9. In this chapter recent documentation on labour market signalling and manpower needs forecasting is reviewed critically. The primary documentation under review dates back no further than 1989 although secondary references are included dating back to 1970. It concentrates on the outputs of the World Bank and ILO since 1989 and includes summary reviews of some key texts, tracing the analytical work which has led to important changes in perceptions and expectations of labour information, as long-standing manpower requirements forecasting/techniques have become discredited.
Changes in the perception and expectations of labour market information (signals and indicators)
10. There are three major issues related to the approach to labour market information which dominate recent international literature on the subject. Firstly, there is the long running dispute between the protagonists of the Manpower Requirements approach - a technique which is still highly regarded by many planners around the world - and labour economists who support the more recently developed signalling methods, designed to produce indicators more related to quality issues than for use in quantitative forecasting.
11. Secondly, there is the problem faced by the laissez faire perspective adopted by countries which still rely on a supply-dominated approach to the preparation of labour market entrants. Once denounced as the major cause of educated unemployment, recent research is tending to favour a balanced understanding between the realities of labour market demand and the obligations of governments to respect social expectations.
12. Finally, a review of the available documentation clearly emphasises the need for training at all levels of the labour market information system. It is not only the data gatherers who need professional enhancement but also the analysts, those responsible for storage and retrieval of information, and, last but by no means least, the end users, be they government macro level planners, politicians, local government officials, enterprise managers, trainers or educators, who need to know how to interpret and use the assembled information.
13. Also evident from the review of relevant literature is the prominence of labour market experiences quoted from three developing countries: Indonesia, Malaysia and Tanzania. While other countries (Mauritius, Philippines etc.) may get passing mention, the country acknowledgements in both the text and the reference lists indicate that most of the prominent labour market economists reviewed here have worked in one or more of the three countries.
Manpower requirements v labour market signalling
14. The term "manpower requirements approach" implies the use of statistical inputs from economic sources in order to establish present and future demand for manpower by educational and occupational categories. This data is then compared with the output expected from the supply of manpower to the labour market - the difference indicating either a surplus, with a consequent need to cut back on education or training in selected subjects; or a deficiency requiring further investment in relevant areas of manpower development (Godfrey 1991). This approach was widely advocated during the 1960s and 1970s and became an immediate favourite with many planners, because of an initial impression of respectability, due in part to its overt reliance on statistical data and an apparent logical process of deduction.
15. Manpower forecasting became a number crunching exercise, occupying the time of large numbers of economic statisticians and social researchers, with claims that indicated an ability to determine the manpower requirements of each occupational area for a particular city, region or country - with an accuracy calculated in some cases down to tens or even units (see Indonesian Education Sector Review 1986, and Philippine National Manpower and Youth Council regional manpower plans as recently as 1991). Such forecasts were, and still are used to influence labour and educational policies at national and/or regional levels and played a major role in justifying budgetary expenditures (Clark 1986).
16. As early as the 1970s, voices were raised in concern over the "slavish adherence to unreal assumptions" (Blaug 1970) but still the practice gathered momentum. By the mid 1980s many more economists were raising alarm signals that the manpower requirements approach was neither accurate nor reliable (Dougherty 1983, Psacharopoulos 1983, Amjad 1985, Hollister 1983 & 1986, Clark 1986). It was found that the approach "distorted the true nature of manpower problems facing the developing world - i.e. the provision of sufficiently productive employment to an increasing number of the labour force consequent on population increase" (Hasibaun 1986).
17. It was also too conceptually restrictive, "failing to recognise that organisations with a functioning feedback system could and did identify shortfalls, taking effective action to overcome them" (Hasibaun 1986). Nevertheless, despite increasing evidence against the technique, the World Bank found that "most [of its] requests for education loans and most project appraisals were based on the grounds of manpower requirements" (Psacharopoulos and Woodhall 1985). Even as late as 1991 some National Manpower and Youth Council (NMYC) officials in the Philippines were arguing that the only thing wrong with the numerical approach to manpower forecasting was the NMYC's previous inability to process the available data fast enough to provide uniform, coordinated recommendations. Hence the inclusion in the World Bank VTP II Project of requests for a mini-computer and large numbers of fast desktop computers for use by regional manpower offices.
