Response to TOC - 2010

U. S. ELECTION ASSISTANCE COMMISSION
OFFICE OF THE EXECUTIVE DIRECTOR
1225 New York Avenue, NW, Suite 1100
Washington, DC. 20005

May 5, 2010

Dear Mr. Hunt:
EAC thanks OMB for its acknowledgement of the improvement in the quality and completeness
of EAC’s 2008 data in its draft terms of clearance for the 2010 Election Administration and
Voting Survey. However, EAC is concerned about repeated references to the quality of the
UOCAVA data, particularly when such issues have been addressed within the report itself, on
EAC’s website, and with OMB and DoD. As the UOCAVA report (and others within the Survey
report series) demonstrates, States have made strides in collecting and reporting election data to
EAC.
EAC interprets HAVA’s mandate for election data collection to require a census approach;
therefore, EAC’s data collection effort does not involve sampling of the States and local
jurisdictions. EAC compiles the data provided from the States for use in its reports (pursuant to
HAVA Section 703, which relates directly to UOCAVA data). In 2008, EAC received data from
all 55 States and territories. EAC notes its census approach in all of its reports, along with
caveats about missing data and cautionary notes about any attempts to draw conclusions based
on incomplete data sets. The fact that EAC dedicates an entire section of its reports (and an
online FAQ sheet) to informing readers about the States’ data and issues related to States’
nonresponse, demonstrates that EAC takes seriously its duty to inform all readers of States’ datarelated issues. As a way of providing context for the numbers appearing in the reports, EAC
performs some descriptive analysis. However, EAC makes the raw data publicly available so
that if readers wish to do more extensive statistical analyses that involve weighting and
imputation, they may avail themselves of the opportunity. EAC will continue to include the
previously noted caveats in its 2010 reports to Congress and on its website.
With respect to data quality, EAC has developed and maintains a rigorous written protocol for
State verification and approval of the data they submit. In addition, EAC has instituted a
procedure in which all data submitted by the States are checked for possible data entry or
tabulation errors. This data checking is conducted by at least two EAC research staff, along with
the EAC contractor performing the data collection and analysis (the 2008 contractor was the
Research Triangle Institute). Upon finding a questionable data point — for example, a State
submits data for the number of ballots counted that is larger than the number of ballots submitted
for counting — EAC goes back to the State in question to obtain clarification. If that State
maintains its numbers are correct even after being notified of an error, EAC believes it has an
obligation to report that State’s data. EAC does not change a State’s submission, but rather
includes the information as provided by the State. EAC will continue to engage in the data

quality and consistency checks described above for its 2010 data collection effort. EAC’s
protocols are available for your review upon request.
As with past reports, EAC will continue to clearly indentify the data as State-reported figures.
For example, page 7 of the 2008 UOCAVA report section II reads, “States reported transmitting
989,208 ballots…” On that same page under section III, “States reported that 682, 341
ballots…were returned and submitted for counting.” EAC will continue to use this language in
its 2010 report.
Thank you for the opportunity to provide additional comment on OMB’s draft terms of clearance
for approval of the 2010 Election Administration and Voting Survey.
Sincerely,

Thomas R. Wilkey