APPRENTICESHIP IN THE BRITISH 'TRAINING MARKET'.

Ryan, Paul ; Unwin, Lorna

Paul Ryan [*]

Lorna Unwin [**]

British apprenticeship, now dependent on the Modern Apprenticeship programme, is compared in this paper to both German apprenticeship and its national predecessor, Youth Training. Modern Apprenticeship shares many of the attributes of Youth Training, and shows some improvement in terms of skills produced. However, British apprenticeship performs poorly, in terms of rates of qualification and completion, as well as in breadth and depth of training, relative to its German counterpart, despite the provision by Modern Apprenticeship of substantial government financial support. The fact that MA resembles YT more than German apprenticeship reflects continuing institutional differences between the two countries, notably the limitations of the training quasi-market in which both YT and MA have operated. The prospects for MA to flourish, let alone perform the educational role that the government envisages for it, are bleak in the absence of institutional development along different lines.

Introduction

Since 1994 a public programme, Modern Apprenticeship (MA, has sought to increase Britain's supply of in-termed ate skills by expanding work-based learning among young people. The programme is oriented towards craft and technician skills, and to the blending of on-the-job and off-the-job training.

The context is one of low skill supplies, and the damage done thereby to Britain's economic performance and social fabric. The national shortfall in productivity and trade performance, relative to Germany and France in particular reflects lower national inputs of skill, particular at intermediate level (Prais, 1995; Oulton, 1996; Mason, 2000). Moreover, the work-based path to intermediate skill, through apprenticeship, is associated with favourable school-to-work transitions, for both individuals and countries (Ryan, 2001a).

Such concerns nowadays permeate government policy (DfEE, 1999a). National targets have been adopted for the acquisition of qualifications by young workers. Progress towards meeting the youth targets has however occurred largely along the full-time, classroom-based route, general and vocational. The work-based route has thus far contributed little: youth employment has shrunk and many participants in labour market programmes have not gained the desired qualifications. An expansion of work-based learning could tap its greater appeal to many young people than that of fulltime, classroom-based learning (Green and Steedman, 1997).

Modern Apprenticeship is enigmatic, welcomed in principle but criticised in practice (Evans et al., 1997; Fuller and Unwin, 2001). The programme may be viewed from two perspectives. The content of its title suggests that MA should be viewed as apprenticeship as the institution is understood nowadays on the continent-i.e., as occupational preparation combined with vocational education. From that standpoint, it appears peculiar and defective (Steedman, this issue; Ryan 2000, 2001b). The capitalisation of its title suggests however that MA should be viewed as a labour market programme, part of the genus developed since the 1970s (Unwin, 1997). From that standpoint, Modern Apprenticeship appears an improvement within national mainstream. We find the 'programme' perspective more appropriate to understanding MA's functioning; the 'institution' perspective, to assessing its contribution to skills.

Our discussion contains four restrictions. Firstly, we concentrate on efficiency aspects, notably national skill supplies, and set aside the equity objectives as universal youth access, that have influence the programme's design. Conventional economic relating labour market outcomes to in MA, is debarred by data unavailability (Payne et al., 2001). Instead, we present evidence on the programme's operation, aspects typically slighted in economic evaluations (Grubb and Ryan, 1999). Secondly, given our focus on intermediate skills, we concentrate on what is now termed Advanced Modern Apprenticeship. Unless otherwise indicated, we set aside lower level programmes, which are currently being included in MA as Foundation Modern Apprenticeship.

Thirdly, given that MA has been reorganised frequently, a particular date must be chosen for its characterisation. We choose April 2001, when the Learning and Skills Council, the national funding channel for post-16 learning, began operations. Most of the data presented for MA below refer however to earlier years, when many changes had yet to be implemented. Finally, we use 'British' as a convenient descriptor for MA despite its inaccuracy. The programme's name and the availability of data differ between the constituent countries of the UK (Canning, 2001), and most of our data are limited to England and Wales. The content and organisation of the programme are however broadly the same across the UK.

The following section outlines the operational, and the third section the institutional, attributes of Modern Apprenticeship, as compared to those of apprenticeship in Germany at present and YT in Britain in the past. The Fourth section links the performance of MA to the quasi-market for public training services of which it Forms part. The fifth section concludes.

Scale and outcomes

A definition of apprenticeship that corresponds to contemporary ideals might be: a structured programme of vocational preparation, sponsored by an employer, juxtaposing part-time education with on-the-job training and work experience, leading to a recognised vocational qualification at craft or higher level, and taking at least two years to complete, after requisite general education.

It is difficult to measure apprenticeship in Britain according to this or any other definition. The absence of statutory regulation debars the continental option of basing the definition on legal status. British surveys of households and employers do indeed inquire about individuals' status in relation to apprenticeship, but the resulting estimates are weakened by the absence of any explicit definition, leaving respondents to fall back on their (undoubtedly variable) personal understanding of it. A second possibility is provided by Modern Apprenticeship, whose design incorporates several apprenticeship-related attributes. The programme supports work-related training programmes for 16-24 year olds that meet all of four criteria: orientation towards the acquisition of a work-based National Vocational Qualification at craft level or above (NVQ3+); attainments in two 'key skills' at Level 2 or higher; compatibility with any other sectorally determined requirements (e.g. other Key Skills or a college-taught vocati onal qualification); and a training 'pledge' signed by the parties, including the employer and the apprentice.

Neither approach to measuring apprenticeship in Britain is highly satisfactory. The Labour Force Survey, which asks respondents if they are taking a 'recognised trade apprenticeship', simultaneously excludes some individuals taking training that meets the conditions suggested above (e.g., it is not 'recognised'), and includes others whose training falls short of them (e.g. not preparing for any vocational qualification). Similarly, the MA headcount excludes apprentices whose employers remain outside the programme, and includes other trainees whose programmes would not meet any standard definition of apprenticeship. Nor do the LFS and MA estimates overlap heavily. In spring 1998, only 45 per cent of respondents describing themselves as apprentices also reported participating in MA (unpublished LFS tabulations).

The results of the two approaches are depicted in chart 1, along with those of a (now discontinued) survey that asked manufacturing employers to enumerate their 'apprentices and other long-term trainees'. The LFS and MA series differ for recent years, but the gap between them has been moderate and fairly stable since 1997 (chart 1). Nowadays, allowing for re-entry, roughly one n eight young British people enters apprenticeship, and apprentices account for slightly more than one half of 1 per cent of total employment (table 1).

Historic and cross-national comparisons show just how low these activity rates are. Apprenticeship declined during 1965-90 from 3 per cent to 1 per cent of manufacturing employment. It declined further during he 1990s (as a share of total employment), falling to around two-thirds of 1 per cent. The growth of Modern Apprenticeship after 1994 may have arrested secular decline, but macroeconomic recovery probably contributed too. The downward trend has not been reversed by MA, whose expansion petered out after 1997 chart 1).

Compared with Germany, the stock of apprentices in Britain amounts nowadays to between one-sixth and one-ninth of the share of 'employment', and the inflow to apprenticeship to less than one quarter of the share of the youth cohort, of their German counterparts (table 1). Germany does indeed have the largest national apprenticeship system, and its coverage of youth has varied considerably in recent years. Nevertheless, apprenticeship remains much smaller in Britain.

What about quality? The quality of work-based training is not easy for external observers to assess (see fourth section, below). Some indicators are however available. Judging by completion and qualification rates, Modern Apprenticeship again appears deficient (table 2). The average leaver participates for only seventeen months. Only one in every two leavers obtains the central qualification, an NVQ3, and even they spend only 21 months doing so. No data are available on completion of the Key Skills requirement, introduced in 1998, nor therefore on overall completion since then. As, however, Key Skills requirements were rarely being met in 1999 and have since been tightened (DfEE 2000e), the overall completion rate probably lies closer to zero than to one half. One-fifth of leavers emerges without any vocational qualification, even a lower level or partial one.

