MANIFESTATIONS OF GENDER INEQUALITY IN EDUCATION: MALAWI. ZAMBIA AND ZIMBABWE.

Introduction

One of the most enduring types of educational inequality is that of gender. At a global level, the gender gap in education has been reduced significantly in many of the countries of the North although it remains extreme in parts of the South, particularly in South Asia and Sub-Saharan Africa (SSA). South Asia has the largest gender gap at both the primary and secondary levels followed by SSA.
In the poorest countries of the world, gender inequality is reflected in lower enrolment, attainment and achievement, and higher wastage rates for girls. While SSA has the lowest levels of education as a region, it has, nonetheless, made the most progress in increasing schooling for girls and women over the past three decades (UNESCO, 1993). Thus, the enrolment of girls has increased at a faster rate than that of boys although starting from a much lower base level. This is largely the consequence of the expansionary education policies followed by post-independence African governments. Even so, gender inequalities with respect to enrolment levels and educational outcomes are still very marked both in absolute terms and in relation to other developing countries. During the 1980s, two-thirds of primary school-aged African children who were out of school were female (Colclough, 1994). While it is true that the enrolment gap between boys and girls has diminished in many SSA countries at the primary level, the education of women and girls remains highly inequitable, particularly at the tertiary level.
Generally speaking, economic recession and structural adjustment policies (SAPs) have over the past decade deepened social and gender inequalities in African societies. Widespread concerns have been expressed about the adverse impacts of SAPs on education sectors mainly in terms of declining overall enrolments for both males and females and deteriorating quality of provision at all levels. Due to the multiple political, social and economic constraints operating against females in SSA, deteriorating economic conditions have had a particularly damaging effect on the education and employment prospects of women and girls (Elson, 1994).

The importance of female participation in education

The very considerable private and social benefits arising from the education of girls are now well documented (see, for example, Floro and Wolf, 1990, King and Hill, 1991, Herz et al, 1991). The level of women's education affects economic productivity, child health and welfare, and influences the length of time daughters are sent to school. As schooling tends to improve the mother's knowledge and use of health practices, each additional year of schooling is estimated to decrease the child mortality rate (Hartnett and Heneveld, 1994). Research reported in two World Development reports (See IBRD 1989 and 1992) confirms that female education reduces fertility, especially where family planning services are available. Female education is linked with later marriage, lower fertility, desire for smaller families, and increased practice of contraception. The relationship is stronger as women's education increases (see Herz et al, 1991).
There also seems to be a strong link in rural areas of SSA between education of women and agricultural productivity although this has not been adequately researched. In the rural areas of Ghana (where currently three-quarters of female farmers have no education), lack of female education limits farm productivity. Inadequate literacy and numeracy skills are also reported to affect adversely the efficiency of women traders (BRIDGE, Ghana, 1994). Another study on Ghana finds that maternal education is the main influence on children's schooling, whether for boys or girls. But the effect on girls is twice as great - with respect to both girls' actual enrolment and the probability that girls will continue to the next grade (Herz et al, 1991).
The positive impact of maternal influence on schooling has been found right across the developing world in different cultural settings. Studies conducted in Latin America indicate that parent's income and own years of schooling have a strong positive impact on children's education with the mothers' educational level appearing to have a greater impact (Bustillo, 1989). An ILO study on vocational training in Zambia (Kane, 1990) also shows that well educated women are more likely to educate their daughters.
The influence of mothers on their children's education is particularly important in the African context where in many countries a relatively high proportion of households have a female head. This includes many of the countries of Southern Africa where male migration is widespread. Many women are the prime movers with respect to their children's education and their own levels of education and command over resources are important factors in their ability to send children to school (Fuller et al, 1994). In rural Botswana where 48% of households have a female head, it has been found that investment in a daughter's schooling is higher for those households. A recent household survey in Botswana also found that the mother's literacy levels and reading practices were closely related to their daughters' level of school attainment (ibid).
