ISBN 92-2-111662-X
ISSN 1020-5322
First published 1999

Foreword

Concerned by increasing poverty and social exclusion around the world, and by the inadequate emphasis so far placed on skills and training as a critical component of any strategy for tackling the problem, the ILO's Training Policies and Systems Branch launched a study on the subject in 1998.

The report, which is here issued as an Employment and Training Paper, examines the current "training crisis" in terms of the limited impact that existing vocational training has had on the poor. It pays particular attention to the failure of training systems as a whole to re-orient their activities towards the poor, although this would be vital to provide them with the knowledge and skills required for a significant increase in their productivity and incomes. More importantly, it identifies the elements for a pro-poor training strategy. In the light of the findings, some weaknesses are identified in the current ILO standards (Convention no. 142 and Recommendation no. 150) on human resources development.

I wish to thank: Paul Bennell for producing this excellent report as an ILO external consultant; Eugenia Date-Bah for developing the analytical framework to launch the study and for supervising the preparation of the report; CINTERFOR, other ILO offices and teams, and a number of other bodies for responding to our request for relevant materials; and colleagues in Geneva for their comments on a draft of this text.

The printing of this report is one of the efforts the ILO is making to widely distribute its findings and concerns on this critical issue, and to share the information with relevant actors in order to support and galvanize action. Preparation is also under way for an expert group discussion based on the document, in order to draw specific conclusions for our future work. More generally, we intend this document to be used to promote debate on the role of training as part of a concerted attack on poverty and social exclusion. It is also designed to help deal with the "information crisis" - the paucity of good quality data about the provision of training to the poor and its impact on their employment and incomes.

Gerry Rodgers
Chief
Training Policies and Systems Branch

Executive summary

In the context of mass poverty in most developing countries, the critical role of training in furnishing badly needed skills to improve productivity, incomes and equitable access to employment opportunities seems particularly obvious and straightforward. Certainly, pronouncements abound on the fundamental importance of skills and capacity building in the development process, especially in the fight against poverty. However, a particularly striking feature of most poverty reduction strategies in developing countries is that the role of vocational education and training (VET) in its wide variety of forms is largely absent. This neglect is puzzling not only because of the extent of absolute poverty in most countries, but also because it is widely accepted that training is an essential instrument of public policy, especially for the most vulnerable groups in society. For many, it is precisely because the vast potential of training has not yet been realised that the role of VET has become so marginalised in most poverty reduction strategies. We are confronted, therefore, by what looks like a major contradiction: Just as governments and donors have begun to give due recognition to the need for concerted efforts to build the human assets/capabilities of poor, training is being accorded less and not more importance.

The main objective of this paper is to analyse the reasons for this alleged failure of national VET systems to provide the main target groups among the poor with the knowledge and skills needed to increase significantly their productivity and incomes. With such a large literature on training to overcome economic vulnerability in both developing and developed countries, one may wonder whether this is really necessary. However, despite widespread concerns about the performance of training institutions, particularly in the public sector, remarkably little attention has been focused on trying to understand the precise nature of this training crisis, and, more important still, what should be done to resolve it.

The first part of the paper focuses on the two main dimensions of this crisis, namely the failure to re-orient 'training systems' to support the poor and the limited outputs and impacts of most training interventions. While in many countries, low impact and limited reorientation are closely inter-related, the failure to separate clearly between the two has resulted in considerable confusion. Given the received wisdom that training for the poor has had fairly limited impact and training systems have not been reoriented to meeting the need of the poor, the key question is 'what is the scope for improvement with respect to both these dimensions of the training crisis?'

The paper discusses why the prevailing mood amongst most expert commentators is so pessimistic. Two types of pessimism are identified. 'Training impact pessimists' maintain that training interventions for the majority of the poor are only ever likely to be effective under the most exceptional circumstances. Consequently, there is little point in trying to reorient public training systems in support of these groups. Instead it is better to concentrate on areas of training that have high pay-offs (which are mainly in the formal sector) and provide other types of support for the poor (such as micro credit, primary education and health services) that have much greater impacts on poverty reduction. 'Training system pessimists', on the other hand, argue that, while the training record has not been good, considerable scope still exists to develop training interventions that can effectively address the skill needs of the poor. However, this can be achieved on a mass scale only if training systems are themselves comprehensively reformed. Their pessimism stems, therefore, from their assessment of the poor prospects for significant re-orientation of national training systems in the foreseeable future. Of particular concern is that, while the number of people living in absolute poverty continues to grow, the capacity of the state to support appropriate training appears to be declining in many developing countries. More generally, given dwindling resources and other pressing demands for training services from other sectors, there is a sense of being overwhelmed by the enormity of the skills challenge in support of the poor.

The report also considers other factors that have further compounded the pervasive concerns about lack of impact and/or re-orientation. In particular, there is considerable confusion about what exactly "training to overcome economic vulnerability" actually refers to and the availability of hard evidence on training provision, outputs and impacts, in itself, amounts to an information crisis. A clear distinction is made between the traditional formal training courses of both public and private sector providers and new types of 'participatory skill development' that emphasise the distinctive nature of skill formation among the poor and are premised on the belief that the poor already have many of the skills they need in order to improve their productivity and incomes. The efficacy of group empowerment is also central to this more radical approach.

The last part of the paper considers some of the main reforms that are needed in order to create a 'pro-poor' national training system. It is argued that the need for fundamental reform of VET provision in most developing countries is compelling and should, therefore, be addressed by governments and all other major stakeholders as a matter of urgency. However, remarkably little serious attention has been devoted to analysing what exactly the main features of a pro-poor training strategy and related national system should be. Current debates are excessively preoccupied with the 'higher' skills needed to achieve international competitiveness in a rapidly globalising world economy. The ILO should take the lead in initiating a more balanced and well- informed dialogue about skills development for the economically disadvantaged and socially excluded. The ILO's own conventions and recommendations on training will need, therefore, to be carefully scrutinised. It is suggested that serious consideration should be given to the formulation of new international labour standards (convention and recommendation) that specifically address training for the poor and other disadvantaged groups.

In thinking about what a pro-poor training strategy should look like, two sets of issues are considered in detail, namely overall resource availability and the development of the training system itself. The main functional components of this system (governance, planning, funding and actual delivery of training services) are discussed.

There is an emerging consensus that skills development for the poor must be part and parcel of community-based economic and political development. Communities need to mobilise around specific "development alternatives" that address key political, social, and economic constraints. Skills delopment should be driven by a 'people-centred' pedagogy' which mximises locally available skills and empowers the poor to learn for themselves. Support for skills development should be directly linked to the actual skills needs of the poor and, invariably, will need to be closely related to on-going production activities. However, many of the key characteristics of market-driven VET reform strategies can and should be incorporated into the design of pro-poor training strategies. In particular, the state should perform a largely regulatory and facilitatory role while actual training provision should, wherever possible, be contracted out to independent training providers.

1. Introduction

This paper explores the role of training in assisting individuals who are economically vulnerable and socially excluded (EVSE) in developing countries. Roughly speaking, almost one in four of the population in the developing world lives in absolute poverty and this number continues to increase rather than decrease. Poverty reduction is now at the top of the policy agendas of most bilateral donor agencies and international development organisations within and outside the United Nations system as well as a growing number of governments. Ambitious targets to halve poverty by 2015 have been set by the Development Assistance Committee of the OECD (see UNDP, 1998; OECD, 1997).

1.1 Training and the poor

In the context of mass poverty in most developing countries, the critical role of training in furnishing badly needed skills to improve productivity, incomes and equitable access to employment opportunities seems particularly obvious and straightforward. Certainly, pronouncements abound on the fundamental importance of skills and capacity building in the development process, especially in the fight against poverty. "Knowledge, skills and competencies of all men and women have become the cornerstone of personal growth and employability, enterprise competitiveness, and society's economic and social sustainability" (ILO, 1997: 5). Statements of this kind are backed up by a large body of research that clearly demonstrates that poverty is directly correlated with the level of human capabilities. Self evidently, therefore, there is enormous need to upgrade the knowledge and skills of the EVSE. According to Ducci, training for the informal sector is "a vast and promising area for future action" (Ducci, 1994:183).

