-Aspen's historic May 5, 2009 IRV election audited as single ballots- 5/5/09 Aspen CO held an instant runoff election (IRV) for mayor and 2 council members. Interpreted contents of each ballot, scanned by True Ballot, were publicly released. Open records requests for a CD of image scans were denied. Aspen has been sued to protect records from destruction and to allow inspection of the scanned ballot files. A Court of Appeals ruling holds that unidentifiable ballots are public records.

Search this and related blogs

Friday, January 29, 2010

Latest Questions for City Attorney regarding elections and Marks v. Koch

[We are collecting questions and answers about Aspen's election and also public verifiability of elections in general on this site. Please contribute your own questions (anonymous questions ok). Also please participate in discussing answers to these questions (answer please under your own real name).  You may use the blog comments for this or contact Harvie Branscomb at 970-9631369.   Thanks]

New Questions asked by Harvie Branscomb based on developments since January 1st:

35: Question:  The Colorado Bar Association says it doesn't  address the condition that the Election Commission (EC) and the City are directly adverse to one another. Why wasn't this possibility considered in the query to the Ethics Committee?

36:  Question: The Ethics Committee replied that it doesn't know if the EC is a client of the City Attorney. Is the EC a client of the City Attorney?  If the City Attorney has no answer for this, how would one find out the answer to this question?

37: Question: If the EC is a client of the City Attorney is that representation subject to the consent of the Election Commission, as the CBA indicates?

38: Question: If the EC refuses the representation of the City Attorney, does it have a right to counsel, paid for by the City?

39: Question: If the EC is not a client of the City Attorney does it have a right to counsel?  Is it expected to pay independently for that counsel?

40: Question: Why would the City want to file a motion for a protective order to prevent Marilyn Marks from obtaining testimony from the True Ballot Inc.  by deposition?  What public interest could be served by blocking this communication?

41: Question: Why is the City Attorney complaining to Judge Boyd about the cost of defending the litigation against the City Clerk for inspection of ballot images when the City has taken steps which have caused an increase in the cost of litigation such as filing extensive briefs, when the cost  of litigation is a direct cost to Marilyn Marks but most of which is an indirect cost to the City covered by salary and overhead?

42:  Question: Why can't the city provide copies of the spoiled ballots which are classified as "other election records"?  Why can't the city provide information about the seals which protect the election records?

[ the Open Records Requests and City responses will be posted elsewhere on this site. H. Branscomb ]

---- previous questions are included below the break ----- (no answers to questions has been received from the city as of Jan 29, 2010)

1.What might a citizen risk by asking questions? From my perspective, two voluntary commissioners were recently dismissed following an unjust smear campaign that prevented their pursuit of election quality improvements. It appears to me that their principal error was to ask difficult questions.

2.It has been reported that the city knowingly skirted the law by holding a secret meeting to discuss spending marketing funds. Were election commissioners railroaded from their roles for a similar but unintentional technical infraction after questioning election impropriety?

3.There were procedural defects reported for the May 5 election, such as less-than-private early voting, keys left in ballot box, insufficient testing, last minute changes to software setup, non-random and incomplete auditing, failure to audit the tabulation in time, etc. Will there be a public discussion of these and other problems, which I believe some council members may already know about?

4.If the city believes that the ballots and digital images cannot legally be removed from the locked ballot box for inspection, wasn't it illegal to open the box on May 7 for a part of an audit?

5. If voted ballots are not allowed to be seen by the public why does Ordinance Sec. 2.26.050 require:  “At a minimum, the City Clerk shall arrange the counting of ballots so that the candidates and their representatives may observe the ballots as they are counted.”

6. What powers do you believe the Colorado law says the Election Commission has under the City Charter and state law? Colorado law appears to give power to the Election Commission for purposes of checks and balances. In your view, who or what entity provides election oversight in Aspen?

7. Should City Council members select two election commissioners while essentially employing the third, knowing their own re-election might depend on an Election Commission decision?

