Always look for the methodology. This poll is bullshit.
The Reuters/Ipsos poll was conducted online in English across the United States. It gathered responses from 2,440 adults, including 946 Democrats and 846 Republicans.
More Americans blame Trump for government shutdown: Reuters/Ipsos poll | Reuters
Couple things:
1. 946 Democrats means approx 40% responses came from Dems and approx 35% came from 846 Repubs, so what do you suppose the results would be if the responses were 50-50? My guess, somewhat fewer blaming Trump and instead more people blaming the Dems.
2. Which means 648 responses came from people who did not align themselves with either party, but if you extrapolate from those numbers I mentioned then what would the results of that poll be if the ratio of those 648 people was also slanted to the Left? Across the US, huh? Wonder how many calls were made to rural areas and red states. And how random were those phone calls, were they to the same people that indicated Hillary Clinton would win back in 2016?
3. Curiously, the poll questions were not to be found in the Reuters story, why is that? Knowing the way some people like to interpret poll data, one wonders exactly what the poll questions were.
4. Curiously, there was a Quinnipiac poll done a few weeks ago of 1,147 voters found that 43% now support a border wall and 54% oppose it. In August, support was at 38%, thereby reflecting a five-point jump in a relatively short period of time. More importantly for the president, 86% of Republicans and 47% of Independents support building the wall, while 51% of those Independents oppose it. Given the margin of error, that's kinda close. 90% of the Dems oppose the wall, go figure. But it's easy to see how the breakdown of who is polled really can influence the results.
