Feminist Knowledge | Other Resources

Sketching a Methodology for Assessing Gender Equity under Adjustment Policies in sub-Saharan Africa. With an introduction to the case study of Madagascar

By Milasoa Cherel-Robson

Acknowledgements

Writing this paper has been made possible by the 2000 Associateship Programme, funded by the Rockefeller Foundation at the African Gender Institute in Cape Town. I am grateful to the AGI staff for their support, to Professor Amina Mama for our enlightening discussions and to my fellow associates for the interesting exchanges that we had, and the wonderful sisterhood that we have shared over the three months that we spent together in Cape Town. I also wish to thank Professor Nicoli Natrass for her discussion of an earlier draft of the paper at the AGI conference on Gender Equity. And finally, I owe special thanks to Yann, for being a very understanding and supportive partner.

List of Acronyms and Abbreviations

ADP Agricultural Development Programme

CGE Computable General Equilibrium

DECRG Development Economics Research Group

DfID Department for International Development

DHS Demographic and Health Survey

EPZ Export Processing Zone

ERP Economic Reform Programme

ESAP Economic Structural Adjustment Programme

EU European Union

FHH Female-headed Households

FOFIFA Foibem-pirenena momba ny Fikarohana Ampiharina amin'ny Fampandrosoana ny eny Ambanivohitra

GAD Gender And Development

GATT General Agreement on Tariffs and Trade

GDP Gross Domestic Product

GEM Gender Empowerment Measure

GNP Gross National Product

HDI Human Development Index

H-O Hecksher Ohlin

HPI Human Poverty Index

IDASA Institute for Democracy in South Africa

IFPRI International Food Policy Research Institute

IMF International Monetary Fund

INSTAT Institut National de la Statistique

LSMS Living Standard Measurement Study

MFA Multi-Fibre Arrangement

NGOs Non-governmental Organisations

NTAE Non-traditional Agricultural Export

OECD Organisation for Economic Co-operation and Development

PRA Participatory Rural Appraisal

SAPs Structural Adjustment Programmes

SDA Social Dimensions of Adjustment

SDF Social Dimensions Funds

SLF Sustainable Livelihood Framework

SPA Service Provision Assessment

SSA Sub-Saharan Africa

UNDP United Nations Development Programme

UNIFEM United Nations Development Fund for Women

USA United States of America

USAID United States Agency for International Development

WB World Bank

WBI Women’s Budget Initiative

WID Women In Development

WTO World Trade Organisation

List of Tables

Table 1: Change in women's share of seats in national parliaments in sub-Saharan Africa,1987–2000

Table 2: Expected outcomes from SAPs

Table 3: Prevalence of violence against women by an intimate partner

Table 4: Visits byAgricultural Extension Agents (percentage of households visited)

Table 5: List of surveys in Africa

Table 6: Contextual analysis

Table 7: Description of the status of women

Table 8: Trends, shocks and seasonality

Table 9: Processes

Table 10: List of events of interest under the period of study

Table 11: Description of the LSMS questionnaire

Table 12: Description of the three household surveys

Table 13: Description of the three community surveys

List of Boxes

Box 1: Gender and Growth: Missed Potential

Box 2: The Living Standards Measurement Study

Box 3: Demographic and Health Surveys

Box 4: Other categories of surveys

Box 5: Tools for gender-sensitive budgets

Box 6: The Women’s Budget Initiative in South Africa

List of Figures

Figure 1: A pyramid of poverty concepts

Figure 2: Sustainable livelihood framework

Figure 3: The asset pentagon

Introduction

Gender inequalities are still pervasive throughout the globe. Statistics show that in no region of the developing world do women experience equality with men in terms of legal, social and economic rights. Many studies have also shown that gender disparities in rights, access to productive resources and participation in public life all have negative impacts on development. Countries that discriminate on the basis of gender have more poverty, slower economic growth, weaker governance and a lower quality of life. To corroborate this assertion of a negative link between gender inequalities and development, sub-Saharan Africa (SSA) is both one of the poorest regions in the world and one where women still face constraints that derive from so-called “cultural” attitudes[1] and religious values. For example, in September 1999, the Supreme Court in Zimbabwe ruled that women were inherently inferior to men, citing “the nature of African society” as the basis for their decision (The Guardian, 25/9/99).[2]

At the beginning of the 1980s, following a decade-long recession, most African countries embarked on World Bank (WB) and International Monetary Fund (IMF) financed structural adjustment programmes (SAPs). The controversy surrounding the appropriateness and the efficiency of these policies in the African economic context spurred a series of studies aimed at analysing their effects on living standards. A concern for the impact of SAPs on women appeared at the same time. Put simply, most of these studies concluded that adjustment programmes generally had negative effects on women. However, the evidence presented was patchy and hence easily dismissed by sceptics, thus making the evidence on the effects of SAPs on women inconclusive.

