Who is counted? The (in)different selection of population figures for measuring malapportionment

Abstract Who is counted? To realize the “one person, one vote” principle and measure malapportionment, practitioners and political scientists have selected a population figure, or population count, arbitrarily or indifferently . Population figures include total population, eligible voters, registered voters, cast ballots, and valid votes. Focusing on population figures for measuring malapportionment, the paper demonstrates that the (in)different selection of the population figure matters theoretically and empirically. Normatively, who is counted brings us to mind who is included in demos. Thus, selecting different population figures mirrors different conceptualizations of the norm of representative democracy. Based on these conceptualizations, our formalization derives the conditions and policy responses causing differences in malapportionment calculated using different population figures. Furthermore, using the most popular database, we reanalyze existing findings to investigate whether they remain robust. The reanalysis simulates the data-selection process by generating $$5000\times 4$$ models sampled from $$2^{28}$$ models using a malapportionment index while (in)differently choosing total population or registered voters as the figure. Our finding suggests that the probability of obtaining the same results as the previous research is like flipping a coin. Therefore, we cannot remain indifferent to the (in)different selection of the population figures.