Ohio Voters May Demand Paper Ballots
Ohio Secretary or State Jennifer Brunner announced yesterday that paper ballots must be provided on request [link to http://www.columbusdispatch.com/live/content/local_news/stories/2008/01/04/paper_ballots.ART_ART_01-04-08_A1_ER8V41M.html?sid=101 no longer works].
Poll workers won’t be told to offer the option to voters but must provide a ballot if requested to help “avoid any loss of confidence by voters that their ballot has been accurately cast or recorded,” a directive from Secretary of State Jennifer Brunner said. The paper ballots would be counted by optical scanners at county elections boards.
The Ohio ACLU is against having paper ballots available in the primary, claiming that not having scanners at the local polling locations is against state and federal laws mandating that voters have to know if they made a mistake such as casting too many or too few votes when filling out the ballot.
But Brunner said after consulting with the attorney general’s office, she thinks the ACLU is “flat wrong” and that voters will be adequately educated to avoid unintended over-votes and under-votes — problems that plagued the punch-card voting system that the electronic machines replaced.
Even so, Brunner told The Dispatch that said she is re- thinking her previous recommendation that no ballots be counted in the precincts, after activists argued that would eliminate a way to verify whether the final results are accurate.
The option for having paper ballots is in response to feedback in response the report issued last month by Brunner’s office revealing several critical vulnerabilities [link to http://www.sos.state.oh.us/sos/info/everest.aspx no longer works] in currently available electronic voting systems. Brunner has also recommended that Ohio move to all paper ballots for the November election and has asked that the state legislature Gov. Strickland approve and fund the change.
The executive report [link to http://www.sos.state.oh.us/sos/info/EVEREST/00-SecretarysEVERESTExecutiveReport.pdf no longer works] is long but very educational and well worth reading, especially the recommendations. The full details are also online [link to http://www.sos.state.oh.us/sos/info/everest.aspx?Section=3180 no longer works] as well. California also recently released their own extensive reviews [link to http://www.sos.ca.gov/elections/elections_vsr.htm no longer works] some of which were leveraged for the Ohio study. I’ve only skimmed portions of it so far, but by all reports, it is also very enlightening.
Speaking of California, the Secretary of State Bowen, has announced some very impressive new requirements [link to http://usablesecurity.com/2007/08/04/california-limits-use-of-dres-and-adds-security-restrictions-on-other-voting-machines/ no longer works] for the use of electronic voting. This is great stuff, that helps deal with the issues of existing machines while still allowing the democratic process to move forward. Hopefully other states will follow suit.
In the UK we still use hand-counted paper ballots and it’s a system which works. The final count may not be 100% accurate (but in cases where the last few votes matter, a great deal of care is taken to ensure that they are counted accurately) but it has the distinct advantage of being a completely open process – there is literally nothing to hide. All candidates and their agents are allowed to attend and observe the count so post-count challenges are remarkably few.
It does help that each individual election usually has its own completely separate ballot paper, and we vote directly on far fewer positions and measures. The most I think I’ve ever experienced in a governmental election was three (local, parliament, European Parliament) – for University student union council I think it was twelve although that was a single ballot paper if I recall correctly.
Usability errors can still occur. A few years ago, the ward boundaries in my town (Reading) were changed and all the councillors had to be re-elected (normally one third of all councillors were elected each year for a four year term, with no election every fourth year). The new boundaries created a new ward to be represented by one councillor, the remaining 15 to be represented by three councillors as before. The authorities decided to run each ward’s election by including all the candidates on the same ballot paper and asking each voter to place three marks on the ballot paper against their preferred three candidates. The candidate receiving the most votes would sit for four years, the second most popular for three years, and the third for two. I would expect most voters to vote for a ‘slate’ of three candidates from the same party, but the results were surprising: in all wards, there was a significant difference in the number of votes cast for the top three candidates, and in two, the third-placed candidate was from a different party to the top two. My feeling is that quite a few voters under-voted.
Counting is also simplified because we use first-past-the-post for most elections (i.e. candidate with most votes wins) but in some, e.g. European Parliament, some form of transferrable voting is used. This is again done with the original ballot papers by physically moving them between piles and recounting.
It seems to me that the US is optimizing their voting system for the rapid, automated collation of results, NOT for accurate recording by the voter of their preference and verifiable collation of the results. It’s not as if you need the results that quickly – there is more than two months between the election date and the inauguration date of the new President. In contrast a change of government in the UK happens on election night, if the result is sufficiently conclusive (as happened in 1979 and 1997 – indeed I think every occasion since 1974, where there was a hung parliament).