However, a particularly striking feature of most government and donor poverty reduction strategies in developing countries is that the role of vocational education and training (VET) in its wide variety of forms is largely absent. For example, in the UNDP's Human Development Report, training is not treated as "basic social service" for all (unlike primary education and basic health care), although it is acknowledged that there is an urgent need "to strengthen the institutional capacity for delivering these services" (UNDP, 1998). Apart from the ILO, the invisibility of training for the poor as a priority issue is equally apparent in most other high profile reviews of poverty alleviation and human resource development which have been produced by both bilateral and multilateral donors (see World Bank, 1995; DFID, 1997; UNICEF, 1998).

This neglect is puzzling not only because of the extent of absolute poverty in most countries, but also because it is widely accepted that training is an essential instrument of public policy, especially for the most vulnerable groups in society. Certainly, the standard definition of 'basic education for all' which emerged from the Jomtien Conference in 1990 does cover "all the skills and knowledge that people need if they are to lead a decent life". These "basic learning needs" include early childhood education, primary schooling, and non-formal literacy and other programmes for youth and adults including vocational training that helps to provide basic life and employment skills (UNESCO, 1991).

For many, it is precisely because the vast potential of training has not yet been realised that the role of VET has become so marginalised in most poverty reduction strategies. We are confronted, therefore, by what looks like a major contradiction: Just as governments and donors have begun to give due recognition to the need for concerted efforts to build the human assets/capabilities of poor, training is being accorded less and not more importance.

Some would go so far as to argue that VET is in danger of becoming a Cinderella sector as donors and governments focus their efforts on basic education and other forms of intervention, most notably microfinance. While the standard definition of 'basic education for all' does clearly include provision of basic vocational skills, this key areas of skill formation has been largely excised from the poverty reduction discussions and debates within the donor community during the 1990s. This process of policy exclusion needs, therefore, to be carefully analysed.

1.2 Study objectives

The main objective of this paper is to analyse the reasons for this alleged failure of national VET systems to provide the main target groups among the poor with the knowledge and skills needed to increase significantly their productivity and incomes. With such a large literature on training to overcome economic vulnerability in both developing and developed countries, one may wonder whether this is really necessary. However, despite widespread concerns about the performance of training institutions, particularly in the public sector, remarkably little attention has been focused on trying to understand the precise nature of this training crisis, and, more important still, what should be done to resolve it.

The following discussion focuses on the two main dimensions of this crisis, namely the failure to reorient 'training systems' to support the poor (Chapters 2 and 3) and the limited outputs and impacts of the training services that have been provided by both public and private sector organisations (Chapters 4, 5 and 6). In Chapter 7, we consider some of the main reforms that are needed both with respect to training systems (in particular governance, organisation and funding) and the actual delivery of training services to the poor.

In a discussion of this kind, there is inevitably a strong 'generalisation imperative'. While generalisations are indeed unavoidable, it is important to emphasise that there is an enormous diversity of experience among national vocational training systems across developing countries. While there are, therefore, no simple 'recipes' and 'magic bullets', it is nonetheless possible to identify not only common areas of weakness in the ability of governments and individual training organisations to promote skill development among the poor, but also the key underlying characteristics of a pro-poor training system.

1.3 EVSE target groups

Poverty is the inability to maintain a minimal standard of living. It consists of two elements. The first is the expenditure necessary to buy a minimal standard of nutrition while the second element varies from country to country and reflects specific national normative concepts of welfare. As societies become wealthier, perceptions of the acceptable minimum level of consumption also change. Consequently, poverty is a context-specific concept and, as such, is very much a moving target (See DANIDA,1996).

The EVSE are an extremely heterogeneous grouping. Being able to identify and target services at specific, well defined groups is one of the main challenges in designing and implementing poverty reduction programmes. The main defining features of the poor are: nature of employment (waged/self-employed/unemployed); gender and age (male/female, adult/youth/child); extent of physical and mental disability (able-bodied/disabled); location (rural/urban/peri-urban); sector (farm/non-farm, specific activity); household characteristics (size, dependency ratios, male/female headed); degree and duration of poverty; specific 'minority groups' who suffer from particular forms of discrimination and/or neglect; and populations affected by war and natural disasters.

It is estimated that there are currently around 1.3 billion people living in absolute poverty in the developing world. Table 1 shows that 56 per cent of the absolutely poor (i.e. those living on less than a US dollar a day) live in South Asia and Sub Saharan Africa (SSA). The poor are heavily concentrated among small, resource poor farmers and those engaged in a wide range of manufacturing and service activities in the largely non-regulated, rural and urban informal sectors (see Tokman and Klein, 1996). There are no reliable estimates of the overall size of the informal sector. Typically, in SSA and South Asian countries, well over 60 percent of urban populations are employed in the informal sector.

Table 1. Population living on less than $1 a day in developing regions, 1987 and 1993

Source: World Bank, 1998 World Development Indicators.

Women and disabled persons are particularly susceptible to poverty. Consequently, policy measures that target both these groups, in particular by reducing labour discrimination and improving human capital should be a central feature of all poverty reduction programmes. While the need to tackle the 'gendered nature of poverty' is increasingly emphasised in policy discourses, "the condition of the disabled is at the bottom of the development agenda" (Ghai in Harriss-White, 1996:i). In India, for example, there are more seriously disabled people than there are seriously malnourished ones. In Asia as a whole, four per cent of the population are seriously disabled (see ILO/ARTEP, 1994). In many countries, particularly those in post-conflict situations, this percentage is much higher. In Uganda, for example, there are 800,000 disabled people.

Among the 'economically vulnerable', there is an enormous range in the level of economic and social well being. At one extreme, are the most marginal groups (including the destitute, beggars, street children) while, at the other, are those with regular, relatively secure sources of income (such as the operators of well established microenterprises) who may occasionally fall below a national poverty line ⁽¹⁾. To date, 'training for the poor' has mainly benefited relatively better-off groups among the EVSE mainly because they are more 'reachable'/ 'investable' and are more likely, therefore, to have identifiable training needs. However, although the most economically vulnerable are generally the hardest to reach, the potential 'pay-offs' of being able to reduce significantly the number of people who are living in the greatest poverty are enormous.

Identifying the contribution that training can make to reducing levels of poverty among the myriad of economically vulnerable groups is a major challenge for researchers and policymakers. For some groups (in particular, illegal child labour), the need for any kind of training intervention may itself be called in to question. Thus, the role of training in poverty reduction must be situated in a wider analysis of the causes of economic vulnerability. With regard to employment issues, there are three sets of explanatory factors that must be disentangled: lack of human capital, presence of labour market discrimination and other distortions, and poor macroeconomic and labour market conditions.

2. The training crisis: an overview

2.1 Dimensions of crisis

There are two basic sets of concerns about VET and poverty reduction. The first focuses on the failure of most targeted training interventions to have any appreciable, sustained impact on livelihoods. As will be discussed below, numerous reasons have been identified for this lack of impact. Most of these relate to the poor quality and relevance of training and/or the inability of trainees to utilise the skills and knowledge they have acquired due to a variety of economic and social constraints. The second set of concerns focuses on the alleged failure of national vocational systems to reorient themselves to meeting the skill needs of the EVSE. While in some countries, low impact and limited reorientation are closely inter-related, the failure to separate clearly between the two can result in considerable confusion.

2.1.1 Poor outputs, limited impact

During the 1970s, there was considerable optimism among policymakers, donors and researchers about the potential impact of vocational training on productivity and incomes for the poor. In particular, the 'discovery' of the informal sector resulted in a wave of recommendations and interventions to support mainly non-regulated micro and small enterprises (MSE). While this optimism had largely dissipated by the mid-late 1980s, there are still those who see the role of training for the informal sector as relatively unproblematic. For example, a recent study by the OECD confidently concludes that "All persons, whether small or micro enterprises, must be helped to acquire minimal training in the trade concerned and in elementary management. Even if an artisan knows his trade, he is often handicapped by ignorance of the simplest management techniques. Knowledge of these techniques can transform a worker into a head of a micro-enterprise. Very short training sessions (a few days) adapted to the sector can be devised. The state could often entrust this task to NGOs" (Morrison, 1995:28). Clearly, however, this begs the question, 'if training is so easy to deliver and the potential pay-offs are so great, why hasn't this happened throughout the developing world'?