The comparison of quality to German apprenticeship again proves adverse (table 2). A large majority (78 per cent) of German apprentices completes training. German apprentices stay longer in training: on average, 32 months for entrants and 35 months for completers. Few German apprentices fail their final examinations, at least after re-sits: 76 per cent of entrants secure craft qualification, in contrast to only one half -- of a much smaller group -- in Britain. MA's qualitative performance is actually worse than this suggests, as German craft qualifications outstrip their NVQ3 equivalents in terms of technical knowledge (Steedman et al., 1997; Steedman, 1998).

We now compare MA also to its early 1990s predecessor, Youth Training (YT) -- also a work-based scheme without fixed training periods, but aimed at semi-skilled rather than craft skills. Quality indicators show many similarities between the two programmes (table 2). As in MA, YT participants left within a year on average, and only a minority completed -- though, at 43 per cent, that minority was almost certainly larger than MA's. Similarly, through 1998--9, both average training duration and the overall qualification rate had yet to exceed YT levels.

MA can however claim two gains over YT. Firstly, it has raised the modal qualification gained from semi-skilled to craft (Levels 1--2 and Level 3+, respectively). The proportion of leavers acquiring a Level 3 qualification rose from one-sixth in YT to just over one half on MA. The comparison is however loaded in MA's favour, given that (Advanced) MA comprises the upper tier of government-supported training, whereas YT included it all. When YT is compared to all work-based programmes in 1998--9, not to MA alone (table 2, cols. 4 and 5), however, an increase in the incidence of qualifications remains visible, particularly at Level 3+. MA appears therefore to have had a causal effect, and not simply selected towards Level 3 training. At the same time, the upgrading has been only partial. MA harbours much lower level training. The share of leavers gaining a Level 3+ qualification (in all programmes) rose only 7 percentage points between 1992 and 1998 (ibid.). Fully 41 per cent of a 1998 sample of MA participants reported seeking only a Level 2 qualification (Payne, 2001, Table 5.10).

Secondly, MA appears to have improved its performance over time. The share of leavers emerging with an NVQ3 doubled, from 27 per cent to 55 per cent, during 1997-2000 (DfEE, 2000a, Table 3). Average training duration rose by one half between 1998-9 and 1999-2000 (table 2). Two reservations must however be registered. Firstly, these indicators are expected to improve passively when a programme expands, as temporary adverse selection among leavers towards early departure and non-completion declines. Secondly, qualification rates rose strongly under YT too (ED, 1994).

Modern Apprenticeship is also highly heterogeneous. A contrast may be drawn between established industrial occupations, with their apprenticeship traditions, and service occupations, in which structured training had previously been rare. MA has had only limited success at reducing the disparity. Participants spend longer in training and more frequently obtain an NVQ3 in the industrial occupations (table 3). Customer services, hospitality and retailing represent the nadir: in 1998-9 an average of only seven months in training in customer services, an NVQ3 acquisition rate of only 16 per cent in both hospitality and retailing (Fuller and Unwin, 2001).

In Germany, training durations and completion rates also differ between sectors (table 4). Heterogeneity is however much less than in MA. A simple measure of heterogeneity proves much larger for the industry-services division in MA than for the industry-artisanal one in Germany (tables 3 and 4). We take the difference between the countries to reflect the difference in minimum training standards rather than the dissimilarity of sectoral definitions.

In sum, British apprenticeship under-performs relative to German apprenticeship, while improving on YT in terms of skills produced. The gap between MA and YT is however small for other attributes, and overall much less than might be expected, given the difference in the programmes' objectives.

Institutional attributes

We now extend the comparison of MA to German apprenticeship and YT to cover skills, governance and financial attributes. A tabular summary is provided in the appendix.

Skills

Some attributes of MA resemble their German counterparts. Both involve a set of collectively defined training occupations ('sectoral frameworks' in MA), with vocational qualifications pitched at craft level or higher. By contrast, educational content and process requirements are weak and absent, respectively, in MA but prominent in Germany, where apprentices must take part-time courses, general as well as vocational, at technical college, and spend the stipulated time in training before qualifying.

The weakness of MA's educational contribution reflects the centrality of the requirement for training aimed at a National Vocational Qualification. NVQs are intended to certify work-related competence, independently of how it has been attained. As such, they have encouraged MA's avoidance of general requirements for part-time education or minimum training periods, and its emphasis on practical skills rather than technical understanding. Moreover, NVQ assessment methods, which rely on the judgements of workplace supervisors, may not even measure practical occupational skills reliably (Eraut, this issue; CBI, 2000). MA's emphasis on NVQs has thus facilitated short stays, low completion rates and rapid completion.

In practice, MA never adhered fully to the methods of the NVQ 'competence revolution' of the 1980s (Wolf, 19 95). From the outset, providers have been expected to provide participants with some off-the-job training, and a requirement for a technical qualification could be inserted at sector level. The expectation can however be met in some occupations simply by providing practical training in another part of the workplace, and even that has been ignored by some employers (Winterbotham et al., 2000; Kodz et al., 2000). A requirement for technical education was adopted for only two engineering frameworks.

The skills contribution of MA has been correspondingly attenuated. Two correctives have been adopted: the 1998 requirement for Level 2 achievements in two Key Skills (communication and application of number), including their impartial external assessment, and the impending requirement for training towards a Technical Certificate, distinct from NVQ3, to demonstrate relevant 'underpinning knowledge'. The government even proposes to develop apprenticeship as part of a ladder of vocational qualifications which both part-time and full-time students may climb. Educational requirements have therefore emerged under MA, indicating that government is finally attending to the educational potential of work-based learning (Green, 1997; Steedman et al., 1998; DfEE, 2000c, 2000e; Blunkett, 2001).

From an educational standpoint, however, these changes are limited. Level 2 Key Skills represent the prior achievement expected of entrants who hold GCSE passes at grades A-C, the traditional requirement for entry to apprenticeship. The content of most Technical Certificates has yet to be determined, but the employer representatives charged with their introduction may well permit informal training at work (e.g., computer-guided study) to suffice in some occupations. Moreover, both requirements can easily be evaded by non-completion.

YT differed from MA in lacking occupational foci and in requiring training to aim at Level 2 only. Both programmes have however been designed around NVQs, emphasising practical skills rather than educational development, and avoiding process requirements in training.

Governance

British-German organisational differences are marked. A fundamental difference is regulation by statute law, present in Germany but absent from Britain, where the alternative provided by default by common law is only partial and obscure. MA functions instead under 'leaflet law': ministerial powers, legislated in the 1970s, to modify labour market programmes such as itself and YT (DeiBinger, 1996; Deakin and Morris, 1998). MA was simply announced in the 1993 Budget and then introduced without any parliamentary debate on the reform of an institution whose antecedents stretch back centuries.

MA shows some administrative resemblance to German apprenticeship. In both countries, responsibility falls to the national education ministry and detailed administration relies on a variety of intermediate committees, organised along cross-cutting sectoral and geographical local lines. The UK's 73 National Training Organisations, which specify MA training frameworks, and the 47 Learning and Skills Councils, which fund youth learning locally, resemble formally the sectoral committees and local Chambers that administer apprenticeship in Germany. These bodies draw on the services of employer representatives in both countries.

