Tracking Public and Individual Opinion of the Affordable Care Act

Understanding the evolution of public and individual opinion during the roll out of the ACA, particularly as it depends on how individuals are affected by the reform, will shed light on the success of this pivotal change in the health care marketplace.  Furthermore, tracking changes in opinions over time contributes to our understanding of how individuals incorporate learning into their preferences and opinions.

We use a series of surveys using the American Life Panel.  Each month respondents are asked to report their opinion of the ACA and of the effects on themselves and the country.  Respondents are randomized into four groups, with the first group receiving the survey in the first week of the month, the second in the second week, etc.  Surveys will be conducted from November, 2013 through April, 2014, with a baseline collected in September, 2013 and a follow up to be conducted in July, 2014. 

Methodologically, this series of surveys differs from other surveys of public opinion of health reform in several important ways.  Most importantly, we use a panel design with the same respondents being surveyed each month.  Other surveys use a repeated cross section design.  This allows us to track how opinions evolve and identify precisely which individuals are changing their minds.  In particular we look at how individual and state characteristics affect the evolution of public opinion.  For example, we consider differences as they relate to state policy, such as the expansion of Medicaid or the use of the federal marketplace website. Problems with the federal website are likely to have a larger effect on individuals from states using the federal website than from states using their own website.  We also consider how individual opinions change as individuals enroll in health insurance for 2014, use health care, and gain experience with the reform. As previously uninsured individuals gain access to insurance their opinions may become more positive, while those whose previous coverage is lost may become negative. Finally, we compare the stability of individuals’ opinions about the ACA to the stability of the voting intentions during the 2012 election.  This allows us to investigate whether stability of opinions is associated with other observed characteristics, including demographic characteristics as well as income and education levels and will help to shed light on how changing experiences contribute to learning.

Preliminary results suggest that while overall public opinion was stable from September to November, individuals’ opinions fluctuated greatly; suggesting that stable means masked the true level of change.  Furthermore, the old and the more educated are less likely to change their minds. Without panel data, it would not be possible to identify precisely what characteristics are most associated with changing opinion.