In the early 1990s, following Elson’s innovative work on male bias in development and in orthodox macroeconomic analysis (1991), feminist economists challenged not only the effects of adjustment policies, but also the assumptions on which they were based. In 1995, this culminated in the publication of a special issue of World Development on gender and macroeconomics. This issue contains key landmark studies on applied feminist thinking to economic models (Çagatay, Elson, and Grown, 1995). It is believed that research findings, the lobbying of NGOs and the overall pressure from the development world have led to changes in the public position of the World Bank on gender and economic reform. Gender concerns are now integrated into poverty assessments. Although this represents a major step forward, feminist economists continue to argue that gender should be included at the policy-making stage.

The purpose of this paper is to attempt to design a methodology for assessing the linkages between gender equity and SAPs. In doing so, it will unpack the issues surrounding gender analysis and Structural Adjustment Policies/Programmes in SSA. It argues that without a truly gendered conceptual framework, SAPs will result only in maintaining or even worsening the status quo regarding gender inequality in SSA. The paper also introduces the case study of Madagascar as an example of the application of the methodology.

Section 2 presents definitions of the terms used, as well as a descriptive overview of SAPs in SSA. Section 3 follows with a discussion of gender analysis, gender relations and gender and economics. Section 4 provides an overview of the previous studies on the effects of SAPs on women in SSA. Section 5 lists the type of data available. Linking SAPs to gendered outcomes necessitates the use of a conceptual framework, indicators of well-being, and specific methods for analysing the linkages between macroeconomic policies and outcomes at the micro level. Section 6 explores some of the different possibilities. The biggest challenge here is to bring gender into the quantitative analysis. Section 7 introduces the case study of Madagascar and Section 8 concludes.

Definitions and Descriptions

This section defines and describes the history of the main concepts used in this paper, namely SAPs, gender analysis and methodology.

General description of Structural Adjustment Policies (SAPs)

The examination and exact listing of policies implemented as part of the adjustment package is the first step of any study looking at the effects of SAPs on gender equity. However, it is also important to note that there can be a gap between what was intended according to policy documents and what happens in practice. Details of policy measures can vary between countries (White and Levy, 1999). This sub-section presents the general components of SAPs.

At the beginning of the 1980s, over 30 African countries adopted the WB-IMF supported SAPs. In most countries, the main purpose of SAPs was to transform state-controlled economies into market-based economies. SAPs usually have two components. In the first stage, stabilisation measures are implemented followed by adjustment policies in the second stage, The main objective of stabilisation policies is to cut aggregate demand in order to reduce the balance of payment deficit, the government budget deficit and the rate of inflation. Once the finances of the economy are put in order, adjustment policies are put in place in order to improve resource allocation, increase economic efficiency and create conditions for sustainable growth.

Stabilisation policies are also called expenditure-reducing policies or expenditure-switching policies. The expenditure-reducing policies are concerned with contracting monetary and fiscal stance in the domestic market. They usually involve increasing tax, reducing public expenditure and raising interest rates. The expenditure-switching policies are best known for their exchange rate depreciation component. They are aimed at switching expenditure from foreign to domestic sources, in order to stimulate domestic production rather than imports.

In spite of the possibility of sequencing problems, structural adjustment policies are put in place in the second stage of the programme. Their main objective is to influence the pattern of growth by improving economic efficiency and the competitiveness of the economy. The main policies are trade liberalisation on both domestic and international sides, foreign exchange liberalisation, financial market liberalisation and privatisation. More specifically, the details of each policy are as follows. International trade liberalisation involves reducing export and import taxes, whereas domestic market liberalisation includes the removal of price controls and the reduction of subsidies on food and fertiliser. Foreign exchange liberalisation eliminates exchange controls, whereas financial market liberalisation translates into the elimination of interest rate ceilings and the withdrawal of credit subsidies. The adjustment phase can also include economic and social infrastructure policies.

Stabilisation and adjustment programmes can also be grouped under the word “reform.” Reform is aimed at transforming the organisational structure of the economy, as well as its administrative structure and its political culture. It requires civil service reform, the installation of transparency, rule of law and, more generally, good governance. In fact, reform deals with the long-term economic and social transformation of the country.