Major differences are however present in the composition and powers of intermediate committees. Significant representation of employees and vocational teachers, alongside employers, is required by social partnership in Germany, but not in the UK. LSCs do contain trade union and educational representatives, but employers dominate numerically. NTOs remain the exclusive preserve of employers. Concerning powers, German employers are required to join and finance intermediate bodies, which examine and certify apprentices and validate employer eligibility to train. Their British counterparts possess none of these powers. NTOs, in particular, are small and poorly funded, lacking systematic financial support from employers who use the skills that they seek to promote (NTONC, 2001).

Finally, in Britain, specialist training organisations are not only permitted to take on the sponsorship of apprenticeship, i.e., responsibility for a training programme as a whole, but have come in practice to dominate it. Specialist trainers, primarily private companies and public colleges, constitute 59 per cent of providers, sponsoring 67 per cent of trainees (on all programmes; table 5). Single employers sponsor only 5 per cent of trainees. In Germany, the employer must take responsibility for the programme as a whole. The implications of the difference are far-reaching (next section, below).

MA shows greater organisational similarity to its predecessor. YT was administered along similar lines by the antecedents of the NTOs and the LSCs. The NTOs' role in setting training frameworks under MA is however new, and LSCs contain more employee and educator representatives than did their predecessors, the TECs (Wood, 1999).

Finance

In German apprenticeship, the public purse bears the entire institutional costs of part-time education at technical colleges. The costs of workplace training are privately born, by the employer and the apprentice. The principle is simple: the taxpayer pays for the college-based, educational part; the employer and the trainee, for the workplace-based part.

The principle has an economic rationale. Education comprises general skills, of potential benefit both to the trainee and to society as a whole (Becker, 1964). It should therefore be financed mostly from public funds, at least through upper secondary level, with students possibly bearing opportunity costs (foregone earnings). Work-based training is expected to contain a greater component of employer-specific skills, of benefit primarily to the relevant employer and apprentice -- who should therefore finance it themselves, on both equity and efficiency grounds. The occupation-specific, transferable component of training (Stevens, 1994) is not however financed according to economic logic: only in construction do all employers in the sector collectively fund the training from which they collectively benefit.

MA also sees substantial subsidies to training: grants now average nearly [pound]7,000 for mainstream 16-18 year old MA completers, ranging from [pound]4,000 in retailing to [pound]12,000 in engineering (table 3, above; LSC, 2001). The MA grant is influenced by two factors: the perceived overall cost of training in an occupation (high in engineering, low in retailing) and the perceived importance of skills supplies in the occupation (ibid.). The degree to which the grant actually covers total training costs remains unclear. It appears to have been set at occupational level as a rough average of the various grants made previously by TECs, rather than the result of research into actual training costs and desirable funding principles. It may well cover the entire direct costs of training (on and off the job) in many service occupations, including retailing and personal services, where an NVQ can be achieved largely or wholly on the job, but only a minority share in industrial occupations where technical instruct ion is required (Hogarth et al., 1996). If so, MA resembles German apprenticeship, which shows low-to-zero net training costs for small firms and artisanal training, but high net costs for larger firms and industrial-commercial training (Bardeleben et al., 1994).

No German-type linkage has however been attempted under MA between training components and public funding. The grant goes towards the entire training programme, not just vocational education or off-the-job training. The principle is pragmatic: that the rate of subsidy be sufficient to generate sufficient apprenticeship places to satisfy a government keen to meet its skills targets. 'It is unrealistic to ask employers to bear all of [the] costs of apprenticeship training other than the fee costs of the off-the-job element' (DfEE, 1999a, p. 55). The principle has not however been followed consistently. The introduction of the Key Skills requirement, with its explicitly educational focus, was not accompanied by any general increase in grants, to the detriment of employer involvement.

A practice closer to the German funding principle has actually emerged by default in parts of MA. Although specialised training providers sometimes pay sub-contracting employers for providing on-the-job training and work experience, that practice, widespread under YT, has become rare under MA (ALP, 2000), leaving the employer to provide on-the-job training without external subsidy. The cost of work-based training then falls wholly to the private parties and that of off-the-job training to government. The resemblance to Germany remains limited even then, as the educational component of training, and therewith the principal justification for public subsidies, remains weak in Britain.

YT also provided grants for work-based training, focused on 16-18 year olds. Real spending per participant in 1992-3 amounted to [pound]3,070 (at March 2001 prices). The rate of subsidy may have been increased by MA, in association with the reorientation of training from Level 2 to Level 3 qualifications. The YT figure is indeed less than half the prospective MA grant for 16-18 year olds of [pound]6,800 (table 3 above), but widespread and early non-completion make the latter an invalid guide to actual expenditure on MA -- concerning which data are unfortunately not available. The introduction of MA has not been accompanied by any increase in the rate of funding for work-based training as a whole: real per capita spending on all programmes actually fell, by 9.3 per cent, between 1992 and 1998 (table 6). Any compositional rise associated with the arrival of MA appears to have been more than offset by the standing requirement for 'efficiency gains' in all programmes (see next section). The ex post rate of subsid y under MA may well not exceed that under YT.

Two other financial attributes are the division of private costs between the employer (or provider) and the apprentice, and the further division of the employer share between those who do and do not sponsor apprentices. Concerning the letter, cost redistribution among employers who do and do not provide apprenticeship, by way of levies and grants, is absent outside construction in both countries, notwithstanding the potential resource misallocation associated with movement of skilled workers between employers.

Concerning the former, the division of costs between the employer and the apprentice in Germany reflects the special status of the apprentice, as neither employee nor student. The corollary -- low apprentice allowances -- sees both parties share training costs, in larger firms at least (Beicht and Holzschuh, 1990). MA appears similar, in that the apprentice and the employer must sign training 'pledge', and the provider formulate a training plan. The pledge is not however a legally enforceable contract. Moreover, 97 percent of participants hold an employment contract, earning a wage that invariably exceeds the fall-back weekly training allowance of [pound]40 (TSC, 2000b).

In YT, only 30 per cent of participants had held employee status in 1992 (ED, 1994, Table 7). The government encouraged employee status on MA in order to strengthen the link between the employer and the apprentice, and to attract young people to a programme with uncertain benefits. Employee status has certainly encouraged employment continuity after apprenticeship: many non-completers remain with the same employer after leaving (Winterbotham et al., 2000). It has also increased training costs, particularly in industrial occupations, with their longer programmes and more expensive facilities (table 3, above) - to the potential detriment of the supply of training places (Marsden and Ryan, 1991).

MA's content, organisation and funding therefore resemble both German apprenticeship and YT in particular respects. The programme is closer to German apprenticeship in terms of its formal skills focus, the formal role of intermediate bodies, and substantial public funding. It resembles YT in its weak educational content, avoidance of process regulation, reliance on specialised providers, pragmatic approach to funding, and lack of a statutory framework. The actual skills content of MA, and the composition and powers of intermediate committees, both resemble YT more than German apprenticeship. The balance as a whole is therefore tilted markedly towards YT.

Britain's 'training market'

We now discuss the factors that hold British apprenticeship back. The problem centres on the market-oriented methods chosen for public support for apprenticeship.

The British 'training market' is actually a quasi-market: i.e., government sets demand by funding the service, but does not supply the service itself, relying instead on independent contractors, private or public (LeGrand and Bartlett, 1993). It has three components: public targets and funding, 'employer leadership' and competitive contracting.

The government determines national skill needs and the expenditure to be devoted to their attainment, as represented by the National Learning Targets and the DfES budget, respectively. The relevant target is 60 per cent of English 21 year olds holding a Level 3 qualification (i.e., two A-Levels or the vocational equivalent) by the year 2002. The actual rate rose quickly in the 1990s, but as it had reached only 54 per cent in the year 2000, the targets have meant continuing pressure to raise youth qualification rates (DfEE, 2000a).

