Why is LMI collected?
What LMI is collected?
Who collects LMI?
What techniques are used for LMI collection and analysis?
Who uses LMI?
Who benefits from LMI?
What are the costs of LMI collection and analysis?
What are the training needs of collectors and users of LMI?
How is the LMI system currently being improved?
199. In this section, the major themes examined in the case studies are considered separately. These relate to the initial research questions posed in Figure One and Appendix Two. The four case studies have been compared with a view to drawing out significant commonalities and pointing up key contrasts between the case study countries.
200. Our findings are drawn from countries in three continents, ranging from less than 4 million people (Costa Rica) to over 118 millions (Bangladesh), with very different economies, levels of GDP and socio-political structures. We have taken care, therefore, to avoid broad generalisations wherever possible. Despite these enormous differences, it has, however, been possible to point to some issues concerning the operation of manpower planning and labour market analysis at local, regional and national levels, and to identify needs and issues concerning labour market signals and indicators which transcend national and international differences.
201. The reasons for the collection of LMI are diverse. No clear patterns were identifiable, except that different organisations collect LMI for different purposes. The intention is to use that information in formulating policies, whether by donor agencies, government departments, training providers, employers or labour organisations. There is strong evidence that LMI is collected as an instrument in government planning for industrial location, incomes policies and the reform of educational and training systems - but equally strong evidence that this information is far less significant in decision making than the influence of political concerns and pressure groups. Because the purposes are so diverse, much data collection is duplicated, and there are very few examples of data being shared by data-collecting organisations. Donor agencies collect LMI as a matter of course in order to justify aid projects. In each of the case study countries the major international aid agencies are instrumental in encouraging the collection of LMI, finding out employer needs before investing in aid programmes which meet those needs, but there are concerns that such information is not thereafter utilised in shaping those aid projects, and that it is not supported by follow-up studies, while the techniques for data collection and analysis are very rarely incorporated into regular government practice.
202. Most of the data collected is quantitative. Census data is used where available, but much effort is put into collecting LMI statistics for specific purposes. Unfortunately, this data is generally collected at too high a level of aggregation for it to be of much use. Its usefulness is further impeded by the lead times involved in aggregating, analyzing and publishing the data. Much of the data lies dormant before analysis, while other priorities absorb the time of the limited numbers of people with the requisite analytical skills. There are serious doubts as to the accuracy of this data, particularly as informants are likely to provide only that information which will support their interests or remain fearful of the true purpose of LMI collection and either avoid responding or give misleading information. In the case study countries, high levels of under-employment mean that there are few needs for LMI as a basis for industrial location policies - high unemployment and under-employment levels mean that industries tend to find their labour supply wherever they might be located, particularly in countries where migrant labour or expatriate labour is also readily available.
203. There is some evidence that the information collected by external agencies from companies is not particularly reliable. While large companies tend to respond to requests for LMI, small and medium enterprises are often less cooperative, particularly where the existence of payroll taxes presents a powerful disincentive for the accurate recording of personnel details. However, the confidential personnel data collected by large companies tends to be both more accurate and more heavily used than other LMI.
204. The evidence for the collection of 'soft' signals and indicators came mainly from the training organisations Industrial liaison staff who keep in close touch with employers acquire motley signals and indicators about new training needs, opportunities for delivering new courses, and technological trends which require changes in the training curriculum. The information is not statistically based and in consequence has little credence outside the particular departments where it is acquired. It is, however, valuable information about the relevance of the current training curriculum, pointing not just to the need for change but to ways in which effective changes might be implemented. 'Key informants' from business and industry also provide some labour market signals where they sit on advisory committees at institutional and governmental levels. Too often, however, this information is not drawn directly from the workplace and is tempered by the political and social concerns of the informant.
205. The collection of signals about job vacancies and wage levels, obtained for example through scanning the newspapers, is used to some extent. We witnessed its effective use in Zimbabwe, where it forms part of the indicators used by the Congress of Trade Unions in supporting its wage claims, while in Bangladesh it is used by BTEB as a signal for priorities in curriculum development. Otherwise, even where such data is acquired, there is uncertainty as to what to do with it.