18. Reasons for the gross unreliability of manpower requirements approaches mainly stem from the inadequacies of its data base. Labour force surveys, the mainstay of the approach, are dependent upon relatively untrained and non-specialist enumerators. Even so the data itself is too aggregated, with much inevitable confusion over the demarcations of occupational classifications and skill levels (Godfrey 1991, Richter 1992). Little agreement exists on the use of acceptable definitions for employment categories such as "supervisors", "technologists", "technicians" or even "skilled workers" (ILO 1991).
19. But the final obstacle to the success of the approach comes from an increasing recognition that most medium and large scale companies operate an internal labour market, preferring to hire unskilled workers and to train them on the job. Increasing recognition that this practice is widespread and preferred by most employers makes the traditional manpower requirements approach redundant. Instead, it highlights the need for broader, more effective general educational preparation rather than the provision of specific, pre-employment vocational skills (Clark 1986, World Bank Policy Paper 1991).
20. The need for a "new approach to manpower planning" emerges in a World Bank Staff Working Paper (Psacharopoulos et al 1983) and in ILO research (Richter & Mason 1985, Richter 1989, Amjad et al 1990). Central to the new approach is the recognition that "planning" is a misnomer, conjuring up thoughts of central mobilisation and disbursement of resources (ILO 1991). "The art of manpower planning is certainly in disarray - 'planning' is becoming a word to avoid" (Psacharopoulos 1991). In place of manpower "planners" there is a new emphasis on manpower "analysts" who are more concerned with equity and efficiency (Richter 1989).
21. Using a variety of systems designed to signal or highlight significant trends in the labour market at either national, regional or locals levels, the resultant indicators are used to make decisions which improve the efficiency and efficacy of the supply side, and which have beneficial influences on the performance of the labour market or the demand side. To achieve this balance requires the use of methodologies designed to take into account such diverse issues as social benefits, educational investment, and training costs, all the while ensuring that the results are responsive to changing economic conditions (Middleton, Ziderman & van Adams 1993).
22. Recognition that movements in such items as wages or terms of employment can serve as important signals of change in the demand and supply of skilled labour was to gain prominence in the mid 1980s. This recognition also established the importance of observing "trends" rather than attempting precise quantification (ILO 1988). More importantly, labour market signalling focused on education and training qualifications rather than occupational classifications, which had been the main concern of old-style manpower planners (Amjad et al 1990, Godfrey 1991). The need for this change in emphasis was supported by rapid technological advancements which made skills more important than occupations in a rapidly changing workplace (ILO 1988). The weakness of the link between skill specialisation and occupational placement, which was the main purpose of the manpower requirements approach, was clearly demonstrated by reverse tracer studies undertaken in the Philippines (Arcelo & Sanyal 1987).
23. In contrast to the previous emphasis on occupations, the new labour market signalling techniques are concerned with economic outcomes measured in terms of wages and employment which can be compared with the costs of specific education and/or training programmes involved (Hinchliffe & Youdi 1993). This provides a basis for determining desirable levels of public and private sector spending on education and training, while the results can be directly linked to productivity (Middleton, Ziderman & van Adams 1993).
24. The identification and interpretation of labour market signals requires a basic understanding of the analytical processes which can be applied to labour market supply and demand. It also implies the availability of reliable labour market data for:
a) guiding private training decisions (basic information required includes - wage rises, job vacancy rates, graduate placements, enrolment data);b) managing training systems (basic information required includes - reverse tracer studies, cost comparisons of alternative paths, rates of return, measurements of cost effectiveness, experiences of new industries, productivity rates);
c) improving labour market efficiency (basic information required - wage levels, non-productivity-related wage differentials, evidence of labour mobility, public subsidies for training, labour market segmentation, labour codes, obstacles to wage flexibility); and
d) planning public investments in training (basic information required - private training capacity, balance of skills demand and supply, market imperfections, structural changes) (van Adams, Middleton & Ziderman 1992).
25. Household and establishment surveys provide the major source of modern day labour market information but these may need adaptation to suit the specific requirements of manpower planning (Richter 1989). Tracer studies, particularly reverse tracer studies of employees and trainees, can be designed to meet specific data requirements. Studies to identify training or educational costs can also be commissioned. CEDEFOP makes a strong case for setting up a network of longitudinal studies, although recognition is given to the difficulties of accomplishing this in developing countries (CEDEFOP 1993). However, a more permanent information pipeline is needed for those responsible for short term planning and policy-making such as college principals or training managers.