Secondly, employer representatives on NTOs set the skill standards (NVQ level and content) in terms of which the national targets are defined (subject to the loose requirement that attainment levels be similar across the part-time and full-time routes -- e.g., that the content of an NVQ3 be 'equivalent' to that of two A-Levels).

The third component is competition for public business. A range of providers -- private as well as public, for profit as well as non-profit -- compete for contracts to provide training. Thus further education colleges compete with training companies to win off-the-job training contracts. Contractual incentives are embodied in output-related funding: payments to providers depend on trainee outcomes -- qualifications gained and jobs found -- not just time spent in training. Under YT, the output -related share of funding, dependent on local TEC decisions, rose to more than one half in some localities. Under MA, it has been reduced and standardised at one-fifth, split equally between the achievement of the NVQ3 and other requirements (Felstead and Unwin, 2001).

Employer leadership and competitive contracting contrast with social partnership and administrative allocations in German apprenticeship. In Germany, skill standards are determined by sectoral committees on which employees and educators are represented along-side employers. The separation between college-based and workplace-based components means hierarchical funding and control through the public education system for the former, and autonomous self-funding by employers and apprentices for the latter (see previous section). Public vocational colleges monopolise part-time education for apprentices.

Britain's use of quasi-markets rather than hierarchies is intended to capture particular benefits. Economic incentives strengthen, particularly when for-profit providers compete. Responsiveness should rise, with the purchaser getting more of what it demands, not simply what the producer wishes to offer. The cost of given services should fall as inefficient providers lose out. Subcontracting should capture economies of specialisation and scale, as the provider contracts out services that other bodies can provide on better terms, such as trainee assessment. Such developments should all increase economic efficiency.

The competitive quasi-market has indeed supported a large volume of training, pushed costs down, and increased variety and flexibility of provision. FE colleges, for example, nowadays offer off-the-job training courses at the workplaces and also time them to suit the employer, departing from the traditional classroom-based, academic year format.

At the same time, market forces do not work well in all contexts. Education and training are leading candidates for the 'market failure' category (Stiglitz, 1994). Two features of Britain's training quasi-market point to underlying inefficiencies: low quality and weak commitment (Unwin, 1997).

Quality

The quality problem is intrinsic when the content of contracts cannot be specified with anything approaching completeness - because of four factors: output is multi-dimensional; its quality dimensions matter; information about quality is asymmetric (i.e., the supplier knows more about it than does the buyer or external agents); and suppliers can substitute quantity for quality (i.e., increase the quantity of output from given resources by reducing its quality).

Such is the case for training, as for education and health services. Training quality denotes here either the amount of training received or the pedagogical merits of a given programme: both influence the amount of skill acquired. Both are difficult for the buyer to observe - particularly in work-based training, where learning and producing are intertwined (Katz and Ziderman, 1990; Ryan, 1994; Acemoglu and Pischke, 1998). Given the informational asymmetry between seller and buyer, contracts are by default written in terms of quantities. The commercial supplier then reduces quality in order to increase profit, even when the buyer and the economy would gain from a higher ratio of quality to quantity.

Low quality has been endemic in public training programmes. In MA, the problem has been observed in three areas. Firstly, there is the bias towards early departure and premature completion by apprentices as a result of the narrowness of NVQ qualifications, the laxity of much workplace assessment, and the incentive to wave apprentices through quickly under output-related funding (Stanton, 1996). The LSC's decision to reduce the output-related share of funding to 20 per cent reflects recognition of the damage already done to training quality (DfEE, 1999b).

Secondly, low quality has been encouraged by a pernicious combination of fiscal and technological factors. Public spending on youth programmes has soared since the 1970s, but so too has the clientele: government programmes, like YT before them, currently guarantee a place to all young people who seek one. In an era of fiscal restriction, government has used competition between providers to drive down unit costs, and thereby cushion the effects of high training volumes on public expenditure. Moreover, training programmes have long been funded, under the misleading rubric of 'efficiency gains', on the presumption that productivity - e.g. the number of qualifications produced from given inputs - can be expected to grow at 2 per cent per annum in both commercial and public provision. From a technological standpoint, however, it is unrealistic to expect that in so labour-intensive a sector, in which personal interaction between trainer and trainee remains important for learning, productivity growth can approximate its long-term rate in the economy as a whole, however competitive the quasi-market (Baumol, 1993; Ryan, 1992). The expectation becomes totally unrealistic when capital spending is cut back as well, as has occurred in further education. When fiscal restriction, stagnant technology, buoyant activity and asymmetric information converge, it is not surprising that quality suffers.

Thirdly, MA activity has been biased towards short, low cost programmes and skills not in short supply, particularly in the service sector, away from more expensive and scarce industrial skills. Pursuit of national skills targets has aggravated the problem. The targets have been broken down by area and assigned to TECs (now LSCs) to achieve. In MA's early years, a TEC priority was to meet the local target. Low cost, Level 2 training programmes previously funded by YT were extensively 'rebadged' under MA, with little or no change of content (DfEE, 1999b). Providers benefited from the higher rate of subsidy; TECs expected to benefit from low cost progress towards their targets (Felstead and Unwin, 2001). The LSC's injection of sector-specific weights (typically 50 per cent greater in industry than in services) into the basic MA grant seeks to counter the bias by offering yet higher grants for apprenticeships in industry (LSC, 2001).

Evidence of low quality abounds in the set of stripped down providers that MA inherited from YT. Most staff are not adequately trained and qualified -- occupationally, pedagogically or both. Indeed, many providers train their own staff only when contractually obliged to do so, e.g., by MA's health and safety requirements (DfEE, 2000a, 2000b; TSC, 2000a). Similar problems are visible among workplace training staff (ERS, 1998; TSC, 2000a). Proposals to require appropriate qualifications for trainers along German lines have been rejected by employer representatives, who cite the associated threat to the supply of places (NTONC, 2000).

The government's response has been a campaign to raise standards, centering on the external inspection of providers by the Adult Learning Inspectorate (formerly the Training Standards Council). The resources devoted to inspection have increased considerably, leading to public recognition of the scale of the quality problem (TSC, 2000a).

An expansion of external inspection is welcome, but its cost-effectiveness remains dubious. Asymmetric information means scope for 'pulling the wool' over inspectors' eyes, particularly in the less accessible, work-based component of training. Some private providers have indeed had their eligibility to contract removed after receiving low-quality grades and failing to improve, but the government's desire to maintain activity levels discourages any general imposition of sanctions on low-quality providers. The competence of some inspectors themselves has been questioned (DfEE, 2000e). TSC's (200b) review of Modern Apprenticeship induced little confidence in the inspectorate's own quality assurance, given its neglect of the evidence of low quality provision presented in other official documents (DfEE, 2000b; TSC, 2000a).

Moreover, YT provided the discouraging precedent of an initial emphasis on quantity, leading to evidence of low quality, and followed by recourse to inspection -- albeit in a less independent and a lower budget format than at present. The results of that effort were derisory (TSAS, 1989; Marsden and Ryan, 1991).

Even when quality proves low, the appropriate market-oriented response is not clear, particularly for public providers. Thus some under-performing further education colleges nowadays face simultaneously a cut and an increase in resources, under output-related funding and the Education Standards Fund, respectively (DfEE, 2001a).

Finally, asymmetric information permits sharp practice among competitors for the business of 16-19 year old learners. Some schools and colleges impede the flow of information to their students about the part-time, work-based route (Coleman and Williams, 1998). They do so partly out of concern for youth welfare, given the quality deficiencies of youth programmes, and partly in response to public pressures to increase full-time participation rates. Institutional self-interest contributes too: the institution's funding benefits when students enrol full-time, even though apprenticeship might suit the young person better.