Faced with the proliferation of studies on the social costs of adjustment initiated by Cornia et al (1987), donors such as the World Bank first maintained that most of these studies were unrepresentative or based on data of dubious quality. However, criticism of SAPs became so fierce that a group of donors led by the Bank was forced to recognise the necessity of assessing the social costs of adjustment and, at the end of the 1980s, began to finance a series of household surveys in Africa. These multi-purpose household surveys typically gathered information on household composition, expenditure, and other characteristics. Data was collected at the individual level for economic activities, education, health and sometimes time-use. Sample sizes varied from about 3 500 households in Mauritania, to roughly 18 000 households in Zambia.

These surveys have been long awaited. Referring to them, Sparr comments that: “Research projects of this scope and complexity, particularly those planned by the World Bank's Social Dimensions of Adjustment Unit, take a few years before they are fully functional and some preliminary results are known. They are very much needed and will take us into a new, much more sophisticated stage of understanding” (1994:20). She also adds that “Many of the early reports tend to be anecdotal and broadly correlate information without applying statistical techniques to separate out and account for other influential factors.” The time has come to reassess the effects of SAPs based on these datasets.

In fact, various studies have already looked at changes in living standards in SSA based on the availability of these new datasets. However, most of them simply ignore gender dimensions, at best including a section on “Female-headed Households” (FHH). It is well known that (1) the so-called FHH are heterogeneous. This category includes such diverse groups as small-scale farmers, recipients of remittances and traders. (2) Gender inequity is also present within “Male-headed Households”. This implies that it is also important to examine intra-household dynamics. Moreover, the first string of gender analysis has come from inequalities within Male-headed Households.

Gender analysis

Following Esther Boserup’s pioneering work on the key role of women in development (1970) and work by other “new” anthropologists such as Tinker (1982) and Maguire (1987), the United States Agency for International Development (USAID) adopted an approach called “Women In Development” (WID). A decade later, the recognition of gender as essential to development analysis culminated in the Special Programme for Africa’s 1998 Poverty Status Report being focused on gender and growth (Blackden and Bhanu, 1998). Evidence on the costs of ignoring gender emerged as illustrated by the cases of Burkina Faso, Kenya, Tanzania and Zambia shown in Box 1. As can be seen from Box 1, allocating resources to women increases output by 10–20%. Similarly, reducing women’s time burden has beneficial effects on smallholders’ incomes, as well as on labour productivity and capital productivity.

Box 1: Gender and Growth: Missed Potential

Burkina Faso: Shifting existing resources between men’s and women’s plots within the same household could increase output by 10–20% – see also Box 2.

Kenya: Giving women farmers the same level of agricultural inputs and education as men could increase yields obtained by women by more than 20%.

Tanzania: Reducing time burdens of women could increase household cash incomes for smallholder coffee and banana growers by 10%, labour productivity by 15%, and capital productivity by 44%.

Zambia: If women enjoyed the same overall degree of capital investment in agricultural inputs, including land, as their male counterparts, output could increase by up to 15%.

Sources:

Taken from Blackden and Bhanu 1998.

Although the WID approach was initially considered a breakthrough, it soon was criticised for perceiving womens’ in terms of their sex – which represents their biological differences from men, rather than in terms of their gender – which epitomises the discriminatory nature of the social relationship between men and women. Critics argue that it is too instrumentalist (Pearson, 1994; Baden and Goertz, 1998). Pearson warns that concern for gender only as an instrument for achieving development goals such as poverty reduction will result in gender equity being considered only as means to other ends. To avoid this shortcoming, interest has shifted to “Gender and Development’”[3] (GAD). This new emphasis argues that women and men’s social relations are determined by their economic, cultural, religious and historical context (Whitehead, 1979), thus echoing the views of many feminists. Indeed, Moser (1993) contends that gender relations are temporally and spatially specific. Similarly, Pearson (1994) adds that the form that gender relations take in any historical situation is specific to that situation and has to be constructed inductively and cannot be read from other social relations nor from the gender relations of other societies. Along the same lines, Chafetz (1990) maintains that processes and structures at macro and meso levels shape gender at household level. Hence, gender stratification systems are interrelated with all other socio-cultural institutions and processes.

