Women working for wages: Putting flesh on the bones of a rural labour market survey in Mozambique

Journal of Southern African Studies

Volumn 32, Issue 2, 2006, Pages 313-333

  Sender, John ^a   Oya, Carlos ^b   Cramer, Christopher ^b  

^a UNIVERSITY OF LONDON   (United Kingdom)

^b NONE

Author keywords

[No Author keywords available]

Indexed keywords

LABOR MARKET; POVERTY; RURAL SOCIETY; WAGE; WOMENS EMPLOYMENT;

AFRICA; MOZAMBIQUE; SOUTHERN AFRICA; SUB-SAHARAN AFRICA;

EID: 33745097760     PISSN: 03057070     EISSN: 14653893     Source Type: Journal    
DOI: 10.1080/03057070600656291     Document Type: Article

References (73)

1. 'Research has cast little light on the role of wage labour in rural income strategies; we know of no empirical study that has explicitly focussed on this issue...': D. Tschirley and R. Benfica, 'Smallholder Agriculture, Wage Labour, and Rural Poverty Alleviation in Mozambique: What Does the Evidence Tell Us?' (Maputo, Ministry of Agriculture and Rural Development, Directorate of Economics Research Paper Series, 41, 2000), pp. 2, 7.
2. The International Labour Organisation (ILO) also stresses the lack of reliable statistics on waged agricultural labour: ILO, 'Decent Work in Agriculture' (Geneva, ILO, ILO-Bureau for Workers' Activities Background Paper, International Workers' Symposium on Decent Work in Agriculture), p. 42.
3. The definition of the 'household' included all people who made a contribution quite regularly to the survival of the household. This definition aimed to capture the most important economic relationships between a group of people, rather than focusing on 'residential' criteria, kinship, or eating habits.
4. One of the very few recent surveys covering female rural wage workers in Mozambique also found that a high proportion (38 per cent) of female workers were divorced or widowed. The survey covered workers dismissed by two privatised cashew-processing factories: C. Vijfhuizen, C. Braga, L. Artur and N. Kanji, 'Gender, Markets and Livelihoods in the Context of Globalisation: A Study of the Cashew Sector in Mozambique' (London, IIED, Main Report, Phase 1: Nampula Province, 2003), p. 15.
5. T. Lawson, 'Modern Economics: A Problem and a Solution', in E. Fullbrook (ed.), A Guide to What's Wrong with Economics (London, Anthem Press, 2004).
6. H. White, 'Combining Quantitative and Qualitative Approaches to Poverty Analysis', World Development, 30, 3 (March 2002)
7. A. Bryman, Social Research Methods (Oxford, Oxford University Press, 2001).
8. A total of fourteen life stories were collected in the three provinces.
9. J. Breman and G. Wiradi, Good Times and Bad Times in Rural Java (Leiden, KILTV Press, 2002).
10. Rushed 'hit-and-run' data collection has been a feature of a great deal of NGO work in rural Mozambique:sEe B. Pijnenberg, 'Keeping It Vague: Discourses and Practices of Participation in Rural Mozambique' (PhD thesis, Wageningen, 2004), p. 145.
11. See E. Allina-Pisano, 'Borderlands, Boundaries and the Contours of Colonial Rule: African Labor in Manica District, Mozambique, c.1904-1908', International Journal of African Historical Studies, 36, 1 (2003), pp. 59-82, p. 67.
12. D. Hughes, 'Refugees and Squatters: Immigration and the Politics of Territory on the Zimbabwe - Mozambique Border', Journal of Southern African Studies, 25, 4 (1999).
13. J. Pfeiffer, 'Cash Income, Intrahousehold Cooperative Conflict, and Child Health in Central Mozambique', Medical Anthropology, 22 (2003), p. 102.
14. In the quantitative survey, 78 per cent of those female respondents who had experienced desperate episodes of hardship reported that, in attempting to overcome their difficulties, they were most likely to resort to casual wage labour. The Participatory Poverty Assessment carried out in 2001 confirms that the practice of ganho-ganho was always associated with the very poorest households: MPF, Avaliaĉão Participativa Da Pobreza: Relatorio de Síntese (Maputo, Government of Mozambique, 2001), p. 22.
15. In the social context of the provinces visited most marriages were reported as união marital, which normally corresponds to 'traditional marriages' that are not sanctioned by the civil registry.
16. In the quantitative survey, 67 per cent of all respondents had experienced crop failures over the previous two seasons.
