In a previous article in this journal (Gilley 2006), the political legitimacy of the state was conceptualised for the purposes of empirical research. The conceptualisation was based on the three dimensions of political legitimacy outlined most fully by Beetham (1991) – namely the legality of the state, the moral justification of the state and the consent enjoyed by the state. Using historical evidence as well as convergent data analysis, it was argued that in aggregating these three dimensions, justification should count roughly twice as heavily as each of legality and consent.

The operationalisation of the concept was then executed using data from the late 1990s and early 2000s (c. 2000) for 72 countries, containing within them 83 per cent of the world's population. The data included substitutive or ‘effect’ indicators as well as constitutive or ‘cause’ indicators (Bollen & Lennox 1991). The results were briefly presented and discussed and became the basis of a subsequent book-length treatment of the empirical study of political legitimacy (Gilley 2009). Other scholars have also drawn upon the results as the basis for analysis of social and political issues (Grimes 2008; Rothstein 2010; Hechter 2009).

The results have also given rise to renewed interest in legitimacy measurement. Power and Cyr (2009) validated, replicated and expanded the measurement for 18 Latin American countries c. 2005, and then used the results to inquire into the explanation for variations across those cases. Seligson and Booth (2009) dissented from the conceptually driven measurement in favour of a correlation-driven one, arguing that legitimacy is only legitimacy if it can be shown to correlate with certain observed behaviours. They also argued that ‘political legitimacy’ should address a wider object than only the state, and included ‘political community’ and ‘economic performance’ in their measure. Their legitimacy calculations for eight Latin American countries c. 2004 (Seligson & Booth 2009: 247) were only modestly correlated (r = 0.55) to the same eight countries in the ‘Gilley replication’ measurement of Power and Cyr, showing the implications of the different measurement strategies.