rather than through planned reform. Studies have found even the poorest
parents have exercised the exit option, moving their children en masse from
free government schools to the private sector. In a private school teachers are
accountable to the manager (the headmaster who can fire them) and through
him to the parents (who can withdraw their children). In a government
school, controversially, the chain of accountability is much weaker, as teach-
ers have a permanent job with salaries and promotions not related to perform-
ance. There are concerns that private schools are inequitable because they
charge high fees so will exclude the poorest and that private schools are of
low quality. Analysis of fieldwork on private education for the poor in Ghana,
Nigeria and India has revealed some surprising results. Schools in selected
parts of these countries were visited to check facilities and the activities of
teacher and pupils were given a standardized ability test. The research also
collected data on background variables of pupils such as household income,
years of parental education, caste/tribe, religion and parental motivation. The
results found that though the proportion varied, private schools enrolled a
clear majority of children, ranging from 60 to 75 per cent of the total. There
was no enrolment difference by gender. Fees were affordable: in India, for
example, for the 4th grade they averaged around 3–4 per cent of the monthly
wage for a breadwinner on the minimum wage and in Africa around 12–20
per cent of the minimum wage. The majority of these schools also offered
free or reduced rate places to children. In Hyderabad, India around 75 per
cent of private schools offered free or concessionary places amounting to
nearly one-fifth of all private school places. In both countries private primary
school teachers were more often teaching and absenteeism was lower than in
their government counterparts. In Hyderabad more than 90 per cent of teach-
ers were teaching in private schools compared with 75 per cent in government
schools. In each of the studies private schools consistently achieved higher
test results than government schools. In Hyderabad mean scores in mathe-
matics were about 22 percentage points higher in private schools than in
government schools and the advantage was more pronounced for English
(not unexpected as private schools were more likely to be English-medium).
In Lagos State, Nigeria the mean advantage over government schools was
between 14 and 19 percentage points in maths and between 22 and 29
percentage points in English. Assuming that children of wealthier and more
educated parents are more likely to attend private school we need to make an
adjustment to these raw scores to find the value added by private education.
After controlling for these effects the differences were reduced, though still
large and in favour of private schools. Finally, private schools provided more
cost-effective education. Average salaries were more than three times higher
in government than private schools and in all studies class sizes were found
to be smaller in private schools. The results showed that the per-pupil teacher
cost is nearly two and a half times higher in gov ernment schools (Tooley and
Dixon, 2006).
130 Sources of Growth in the Modern World Economy since 1950