Trump approvals LV, RV, vs All

Amazing difference when you look at who the pollsters are polling.

Let's start with the numbers according to RCP approval composite averages:

• Overall: 47.5 (minus 2.2)

• All Americans: 44.1 (minus 7.4)

• Registered Voters: 47.6 (minus 2.4)

• Likely Voters: 54% (plus 8.7)

So there is not just a little bit of difference here. There is a large and significant difference between polls depending on who is being polled. That being said, the pollsters who are including "all Americans" in their sample include pollsters who have shown an objectively anti-Trump bias. Specifically including Yahoo and Reuters/Ipsos as the only pollsters showing Trump negative by double digits. These are not just an outlier, but well outside of the overall averages.

If you are wondering, Rasmussen, Trafalgar, and Insider Advantage are among the pollsters who are using the likely voter samples. Many will argue that these are "pro-Trump" pollsters; which is to say that they have been more closer to reality than most other pollsters. These are pollsters who actual showed Trump with numbers that corresponded to his election results. So I am going to argue that the likely voter pollsters are more accurate both historically and based on how they are sampling their respondents.

Why do we like likely voter polls? Not only do these polls show better reflection of what we can expect at the voting booth. But at the end of the day, consistent voters are simply more informed than non-voters, and the more informed someone is the more I trust (and care about) their opinion.