8. Are citizen members of the Election Commission expected to avoid representing their party's interests? Are candidates expected to refrain from advocating to commissioners?

9. If future election commissioners learn that election officials have made a mistake do they have any ability to overrule the officials? Could they become subject to another retroactive recall for questioning the city attorney's advice? What protects them from a smear campaign which renders them impotent if they disagree with City Council? 

10. Recognizing that irregularities occurred, and since the Election Commission asked, why wasn't the Election Commission allowed outside counsel?

11. Does the city inform its citizen volunteers regarding the rules regarding open records and open meetings? Were Kathryn Koch's meetings with Elizabeth Milias or with Chris Bryan (but not together) more acceptable than the few meetings Milias and Bryan had without Koch?

12. Are the recently applied standards for Election Commission meetings and records applied to the Housing Frontiers Board and the Financial Advisory Board? Do these groups announce meetings or take minutes? 

13. When were the two election commissioners dismissed, or did their term simply end in July? As a commission member by law, is Kathryn Koch now a quorum by herself and therefore perpetually in a public meeting subject to CORA?

 14. Is it possible for someone to overturn the election (not that I want to) considering that the deadline for contests is over?

[the following questions are contributed]

15. Why was the software configuration changed (presumably to the Cambridge method) ,the evening before the polls opened without notifying the press, candidates or public?

16. After the tabulation error was found by True Ballot, why wasn’t the public notified for 10 days (past the recount deadline)?

17. Was a post election public audit to verify the IRV tabulation (not just the interpretation of individual ballots) not performed? A full hand count to verify the IRV process was anticipated on several occasions by council members prior to the election.

18. The City stated in a formal press release that there was a “staff audit of the IRV process.” Did a “staff audit” of the IRV process take place? What did it consist of? Which contests were audited and to what extent?

19. Given that the City has stated that someone could use the ballot images to match the ballots contents with the poll book list and identify how voters voted, what does the City say about the ability to identity of identically sequenced ballot data strings?

20. The Charter calls for a majority of votes cast to win a council seat, but neither Torre nor D. Johnson achieved a majority of voters’ votes. Does the language in the Charter actually mean that a majority of voters voting need not have voted for the winner to satisfy the law?

21. Why was the public told that a lower ranking on IRV cannot hurt a higher ranked candidate’s chances of winning, when the Aspen method does create this problem for the City Council contest?

22. Given that the Diebold machines were used to officially count the Art Museum vote, why did the City tell us that the Diebold machines were not used for official counts?

23. Why was the early voting ballot box in City Hall not always locked ?

24. Does the law allow early voting for City elections? If not, why were the absentee voting procedures not followed for the 803 “early voters?”

25. Q. With respect to the IRV ballots, what definitions does the city use for each of the following terms:


26. Q. With respect to the definitions, above, where in the law is each definition specified?

[ Only the city can answer these questions. H. Branscomb]


27. How many ballots were accepted by the AccuVote machines but rejected for any IRV round due to improper marking?
[Accuvote devices were programmed to reject over votes only on the first rank of the mayor contest and first and second rank of the city council contest.   I believe that 51 ballots were both accepted by AccuVote and subsequently may have been rejected by TrueBallot during some round.  8 max for mayor and 43 max for CC.  It is difficult to verify these numbers as some official documents are not internally consistent. In some cases the ballot errors may not have been reached in any IRV round.  Many documents which were available on the city web site have been removed in a massive removal of open records which occurred when the city web site was upgraded after the election. Much previous public access to records has now  become much more difficult.   H. Branscomb]

28. Why do Aspen rules require #1 and #2 choices to be ranked to be eligible for IRV counting, when common IRV practice is to ignore gaps. For example, if only #3 and #5 are ranked these are usually regarded as #1 and #2 choices.
[One of the Aspen rules allows for the highest next rank to be considered, except in case of the first round of city council tabulation which is a special case. The first round of city council tabulation was *not* conducted as “instant runoff voting”.  In that special case duplicate voting for the same candidate in first and second rank led to the second rank for the same candidate being eliminated.  This type of detail demonstrates the complexity of the actual tabulation. H. Branscomb]