Research has shown that both the social and private rates of return to girls' primary education are generally high. For this reason, King and Hill (1993) recommend that governments should invest in basic education for girls in both formal and non-formal settings. Recent World Bank studies have suggested that returns to secondary education for girls and boys are now comparable to those for primary (Herz et al, 1991). However, rates of return estimates for SSA countries should be treated with caution since the data and/or the methodology used in many country studies are seriously flawed (see Bennell, 1995).
Due to the undoubtedly positive impacts of female education, a somewhat instrumentalist view has emerged from the World Bank and some other aid donors regarding gender and education. Primary education is now seen as a particularly 'good investment' both generally and from a gender perspective. However, the factors militating against the education of African girls cover political, social and economic factors. While it is important that the economic and social benefits are highlighted, the key issue remains one of equity, namely that education must be recognised as a basic human right. Furthermore, it is important that women are not merely the 'objects' of policy but rather that they themselves become part of the process of transformation of not only education but social attitudes. Thus, investing in girl's education cannot be a 'quick fix' solution to economic and social problems, since any kind of progressive change will need to involve a long term process of political empowerment as well.

Educational expenditure patterns

Any analysis of enrolment patterns must be placed in the context of government spending on education, and the differential spending between primary, secondary and tertiary educational levels. As elsewhere in SSA, the education policies of the three countries under consideration have been severely affected by financial stringency in the 1990s.
Primary: In 1990, Zimbabwe spent approximately eight times more (in US dollar terms) on each primary pupil than either Zambia or Malawi (see Table 1). Zimbabwe's expenditure per pupil was at its highest in 1980, fell in 1985 but then rose again in 1990 (although not reaching the 1980 figure). However, Zimbabwe since 1990-1991 has seen a reduction in real per capita government expenditure of 32% while government spending on secondary and higher education has fallen but not as fast (Chisvo, 1994). Malawi's expenditure per primary pupil fell slightly during the 1980s and was only $15 in 1990. Since 1991, however, Malawi has been involved in a major expansion of the primary education sub-sector which renders this disparity less extreme. Due to deep financial cuts, spending on primary pupils in Zambia fell consistently throughout the 1980s, so that by 1990 it spent only $16 per pupil, about half the 1985 figure and the same as Malawi. Primary education in Zambia has borne the brunt of recent reductions in education expenditure, while spending on tertiary education has been maintained and even increased (Kelly, 1994).
Secondary: While the differences between the three countries with respect to public expenditure per secondary school student are not so marked as at the primary level, they are still relatively large. In 1990, Malawi spent $41 and Zambia $95 per student, far less than the $233 per pupil in Zimbabwe. Moreover, the drop in spending per secondary school pupil between 1985 and 1990 was not nearly so drastic in Zimbabwe as it was in Malawi and Zambia (see Table 1).
Tertiary: In Zimbabwe, real expenditure per tertiary-level student almost doubled between 1970 and 1980, but during the 1980s, expenditure per student fell by more than half, so that by 1990 it had the lowest per capita tertiary expenditure of the three countries, the highest being Malawi at $1 782 per student. It should, however, be borne in mind that Malawi has a much smaller proportion of the relevant age cohort attending tertiary institutions than Zambia and Zimbabwe. In both Zambia and Zimbabwe, as we have seen, the cutbacks in expenditure on the tertiary sector have not been severe. In Zimbabwe, real per capita spending on higher education declined by only 4% between 1990/91 and 1993/4 (Chisvo, 1994). This seems curious given the prioritisation of many aid donors, (notably the World Bank) of primary education while pruning expenditure at the tertiary level.
