The Role Of Health Care In The 2020 Election

KFF is tracking the role health care is playing in voters’ decisions throughout the 2020 primary and general elections. The latest KFF Health Tracking Poll finds health care among the top issues for all voters as well as the crucial group of voters who have not yet made up their minds about who to vote for in 2020. Throughout the Democratic primary, KFF is also analyzing state-level data from AP VoteCast, a survey of primary voters and caucus-goers in the Democratic primary contests conducted for seven days, concluding as the polls closed in each state.

As of March 18, 2020, health care was either the top issue, or among the top issues, for Democratic primary voters and caucus-goers in 17 states that have held their primary contests where there is AP VoteCast data available. Using the interactive below, you can find out which candidate won the most delegates in each state, the share of voters who said health care is the most important issue facing the country, as well as the share of Democratic voters who favor a single-payer health plan, similar to the one being proposed by front-runner Sen. Sanders, or a proposal, similar to the one being proposed by former Vice President Biden, in which all Americans would have the option of having a government health insurance plan.

.

Methodology of AP VoteCast

AP VoteCast is a survey of American voters conducted by NORC at the University of Chicago. Interviews begin six days before the day of the primary election or caucus in each state, and are conducted until polls close in the state. Interviews are conducted both online and by telephone (landline and cell) in English and Spanish. All states include a probability-based sample of voters drawn from Catalist LLC’s registered voter database and matched to a registered voter database maintained by L2, which provides additional phone numbers for voter records. For some states, the probability sample is supplemented with online interviews with self-identified registered voters selected from non-probability online panels managed by Lucid or Dynata

Weighting involved multiple stages and are done separately for the probability sample and the nonprobability sample (when applicable). First, the probability-base sample is weighted to adjust for disproportional nonresponse. Then these weights are adjusted to population totals of registered voters in each state using a combination of the 2018 CPS Voter Supplement, the 2018 Census Bureau’s ACS, and the Catalist voter file. For the nonprobability sample, the respondents receive a calibration weight to ensure the nonprobability sample is similar to the probability-based sample on key demographic variables such as ideology. All respondents are then weighted to improve estimates for sub-state geographic regions using a small area model. Finally, the survey results are weighted to the actual vote count following the completion of the election.

Numbers of interviews in each state and margins of sampling error (adjusted for design effects) are shown in the table below. For subgroups, the margin of sampling is higher.

More details about the AP VoteCast methodology can be found here.

Table 1: Sample sizes and margins of sampling error
Total sample Total sample MOSE Non-probability sample
Alabama 1,193 +/- 4 percentage points N
Arizona 2,105 +/- 4 percentage points N
California 4,023 +/- 3 percentage points Y
Colorado 3,006 +/- 3 percentage points N
Florida 3,412 +/- 3 percentage points Y
Illinois 2,738 +/- 3 percentage points Y
Iowa 3,036 +/- 3 percentage points N
Massachusetts 3,085 +/- 3 percentage points Y
Michigan 2,460 +/- 3 percentage points Y
Minnesota 1,337 +/- 5 percentage points N
Mississippi 1,091 +/- 4 percentage points N
Missouri 1,891 +/- 4 percentage points N
New Hampshire 3,111 +/- 3 percentage points N
North Carolina 2,706 +/- 4 percentage points Y
South Carolina 1,499 +/- 4 percentage points N
Texas 3,268 +/- 4 percentage points Y
Virginia 2,604 +/- 3 percentage points Y