Why no standard sample weigths in the LISS panel

We do not standardly deliver sample weigths for LISS data, for the moment. The reason is, firstly, that for paneldata there are several possible target populations to weight to (for example: the population at the present, the population at the time of the start of the panel, or weights to compensate attrition or nonresponse) and, secondly, due to the way the LISS panel is used, there are also different possible variables to take into the weights. See http://www.lissdata.nl/dataarchive/study_units/view/3 for the LISS panel Background variables.
 

The choice of weighting variables is strongly dependent on the topic and purpose of a specific study (and hence different every month, in the case of the LISS panel). Hence, there is no single, standard way to construct weighting variables for LISS data. We can calculate them a la carte, in which case you would have to indicate in detail for which variables you would like to weight and to which target population. This is not, however, cost-free.
It is always possible to calculate them yourself, using population statistics from Statistics Netherlands (see http://statline.cbs.nl/statweb/?LA=en ).
Finally, we closely monitor the composition of the LISS panel and recruit refreshment samples regularly. In combination with the fact that the panel is based on a true probability sample, this should assure the representativity of the panel.
 
Note: A working paper of Lynn and Kaminska (ISER, University of Essex, UK), titled: “Criteria for developing non-response weight adjustments for secondary users of complex longitudinal surveys”, describes the complexity of constructing weights in panel studies. You can obtain the paper from the authors or by contacting a.c.scherpenzeel@uvt.nl.