A recent dispute in political science raises issues about the objectivity of measures of democracy. Political scientists Little and Meng argue that democracy indices using country experts as coders show a greater degree of democratic backsliding than measures that are objective. They worry that this discrepancy may reflect coder bias. I distinguish three aspects of objectivity and offer a reconceptualization of objectivity as coherence objectivity. I argue that coherence objectivity is better suited for evaluating measures of social science concepts like democracy than the understanding of objectivity implicit in Little and Meng’s discussion.