The biggest problem with polling in 2016 was not flawed methodology or the naively confident aggregation of polling into probabilities of outcomes. The issue was that everyone was so centered on the horse race.
Pollsters and data journalists have been much smarter in 2018 about covering voters’ preferences for policy as they debate how they translate that into vote choice.
Understanding where the people are on policy can help create a more informed electorate. However, it also leads to a more informed elite. For instance, both parties have learned important things about their voters as they head toward election day.
Democrats are learning that people love expanded access to health care, and Republicans are learning that massive corporate tax cuts have extremely limited appeal.
Once they start adapting new methodology, it will only be easier to explore beyond the horse race into more policy and value-frames.