26. Promoting local linkages and information networks between suppliers and employers is essential (Amjad et al 1990, Godfrey 1991). One example of such a local system is based on "key informants"; people who can be used to regularise the data flow. Key informants are selected persons whose occupation, position or responsibility, in the public or private sector, give them intimate knowledge of local manpower and employment trends. Identifying such persons and designing a system of structured questionnaires and interviews which are to be conducted at regular intervals, produces a low cost but effective and responsive labour market information system (Richter & Mason 1985, Richter 1991).
27. Another similar technique which can also provide regular data relevant to labour market analysis is the "industrial targeting strategy" in which particular enterprises are selected to be representative of a particular industrial sector or sub-sector. Each selected enterprise is encouraged to make regular returns of selected data from which overall trends and indicators can be deduced (Hollister & Kam 1990).
28. Although most international institutions favour the new and more flexible approaches to manpower data with its emphasis on analysis rather than planning per se, there remain those who still argue that the manpower requirements approach is not entirely negative.
Given better resources and a fresh start, it is argued, there are good results which may be obtained from the quantitative planning approach (Colclough 1989). Others suggest that since so many planners still adhere to its principles, some attempt must be made to salvage what is usable in the technique and incorporate those parts into the less dogmatic signalling process (Godfrey 1991).
Market forces and un-checked supply driven systems
29. Over the past decade there has been much criticism of countries which allowed their educational systems to concentrate on social expectations rather than specific labour market demand. The prevailing parental view that blue collar jobs are of lesser value than white collar jobs has been blamed for a surfeit of over-qualified but unemployed (or under-employed) workers throughout Africa, Asia and Latin America (Muqtada 1991) Public educational systems frequently offer subsidised programmes at various levels of post-secondary education, irrespective of employment opportunities. Evidence of Sri Lankan taxi drivers with PhDs (Kelly & Culler 1990) or Pakistani polytechnic graduates remaining unemployed five years after graduation (Asian Development Bank 1989) are quoted to show the dangers of an educational system which fails to take cognisance of labour market signalling.
30. Society's perception of the relative value of different forms of education does influence government policy (Dougherty 1987) and there are many examples of governments which continue to operate a laissez faire policy relating to tertiary education enrolment, mainly because voters demand subsidised opportunities to enter non-vocational programmes at colleges or universities (ILO 1991). Such countries have no system in place by which to gather or analyze labour market information because they see no need for such data (Asian Development Bank 1994). But social expectation is not the only cause of supply-driven educational systems. Countries following all or some of the paths of centrally planned economies have often regarded all levels of education as a "right" for all, or at least as many as the system could stand, irrespective of employment demand. "Equity" has become a justification for much expansion in educational buildings, equipment and facilities (Mazumdar 1989). World Bank labour market economists making their first visits to China were surprised to find the ultimate supply-side system, where local state controlled enterprises were told how many graduates from vocational schools, technical colleges or universities should be taken on as employees each year (Noah & Middleton 1988).
31. ILO identified three main reasons for the unquestioned support given to investments in human capital formation, particularly by governments in South Asia:
a) social benefits - including externality and spill-over effects;b) enhanced employment opportunities and an apparent efficient utilisation of manpower; and
c) the belief that the private sector would shy away from financing certain aspects of education, specially in countries with low levels of social infrastructure (ARTEP 1991).
But research has shown that supply driven systems of education and training have led to social unrest when graduates fail to find employment which match their expectations (APSDEP 1992); that large numbers of over-trained graduates represent a very expensive and inefficient use of human capital; and that the private sector will respond (albeit sometimes slowly) if market forces are left to operate freely (Dougherty 1987). Some countries have blamed the acknowledged inadequacies of previous manpower planning techniques for their failure to include elements of labour market demand in their educational and training policies (Lucas 1991).
32. Unfortunately, as in so many aspects of educational policy, the pendulum is now swinging too far in the opposite direction. There is the current view, widely held in South-East Asia, that casts any form of central planning or control of educational investment as a left-over of socialist dogma, which should be dismantled at all costs. It is therefore very difficult if not impossible to convince some governments of the need for integrated human resource development (HRD) planning (Muqtada 1991).
