Wednesday, October 31, 2012

Post-Election Report, Part 2 - Results

The Central Electoral Commission has not yet processed all of the protocols, but 98% are completed. At this point in time, the results are the following:

Party PR Vote % PR Seats SMD Seats Total Seats
Party of Regions 30.22 73 115 188
Batkivshchyna 25.37 61 43 104
UDAR 13.86 34 6 40
Communist Party 13.26 32 0 32
Svoboda 10.36 25 11 36
United Center
0 3 3
People's Party
0 2 2
Oleg Lyashka Party
0 1 1
0 1 1
0 43 43

Total 93.07 225 225 450

The final vote processing could change SMD results, so those totals may shift slightly. To form a coalition, the Party of Regions needs 38 deputies to reach the "magic number" of 226 for a majority.* Cooperation with the Communist Party will not be enough to establish a majority; Regions will need another party and/or independent candidates to sign on to the agreement. Batkivshchyna and Svoboda are not potential coalition partners, leaving only UDAR. The failure of UDAR to sign a pre-electoral coalition agreement with other parties certainly implies that it might be willing to negotiate with Regions, although some comments from the party suggest otherwise.

Negotiating with independent candidates will be costly for Regions, both in terms of resources and longer-term coalition stability. Independents will require payoffs to join and maintain allegiance to any coalition. While Regions has resources, this parliament is likely to be less stable and more contentious than the most recent convocations. The Rada may not return to the chaos of 1998 when the first task of selecting the speaker was a drawn-out and controversial process, proving to be a harbinger for the whole convocation, but it is likely to be more fractious than the last two convocations.

*The math for an opposition coalition is less promising. Batkivshchyna, UDAR, and Svoboda - if they could make an agreement - are still 46 seats shy of the number needed. Because the Communists are not a feasible partner, they would need to attract nearly all of the independents (which is essentially an impossible task).

Action in District 211

Yesterday, I visited the District Electoral Commission managing the compilation of votes for Constituency 211 in Ukraine's parliamentary election. The race has been tense and tight. At this moment, with 81% of the protocols completed, the Batkivshchyna candidate (Ter'okhin) leads the Party of Regions candidate (Lysov) by just over 1,800 votes. Because it is Wednesday, and the elections were held on Sunday, it would be reasonable to ask why the results are not yet finalized. My visit to the district provided me with valuable insight on this question.

When I arrived in the district around 1 pm, all was quiet. Observers are allowed to attend the meetings of the commission held in a small auditorium, so I sat in a front-row seat. I introduced myself to a polling station director who was awaiting the processing of his materials. Before polling station officers complete their work, it must be accepted by the District Electoral Commission. During our conversation, he had a visit from a frustrated colleague: the DEC* was rejecting their documents because the parties were not placed in the proper order on the protocols. He was especially annoyed because the DEC had been reviewing the materials for some time - why did it take them so long for this decision?

After he left to reconvene his polling station commission and re-write the documents (all protocols are hand-written), I struck up a conversation with a representative of the Committee of Voters of Ukraine. CVU is a well-established domestic NGO that monitors elections. He had been stationed at the DEC and told me that the initial processing was quite slow. The first polling station to turn in documents was a small hospital, and the DEC took several hours on its materials alone. When I asked him about other observations, he noted that he had received reports of vote buying in the outlying regions of Kyiv, but had no concrete evidence.

Mid-late afternoon, the action picked up. The DEC decided to re-convene its meeting to address several issues, including the disposition of materials from PEC 800029. The members of that commission were agitated, and I came upon a conversation they were having with other observers in the hallway before the meeting. They claimed that the DEC was attempting to alter their final results.

When the meeting began, a member of the DEC (I believe it was the deputy director) began to read a resolution that would call for recounts in around 29 polling stations. At this point, the meeting descended into chaos and protests. During the chaos, the PEC director climbed onto the stage, knocked off the top of the box containing her station's materials.

She and her colleagues loudly proclaimed that the DEC had illegally opened the boxes and were tampering with the contents. She eventually occupied a spot on the commission table and would later declare that she was going to begin a hunger strike.

As more chaos ensued, the militia arrived to clear the stage.

Eventually, the hall began chanting for the original protocols to be certified.

Apparently, one of the members of the DEC suffered a heart attack on the previous day and he returned from the hospital to participate in the voting. In addition, a member of parliament and the media arrived on the scene to participate or chronicle the goings-on. The candidate who is at the center of the controversy - Ter'okhin - also arrived, interacted with observers and the press, but was careful not to directly engage in the debate (as he indicated that it would be improper for him to personally interfere).