Total government expenditure on education: Educational expenditure expressed as a percentage of GNP was nearly four times higher in Zimbabwe in 1990 than in Malawi and Zambia. Zimbabwe's expenditure on education as a proportion of both government expenditure and GNP rose consistently between 1970 and 1990. On the other hand, public expenditure on education in Malawi as a percentage of total government expenditure and GNP has remained consistently low. Zambia has experienced the largest fall in education as a proportion of government expenditure and GNP between 1985 and 1990 (see Table 1a). Real spending on education stands at less than half what it was in 1984 and by 1994 education accounted for only 2.5% of the GNP (Kelly, 1994)

Enrolments

Primary: Both primary and secondary schooling have expanded phenomenally since independence in Zimbabwe, which consequently has high gross enrolment ratios (GER). In contrast, GERs in Malawi for both primary and secondary education (at 69% and 3.9% respectively) were the lowest for 1990 (see Table 2). However, primary school enrolments in Malawi are now expanding very rapidly as a result of the new policies introduced by the democratic government (which was elected in May 1994). In particular, primary education was made free in October 1994. Between July and September 1994, over one million new children were pre-registered and 19,000 new teachers recruited. In October 1994, with over three million children estimated to be in primary school, the GER was 132% (38% of the projected enrolment was in Standard 1 and from all ages). The new policy increases access to education but poses severe problems of quality delivery, particularly in the light of a substantial shortage of fully trained teachers and classrooms (ODA, Malawi Primary Community Schools Project, 1994). In a way, this situation is a replay of the conditions facing the education sector in Zambia and Zimbabwe after their rapid expansion of primary and secondary schooling in the 1970s and 1980s respectively.
The problems of access to primary education have largely been overcome in Zimbabwe. Each of the three countries now has nearly equality of overall enrolments between girls and boys at primary school (see Table 3). By 1992/3, girls in Malawi made up 51% of enrolments in Standard 1 but in Standard 8 their share had dropped to 35%, mainly because of relatively high drop out rates among girls in the lower grades of primary (GABLE, 1995). In the senior grades of primary school, there is a greater propensity for boys to repeat years in order to increase their chance of gaining entry to secondary school (Zeitlyn, 1994). However, Malawi admits a much smaller proportion of students of the relevant cohort to secondary school and the average transition rate from Standard 8 to Form 1 is approximately 8% (in 1992) with the transition rate being higher for girls than boys although boys do have higher scores than girls on the Primary School Leaving Exam (Hyde and Kadzamira, 1994).
Primary and secondary school enrolments also grew rapidly in Zambia during the 1980s but this expansion was not properly resourced. As a result, the quality of educational provision declined dramatically. For example, in the early 1990s there were only 20,000 classrooms for 1.5 million primary school pupils and some large schools ran three or four sessions a day (IBRD, 1992). The gender gap in primary enrolment does seem to have narrowed in Zambia in recent years, with girls as a proportion of the total number increasing from 47% in 1987 to 49% in 1992. Similarly, improvements in girls' enrolment were made in Standard 7 where the proportion of girls rose from 43% in 1987 to 48% by 1992 (Maimbolwa-Sinyangwe, 1994).
The rather favourable overall primary school enrolment ratios for all three countries at the national level mask severe regional inequalities in both the quality and quantity of educational provision. Available evidence suggests that early drop-out and wastage is considerably higher in remote rural areas than in urban and peri-urban areas. In the large scale farming areas of Zimbabwe, due to various logistical and household constraints, the drop out rates for children are three times the national average for eight year olds and five times the national average for 11 year olds. Worse still, the rate of drop out for girls is estimated to be on average three times greater than that of boys overall the age categories (Nyagura, 1995). These particularly high drop out rates are due mainly to the strong demand for child labour to support households.
There are also large differences between rural and urban areas in Zambia. In rural areas the GER is 69% while in urban areas it is 100.8%. In both rural and urban areas, however, the net enrolment rates for girls are almost identical with those of boys (Kelly, 1994). In the provinces with the highest levels of poverty: Eastern, Northern, North-West and Western have the lowest GER and the highest drop out rates (UNICEFb, 1995).