Most of the leading experts on vocational training in developing countries are in broad agreement that formal training has had little impact in overcoming economic vulnerability among the poor, and that, as a result of 'unintended consequences', the overall impact of some interventions may even be negative. In their recent book on training for self-employment, Grierson and McKenzie state that "most training systems offer only limited support to those seeking work or self employment. Indeed it is commonly accepted that the inability of vocational training systems to serve labour markets is a problem of crisis proportions...It has yet to be demonstrated that vocational training institutions can re-orient themselves to the specialised field of enterprise development" (Grierson and McKenzie, 1996: 15). Most other leading commentators have reached similar conclusions (see Box 1).

It is important to emphasise that this lack of impact is not just confined to the developing world. For example, in his submission to the G7 Employment Conference, the Director-General of the ILO noted that "the economic and social returns to standardised labour market training measures for vulnerable groups is low" (ILO, 1997:4). Similar conclusions have been reached by numerous other reviews of the role of special training programmes in reducing poverty in the advanced industrial economies (see, for example, Lalonde, 1996).

2.1.2 Lack of provision and system reorientation

It is widely argued that training systems in developing countries should meet the training needs of the poor in an effective and equitable manner. "The bulk of new jobs are being created in micro and small enterprises. Consequently, the training system should prepare people to be productively employed in these sectors" (ILO, 1998:57). The continuing lack of training opportunities for the poor and disadvantaged is, therefore, a constant refrain in the VET literature.

Public training institutions, in particular, are criticised for being both elitist and ineffectual. The principal beneficiaries are urban males from relatively well off background who attend training institutions in order to acquire qualifications that give them access to high paying professional and technical jobs in the formal sector. These institutions, it is argued, have neither the capacity nor the incentives to re-orient their training services to support the poor. Non-governmental organisations (NGOs) are 'closer to the people' and are generally better able to support the training needs of the poor. However, they remain at the margins of the training system and lack the resources to make a large-scale, sustained contribution.

Inadequate institutional adjustment is seen as a universal problem. Even in Latin America which, as a region, has probably the best record for providing training to the poor, there have been "long delays by government in formulating strategies for the informal sector" (CINTERFOR, 1998). Claudio de Moura Castro is even less sanguine: "Solid training institutions failed to adjust and the misfit between supply and demand became endemic" (Castro, 1996: 2).

Given the level of concern about the limited reorientation of VET systems, it is surprising that so little attention has been devoted to analysing the various constraints that have prevented any such reorientation occurring. Even more serious, few, if any, commentators are prepared to say just exactly what this reorientation should entail. The result has been a vague and, at times, rather sterile debate which has lost any real sense of purpose and direction since it was first initiated over a decade ago.

2.1.3 The potential for change

Given the received wisdom that training for the poor has had limited impact and training systems have not reoriented to meeting the need of the poor, the key question is 'what is the scope for improvement with respect to both these dimensions of the training crisis?' Again, the prevailing mood among leading commentators is decidedly pessimistic. Broadly speaking, two types of pessimism can be discerned.

'Training pessimists' (for want of a better term) maintain that training interventions for the majority of the poor are only ever likely to be effective under the most exceptional circumstances. Consequently, there is little point in trying to reorient public training systems in support of these groups. Rather it is better to concentrate on areas of training that have high pay-offs (which are mainly in the formal sector) and provide other types of support for the EVSE (such as micro credit, primary education and health services) that have much greater impacts on poverty reduction.

'System pessimists' believe that, while the training record has not been good, there is still considerable scope to develop training interventions that can effectively address the skill needs of EVSE. However, this can be achieved on a mass scale only if training systems are themselves comprehensively reformed. Their pessimism stems, therefore, from their assessment of the poor prospects for significant re-orientation of national training systems in the foreseeable future. Of particular concern is that, while the number of people living in absolute poverty continues to grow, the capacity of the state to support appropriate training appears to be declining in many developing countries. More generally, given dwindling resources and other pressing demands for training services from other sectors, there is a sense of being overwhelmed by the enormity of the skills challenge in support of the poor.

More optimistic voices tend to be drowned out by the prevailing mood of despondency about the potential role of training in overcoming economic vulnerability. The 1998-99 World Employment Report, which focuses on training, cautions against undue pessimism on this issue. It concludes that "if training is complemented with credit at low rates of interest through a decentralised system of loan delivery and collection, it is possible to make a real difference to incomes in the informal sector" (ILO, 1998). However, while obviously well-intentioned, this hardly amounts to a coherent strategy that addresses all the many concerns about the potential of formal training activities to reduce poverty. Indeed, it is precisely this lack of vision, of simply not knowing what to do, which is perhaps the most worrying aspect of the "training crisis".

2.2 Contributory factors

There are a number of additional factors that have further compounded the pervasive concerns about lack of impact and re-orientation. In particular, there is considerable confusion about what exactly "training to overcome economic vulnerability" actually refers to and the availability of hard evidence on training provision, outputs and impacts continues to be 'lamentable' (CINTERFOR, 1998).

2.2.1 Training provision, outputs and impacts

There is an extraordinary lack of good quality, comprehensive data about the provision of training to the poor and the outputs and impacts of this training effort which, in itself, amounts to an information crisis. Not surprisingly, therefore, most attempts to review the global experience of training for the poor are characterised by sweeping, unsubstantiated observations, generalisations and recommendations and chronic anecdotalism,⁽²⁾ with most reports recycling the same examples of successful and unsuccessful interventions⁽³⁾.

Both ministry-based training institutions and NGOs and other private sector providers are equally culpable. Despite the alleged superiority of NGO training interventions, there is no evidence to suggest that they are any better at monitoring and evaluating outputs and impacts. Out of a total of 30 international NGOs who were asked as part of this study to provide evaluations of training and other skill development projects, not one was able to furnish any robust evidence.

Given this paucity of information, it is not possible to draw any solid conclusions about recent trends in the provision of training to well delineated target groups. However, a proper enumeration of all types of training activities that directly and indirectly impact on the poor would almost certainly reveal a far greater level of training provision than is commonly assumed.⁽⁴⁾.

It is frequently stated that social and private returns to training for the poor are low, but there is little or no hard evidence to support these claims. The alleged limited net benefits to training are in striking contrast to the claims that are repeatedly made with respect to primary education and other interventions in support of the poor (most notably micro credit). What is interesting here is that the evaluation playing field is so uneven. Since most training interventions for the poor are assumed to have measurable and fairly immediate income and productivity pay-offs, it is normally expected that they should be rigorously evaluated. However, usually intractable methodological problems make it virtually impossible to measure the actual impacts of individual primary education projects. Thus, faute de mieux, as long as the outputs of primary education projects are reasonably good, it is usually assumed that the medium-long term impacts will be broadly in line with the social rates of return and externality evidence.

Moreover, since poverty is correctly seen as the consequence of unequal relations in different institutional arenas (the household, community and the state), it is widely believed (at least implicitly) that primary education (particularly for girls) has greater potential for changing these social relationships than training. In fact, because of the prevalence of stereotyped training for women, these kinds of training activities have been criticised for actually reinforcing the subordination of women.

There are a number of reasons for the paucity of information about training outcomes and impacts. These include fragmented provision, the complexity of evaluation methodologies and lack of commitment.

Fragmented provision: Even in quite small countries, training activities in support of the poor are widely dispersed across very large numbers of public and private organisations. Not only is reporting by local NGOs to parent organisations or funders partial and rudimentary, it is equally rare for governments to maintain any kind of basic data base on formal training courses and other types of training services offered by individual ministries.

Private sector training centres have burgeoned in many countries during the last decade. Public training provision has been unable to keep pace with demand (especially among school leavers) and governments have lowered regulatory barriers to entry for aspiring training entrepreneurs. Many private training centres are not registered and, even among those that are, little is usually known about the training that is offered and the clienteles that are served. It has been suggested that 'back yard' training is affordable for the poor and widespread, particularly in urban areas. However, this training is often indistinguishable from on-the-job training, and formal provision (where fees are payable for discrete 'courses') is probably quite limited, especially in rural areas.⁽⁵⁾

Evaluation methodologies: Evaluating training programs is very complex. This gives rise to a number of serious methodological problems which, invariably, are not properly dealt with. In particular, attribution of the precise impact of training activities is virtually impossible when a package of services are made available (credit, tools, technical assistance, training courses). And, given the current "fascination with micro-finance" (Buckley, 1997) there is a real danger that credit may take all (or certainly more than its fair share) of the credit! Selecting a control group in order to compare training and no-training outcomes is also often very difficult and is frequently not done well. Similarly, the nature and incidence of 'displacement effects' are usually ignored altogether.

