being registered in higher education. Among AE participants, this is 67 per cent
in the group with three years of upper secondary school, 18 per cent among
those with a two-year upper secondary level and 12 per cent among those with
only compulsory school. Those registered at komvux after 1999 are also ex-
cluded to allow for at least two years of “undisturbed” wage earnings post-
komvux. Further, AE participants are restricted to those who have registered
annual wage earnings above SEK 20,000 at least once prior to enrolment. This
is to avoid absurdly high percentage increases in the annual wage earnings fol-
lowing AE. For the comparison groups, the condition is set that there should be
at least four observations of annual wage earnings above SEK 20,000. This is
somewhat arbitrary, but the intention is to exclude individuals who essentially
are without attachment to the labour market. Of the total sample, 49,675 indi-
viduals remain, of which 24.8 per cent at some stage were registered at kom-
vux. In Table A.1 and A.2 in the Appendix, further descriptive statistics are
presented.
Table 3 presents frequencies of AE participants as well as non-participants,
dividing the AE individuals into six groups based on credits accumulated from
1988 to 1999.7 The fraction of AE participants with zero credits represents 24.7
per cent of the total number enrolled in AE, meaning that a binary variable in-
dicating registration in AE would be partly misleading as a proxy of the adult
studies conducted. These individuals interrupted on average 40.7 per cent of
their courses. The remainder either had grades not reported or a reported grade
below three (or from 1993 ‘fail’).
7 AE courses from the schooling year 1992-93 were, as mentioned, underreported and only a
small fraction of these had grades attached. All courses were therefore counted as passed.
IFAU - Does adult education at upper secondary level influence annual wage earnings?
15