Regional variations in enrolment are also significant in Malawi; in the northern region in 1990/91, 80.8% of eligible children attended primary school whereas in the less developed south only 45.5% attended (Hyde, 1992). Policy makers concerned with equity of provision and gender inequalities need to focus their attention on these disadvantaged rural areas.
In Zambia and Zimbabwe, much higher proportions of students are admitted to secondary school than in Malawi, with GERs for 1990 of 21.3%, 50.0%, and 3.9% respectively. The primary drop out rates in both Zimbabwe and Zambia are only a little higher for girls than boys. For example, in Zimbabwe the primary drop out rate for the 1985-1992 cohort was 28% for boys and 29% for girls, although the proportion of Grade 7 enrolment moving on to Form I was 73% for boys but only 64% for girls in 1992 (UNICEF, 1994). A similar level of drop out of girls occurs between '0' level and 'A' level. In Zimbabwe, it 'would appear that primary school completion rates are deteriorating for both sexes but to a greater extent for girls (UNICEF, 1994). Ministry of Education Statistics show that through grades 1-7 more girls are dropping out than boys (UNICEFb, 1995).
Secondary: The secondary and tertiary education sectors in Malawi are relatively very small. Only 31.000 students were enrolled at secondary schools in 1990 compared with 672,000 in Zimbabwe and 195,000 in Zambia (see Table 3a). Secondary education GERs were 3.9% in Malawi and 50.0% in Zimbabwe in 1990. Despite the rapid expansion of provision, primary education is still considered to be terminal for most Malawians.
Table 3a shows that in both Malawi and Zambia, the proportion of girls attending secondary school increased steadily during the 1980s. In Zimbabwe, this proportion fell from 42% in 1980 to 40% in 1985 but rose again to 46% in 1990. The rate of expansion of secondary schooling in Zimbabwe during the 1980s decade was phenomenal; primary school enrolments in Zimbabwe increased three times while secondary school enrolments increased tenfold. This expansion as well as the abolition of tuition fees at primary level benefitted both girls and boys (Gordon, 1993).
In Malawi, the average transition rate from Standard 8 (primary) to Form 1 is currently 8%. In Malawi's mixed secondary schools, boys outnumber girls by approximately three to one (secondary school places are reserved for girls). Approximately the same ratios prevail in Zimbabwe and Zambia. Even though there was a slight improvement in the proportion of girls in secondary grades 8 to 12 in Zambia between 1980 and 1989 (i.e. from 35.1% to 38.7%), the drop in the proportion of girls attending grades 8 (41.2%) and 12 (33.9%) in 1989 remains significant (IBRD, 1992, Annex 1a). For those Zambian girls who do manage to secure places at the secondary level, most of them are channelled into non-technical disciplines and traditionally female subjects such as home economics and social sciences. Boys, on the other hand. are encouraged to study technical subjects similar to those offered in the boys only technical high schools which existed up until 1993 (Republic of Zambia, 1995). As indicated earlier, the overall figures for primary and secondary enrolment disguise the progressively lower participation of girls in the upper grades.
Tertiary: The tertiary education sector in Zimbabwe is the largest with 49,400 students in 1990. Enrolments grew particularly rapidly between 1980 and 1985. In contrast, in Malawi and Zambia, enrolments stagnated in the early 1980s but increased, especially between 1985 and 1990 (see Table 4). However, during the 1990s, enrolments in all three countries have remained fairly static given the financial stringency facing governments and the prioritisation of primary education by donors.
