Questions on income in surveys are prone to two sources of errors that can cause bias if not addressed adequately at the analysis stage. On the one hand, income is considered sensitive information and response rates on income questions generally tend to be lower than response rates for other...

Applications of multiple imputation have long outgrown the traditional context of dealing with item nonresponse in cross-sectional datasets. Nowadays multiple imputation is also applied to impute missing values in hierarchical datasets, address confidentiality concerns, combine data from...

Using data from 49 European regions covering 2005-2012, this paper finds that the estimated effect of cohort size on employment and unemployment outcomes is very sensitive to the age range of the sample. We argue that this is because the identification strategy commonly used in this literature...