Low-skill Work in Flux

This paper addresses the question of how far traditional assumptions about low-skill jobs are still appropriate today. Are we really dealing with activities without any particular skill requirements? How do firms proceed in filling such posts? What role is being played in this respect by ?atypical? employment relationships? The analysis is based on an evaluation of available studies on recruitment problems and low-skill jobs and case studies of recruitment into low-skill jobs in several service industries. There are several indications that job requirements are changing and becoming more differentiated, which would suggest that low-skill work is in a state of flux. Firms adopt a range of different recruitment strategies in order to fill such posts, among which flexible employment relationships and so-called mini-jobs play a role.


Introduction
The starting point for this paper is the highly controversial debate on low-wage employment that has been going on for years. Participants in the debate, from both the academic and political spheres, have frequently pointed to low-wage work, particularly in the service sector, as the most likely potential source of new jobs. It is usually implied that workers with low levels of formal qualifications, who make up more than half of the unemployed population in Western Germany, could benefit particularly from an increase in the supply of low-wage jobs. That many such jobs, even full-time ones, do not provide a living wage, and for this reason tend to be characterised as 'precarious', would seem to be beyond all question.
Excessively high wages and/or non-wage labour costs are frequently adduced as the principal obstacle to the development of low-wage jobs. In addition, it is suggested that recruitment into such jobs, whose pay levels are typically located around the bottom end of the income scale, is made more difficult or even prevented by the fact that unemployed workers have little incentive to accept them because of high transfer payments. The wider range of reasons that might militate against the establishment and filling of low-wage jobs is frequently ignored. The Institut der Deutschen Wirtschaft (2004: 5), for example, states that, "The skills mismatch (…) that otherwise plays a role ceases to be an explanation in the case of low-wage occupations -they are indeed characterised by the very moderate demands made of job-holders. If necessary, any gaps in knowledge can be filled by crash courses". The current reforms being made to labour market policy follow this line of reasoning, to the extent that a reduction in unemployment is to be achieved by, among other things, tightening the reasonableness criteria and cutting the level of transfer payments (the so-called Hartz IV programme). At the same time, a number of new instruments have been introduced with a view to widening the range of flexible employment forms, which tend to be characterised as 'precarious' and are associated with low pay and, in some cases, with reduced social protection (mini-jobs, personnel service agencies (PSA) and the "Ich-AG" as a subsidised form of self-employment).
It is quite striking that there is a relative scarcity of empirical studies of corporate employment policy in this segment of the labour market. This paper aims to fill this gap and to analyse the extent to which the argument outlined above reflects reality at firm level: • To what extent do firms actually encounter problems in recruiting workers for low-skill jobs? • What requirements must applicants for low-skill jobs fulfil? How do firms go about recruiting and selecting applicants? What role do precarious employment relationships play in this regard? • Does the equation 'low-skill job = low requirements = opportunity for poorly qualified workers' still apply? To this end, the results of company surveys will be evaluated in order to identify what they can tell us about the extent and possible causes of recruitment problems in lowskill jobs (section 2). This evaluation will provide some initial indications that the de-mands on job applicants are becoming more differentiated and partly increasing. These developments have so far attracted little attention. As we show in section 3, this applies both to studies based on segmentation theory, which characterise jobs with low requirements as 'peripheral' or 'unspecific', i.e. as jobs anyone can do, and to the arguments currently being advanced in the debate on employment policy, which amount to the equating of 'low-skill jobs' with 'low requirements'. On the other hand, the few studies that have been carried out of the skills and qualifications required in semi-skilled jobs support the argument that low-skill work is changing. These findings provide good grounds for subjecting the prevailing assumptions about corporate personnel strategies with regard to low-skill jobs to critical scrutiny. In section 4, we present the interim results of an on-going study of recruitment processes in 'low-skill' service activities, the aim of which is to investigate how firms actually go about recruiting and selecting workers for semi-skilled activities. In some cases, precarious employment relationships, such as mini-jobs, are used as a means of settling new recruits into jobs or checking whether they meet the employer's requirements.

