporting results in some places the way we’ve done it for decades. Our breakthrough technology is the fax machine. [Actually towns can also use email, now that the law requires towns to provide email to all registrars]

Some towns make up their own reporting forms rather than use the state’s, and have state troopers, who usually have better things to do, drive their results to Hartford. [As allowed by antiquated state law. We also note that forms do have to be customized for each town, and sometimes for each district, since there are different offices and numbers of candidates on the ballot]

As a consequence, Connecticut can’t get reliable results from some towns on election night, or even the next day. This drives the media nuts, of course, but more important, it’s a disservice to the public. Voters would like to see the official tally on who won and by how much. Is that too much to ask? [Once again the current system may delay results a few hours, but not even a day. I am sure most state troopers could get to Hartford in less than a couple hours]

Secretary of the State Denise Merrill shares our pain, agrees that the present system leaves much to be desired and believes she has a solution. She said her office has been developing a software program over the past 30 months that allows instantaneous reporting of election results — “just type in the numbers and hit send.” She said 40 towns field-tested the program in Tuesday’s election. It worked well in half of them and had some bugs that need to be worked out in the others. She hopes to have it in place in all towns by the 2014 elections. [We would like to see the Secretary and the Courant Editorial Board close a polling place, get the data in via smart phone, or close absentee ballots and report via laptop. We will help time them and transparently get it up on YouTube]

Ms. Merrill’s office made more results available more quickly this year by scanning the paper forms that were faxed in and posting them on the website. Some of these were hand-written with cross-outs (see Waterford, for example), making them barely legible — more evidence that the present system is hopelessly antiquated.

Yes, Florida Does Voting Right

Ms. Merrill would do a great public service by proposing a law similar to Florida’s, requiring quick posting of at least preliminary election results. At present towns have until 6 p.m. the following day to get their results in, and many don’t make even that expansive deadline. The chance for error is magnified as numbers are transcribed once or twice, added up, faxed in and typed into the state system. The new software does the addition and requires only one input, reducing the chance for error.

Some election processes are hard to change; some local officials like things as they are. Ms. Merrill should push ahead and drag the state into at least the 20th century. [The Courant and Ms. Merrill should set reasonable expectations of the system and election officials, based on the results of tests, and then change the law, negotiating with election officials]

To understand more details, you can listen to the training and/or view a PowerPoint presentation on the Secretary of the State’s election reporting system at her web site: <view/listen>

UPDATE: 11/10/2013. A column and op-ed in the Courant today:

First, an op-ed by Karen Cortes, a conscientious(*) registrar from Simsbury, Antiquated Systems Stall Election Results <read>

She mostly echoes our concerns, yet there are several areas where we diverge:

I do not agree that automation provides a total solution and that getting results immediately is desirable. No matter how well the collection system is automated, there is an need for checking and rechecking at the origin of hand count and write-in results in particular, time should be taken to make sure the check-in list counts match the total ballots counted. Electronic data entry checked well can save a few hours and some redundant work for the Secretary of the State as well.
Electronic data transmission from our election machines is risky and not a cure for errors. Connecticut wisely does not connect our machines to the Internet or phone system, to protect against viruses and attack.
Blindly submitting electronic results, bypasses the careful checking that the scanner was used properly and did not miss votes or double count them – that has happened in Connecticut, even over turning an election.
I would not hold out NJ as a good example. NJ uses DRE (touch screen) voting which in NJ are total uninhabitable and proven to miscount.
As for Virginia, this year they are a poster-state for blindly accepting machine results. The results from one county were blindly reported, were obviously incorrect, and may, if corrected, result in the change in a result. See <BradBlog> As for me, I will opt for taking the time for accurate date entry, and checking for reasonably accurate results in initial reports.

Also a column from Jon Lender, with views closer to our own: Computerized Vote-Tally System Tested: Merrill Gives It C+, But Local Registrar Says It Flunked <read>

* CORRECTION: Good grief! An earlier version said ‘contentious’. We regret the error. At least we have proven the need for checking and rechecking.

Op-Ed: Internet Voting Security; Wishful Thinking Doesn’t Make It True

This was a simple online poll that was easily compromised. Internet voting vendor software will be harder to compromise, but this shows that computer security is hard and claims must be proved. Before we entrust critical public functions such as voting to such software, the public deserves a solid demonstration that such claims are truly substantiated, and policy makers need to be schooled in a proper skepticism about computer security. That has not yet happened.

“Internet Voting Security; Wishful Thinking Doesn’t Make It True”

Duncan Buell

On March 21, in the midst of Kentucky’s deliberation over allowing votes to be cast over the Internet, its daily poll asked its readers, “Should overseas military personnel be allowed to vote via the Internet?” This happened the day before their editorial rightly argued against Internet voting at this time.

