(Chaudhury et al., 2006). Higher-ranking staff such as headmasters and
doctors were more often absent than lower-ranking ones. The bigger problem,
though, was the lack of accountability among teachers/medical staff to other
groups (parents and pupils/health service users). Many developing countries
have highly centralized and formalized government systems for recruiting
teachers and health workers. Such workers are typically unionized and so
politically influential. Recruitment, salaries and promotion are largely deter-
mined by educational qualifications and seniority, with little scope for
performance-based pay. Disciplinary action is rare and teachers and doctors
almost never fired. The main sanction is transfer to an undesirable location,
though one study found that only in 18 of 3,000 cases was a teacher reported to
have been transferred for repeated absence (Chaudhury et al.,2006).
One possible solution is to improve the technical ability to monitor teach-
ers combined with attendance incentives. The NGO Seva Mandir imple-
mented a teacher incentive programme in government schools in rural
villages of Rajasthan, India. Before the study in August 2003 teacher absence
was about 35 per cent. In September Seva Mandir gave 57 randomly selected
schools a camera and instructions for one student to take a picture of the
teacher and other students at the start and close of the school day. The cameras
had a tamper-proof date and time function. Each teacher was then paid a
bonus for attendance exceeding a specified number of days per month. In 56
comparison schools,teachers were paid a fixed rate for the month. The
programme resulted in immediate and long-lasting (four years later) improve-
ments in teacher attendance rates. Teacher attendance was tracked over 30
months; programme schools had 21 per cent absence, compared with 42 per
cent in comparison schools. This experiment showed that teachers are respon-
sive to financial incentives. There was strong evidence that once in school,
teachers also increased teaching time. Classroom contact time increased by
an estimated 30 per cent in the treatment schools. Pupils in the treatment
schools were also more likely to later switch to a government primary school
which required them to pass a learning competency test. This was evidence
that the quality (as well as quantity) of education had increased (Duflo et al.,
2012). Despite the evident success of the programme there is doubt about the
extent to which this policy can be replicated. Government teachers tend to be
politically powerful and would be unlikely to agree to be subject to such close
monitoring. An example in the section on health (see p. 142) of a similar
programme applied to nurses in Rajasthan shows how the scheme was under-
mined by politics.
Empowering parents and communities
An alternative is to give greater control to potential beneficiaries (maybe
parents in the case of education or the local community in the case of health
services). There are two pre-conditions for such a scheme to work. First,
beneficiaries must demand the service so they have an incentive to monitor
128 Sources of Growth in the Modern World Economy since 1950