Recruitment difficulties for low-skill jobs?
The fact that there is a large number of such vacancies registered with the Federal Employment Agency (BA) is frequently adduced as evidence of recruitment problems in the low-skill segment of the labour market, since it supposedly points to the existence of a 'mismatch' (e.g. Institut der deutschen Wirtschaft 2004: 5). However, empirical studies that have addressed this question are few and far between. The majority of them are regional surveys of recruitment problems that either concentrate from the outset on 'skilled workers' (cf., for example, Kölling 2003) or only partially differentiate the results in terms of skill requirements, if at all. In some cases, the questions are formulated in such a generalised way that it is unclear whether the answers reflect actual experiences or politically biased opinions. This may very well be one reason why the results in some cases differ considerably from each other, as is shown below.

Extent of recruitment problems
The vast majority of the studies that have been conducted to date conclude that there are indeed problems in recruiting workers for low-skill jobs but that, overall, they occur less frequently than in the search for skilled workers (Klös 2001: 46;Strotmann/Volkert 2002: 15).
However, estimates of the extent of these recruitment problems vary significantly. Thus one survey concluded that one in five jobs for which no training is required could not be filled (IHK Stuttgart 2001) whereas some other surveys estimated recruitment obstacles as being lower (c.f. for example Klös/Egle 1999: 108). Moreover, jobs with low skill requirements (or for 'low-skill' workers, as some studies put it) account for only a relatively small share of total vacancies. A survey conducted in the Rhine-Main area found that the share of vacancies for low-skill workers in total vacancies was only 8% (and had also declined from 12% in 2001) (Schmid et al. 2003: 2). This is considerably lower than the share of people who have not completed a vocational training course in the population as a whole, which stands at about 17%. A com-pany survey carried out by IHK Koblenz (1998: 8), on the other hand, produced a completely different result, with 30% of all vacancies being jobs for unskilled workers.
Only one study includes an explicit question about the experiences of companies that had actually filled vacancies for low-skill workers. Of these companies, 13.4% stated that they had had great difficulties in filling their last vacancy of this kind; however, more than half said they had had 'some problems'. Nevertheless, it is not clear what these difficulties actually consisted of. In any event, there is no direct link with the number of applications received, since the firms surveyed had received an average of eight applications for the vacancy they had advertised, although this average value conceals considerable variation. Companies in the distribution and hotel and catering sectors, in particular, complained that they had received very few applications, as did smaller firms with up to 20 employees (Strotmann/Volkert 2002: 18ff).

Requirements and selection criteria
Only a very small number of studies provide more detailed information on actual job requirements and companies' selection criteria. The most detailed, once again, is the survey conducted by Strotmann and Volkert (2002: 21). Taking the last low-skill job filled as a reference point, they found that potential employees were subject to a wide range of requirements. Reading (60.7%) and writing (53.3%) had consistently high shares of mentions, but the importance of other requirements varied considerably. In 31.1% of the jobs, heavy physical work played an important part, although in 29.3%, it 'never' played any part at all; the range of responses to questions about the need for arithmetical and/or mathematical skills was equally wide, with 36.3% of the jobs requiring 'daily' use of such skills and 37% 'never' requiring them. In 19.5% and 27% of the jobs respectively, daily telephone or face-to-face contact with customers played an important part, while 25.5% of the jobs required the use of computers. In these job categories, however, the share of jobs in which this was not part of the requirements profile, at around 60%, was very high. This indicates that the general 'low-skill' category includes jobs with a very wide rage of different requirements. However, it would seem to be debatable whether they should all be classified as 'low-skill'.
In response to the question of what selection criteria were important for the last low-skill job filled, reliability and punctuality were the two that received the most mentions (> 80%), with motivation (a good two thirds) and knowledge of the German language (just about two thirds) not far behind (Strotmann/Volkert 2002: 22). These findings are consistent to a large extent with those reported by Klös (2001: 38), although these results are not differentiated by job level. The firms surveyed in this study considered commitment and motivation as well as reliability and punctuality to be important criteria in the selection of unemployed job applicants. However, these findings are somewhat contradicted by the statement that an absence of skills or qualifications was by far the most frequent reason for rejection, ahead of inadequate motivation.
The survey results point sporadically to an inadequate match between job requirements and conditions in the wider environment outside the firm. This applies, for example, to the inadequate availability of women with children for part-time work in the afternoons (DIHK 2001: 13) or the problem of workers without access to a car applying for jobs in locations that cannot be reached by public transport.