One of the multiple choice answers was “Yes, it can be made just as secure as any balloting system.” This brings up the old adage, “we are all entitled to our own opinions, but we are not entitled to our own facts.” The simple fact is that Internet voting is possible – but it is definitely NOT as secure as some other balloting systems. This is not a matter of opinion, but a matter of fact. Votes cast over the Internet are easily subject to corruption in a number of different ways.

To illustrate this point, two colleagues of mine wrote simple software scripts that allowed us to vote multiple times in the paper’s opinion poll. We could have done this with repeated mouse clicks on the website, but the scripts allowed us to do it automatically, and by night’s end we had voted 60,000 times. The poll vendor’s website claims that it blocks repeated voting, but that claim is clearly not entirely true. We did not break in to change the totals. We did not breach the security of the Courier-Journal’s computers. We simply used programs instead of mouse clicks to vote on the poll website itself.

Some policy makers are wishing that the net were secure and the security promises of vendors were true, and they are not listening to the computer experts who know otherwise. Why shouldn’t we entrust computer voting security to government and its vendors? Ask that of South Carolina taxpayers; hackers have shipped overseas all tax records and identifying information from 1998 to 2012. Wishful thinking is dangerous when it causes us to fail to protect our best interests; we must defend our data just as we defend our shores.

This was a simple online poll that was easily compromised. Internet voting vendor software will be harder to compromise, but this shows that computer security is hard and claims must be proved. Before we entrust critical public functions such as voting to such software, the public deserves a solid demonstration that such claims are truly substantiated, and policy makers need to be schooled in a proper skepticism about computer security. That has not yet happened.

There is an irony in hacking an online poll about whether voting can be hacked. But it points to a much-needed dialogue between policy makers and computer security experts. Elections are too important to be entrusted, without proof, to the marketing hype of an Internet voting company. The nation’s real elections should be decided by the voters in the nation’s jurisdictions, not by whichever entity – foreign or domestic – happens to have the best software bots running on any given biennial Tuesday in November.

As Professor Buell points out “Internet voting vendor software will be harder to compromise, but this shows that computer security is hard and claims must be proved.”. That has been tested once, in Washington, D.C. and the result was exposure of a clearly insufficient Internet voting system.

For now we await vendors willing to subject their systems to ongoing rigorous professional and open public adversarial testing. We admit it will take a lot to satisfy us that systems are sufficiently secure from outsiders and insiders. But it seems vendors are hardly willing try.

PS: Not so long ago another newspaper’s poll was compromised, by parties and methods not disclosed, will little lessons learned by the newspaper.

[Why NOT] Let Overseas Military Fax Votes Home ?

Connecticut does need to improve the voting process for military voters — but Internet voting is not the answer.

Every day, headlines reveal just how vulnerable and insecure any online network really is, and how sophisticated, tenacious and skilled today’s attackers are. Just last week, we learned that the U.S. has already experienced our first-ever documented attack on an election system, when a grand jury report revealed that someone hacked into the Miami-Dade primary elections system in August 2012.

Earlier in March, the Courant ran an editorial in favor of online, email, and fax early voting: Let Overseas Military Fax Votes Home <read>

We have testified twice this year against this risky, expensive, unconstitutional concept, providing information on safer, economical alternatives <here> <here>

Today, to its credit, the Courant published an op-ed by Pam Smith of VerifiedVoting. org, articulating Why Not: Internet Voting Puts Election Security At Risk <read>

Connecticut lawmakers are considering legislation to allow military voters to cast ballots over the Internet. The intention of this legislation is well-meaning — Connecticut does need to improve the voting process for military voters — but Internet voting is not the answer.

Every day, headlines reveal just how vulnerable and insecure any online network really is, and how sophisticated, tenacious and skilled today’s attackers are. Just last week, we learned that the U.S. has already experienced our first-ever documented attack on an election system, when a grand jury report revealed that someone hacked into the Miami-Dade primary elections system in August 2012.

A chilling account in The Washington Post recently reported that most government entities in Washington, including congressional offices, federal agencies, government contractors, embassies, news organizations, think tanks and law firms, have been penetrated by Chinese hackers.

They join a long list that includes the CIA, FBI, Department of Defense, Bank of America, and on and on. These organizations have huge cybersecurity budgets and the most robust security tools available, and they have been unable to prevent hacking. Contrary to popular belief, online voting systems would not be any more secure.

Not surprisingly, a senior cybersecurity official with the Department of Homeland Security warned election officials last year that online voting is premature and not advisable at this time. The National Institute of Standards and Technology (the federal body tasked with researching Internet voting) issued a statement shortly after, warning that secure Internet voting is not feasible with the tools currently available. Because the agency determined that Internet voting cannot be done securely, it has not developed testing or certification standards for systems.

So why are state lawmakers considering online voting for the military?