The DEC took a vote to rescind the decision to recount PEC 800029's results, but the vote to rescind failed. When I left the DEC around 7:30 pm, the PEC and DEC were locked in a standoff, with neither side willing to yield. Formally, the DEC voted to recount and the PEC, as well as others, were lodging complaints. As of this moment, the protocols for PEC 800029, and many others in DEC 211, are not yet uploaded to the CEC website.

What I witnessed was not theater, but a clash over the meaning of free and fair elections.

*I have used TEC and DEC in tweets and posts interchangeably. Territorial Electoral Commission is often used as a generic term for mid-level commissions. But, in Ukraine, the better acronym is DEC since the commissions are labeled District Electoral Commissions. Apologies for any confusion.

Tuesday, October 30, 2012

Post-Election Report, Part 1 - Preliminary Results and Turnout

District electoral commissions continue to work through the protocols and report results to the Central Electoral Commission. As of 9:00 am (Kyiv time) on Tuesday, just over 86% of the results have been entered into the CEC database. Some districts have been slower than others and the failure to have more reports in at this point is a bit troubling. While the final results will vary a bit as those 14% are added to the final tally, the party list results have produced some interesting surprises. The party list vote thus far shows the following:

Party % #
Party of Regions 31.83% 5 554 712
Batkivshchyna 24.22% 4 226 529
Communist Party 13.88% 2 423 665
UDAR 13.47% 2 350 751
Svoboda 9.61 1 677 042
Ukraine Forward! 1.61% 282 037

Before addressing the surprises, a couple of comments on expectations that were fulfilled. The Party of Regions performed best on the list, followed by Batkivshchyna. Ukraine Forward! failed to pass the 5% threshold and Svoboda succeeded. UDAR, in a surprise, did not finish a solid third, but is currently in a virtual tie with the Communist Party. Klitschko had been receiving substantial attention in the days before the vote and his ascent to second place was considered to be a possibility. Instead, it appears that many opposition voters may have selected Svoboda as an alternative. Once the final results are in, I will address these issues in greater detail.

Because precinct electoral commissions are required to report on turnout throughout the day, the official reports on citizen activity should be much closer to final. Overall, turnout in parliamentary elections just about matched its post-Soviet nadir of 57.94% in 2007; turnout on Sunday is reported to be 57.98%. 

Turnout by District

Turnout data reveal some interesting regional patterns of activity, which suggest variation in mobilization success by competing forces. The map above shows district-level turnout across Ukraine as of 8pm on Sunday when the polls closed (using natural breaks (Jenks) for categories). Activity is notably stronger in western Ukraine (except Zakarpattya) through the central region to Kyiv, with scattered districts showing high levels of activity in Ukraine's east. At the oblast level, western Ukraine exhibits the highest level of mobilization, with L'viv (67%), Ternopil (67%), and Volyn (66%) leading the way. The lowest turnout was in Ukraine's south, with Crimea (49%), Odesa (50%), Sevastopol (50%), and Kherson (51%) at the bottom of the list. Given the level at which Ukrainian citizens express dissatisfaction in surveys with most politicians and institutions (see the most recent IFES survey, for example), voter apathy is not surprising. The imprisonment of opposition leaders and enactment of the language law, however, may have served as a stronger mobilization tool for western interests than for those in the east and south. I will explore variation in party performance by region in a later post (once all the results are in), especially Svoboda's performance outside of its core western regions.

In addition to looking at final turnout figures, it can be interesting to assess the change in turnout during the day. The next two maps show the difference between reported turnout at noon and 8pm, and at 4pm and 8pm in standard deviation terms. 

Difference in Turnout, Noon-8pm (Standard Deviation)

Difference in Turnout, 4pm to 8pm (Standard Deviation)
Patterns of turnout may vary across space based on the work and personal lifestyles of the population. For example, when I was living in L'viv during the 2007 parliamentary election, I noticed a surge of voting in the late morning and early afternoon. Since it was a Sunday, I asked why and was told that locals come to vote after church whereas voters might show up in larger numbers earlier in other regions. I never tested this proposition empirically, but it is plausible. Indeed, the results on the first map would be consistent with it. 

The second map shows an outlier - district 101. If one looks at the four hour increments (noon-4, 4-8), the pace of voting tends to taper off as the end of election day approaches. In Ukraine, it increased in three districts: 66, 101, and 153. In 66 and 153, in increased by about 2 percentage points. But, in district 101 the difference was an impressive 19.4 percentage points. 

Proportion of Voters Casting Ballots by Noon 

Proportion of Voters Casting Ballots by 4pm

The next two maps look at the data differently, taking the final turnout as the total. The first map shows the proportion of voters who cast ballots by noon, and the second shows the proportion by 4pm. Once again, voters in western Ukraine showed up later in the day relative to their counterparts elsewhere. A couple of districts in other parts of the country - again, notably 101 - are unusual in their turnout. In District 101, only 40% of those who are reported to have voted showed up at the polls by 4pm (the mean for all districts was 78%). This district in Kirovohrad has also only reported 29% of its protocols according to CEC records at the moment. It is particularly worthy of additional scrutiny.

In sum, turnout is low and shows regional variation. Some of this variation may be explained by variation in patterns of life, but other aspects of the variation require more investigation. 

Monday, October 29, 2012

Notes from the Field

This year's election in Ukraine marked my tenth official election observation deployment and the day featured standard and new experiences. Per standard procedure, I was connected to a team including another observer, driver, and interpreter. Everyone on the team was professional and we approached all of our tasks with appreciation of the importance of yesterday's vote.

After a preliminary scan of polling places on Saturday, we planned for E-day. The general idea was to visit one polling station in Kyiv so that our driver could cast his ballot, then move into Brovary for the rest of the day. In total, we visited 14 different polling stations, spending around 30 minutes or so at each site.

Perhaps the most notable new feature in Ukraine's parliamentary elections was the introduction of webcam monitoring. We found two webcams in each polling station, one directed toward the registration desk and one directed toward the ballot boxes. Due to the configuration of some polling places, however, cameras were sometimes located in awkward positions. In the photo below, you see a camera directly placed over a voting booth, with a sign noting that video surveillance does not cover activities in the booths. While the camera was indeed pointing in another direction, voters might find this particular juxtaposition disconcerting.

Each polling station also had a laptop in a locked metal box where the feed was connected. We observed the functioning of the webcams in a few polling places and the laptops were recording images in real time. When our interpreter Katya spoke with a family member so that we could test the cameras, we found that the cameras in that site had a significant delay of several minutes before images found their way online.

Our poll opening was a bit hectic as the commission was stuffed to the gills with observers as you can see above. According to the director of this station, none of the commissioners had ever participated in an election on a polling station commission before. I have never encountered a completely green PEC and some of the early chaos may be explained by the lack of experience. The morning commenced with a flurry of activity at the last moment, but the station opened on time for voters to begin casting ballots.

A general rule in observing elections is that the team first introduces itself to the polling station officers and is registered by the Secretary. One team member begins gathering data from the officers while the other begins to mull around and look for areas where specific problems could emerge. I typically look for things like unusual patterns of signatures on voter lists (or problematic stamps), issues with ballot box seals, the accuracy and efficiency of the commission in issuing ballots, whether or not voters are having problems with ballots and how questions are addressed, and if anyone seems to be influencing voters in any way. Outside the polling station, I look for any signs of partisan activity (or other questionable activity) and if the basic materials required are in place (e.g., official signs with the party lists).

Observation involves quite a bit of standing around. To the extent possible, I try to engage with other observers to see what they have found on the ground. While the report that our team files focuses on what we see and hear directly, reports from other observers are often quite helpful. We spoke with local party observers, media representatives, international observers, and even one of the candidates throughout the course of the day.

In past elections, I have on occasion witnessed direct acts of fraud or intimidation, including efforts to stuff the ballot box with mobile votes from psychiatric patients (Ukraine 2002), biased commission decisions favoring one candidate over another (Ukraine 2004), likely voter "carousels" (Ukraine 2004 and Azerbaijan 2008), and egregious efforts to forge voter lists and intentionally delay the vote count (Ukraine 2010). I have also witnessed commissions working in good faith, but making errors that rendered the reconciliation of vote counting impossible (Russia 1999), or violating election provisions because of inexperience.

The major organizations monitoring the election are in the process of releasing final reports and they have more complete information than a single team that visited a dozen or so sites in a single region. While we encountered some suspicious activities, such as polling station officers unwilling to identify which party they officially represent at the polls, and accusations from other observers of tampering with the storage place for ballots (which we did not witness), most of the problems we encountered were procedural and likely due to training issues.

Above you can see an example of a minor infraction that was unaddressed. I watched a young girl waving a Socialist Party flag while accompanying her father to the polls. She waved around the lovely pink banner while her father got his ballot, completed it in the booth, and dropped it in the box. Was she attempting to agitate for the SPU? Unlikely, but a polling station commission member should have discreetly asked her father to store the flag while in the polling station. It is a little thing, but partisan agitation is illegal and the polling station members either didn't see it (hard to imagine) or thought it was "no big deal" because it was a small child. Fortunately, most of the issues we found were on this end of the scale of malice.

We attended the overnight vote count at a track and field training facility. Two commissions occupied the same space but they varied in how well they understood and applied the rules. Ukrainian law specifically spells out the steps to take in the vote counting process and if these steps are followed, most problems can be avoided. Unfortunately, our commission mixed up steps, did not check its math through the night, and tried to hurry up a bit on the process. The result? By our departure time around 6:00 am, the commission could not reconcile the numbers to complete the needed documents and had descended into bickering. We ended up leaving without the protocol - a document we always try to procure - because the commission simply was not able to address math issues that had been apparent to many all night long. Below are some final images featuring the steps of the vote counting process. I'll post more about the results and implications as the CEC finalizes its data.

Saturday, October 27, 2012

Preparing for Observation

Short-term election observation requires a substantial amount of patience, fortitude, and understanding. Despite what officials in a certain US state claim, the purpose of observation is not to interfere with the electoral process. But, the impression that observers have sinister intent is not limited to fringe cases. In the ten missions in which I have participated (including the ongoing deployment in Ukraine), I have regularly been asked about the purpose of observation and challenged to defend my qualifications, or the authority of my organization or home country to engage in the process.

Observing elections is something like being an efficiency expert, or perhaps the observer in the film Kitchen Stories. You spend a significant amount of time watching people do their jobs, asking questions, and taking notes. Understandably, it can be quite disconcerting to be the one who is watched and about whom notes are taken. It is quite natural to be defensive or sensitive about being monitored and evaluated by people you do not know who gather information about you for a short period of time.

The ultimate goal is to better understand the quality of the process by watching its parts, gathering and aggregating data, and finally rendering a general assessment. Organizations vary in how they function on the ground and how they process data, but the broad contours of  the process are similar among monitoring groups with which I am familiar. To make a general assessment, groups - especially those with large numbers of observers - may attempt to randomly assign teams to polling stations. I suspect that these assignments are not fully randomized as elements of convenience (e.g., issues related to transportation) are factored in. However the assignment is arranged, observers are deployed, collect data, report the data, the data are processed, and a summary is presented to the public. Regardless of the assessment, the evaluation should help improve the electoral process next round.

My team is assigned to TEC 97 near Kyiv. According to data from Maidan Monitoring, TEC 97 is a district where many pre-election violations have been reported relative to other districts. The map below shows reported violations across the country, including only those reports where a constituency could be identified.* The SMD race is competitive and one where Batkyvshchyna and UDAR did not successfully coordinate. In sum, 24 candidates are contesting, but the Batkyvshchyna, UDAR, and Party of Regions candidates will probably be the most competitive.

As I have stated in other entries, I am not posting assessments of the election process until the balloting is over and organizations have evaluated data from across the country. What my team finds in Brovary tomorrow may - or may not - represent what is occurring across Ukraine and it would be inappropriate for us to comment publicly on the process before the "big picture" is clearer.
Map of Reported Violations, Using Maidan Monitoring Data as of 10/27/12.
Thanks to Serhij Vasylchenko for the map shapefiles.

*I will update the map with Maidan Monitoring data after election day (when more alleged violations are likely to be entered).

Election Eve

A few hours ago, official campaign activities ended for Ukraine's 2012 parliamentary elections. The Party of Regions held a rally not far from where I am staying in Kyiv, and the mood was celebratory. As Ukraine moves forward to election day on Sunday, a few aspects of the election should be given special attention.

  •  Does Batkyvshchyna receive more party list votes than UDAR? Most of the polling prior to the election suggested that Batkyvshchyna should edge out UDAR for second place on the party list tier. However, UDAR has been receiving significant attention since the polling moratorium on October 18 that may affect its ability to mobilize voters. Relative party list performance is primarily important for its symbolic value. It is the first set of results that will be reported via exit polling around 8:00 pm on Sunday night. In 2002, the opposition led by Our Ukraine was able to make a claim of victory given its strong PR showing. When all of the ballots were counted, and independents elected in SMD aligned themselves in party factions, regime supporters retained their dominant position. This year, however, the Party of Regions is likely to perform best in PR, but the final disposition of parliament will be determined in the districts.
  • Does Svoboda and/or Ukraine-Forward! pass the 5% threshold? Because votes will be wasted on parties that do not pass the threshold, the seat allocation in PR will be greater than the percentage of votes that the parties receive (That is, if the Party of Regions, for example, receives 30% of the vote, it will subsequently receive greater than 30% of the 225 seats allocated to PR because parties that do not pass 5% will not participate in seat allocation.). If these parties succeed on the party list, they may become significant players in making or breaking coalitions. The odds seem much better for Svoboda than Ukraine-Forward!, but either or both could pass the threshold.
  • Do coordination failures hurt the opposition? As I have noted before, the opposition has arguably performed better in terms of pre-electoral coordination than in past elections. Although Batkyvshchyna and UDAR accused each other of not living up to their agreements in good faith, especially regarding coordination in Kyiv, they cooperated in many districts. SMD races are much less predictable and are easier to manipulate, so coordination failures could prove costly.
  • How engaged are Ukrainian citizens in casting ballots? Turnout is an indicator of engagement, and it may be low. Turnout has been falling in parliamentary elections, with 75.81% in 1994, 69.64% in 1998, 65.22% in 2002, 58.97% in 2006, and 57.94% in 2007.* Parties have been attempting to mobilize, but public opinion data suggest that voters perceive limited efficacy in their participation.
  • How do observers assess the quality of election administration, especially in the SMD contests? The elections will be closely scrutinized, with thousands of international and domestic observers deployed all over the country and webcams in every polling station. If mischief occurs, it is more likely to be associated with SMD contests and is consequently more likely to be diverse in its forms. Pre-election reports have emphasized concerns about the potential for vote-buying, but the entire menu of manipulative activities is potentially available should unscrupulous actors be interested in committing fraud.
  • How do Ukrainian citizens respond to the outcomes? Inevitably, some participants will claim that fraud undermined the process. Indeed, a recent IFES survey noted that over half of Ukrainian respondents expect fraud to be a part of this year's contest. Large-scale mobilization does not appear to be in the works as it was when I was here for the elections in 2004. But, I expect to see victory rallies and protest rallies nearby on Monday.
I will be tweeting about the election via @erikherron. Because I am serving as an official observer, I will not comment on election quality during election day. I will make notes about what is taking place on the ground, and will post more details after the organization's press conference on Monday.

*Turnout figures are from Ukraine's Central Electoral Commission, except for data from 1994 which come from International IDEA. The IDEA data consistently report higher turnout than the CEC, but they also show a general decline in participation over time (with an uptick in 2002).

Monday, October 22, 2012

Observations about Ukraine's Upcoming Election, Part 3

In less than one week, Ukrainian citizens will cast their ballots in the third parliamentary election in six years. The moratorium on printing polling results began on October 18, and most of the polls that I read suggested that the likely party list outcome is: Party of Regions>Batkyvshchyna>UDAR> Communist Party>Svoboda>Ukraine-Forward!. The prohibition on polls does not prevent bookies from giving odds, with the Party of Regions identified today as the odds-on favorite to win the most seats. A couple of key party list questions remain:

1) Who will be the "last party in" above the 5% threshold? The polling suggests that Svoboda may make it and Ukraine-Forward! will not (despite the latter's protests). The United Opposition seems to be reasonably confident in Svoboda's chances as it made post-election coalition plans public.

2) How will SMD seats be distributed? If we look at the last election held under the mixed system, the party list and constituency seats varied substantially. The party-of-power in 2002, the For United Ukraine electoral bloc, overperformed in SMD relative to PR and some of this may have been due to fraud in the districts. [For example, I demonstrate with my colleague Paul Johnson in a book chapter about special polling stations in 2002 that the party-of-power disproportionately benefited from special voting, especially in prisons.]

The opposition has been more successful in pre-electoral coordination in the districts in 2012 than in earlier elections, with opposition parties reducing the number of districts in which they compete. This phenomenon is not new, I wrote about coordination in a Party Politics article in 2002, but it seems to be more deliberate and extensive in 2012. Coordination failures have prompted some sniping, with Batkyvshchyna and UDAR interpreting differently their agreement on coordination in Kyiv. [According to stories, the United Opposition wanted to divide districts based on polling of likely winners and allow competition where Batkyvshchyna-UDAR head-to-head competition would not yield a winner from outside of those two parties. UDAR's position is that the United Opposition failed to live up to the agreement.]