Who is appointed?
Analyses of the structural characteristics of employees in jobs that do not have any particular specialist requirements show that by no means all of them have low levels of formal qualifications (defined as not having completed a course of vocational training) (cf., for example, Solga 2000: 4). This is also consistent with the findings reported by Strotmann/Volkert (2002: 29). No fewer than 39% of the workers who had been recruited for low-skill jobs had in fact completed a course of vocational training; 7% had the upper secondary school leaving certificate (Abitur) or the entrance qualification for advanced technical college (Fachhochschulreife), while a good 5% were university graduates. The authors conclude from this that "low-skill jobs are now no longer also necessarily jobs for workers with low skill or qualificational levels".
The other surveys provide no information on this matter. On the other hand, experiences from various in-work benefit ("Kombilohn") pilot projects in recent years show that 'low-skill jobs', which are indeed the kind the subsidies are intended primarily to generate, cannot automatically be equated with employment opportunities for workers with low levels of formal qualifications. Thus in the experimental programme "Mainzer Modell", for example, the share of participants who had not completed a course of vocational training was only around 36% (Kaltenborn et al. 2005).

Interim conclusion
Analysis of the available company surveys on the extent and causes of problems with recruiting workers for low-skill jobs has shown that, in some cases, the results differ considerably from each other. Clearly, however, there is little, if any, evidence of a general lack of applicants, in contrast to what is frequently assumed in the current debate. And while it is true that the applicants' motivation plays a relatively important role for virtually all the companies surveyed and is also very frequently mentioned as a criterion for or an obstacle to selection, it is often unclear whether respondents are really alluding here to the financial incentives for employees to accept a job at all or to the need for them to be motivated and committed to their actual work.
Some survey results also raise the question of the extent to which classifying a job as 'low-skill' is (still) actually synonymous with a job that makes no particular demands of the person holding it. Clearly, the share of workers with formal qualifications being recruited to low-skill jobs is by no means small. And the spectrum of requirements, from technical skills (e.g. 'working with computers') through personal characteristics (e.g. physical fitness, 'well-groomed appearance', age) to social competences (e.g. reliability, punctuality, customer contact etc.), seems to be fairly wide and differentiated.
On the other hand, the available surveys provide no information at all as to the share of vacancies involving flexible or precarious employment forms. We will return to this question in section 4, when we turn to our own empirical investigation of recruitment strategies.

Low-skill jobs -work for anyone?
In this section, we focus on the assumptions about low-skill jobs that are made by both labour market theorists and policy-makers. Labour market experts who draw on segmentation theory speak of 'peripheral', 'unspecific' or even 'undemanding' jobs, i.e. jobs that can be done by virtually everybody. And in some of the arguments being advanced in the current employment policy debate, it is assumed that there are still jobs that can be done by anyone. However, empirical studies suggest that changes are taking place in low-skill jobs that tend to give the lie to such sweeping statements and assumptions.