17. Reductions in expenditure on public sector wages clearly have knock-on effects on the real wages of female domestic servants in rural areas. These effects are rarely recognised when assessing the impact of orthodox fiscal policies on the rural poor.
18. Most of the female respondents in the quantitative survey reported that no-one in their households had eaten these high value foods in the last seven days. The proportion of these households never consuming meat was 81 per cent, chicken 70 per cent, eggs 78 per cent, milk 88 per cent and fresh fish 64 per cent.
19. The median wage of the female domestic servants in the quantitative survey was MT150,000.
20. Primary education in Mozambique consists of seven years of schooling divided into two levels: Level 1 (EP1) ends at Grade 5 and Level 2 (EP2), which consists of Grades 6 and 7.
21. Attendance at mother-child clinics in Manica is costly, even when the services are officially free. See R. Chapman, 'Endangering Safe Motherhood in Mozambique: Prenatal Care as Pregnancy Risk', Social Science and Medicine, 57 (2003), p. 365.
22. The chances that a child from a poor rural family will be immunised or treated appropriately for diarrhoea are relatively small: S. Chao and K. Kostermans, 'Improving Health for the Poor in Mozambique: The Fight Continues' (Washington, Africa Region Human Development Working Paper Series, World Bank, 2002), p. 18.
23. The differences between the immunisation rates of poor and wealthy households in Mozambique are very large in comparison to most other sub-Saharan African countries, only exceeded by the gaps in Niger and Chad: See D. Gwatkin and G. Deveshwar-Bahl, 'Immunization Coverage Inequalities: An Overview of Socio-Economic and Gender Differentials in Developing Countries' (http://poverty.worldbank.org/library/view.php?id=8970, in Poverty Net: Recent Papers on Poverty, Health, Nutrition and Population by World Bank Staff, World Bank, September 2001), p. 18.
24. In 2004, the National Director of Planning in the Education Ministry reported that although 731,000 children were enrolled in Grade 1, only 197,000 pupils could enrol in Grade 6. He also said that the average pupil:tEacher ratio was 64:1. See AIM Reports, www.poptel.org.uk/ mozambique-news/newsletter/aim268.html, AIM Reports, 19 January 2004.
25. Two-thirds of the female respondents in the quantitative survey had children younger than ten years old. About 40 per cent of these women reported that their children either had no one to look after them while they were at work, or that they were cared for by an older sibling.
26. The amount of cash required to send a child to primary school in rural Manica is approximately MT5,000 per year for the fees, plus another MT280,000 for a uniform and around MT 40-50,000 to buy schoolbooks etc. With four children at primary school, Tomasinha will have to find the lump sum equivalent of almost six months' wages to meet these costs, if uniforms become compulsory.
27. Adolescent fertility rates (15-19 years) are very much higher in poorer than in wealthier households in Mozambique and probably more than half of rural girls aged between fifteen and nineteen in Mozambique are either pregnant or mothers (Chao and Kostermans, Improving Health, p. 5
28. UNICEF, 'Provincial Data. Mozambique 2004' (http://www.unicef.org/mozambique/est_geral.htm).
29. See also A. Granja, F. Machungo, A. Gomes and S. Bergstrom, 'Adolescent Maternal Mortality in Mozambique', Journal of Adolescent Health, 28 (2001), pp. 303-306.
30. The only state-operated agricultural credit programme in the area, GAPI (Small Investments Support Office), concentrated on lending to rather large businesses and had only approved a total of fifteen loans in 2003, few of them for farming. The average size of these loans was about MT500 million.
31. Mozambique has one of the highest rates of infant and under-five mortality in the world and these rates are very much higher in the poorest households, such as Maninha's, than in richer households: See Chao and Kostermans, Improving Health, p. 5.
32. The risks of post-neonatal and child mortality have also been shown to be significantly higher in Mozambique when, as in Maninha's case, a child's father has not been to school: G. Macassa, G. Ghilagaber, E. Bernhardt, F. Diderichsen and B. Burstrom, 'Inequalities in Child Mortality in Mozambique: Differentials by Parental Socio-Economic Position', Social Science and Medicine, 57 (2003), pp. 2,255-64.
