Sentiment and expressive responding in economic data

There has been an ongoing issue with the difference between what American consumers are feeling about the economy versus what is showing up in the economic numbers. By the numbers, the US economy is doing well. Ask the consumers and you will get a different story; a very different story. The view from consumers when you get into specifics is very wrong as seen from this FT survey.

What is driving this large divergence? There is now the argument that consumers are engaging in expressive responding. Expressive responding has been noted in political science research. Some respondents will provide false answers as an expression of political views to strengthen their point of view. For example, there is a large difference between Republican and Democrat consumers. Why is this happening now? And, why is this just happening with US economic data? Is expressive responding misinformation? How do you identify this behavior in a survey? Should economic surveys ask about politics and then eliminate any partisan participants. 

How much weight should you now place on survey data? Unfortunately, survey information or sentiment plays a key role in forming market expectations. This is not going to be solved in the near-term.