Via InstaPundit, Friday’s Reuters-Ipsos poll shows significant movement toward President Trump and the GOP, with the president’s approval rating among registered voters at 49% and Democrats +5 on the generic ballot. The Ipsos people caution that this week’s poll is probably an outlier:
This week’s Reuters/Ipsos Core Political release presents something of an outlier of our trend. Every series of polls has the occasional outlier and in our opinion this is one. So, while we are reporting the findings in the interest of transparency, we will not be announcing the start of a new trend until we have more data to validate this pattern.
You can hear the pollsters’ disappointment in this week’s crop of respondents. I’m certainly glad they decided to “[report this week’s] findings in the interest of transparency” rather than suppressing them!
In fact, though, there isn’t any mystery about the results. This week’s poll sampled 579 Republicans, 556 Democrats, and 163 independents. That is quite a few Republicans. In contrast, the Reuters-Ipsos survey of April 25 included 551 Democrats, 484 Republicans, and 157 Independents. Similarly, the April 18 poll included 559 Democrats, 459 Republicans, and 162 Independents. Ipsos says its poll is “weighted to the U.S. current population data by gender, age, education, and ethnicity,” but not, apparently, party affiliation. Naturally, then, President Trump and the GOP do better when the survey sample includes more Republicans.
So this poll may indeed be, as Ipsos claims, an outlier. On the other hand, maybe more people are beginning to identify with the GOP because they don’t want to be associated with a party of vicious haters. Ask me in November.
This illustrates the futility of obsessing over polls–although, of course, many of us can’t help it. The answers you get depend on whom you ask, and the only important question is who will show up at the only polling places that matter.