1 Introduction
It is commonly assumed that university students benefit from attending lec-
tures. This assumption, however, needs to be tested, as developments in
information technology are increasingly calling for a reassessment of the
traditional approach to university education, largely based on physical at-
tendance of lectures and classes, and a number of alternative “weightless”
educational models, based on distance learning, are being introduced. Nev-
ertheless, as pointed out by Romer (1993), until recently there was relatively
little evidence about attendance and its effects on student learning.1
In the past decade, a number of studies have examined the relationship
between students’ attendance (or absenteeism) and academic performance,
generally finding that attendance does matter for academic achievement (see
e.g. Durden and Ellis (1995), Devadoss and Foltz (1996), Chan et al. (1997),
Marburger (2001) Rodgers (2001), Bratti and Staffolani (2002), Dolton et
al. (2003), Kirby and McElroy (2003)). This kind of evidence has led some
authors to call for measures to increase student attendance and even to con-
sider the possibility of making attendance mandatory in some undergraduate
2
courses.2
The main problem in assessing the effects of attendance on academic per-
formance is that attendance levels are not exogenous, given that students
choose whether to attend lectures and classes, and that this choice is af-
fected by unobservable individual characteristics, such as ability, effort and
motivation, that are also likely to determine performance: better students,
who are more able, work harder or are more motivated, tend to have higher
attendance levels, other things being equal. This implies that estimates of
the impact of attendance on academic performance are likely to be subject
to omitted variable bias.
Most existing studies either brush aside this problem or attempt to dis-
entangle the impact of attendance on performance from unobservable ability
and motivational factors by including in the set of regressors proxies of ca-
pability (students’ grade-point-averages, scores on college entry exams, etc.),
effort (homework-assignment completion) and motivation (students’ self re-
ported interest in the course). However, such indicators are generally an
imperfect measure of ability and motivation. As a consequence, OLS esti-
1 “Even though teaching is a very large part of what we do, we know very little about
many aspects of instruction and learning” (Romer, 1993, p. 214).
2See Romer (1993) and the following discussion in Brauer et al. (1994).