Despite the fact that university entry requirements for girls are lower than for boys in Zimbabwe and Zambia, access to tertiary education is even more limited for girls than at the secondary level. In 1990, the proportion of female university students were as follows: Malawi (24%), Zambia (28%), and Zimbabwe (27%) (see Table 4). In the 1992/93 academic year, female students accounted for only 19.3% of total enrolment (Kelly, 1994). However, overall enrolment figures at tertiary mask the high concentrations of female students in teaching, arts and humanities. In 1990, females comprised only 24% of science students in Malawi and 8% in Zambia (DAE, 1994). In Zimbabwe, this figure was an incredible 1% although by 1993 female student enrolments in science had risen to 20% (Sibanda, CAMFED Conference, 1995). The particularly low enrolments of women in science subjects is the consequence of a gendered subject specialisation as early as primary and lower secondary school. This means that by the end of secondary few girls have the requisite entry qualifications to study science or vocational subjects such as engineering or medicine.
What is particularly apparent is that the gender gap in enrolments between boys and girls becomes progressively more pronounced from the end of primary up to tertiary in all three countries. Although very little data are available for the past few years, it seems likely that SAPs have had a negative impact on the numbers enrolled at primary, secondary and tertiary sectors (particularly in rural areas) in both Zambia and Zimbabwe. It is clear that primary GERs in Zimbabwe and more particularly, Zambia dropped between 1985 and 1990 (DAE, 1994 and Kelly. 1994). More research is certainly needed in this area. In the light of present financial stringency, educational expansion has slowed considerably in Zimbabwe and more particularly, in Zambia. Malawi could also find its current expansion programme curtailed as a consequence of its present financial difficulties.

Performance and attainment

The mean educational attainment of girls in SSA is low because enrolment is low compared with boys and wastage is high (Hyde, 1993). While it is well known that girls' educational attainments are, in aggregate terms, considerably lower than for boys in SSA, this is not easy to document for many African countries. Adequate records of school-based performance by sex are not kept on a consistent basis for all schools in many countries. However, some research has been conducted on exam performance in Zimbabwe, Malawi and Zambia. Enrolment figures by grade can also give an indication of the stage at which girls drop out of the school system although it is not possible with the present data to determine the reasons for their departure.
Wastage includes both grade repetitions and drop outs. The reasons for wastage are complex and will be explored more fully in the next section. Statistical information about drop outs and exam performance can help illustrate the problem of gender inequality. Detailed classroom based research is needed to illuminate the precise nature of the attainment difficulties faced by girls.
In all three countries, girls perform worse in all examined subjects except local languages and, in the case of Zambia, English. Primary school leaving exams are a way of screening out those with weaker academic grades. Malawi, Zambia and Zimbabwe all have leaving exams at the end of primary and the pattern of girls' performance is consistently worse that of boys in most subjects. A situation which is then repeated at the end of secondary school for those girls who have made it that far.
Zimbabwe: Up until 1993, drop out rates for girls in every age cohort were higher than for boys (Gordon. 1993). Since then, however, the drop out rates for the primary grades have been similar for both girls and boys (UNICEF, 1994). In 1991, the Ministry of Education in collaboration with the International Institute of Education Planning (IIEP) found no difference in achievement levels between boys and girls at the primary school level (MEC, 1994). However, in the same year, Riddell and Nyagura carried out a study of academic achievement among boys and girls in secondary school and found that girls' achievement was significantly lower than boys, particularly for maths. They recommended that more observational studies were needed in order address this deep seated problem through improved teaching practices (Nyagura and Riddell. 1991).
In Zimbabwe, primary school completion rates declined for both boys and girls between 1985 and 1993, although girls tended to drop out in larger numbers than boys at all the transition points in the school system (UNICEF, 1994). The poorer performance of girls through the school system is reflected in '0' level results from Zimbabwean secondary schools. It should be pointed out that these examinations are the potential passport to the job market or higher education and are, therefore, ferociously competitive. The overall '0' level pass rate in Zimbabwe has declined as school enrolments have increased. In 1986 and 1987, 15% of boys but only 7% of girls passed five or more subjects to earn a Cambridge School Certificate. The performance of boys in individual subjects was better than that of girls with the exception of three areas: the vernacular language, physics and chemistry. However, girls did not do well in maths and other science subjects (Dorsey, 1989). It is clear that since 1980 girls have consistently performed worse than boys in the '0' level exams (Gordon, 1993).