33. Much evidence has now been amassed which supports the importance of external market forces and downplays the role of planning. In the Bombay labour market it has been found that it is the size of the firm which is the major determinant of wages, not the educational attainment of individual workers (Mazumdar 1989). Deliberate expansion of education and training investment to attain a shift in comparative advantage has been seen by some to be a misplaced concern. No causal relationship can be established, although some country examples appear to show associational relationships (Berhman 1990).
34. Even if market forces are recognised as an important element in developing human capital, there is still a need to identify the trends, prepare the indicators, and hoist the signals in a way that will be useful and valuable to both suppliers and end-users of manpower development. This means that there is a strong role for some form of labour market analysis which seeks to reconcile supply and demand (Godfrey 1991, ILO 1991). While there may be no conclusive evidence on whether manpower planning and economic growth are causally related, several studies show a strong associational relationship between the two. "The recent contentions on HRD planning are related more to what constitutes a balanced and meaningful HRD strategy, and what role the public sector should play in the creation (supply) and utilisation (demand) of manpower" (Muqtada 1991).
35. In the search for a balanced perspective, the World Bank policy paper on vocational education and training (1991) has come in for strong criticism from some quarters for failing to reconcile the opposing requirements of social equity and efficient productivity (ARTEP 1991). Based on the idea that maximum effectiveness and efficiency can be obtained by providing training to those who are already in employment, the policy paper appears to hold little prospect for the many millions who have yet to find their way into the formal labour market. Labour market signalling must surely hold the key in efforts to redress the balance between supply and demand (Mazumdar 1989).
Training those concerned with labour market information processes
36. The shortcomings of the largely discredited manpower requirements approach were in part attributable to the absence of reliable data and the inability of those concerned either to gather or to analyze the necessary information (Godfrey 1991, ILO 1993). There is a danger that similar problems could beset the current labour market signalling techniques unless there is a substantial investment in preparing those tasked with assembling, storing, and retrieving relevant data, analyzing the contents to determine significance, and interpreting the results for appropriate policy formulation (Richter 1989, Godfrey 1991, Muqtada 1991, ILO 1993). "Manpower planners need to be firmly grounded in analytical techniques such as multiple regression analysis, survey research methods, and basic inferential statistics in order to develop and analyze those systems" (van Adams, Middleton and Ziderman 1992). It is necessary "to establish a viewpoint or a broad understanding of the main issues in employment and manpower information. Such an understanding of the issues is important for rational decision-making and information in the formulation, implementation and monitoring of policies" (ILO 1990).
37. In a world where the economic and social environment is constantly changing and evolving, analytical work is essential for the development of sound educational and employment policies (ARTEP 1991). However, the work must go beyond mere written descriptions of the past or the verbal depiction of previous trends or policies. It must also provide detailed qualitative and quantitative examinations of the constraints. It must analyze the relevance to current or future policies, assess alternative strategies, and suggest alternative policies deemed more appropriate to future needs (Psacharopoulos 1988). Some detailed proposals for analytical tools relevant to determining labour market indicators for use in the education sector were set forth by Mingat & Tan (1988). However, the need to expand the repertoire of techniques has been stressed (World Bank 1991), particularly the urgent requirement to develop techniques for rapid data gathering and assessment specifically designed for short term use in the field. Such procedures should be designed for rapid assimilation by relatively inexperienced enumerators or field staff in developing countries (Hurley & Richter 1991).
38. "As a contributor to the development of a national system of employment and manpower information which will have to serve a wide variety of purposes, you will have to be concerned with the content and uses of the products of such a system, and you will be concerned with the organisational and technical questions involved in order to improve your knowledge and skills" (preface to Employment and Manpower Information in Developing Countries - A Training Guide, ILO 1990). The need for training is well recognised: some of the tools have been assembled. Putting such training programmes into effect remains a major task.
39. The OECD research and development centres have been exploring the impact of economic and technological changes on the relationships between education and the economy. In particular, the effects of market liberalisation are recognised as creating fundamental changes in the supply (education/training) and demand (employment) sides of this relationship. The processes of economic and technological reform are changing demands. New needs are for more flexible working structures and for HRD policies which emphasise on-job and just-in-time training (Turnham 1993).