'Jobs that anyone can do' -in theory…
Analyses based on segmentation theory divide the labour market into various segments (internal/external or primary/secondary) that have different characteristics in terms of employment stability, skill requirements and working and employment conditions. Particularly in those studies that applied the segmentation theory developed in the USA to the German labour market (e.g. Lutz/Sengenberger 1974;Sengenberger 1975 andLutz 1987), workers' skills and qualifications and/or the skill requirements associated with particular jobs function as important differentiating criteria. In this 'firm-based segmentation approach', the labour market is divided into three submarkets: the market for firm-specific skills, otherwise known as the internal labour market, that for occupational skills or qualifications, known as the occupational labour market, and the market for non-specific skills, also known as 'Jedermannsarbeitsmarkt'. In this last segment, firm-specific or occupational skills and qualifications by definition play no role: "In the extreme case of a job anyone can do ('Jedermannstätigkeit'), an individual's work capacity is irrelevant to the completion of a transaction in the unspecific labour market, since only the most general capabilities (such as the minimum skills required for civilised living) are required, which the vast majority of those on the supply side of the labour market can be assumed to possess" (Lutz 2002: 19; a similar argument can be found in Henneberger/Kaiser 2000: 12).
What jobs can be described as 'extreme cases' and what relevance the term still has for current practice are questions that academic debates of recent years have scarcely touched on. True, the question of how far the categories of segmentation theory are able to capture the heterogeneity of labour market structures in a sufficiently differentiated way has repeatedly been the object of critical scrutiny and prompted further development. In many cases, the impetus for this has been the application of the categories to changed labour market structures. Ultimately, the particular characteristics of the East German labour market set in train a thoroughgoing modification of the concepts used in segmentation theory (Lutz 2002;Köhler et al. 2004a). However, the central question that concerns us here, namely the extent to which assumptions about the unspecific labour market are still relevant to today's so-called 'low-skill' jobs, seems to us to have played a subordinate role.
With the definition of 'undemanding jobs' as those requiring a training period of up to two months (Köhler et al. 2004b: 71) and the identification of a low-skill segment within internal labour markets in which, in contrast to low-skill workers in the external market, workers are given firm-specific on-the-job training (Henneberger/Kaiser 2000: 18ff), the importance of skills outside corporate employment strategies for the intermediate and upper skill levels has now been explicitly recognised. In principle, this opens the way for a differentiated investigation of the bottom segment. However, even this would not appear to offer protection against generalised classifications of individual jobs. Henneberger/Kaiser (2000: 50), for example, state in their analysis of future prospects: "The increase in service activities (…) may also lead to an expansion of the low-skill segment in the external labour market. This would be likely if jobs are created in personal services, which they have been in large numbers in the USA for a long time. If such a trend were to emerge in the German-speaking countries, it could help to reduce unemployment among low-skill workers."

… and policy
This quotation brings us full circle to the starting point of our investigation, namely the employment policy debate on low pay and low-skill jobs in the service sector. It is often assumed that an increase in the number of jobs in this sector could particularly benefit workers with low levels of formal qualifications, who account for a highly disproportionate share of unemployment. However, the question of which particular services and activities are being alluded to here frequently remains unanswered. In many cases, reference is simply made to the relatively low level of productivity in these activities or to customers' reluctance to pay for the services in question; in some cases it is merely pointed out that the shares of employment in the segments in question are lower in Germany than in other countries, such as the USA (Klös 1997).
When particular areas of activity besides retailing, repair and maintenance activities, hotels and catering and tourism are mentioned, the main ones identified -typically without any further clarification -are "community and personal services" (Zukunftskommission der Friedrich-Ebert-Stiftung 1998: 253), "personal, social and household services" (Streeck/Heinze 1999: 155) and "low-skill, low-productivity and usually personal services" (Fels et al. 1999: 2). More precise details are provided by, for example, the Institut der Deutschen Wirtschaft (2004), which mentions nursing and care services and private households, and the Kommission für Zukunftsfragen der Freistaaten Bayern und Sachsen, which also mentions care of the elderly and of children. This list makes it clear that activities traditionally carried out predominantly by women are often characterised as 'low-skill' (cf., for example, Meifort 2003: 35). This continues the tradition of underestimating abilities and skills that in many cases are regarded as typically feminine and which are often assumed to have no particular technical or professional content. This point of view is represented unambiguously elsewhere in this latter report: "These services are 'low-skill' in the sense that people can provide them without any particular knowledge or skills" (Kommission für Zukunftsfragen der Freistaaten Bayern und Sachsen 1997a: 117).
Such clarity is not to be found in the report published by the Hartz Commission. Nevertheless, here too, with regard to the objective of reducing clandestine employment in private households and thereby creating more legal employment, the report states that, "Subsidised forms of self-employment for individuals and families, as well as 'mini-jobs', provide an attractive framework for the provision of services. There is also a right to low-skill work. Many people do not wish or are unable to take part in further training" (Hartz et al. 2002: 41).
In all these statements, there is an implicit assumption that no particular skills are required in order to do the jobs on offer in many service industries. If skills and qualifications are by definition irrelevant, then wages and labour costs become the focus of reform efforts. In our view, however, this assessment falls short of the mark. In what follows, we will attempt to justify this statement in greater detail.