In a sample of secondary schools in Zimbabwe, Bennell compared different groups of children in terms of their performance in the '0' level examinations and their job market experience. He observed that among the 1985 group of Form IV school leavers, 31% of the males and 15% of the females passed more than 5 '0' levels whereas the corresponding pass rates for the corresponding 1988 group were 19.9% and 8.6% respectively. This decline in performance for both sexes was put down to the deteriorating quality of provision (Bennell, 1994). Due to the importance attached to certification, a high proportion of students retake their '0' levels (between 15-20% of the school leavers surveyed by Bennell). In 1993, only 5% of girls as opposed to 16% of boys obtained 5 passes or more (UNICEF, 1994). Thus', it would appear that the performance of girls in the crucial '0' level exam seems to have deteriorated still further vis-a-vis boys during the 1990s.
In the November 1992 '0' level exams, there were 103,753 male candidates and 76,678 female candidates (42.5% of the total). Apart from Shona and Ndebele, in all subjects taken by more than 2000 candidates, boys did significantly better that girls. For most subjects, the male pass rate was, in fact, double or three times the rate for girls. These disparities in examination performance are puzzling because the girls who reach Form 4 are more highly selected than comparable boys (IBRD 1992a). Possible reasons for this poor performance are the negative attitudes of teachers and the hostile learning environment in secondary schools for girls.
Depressingly, Bennell and Ncube's research on Zimbabwean school leavers using two cohorts in the 1980s, shows that even among the best qualified school leavers with 5 or more '0' levels, 71.8% of the 1985 male school leavers were in wage employment by early 1991 compared to only 52.9% of their female counterparts (Bennell and Ncube, 1994). Gender segregated job markets and prejudice against women entering traditionally male preserves make access to well paid employment difficult even for those girls who do overcome the hurdle of formal schooling. Labour market prospects undoubtedly influence the decision of parents to send their daughters to school and to keep them there. According to recent research, girls who leave the education system are much less likely than boys to find their way back into either formal or non-formal education systems (UNICEF, 1994).
Malawi: A similar pattern of poor achievement and low persistence for girls can be found in both primary and secondary schools in Malawi. Throughout the school system, girl's performance lags behind that of boys. Girls are more likely to repeat a grade in primary school (except in Standard 8 where the reverse is true) and pass rates for girls are lower than those of boys. In the primary school leaving examinations, girls do worse than boys in all subjects examined (Kadzamira, 1994).
Several studies in the 1980s on sex differences in academic achievement and exam performance (at primary and junior secondary level) have revealed that sex differences in maths and science exist from primary school onwards. Boys outperformed girls in the Malawi School Certificate of Education exam in geography, history and English although girls outperformed boys in Bible Knowledge and Chichewa (Kadzamira. 1987 and 1988). Furthermore, fewer girls than boys obtained passes with distinctions and credits in English, maths and physical science.
There is evidence to suggest that girl's academic performance has not improved in Malawi since the early 1980s. in fact, for maths and science the gap between girls and boys' examination performance actually widened during the 1980s (Mwanza, 1990). Despite fluctuations, the difference between the pass rates of girls and boys at the JCE exam increased from 24.2% in 1985 to 28.3% in 1989 (Hyde, 1992). Some studies have attempted to analyse exam results by type of school i.e. single sex or mixed. The results usually indicate that girls in single sex schools generally perform better than girls in mixed schools (Kadzamira, 1987 and Hiddleston, 1991). Although nearly half of the girls taking MSCE exams attended mixed secondary schools, over 70% of the girls entering Chancellor College come from single sex schools. Despite this improvement in performance for girls in single sex schools, their pass rates are still 10% lower than boys at single sex schools and 5% lower than boys in mixed schools (Hyde, 1992).