Reflecting on the new directions taken in gender analysis, Kabeer (1994) remarks that the shift from “women” to “gender relations”, as the key focus of analysis in development was an attempt by some feminist scholars and practitioners to bring power relations between women and men into the picture. She advocates a relational analysis of gender inequality, which involves looking at the institutionalised basis of male power and privilege and aims to uncover the operations of male power within the purportedly neutral institutions within which development policies are made and implemented. However, she also adds that while the terminology of gender, gender roles and gender relations has been widely adopted, their implications have not been fully worked through. In certain cases, this has provided an excuse to abandon any measures intended specifically to benefit women; the reasoning being that the focus on gender relations means that women-focused policies and projects go against the spirit of gender analysis.

Kabeer’s point is shared by many others (e.g., Ellis 2000) who go further by saying that the development profession has found it much easier to assimilate gender through its link to other development objectives rather than to challenge directly the social and institutional mechanisms by which gender inequalities are perpetuated over time. Gender equality is seldom set as a leading goal in its own right. However, in spite of this dominant view, there is a burgeoning literature and an emerging movement pushing for the consideration of women’s economic and social rights as an end in itself (see for example Day and Brodsky, 2000).

Regarding gender relations in the African context, Mama (1997) embraces the context-specific view by emphasising the need to account for diverse realities and identities. She insists that cultures are not homogenous and that people’s identities result from the interaction of divisions of gender, class, ethnicity and other categories of social difference within a historical context. Like Kabeer, she also believes in the primacy of relations of power. Current evidence on power inequality, as illustrated in table 1, fully supports this view.

Imam (1997) also adheres to the use of the social relations of gender as the focus of analysis rather than “women” or “men.” She welcomes the now general recognition that “women” form a heterogenous group, which ranges from rural peasant women to urban middle-class women, but also points out that regardless of the differences in the life situations of these groups of women, there are systematic similarities, all other things being equal, between women taken as a group and men as a group. She contends that the power, access to resources and autonomy of rural peasant women are usually lower than those of peasant men, as are those of middle-class women compared to those of middle-class men. Both peasant and middle-class women in northern Nigeria, for example, are discriminated against at the societal level regarding issues such as inheritance rights and access to education. Men (peasant or otherwise) usually receive more favourable treatment on these issues. Consequently, she argues that attention needs to be paid to the nature of the relationships between gender groups rather than absolute categories of women or men. She maintains that the crucial question is to understand the “interweaving of class, gender, imperial relations, etc. Following Brewer when discussing black feminism, she believes that the mode of analysis is not “race + class + gender”, but “race x class x gender”.

To illustrate gender inequalities in power, Table 1 shows the change in women’s share of seats in parliament between 1987-2000 in sub-Saharan Africa. In 22 countries out of 30, women’s share of seats in parliament increased, and in 8 countries, it decreased. In 1987, it varied from 0% in the Comoros and Djibouti to 16% in Mozambique. In 2000, it increased to 2% in the Comoros, whereas South Africa had the highest share with 30%. However, in 15 countries, women had less than 10% of the seats in parliament. Appendices 1 to 4 further confirm gender disparities in secondary level enrolment ratio, share of employment in industry and services and share in self-employment.

Table 1: Change in Women's Share of Seats in National Parliaments in sub-Saharan Africa, 1987-2000

Source: Inter-Parliamentary Union Website: http://www.ipu.org; http://www.un.org/Depts/unsd

As stated above, the aim here is to design a methodology for analysing the linkages between gender equity and SAPs. Before turning to the review of existing literature on theoretical and empirical work on gender and SAPs, I first clarify what is meant by “methodology” and why it is deemed essential to my approach.

Methodological starting point

It can be argued that it is key methodological questions that shape the character and the conclusions of a research study. Given that policies are influenced by the conclusions of research studies, methodologies occupy an essential role in the process of policy-making. A methodology can be defined as a system of methods and rules that facilitates the collection and analysis of data. It offers a basis for choosing an approach, which consists of theories, ideas, concepts and definitions of the topic. There are different ways of perceiving what reality is (ontology), different possibilities of asserting what can be accepted as real (epistemology), different approaches for validating and confirming this understanding and consequently different techniques for collecting data. There are two main methodological traditions: a positivistic approach based on hard statistical data and an interpretative approach based on soft data. Economists often use a positivistic approach in the analysis of the effects of SAPs, whereas other social scientists favour the interpretive approach.