Low-skill jobs in flux
In our view, one fundamental shortcoming of the debate on low-skill jobs is that there has been no attempt to compare the assumptions that are frequently made with firms' skill requirements, the specific characteristics of the jobs concerned and current developments in these areas of work. "The lack of attention paid to firms in addressing these (…) fundamental questions constitutes a considerable shortcoming. Ultimately, it is firms that determine present and future requirements and hence also the employment opportunities of low-skill workers." (Strotmann/Volkert 2002: 5) Closer examination indeed reveals a good deal of evidence to suggest that even so-called 'low-skill' jobs are changing and that the demands made of workers are increasing. One important driving force in this regard is, in our view, the on-going process of tertiarisation. In service industries, supposedly 'low-skill' jobs frequently involve direct contact with customers, which places increased demands on employees' social competences, since they have to react flexibly and capably to a wide range of different expectations (Weinkopf 2002). Recent publications on so-called 'interactive service work' suggest that such work can be standardised only to a limited extent. One of the consequences of this is that service quality is strongly influenced by employees' behaviour and competence (Voswinkel 2000).
However, low-skill jobs in the manufacturing sector are changing as well. In the past, work in manufacturing industry was frequently divided into planning, organisational and execution tasks, with each set of tasks being performed by different categories of workers (Taylorisation). Today, the influence of new production systems, customeroriented manufacturing and so on is clearly spreading to low-skill jobs as well: "In order to fulfil the demand for flexibility, planning, supervisory and executions tasks are increasingly being combined. (…) The basic innovation is that low-skill tasks and the associated requirements profiles are no longer defined by individual jobs but by the work environment as a whole" (Zeller et al. 2004a: 32f).
And even if individual activities considered in isolation can still be classified as 'low-skill' (e.g. cleaning, simple clerical work, warehouse work, driving and courier services), nevertheless increasing demands often arise out of the fact that such activities, which require only a limited number of hours' work or are carried out irregularly (particularly in small and medium-sized firms), are being combined (Kay/Kranzusch 1999: 5;Falk/Klös 1997). In our view, this reflects the considerable importance attached to functional flexibility, which has hitherto been regarded as characteristic of the 'German model'. Comparisons with the US model, which is characterised by a more pronounced division of labour, show that this model does indeed create more employment opportunities for low-skill workers but at the same time requires more skilled personnel for supervisory and guidance tasks (Schettkat 2002: 64).
In sum, we can say that forces driving change in low-skill work can be identified at various levels: shifts in work content (more services, more customer contact) and a broadening of activity profiles, changing technical or professional requirements ("both professional competence and process competence" - Zeller et al. 2004b: 51) and the closely related trend towards greater demands on workers' social competences: "It is no longer simply a question of carrying out a certain work task as per instructions; rather what is required of workers is self-responsibility and the commitment to ensure that an entire series of operations is performed smoothly. Consequently, workers have to take into account the wider environment in which they perform their own duties, as well as the processes upstream and downstream of their own work." (Böhle 2004: 102)