Girls' performance in maths and science is particularly problematic in both primary and secondary schools. However, Hiddleston's study undertaken in 1991 of undergraduate girls from single sex schools taking maths and science at Chancellor College (University of Malawi) indicates that girls moved from being below average in their first year to above average in their fourth year even though they had poorer scores on entering college (Hiddleston, 1994). Similar observations have been made in Kenya. This change in the pattern of girls' performance in maths and science is significant and the learning environment at university should be closely examined. More research is needed generally on how pedagogical practices affect girls and women in different contexts.
Zambia: There appears to have been little systematic research on learning achievements in primary schools in Zambia and apparently the country does not have any good indicators of learning achievement (Kelly, 1994). In all of the six papers of the primary school leaving exam the mean scores for boys are higher than those of girls and the difference is greatest in social sciences and maths (ILO, 1990). At Grade 9 the overall picture is of weak performance for both girls and boys although there is a significant gender gap with only one-third of girls getting a full certificate compared with half of all boys. The performance of girls in the end of primary exams has deteriorated in recent years whereas the performance of boys remains unchanged (Kelly, 1994). Very similar patterns of poor female retention and performance would appear to exist in Zambia as in Zimbabwe and Malawi. Thus, girls fail more frequently and do worse than boys in every exam at primary and secondary level except for English literature in the school leaving exam (ILO, 1990)
A higher proportion of girls drop out at all levels than boys, although the drop out rate at the higher grades is more pronounced. Information on the progression of one cohort from 1978-1989 shows a sharp fall off for both boys and girls after grade 7, with the gap between boys and girls widening from grade two onwards (IBRD, 1992). Of those who enter grade 1, only 11% of boys and 3.8% of girls complete grade 12 which is the end of the secondary cycle (Ibid). The highest repetition rates occur at standard 9 where pupils take an examination to qualify for further studies at higher grades. Repetition rates at secondary level ranged from 2% in the Central Province to 6.7% in the Western Province in 1994 (Republic of Zambia, 1995). For all the years between 1980 and 1992, completion rates for girls are lower than those for boys, sometimes by more than 25 percentage points (Kelly, 1994).
A major programme of gender research in Zambia that includes surveys of the factors contributing to gender inequalities in education is currently underway as part of the World Bank's ongoing Education Rehabilitation Project (IBRD, 1992). This should be of considerable assistance in pinpointing more precisely the reasons for girl's low attainment throughout the schooling system.

Literacy

A key indicator of the success of past educational policies and practices is the rate of literacy for the population as a whole and women in particular. Zimbabwe and Zambia present a very similar picture: overall literacy rates in 1990 were 66.9% in Zimbabwe and 72.8% in Zambia and female literacy rates were 60.3% and 65.3% respectively (UNDP, 1993, UNESCO, 1993). Kelly stresses that the relatively high adult literacy rate in Zambia is a considerable achievement given the level of under-resourcing of the schools over the last two decades (although he cautions that the female literacy rate of 72% for 1990 from the Census could well be an over-estimate (Kelly, 1994)). In contrast to Zambia and Zimbabwe, the literacy rates in Malawi are among the lowest in SSA with a very large gender gap: the female literacy rate was estimated at 17.7% in 1990, less than half the rate of males (see table 5). These figures for Malawi are an indication of the very low rates of education spending in the past and consequent inadequacy of both formal and non-formal education provision. They highlight the need not only for the building up of primary education but also the development of an efficient, community based adult education system.
As in the case of enrolment ratios, there are huge differences in literacy rates in all three countries between urban and rural, with the more remote rural areas registering much higher rates of illiteracy. In Zambia, for example, according to the 1990 Census, the percentage of persons aged 1 5 and above who cannot read and write is 55% and 21.5% for rural and urban females and 32.5% and 9.2% for rural and urban males respectively. The disparity between male and female is larger in rural than urban areas which reflects the better male access to schooling and employment opportunities.