Tracer surveys of individuals once they have completed their training are essential. However, formal surveys of representative samples of graduate trainees are rare and low response rates invalidate the results of most that are undertaken.

Evaluation commitment: Most donor-supported training projects are formally evaluated. However, evaluations are usually undertaken too soon after projects have finished to be able to assess the extent to which there have been sustainable impacts on the well being of trainees. Rigorous evaluations are too time consuming and expensive for most projects. Moreover, there is usually little incentive to undertake them since it is the immediate post-project evaluation that is most critical in securing additional resources and project extensions.

2.2.2 The concept of training

The general failure to clarify precisely what activities should be included in "training to overcome EVSE" has resulted in considerable confusion and vagueness. The conventional dictionary definition of training is "to prepare for a performance or a task(s) by instruction". Generally, this involves a trainer instructing trainees in a formal classroom or workshop setting. The traditional concept of 'vocational education and training' is very much in accordance with this view of training. The main objective of VET is to furnish the technical and management skills and help develop appropriate attitudes for specific occupations and jobs: It is the "ordered and systematic transmission of skills and dexterities and of technical know how for workers in skilled and semi-skilled occupations" (CINTERFOR, 1997).

Governments in developing countries have established networks of VET institutions in order to supply the high and middle level 'manpower' needed to meet ambitious objectives laid out in development plans and elsewhere. In particular, national manpower development has been inextricably linked with state-led import industrialisation strategies. Consequently, public sector VET has come to be closely associated with the widespread failure of this particular development model. Moreover, as formal qualifications have been increasingly used by employers to screen potential job applicants, chronic credentialism has become pervasive as the supply of job seekers has far exceeded the jobs that are available. There is a strong feeling, therefore, that public sector VET provision is increasingly socially wasteful and, ultimately, dysfunctional.

It is clear, however, that training can embrace any instruction, advice or other type of purposeful activity that seeks to enhance the capabilities of targeted individuals and/or groups through the provision of relevant knowledge and/or development of specific skills. Income generating projects, for example, that enable groups of EVSE to learn new social and technical skills may not include any formal training courses, but assistance provided by an NGO or other provider can facilitate the development of these skills. The problem with broadening the definition of training in this way is that it may become too all-inclusive as a concept and the notion of a 'national training system' could lose any operational meaning since all learning modalities are included.

The conventional concept of training is, in many ways, being superseded by a much wider definition which focuses on activities that promote learning and skill acquisition through empowerment and capacity building but which are not considered to be training per se. In other words, the training function is itself becoming invisible, not so much because it is being ignored or excluded (hence the 'training crisis'), but because it is being more closely integrated or 'embedded' in a range of financial and non-financial interventions that seek to achieve sustainable improvements in the livelihoods of the poor. Many organisations are reluctant to describe what they do as 'training' because, with the growing emphasis on individual and community empowerment, the notion that poverty reduction entails a simple transfer of a discrete body of knowledge and skills from the 'haves' to the 'have nots' is politically and intellectually objectionable. Traditional training approaches are, therefore, being fundamentally questioned, to the extent that there is a denial that training is taking place at all!

In particular, with the advent of the New Poverty Agenda in the early 1990s, supporting microenterprises has become a major focus of governments, donor agencies and NGOs. While skills development is key objective of the microenterprise programmes/projects, most enterprise development organisations do not regard themselves as training organisations. In a very real sense, therefore, because the training function is losing its institutional identity, many specialised training institutions are faced with an identity crisis.

Whatever concept of training is actually adhered to, it is clear that 'training to overcome economic vulnerability' embraces a much wider set of skills than just conventional technical and managerial competencies. These include basic literacy and numeracy, social and political 'awareness '(gender training being a prime example) and life skills. Interventions that facilitate 'personal development' by raising self-esteem, confidence and motivation are the main objective of many NGO interventions. Similarly, it is generally accepted that enterprise development and income generating projects require more complex combinations of skills with much heavier emphasis on social and management skills rather than narrowly defined technical competencies.

2.2.3 Training objectives

Training policy objectives with respect to the poor are frequently poorly defined. Social exclusion is a complex theoretical concept referring to causal mechanisms producing poverty. Translating this concept into practical, poverty reduction policies has proved to be difficult in most countries (see Gore and Figueiredo, 1997).

The labour market is a critical mechanism for inclusion and exclusion. Policies to combat labour market exclusion focus on the elimination or reduction in discriminatory practices and improvements in the human and social assets of the poor. Certain groups, most notably women, disabled persons and minority groups, continue to be seriously under-represented in many occupations. Thus, as part of a comprehensive equal opportunities programme, education and training policy can promote more equitable labour market outcomes. This can be done by improving the access of these groups to secondary and tertiary education institutions and, where necessary, by also providing various forms of support once enrolled so as to boost graduation rates. However, the impact (at least in the short term) of these types of programmes on poverty reduction is likely to be fairly minimal because, in most countries, non-poor individuals tend to be the principal beneficiaries. Furthermore, the formal sector absorbs only a small fraction of the economically active labour force (typically 5-20 per cent in most South Asian and sub-Saharan African countries) so that only tiny proportions of the discriminated groups are likely to be affected.

There are two principal types of training provision to counter poverty. First, there are training activities directly targeted at specific groups of EVSE as the principal beneficiaries (e.g. poor farmers, roadside mechanics, women's income generating projects in a certain area). And secondly, there is training that has as its principal objective the improvement of services for the poor provided by institutions, especially those that are central to poverty reduction programmes. Many of these services are intended to improve directly the skills and knowledge of the poor and, as such, they can be considered to be forms of training.

Surprisingly, the literature on training for the poor focuses almost exclusively on the direct provision of training to the poor and largely ignores the role of training in improving the provision of basic services for the poor. This is a serious omission because the failure of many of the services provided by the state to benefit the poor (particularly women in rural areas), is a key characteristic of the 'training crisis' in many developing countries. The main concern here is that occupational structures in key areas of service provision are too top-heavy with a disproportionate number of professionals providing relatively sophisticated services to mainly non-poor, urban clienteles. Attempts in the past to create more bottom-heavy occupational pyramids have invariably been strongly resisted by professional associations and other powerful vested interests ⁽⁶⁾ (see Box 2).

2.2.4 National training systems

While constant reference is made in the literature to 'vocational training systems', it is rarely made clear what exactly is meant by training system. A system is "a set of things considered as a connected whole". However, in most countries, the degree of connectedness between training institutions in different ministries is so minimal that it is difficult to conceive of a public vocational training system, let alone a system that embraces all institutions that, in one way or another, are concerned with the provision of training services. National training councils are typically weak (advisory) institutions and are usually preoccupied with public sector training provision for technical and commercial occupations in the formal sector. In many ways, therefore, the lack of effective national training systems is a key feature of the training crisis.

More often than not, the term 'training crisis' only refers to a particular segment of the training system, namely post-secondary training institutions that were originally established to meet the 'manpower needs' of 'modern industry and commerce' and are usually part of Ministries of Labour and/or Higher Education. As noted earlier, these training institutions have been widely criticised for failing to provide good quality, relevant and cost-effective pre-employment and job-related training.

The response of enterprises to inadequate public sector training provision has varied. The private sector in some countries has largely turned its back on the public training system and internalised the training function as much as possible. More generally, industries as a whole have tried to take greater control of the training process. The result is that sectoralisation of training provision with the creation of lead bodies, industry training boards etc. is a major trend world-wide. Once individual enterprises and industry organisations gain some measure of control over public training resources and are given the freedom to choose where to train, they are increasingly opting for private sector providers.

The loss of the traditional clienteles among public sector training institutions is bad enough. But, as discussed earlier, the training crisis is frequently compounded by the lack of interest and/or capacity of these institutions to re-orient their activities towards the poor.

While this particular group of VET institutions is struggling to survive in many countries, how true is this for training institutions in other parts of the 'training system'? It is necessary, therefore, to examine carefully the activities and resource commitments of all organisations that explicitly seek to promote skills development among the poor. Only then is it possible to reach any meaningful conclusions about the scope and efficacy of this training effort.