Commitment

The second deficiency of Britain's training quasi-market is low commitment to the achievement of public education and training goals among participants, both employers and apprentices.

MA sought to make the employer central to apprenticeship. The apprenticeship 'pledge' was to be signed by the young person, the employer and the TEC (ED, 1994). Employee status was to strengthen further the bond between employer and trainee. Employer sponsorship of MA has however proved rare, particularly among large national companies. Most employers who are involved function at the end of the sub-contracting chain, as suppliers of work-based training only. They also tend to be small -- nearly three quarters of workplaces with MA places have fewer than 50 employees -- and to train only one or two apprentices at a time.

The low profile of the employer is complemented by the primacy of the specialised provider (table 5, above). Two decades of youth programmes have seen the consolidation of a sector with expertise in organising activity within the quasi-market's rules and funding (Unwin, 1999).

The employer-provider relationship is often problematic, with responsibility for apprenticeship prone to falling between the two stools. Specialised providers report employers as often showing 'relatively little commitment to MAs', as 'playing a passive role' and as using providers as 'simply a recruitment agency for 16-18 year olds'. Non-completing apprentices often report having encountered little effort by the employer to explain their programme (Winterbotham et al., 2000). Conversely, many employers criticise the training offered by external providers (Kodz et al., 2000).

Many apprentices also show low commitment, abandoning training for regular employment, typically for a pay increase, offered sometimes by the employer providing their placement. Some apprentices report having drifted into MA with inadequate information or without much thought, or simply as a stop-gap activity, and limited concern on that score on the part of the provider (Coleman and Williams, 1998; Winterbotham et al., 2000).

Low employer commitment contributes to MA'S quality problem. Inspection ratings of training quality tend to be higher when the sponsor is an employer and when a further education college provides the off-the-job component - i.e., training arrangements resemble those in German apprenticeship (TSC, 2000a). Employer commitment proves important also to the success of the Key Skills requirement, but is absent more often than not: 'in the minority of situations where employers were seen to be committed, training providers and trainees tended to take [it] seriously' (DfEE, 2000e, p.10).

Weak employer involvement may be understood partly in terms of aversion to the bureaucratic requirements of youth programmes (Wood, 1999; CBI, 2000; Sims et al., 2000). Such complaints appear more frequent in MA than in German apprenticeship. The resulting paradox -- that the burden of regulation should strike employers as greater in the more market-oriented system, with the apparently larger public subsidies -- may have a simple explanation. Allocations made through quasi-markets not only rely on self-interest but also intensify it and narrow it. Employers and apprentices often use MA opportunistically, as a vehicle for their immediate economic interests, rather than as a programme for education and skills. Many employers adopt a cafeteria approach, selecting the components of a training framework that suit their own limited requirements, claiming the corresponding funding, discarding the remainder -- and leaving the government to re-jig the incentives and rewrite the regulations in vain attempts to discoura ge their behaviour (Winterbotham et al., 2000; DfEE, 2000e).

The decision to inject the widely resented Key Skills requirement into MA after the programme had already been running for three years also encouraged opportunistic responses. An issue of principle was also present: having been told that MA was intended to cater to their immediate demands for occupational skills, many employers objected to its subsequent extension to perform what they saw as the task of compulsory schooling (CBI, 2000; DfEE, 2000e).

More generally, the commercialised British approach promotes low trust between government, as purchaser of training, and providers, as sellers. In order to ensure that public funds are not wasted or purloined, auditing procedures are adopted, the multiplicity, obscurity and intrusiveness of which deter employer engagement with MA.

We do not suggest that in Germany employers and young people opt for apprenticeship out of altruism rather than self-interest, nor that vigilance over the use of public funds is absent. On the former issue, the avoidance of a commercialised training market, together with the integration of apprenticeship with upper secondary education, promotes in Germany a more long-sighted and broader concept of self-interest. On the latter one, under 'associational self-government', statutorily defined administrative powers over apprenticeship are devolved to intermediate committees, comprising educators and trade unionists as well as employers, and faced by limited reporting requirements. These arrangements both rely upon and foster commitment -- including that of employers to apprenticeship (Streeck, 1992; Crouch et al., 1999).

An irony here is that the British training market, which gives employers 'unprecedented influence' over public policy, elicits so little commitment from them (DfEE, 1999a, p.10; Keep, 1999). Another is that statutory regulation should be associated in German apprenticeship with a lighter regulatory burden on employers than is Britain's market-oriented system.

Britain's commercial methods also damage efficiency by undermining the potential contribution of professionalism. Professional motivation is often viewed in official circles as a cover for producer control and slack working. Its orientation to interests of the client and to service quality can however reduce goal divergence between providers, apprentices and government, relative to that in the quasi-market, and make it easier and cheaper to attain desired quality attributes (Matthews, 1991). The quasi-market not only squanders professional motivation, it also undermines it by putting profit first.

The problem is acute in further education. Set into competition with commercial providers, it has been pulled down to the level of funding and quality at which they have to operate. Declining relative pay, the casualisation and growth of part-time employment, intensified external demands and monitoring have created serious recruitment and motivation problems in a service central to the educational potential of apprenticeship. Many instructors retain a professional orientation, but find they have to manipulate the rules of the quasi-market in order to help trainees (Felstead and Unwin, 2001). These problems are largely side-stepped in policy documents, and even in the reports of the inspectorate (FEFC, 2000). A need to 'rebuild the technical instruction that once symbolised the very best of our industrial training system' has been explicitly conceded (DfEE, 2001b, p.1). Proposals for colleges to compete for the honour of 'Pathfinder' status miss the point, however (ibid.).

Marketing

The commercialisation of training policy also sees government imitation of the sales promotion techniques of the private sector. Public programmes have been the object of repeated reinvention and extensive marketing. Such 'rebranding' comprises however changes of name and presentation more than of content and method. The leading example is the six re-launches that converted the Youth Training Scheme of 1982 into Foundation Modern Apprenticeship in 1999. YT constituted only one stop en route, but its institutional attributes have been broadly present throughout.

In the private sector, rebranding can work when it is associated with the maintenance or creation of a high quality product, but otherwise it tends to backfire (Aaker, 1996). Widespread recognition of low quality in youth programmes may therefore explain general indifference to their various re-launches. The creation of Modern Apprenticeship in 1994 was a more promising case, attempting as it did to link the upgrading of a youth programme to apprenticeship's reputation for intermediate skills training. The potential has however been undermined by two factors: evidence of low standards, and the rebranding of all Level 1-2 youth training by 2002, as Foundation MA.

Clumsy imitation of the methods of the private sector damages the product. Many employers report ignorance of, or bafflement over, MA's content, particularly with the advent of the Advanced/Foundation dichotomy. Others object to an excessive 'sales pitch' in the promotion of MA. The battered reputation of MA and NVQs suffers further (CBI, 2000; Sims et al., 2000).

The British and German approaches to universalising youth apprenticeship differ radically. Germany has maintained craft requirements and used pre-apprenticeship programmes to raise steadily the share of young people that can prospectively attain them (Wagner, 1999; Franz et al., 2000). Britain has opted instead to abandon craft requirements, and with a stroke of the pen to include in 'apprenticeship' the many young people who remain incapable of meeting them. 'Apprenticeship' now promises to become simply a synonym for publicly funded work-based learning, to the detriment of its historically hard-won reputation for skill and quality.

Against the flexibility-related advantages of Britain's training quasi-market must therefore be set its deficiencies of quality and commitment, as well as the low cost-effectiveness of attempts to counter those defects.

