1.Resigned satisfaction (RS)
We used four items from the Short Form of the Work Satisfaction Questionnaire This resigned satisfaction scale has a coefficient of reliability of .684. Moreover, following a model generating procedure of CFA, this measure of resigned satisfaction has a relatively good level of reliability (Chi2 = 30.625; df = 2; p-value = .000; TLI = .935; CFI =.987; NFI = .987;RMSEA = .064). The standardized factor loadings are above the usual threshold (0.51–0.70).
2.Red tape (RT)
Many different operationalizations of red tape compete in the relevant literature. For the purposes of this research, we focused on individualistic perceptions of red tape gleaned from previous research .According to the literature, ‘red tape is best accounted for by the meaning ascribed among those who interact with, and are affected, by rules and procedures within an organization. Red tape or procedural constraints were measured using three items adapted from Wright’s study, which assessed the extent to which respondents felt constrained by organizational rules and management practices. This scale of red tape has a coefficient of reliability of .663.
3.Public service motivation
Measures of PSM are varied and their inconsistency is one of the ‘big issues’ in PSM research. Even if its multidimensionality were set from the beginning most of the studies use a collapsed measure of PSM, summing up all PSM items in one single measure. Following the example of Bright and Vandenabeele. we decided to use a separate measure for each of the four dimensions. A set of 11 items inspired by Perry (1996) was used. Two items form the ‘attraction to policy-making’ (PSM_POL, alpha = .915) and the ‘self-sacrifice’ (PSM_SS, alpha = .577) dimensions; four form the ‘commitment to the public interest’ (PSM_CIV, alpha = .646) dimension and three form the ‘compassion’ (PSM_COMP, alpha = .679) dimension. Cronbach coefficients were relatively low, but they mostly stayedwithin the range of other studies .Summative indices were created for each dimension for the purpose of correlation and regression analysis.