Firms' recruitment strategies
In classic segmentation theory, one consequence of the low skill requirements in the bottom segment of the labour market is that employees can move relatively easily between firms and no significant costs are incurred for dismissal or initial skill adaptation training: "Since workers from the unspecific segment of the labour market have no firm-specific training, they can be dismissed without cost and also hired again at the relevant market wage rate for non-specific skills" (Henneberger/Kaiser 2000: 14, see also Sengenberger 1975. However, few empirical studies have as yet paid much attention to firms' recruitment procedures in this segment of the labour market. In our view, the trends outlined above are sufficient evidence of the need for an upto-date empirical survey. If it is assumed that specific capabilities and qualities are increasingly being demanded of employees even in low-skill jobs, the assumptions about recruitment processes referred to above begin to lose plausibility, which in turn raises the question of how firms actually do go about finding suitable personnel. This question is also of relevant to policymakers, since measures intended to improve the employment chances of low-skill workers may well prove fruitless without more precise knowledge of firms' requirements and recruitment strategies. Against this background, we have been carrying out a study since mid-2003, commissioned by the Federal Ministry of Economics and Labour, which uses case studies in 25 companies in four service sectors (elder care, the retail trade, commercial cleaning and the hotel trade) and a total number of 107 interviews in order to identify the strategies firms adopt when seeking workers to fill 'low-skill' job vacancies. Of course, we have also been trying to ascertain the extent to which they experience difficulties in getting applicants for such vacancies. None of the firms surveyed has reported serious problems; on the contrary, most of them complain of 'being flooded by applicants'. In so far as they have reported any problems at all, these have mainly concerned the number of applicants unable to offer the work schedules or volume of hours they are looking for.
As far as recruitment is concerned, our preliminary results show that firms adopt very diverse strategies, which differ from each other in the following respects (among others): • the scope and specific nature of the demands made of the employees concerned (skill requirements); • the extent to which firms invest in initial skill adaptation training for newly recruited employees (length and intensity of initial skill adaptation training); • the point at which firms conclude the selection process: are the selection criteria applied at the point of recruitment or during or after the initial training and probationary period? (timing of selection). These three criteria can of course be combined in many different permutations, but there are three that best represent the various strategies adopted in the firms we surveyed. They are shown in Table 1 and are described in greater detail below.

'(Self-)selection'
Companies that adopt this recruitment strategy largely eschew the use of specific selection criteria when recruiting workers. They assume that anyone 'of good will' is capable of carrying out the tasks in question. The initial skill adaptation training period is correspondingly short. Nevertheless, this does not mean that each person recruited also stays with the company. The actual decision as to who is offered a permanent position is taken within the first few weeks. New recruits are frequently dismissed during this period. Both employers and employees take part in the decision-making process. Employees decide during the initial skill adaptation training and probationary period whether they are able and/or willing to carry out, on a long-term basis, the tasks required of them on the terms and conditions laid down by the employer. The main reason why employers, for their part, dismiss new recruits is that they fail to carry out the routinised tasks as quickly as required. In this sense, it is not so much skill requirements as efficiency or productivity criteria that are the decisive factor. Some of these companies recruit for preference from their temporary or seasonal workers or holiday replacement staff, who have already proved themselves in this respect. This type of strategy is the one that corresponds most closely to the personnel policy strategy in the unspecific segment of the labour market described above. It is particularly common in typical 'back-office' activities that do not involve any significant customer contact, such as maintenance cleaning, shelf stacking in retailing or goods packing in the mail order business. However, the example of maintenance cleaning makes it clear that meeting the required efficiency or productivity criteria is by no means merely a question of motivation (Jaehrling 2004) and that the corresponding 'efficiency' or 'productivity' cannot readily be distinguished from competences. The ability to carry out simple, routine tasks at high speed while at the same time meeting minimum quality standards requires certain routines that have to be learnt on the job. Whether such tasks can actually be performed by just anyone also depends on whether workers are granted the time to acquire this ability or whether they are required to meet the performance criteria in full from the outset.