206. Formal, quantitative labour market data is mainly collected by research units within government departments, and especially within Labour and Planning ministries. In each of the countries visited, we found evidence of several ministries separately collecting their own labour market information, without reference to each other and with much duplication of effort. The approaches are highly formalised, and are aggregated for the purposes of long-term manpower planning. The rivalry between Labour and Education ministries is a phenomenon of central government throughout the world (to some extent reinforced by the preferences of international donor agencies). With respect to LMI, this is compounded by the interests of governmental statistical offices and we found examples where the conflicting priorities and interests of these government departments impeded the effective collection and use of LMI.
207. As we have noted, there is often a significant time lapse before the publication of findings from data analysis. In some countries, specialist research units within the universities and specialist institutes are contracted by government departments and international donor agencies, as are consultancy firms, including, where they have local offices, the (expensive) major multi-national accountancy/consultancy houses. We found little evidence of speculative investigations to collect and then sell on information. As indicated above, the Zimbabwe Congress of Trade Unions collects its own information, as a basis for both pay negotiations and internal training provision.
208. Non-governmental agencies were identified as effective collectors (and users) of LMI. In both Costa Rica and Bangladesh, NGOs were the only organisations working effectively with the informal sector. They identify the distinctive needs of microbusinesses and disadvantaged groups such as the handicapped and female workers, notably developing strategies which avoid locking women into traditional (and particularly poorly paid) areas of work.
209. Some of the most effective examples of LMI collection and use were found within training institutions, where individuals had been released from training duties to liaise with employers and identify ways in which training might be made more relevant. The data collected was labour market intelligence rather than hard statistical data, but it had the double benefit of improving the curriculum and enhancing the credibility of the training organisation with employers.
210. The structure of advisory committees based either on individual training institutions or on industrial sectors (e.g. Cost Rica's camaras) provide a potential ready-made system of 'key informants'. However, as indicated above, these have their limitations in providing useful insights into labour market needs, as the information circulated for and at such committee meetings depends on the particular interests of the current industrial representatives on the committees. The chambers of commerce themselves collect some LMI, mainly to use in trying to persuade the government to provide or retain tariff protection, tax regimes or wages policies which favour their sector. However, they are unwilling to share their information with other data collectors. There is, in consequence, little discussion between analysts as to the relative effectiveness of particular approaches to data collection and analysis, and no cumulative body of experience and expertise which improves the processes of labour market analysis.
211. The most common investigative technique is the postal questionnaire survey, supplemented by census information. Telephone interviews with companies are employed in countries with good telephone services (such as Costa Rica). Newspaper scanning for employment news and job advertisements is undertaken in Bangladesh and Zimbabwe. The tendency is for the development of long and detailed questionnaire surveys (see Appendix Three) whose very complexity inhibits completion and return rates.
212. Data analysis demands computer power which was either lacking in the countries visited, or was not matched by the systems design expertise which could structure data analysis to meet specified requirements. The plea for better computer facilities (met in each of the case study countries) is understandable, but such provision is pointless unless the outcomes of such analysis are both accurate and useful. In each of the case studies (except Cambodia) staff are available with the technical expertise to process data using modern computerised technologies. Indeed, in Costa Rica these skills are in sufficient supply for the development of a rapidly growing computer industry processing data imported from North America. The weaknesses lie in the quality of the LMI data being processed, the time taken to complete the analysis, and - most important of all - the uses made of the findings.
213. Less formal 'intelligence' is gathered by training organisations particularly where they deploy staff to go out to employers, seeking information on new training needs, work placement possibilities, and technical developments within the industry which might influence the training curriculum. All organisations encounter such information. The most effective ones are able to make use of it to recruit more trainees, modify their curriculum and obtain industrial support and sponsorship. We found few examples of tracer studies being used, although they are commonly described as highly effective techniques.
214. Given the problems outlined above - inaccurate data, poor return rates, long delays while it is processed and high levels of aggregation in the published findings - it is not surprising that we found only limited examples of LMI being used to enhance the short and medium term responsiveness of TVET. The most effective uses were by organisations whose LMI collection and analysis had a specific and bounded purpose, as with the Zimbabwe Congress of Trade Unions for pay negotiations and internal training provision, and (a few) departments within training organisations to enhance the quality of their training provision. A key problem is that information is useful only in a political economic climate in which decisions are based on information.
