Rajaraman, Indira creator Saha, Debdatta text xu| 2008 continuing ng p.223-33. This article attempts a purely empirical approach to the issue of optimal civil service size by looking for a systematic relationship between observed size and underlying explanators. The residuals from such a fitted relationship then serve as a first approximation to whether the government in question is overstaffed or not. The data are from Indian subnational states, which as members of a single federation, follow the same definitional conventions with respect to the civil service. the contribution of the article is methodological, and addresses an issue of central importance, since staff downsizing conditionalities attached to externally funded structural adjustment programmes are not analytically underpinned. The article also explores issues surrounding the composition of the civil service. There is no evidence that overstaffing, as determined from the residuals of the estimated specification, is the joint outcome along with staff composition, as proxied by mean salary, of job creation pressures at the lower salary scales. This suggests that correction of overstaffing might be possible without political opposition, while at the same time leaving unexplained the variation in staff composition across states. - Reproduced. Civil service Public Administration and Development 180718