I have been looking at the GHS2001client1 data. There
are income variables there such as grearn grearn1 grearn2. Grearn is supposed
to be gross weekly earnings from the labour market. Out of the 7591 respondents
with valid responses, only 2.4% have weekly earnings less than 2000, 10% have
weekly earnings less than or equal to 6200, but if you look at grearn1 and
grearn2, the figures are much more sensible. Could you tell me what is going
on with grearn and whether we can use it at all? What exactly, is it measuring?
The reason that GREARN appears to
have strange values is because it is recorded in pence not pounds. GREARN
1 and 2 are more sensible because they are recoded into pounds (or at least
the value labels are in pounds). You can check this by doing a tabulation
of GREARN by GREARN1 or GREARN2. For example, the case which is recorded as
8089.74 at GREARN is in category 2 (50.01-100.00) at GREARN1. Which means
that 8089.74 is actually 80.8974 pounds (i.e. 8089.74 pence /100). The variable
labels in the latest version of the file held at the UKDA reflect this (not
sure if older versions do). GREARN is labelled 'Gross weekly income (pence
per week)', whereas GREARN1 and GREARN2 are labelled '#'.
