Full Data & Results

The data and results from the Treasure State 2020 Election Surveys administered by the HELPS Lab at Montana State University - Bozeman were made available to the public in February 2022 in a comprehensive manner. The posted files include:

The dataset is available in STATA format through OPEN ICPSR, which hosts all the documents above, as well. The suggested citation for these data is:

Parker, D. C. W., Raile, E. D., & Shanahan, E. A. (2022). 2020 Treasure State Election Surveys [Data file]. Retrieved from http://helpslab.montana.edu 

Further, an in-depth report explores gaps between pre-election and actual results in political races.

Original Pre-Election Survey Results

Available materials for the pre-election survey include:

Brief Summary of Methods

The Treasure State 2020 Pre-Election Poll was conducted by the HELPS Lab of Montana State University-Bozeman between September 14 and October 2, 2020. The original population for the poll was Montana voters who registered by August 20, 2020, and were deemed active by the Montana Secretary of State. We stratified the sample by state house districts, and then drew a random sample of 9,000 voters proportionally from these strata. Sampled individuals received a questionnaire by mail and were asked to return the questionnaire via a prepaid business reply envelope. Respondents returned 1,789 surveys, a response rate of 20.2% based on 8,836 deliverable addresses. The Treasure State 2020 Post-Election Poll, also conducted by the HELPS Lab, used a panel design that returned to the individuals who responded to the Pre-Election Poll. This poll was in the field from November 4-25, 2020. Of the 1,789 original respondents, 1,069 completed the post-election questionnaire, for a response rate of 59.8%. The procedures for the Post-Election Poll otherwise paralleled those for the Pre-Election Poll.

The number of individuals voting in the 2020 general election in Montana was 612,075, according to information from the Montana Secretary of State. We have adjusted the overall survey margin of error (MoE) for design effect owing to our subsequent use of weights in calculating results. This adjustment has the consequence of making the margin of error substantially larger than it otherwise would be. For the pre-election sample, this results in a MoE of +/- 3.7 percentage points, using a confidence level of 95% and a response distribution of 50% (the most conservative estimate) with the given population size. For the post-election sample, the corresponding MoE is +/- 6.0 percentage points. We note that these figures are overall approximations given sampling stratification and different response options and response patterns for specific questions.

This poll was funded by over 100 individual donors to the Montana State University Alumni Foundation and by Montana State University. If you have questions concerning the survey methods, please contact the HELPS Lab (helpslab@montana.edu).