Conclusions

The weaknesses of contemporary British apprenticeship, in comparison to its German counterpart, are easier to understand if its principal ingredient, Modern Apprenticeship, is viewed as a labour market programme rather than an institution. MA evolved out of the youth programmes typified by Youth Training, to which we have also compared it. MA has aimed higher than YT, but the organisation and operation of the two programmes have much in common.

MA has increased the contribution of youth programmes to national skill supplies. Like YT before it, the programme contains some excellent training. Overall, however its contribution has been limited, leaving MA well short of the mark set by German apprenticeship. Rates of qualification and completion remain low, as does employer involvement. Apprenticeship activity appears not to have increased, despite an unprecedented rate of subsidy. Opportunities to secure high quality vocational preparation remain hard for young people to find. We link these failings to the commercialised training quasi-market within which MA operates, as did YT before it.

Two further questions may be raised. Firstly, are the public funds devoted to Modern Apprenticeship worth it? MA can indeed claim to have extended systematic youth training into sectors and occupations from which it had been largely absent, and to have reduced gender and age imbalances in access to apprenticeship. At the same time, much of the training that MA has developed is marred by being skilled only in name, and by low rates of qualification and completion. Moreover, in industrial occupations, with their greater training costs and skill shortages, activity is stuck near the low levels of the mid-1990s. Deadweight appears extensive.

Secondly, do current reforms (Blunkett, 2001) promise marked improvement? On paper, they are highly attractive, proposing to build apprenticeship into a ladder of ascending work-based qualifications, linked to the wider ladder of vocational qualifications. Climbers on the vocational ladder are to be encouraged to step to and from the general ladder. The ambition aligns with recent developments in continental apprenticeship and the rising supply of more qualified school-leavers at home. Institutional support for so radical a departure remains inadequate, however. NVQs are educationally impoverished, and requirements for Key Skills and Technical Certificates are easily side-stepped by employers and apprentices. Many training providers, including further education colleges, lack qualified staff. The educational content of Foundation MA remains weak.

At the root, the training quasi-market constitutes an obstacle rather than an asset. Apprenticeship's contribution to intermediate skills will continue to falter until better ways of ensuring quality and eliciting employer commitment are developed. Continental apprenticeship suggests promising alternatives.

(*.)University of Cambridge [King's College, Cambridge CB2 IST, e-mail: pr23@econ.cam.ac.uk].

(**.)University of Leicester [Centre for Labour Market Studies. 7 Salisbury Rd, Leicester LEI 7QR, e-mail: l.unwin@le.ac.uk]. We thank officials of the Department for Education and Skills, LSCs and further education colleges for helpful discussions, the DfES and ONS for access to unpublished data, and S.J.Prais for suggestions and encouragement. The views expressed in this paper may well diverge from those of the DfES.

REFERENCES

Aaker, D.A. (1996), Building Strong Brands, New York, Free Press.

Acemoglu, D. and Pischke, J.-S. (1998), "Why do firms train? Theory and evidence', Quarterly Journal of Economics, 63(l), pp. 79-119.

ALP (2000), 'Response to "Vocational Education and Training: a Framework for the Future", London, Association of Learning Providers' (www.uuy.org.uk/projects/ten/newslett/2000-may).

Bardeleben, R. von, Beicht, U., and Feher K. (1994), 'Kosten und Nutzen der betrieblichen Berufsausbildung' Wissenschaft und Praxis, 23, 3, pp. 3-17.

Baumol, W.J. (1993), 'Health care, education and the cost disease', Public Choice, 77, pp. 17-28.

Becker, G. (1964), Human Capital, New York, Columbia University Press.

Beicht, U. and Holzschuh, J. (1990), Die Ausbildungsvergutungen in der BRD, Bonn (BIBB).

Blunkett, D. (2001), Education into Employability: the Role of the DfEE in the Economy, London, DfEE.

BMBF (2000a), Berufsbildungsbericht 2000, Bundesministerium fur Bildung und Forschung, Bonn.

--(2000b), Grund- und Struktrudaten 1999/2000, Bundesministerium fur Bildung und Forschung, Bonn.

Canning, R. (2001), 'Vocational education and training in Scotland: emerging models of apprenticeship', in DeiBinger T. (ed.), Berufliche Bildung als Orientierungs- und Anpassungsproblem. Analysen zur Vorbildfunktion von Berufsbuildungssystemen und berufsbildungspolitischen Entwicklungsperspektiven. Baden Baden, Nomos.

CBI (2000), Results of CBI Survey of Members on Modern Apprenticeships 2000, London, Confederation of British Industry, Human Resources Directorate.

Coleman, N. and Williams, J. (1998), Evaluation of Modern Apprenticeships: 1998 Survey of Young People, Research Report 93, London, DfEE.

Crouch, C., Finegold, D. and Sako, M. (1999), Are Skills the Answer? Oxford, Oxford University Press.

Deakin, S. and Morris, G.S. (1998), Labour Law, second edition, London, Butterworths.

DeiBinger T. (1996), 'Germany's Vocational Training Act; its function as an instrument of quality control within a tradition-based vocational training system', Oxford Review of Education, 22(3), pp. 317-36.

DfEE (1999a), Delivering Skills for All. Second Report of National Skills Task Force, London, Department for Education and Employment.

--(1999b), Learning to Succeed: a New Framework for Post-16 Learning, London, Department for Education and Employment.

--(2000a), National Learning Targets for England for 2002, London, Department for Education and Employment.

--(2000b), Learning to Succeed: Raising Standards in Post-16 Learning. Building Practitioner Skills and Qualifications in Work-Based Learning Funded by the LSC and the Employment Service. London, Department for Education and Employment.

--(2000c). Modern Apprenticeships: Consultation Document, London, Department for Education and Employment.

--(2000d). 'Advanced Modern Apprenticeships: comparisons of pay during and after training', unpublished paper, London, Department for Education and Employment.

--(2000e), Delivery of Key Skills in Modern Apprenticeships, Quality and Performance Improvement Dissemination Report 89. Sheffield, Department for Education and Employment.

--(2000f), Department Report. The Government's Expenditure Plans 2000-01 to 2002-03, London, Department for Education and Employment, HMSO.

--(2001a), Raising Standards in Further Education. London, Department for Education and Employment.

--(2001b), Centres of Vocational Excellence: Heralding a New Era for Further Education. Consultation Paper, London, Department for Education and Employment.

DfES (2001), 'Work-based training for young people and work-based learning for adults: volumes and outcomes', First Release, 22 June, SFR 28/2001, London, Department for Education and Skills.

ED (1993), 'YT Outcomes', Labour Market Quarterly Report, November, London, Employment Department.

--(1994), Department Report. The Government's Expenditure Plans 1994-95 to 1996-97, Employment Department, Cm 2505, London, HMSO.

ERS (1998), Evaluation of Modern Apprenticeships: 1998 Survey of Employers, Research Report RR94, London, Department for Education and Employment.

Evans, K., Hodkinson, P., Keep, E., Maguire, M., Raffe, D., Rainbird, H., Senker, P. and Unwin, L. (1997), Working to Learn, London, IPD.

FEFC (2000), Quality and Standards in Further Education in England, 1999-2000, London, Further Education Funding Council.

Felstead, A. and Unwin L. (2001), 'Funding post-compulsory education and training: a retrospective analysis of the TEC and FEFC systems and their impact on skills', Journal of Education and Work, 14(1), pp. 91-111.