First, there is a mistaken perception that because we can shop and bank online, we should be able to vote online securely. But shopping or banking online are far from secure. Banks and online merchants lose billions every year to online fraud. They factor this into the cost of doing business.

There is, however, no acceptable level of vote fraud or manipulation. Moreover, elections have unique properties that are unlike banking or e-commerce. In a financial transaction, both parties can check each online transaction by reviewing a statement or receipt. But we vote by a secret ballot. Neither the voter nor the election official can verify that a ballot has been received the same way it was sent. This makes online voting especially susceptible to undetected hacking.

Second, we have seen a big push for Internet voting (including via email and digital fax) because the vendors of online voting systems have targeted state lawmakers and election officials with aggressive marketing and sales campaigns. The vendors have made extraordinary claims of security and auditability — all of which are unsubstantiated by any publicly reviewable research or documentation. None of these systems are subject to any standardized security testing or certification and claims of security are backed only by the vendors’ words.

There are things we can do to improve the voting process for our military voters without risking the integrity of their ballots or the security of our elections.

We can:

1. Move registration deadlines closer to the election. Virginia did this in 2012 and it paid big dividends, allowing service members to receive and return their ballots up to Election Day.

2. Allow our troops to use the Federal Write-In Absentee Ballot to register and vote. Many states now allow service members to do both. It’s the military voter equivalent of Election Day Voter Registration. This widely used, practical reform would make absentee voting much easier for service members stationed outside of Connecticut.

3. Count military ballots postmarked by Election Day, and received seven days after. This would still give election officials enough time to count the ballots before certification, and would give our troops an opportunity to vote on Election Day.

Our legislators are right to act to improve voting for our brave men and women in uniform, but online voting is not yet the answer. Instead, the General Assembly should look to make the voting process easier and more accessible with some simple, common-sense improvements.

Others Weigh In On Election Reforms

Last week Secretary of the State Denise Merrill weighed in at her press conference on, as the Courant headlines, how “Reforms Could Boost Voter Participation”. Also weighing in was Melissa J. Russell, President of ROVAC (Registrars Of Voters Association, Connecticut).

Last week Secretary of the State Denise Merrill weighed in at her press conference on as the Courant headlines, how Reforms Could Boost Voter Participation <read> <press release>.

Merrill noted that four of the states with better voter turnout than Connecticut last month — Minnesota, Wisconsin, New Hampshire and Maine — all allow voters to register on election day. Beginning in Nov., 2013, Connecticut voters will be able to do so as well.

Early voting, voting by mail and no-excuse absentee voting would also boost turnout as well as reducing wait times at polling places on election day, Merrill said.

By our references Connecticut ranked 16th and 17th in voter turnout in 2008 and 2010. This year we ranked 7th – much more than respectable in gains and in absolute rank, especially considering Hurricane Sandy! In fact, Connecticut ranked ahead of many states with early voting and long lines such as Ohio and Florida, not to mention 7 of eleven states with Election Day Registration. We reiterate several themes we have discussed previously:

According to statistical analysis and a summary by Doug Chapin at the Secretary’s Election Performance Task Force, which she attended: Early voting, by mail or in-person tends to decrease turnout. In other states, Election Day Registration has increased turnout and mitigated the negative effects of early voting on turnout.
Early voting in person can increase convenience, but would be costly to implement, especailly in Connecticut with our independent 169 town voting system. Seven days of early voting would increase election day costs for such towns by close to a factor of 7.
Early voting in person can have high integrity, if we are willing to pay for extra security when elections are closed each day.
There are better solutions than early voting for long lines on election day. Our scanners can handle thousands of ballots, yet problems this year were caused by a shortage of checkin lines. Each additional line requires one extra official and perhaps a few more privacy booths – much more economical than in person early voting. And it works! Early voting is no panacea – see FL and OH – and during the last two elections storms hit during the likely early voting period, lessening its probable value.
Increase mail in voting increases the risk of documented fraud. Mail-in or absentee voting seems to be the source of the easiest and most frequent fraud in U.S. Elections.
We support Election Day Registration, but point out that the method chosen for Connecticut is not like the successful methods of other states which increased turnout, and is likely to lead to long lines.

Also weighing in was Melissa J. Russell, President of ROVAC (Registrars Of Voters Association, Connecticut), where she “looks forward to working closely with the Board and the Committee chairs to improve the professionalization of ROVAC” Op-Ed: Registrars are continually working to improve elections system <read>

We would hope that the ROVAC President, Board, and Committee chairs will not only strive to become more professional, but also to work cooperatively with the Secretary of the State, her office, legislatures, and advocates in improving our election system.

The word “archaic” has been bandied about, along with censure of the long lines voters faced in West Hartford, Hartford and Manchester. In many of these stories, the registrars of voters across the state are taking the brunt of this criticism.

