Let us act deliberatly to actually improve elections

We are amazed by the number of election integrity issues raised by this election and the flurry of suggestions for improvement, led by the Hartford Courant.  Yet in all the excitement and rush to judgement and improvement, among the good intentions and good ideas, there is also a misunderstanding of the system, ideas that are not feasible, uninformed, and that would make a worse system.

In fact, its too early to rush to judgement with knee-jerk reactions, without deliberate, complete plans for reform.  Yet, we have attempted to counter errors and support good directions in letters to the editor and blog comments. Here we archive some of those articles, editorials, and some of our (edited) comments.

On Sunday the Courant had an article, an investigative report, an opinion piece, and an editorial mostly focused on Hartford’s mess-up, and the allegation that our votes are reported slower than the rest of the nation. The article and the investigative report are a true service, the best of factual journalism. Not so much the editorial and opinion piece<read>

Our response was a letter to the editor of the Courant, addressing the editorial and opinion piece, withing the limits of the 200 words allowed by the Courant, now published online:

Vote Count Accuracy More Important Than Speed

I agree in part with the editorial “Where Were The Results?” [Nov. 9]. Our current system of local registrars is antiquated. We should do for elections what we have done for probate: regionalize, professionalize and economize. Yet, change should not include blind pursuit of speed over accuracy or risk tampering with elections.

Beyond tampering with results, connecting scanners or memory cards to the Internet risks that the scanners can be infected to compromise future elections. Recognized computer scientists and security experts agree. Based on UConn’s recommendations, the external ports on our voting machines are required to be sealed.

Contrary to the opinion piece by Brandon Finnigan, “Learning Who Won Takes Too Long” [Nov. 9, Opinion], some states are more organized and careful than Connecticut in reporting reasonably complete results. Los Angeles County is the nation’s largest election jurisdiction, managed by a professional election administrator, Dean Logan. California has about 50 percent its votes cast by mail. As the Los Angeles County website states, mail ballots received by Election Day and some others are counted over a 28-day period after election night.

Let’s regionalize. Let’s improve the system. And let’s lighten up on getting results, any results, without regard to their accuracy.

Luther Weeks, Glastonbury

The writer is executive director of CTVotersCount.org, an election issues advocacy site.
Copyright © 2014, Hartford Courant

What follows are some of my comments online on those articles and opinions, edited for grammar and completeness.

Two of the articles complained that results were too slow, including an “Expert” from Southern California who wrote the opinion piece. I said:

California counts absentee ballots for up to 28 days after the election. In 2008 Minnesota took at least a couple of months to determine the winner of the Frankin-Coleman race for the Senate. Connecticut would rush to complete our recanvass in eight days…

Connecticut accepts no absentee votes after 8:00pm on election night. In CA, where about half the vote is mail-in they continue counting them for days and weeks after the election. They are not required to be done for 28 days – I know that LA county, the largest jurisdiction in the U.S. goes quite a while. The only thing we count later (if we do) is the provisional ballots.

Connecticut  is far from the slowest. Take Alaska’s election counting, please:

Alaska will begin counting more than 53,000 absentee and questioned ballots on Tuesday[Nov 11] in an effort to resolve the state’s unsettled contests for the Senate and for governor. Democratic Sen. Mark Begich trailed Republican challenger Dan Sullivan by about 8,100 votes after Election Night…The race for Alaska governor is actually closer than the Senate contest. Independent candidate Bill Walker, aided when the winner of the Democratic primary bowed out of the race to run as Walker’s lieutenant governor, led incumbent Republican Gov. Sean Parnell by about 3,000 votes.

So, Connecticut is “among the last in the nation to get election results”, yet days, maybe weeks ahead of Alaska, California, and Colorado:

It’s a week after Election Day and they’re still counting votes in Colorado, where some are blaming a new state law that replaced polling booths with mandatory mail-in ballots. Top-ticket races have been decided—Democratic Gov. John Hickenlooper was re-elected and Republican Cory Gardner unseated Democratic Sen. Mark Udall—but the vote totals in a dozen state House and Senate races remain unknown

To the suggestion that we should connect our scanners to the Internet because Massachusettts has found no problems, I said:

Massachusetts has never had an election audit until this year. So there is no evidence of the reliability of their processes. In fact, half the states have no audit at all and there is very little to go on to know if there ever was or will be errors or fraud in those states. It could always be that paper ballots and audits actually deter fraud and reduce error. NY did find errors a couple of years ago in their audit, as we have in Connecticut. Respected computer scientists like Professor Shvartsman at Uconn agree there are risks, that is why he advised the state to seal the ports on our scanners. Finally, even though we counted all the ballots in Bridgeport in 2010, the system was never able to recognize the actual results there or to investigate the difference between voters signed in and the number of ballots there.

At CTNewsJunkie: Problems at the Polls Highlight Limitations of Locally Operated Election System <read>.

The main point  was that the the new election night reporting system by the Secretary’s office will solve many of the problems – it may solve some, but certainly not all.  I commented:

I hope the new reporting system works and is workable. The previous system prototyped twice required too much data entry by each moderator, many after a 17-24 hour day. An improved system will get the data earlier and more accurately, provided time is still taken to double check it in each town before entry and if the system allows the efficient entry of data by refreshed people and their entry is also double checked.

Anyone who pronounces such a system as ready for general use by moderator’s should be required to use it to enter central absentee results after a long long day.

There is much to improve in the whole system especially by regionalization. Yet, there is much that could be done in the name of speed and modernization that could make things worse rather than better.

Human nature being what it is, we tend to believe that any change we are in favor or against will or would have prevented this problem. e.g. see “911 Patriot Act”.  Several commenters indicated that if we had kept lever machines or went to touch screens, we would not have all this problems counting absentee votes. I commented:

As an experienced central count absentee moderator I would like to point out a few things:
1) Our absentee procedures are no more complex than other states. Actually simpler than those that require that signatures be matched with those on file.
2) When we had lever machines, and if we changed polling place technology, we would still need paper absentee ballots.
3) The only difference with absentee ballots now is that we have scanners to help count them faster and more accurately. Much more quickly than by hand counting.
4) Absentee ballot counting can start at 10:00am, when 95+% are available. In a well organized operation, all there is to do at 8:00pm is to process a few that came in at the end of the day, print machine tapes, and complete paperwork. That should not take more than a couple of hours.
5) You can pretty well predict the number of absentee ballots based on those that are requested, so you can staff accordingly to get the job done.
Finally, there are good reasons we have voter verified paper ballots in polling places as well as for absentees- they provide a much more secure and auditable vote, over all its a much less costly technology, its less likely to cause long lines (check-in is a separate issue), and paper ballots can be used despite machine and power failures etc.

Ironically, some who complain about the results of the knee-jerk, partially helpful, Help America Vote Act, also propose knee-jerk action this time.

Bottom Line for now:

  • There were many problems highlighted by this election
  • There is a lot to fix, things that voters should not put up with
  • But like some hurricanes, we missed the big one – by Wed we knew who the winners were, a few voters and votes were lost, and the media had a field day complaining about their not getting results fast enough.
  • Deliberate action based on a bit of experience, facts, and research can lead to positive improvement. Lets do that, not forget the problems and not forget do the work.
Facebooktwittergoogle_plusredditpinterestlinkedintumblrmailFacebooktwittergoogle_plusredditpinterestlinkedintumblrmail

Leave a Reply

You must be logged in to post a comment.