Rates and the judgment of government performance

Governments around the world regularly report policy outcomes to citizens in the form of rates, such as rates of crime or disease per 100,000 people. The choice of a base number used in reporting (for example per million, per... [ view full abstract ]

Governments around the world regularly report policy outcomes to citizens in the form of rates, such as rates of crime or disease per 100,000 people. The choice of a base number used in reporting (for example per million, per hundred thousand or per thousand) is a matter of judgment for politicians and public managers. We ask, does the choice of base use for a rate influence citizens’ substantive perceptions of risk and of conditions in society? We focus on real US government statistics regarding violent crime (from the FBI) and infant mortality (from the CDC). In other contexts, it has been shown that people tend to judge the likelihood of an event as greater when presented as a ratio with a large numerator and denominator than when presented as a mathematically equivalent but smaller numerator and denominator. We hypothesize that ratio bias affects citizens’ perceptions of risk and evaluations of societal conditions in the context of real government statistics. We use two experiments (on crime and infant mortality) randomly to allocate participants to treatment groups receiving different versions of the same information, and we also present the information as an absolute number. We find evidence of ratio bias, especially with lower perceived risk and more favorable assessments of conditions where a ratio is presented per inhabitant. The findings suggest that citizens’ substantive judgments about risks and conditions in society can be influenced by the ratios governments choose to use in reporting official statistics.