Blogs

Yesterday’s political headlines in the United States regarding Mitt Romney winning the Republican Caucus in Iowa had as many lessons for technology purveyors as for the purveyors of political commentary. The news media debated endlessly the meaning of the elections. Were the voting Republicans supporting a politically “moderate” person over more “conservative” candidates? Did money or negative ads make a difference in the voting? To what extent were Iowa voters comparable to the generic US voters that will eventually determine the next president? And so on.

What seized my attention though was a Fox News cable television claim that some of the votes in a certain district were cast by recording the votes on paper ballots, and then belatedly reporting the votes to the central election organization tallying the overall voting results. Afterward the paper ballots were to be discarded. Iowa apparently has a long history of using a process where “… caucus workers will hand out "ballots," which are usually slips of paper on which voters write a candidate's name. Different precincts use different rules with differing levels of secrecy regarding how voters hand in their ballots, but the caucus chair announces the winner at the precinct after the votes are tallied.” (http://www.globalpost.com/dispatch/news/regions/americas/united-states/111230/iowa-caucus-5-key-facts)

However, according to a CNN cable news story, Edith Pfeffer, the chairwoman of the Clinton County Republican Party, played a major role in the final counting of the caucus votes. “Pfeffer explained that she had been asleep until friend and colleague Carolyn Tallet, president of the Clinton County Republican Women's Club, woke her. Tallet told Pfeffer that it was suspected that ballots from Clinton County had not been counted, and that these ballots could decide the winner.” (http://www.washingtonpost.com/blogs/blogpost/post/iowa-caucus-results-edith-and-carolyn-celebrities-after-solving-mystery-of-missing-votes/2012/01/04/gIQAAlMYaP_blog.html)  After phone calls, television interviews and a new tallying of the results,Edith and Carolyn became momentary TV stars.

Apparently they were contacted about the need to send in their votes, when it appeared the central election organization had never actually received or tallied the Clinton County votes. “They thought there was a computer glitch in Des Moines, …” (http://www.theatlanticwire.com/politics/2012/01/edith-and-carolyn-were-stars-cnns-coverage-last-night/46941/)  Suddenly, with some scrambling through trash cans according to some accounts, the ballots were found, and re-reported to the central coordinating election organization. It made for great insights into both grass roots participation in the American election processes and a stark reminder of how risky some manual information processing workflows can be. These late votes were some of the ones that enabled Mitt Romney to squeak by Rick Santorum to an amazingly small eight vote lead out of tens of thousands of votes that had been cast. (http://www.foxnews.com/politics/2012/01/03/gop-candidates-await-iowa-verdict/)

Many Americans will recall the paper ballot “hanging chads” that caused enormous consternation and disagreement during the 2000 presidential election contest between George Bush and Al Gore. (http://en.wikipedia.org/wiki/United_States_presidential_election,_2000and http://en.wikipedia.org/wiki/Chad_(paper))  In fact, the paper balloting processes themselves were highly debated for years and questioned extensively with respect to reliability, openness, chain-of-records-custody and other issues. Some states made paper-based balloting illegal. The Iowa balloting process is still paper based in many locations and is not inherently inaccurate or subject to compromise if the accounting and the arithmetic are done accurately. (http://www.blackboxvoting.org/).  Accurate records must dominate the process – 1) the registered voter listing, 2) the listing of actual voters, 3) chain of records custody trustworthiness, and 4) records of actual votes counted.

However, you would think with all of these records floating around, there would have been a successful attempt by now to completely automate the voting processes of most elections. Unfortunately, there are continual challenges including local cultures and customs, financing of systems, and more than a few electronic records issues. In addition, it seems that paper-based processes as a cultural norm can be transcended more easily than the continual security risks posed by open computer networking systems.

In fact, it was speculated that the “Occupy” Whatever folks would try to attack the Iowa Caucus process, through denial of service or SQL insertion techniques, though this never materialized. (http://biggovernment.com/cjohnson/2011/12/21/hacking-the-iowa-caucuses/). Considering the continually documented instances these days of data breaches experienced by credit card, financial and health care institutions, it would seem attacking election records systems would be an expected natural goal for hackers to attempt. Being able to thwart these attacks would be a huge win for technology system vendors.

Issues regarding the riskiness of electronic records in the voting process have been debated and documented over many years. (for example - http://euro.ecom.cmu.edu/people/faculty/mshamos/paper.htm).  Many issues evolve out of cultural resistance and election workflow related rather than technology embedded concerns. However, it would seem that technology systems vendors would have a major marketing and social responsibility opportunity to collaboratively work together to eliminate the technology concerns in this arena. After all, it gets great nationwide press coverage when a couple of charming elderly ladies in the Iowa farm belt of America during an election can still blame it on a “computer glitch.”

Despite the obvious human factors relevant in the creation of a viable electronic election records system, just taking on this challenge should get great press for technology vendors. Who knows? Maybe it takes a computer to figure out the way to create a secure and implementable election electronic records system. Maybe we should just ask IBM’s Watson for the answer.