29. Is there a legal challenge process for the voters who did not find their string in the published list of ballot strings?
[This is obviously a question for the city, but here is my attempt to answer: voters who have noted they did not find their strings have thus far been ignored (as one might expect).  If voters were given the right to pursue a contest of the vote on the basis of their vote missing from the string data, they would need a way to prove their original vote.  Proving their own original vote would require positive identification of both ballot and voter together at the time of voting, which in most cases would probably violate the Colorado Constitution.  H. Branscomb]

30. Why does the spreadsheet of ballot strings only show "clean" strings without the original strings required to check the cleaning process?
[The spread sheet in question was extracted (probably by me) from the much more complex Microsoft Access database with many tables including those containing strings before and after “cleaning”. There were at least 4 phases in which vote interpretation was modified- three entirely by software and one with human interaction.  Again the intermediate values and scripts which handled some of this "cleaning" in the database is proof of the "complexity" of the IRV tabulation method. One reason to study the ballot images is to determine what the “original strings” representing the voter marks should be.  The database does contain strings representing the mark detection software’s interpretation of the marks.  But these cannot be verified without access to the ballot images. H. Branscomb]

31. Are any materials from the L&A tests (both AV-OS and IRV) available to the public? Were any incorrectly or poorly marked ballots tested?
[We have seen no documentation which may have been created during the LAT.  These were requested by open records request and not received.  There is no evidence that poorly marked ballots were tested.  In some elections in Colorado, citizens have been prevented from poorly marking ballots during LAT.  In one case, the tester was removed from all testing and the law subsequently changed to allow such removal of testers. H. Branscomb]

32. Since the software counting method had to be adjusted during the course of election night, isn't it fair to say that the system used was not certified?
[There is no question that the system used (True Ballot, Inc.) was not certified.  Even the Diebold/Premier system 1.94W could be said to be not certified, given that it was grandfathered in a way to avoid the determination by the judge in Conroy v. Dennis that no voting system in Colorado had been certified prior to 2007.  Unlike other systems, Diebold 1.94W was not resubmitted for certification in Colorado in 2007.  Now, thanks to HB 09-1335 all systems in Colorado regardless of their certification status that were used in the 2008 election are grandfathered for use in Colorado forever. H. Branscomb]
 33. Was the True Ballot tabulation system required to be certified by law?

[ This question has been directed to the secretary of state's office.   I have already been informed by them that they have no jurisdiction over municipal elections. Of course, the entire Constitution and some of Colorado Revised Statutes do apply to municipalities including "home rule" municipalities like Aspen Colorado. H. Branscomb]

34: Question: If you can get an image of everyone's ballot, will you be able to see who the voter was?  I know that when I vote there seems to be a number tied to that ballot and they ask for an ID when I show up.  Also, in the last mail only election I had to put my name on the envelope the ballot came in.  What's to prevent the person who gets the ballots from looking at each ballot and who sent that ballot in?

[ No, barring something completely unexpected, the ballot images on the True Ballot CD have nothing on them which would identify the voter any more than information already available to the public ( the strings containing the interpreted vote patterns, and the associated file names which are not unrelated to the order of voting ).  The number on the stub of the ballot is tied to a particular voter but it was removed before the ballots were scanned.  The information on the envelope is of course removed before the ballot is counted.  After separation, the envelopes remain separate from the ballots and of course separate from the ballot images. At the time the absentee votes were counted however, the envelopes were removed from the ballots in the same room and nearly at the same time as the ballots were counted. Thus it is possible that election judges were aware of the identity of the voter at the time the ballot was unfolded and placed into the Accuvote voting machine. It is only proper procedure which prevents the person opening the envelope from knowing how that voter voted.  In well run elections the two steps occur in different rooms, or at least at completely separate times.  H. Branscomb]

No comments: