Polling Center Analysis

Democracy International's analysts have been taking a closer look at the election commission's August 18 announcement on polling centers. Here's what they found:

Introduction

Consistent with the official electoral calendar, the Independent Election Commission of Afghanistan (IEC) announced on August 18 the numbers of polling centers they plan to open in each province for the September 18 Wolesi Jirga elections. This memo provides additional information and analysis on the information released by the IEC.

The vetting process

The IEC developed a list of 6,835 polling center locations, comprising 19,945 polling stations. This list was provided to Afghan and international security forces for their analysts to assess the safety of opening polling centers in these potential locations. After review, the IEC decided that 938 polling centers (13.7% of the initially planned locations), consisting of 2,448 polling stations, were too insecure for polling to take place. With this reduction, the IEC is now planning to open a total of 5,897 polling centers containing 18,762 polling stations for the September 18 vote.

Inconsistency in the IEC's August 18 Press Release

DI uncovered one inconsistency in the IEC's August 18 press release. The IEC reported that 938 polling centers, consisting of 2,448 polling stations, were deemed too insecure for voting to take place. This resulted in 18,762 polling stations being planned to open on Election Day. They also reported an initial number of potential polling stations of 19,945. The difference in the original number of polling stations (19,945) and the reported number of polling stations they are planning to open (18,762) equals 1,183 and is inconsistent with the reported 2,448 polling stations which will not open. DI has inquired with the IEC regarding this inconsistency.

The effect on voters

The IEC has stated they will open additional polling stations in nearby polling centers to accommodate those voters whose polling center has been closed for security reasons. Nevertheless, there will certainly be voters who will now be too distant from the nearest open polling center to cast a ballot.

More polling centers generally permit greater access for voters but the larger number further taxes security assets. The IEC has to judge whether the level of security for a given polling center is sufficient to permit voting or if the opening of that polling center in such weak security circumstances simply provides an opportunity for political capture of the polling center and the fraud that results.

How this decision compares with the number of polling centers closed in 2009

In 2009, the IEC faced enormous pressure to open fewer polling centers, particularly during preparations for the aborted runoff campaign. Advocates of this stance argued that an insecure polling center which is highly susceptible to fraud is essentially worse than disenfranchising some voters.

In 2009, the IEC planned to open 6,524 polling centers. 5,438 (83.4%) were included in the final results but how many of these actually conducted polling is still not known. As a result, it is not possible to directly compare the current plan to open 5,897 polling centers with a reliable figure on the number of polling centers that actually functioned in 2009. Further, it is likely that the fluid security situation will result in the closing of more polling centers before Election Day while it is unlikely that new polling centers will be added.

Implications for candidates

The implications for candidates may be serious, depending on the number of polling centers closed in a particular district. The closing of polling centers could disproportionately affect those candidates whose base of support is in areas with high numbers of closed polling centers.

PC and PS Breakdowns

Of the 18,762 polling stations the IEC plans to open on September 18, 57% are reserved for males, 38% for females, and 5% for Kuchis. In 33 provinces the number of planned male polling stations outnumbers the number of planned female polling stations. Only in Bamyan does the number of planned female polling stations (165) outnumber the number of planned male polling stations (164). In many provinces, the number of planned male polling stations significantly outnumbers the number of planned female polling stations. In seven provinces, more than 25 percent more male polling stations are planned to open than female polling stations (Helmand, Kabul, Kandahar, Kapisa, Khost, Kunar, Kunduz, and Zabul). The largest differential exists in Helmand province, where planned male polling stations outnumber planned female polling stations by 43.6 percent (229 stations).

Comparison with 2009

In order to investigate further, DI compared the summary statistics for the 2010 polling center distribution to the summary statistics of polling centers from 2009 which were planned to open as well as those which were eventually included in the final certified results.

In 2009, the IEC planned to open 6,524 polling centers. 5,438 polling centers were eventually included in the 2009 final results, 459 less than are planned to open in 2010. In only three provinces – Daikundi, Helmand, and Zabul - does the IEC plan to open more polling centers than they had planned to open in 2009. In 11 provinces the IEC plans to open fewer polling centers than were included in the final certified results in 2009 and in 21 provinces the IEC plans to open more polling centers. In 2 provinces the IEC plans to open the same amount of polling centers that were included in the final certified results. Theses comparisons are listed in Table 1. In three provinces – Kandahar, Paktika, and Wardak - the IEC plans to open more than 50 additional polling centers than were included in the final certified results in 2009.

While polling station level data has been posted to the IEC's website it has not been done so in a format conducive to data analysis. DI will provide such analysis when the data is made available by the IEC.