Political approval is a survey based measurement of whether people approve or disapprove of the performance, conduct or handling of responsibilities by a political actor or institution. The object being evaluated can be a political leader, government, president, prime minister, parliament, party, mayor or another political institution, depending on the wording of the question. An approval measure is therefore not inherently a measurement of voting intention.
A person can approve of a leader's performance while intending to vote for another party, or disapprove of a government while continuing to support the governing party. The precise meaning depends on the question wording, the political office or institution being evaluated and the population surveyed. Approval is consequently a specific attitudinal construct rather than a general synonym for political support.
Approval ratings are usually generated through survey questions offering respondents some form of approve or disapprove response, sometimes with additional categories such as strongly approve, somewhat approve, somewhat disapprove and strongly disapprove. Some surveys also include a don't know, no opinion or neither category. The treatment of these responses affects the published percentages. For example, an approval percentage might be calculated among all respondents or among respondents who provide a substantive answer, depending on the pollster's methodology. Comparisons between approval ratings should therefore examine the exact question and denominator rather than assuming that all approval percentages have identical definitions. AAPOR emphasises the importance of question wording, response options and methodological transparency when interpreting survey results.
Political approval is different from favourability, popularity, voting intention and electoral support. Favourability commonly asks whether respondents have a favourable or unfavourable view of a person or organisation, whereas approval normally asks whether respondents approve of performance or conduct. Voting intention asks what respondents say they would choose in an election, while approval asks for an evaluation of performance or handling of responsibilities.
Party identification measures an individual's political affiliation or identification and is not necessarily equivalent to approval of a party's current performance. These concepts can correlate, but they measure different things. A political leader can have relatively stable personal favourability while experiencing changes in approval of current performance, and a party can retain electoral support among people who express criticism of particular government actions.
Approval ratings are time sensitive because evaluations can change in response to political events, policy decisions, economic conditions, crises, media coverage, leadership changes and other developments. A poll conducted immediately after a major event can produce different responses from one conducted before it. This does not necessarily mean that one survey is inaccurate; the population's opinions may genuinely have changed, or differences may reflect sampling variation and methodological factors.
AAPOR recommends paying attention to when interviews were conducted relative to recent events because political attitudes can be affected by circumstances occurring around the fieldwork period. Approval ratings should therefore always be interpreted as measurements taken during a particular period rather than as permanent characteristics of a political actor.
Approval ratings can also vary substantially between population groups. Pollsters may examine approval by age, education, geography, political identification, gender, socioeconomic characteristics or other variables included in the questionnaire. Such subgroup comparisons can reveal different evaluations across sections of the population, but subgroup estimates are based on fewer observations and therefore generally carry more uncertainty than the overall result. AAPOR specifically warns that sampling error is larger for subgroup estimates and recommends identifying the number of respondents in the subgroup when such findings are reported. Consequently, an apparent difference between two small groups should not automatically be interpreted as a precise or stable difference in public opinion.
Approval ratings can be expressed in several related ways. The approval percentage is the proportion approving under the pollster's stated calculation, while the disapproval percentage represents those expressing the opposite evaluation. Some analysts also calculate net approval by subtracting the disapproval percentage from the approval percentage. Net approval is a derived statistic rather than a separate direct response, and its interpretation depends on how approval and disapproval were measured.
If a substantial proportion of respondents expresses no opinion, the net figure can obscure the size of that group unless the underlying percentages are also reported. Comparing net approval between polls requires attention to the questionnaire, denominator, fieldwork dates and treatment of undecided or non substantive responses.
Approval polling has methodological limitations because evaluations are sensitive to wording and context. Asking whether someone approves of a leader's overall performance can produce a different result from asking whether they approve of the leader's handling of a particular issue. The order of preceding questions can also influence how respondents interpret an approval question. AAPOR identifies wording and question sequence as factors that can affect survey results, making it important to examine the exact instrument before comparing ratings.
Approval is also subject to the general problems affecting surveys, including sampling error, nonresponse, coverage, self selection and weighting. A reported approval percentage is therefore an estimate produced by a particular measurement process rather than an objective numerical property possessed independently by the political actor.
Political approval should not be interpreted automatically as a prediction of electoral behaviour. An approval rating describes how respondents evaluate performance or conduct at the time of measurement, whereas an election involves choices among candidates or parties under particular electoral rules. A person may approve of a government's handling of one area while intending to vote against its governing party, or may disapprove of a leader but still prefer that leader's party to the alternatives.
Approval can be an important indicator of public sentiment and can be tracked over time, but its relationship with future voting depends on the political system, competing choices, campaign circumstances, turnout and many other factors. Polling organisations therefore treat approval and voting intention as separate measures with distinct questions and interpretations.