40. However, on the supply side changes are much more sluggish. The formal education and training institutions are inhibited in responding to changes in demand, largely but not entirely because of government education policies which constrain responsiveness at local levels, a problem which cannot readily be addressed (Bengtsson 1993, Turnham 1993). Turnham's view is that national "policy makers rarely take radical action unless forced to do so by unavoidable pressures emanating from inside or outside the system". The liberalisation of global markets is providing some of those pressures on the demand side. Inappropriate models of the supply-demand relationship make relevant interventions difficult. In consequence, Bengtsson points to the migration of traditional supply-side functions - training and scientific and technological research - away from educational organisations to the major employers. Conversely, the supply functions of public sector agencies become diffused when they are used as "employers of last resort" to absorb high levels of graduate employment (an important reason for INA's decline in Costa Rica - see paragraph 113). This then draws attention to the sclerotic nature of those organisations, and to forces as part of liberalisation policies to privatise (or abolish) their functions.
41. The conclusions are less than optimistic. Turnham points to the powerful forces resisting reform, and the unusual circumstances supporting reform in the successful East Asian economies. Bengtsson looks to a revival of interest in earlier models of lifetime learning, driven by employers' HRD policies and a more efficient training market in which both public and private providers compete. Implicit in both arguments is the need for the kinds of LMI which will inform decisions by individual workers and employers as well as government departments, will counter resistance to reforms by those who have most to gain from them, and will enable the efficient operation of national and local markets in training and education.
The current status of LMIS
42. Richards & Amjad (1994) bring together recent research and current thinking on labour market information systems. Editors of conference proceedings always have a hard task to achieve sequence and coherence; in this case the difficulty is reinforced by the apparent uncertainty of contributors as to the nature and application of manpower planning and analysis. Nonetheless it illuminates the need for some kind of decisive, tidy, user-friendly guide for actual and potential practitioners.
43. Martin Godfrey in his paper entitled Planning for vocational education, training and employment: A minimalist approach, makes the point that "what is needed, by students, their parents, teachers, trainers, jobseekers, employees and employers, is reliable and timely information about the state of the various labour markets for qualified people...Meeting these information needs will be one of the main functions of planners in this field." Labour market analysis is not in fact an arcane preserve of central government. In his introductory paper on Issues in Manpower Analysis the-co-editor, Peter Richards, points to the provision of information as an important government role. "That forecasting is essential in setting up training courses and facilities does not necessarily imply that governments must undertake it. Governments are, however, uniquely placed to collect and analyze information... The government can pass on to employers, and indeed make public in general, its assessment of where the economy is leading and which activities are likely to be favoured."
44. The point comes up again in the paper by Robert Lucas on Manpower planning and economic development. "... if the State genuinely possesses more complete information about future conditions (over which they exert some control), and is either unwilling or unable to disseminate this information to the public, then a case can be made for state intervention in guiding skill acquisition..." Nonetheless, he continues, "for the state to select fields of specific vocational education... is highly risky".
45. Godfrey gives practical, if basic, advice on the collection of information. "This may involve... baseline studies, but, initially at least, will be more a matter of talking informally to employers, trade unionists, researchers, journalists, educators, trainers and government officials..." He stresses this later by saying that "[planners] have to get out of their offices and develop a 'feel' for the relevant training programmes and labour markets..." In this respect George Psacharopoulos (World Bank) would probably disagree. In his paper From Manpower Planning to Labour Market Analysis he is sharply critical of key informants. "These are people who are supposed to know a particular sector inside out... the fact remains that the information they provide is based on their opinion. It does not, and cannot, represent an exact description of hard facts about the future."
46. Godfrey also, in effect, provides a reasoned check-list of components of labour market information: unfilled vacancies, unemployment, expatriates, students going abroad, applications for training places in relation to the number of places, the impact of training on earnings over time. This is in fact the only systematic description in the book of what LMI might comprise. A session on LMI in the 'Proceedings of the Workshop' did not, according to the brief report in this book, deal with this point.