215. As indicated above, training organisations use informal 'intelligence' for a variety of purposes. However, the overall findings across the case studies are that only a few enterprising institutions are able to make effective use of this information, and we found none making effective use of central government LMI. Despite a worldwide move towards greater institutional autonomy in public sector education and training, we found that institutions were inhibited in using LMI, even where it exists in each of the case study countries. In Costa Rica data was collected by departments within institutions, which then modified the curriculum in accord with the findings - but had to do so discreetly because they were operating semi-illegally by deviating from government set syllabuses. The use of such market intelligence was most effective where used by training consultants employed by international aid projects to develop new curricula and modify existing programmes. In these circumstances, the consultants and their local counterpart trainers work with the support of central government agencies and are not constrained by formal government syllabuses.
216. Large companies undertake and use their own internal labour market research in determining personnel policies including pensions schemes, recruitment strategies, the form and duration of contracts and in-house training provision. This information is confidential, and it tends to be both up-to-date and accurate, derived directly from personnel records and staff surveys. It is unfortunate that, because of this confidentiality, an important source of LMI is unavailable to TVET institutions and others who supply these companies with their labour.
217. The benefit most commonly cited of LMI at national and regional levels is in deciding whether and where to site new training institutions or faculties/departments within them. Such decisions are, however, so strongly shaped by political factors that LMI is by no means the major determinant. There are numerous examples (drawn from both the developed and the developing world) of TVET institutions located primarily for short-term political reasons, even where rudimentary LMI pointed to the impracticalities of such a location. LMI has a temporary beneficial effect where it is used (as at present in Costa Rica) as evidence in reforming the national education and training curricula. Resources are rarely released to support such revision unless there is evidence that LMI has been collected. However, these short-term improvements are soon countered by the detrimental effects of an inertia deriving from the lack of both resources and political will which 'freezes' the new curriculum until it is so out of date that a new bout of reform is demanded.
218. Where aid agencies investigate LMI, the findings are used (in Bangladesh and Costa Rica) to fund new training facilities including workshops and equipment in vocational training institutions developed to meet those needs. However, it is disappointing that the now obligatory labour market analyses, which precede investment by the major international donor agencies in TVET systems, are used primarily in order to justify the initial investment project. We found only a few examples where that LMI was used thereafter to inform the key decision-makers in the local system as to the main findings and their implications. Nor did we find many examples of the continued collection and use of LMI during the development projects in order to monitor their impact and adjust their direction and structure.
219. We failed to obtain useful evidence of the costs of LMI collection and analysis. In most organisations the real costs are not identified, particularly where the work is undertaken by in-house research units who also undertake a variety of other survey work. Even where the costs are known, as when external consultants are employed, this is confidential information, which was not made available to us.
220. The needs of collectors/analysts and users of LMI are very different. Those responsible for the organisation of data collection and analysis need both quantitative and qualitative skills, including the ability to identify those techniques which work best in particular circumstances. Even where research organisations can be equipped with requisite data collection and analysis skills, they also need to have the clout to be able to make use of their findings either through their own training provision or by commissioning appropriate training with relevant training institutions. Organisations with a national reputation and role, such as BIDS and INA, have the potential to provide this. In Zimbabwe the Central Statistical Office has its own training unit, while training is also provided by ARLAC.
221. End users of LMI need training in order to make use of available data effectively and - just as importantly - to influence data providers to collect the data that users need, and present it in ways relevant to the needs of the end-users. Such training is in part technical, including the ability to draw inferences from statistical data, but is very largely attitudinal. The shift towards greater institutional autonomy implies that institutional managers will enhance their skills in order to take advantage of that autonomy. These enhanced skills include marketing and curriculum development capabilities, which together enable TVET institutions to act more responsively to their main clients, and in particular the labour markets which absorb their trainees and graduates.
222. International agencies are the main sources of intervention for improvement. The African Regional Labour Administration Centre (ARLAC) provides support for African states and is based in Zimbabwe. It has reviewed the methods employed by the Zimbabwean government and recommended ways of improving both the collection and analysis of data. In Costa Rica a national process of economic review around the Science and Technology Plan is drawing attention to the needs for effective data collection and analysis, in a country which looks forward to becoming an information exporter, dependent on the data processing skills of its people.