In nearly all developing countries, a significant proportion of the poor are smallholder farmers. And yet, the provision of training services, most notably agricultural extension, to these individuals is rarely considered in discussions of skills training. Similarly, much of the pre-employment training that is undertaken by the core tertiary VET institutions discussed above is for public sector occupations which provide social and economic services that should benefit the poor. Many of these personnel do not necessarily have an explicit training role as instructors, but the very provision of these services facilitates the development of certain skills. For example, road engineers involved in rural feeder road programmes not only impart skills to his/her technicians and other staff, but also to local contractors and the beneficiary communities themselves.

3. Training priorities, resources and reorientation

3.1 The public sector

"While there is long history of poverty-focused training in developed industrial economies, it is still relatively rare in the large majority of developing countries where most of the poor live" (Malik, 1996:46). This seems particularly ironic given that most of the world's poor live in developing countries. The following discussion looks at why public sector training priorities continue to favour non-poor groups. We shall focus in particular on the design of poverty reduction programmes, overall resource availability and competing claims over training resources from other sectors and groups.

3.1.1 Training and poverty reduction

To a considerable degree, limited, often tokenistic public sector training for the poor is symptomatic of the weak overall commitment to eliminating poverty that is displayed by many governments. A major World Bank study on poverty in Sub-Saharan African countries concluded that, in the mid-1990s, less than quarter of governments in the region were strongly committed to poverty reduction (World Bank, 1996). This lack of commitment is manifested in a variety of ways. With respect to the informal sector, governments have been criticised for adopting a 'minimalist' approach. Thus, in Asia, "few governments have tried to understand the constraints facing the informal sector... Besides easing access to credit, it is assumed that the sector will somehow manage and develop without any intervention from governments with NGOs taking increased responsibility" (ILO, 1995:7).

While it is generally acknowledged that creating an enabling macro-economic and legal environment is often likely to have a much greater impact on MSE development than specific promotional measures, deeply entrenched political and social forces prevent governments from doing this. Faced with this situation, "it is far easier and more politically visible to construct a few subsidised shelters to protect a few lucky informal sector mechanics than to liberalise the regulations which continue to constrain the majority" (Harper,1996:107). There are, therefore, real worries that government training provision is a way of avoiding hard political decisions about creating truly 'enabling environments' for the poor in the informal sector.

In Latin America, many national vocational training institutes have set up specialist divisions to respond directly to the training needs of the poor and disadvantaged in the informal sector. However, in SSA and South Asia, public institutions "have found the idea of serving new target groups much more problematic" (King, 1996: 42). But, it is precisely in these two regions where the incidence of acute absolute poverty is greatest (see Box 3).

National training boards and committees are usually only advisory and their composition is heavily weighted in favour of representatives from other ministries and public bodies with minimal representation, if any at all, from organisations that protect and promote the interests of the poor. These governance structures generally lack any real influence over training policies and related resource allocations.

The design of poverty reduction programmes is also a key issue. Safety net programs tend to absorb the bulk of government resources for short term poverty reduction. This includes such measures as family assistance and/or cash social assistance, food transfers, public works, and income generation. Generally speaking, skills development for beneficiaries is not seen as a major requirement in order for these interventions to be successful.

Political complacency has also been an important explanatory factor. In some countries, politicians and policymakers genuinely believe that concerted efforts are being made to train the poor. In particular, technical and vocational secondary schools are often seen as providing skills for the poor. In most of Latin America, "the establishment of (national vocational) training institutes constituted the instrument of public action in the education plane towards the most disadvantaged" (CINTERFOR, 1998).

It is important to acknowledge, however, that the need for the public sector to play a leading role in skills development in the smallholder agricultural sector is generally accepted by most governments and donors. Perhaps the best example of this commitment is the World Bank- inspired Training & Visit extension system (T & V) which has been adopted in over thirty countries during the last 10-15 years. World Bank lending for T&V projects has totalled over US$4 billion since the early 1980s. However, the Bank policies concerning the role of training for the poor in other sectors are, at best, ambiguous and, at times, decidedly negative. This is particularly the case for core VET institutions which, typically, are the sole or joint responsibility of ministries of labour, education and higher education (see Bennell, 1996).

During the 1990s, the World Bank has consistently encouraged governments, particularly in low income countries, to reduce the share of VET in the overall education budget. Thus, one of the main objectives of the World Bank's VET reform agenda is to privatise funding and provision of VET as much as possible thereby ensuring that primary/basic education receives maximum possible government funding. In practice, however, because of the fungibility of budgetary resources, the exact opposite may be happening. In the early 1990s, the Bank was concerned that "educational policy reforms are being undermined because donor support for primary education allows governments to increase spending on VET" (World Bank, 1993:4).

3.1.2 Training for the formal sector

Despite oft-repeated government pronouncements about the need for concerted efforts to improve the skills of the poor, responding to formal sector training needs has remained the top priority for most public sector training institutions during the 1990s. If anything, this bias has increased during the last decade as concerns have mounted about the need to re-shape training systems to meet the challenges of new (post-fordist) production and organisation technologies, international competition, and globalisation. Recognition that international comparative advantage in most of the key growth sectors is now largely determined by human resource endowments means that training systems must be at the forefront of the process of skills-driven economic modernisation. In short, the need to ensure national inclusion in a rapidly globalising economy has taken precedence over the continued exclusion of the poor from training provision.

There is a mass of evidence that demonstrates convincingly that good quality, relevant training does make a major difference to improving the productivity of formal sector enterprises, especially those producing tradable goods where attaining international competitiveness is critical for sustained export-led growth. In contrast, training needs for MSE are often difficult to identify and effective training demand in many activities remains limited. It is hardly surprising, therefore, that most governments are reluctant to alter significantly resource allocations in favour of the informal sector.

These new demands for increased training provision come at a time when budgetary support for public training institutions in many countries has been declining rapidly as a consequence of stringent fiscal belt-tightening. As a consequence, core VET institutions have had their work cut out just finding ways to maintain their current training activities, let alone think about developing training for clienteles who are generally unable to pay. With increased de facto and/or de jure autonomy, public sector training institutions are seeking to generate new sources of revenue from both enterprises and individuals. Unless, therefore, governments and other donors make available earmarked funding for training the poor, it is unlikely that these institutions will reorient their training activities in support of these groups.

Another important consideration is that the design and delivery of training for the formal sector tends to be relatively straightforward compared to training for the poor which, typically, needs to be closely integrated with other support measures which are usually provided by other institutions. These additional costs and risks of providing training for the poor can be a major disincentive for public training institutions.

In many countries (particularly in Latin America), the funding of public sector training institutions relies to a large extent on training levies collected from registered, formal sector enterprises. This has meant that governments have often felt constrained in the extent to which levy resources can be 'diverted' to meet the training needs of informal sector enterprises, nearly all of whom make no levy contributions (see Castro, 1996).

3.1.3 Market-driven training reforms

During the 1990s, the World Bank has taken the lead in promoting the benefits of pro-market reforms for VET. The main objectives of these reforms focus on the need to promote enterprise-based training in both the formal and informal sectors, encourage greater private sector participation, increase cost recovery, and relatively less public sector training provision (see World Bank, 1991). To date, however, it has proved considerably more difficult to privatise VET than was originally anticipated which has meant that the state has continued to be heavily involved in the funding and direct provision of VET for the formal sector. And, in the poorest countries, even the leading proponents of reform accept that "central governments must play the central role in financing and providing training" ( Middleton et al, 1993:265).

The precise implications of market-driven VET reform for skills development among the poor have yet to be fully investigated. However, given that most training for the poor will have to be publicly funded, a pro-poor training strategy could easily be undermined by concerted efforts to privatise the funding of training. Low levels of demand for training among microenterprise operators and workers coupled with minimal ability to pay obviously limit the scope for cost recovery. The failure of World Bank-sponsored initiatives to introduce significant cost recovery measures for primary schooling in a number of low income countries during the mid-late 1980s provides a salutary warning in this regard (see Bennell, 1996 and Penrose, 1998).