Franz, W., Inkemann, J., Pohlmeier, W. and Zimmerman, V. (2000), 'Young and out in Germany: on youths' chances of labour market entrance in Germany', in Blanchflower, D.G., and Freeman, R.B, (eds), Youth Employment and Joblessness in Advanced Countries, Chicago, University of Chicago.

Fuller, A. and Unwin, L, (2001), From Cordwainers to Customer Service the Changing Relationship between Apprentices, Employers and Communities in England, Research paper, SKOPE, University of Warwick

Gospel, H. (1995), 'The decline of apprenticeship training in Britain', Industrial Relations Journal, 26, 1, pp. 32-44.

Green, A. (1997), 'Core skills, general education and unification in post-16 education', Working Paper, Post-16 Education Centre, London, Institute of Education.

Green, A. and Steedman, H. (1997), Into the Twenty First Century: an Assessment of British Skill Profiles and Prospects, Centre for Economic Performance, London School of Economics.

Grubb, W.N. and Ryan, P. (1999), The Roles of Evaluation for Vocational Education and Training, ILO/Kogan Page.

Hogarth, T., Siora, G., Briscoe, G. and Hasluck, C. (1996), The Net Costs of Training to Employers, Research Study RS3, London, Department for Education and Employment.

Katz E. and Ziderman A. (1990), 'Investment in general training: the role of information and labour mobility', Economic Journal, 100, pp. 1147-58.

Keep, E. (1999), Employers' Attitutudes towards Training, Skills Task Force Research Paper IS, London, Department for Education and Employment.

Kodz, J., Tackey, N.D., Pollard, E., Dench, S., Tyers, C. and Dewson, S. (2000), Modem Apprenticeships and National Traineeships: Skills Utilisation and Progression, Research Report RR204, London, Department for Education and Employment.

LeGrand, J. and Bartlett, W. (eds) (1993), Quasi-Markets and Social Policy, London, Macmillan.

Leman, S. and Williams, P. (1995), 'Apprentices and other long-term trainees -- data from the LFS and other surveys', Employment Gazette, 103, 2, pp. 67-74.

LSC (2001), Learning and Skills Council Operations Guide, Version 3, 11 April, Learning and Skills Council

MAAC (2001), 'Provider models', unpublished paper, Modern Apprenticeship Advisory Committee, Sheffield, Department for Education and Employment.

Marsden, D.W. and Ryan, P. (1991), 'Initial training, labour market structure and public policy: intermediate skills in British and German industry', in Ryan, P. (ed.), International Comparisons of Vocational Education and Training far Intermediate Skills, Lewes, Falmer Press, pp. 251-85.

Mason, G. (2000), 'Production supervisors in Britain, Germany and the US: back from the dead again?', Work, Employment and Society, 14, 4, pp. 625-45.

Matthews, R.C.O. (1991), 'The economics of professional ethics: should the professions be more like business?', Economic Journal, 101, pp. 737-50.

NTONC (2000), "'Raising standards in post-16 learning": NTO National Council response', London, NTO National Council, 27 October.

--(2001), 'Towards a stronger sector skills framework: response by the NTO National Council Board to the UK government's consultation, "Building a Stronger Network"', London, NTO National Council.

Oulton, N. (1996), 'Workforce skills and export competitiveness', pp. 201-30 of Booth, A. and Snower, D. (eds), Acquiring Skills: Market Failures, Their Symptoms and Policy Responses, Cambridge, Cambridge University Press.

Payne, J. (2001), Work-Based Training for Young People: Data from England and Wales Youth Cohort Study, Research Report RR276, London, Department for Education and Skills.

Payne, J., Riley, R. and Coleman, N. (2001), Feasibility Study for the Long Term Evaluation of Modern Apprenticeships, RR290, London, Department for Education and Skills.

Prais, S.J. (1995), Productivity, Education and Training, Cambridge, Cambridge University Press.

Ryan, P. (1992), 'Unbalanced growth and fiscal restriction: public spending on higher education in advanced economies since 1970', Structural Change and Economic Dynamics, December, 3, 2, pp. 261-88.

--(1994), 'Training quality and trainee exploitation', in Layard, R., Mayhew, K. and Owen, G. (eds), Britain's Training Deficit, Aldershot, Avebury, pp. 92-124.

--(2000), 'The institutional requirements of apprenticeship: evidence from smaller EU countries', International Journal of Training and Development, 4, 1, pp. 42-65.

--(2001a), 'The school-to-work transition: a cross-national perspective', Journal of Economic Literature, March, 39(1), pp. 34- 92.

--(2001b), 'Apprenticeship in Britain: tradition and innovation', in DeiBinger, T. (ed.), Berufliche Bildung ols Orientierungs- und Anpassungsproblem. Analysen zur Vorbildfunktion van Berufsbuildungssystemen und berufsbildungspolitischen Entwicklungsperspektven. Baden Baden, Nomos.

Sims, D., Golden, S., Blenkinsop, S. and Lewis, G. (2000), Barriers to Take-up of Modern Apprenticeships and National Traineeships by SMEs and Specific Sectors. Research Report RR 205, Sheffield, Department for Education and Employment.

Stanton, G. (1996), Output-related Funding and the Quality of Education and Training, Institute of Education, London University.

Steedman, H. (1998), 'A decade of skill formation in Britain and Germany', Journal of Education and Work, 11 (1), pp. 77-94.

Steedman, H., Green, A., Bertrand, O., Richter, A., Rubin, M. and Weber, K. (1997), Assessment, Qualifications and Standards: the UK compared to France, Germany, Singapore and the US, Centre for Economic Performance, London School of Economics.

Steedman, H., Gospel, H. and Ryan, P. (1998), Apprenticeship: a Strategy for Growth, London, Centre for Economic Performance, London School of Economics.

Stevens, M. (1994), 'A theoretical model of on-the-job training with imperfect competition', Oxford Economic Papers, October, 46 (4), pp. 537-62.

Stiglitz, J. (1994), Whither Socialism?, Cambridge MA, MIT Press.

Streeck, W. (1992), Social Institutions and Economic Performance, London, Sage.

TSAS (1989), Review of Activity, Training Standards Advisory Service, Sheffield, Training Agency.

TSC (2000a), Reaching New Standards: Annual Report of Chief Inspector, London, Training Standards Council.

--(2000b), Modern Apprenticeships, London, Training Standards Council.

Unwin, L (1997), 'Reforming the work-based route: problems and potential for change', in Hodgson, A. and Spours, K. (eds), Dearing and Beyond: 14-19 Qualifications, Frameworks and Systems, Kogan Page.

Unwin, L. (1999), Jungle Trekking: Vocational Courses and Qualifications for Young People. Research Paper 5, Skills Task Force, London, Department for Education and Employment.

Wagner, K. (1999), 'The German apprenticeship system under strain', in Culpepper, P.D, and Finegold, D. (eds), The German Skills Machine: Sustaining Comparative Advantage in a Global Economy, New York, Berghahn.

Winterbotham, M., Adams, L and Lorentzen-White, D. (2000), Modern Apprenticeships: Exploring the Reasons for Non-completion in Five Sectors. RR217, London, Department for Education and Employment.

Wolf, A. (1995), Competence-based Assessment, Buckingham, Open University Press.

Wood, S. (1999), 'Building a governance structure for training? Employers, government and the TEC experiment in Britain', pp. 363-402 of Culpepper, P.D, and Finegold, D. (eds), The German Skills Machine: Sustaining Comparative Advantage in a Global Economy, New York, Berghahn.