UDAR seems to be the emerging phenomenon in the final week of the election, with journalists and observers offering opinions about the likelihood that UDAR will "beat" Batkyvshchyna in the party list and finish second, questions about Klitschko's status as a "legitimate" member of the opposition, and interpretations of Klitschko as (perhaps) the next big thing in Ukrainian politics.

Unfortunately, data from Ukraine are not solid enough (or plentiful enough) to permit analysis like Nate Silver conducts in FiveThirtyEight for US elections, nor are the bookmakers' odds comparable to Intrade's election market. In addition, Ukraine has many undecided voters due to the party system's inchoate nature and consequently more limited party affiliations, making both mobilization and persuasion important factors as election day approaches. Layer on top of these factors the potential for falsification, and the opinions about UDAR's likely performance become little more than guesses. We know that in most polls, over most of the time period covered, Batkyvshchyna finished ahead of UDAR. With attention on Klitschko, but with no polling, it is hard to gauge if the parties' relative performance will change on election day. In many ways their relative placement on the party list doesn't matter - what will be critical is the total number of seats that the two parties hold when constituency races are considered and how these seats empower them to make or break coalitions.

Klitschko has been accused of not being a "true" member of the political opposition. However, in his position it probably pays to be coy. Neither the Party of Regions nor Batkyvshchyna is likely to have enough seats to gain a majority even with minor partners (say, if Svoboda passes the threshold and joins Batkyvshchyna or the Communists join the Party of Regions in coalition). UDAR could be positioned as "kingmaker" in this scenario and strategically withholding information is in UDAR's (and Klitschko's) interests at this point. Of course, it is also possible that independent candidates elected in the districts will hold the key to a majority, and they will extract a hefty price for their support.

A final note: is Klitschko the "next big thing?" Not that long ago, Serhiy Tihipko was the "next big thing" and while he is still a player, he was quickly co-opted. In UDAR's case, its post-election relevance - especially as a key player in coalition-making - will strongly influence Klitschko's staying power as a major factor in Ukrainian politics.

I am heading to Ukraine in a couple of days and will post and tweet (@erikherron) when possible. I will post after the election but it may take some time, depending on my Internet access and travel schedule.

Friday, October 12, 2012

Who Should Observe?

Georgia's recent election yielded many surprises, including the strong performance by Georgian Dream and the relatively quick and decisive announcement of failure by President Mikheil Saakashvili. A story that received scant attention was the decision by the Central Electoral Commission to deny registration to the US-based NGO, Committee for Open Democracy, that was conducting observation and analysis of election administration.*

According to the COD at its press conference, the primary reason for denial was the organization's resistance to disclosing its donor list. The Economist chided COD for not revealing its funding sources, suggesting that the case provides "insight into the potential pitfalls of private companies working within a sensitive sphere which has traditionally been the preserve of governmental and intergovernmental actors." COD is a non-profit, but it occupies a unique space between government-sponsored observation missions and large 501(c)(3) non-profits, like the Carter Center, that also engage in election observation. It seems that the unspoken question regarding registration related to the organization's objectivity and whether or not private donors would influence the content of the election analysis. Of course, this question could also be directed at all other brands of election observation groups.

The landscape of election observation has been changing, and organizations like COD are but one example of this change. No longer the sole purview of governments or international governmental organizations, the practice of observation is becoming more diverse. In addition to the privately-financed COD, crowdsourced observation projects, like Georgia's own Elections Portal, and Maidan Monitoring in Ukraine, offer another approach to observation and different source of data.

Does more observation lead to "better" observation? Assuming that a proper definition of "better" observation could be developed, this is an empirical question that the observation community could - and should - address. Indeed, if more organizations made their data publicly available, scholars and journalists could potentially assess "house effects" akin to those in public opinion polling and develop more precise measures of electoral integrity. While publishing data would require safeguards to protect individuals in the polling stations, the data could be made anonymous and could better inform all interested parties about how assessments of election-day activities are crafted.

*For full disclosure, I served on the COD's Ukraine local election mission in 2010 and anticipate participating in future deployments. The mission was managed professionally in 2010 and the report was careful to speak only to the observations on the ground.