Many rational choice theories posit that rational decision makers assign subjective values to all available choice options and choose the option with highest subjective value. Choice options are usually composed of multiple attributes, e.g. healthiness and taste in dietary choice or risk and expected returns in financial choice. These attributes have to be integrated into a single subjective value. Subjective value maximizing choice requires choice consistency, i.e. consistent weighing of the choice attributes across choices. However, empirical work suggests that perfect choice consistency is often violated, for example when decision makers weigh choice attributes differently across multiple decisions. Some researchers propose to extend certain bounds of rationality or to abandon the concept of rationality as adherence to consistency principles altogether. A more conservative stance assumes that perfect consistency can be violated by decision makers in practice, but that consistency principles still can explain large parts of behavior. In a review of the recent literature, we identify factors for compromised consistency relative to baseline conditions. Broadly, we distinguish between undynamic trait factors and fluid state factors. We find evidence for an influence of age, education, intelligence, and neurological status. In contrast, choice consistency appears to be relatively robust to the influence of sex, personality traits, cognitive load, sleepiness and blood alcohol levels. We conclude, that, according to the current state of the literature, only fundamental differences in decision makers, that is, trait differences, have a significant impact on choice consistency.