3.1.4 Overall resource availability

The extent of public sector training for the poor is also strongly influenced by resource availability and the overall incidence of poverty. In Latin America, national vocational training institutes are relatively well resourced, but it would appear that only in a few countries (most notably SENA in Colombia) have training priorities and related resource commitments been changed significantly in favour of the poor (see CINTERFOR, 1998, Castro, 1996). While the overall incidence of absolute poverty is considerably lower than in Sub-Saharan Africa and South Asia, the political influence of the poor over training policy and practice has generally been greater. Consequently, the potential for public sector training systems in Latin America to support skill development for the poor in a concerted and comprehensive manner is likely to be considerably greater than in other regions.

In Sub-Saharan Africa, most governments are simply too poor to be able to fund major training programmes for the poor. Most training resources are allocated to a small number of public training institutions which train mainly school leavers for skilled occupations which, in most countries, are heavily concentrated in the public sector. Thus, the provision of public VET in SSA and other low income countries tends to be particularly inequitable (see Godfrey, 1998). And, because pre-employment occupational training by ministry-based training institutions accounts for the lions share of public sector training provision, the number of training institutions which could be re-oriented is usually quite small.

The AIDS pandemic will also limit the scope for training re-orientation. Over 20 per cent of the economically active population are (or are expected to become) infected with the AIDS virus in over 15 African countries (see Lowenson and Whiteside, 1997). Consequently, demands on the training system to replace workers employed in the formal sector who will die or become seriously ill will increase enormously over the next ten years.

As noted above, with pervasive economic adjustment, African governments have come under intense pressure by the IMF and the World Bank to privatise the funding and provision of VET mainly because it is not regarded as a 'basic social service'. Evidence is limited, but the share of VET in national education and training budgets has probably fallen very considerably since the late 1980s in most SSA countries.⁽⁷⁾

Given the very high incidence of poverty in South Asia, public sector training provision for the poor has been relatively tiny. Even in India, where there has been a series of nation-wide skills development programmes for poor and disadvantaged groups (including TRYSEM), only 5-6 per cent of urban target groups have been reached (see Malik, 1996).

3.2 The private sector

Little is known about the extent to which private sector training provision benefits the poor and even less is known about recent trends. In the past, NGOs have focused heavily on providing conventional training for the poor, but this may be changing for two main reasons. First, there is now a greater focus on income generating and advocacy projects that, invariably, have more limited formal training inputs. And secondly, in some countries (e.g. Zimbabwe and South Africa), large cuts in donor support to training NGOs are forcing them to commercialise a significant proportion of their training activities in order to survive. Given the inability of poor clienteles to pay for training services, they are, therefore, being increasingly excluded. Detailed research is needed in order to establish just how widespread this reorientation of NGO training is.

4. The demand for training

As is well known, the supply of training does not usually create its own demand. Clearly, therefore, training provision for the poor has been powerfully shaped by the nature of the demand for training among targeted groups, in particular in the informal sector. Lack of effective demand is a key reason for both the limited training provision for the poor (and hence outputs and impacts) in most countries as well as the overall failure of national training systems to reorient their activities in support of the poor.

4.1 The potential for training interventions

There is a tendency to over-estimate the extent of self-employment among the poor in both the urban and rural informal sectors as well as the scope for improving productive independent economic activity. If anything, this tendency has increased in recent years with the emergence of 'private sector development' as the major long-medium term objective of economic reform programmes, particularly among donors. However, the fact remains that most informal sector activity will continue to be "the last resort for the desperate rather than a panacea for employment problems" (ILO, 1997:).

Breman eloquently describes the situation of the urban poor in Gujarat, India. In situations such as these, where "the great mass of people" are in wage employment and "have no prospect of improving their position" (Breman, 1995: 97), there is virtually no demand among the poor for formal training. In particular, for those at the bottom of rural society, the formal sector is an "impregnable fortress". With very simple technology and an abundant supply of cheap labour, not only can workers be trained at minimal cost on the job, but there are no incentives to invest in more skill-intensive technologies. In such brutal labour regimes, the degrading effects are such that employers regularly replenish their workforces with new recruits from rural areas.

It is clear that the informal sectors in other countries have much higher proportions of self-employed labour and tend, therefore, to be less exploitative. However, the value of detailed anthropological research of the kind undertaken by Breman is that it highlights the immense and complex array of social, economic and political problems that have to be surmounted in order to achieve sustainable reductions in mass poverty.

Another key development in India and other countries is that "the organised sector is disorganising itself" precisely in order to exploit cheap labour to the greatest extent possible. This process of "informalisation of the formal sector" is likely to become more pervasive as the benefits of non-regulation exceed those from regulation. Probably, the greatest training need in these situations is to develop the capacity of specific groups of poor workers to organise themselves in order to limit the degree of exploitation they are subjected to, in particular by increasing wages and generally improving conditions of work. The success of the Self-Employed Women's Association in Ahmedabad is widely cited in this regard. However, initiatives to replicate this type of project do not appear to have been widespread.

4.1.1 Survival enterprises

In simple numerical terms, 'survival' enterprises predominate in most informal sectors. The general view is that the skill requirements for most tasks undertaken in this type of enterprise are minimal and/or are relatively easily acquired on the job. Women are particularly heavily concentrated in very low skill activities (most notably street vendors and food preparation). It is difficult, therefore, to see how conventional training services could significantly increase productivity and/or incomes in these kinds of occupations. "Within the household subsector and independent services subsector, training, in itself, has little impact in breaking the low income trap" (ILO, 1998: 103). In addition, "economic compulsions" and "acute vulnerability on a daily basis" keep the poor (and poor women, in particular) out of conventional training. Although most labour force surveys show high levels of "under-employment", most poor people are, in fact, too busy working to have time to enrol on training courses of any kind.

Faced with so little scope for improvement among existing activities, more 'transformative' approaches have often been tried by both governments and NGOs. The main objective here is to provide a critical mass of skills and resources to targeted individuals and groups that will enable them to transform their livelihoods. Again, however, most interventions of this kind have not been successful mainly because the resources and time needed to start new enterprises are well beyond the means of the poorest and, more generally, they have failed to address adequately the complex array of constraints that keep the poor in poverty.

4.1.2 Enterprises with growth potential

Most training strategies in the informal sector have targeted manufacturing microenterprises that are considered to have some growth potential. However, even within this relatively better-off segment of the informal sector, the effective demand for training has frequently been found to be quite limited. ⁽⁸⁾

Lack of effective demand: Innumerable surveys of informal sector enterprises have shown that formal training and skill constraints are perceived as being relatively unimportant by most entrepreneurs and workers. Since they have to be 'persuaded' to become training clients, the chances of success are limited from the outset. The low level of formal education is also frequently cited as a key factor limiting the demand for training as well as the overall trainability of operators and workers.

The ILO has had considerable experience in trying to graft training projects onto indigenous apprenticeship systems, particularly in SSA. One of the main lessons that has been drawn from the ILO's 'self-training projects' in Francophone Africa is that, initially, informal sector artisans do not usually regard skills training as a priority. They only become aware of its usefulness once they have gained access to new markets and have negotiated favourable terms of purchase for key raw materials (see Maldonado, 1989). Similarly, the results of a large number of country surveys in SSA during the 1990s (especially those conducted under the auspices of USAID's GEMINI programme and the World Bank) have repeatedly shown that training is not perceived as a major problem viv-a-vis other operational constraints. While training is invariably identified as being desirable, when respondents are asked to specify particular problems facing their enterprises, training and/or availability of skilled labour are nearly always ranked as being of very minor importance (see, for example, Mead, 1990; Parker et al, 1992;). While this does not mean that appropriate skills training could not significantly improve productivity, the overwhelming perception among informal sector operators is that its role is limited.

Most surveys identify credit and access to markets as the most critical constraints. Because replication rather than intensification of activities is generally regarded as the main source of short-term growth, the provision of credit is seen as having the greatest potential for improving incomes. However, because relatively small amounts of credit are, on their own, unlikely to lead to the adoption of new techniques and technologies, the scope to improve livelihoods without significant skills upgrading is likely to be quite limited in the medium-long term.⁽⁹⁾

The general consensus is that, unless a number of key constraints can be effectively tackled, there is little point in providing training to most entrepreneurs and workers in the informal sector. However, whereas training on its own is not usually effective, this is not generally the case for other services, most notably microcredit. There is broad agreement, therefore, that training should be integrated in a package of services. Where these are not affordable or the expertise is not available to design and implement what are typically complex, multi-input interventions, then there has been a marked preference for 'targeted' services such as credit. These can be delivered on their own and have lower opportunity costs than most formal training activities.