Regarding studies on the effects of SAPs, for many years, it has been implicitly assumed within leading development institutions such as the World Bank that serious research had to be based on complex statistical and econometric methods. More recently, this attitude has changed. Many WB poverty assessments now include a participatory poverty analysis, and the latest World Development Report (WDR) on Poverty represents an international landmark study on the importance given to the analysis of the voices of the poor (World Bank, 2001).[4]

Gender and Adjustment: Theoretical Literature

Gender in SAPs

Although, the implementation phase of macroeconomic policies have usually ignored gender dimensions, there have been attempts to incorporate gender into the theoretical analysis of SAPs. Using tools of analysis from neo-classical economics, such as informational biases, factor market rigidities and market distortions, Collier (1989) has been one of the first to show how gender biases can ultimately hinder the predicted effects of adjustment policies. Collier’s analysis runs as follows.

It is expected that the “supply response” which follows the implementation of adjustment policies would result in a shift of resources – essentially labour and credit – from non-tradables to tradables, and that this would correct what was seen as initial misallocation of resources between sectors. However, due to rigidities stemming from gender roles and inequitable gender relations, Collier argues that resources may be allocated between sectors in a skewed manner. Furthermore, in spite of a rise in the price of tradables, these constraints can hinder the mobility of labour between sectors and hence ultimately reduce the expected supply response from adjustment policies.

Collier (1989) identifies four distinct processes arising from specific “social conventions” which explain why women face differential constraints in the economic arena. The first form of discrimination operates in the labour and credit markets. The second source of inequality arises from gender-specific roles in production, which get transmitted to younger boys and girls. The third is embedded within the household, where rights and obligations between husband and wife are asymmetric. Lastly, he believes that the heavy burden of reproduction takes its toll on women’s health and time.

Collier’s model has been criticised by Miller and Razavi (1998). Miller and Razavi maintain that although helpful, Collier’s model is problematic, as it does not offer any explanation to how gender differentiated roles are created and maintained. Furthermore, they argue that by putting gender asymmetries in rights and obligations in terms of “social convention”, the model restricts itself to a static picture. On the contrary, they allege that these asymmetries are created and re-created through the everyday relations between men and women, and are “actively mobilised by those in power against those without power.”

Meanwhile, in their study of gender and growth in SSA, Blackden and Morris-Hughes (1995) highlight another major source of gender biases in structural adjustment packages that arise from the invisibility of women’s work, from asymmetrical rights between men and women (as defined by the gender division of labour) and from the linkages between the paid and the unpaid economic sectors. They believe that if these gender biases are not removed, adjustment policies might not result in the intended goals.

In spite of the diversity of views on the exact sources of gender inequalities, most authors agree that not integrating gender analysis into SAPs will only result in non-optimal, non-expected outcomes. Such scepticism partly originates in the feminist economists’ critique of neo-classical economics. We briefly expose the main arguments on which this critique is based before reporting the main findings of the empirical literature.

Feminist critiques of neo-classical economics

Feminists have denounced the gender bias in neo-classical economic theory, which takes gender relations as given. As Sparr (1994) argues, neo-classical economics is not a value-neutral science. She points out that it is ahistorical and wrongly assumes that all societies are fully monetised and market-oriented. Furthermore, neo-classical economic theories ignore the implications of male-female power dynamics on macroeconomic outcomes. They not only neglect the implications of men and women’s unequal political and economic powers but also do not account for labour inflexibility and constraints.

Furthermore, neo-classical economic theories dismiss the fact that gender power relations operate not only at the household level, but also at the community level, in markets and local governments. At the micro level, neo-classical economics conceptualises inequality between males and females as a matter of differences in preferences and resource endowments. As a result, differences in men and women’s labour allocations are taken as the consequence of the lower returns of women’s labour on the labour market and their preference for housework and women’s unequal access to leisure time is attributed to their greater taste for altruism. At the meso level, neo-classical economic theories assume that markets and firms are gender-neutral. Although the existence of gender discrimination is acknowledged, it is seen as self-correcting within the logic of markets and hence no guidelines are offered for any corrective action. And finally, at the macro level, women’s work in the “reproductive economy”, which represents the unpaid production of human resources, is simply rendered invisible.

