Why is this important?
- This contrasts a bit from how we typically view public opinion stability. Earlier evidence shows greater length of time in between opinion measurement doesn’t increase instability (see Figure 4 here).
- In this CCES panel data, with waves in 2010, 2012, and 2014, opinion stability levels are also fairly similar over two years or four year, but not always. The above graph shows differences for some opinions.
- Differences in stability over different time spans generally 1) cast doubt on the Converse’s (1964) characterization of “nonattitudes” and random responding (under which we’d expect no difference in stability) and 2) support the possibility of genuine attitude change explaining instability levels.