Indigenous skill formation: The prevalence of relatively well developed indigenous training systems among MSE in many countries, particularly in West Africa and South Asia is a key factor accounting for the low level of demand for formal training. The authors of the World Bank's influential VET Policy Paper go so far as to suggest that "traditional apprenticeships provide most of the training needed in the informal sector in most countries" (World Bank, 1991:60).

Growing recognition of the importance of indigenous training systems has had a major impact on government and donor training policies and practices. Where these indigenous training mechanisms are in place, it has been argued that "there is no a priori case to be made for specialised training interventions" (Middleton et al, 1993: 167). While there is scope for improvement, there is understandable concern that poorly conceived training interventions could easily undermine the traditional master-apprentice relationship on which the entire system is based. This has, in turn, dampened the interest of both governments and donors in getting too heavily involved in improving indigenous training systems. Informal credit systems, on the other hand, are almost universally inefficient and ineffective which has provided the necessary justification for the provision of cheaper, alternative forms of microfinance that are targeted at the poor.

Formal sector experience: Where MSE entrepreneurs and workers have been previously employed for relatively long periods in the formal sector, this is also likely to lower the demand for formal training. For example, King argues that in most of South America, "the majority of business people who receive training, credit, and other support services have only been able to start their own businesses only after many years of employment in other (formal sector) businesses" (King, 1996).

Direct training costs: The costs of formal training courses is frequently cited by MSE operators as an important reason for not taking up training opportunities. For example, in a large survey of MSEs in Kenya, over 40 per cent of respondents stated that they could not afford training because of 'limited resources' (see Table 2).

Trainability: Illiteracy among the poor significantly lowers their trainability. In South Asia, in particular, a high proportion of workers in the urban informal sector are illiterate rural migrants (see ILO, 1997). Acquiring basic literacy and numeracy skills is, therefore, generally regarded as being of higher priority than conventional VET. There is a growing group of NGOs which see the "transformative potential" of well designed, mass literacy programmes as being the key to poverty reduction. In recent years, Actionaid's REFLECT literacy projects have become perhaps the most well known example of this approach.

Table 2. Reasons for not being involved in formal training activities among manufacturing enterprises in Kenya: Percentage of survey respondents stating that specific reason "important" or "very important", 1994.

Source: World Bank, Regional Programme for Enterprise Development.

4.2 Women

4.2.1 The gendered nature of poverty

Over two thirds of those living in absolute poverty are women (UNDP, 1998). As noted earlier, women are very heavily concentrated in the most marginal survival enterprises (often working at home) and in wage employment in secondary labour markets that are characterised by low skills and high turnover. In Sub-Saharan Africa, they also undertake the bulk of agricultural production. The 'training crisis' is, therefore, overwhelmingly linked to the economic and social vulnerability of women and particularly the multiple constraints that prevent them from exploiting training opportunities.

In all societies, there are four basic institutional arenas - the market, state, community and the household. Each arena is characterised by specific rules, norms and practices which determine structures of entitlements and disentitlements for individuals and groups of individuals. There has been a pervasive failure to ground training policies and practices for women on a sound understanding of these institutional processes. More innovative training, especially in 'non-traditional areas,' has frequently failed to achieve significant and sustainable impacts precisely because it challenges "the norms of gender propriety which constitutes a risk which poorer households are unwilling to take on" (Kabeer, 1997:5). The focus of most training on productivity has 'redistributive connotations' that are threatening to gender relations.

Poor women are already "hemmed in by a complex bundle of risks" which seriously limits their degree of individual 'agency' with respect to most decisions that affect their own well being and that of their children and other household members. However, as far as training is concerned, the degree of female agency varies very considerably between regions and countries. It is especially limited in the 'patriarchal belt' that stretches across North Africa, the Middle and Near East and South Asia. The 'invisibility' of women in these regions is acute. But is also the case that nearly one-third of households in the developing world as a whole are female-headed. While women in these households are much freer to decide what education and training is desirable for themselves and their dependants, higher levels of poverty in this type of household may prevent women from availing themselves of training opportunities.⁽¹⁰⁾

4.2.2 Training provision for women

The identification of women's training needs has often been flawed because "women are rarely treated as knowing what they need" (ibid: 30). The available evidence tends to show that poor women in most developing countries are usually most interested in skills training that meets their own immediate 'practical gender needs' as opposed to longer term, "strategic gender needs" that directly tackle the basic underlying causes of female subordination (see Moser, 1989). As a result, most formal training has been closely related to gender-stereotyped tasks and occupations. Once again, this highlights the fact that training provision for women is itself part of "deep-seated, culturally sanctioned forms of gender inequality".

The demand for training among poor women is not only low but is likely to be falling where poverty and economic pressures on households are increasing. Extensive research has shown that women have usually borne the brunt of household adjustment in the wake of economic crisis and/or macroeconomic policy reform. In these situations, training has rarely been a central part of the coping strategies of poor women.

The control over and allocation of household resources to training is determined by a similar set of factors as for formal education. However, it is the nature of the relationship between the woman and her spouse that is crucial for determining training demand and outcomes. Given the subordinate position of women coupled with very limited household resources, it is unlikely that training for adult female members will be accorded high priority. Furthermore, informal systems of skill transmission in most of the manual trades are generally from father to son. Traditionally, "apprenticeship is an important part of the socialisation process through which masculine identities are constructed" (Kabeer, 1997: 7). Thus, supporting indigenous training systems could actually undermine gender and development objectives.

With such limited room for manoeuvre, there is a widespread feeling that the potential long-term social and economic benefits of improving opportunities for girls in the formal education system are much greater in most countries. In simple terms, " daughters are the future, not mothers". (Godfrey, 1997: 18)

4.3 The impact of economic liberalisation

The potential impacts of economic liberalisation on VET are twofold: change in incentives to invest in training and the availability of public funding for VET. Early arguments in support of economic liberalisation claimed that the removal of labour market 'distortions' (in particular, minimum wages and hiring and firing restrictions) would lead to significant increases in effective demand for training by both individuals and enterprises. Moreover, with the ending of industrial policy regimes that strongly favoured large enterprises, major increases in effective training demand from a rejuvenated MSE sector were also confidently expected.

Just as with formal sector enterprises, changes in the level and type of training activities in the informal sector have to a large extent been determined by the way in which adjustment programmes have impacted on specific areas of activity, in particular those that are more skill intensive. With 'private sector development' firmly established as a key objective of all economic reform programmes, governments and donors are placing more emphasis on the need for effective skill development programmes for MSEs. This is especially so in SSA where the bulk of economic activity is accounted for by these kinds of enterprise.

The evidence is fairly scanty, but what is available suggests that micro-enterprises (i.e. those with fewer than five workers) have often been negatively affected by depressed demand, increased import competition and generally higher levels of uncertainty that have typically resulted from economic reform, at least in the short-term. Not surprisingly, therefore, the expected increases in effective demand for training have not materialised. Returns to activities among survival enterprises in SSA have declined mainly as a result of a flood of new entrants unable to find employment in the formal sector. Barriers to entry remain generally low for most activities (see Dawson, 1993; Dawson and Oyekinka, 1993; Dike, 1995; Gallagher and Yunusa, 1996). Thus, "most enterprises continue hand-to-mouth as increases in demand are quickly competed away" (Steel and Webster, 1991:2).

Among small enterprises (i.e. those with 5-49 workers), much higher capital and skill requirements have tended to restrict increased competition. Consequently, the overall gap in income (and thus returns to training) between relatively high and low income activities may well have widened still further. In some countries, better educated, "middle class" entrants are capturing the more lucrative and skill intensive MSE activities and men are also taking over the relatively few higher income activities where women previously predominated.

The impact of economic liberalisation on traditional apprenticeship systems has not been extensively researched. In Nigeria, however, Gallagher and Yunusa found that fewer school leavers want to become apprentices and the capacity of MSEs to offer training has also declined (op.cit.).