33. Maninha's fears were not unreasonable. Research at Maputo Central Hospital suggests that violent deaths make an important contribution to Mozambique's high maternal mortality rates. See A. Granja, E. Zacarias and S. Bergstrom, 'Violent Deaths: The Hidden Face of Maternal Mortality', British Journal of Obstetrics and Gynaecology, 109 (2002), pp. 5-8.
34. These abuses do not only take place in rural Mozambique. A woman working for wages in an urban garment enterprise had her nose cut off by her husband when she refused to agree to stop working and remain at home: K. Sheldon, Pounders of Grain: A History of Women, Work, and Politics in Mozambique (Portsmouth, NH, Heinemann, 2002), p. 156.
35. The research team interviewed a large number of employers and found that task and time rates vary between very similar employers and from season to season, as well as within seasons and between specific tasks and workers.
36. About 20 per cent of all the female respondents in Manica covered by the quantitative survey had migrated from other districts and/or provinces, and many more described episodes of rural - rural temporary migration.
37. Migration by Mozambicans in search of agricultural wage employment in Zimbabwe allowed some Mozambicans to escape from poverty, as will be suggested below, although others have experienced brutal discrimination and, as in the case of Rita, have little to show for their work across the border: See B. Rutherford, Working on the Margins: Black Workers, White Farmers in Postcolonial Zimbabwe (London, Zed Books, 2001), p. 126
38. Human Rights Watch, 'Fast Track Land Reform in Zimbabwe', Human Rights Watch, 14, 1A (March 2002), pp. 1-44.
39. The formal questionnaire did not attempt to investigate the causes of child deaths and this issue was not pressed during interviews for the life stories.
40. About 95 per cent of female respondents in the quantitative survey lived in households that owned no cows; 77 per cent owned no goats or sheep; and more than 42 per cent did not even own any chickens.
41. In 2003, a very small proportion of all the 30 micro-finance initiatives in Mozambique provided credit for farming (AIM, www.poptel.org.uk/ mozambique-news/newsletter/aim247.html, AIM Reports, 25 February 2004). Less than one per cent of the respondents in the quantitative survey claimed to have received any credit at all from an NGO, a Co-operative, or a Government Office.
42. The literature that judges Mozambique to have achieved a successful privatisation programme does not pay a great deal of attention to those poor rural women who have borne the costs of the disposals of state assets in the 1990s: C. Cramer, 'Privatisation and Adjustment in Mozambique: A "Hospital Pass"?', Journal of Southern African Studies, 27, 1 (March 2001), pp. 79-103.
43. Permanent workers are legally entitled to various benefits, including the right to paid annual holidays. A typical employer strategy is to lay-off casual workers on the 25th of the third month of employment (for three days) and then to offer them a new written casual contract. Large-scale and foreign-owned agribusinesses adopt this practice to prevent female workers demanding permanent status after three months' continuous employment. Amalia had been working as a seasonal worker at Zonue for a period of more than three years without achieving the status of a permanent worker. Less than 4 per cent of all female wage workers in the survey were given paid holidays by their employers, and less than 10 per cent were given paid sick leave or any medical benefits. About 3 per cent had paid maternity leave.
44. This means that she would have had to work for 29 days per month to achieve the legislated minimum monthly wage of MT514,000.
45. The very limited cash earnings that women can achieve through self-employment on their own plots is confirmed in the quantitative survey as well as in another survey in the Sussundenga district: J. Pfeiffer, 'Cash Income', pp. 96, 115, 117.