In reaction to these assumptions, feminist economists provide counter-arguments at the micro, the meso and the macro levels. At the micro level, they underline the role played by intra-household markets and bargaining. Palmer (1991) thinks that in the African context, intra-household markets are biased against women. Examples of this include the favouring of boys’ education, as opposed to the education of girls, and also men’s control over household resources. At the meso level, feminists attack the assumption that market institutions are gender-neutral, and instead argue that they are “bearers of gender” as gender hierarchies and norms determine the social relations between employers and employees, buyers and sellers. Looking at the engagement of women in South Asian and West African markets, Harris-White (1998) concludes that in both “male” and “female” marketing systems, marketplaces and staple food institutions are highly gendered complexes of social institutions in which male domination is reproduced. She finds that the dynamics of gender relations as well as the “bearers of gender” within both social institutions and economic models clearly change in complex ways over time. And finally, tackling the macro level, Elson (1998a, 1998b) presses for the inclusion of gender in economic models arguing that “male bias” introduces flaws to the assumptions on which macroeconomic models are based. For instance, they argue that economic models wrongly assume that the reproductive economy is assumed to continue to function, irrespective of the changes in the rest of the economy. Elson is aware of the difficulties in introducing gender in economic models. She maintains that for feminist analysis to have any real effect on the construction of macro policy, there should be less talking among scholars themselves and more engagement with policy-makers in key ministries and within international organisations such as the World Bank and the IMF.

Empirical Studies on the Effects of Structural Adjustment Policies on Women in sub-Saharan Africa

Intended effects of SAPs

In this brief overview of empirical studies on gender and SAPs, I will focus on their aims, the country-specific policy context, the data they use, their methodological principles and their results. This section will provide details of six case studies. To indicate willingness to learn from cross-disciplinary approaches, the selection comprises two studies that use a combination of quantitative and qualitative methods, two studies that are based on qualitative analysis, one literature review and document analysis and one study based on statistical methods.

As a starting point, Table 2 presents a summary of the theoretical predictions of the effects of adjustment policies on different occupational groups based on Killick’s work (1995). At first glance, it appears that SAPs are foreseen to have more negative effects than positive ones. Indeed, there are 27 N, 8 P and 6 U in this Table, with the worst hit expected to be wage-earners. All the policies, ranging from devaluation to credit freeze, are expected to have a negative impact on this group. In contrast, devaluation, import and export liberalisation policies are expected to have beneficial effects on cash-crop producers and smallholder farmers. Hence, SAPs are expected to benefit farmers and reverse the initial bias of government policies towards urban areas. Since women are part of each of these groups, we can conclude from Killick’s analysis that women farmers are expected to benefit from SAPs, whereas urban women are expected to lose out.

Table 2: Expected outcomes from SAPs

P: Positive effects

N: Negative effects

U: Uncertain

Adapted from Killick (1995)

Real effects of SAPs on women

However, in spite of these intended effects, many studies have shown that African women farmers have not benefited from SAPs (Gladwin, 1991). We first discuss Ongile’s “before and after” study (1991) of household-level factors, especially gender relations, which may “influence the response to, and experience of adjustment.” (1999:47). The study is based on primary and secondary data sources, with a particular emphasis on the tea sector, the country’s largest foreign exchange earner. The author used a 1985-86 study on “Gender Dynamics in Contract farming: Women’s Role in Smallholder Tea Production in Kericho District, Kenya” as a source of baseline data. This provided the basis for comparison of farming practices before and after the adoption of some key agricultural adjustment measures. Another household survey was carried in 1995-96. Ninety-four of the 120 households present in the 1985/86 sample were located and interviewed. The sample included tea-producers and non-tea producers. Both surveys used structured questionnaires for both male and female farmers. Using a combination of quantitative and qualitative techniques, Ongile (1999) argues that due to their different roles and activities in agriculture, men and women are likely to react differently to economic reforms. She demonstrates that gender-based constraints can restrain supply response and hence result in lower production levels than expected.

Exploring women farmers’ perceptions of the costs and benefits of structural adjustment programmes for both tea and non-tea producing households, Ongile points out that there was no policy reform measure targeted at tea prices as part of the structural adjustment programme in Kenya. The smallholder’s incentives were therefore dependent on fluctuations of the exchange rate and world market tea prices, and on the level of tea payments that farmers would receive from the Kenya Tea Development Authority (KTDA), a state-run corporation responsible for overseeing the smallholder sector. However, in spite of the absence of specific policies related to the tea sector, the author argues that the standard of living of tea-growers would still be affected by policies such as the introduction of user charges in health and education, civil service reforms (because of cuts in remittances from relatives in the civil service), rises in fertiliser prices, the general rise in the price of consumer goods and also by price movements in the maize sector, since farmers generally grow both crops.

The study also brings to light some interesting findings regarding perceived changes in the standard of living. Simple questions regarding the respondents’ perceptions and opinions of changes were asked. Out of a sample of 88 women, half of them felt worse off than in 1985/86, 31% felt better off and 18% felt the same. The perception varied greatly between and within hous