the period 1990-1994. Data were collected quarterly and covered earnings and
hours worked, both before and after the displacement. Some 15 per cent of
their sample was at some stage enrolled in studies at a community college and
transcripts of their credits attained were available. The estimated effects of
credits equal to a year of studies were positive; 9 per cent for men and 13 per
cent for women. They found no evidence of decreasing returns with age, a re-
sult in line with Leigh and Gill (1997) but contrary to Light (1995) and Monks
(1997). However, compared with Sweden, the US has greater wage differences
and less generous financial conditions for adult students, as well as a wider dis-
persion in knowledge of reading, mathematics etc. These are all factors that
would increase the expected payoff to AE.
4 Data and descriptive statistics
This section describes the data in this study. To start with, there is a short pres-
entation of the raw data before it is divided into two separate samples that are
presented in subsequent sections. The same division of the data is later made in
the empirical section. This is motivated by the fact that studying the effects of
AE is a very different task if individuals go on to further education than if they
do not. For this reason, the first part focuses on those who have not been regis-
tered in higher education. A fraction of these have participated in AE and the
origins of any effects on wage earnings are then relatively straightforward to
interpret. The second part deals only with individuals who at some point up to
2002 had been enrolled in higher education, with or without prior AE.
The register data originate from several sources. Komvuxdatabasen provides
data on adult education at komvux and contains information on course subjects,
credits, interruptions and grades attained. This detailed level of analysis has not
been used in earlier evaluations and represents one of the major contributions
of this report. Data on individual characteristics, transfer payments and wage
earnings from 1990 to 2002 are supplied by Louise (Longitudinal data on edu-
cation, income and employment). Data on higher education is collected from
The Register of Universities and University Colleges ( Universitets- och hog-
skoleregistret). The highest attained education of each individual is reported
each year by Statistics Sweden in their Standard Classification of Education in
Sweden (Svensk utbildningsnomenklatur, SUN). The SUN codes referred to in
10
IFAU - Does adult education at upper secondary level influence annual wage earnings?