It is possible that more intense competition can also undermine individual incentives for pursuing common interests and undertaking collective action. Research on 'industrial districts' in both developed and developing countries highlights the crucial importance of social and economic relationships of these kinds in achieving high levels of productivity among MSEs. However, in the context of acute economic crisis, it is just as likely that informal associations are "collapsing rather than springing-up" with households rather than the community becoming the focus of resource mobilisation and coping strategies. This clearly has major implications for the current emphasis given by both governments and NGOs to encouraging community/associational-based efforts (including skill development) to improve livelihoods among the poor.

The share of formal sector employment in the total labour force has been contracting in most developing countries. In some low income countries, particularly in Africa, the number of people in formal sector employment has also fallen significantly in absolute terms. Data is generally not available, but the resulting increase in job competition has probably made it more difficult for women to gain access to non-traditional occupations and has, consequently, limited the role of more equitable training provision in redressing gender imbalances.⁽¹¹⁾ In contrast, in the developed industrial economies, increased education and training for women has enabled them to make major in-roads into rapidly expanding occupations since the early 1980s (see Box 4). With such marked differences in labour market conditions between developed and developing countries, great care must be taken in drawing lessons from the policy experiences of OECD countries.⁽¹²⁾

5. Training outputs and impacts

5.1 Is there a poverty reduction crisis?

To what extent are the disappointing outputs and impacts of training interventions in support of the poor symptomatic of a much wider problem, namely the failure of government and NGO efforts to reduce significantly the level of poverty in most countries? Surprisingly, however, most of the literature on training for the poor makes little or no effort to situate this particular intervention in the wider context of the overall efficacy of poverty reduction strategies. By focusing exclusively on training provision for the poor, there is, in fact, a danger of losing sight of the general shortcomings that have characterised poverty reduction programmes world-wide.

Two sets of reasons for low impact can be identified, namely weak institutions and the lack of 'voice' among targeted beneficiaries. A number of key findings and lessons have emerged from recent reviews of poverty reduction policies and strategies.

· Bureaucrats usually have a strong incentive to target just below the poverty line in order to maximise the number of individuals who are lifted out of poverty. Programmes that use communities and other locally based groups generally have much better records in targeting the poorest.

· While it is commonly believed that the provision of credit as part of income generating projects has been widely effective, there is little hard evidence to substantiate these claims. According to Subarrao et al "the evidence for sustainable income generation is only strong for the Grameen Bank (in Bangladesh)...The evidence from other programs is not so strong…the more successful programs are extremely small in outreach. If the opportunity costs of donor funds was also taken into account, many programs (including the Grameen Bank) would not yet be economically viable" (World Bank, 1997:89). Similar concerns have been expressed by Buckley with regard to microfinance schemes in SSA (see Buckley, 1997).

· Regardless of overall political commitment to reduce poverty, few central government agencies have the necessary skills and orientation to foster continued interaction with a wide range of small and frequently scattered beneficiary groups.

· Most delivery mechanisms have serious weaknesses. This results in services that are "unavailable to the poor, neither needed nor desired by the poor, captured by the non-poor, of low quality, unsustainable, cost-ineffective, or delivered more slowly than necessary to respond to urgent needs" (ibid:93).

· The use of social funds can be very effective when they are strongly demand-driven (by groups and communities), rely heavily on non-state actors for the delivery of services (including training), and are autonomous from line ministries. However, this rarely happens, especially in low income countries.

· High transaction costs incurred by the poor prevent them accessing a wide range of services that are made available. Travel time and costs are a major issue in many poverty reduction programmes.

5.2 Public sector training

5.2.1 Training outputs

Despite a chronic lack of supporting evidence, most training for the poor provided by public sector training institutions has been widely criticised for being inaccessible, irrelevant and of poor quality. These limited training efforts, it is argued, have been based on a simplistic modernisation paradigm which, drawing heavily on human capital theories, identifies the skill deficiencies of individuals who are poor and disadvantaged as the key constraint which, once addressed, will result in major increases in productivity and incomes. Thus, the belief in individual agency - the ability of each individual to overcome their state of poverty - is of central importance.

Training projects and programmes for the poor have generally replicated the policies and practices of supply-driven training for the formal sector. As a consequence, they have had the following characteristics:

· A largely 'top-down' process of skills transfer with little or no involvement of trainees in the identification of training needs and the design of training programmes. As passive recipients, there has been little sense of ownership and, by failing to recognise the knowledge and skills of the poor, training has been a disempowering, even "infantilizing" process.

· Attempts to forge training partnerships with other organisations and groups have been rare. There have only been limited attempts to provide training as part of an integrated package of services. Poor communication and frequent "turf wars" among government ministries responsible for these services are endemic in many countries.

· Most training has been delivered at training institutions. In common with the mainstream programmes of these institutions, there has been a heavy emphasis on longer duration pre-employment courses for unemployed youth and other disadvantaged groups, especially the disabled. In many countries (particularly in SSA), governments have preferred to establish a parallel network of mainly rural based training institutions specifically intended for training for self-employment (e.g. Youth Polytechnics in Kenya, Youth Training Centres in Zimbabwe, Brigades in Botswana).

· Traditionally male-dominated artisan training courses (plumbing, metalwork, carpentry etc.) have predominated in most countries. Training in social and business skills has been fairly limited. In her review of programmes of assistance for women entrepreneurs in Africa in the early 1990s, Kuiper concluded that the acquisition of business skills was a "seriously neglected area... the skills taught remain limited to vocational skills and, in the case of women's groups, organisational skills" (Kuiper, 1991:61). Lack of business training has had serious consequences for income generating projects. Other authors have come to similar conclusions (see Bakke-Seeck, 1996; Burckhardt, 1996).

· Fees drive away the poorest. In Ghana, for example, even at government-funded vocational training centres in remote rural locations, the majority of students come from relatively well off urban backgrounds (see Bennell, 1998). Given the massive excess demand that exists in most countries for post-school education and training, training intended for the poor and other disadvantaged groups is invariably 'captured' by better qualified school leavers.

· The provision of short courses for MSE operators and workers remains very limited. Instructors usually have little or no understanding of the problems of doing business in the informal sector. Governments have often created parallel organisations for the development of MSEs (especially Enterprise Development Institutes), but these tend to cater for non-poor clienteles. Training targeted at individuals, (especially women in survival enterprises), has been the exception rather than the norm.

· The design of training projects, especially in post-conflict situations, has tended to be too rushed so that planning ends up being carried out simultaneously with implementation. The ILO-supported 'Start Your Own Business ' Programme in Mozambique is a good example. Training objectives were seriously underfunded and insufficient information was collected in order to able to identify properly training needs in local labour markets. Serious shortages of trainers in rural areas meant that training was generally of poor quality and the skills acquired could not be effectively utilised mainly due to lack of credit and market opportunities (see Bryant, 1997).

· In SSA, training programmes for the poor have often been the result of donor initiatives. As separate projects and programmes with their own funding and management structures, they have rarely been effectively institutionalised on a sustainable basis.

5.3 Training impacts

Despite the lack of evidence, it is widely argued that the impact of public sector training for the poor has been minimal in most countries. Typically, unit training costs are relatively high with small enrolments and low completion rates. The intensive involvement of international experts in many projects has made them especially expensive. In youth training programmes, relatively few trainees have become self-employed. In Nigeria, for example, by the early 1990s, only 2 per cent of the over 100,00 apprentices trained through the government's Open Apprenticeship Scheme had managed to start their own businesses mainly because of the high cost of equipment (see Gallagher and Yunusa, 1996). In Zimbabwe, only three per cent of students graduating from Youth Training Centres in the early 1990s became self-employed (see Bennell, 1992).

There is a broad consensus that while smaller training programmes aimed at groups facing only moderate problems in the labour market have been found to yield positive results, broad and untargeted interventions have been universally ineffective.

5.3.1 School-based VET

There are widespread concerns among educationalists and education economists, especially in the donor community about the efficiency and effectiveness of specialised school-based VET. However, the political appeal of this type of training provision endures. With the chronic lack of institutional capacity to provide post-school VET coupled with pervasive concerns that formal schooling is too academic and engenders 'inappropriate attitudes', it appears quite sensible and rational to try to impart key vocational skills while children are in school.

Table 3 Vocational secondary school enrolments as a percentage of total secondary school enrolments in developing countries, 1980 - 1992/LYA.