A new AARP survey by the Harris Poll examined what's driving women 50 and older ahead of next year's elections and found health care on top. The survey also found that older women’s concerns about Trump are eroding, but not upending, his support with Republicans and independents.
Why it matters: As the House of Representatives prepares to impeach the president, the priorities for this group of high-propensity voters are closer to home and different from what their male counterparts care most about.
Health care was their concern, followed by immigration, terrorism/national security, guns and climate change.
- Health care was an even-more pronounced top concern for 47% of rural and small-town women, and 43% of Hispanic women.
- 4 in 10 women say they can't afford to pay for their health care, resulting in consequences such as skipping or rationing treatments and medication and going into debt.
- 4 in 10 women overall — and about half of black and Hispanic women — said they would switch companies for policies more sympathetic to family caregivers.
- Retirement, college affordability, the economy, distrust in national leadership and worries about America's standing in the world also drove their concerns.
- For the age 50+ men surveyed, immigration was a greater concern than health care.
As for the 2020 election, 17% of the overall sample of women (compared with 12% of men) said they weren't yet sure whether they'll vote for Trump, the eventual Democratic nominee or another option.
- 79% of the women who identified as Republicans plan to vote for Trump.
- 44% of women who consider themselves independents said they were undecided.
Methodology: This Harris Poll survey, conducted online Nov. 8–25 for AARP, covered 1,924 women and 1,227 men ages 50 or older. The sample of women included 969 Democrats, 513 Republicans and 395 independents, with the majority of the women in the 50–64 age group. About 4 in 10 identified as AARP members.
The researchers said they are not attributing a margin of error to the findings because it's not a probability-based sample — it's based on people who agreed to participate in the survey.