One of the most visible services performed by public opinion research is to measure the standings of public officials. Most often these polls take the form of an ‘approve’ or ‘disapprove’ of the way in which public officials handle their jobs. The most likely to be rated is the President of the United States, but other officials in federal and state governments often are also.
Presidential approval ratings often correlate with economic performance, integrity, race, and external factors such as war and terrorist attacks. Some commentators argue that the ratings respond to short-term forces such as speeches or scandals. However, others have argued that presidential popularity follows a pattern, which means the level of approval is the product of pattern of “expectation/disillusionment cycles”. Consequently, the rating may not be as responsive to short-term forces as some people might think.
Researchers generally measure presidential approval by asking something similar to: “Do you approve or disapprove of the way [the current president] is handling their job?” In addition to asking about whether or not the respondent approves or disapproves, the question often measure the intensity of the opinion. An approval score is created by adding up the proportion of respondents who “strongly” or “somewhat” approve. This is the score that we see reported more often in media; the changes in the number are often taken as evidence of some reaction by the public to what the president or other officials are doing.
The standard score, while an important measure of the standing of a public official, is not the only way to look at approval ratings. Simply adding up the approvals overlooks other information that could provide insights into the rating of the official. There is usually nuance in the data, and the most popular way of interpreting it neglects information that could be used to understand more fully how the public views these officials.
“Simply looking at approval and ignoring disapproval skews the perceptions of the popularity of public officials.”
For example, using data from the October Utah Political Trends Panel, Governor Gary Herbert has the highest public approval, and Representative Rob Bishop has the lowest. But what about disapproval? The person with the lowest level of disapproval is Lt. Governor Spencer Cox. The official with the highest level of disapproval is President Trump. Simply looking at approval and ignoring disapproval skews the perceptions of the popularity of public officials. Another factor to consider is the proportion of respondents who answered “don’t know.” People in this category may lack an opinion about an official or could be indifferent toward the official. Either way, this information provides a more comprehensive understanding of just how popular the official is.
Simply aggregating the disapproval and ignoring the approval makes the same mistake as adding up the approvals. To avoid skewing perceptions, we also look at what we call a “Net Approval Rating” that looks at the difference between the total amount of approval and total amount of disapproval^2. Using this measure to look at institutions rather than individuals shows that the lowest net scores are reserved for both the Democrats and Republicans in Congress. Research in political science indicates that citizens do not like the rancor of politics, so it should come as no surprise that these two groups score so low on this measure.
It is also possible to create a ratio, which converts the data into a relational measure. We can also find the intensity of opinion by subtracting the percentage of those who “strongly disapprove” from the percentage of those who “strongly approve”.
While each unique way of conceptualizing approval provides a certain type of knowledge about the standing of the official, visualizing the extent to which the unique measures may cluster provides and overall assessment of the standing of the official. The graph can also give some insight into the way each measure captures something about the concept of approval, providing a reliability check.
The figure displays how several officials cluster toward the top, an indicator that the officials really are all doing well. Some are in the middle, and at the bottom are those officials who consistently seem to score lower on the approval dimensions. This means they have work to do. The data does not indicate why the officials are in these locations, just that in the eyes of the public they are doing well or not so well.
A one-size-fits-all assessment of favorability leaves out other information. The analysis in this blog makes a case for looking at the other information in order to build a more complete understanding of what the public thinks about their officials.