46. In the month preceding the interview, 35 per cent of the respondents in the quantitative survey had purchased maize, and almost 90 per cent had not managed to sell any of the produce grown on their plots during the past year or only 'very little'. An estimate of the returns to maize production using recommended levels of inputs indicates that, depending on the timing of sales, returns per labour day in maize production can be lower than the local average daily wage rate. See J. Howard, E. Crawford, V. Kelly, M. Demeke and J. Jeje, 'Promoting High-Input Maize Technologies in Africa: The Sasakawa-Global 2000 Experience in Ethiopia and Mozambique', Food Policy, 28 (2003), p. 342.
47. Asset indices are a useful and more robust guide to poverty analysis than poverty lines based on dubious expenditure per capita data. See the references in J. Sender, 'Women's Struggle to Escape Rural Poverty in South Africa', Journal of Agrarian Change, 2, 1 (January 2002), p. 13
48. also D. Gwatkin, S. Rustein, K. Johnson, R. Pande and A. Wagstaff, 'Socio-Economic Differences in Health, Nutrition, and Population in Mozambique' (Washington, HNP/Poverty Thematic Group of the World Bank, May 2000), Annex. A and B.
49. On the under-reporting of intimate partner violence in surveys, see L. Heise, A. Raikes, C. Watts, and A. Zwi, 'Violence Against Women: A Neglected Public Health Issue', Social Science and Medicine, 39 (1994), pp.1,165-1,179
50. and R. Jewkes, J. Levin, and L. Penn-Kekana, 'Gender Inequalities, Intimate Partner Violence and HIV Preventive Practices: Findings of a South African Cross-Sectional Study', Social Science and Medicine, 56 (2003), p. 131.
51. R. Galli, People's Spaces: Land and Governance in Mozambique (Lanham, Maryland, Lexington Books, 2003), pp. 10-11.
52. The persistent tendency to define Mozambique's peasantry as 'a (more or less) homogeneous mass' has been emphasised elsewhere: M. Wuyts, 'The Agrarian Question in Mozambique's Transition and Reconstruction' (Helsinki, WIDER Discussion Paper, 2001/14, 2001), p. 1.
53. The high degree of economic inequality within localities (villages and neighbourhoods) in Mozambique, which accounts for most of the inequality in the country as a whole, is shown in MPF, 'Mapeamento Da Pobreza Em Moĉambique: Desagregaĉão das Estimativas de Pobreza e Desigualdade Aos Niveis de Distrito e Posto Administrativo' (Maputo, Government of Mozambique, 2003), p. 3
54. and C. Elbers, P. Lanjouw, J. Mistiaen, B. Ozler and K. Simler, 'Are Neighbours Equal? Estimating Local Inequality in Three Developing Countries' (Helsinki, WIDER Discussion Paper 2003/52, 2003), pp. 11 and 14.
55. Female literacy was found to be an important determinant of standards of living in Mozambique in a model using 1996-7 IAF data: See K. Simler, S. Mukherjee, G. Data and G. Datt, 'Rebuilding After War: Micro-Level Determinants of Poverty Reduction in Mozambique' (Washington, DC, International Food Policy Research Institute, Research Report 132, 2004), p. 47.
56. S. Handa, 'Raising Primary School Enrolment in Developing Countries: The Relative Importance of Supply and Demand', Journal of Development Economics, 69 (2002), pp. 103-28, p. 110
57. C. Lloyd and P. Hewett, 'Primary Schooling in Sub-Saharan Africa: Recent Trends and Current Challenges' (Washington, DC, World Bank Policy Research Division Working Paper 176, 2003), Appendix, Table 2.
58. A total of 213 female respondents were 20 years old or younger. Of these respondents, 90 had become pregnant as teenagers and had already given birth to one or more children. The median years of education completed by these respondents was three, whereas the median for the remaining 122 female respondents in this age range (who had not given birth as teenagers) was significantly higher - four years of completed education. Female respondents who became pregnant as teenagers lived in households with low possessions scores.
59. Estimates of the factors affecting young children's nutritional status in Mozambique (using data from a national survey) show that maternal education has a positive and significant effect on height-for-age Z-scores: See J. Garrett and M. Ruel, 'Are Determinants of Rural and Urban Food Security and Nutritional Status Different? Some Insights from Mozambique', World Development, 27, 11 (November 1999), pp. 1,955-75.
60. Some of the complexities involved in attempting to establish the underlying causal processes are discussed in J. Sender, 'Rural Poverty and Gender: Analytical Frameworks and Policy Proposals', in H-J. Chang (ed.), Rethinking Development Economics (London, Anthem Press, 2003).
