“…In the welfare state retrenchment discussion, the differences between using information on social expenditures and replacement rates are often referred to as a ‘ dependent variable problem ’ (see, e.g., Clasen & Siegel, 2007; Green‐Pedersen, 2004). Nonetheless, in the existing literature which addresses explicitly pension generosity issues, replacement rate seems to be the key indicator reflecting the analysed phenomenon (see, e.g., Kuitto et al, 2021; Sanchez‐Romero & Prskawetz, 2020; Zohlnhöfer et al, 2013), although it is also not free from flaws (Wagner & Elder, 2020; Wenzelburger et al, 2013).…”