[Graph omitted] Table 1. Number of apprentices in Britain and Germany, 1998 UK Publicly funded apprentices (AMA) England & Wales 1. Flows [*] apprenticeship entrants 87.4 ('000) youth population 643.8 cohort share (%) 13.6 2. Stocks apprentice numbers ('000) 118.8 employment 23,967 employment share (%) 0.50 Germany Self-classed Registered apprentices apprentices (LFS) Great Britain Bundesrepublik 1. Flows [*] n.a. 587.5 .. 934.0 .. 62.9 2. Stocks 176.9 1658 26,274 35,860 0.67 4.42 Sources. BMBF (2000a.), U1, S2; (2000b),pp.359 German version), 184 (English version); Dfes (2001),Tables 1, 2; ONS (2001) Table 5.3; ONS, unpublished tabulations from the UK Labour Force Survey; Labour Market Trends, July 2000, Table A.4 and previous equivalents. Notes: n.a.: not available (*)includes re-entrants (8% of AMA total in 2000-01). Inflow and population data for Germany are for 1997. Youth population is number of 17/18 yearolds (Germany) and one-fifth the number of 14-19 year olds (UK). LFS apprentices are individuals who report themselves as such (16-24 year olds only.) 'Employment' includes the self-employed, the armed forces and unpaid intra-family workers. 'Employment share' includes apprentices in the denominator, which requires them to be added in for Germany but not for the UK, where the great majority of MA participants have employee status. Table 2. Length of stay and rates of completion and qualification in MA and YT in Britain and in apprentceship in Germany (1) Apprenticeship Germany 1997 Average lengh of stay (months) Stipulated 36.5 Completers 35.2 All leavers 32.3 Completion (% leavers) 78 Highest VQ gained (% leavers) Full Levels 3-4 76 1-2 0 Partial only 0 Any 76 (2) (3) (4) MA YT England Great Britain 1998-9 1999-2000 1992-3 Average lengh of stay (months) none none none n.a. 20.5 n.a. 11.5 [*] 17.4 11.8 Completion (% leavers) 38 n.a. 43 Highest VQ gained (% leavers) 38 50 16 23 21 42 7 7 13 68 78 71 (5) All programmes England 1999-2000 Average lengh of stay (months) n.a. n.a. n.a. Completion (% leavers) n.a. Highest VQ gained (% leavers) 23 29 7 59 Sources: (1), (3) DfES (2001), Tables 3,5 LSC (2001), ED (1993), p.9, and Unpulished data, DfES Trainee Database; (2) BMBF (2000a, previous year), Tables 33, 34, 43. Notes:(*)average for largest ten frameworks. MA length of stay and completion data are for England and Wales. The MA/YT comparison is slightly biased in favour of MA by intervening changes in the leaver status of participants who change training proiders in the statistics. Length of stay. German average is for entrants, assuming that leaving rates by year of training programme are constant within programmes. Completion. In MA, acquisition of NVQ3 only; in YT, acquiring intended qualification; in German apprenticeship, having observed the apprenticeship, contract for the stipulated duration. True completion data not available for MA subsequent to introduction of requirement for Key Skills certification (September 1998). Qualification. In MA, the qualifications earned by the 4% of MA leavers who earn a non-NVQ qualification in 1999-2000 are assumed equally divided between Levels 1-2 and 3+ (and to have had the same incidence in 1998-9); Key Skills data are not available. 'Partial qualification' denotes completion of at least one module or unit towards a full qualification. The Berufsabschlu[beta] is assumed equivalent to NVQ3. Table 3. MA attributes by sector, ten largest training frameworks; England and Wales (1) (2) (3) Share of Training all duration participants Standard [**] Actual 3/2001 2001-2 1998-9 (%) (months) (months) Industry 34 39.1 19.3 Services 42 24.0 11.6 Both 76 30.8 15.2 Haterogeneity [***] 48 52 (4) (5) (6) NVQ3 Pay at LSC grant [*] completion leaving basic rate to 10.99 2000 2001-2 (% leavers) ([pound]) ([pound]) Industry 47 140 9,892 Services 28 115 4,292 Both 37 126 6,812 Haterogeneity [***] 50 20 112 Sources: (1) DfES (2001), Table 7; (2), (6), LSC (2001); (3), (4), unpublished data, DfES trainee database; (5), DfEE (2001 d). Notes: Industrial frameworks are engineering, motor, construction, electrical contracting; service ones are business administration, customer services, hotel and catering, hairdressing, health and social care and retailing. (*)for 16-18 year old entrants without previous training in same sector/occupation; lower grants payable for older and previously trained entrants, higher ones for disadvantaged entrants in particular localities. (**)basis for timing the payment of LSC grants, from 3/01; (***)absolute difference between two sector groups as percentage of ten-sector mean. Table 4. Attributes of German apprenticeship by sector, 1997 Share of all Completion Qualification apprentices (% entrants) (% entrants) Industry and commerce 45.3 81.9 78.0 Artisanal 38.9 73.7 67.1 All 100.0 78.2 73.4 Heterogeneity [*] 11 15 Source: BMBF (2000, previous year), S1, 2, U34, 37, 40. Notes: (*)Absolute difference between two sectors as percentage of all-sector mean. Total 5. Attributes of MA providers, March 2001: England Number of providers Single employer 185 Specialised training Training company 406 organisation FE college 244 Other Non-profit/charity 94 Employer group: Chamber/GTA [**] 63 Local authority 98 TEC (direct contract) 13 All 1102 Mean no Share % of trainees [*] Providers Trainees [*] Single employer 53 16.8 5.1 Specialised training 254 36.8 46.8 organisation 190 22.1 20.0 Other 218 8.5 10.8 311 5.7 8.6 127 8.9 6.0 498 1.2 2.7 All 186 100.0 100.0 Source: MAAC (2001), Tables 1 and 2; classifiable providers only (94% of all providers). Notes: (*)all work-based training programmes; (**)Chamber of Commerce or Group Training Association. 'Provider' denotes sponsor or organiser of the training programme; 'employer' has as primary business a product or service other than training. Table 6. Public expenditure on work-based programmes: England YT All programmes 1992-3 1998-9 Public expenditure (out-turn) 617 712 ([pound]m) Number in training ('000) 247 269 GDP deflator 85.4 100.0 Real expenditure per 2.92 2.65 participant ([pound]k [*]) Sources: ED (1994), Table 7: DfEE (2000f), Annexe Bii; DfES (2001), Table 2; CSO, Economic Trends Annual Supplement 2001, Table 1.1. Note: (*)at 1998-9 prices. Appendix. Comparative institutional attributes Attribute Germany Skills Training occupations yes: 356 occupations defined Minimum skill level craft Mandatory educational yes: vocational and content general Process requirements yes: duration, methods Governance Specific statutory yes regulation Meso-level committees yes: sector committees, Chambers mandatory representation a. employers yes b. unions, educators yes powers a. employer yes contributions b. certification of yes trainees, employers Specialised providers may sponsor training in place of no employer Finance Public subsidies to a. vocational full (direct costs) education b. work-based training none Apprentice status apprenticeship contract; employment contract prohibited Cost redistribution construction (voluntary) amongst employers Britain: MA Britain: YT Skills yes: 73 frameworks no craft (level 3) semi-skilled (level 2) remedial: Level 2 no Key Skills no no Governance no no yes: NTOs, LSCs [*] yes: NSTOs, TECs [**] yes yes no no no no no no yes yes Finance n.a. n.a. yes yes apprenticeship trainee status; 'pledge'; employment employment contract contract encouraged permitted construction none (statutory) Sources: Ryan (2000), (2001b). Italics within a row denote high similarity. (*)National Training Organizations (sectoral) and Learning and Skills Councils (local) (**)NSTOs: Non-Statutory Training Organisations, Lead Bodies and Industry Training Organisations (sectoral) and Training and Enterprise Councils (local).