61. Moreover, even open-ended questions concerning the reasons for household members' move to their present homes failed to elicit responses concerning witchcraft. However, as seen in the case of Amalia, witchcraft accusations and threats could be decisive in explaining patterns of mobility. 47 The quantitative survey data indicated that by far the most important channel for obtaining employment was through 'relatives and friends'. This recruitment process has obvious cost and disciplinary advantages for employers in rural labour markets characterised by excess supply. See M.J. Wells, Strawberry Fields: Politics, Class, and Work in California Agriculture (Ithaca, Cornell University Press, 1996), pp. 153, 163. Anita has recently recruited her divorced sister to become the maid/cook for the farm manager.
62. The quantitative survey data indicated that by far the most important channel for obtaining employment was through 'relatives and friends'. This recruitment process has obvious cost and disciplinary advantages for employers in rural labour markets characterised by excess supply. See M.J. Wells, Strawberry Fields: Politics, Class, and Work in California Agriculture (Ithaca, Cornell University Press, 1996), pp. 153, 163. Anita has recently recruited her divorced sister to become the maid/cook for the farm manager.
63. The magnitude of the demand generated by other wage workers for domestic servants in rural areas is rarely discussed in the literature on Mozambique, or in other sub-Saharan African countries. However, this type of labour market multiplier is undoubtedly important, as suggested by the finding that over 100 of the female respondents in the quantitative survey were paying someone to look after their young children while they were at work, either in cash or kind. The quantitative survey excluded numerically very important categories of higherstatus female employees in rural Mozambique, such as teachers, nurses and other state employees; most of these workers employ maids. The number of women working for wages as domestic servants in rural Mozambique is likely to be very large and severely underestimated in all official labour market statistics.
64. 'In Manica province on the border with Zimbabwe, at least 60 white Zimbabwean farmers who lost land in Zimbabwe are opening farms to produce tobacco and flowers for export': J. Hanlon, 'Renewed Land Debate and the "Cargo Cult" in Mozambique', Journal of Southern African Studies, 30, 3 (September 2004), p. 609.
65. Tschirley and Benfica, 'Smallholder Agriculture', p. 4.
66. An earlier article co-authored by Tschirley and Benfica provides some evidence from rural Mozambique to suggest that the poorest tercile of households are more dependent on 'net sales of labour' than the richest tercile of households: H. Marrule, R. Benfica, P. Strasberg. D. Tschirley and M. Weber, 'Reflections on Poverty and Prospects for Growth in the Mozambican Rural Sector' (Maputo, USAID/Mozambique, Policy Synthesis for Cooperating USAID Offices and Country Missions, 44, 1999), Tables 1-3.
67. Non-sampling errors in our own survey included filling in the wrong code for marital status as well as undercounting the number of household members in those households with unusual residential patterns. Although, in theory, large samples may attenuate the significance of sampling errors (which can be calculated) and some nonsampling (measurement) errors, the key question is what we can possibly assume about other measurement errors, their magnitude and randomness: A. Deaton, The Analysis of Household Surveys: A Microeconometric Approach to Development Policy (London, Johns Hopkins University Press, 1997), pp. 99-101; 151.
68. There appears to be 'a considerable degree of measurement error for a number of variables on which data were collected in the IAF survey': Simler et al., 'Rebuilding After War', p. 78.
69. These are defined as employing less than ten workers in the peak season.
70. Sender, 'Rural Poverty'.
71. See D. Boughton, D. Tschirley, B. Zulu, A. Osorio Ofico and H. Marrule, 'Cotton Sector Policies and Performance in Sub-Saharan Africa: Lessons Behind the Numbers in Mozambique and Zambia' (Conference Paper, 25th International Conference of Agricultural Economists, Durban, AEA, 16-22 August, 2003)
72. and NEPAD, Comprehensive Africa Agriculture Development Programme (Rome, FAO, 2003).
73. Simler et al., 'Rebuilding After War', p. 80.