It is unfortunate that the “other side” of the story is not being told. In the vast majority of polling places in the vast majority of towns, the election went smoothly, with fast moving lines, cheerful service to the voters on the part of the poll workers, and swift reporting of results at the end of the night. Large cities such as Bridgeport, Stamford and Norwalk handled the large turnout of voters with no major problems.

Here we agree that by and large most voters, in most towns, voters were well served. Registrars deserve credit along with courteous dedicated poll workers and the Secretary of the State. We can be proud of selecting economical, paper based optical scanners for our elections, and the push from the Secretary to purchase enough ballots to avoid another Bridgeport. Perhaps the shortage of checkin lines and enough staffing for Election Day Registration will be solved to avoid repeated problems of polling place lines, and now problems similar to those experienced for Presidential balloting.

We applaud ROVAC’s endorsement of electronic checkin also championed by many beyond ROVAC. We wonder if they will endorse and lobby their municipalities to pay for it along with the associated internet connectivity required to reach its full potential. We have often heard registrars complain of tight and reduced budges, unable to cover small obviously needed supplies, such as laptops to calculate audit results, a far cry from the requirements of automated checkin.

It is also unfortunate that the “archaic” system is being blamed on the registrars and their poll workers. The Registrars of Voters Association of Connecticut (ROVAC) educates registrars all over the state, both at statewide conferences held twice a year, and in county meetings held three to four times a year. Consider some of the improvements that registrars are currently using and developing in their towns: the use of Skype to communicate with various polling places in the cities, leaving phone lines open for the public; the use and development of electronic check-in books, where a voter is checked in not on a paper list of many pages, but with several clicks of a mouse on a laptop, or the scanning of a barcode next to a voter’s name; and a post-election audit system that uses high-speed scanners rather than teams of people hand inspecting and counting ballots.

ROVAC also supports the use of technology that is already available for our current tabulator system to report results: using either the ports on the back of our tabulators to send results to the secretary of the state’s office instantly, or placing the memory cards that each tabulator houses into an “ender machine” that will read the information on the card and send it electronically to the secretary of the state.

Such use of this technology, which is used in other states without problems, would speed up the reporting of the election results tremendously, while virtually eliminating the mistakes that come from bleary-eyed election workers attempting to read and accurately record numbers by hand onto a Head Moderator’s Return, which then gets faxed to the secretary of the state’s office.

Here we have a couple problems. First, opening the ports is one of the more dangerous options for our election equipment, endorsed by UConn and eliminated by the previous Secretary of the State. Moving the card to an ender machine, would still expose the card and data to security risks. And this will not do the job. Those bleary eyed workers still need to submit the additional information by hand, such as hand counted regular ballots, write-in votes, and various special ballots that cannot be scanned.

We do support auditing by appropriate independent machines, with proven processes that are combined with transparent procedures that can verify the results to the public. We are concerned that instead, we will end up replacing one black-box with another black-box system that is insufficient, and completely voids the potential value at considerable cost. We would hope that ROVAC, the Secretary of the State, CTVotersCount, and other advocates get the opportunity to work together to propose legislation, test, and implement a system that provides better audits, while finding ways to pay for the valuable provided.

USA Today: Electronic voting – The Real Threat

A USAToday Editorial, and an opposing view: Editorial: Electronic voting is the real threat to elections <read>

Imagine how easy voting would be if Americans could cast ballots the same way they buy songs from iTunes or punch in a PIN code to check out at the grocery store: You could click on a candidate from a home computer or use a touch screen device at the local polling place.

It’s not entirely a fantasy. In many states, some voters can already do both. The process is seductively simple, but it’s also shockingly vulnerable to problems from software failure to malicious hacking. While state lawmakers burn enormous energy in a partisan fight over in-person vote fraud, which is virtually nonexistent, they’re largely ignoring far likelier ways votes can be lost, stolen or changed.

How? Sometimes, technology or the humans running it simply fail

The article goes on to list some of the failures of electronic voting, many of which have been covered at CTVotersCount.

Fortunately our Legislature has not wasted time on raising Connecticut’s adequate voter I.D. law to the level of voter suppression. Unfortunately, the Legislature has continued to ignore science, experience, and the Constitutional requirement for preserving the secret vote – this spring they voted for email voting, in a provision inserted in a unrelated bill, without public hearings, Governor vetoes bill with email/fax voting “rat”

The opposing view from Bob Carey, Opposing view: Paper voting system is broken <read>

Carey is not a scientist. For years he has been advocating for more effective military voting, he knows much progress has been made to make paper voting effective for military and overseas voters, and that the military has been negligent in following the law to provide an officer to assist military votes at each base. He should also know how vulnerable even the defense department networks remain. And local control of elections requires using the regular internet for voting, managed by officials, largely no more technical than Mr. Carey.