American support for the war with Iran has fallen to 31 percent in a Reuters/Ipsos poll of 1,215 adults conducted over four days, with a margin of error of about three percentage points, Al Jazeera reports.
Support was 37 percent in March and 34 percent earlier in August. President Donald Trump's approval rating is 33 percent, described in the poll as a record low across both of his terms.
Read the movement carefully
A three-point margin of error means the drop from 34 to 31 since earlier this month is at the edge of what this poll can distinguish. The fall from 37 in March is more robust, and the direction has been consistent across three readings.
The clearer signal is inside the president's own coalition. Republican support for the war has gone from 77 percent in March to 69 percent now. A war can be sustained politically on minority public support; it is much harder when the movement is among the voters who are supposed to be reliable.
What the war costs voters directly
Gasoline has risen by more than a dollar a gallon since the fighting began on February 28, when the United States and Israel struck Iran.
That is the mechanism by which a distant war becomes a domestic opinion figure, and it is the reason the House speaker said last week that it "would be helpful to get that thing resolved" before November, while insisting his party can win regardless.
Eighty-three percent of respondents expect the conflict to last a long time, up from 80 percent earlier this month. Expectations of a long war and falling support for it tend to move together: people withdraw backing from something they no longer believe will end soon.
The midterm number, and its limits
Among independents asked how they would vote in the midterms, 33 percent said Democratic and 19 percent Republican.
Those figures do not sum to 100, and the source does not say what the remainder chose or whether undecided voters were offered. A 14-point gap among independents is substantial, but subgroups within a sample of 1,215 carry a wider margin of error than the headline, and we have not been given the subgroup size.
What we could not establish
We could not establish the exact dates of fieldwork, the number of independents sampled, what the remaining independents answered, or how the questions were worded. We verified this account from a single publication, and all figures here are one pollster's.
