I have a question concerning the microdata of Eurostat SILC. We are newly working the data and noticed that if we calculate percentages (prevalences) for certain variables, we do not get the same results as those that we can access via the downloaded data from the Eurostat database.

For example: The percentage value of HH081=3(no bath or shower) from the microdata should correspond to the value from table ilc_mdho02 from the Eurostat database. But this is not the case, e.g. Romania 2019 Eurostat-DB=22.8 and Microdata=24.46. Taking into account the household weights DB090 does not resolve this issue and still results in a percentage of 23.87.

Does anyone know how exactly the data is processed? We already contacted Eurostat, but that did not really help. We also have the official guideline, how to process the data, but do now get the same results as those shown in the webside. It would be strange to report other values as those on the webside, an we are not sure, whether we need to weight or clean the data in some other way than what we already did.