Cablegate: Pre-Election Polling Plans

VZCZCXRO4455
PP RUEHDBU RUEHPW RUEHSL
DE RUEHBUL #1714/01 1830244
ZNR UUUUU ZZH
P 020244Z JUL 09
FM AMEMBASSY KABUL
TO RUEHC/SECSTATE WASHDC PRIORITY 9820
INFO RUCNAFG/AFGHANISTAN COLLECTIVE PRIORITY

UNCLAS SECTION 01 OF 03 KABUL 001714

SENSITIVE
SIPDIS

DEPARTMENT FOR SRAP, SCA/FO, SCA/A, EUR/RPM
STATE PASS TO AID FOR ASIA/SCAA
USFOR-A FOR POLAD

E.O. 12958: N/A
TAGS: KDEM PGOV PREL AF
SUBJECT: PRE-ELECTION POLLING PLANS

REF: A. SECSTATE 62422
B. KABUL 1557
C. KABUL 1506

1. This is the sixth in a series of cables updating and
providing additional details on Embassy plans and efforts in
support of the Afghan presidential and provincial council
elections, per the request in reftel. This cable addresses
post efforts on pre-election polling.

-------------------
IRI Elections Polls
-------------------

2. USAID will sponsor two nationwide polls through IRI,
similar to the one released in mid-June. IRI plans to have
the polls in the field (approximately 60 questions each with
a sample size of around 3,200) during July and will present
them before August 20th. In conjunction with these polls,
IRI is sponsoring two focus groups in July and will present
the results in August. IRI uses the polling company Williams
and Associates which works with Lapis Research and Kabul
Group to conduct their fieldwork. Potential IRI focus group
topics are undetermined at this point.

3. In the event of a runoff, IRI plans to run one nationwide
poll and one focus group before the second round of voting in
October. Both the poll and the focus group will be similar
to those described above. At this time, IRI has no polling
planned for 2010 parliamentary elections.

----------------------------------------
Glevum Elections Research & Focus Groups
----------------------------------------

4. Using an existing contract for election support with
IFES, USAID is also sponsoring two Glevum Associates
nationwide polls using the Afghan Center for Socio-Economic
and Opinion Research (ACSOR) to conduct the fieldwork. The
Glevum polls will include samples of 3,400 Afghans and
approximately 20 questions. The questionnaire covers
presidential candidate ratings and favorability, provincial
issues, provincial council role, local governance and
influence, and the perception of credible, secure, and
inclusive polling. Both surveys will go to the field in
July, two-weeks apart. Glevum will return the results of the
first poll on August 5. It will return the second poll on
August 15. The Glevum poll will also have an educational
aspect, since the pollsters will use a paper in the style of
the ballot to show candidates' pictures and symbols to elicit
favorability ratings of each candidate.

5. Glevum will also run eight focus groups in mostly Pashtun
dominated provinces: Farah, Helmand, Kandahar, Uruzgan,
Ghazni, Zabul, Paktika, and Nangarhar. The Glevum focus
groups will bring shura members from several districts
together to talk about regional issues, potential solutions,
and relevant candidate platforms. The Pashtun regions were
selected because of their strong support for Karzai in the
last election, IRI assessed decline in support for Karzai,
and perceived influence in the coming election as the
dominant ethnic group. Glevum focus groups will be conducted
in July with results available during the first week of
August.

---------------------
INR Elections Polling
---------------------

6. State/Intelligence and Research (INR) will conduct two
nationwide polls (4,000 and 6,000 respondents) before the
election. Results from INR's June survey, conducted by
ACSOR, will be available in late July. It includes open
questions about presidential candidates and favorability
ratings for several candidates and non candidates. Results
from INR's July survey will be available in early August.
The July questionnaire remains in development but will
include roughly 40 closed-ended and 10 open-ended questions.
INR will likely award the July contract to either ACSOR or
the Opinion Research Center of Afghanistan (ORCA) to conduct
fieldwork. For all of its surveys, INR obtains the data sets
from the field firm and analyzes the results internally.

-------------
Other Polling
-------------

7. ISAF's Afghanistan Assessments Group (AAG) expects the

KABUL 00001714 002 OF 003

return of its nationwide quarterly survey in early July. The
survey covers a range of issues beyond elections but includes
questions about voter registration and likelihood of voting
on August 20. RC-East is conducting regional elections
polling and expects the results in July. This poll includes
a question about the extent to which the respondent will
support the elected government if their preferred candidate
does not win.

8. CSTC-A is also conducting nationwide polling focused on
security and related to the ISAF AAG poll. It also includes
some elections registration and process questions. Its
results are expected in mid-August.

9. USAID also sponsors The Asia Foundation poll, which is an
annual poll of overall conditions in Afghanistan focused on
attitudes toward democracy and rule of law. Although the
poll will be in the field throughout the summer, its results
are not expected until October 2009.

10. The Broadcasting Board of Governors (BBG) will conduct a
nationwide survey (of approximately 2,200 Afghans) in
September focused on political views and media use. The BBG
works with Intermedia to supervise their surveys, which in
turn works with a local field firm (in the past, ACSOR) to
conduct its fieldwork. BBG expects results from its next
survey in early October. In case of a second round of
elections, the poll will likely cover elections topics.

11. Gallup will also conduct a national survey in July,
supposedly including elections topics. They have used ORCA
for previous surveys. Discussions with UNAMA, UNDP-ELECT,
and other diplomatic mission staff in Kabul reveals no
additional polling plans are currently in place in advance of
the election.

--------------------------------
Polling Companies & Capabilities
--------------------------------

12. There are several polling companies, of varying
capabilities, operating in Afghanistan: 1. The Afghan Center
for Socio-Economic and Opinion Research (ACSOR Surveys or
ACSOR) based in Kabul and managed by D3 Systems Inc. in
Vienna, Virginia; 2. Opinion Research Center of Afghanistan
(ORCA) based in Kabul; and 3. Williams and Associates, based
in Salem, Massachusetts, which works with Lapis Research,
based in the U.S. and Kabul Group, based in Kabul.

13. ACSOR has carried out the bulk of the nationwide surveys
in Afghanistan since 2004. ORCA, which is run and staffed by
several former ACSOR employees, has worked most recently with
Gallup. Kabul Group has focused mainly on qualitative
research in the past, but is currently developing its
nationwide survey capacity in conjunction with Williams and
Associates and Lapis Research. This consortium is associated
with Moby Media.

14. Each firm has local, full-time employees familiar with
the area and rely on both male and female interviewers to
obtain more comprehensive coverage. Each firm translates
questionnaires into Pashto and Dari to conduct interviews
face-to-face in respondents' homes using their local
interviewers who are familiar with the local dialect and
culture. However, the firms' sampling, interviewer training,
quality control procedures, questionnaire design, and data
entry and cleaning capacities differ.

-----------------
Evaluating Polls
-----------------

15. Although public opinion surveys in Afghanistan are
increasing in sophistication and number, many challenges
remain for conducting surveys that meet international
standards. Beyond routine research challenges such as
methodology or sample design, key obstacles in Afghanistan
include limited on-the-ground experience with survey
research; extensive damage to infrastructure
(roads/telecommunication) which impedes access to the rural
population; insecurity/violence; a high level of illiteracy;
and cultural and religious values that sometimes hinder
access to women, promote biased (regional or ethnic) or
socially desirable responses. Exacerbating the difficulty of
designing and evaluating studies is the lack of definitive
population data, as Afghanistan has not conducted a national
census since 1979.

KABUL 00001714 003 OF 003

EIKENBERRY

© Scoop Media