Several reasons have been assigned for the rampant absenteeism among teachers in primary schools � a problem that plagues not just India but most developing countries. These include the complete absence of any monitoring mechanism, the disincentivising effect of poor infrastructural facilities and poor compensation, particularly at the primary level and lack of intrinsic motivation. Policies to tackle absenteeism in schools have, therefore, tended to revolve round improving infrastructure, compensation packages etc.
However, an empirical analysis based on a theoretical model developed by this researcher showed that the decision to attend or not is in fact a joint decision taken by the teacher and the student, which implies that the decision of one affected the other. Thus there exists a simultaneity between teachers� decision to attend or not and decision of the students. A teacher would not like to come to an empty class and a student prefers a less truant teacher. This finding has significant implications for policy making.
We employ Becker�s marriage market framework and propose that the teachers and students produce a �shared good� when they attend classes. A shared good, in this context, can be thought of as the benefit of a more educated society, which accrues to the society as a whole. This shared good is produced only when both the teacher and her student is present. A utility maximisation exercise is performed by the teacher and student given a set of constraints where the utility functions depend on the shared good, among other things. At equilibrium, we show that teacher�s and student�s attendances move in the same direction. Subsequently, we also establish the same result using a game theoretic approach. Here, the teacher and the student play a guessing game under full information where one tries to guess whether the other is going to attend class or not and subsequently makes one�s own decision. We find a continuum of Nash Equilibria, where attendance of one is equal to the attendance of the other. Each of the teachers and students thus stands to lose if they choose an attendance level different from the other.
We verified the above results empirically on the basis of a survey of primary schools in the North-West Frontier Province in Pakistan conducted by the World Bank. We estimated the following two equations simultaneously using seemingly unrelated regression.
is the teacher�s attendance,
is the child�s attendance
is the predicted attendance of the teacher obtained from the exogenous variables of the teacher�s equation only
is the predicted attendance of the children obtained from the exogenous variables of the child only,
and are the vector of exogenous variables and the disturbance of the teacher�s equation and
and are the vector of exogenous variables and the disturbance term of the child�s equation.
We find that the coefficients of the predicted attendances are positive and significant at the 5 per cent level for both equations indicating that the attendance of one is directly proportional to that of the other. Thus, the decision to attend or not is a simultaneous one. We also observe that is significantly influenced when the predicted attendances are included in the respective equations. It increases from 0.0775 to 0.2795 in the teacher�s equation and 0.0583 to 0.2479 in the child�s equation when predicted attendance is included among the repressors, implying that it plays a significant role in determining teachers� and students� attendance.
Our study shows that teacher absenteeism should not be treated in isolation. Policies should target both the groups simultaneously and not each group separately. In this regard, one of the greater focus areas to enhance attendance should be mandatory and active participation of parents in school management. That the active participation of parents in school management markedly improved attendance of the teachers as well as students was borne out in an experiment run by World Bank in Nicaragua. Involving local participation in school management may, therefore, prove an important factor in tackling absenteeism in schools than policies and perhaps needs to be encouraged along with measures such as improving infrastructure and compensation packages.