47. In his second paper, on Information needs for the transition to labour markets, Godfrey discusses employment and unemployment information in Hungary, and describes a pilot survey carried out in China in 1987. Although under central planning in Hungary huge volumes of information were collected, it was not in a form which made multivariate analysis feasible. The Chinese survey combined establishment details with records of individual workers. Tiny in Chinese terms, it covered 109 enterprises and 36,000 workers; and it was learnt that fewer workers needed to be analyzed, but the number of enterprises needed to be much larger.
48. George Psacharopoulos points out that enterprise surveys cover only the small fraction of the population [in developing countries], which works in modern establishments. "The best place to obtain... information is in the household." This is one of 17 "Do's and Don'ts" in his paper From Manpower planning to labour market analysis. Inter alia he also favours wage measuring, tracer studies, educational profiles (instead of occupational categories), and cost-benefit analysis. Manpower planners should consider questions of equity and poverty as well as production efficiency, and economic efficiency as well as technical efficiency. Training should be general rather than skill-specific. In this respect he cites a US Department of Labor report of 1991; this indicates "five competencies" in which secondary school leavers should be proficient: "ability to allocate resources; interpersonal skills; ability to assess information; ability to understand work systems; ability to deal with new technology". This matches Lucas's point "that the more successful vocational schools have actually emphasised the academic portion of the curriculum..." But both contributors seem to duck the difficulty that education and training become continuously more specialised: even the choice of subjects for GCSE contains the germ of an occupational area. Be that as it may, Psacharopoulos follows by preferring firm-based training, cost recovery (rather than free education and training), private institutions; and finally short-term time horizons, a gradual and marginal approach, and the correction of labour market distortions.
49. When Godfrey comes to planning training in his first paper, he separates programmes with short lead times' (which he seems to equate with short programmes) from those with long lead times. "To some extent the planning of short programmes can be handed over to those who run them"; but an elaborate six-step process is indicated for long programmes. He then has to defend this against the possible objection that the process has "a family resemblance" to "the manpower requirement forecasting approach criticised earlier". It might also be objected that the division of programmes into those with short and long lead times (or short and long) is to say the least an oversimplification. This is a surprisingly weak part of an otherwise very helpful contribution, but he does at least face up to a key problem.
50. Two further papers discuss African experience of manpower planning. Eleazar C. Iwuji holds that it has been ineffective in Sub-Saharan Africa, not being trusted or understood; there is little communication between manpower planners and educational planners. (This is a valuable point, to be remembered in the design of any future programme.) Christopher Colclough, writing on Botswana, shows that skill shortages have become more intense in the period 1965-90 and asks whether this implies failure of manpower planning. He bravely concludes that "the story is by no means one of unmitigated disaster", but it is certainly a cautionary one for us. Manpower planning makes a fool of itself when over-ambitious; but labour market information is still an indispensable basis for plans of many kinds.
Conclusions
51. The current aim of most governments is to make education and training more effective, an objective which requires new policies and new methodologies, as well as fresh investments in new facilities. But the first step must be to devise better policies calculated to introduce greater efficiencies than those which have been used before. To design improved methodologies, which can be more effective and relevant to the process of learning, will require new strategies backed by new analytical techniques.
52. A review of the priorities for upgrading labour market information has been recently undertaken by ILO in collaboration with the Danish agency DANIDA. Their joint report highlights the need for a better understanding of the process and use of labour market information, the need to establish key informant systems, the need to maintain a balance in which the social dimension is not lost or diminished, and the need to train trainers in the techniques of LMI generation, analysis and dissemination (ILO 1993).
53. The main conclusions of a four year programme of research and consultation on vocational and technical education and training by the World Bank's Human Resources Department point to the essential needs.
"The analytical base for training policy and investment has traditionally included an assessment of skills demand, most often through a manpower requirements forecast, and a plan for expanding skills supply, generally through the establishment of public training institutions This kind of analysis overlooks the more dynamic aspects of economic environment - among them the incentives for employers and workers to investing skills training. It also ignores the actual and potential contribution of employers and private training. Needed instead is a general framework that identifies the key questions and kinds of information needed to deal with both the economies of skills demand and the dynamics of training supply in the private and public sectors To set up such a framework, policy makers should try to answer three crucial questions. What is the economic context of employment and skills demand? How extensive and effective are the various forms of training? Are current training policies and planning practices able to adjust training supply to a changing economic context?....Labour market signals have become the key tool in planning" (World Bank 1994).
