The “folk theory” of democracy is the commonly held notion that citizens develop a preference for policy, select politicians that reflect these policy preferences, and therefore have some level of meaningful control over government. Achen and Bartels believe that the folk theory serves as a reassuring vision of how democracy should function, but does not accurately represent the way voters actually act. The authors argue that the average voter has limited political knowledge and that they are usually more influenced by identity and partisan loyalty than informed preferences when casting their votes.

The group theory of politics suggests that the primary determinants of political action are the social groups and collective identities to which individuals belong, as opposed to individual preferences. Achen and Bartels reference the work of Arthur Bentley, arguing that through their involvement with various social groups, individuals gain their beliefs, allegiances, and political views. As a result, group theory provides a basis for the authors’ realist alternatives to classical democratic theory. By focusing on groups rather than individuals, the authors emphasize how groups compete and interact within political institutions and among elites.

Identity-based voting is the process by which individuals cast ballots for political candidates or parties due to common social identities (e.g., race, religion, class, nation), as opposed to casting votes based upon policies. According to Achen and Bartels, identity-based voting is much more prevalent and has a greater impact on vote choice than policy-based decision making. For this reason, the authors portray identity as a fundamental source that affects not only vote choice but also one’s view of politics and interpretation of events.

Partisan identification is a durable psychological commitment to a particular political party that shapes a voter’s political perceptions and behaviors. Through empirical research and case studies, Achen and Bartels demonstrate that partisanship typically precedes policy preferences and shapes what voters perceive when they consider candidate positions on issues. The authors thus suggest that a voter’s allegiance to a particular political party shapes their political opinions, rather than voters choosing candidates based on specific policies.

Retrospective voting refers to the theoretical model in which voters judge incumbents based upon prior performance and reward or punish them accordingly. In contrast to the folk theory of democracy, retrospective voting is viewed by many as a more realistic representation of voter behavior. However, Achen and Bartels assert that the retrospective voting model rests on unrealistic assumptions about voter ability. Consequently, Achen and Bartels demonstrate through analysis that retrospective evaluations made by voters are frequently biased or uninformed, thus limiting its potential as a means of providing democratic accountability.

Selection models of retrospective voting propose that voters can determine whether an incumbent has skill or not based upon observable outcomes. Voters aim to re-elect competent incumbents who are likely to achieve good results in the future. Achen and Bartels argue that this type of selection model assumes that voters are able to differentiate between competency and luck. Since external random factors frequently affect outcomes, distinguishing between skill and luck becomes increasingly difficult for voters.

Sanctioning models of retrospective voting propose that voters award or penalize incumbents to encourage efforts by the incumbent. Achen and Bartels point out that, like selection models, sanctioning models are also limited because randomness and incomplete information reduce the relationship between effort, outcome, and electoral sanctions.

The “specious present” refers to the tendency of voters to focus disproportionately on current events when judging political leaders, illustrating how election outcomes are frequently determined by short-term economic conditions rather than long-term performance. This temporal bias limits the extent to which voters make informed decisions based upon a complete history of governance.