'In-house training'
This type of strategy goes hand in hand with a longer period of initial skill adaptation and also, if necessary, more highly formalised forms of initial training. The requirements in terms of professional or social competences are not necessarily any more rigorous at the time of recruitment than in the case of the '(self-)selection' strategy, but they are more comprehensive in the long term. The probationary period is used relatively intensively in order to train the new recruits and at the same time assess their suitability. Here too, therefore, the actual decision on whether or not to appoint permanently is postponed until the period after recruitment. If new recruits do not fulfil the requirements, they are given their notice during the probationary period. Since wrong decisions give rise to costs even when the initial training period is short, some of the firms that pursue this strategy try to minimise such costs by giving preference to applicants who have already proved themselves. They may already have been employed as hourly-paid temporary workers, holiday replacements, seasonal workers or, in some cases, as temporary agency workers, that is on contracts that can be described as 'precarious' because of short working times or a short contractual period. In this case, firms are able, when necessary, to draw on a pool of labour that has not only work experience but also firm-specific knowledge at the time they are offered permanent employment.
In practice, we encountered this strategy most frequently in the care sector. It is facilitated and encouraged by the increasing use of trial work periods and work placements as well as by more recent labour market policy initiatives, such as PSA; these initiatives make workers available for trial periods at relatively low cost, sometimes even at no cost at all, and are being intensively used by employers, particularly for auxiliary jobs in elder care.

'External sourcing'
One of the characteristics of this strategy is that the skill requirements are more specific but are not accompanied by longer initial skill adaptation periods. Instead, the competences required are purchased in the market. In consequence, applicants must already be in possession of them when they take up their new position. In contrast to the two other strategies, firms are actually confronted by the challenge, identified above in theoretical terms, of having to assess which applicants meet their requirements without the aid of standardised qualifications. Several firms try to resolve this difficulty by recruiting from particular groups of people who are assumed to have the required competences (e.g. students and women returning to the labour market). For some jobs, qualifications that validate particular abilities or specialist knowledge are required (e.g. a fork lift driver's certificate). In addition, many potential employers attach considerable importance to work experience. This need not necessarily have been acquired in the same area of activity, but may equally well involve experience or even vocational qualifications in related or even unrelated areas of work. The social competences that are important in service work, such as an ability to work independently or to interact with customers, are frequently regarded as particularly transferable.

Interim conclusion
The basic types of corporate recruitment strategies outlined here show that the requirement profiles of unskilled and semi-skilled jobs differ in the importance attached to productivity and skill requirements and personal characteristics. Only in the first strategy do the skill requirements tend to be on the low side. In this sense, the '(self-)selection' strategy is the one that most closely matches the characteristics typically attributed to jobs in the unspecific sub-market. However, it is clear from the company case studies that employers make demands related to efficiency, physical fitness and other personal characteristics such as age and appearance that can lead to the preclusion of certain applicants.
The length of the skill adaptation training period varies, depending on the requirements of the post to be filled, on the one hand, and the requirements stipulated for individual applicants, on the other. A short skill adaptation training period of up to two months is not in itself a suitable criterion for classifying jobs as 'unspecific' or 'undemanding' (as suggested by Köhler et al. 2004b: 71). The requirements that have to be met on recruitment must also be included. In the 'external sourcing' strategy in particular, these requirements are frequently by no means inconsiderable.
The third differentiating criterion concerns the timing of the actual decision on hiring and/or continued employment. In the 'external sourcing' strategy, this decision tends to be taken at the time of recruitment; in the other two strategies, in contrast, it is taken until later, after a probationary period. This probationary period may be structured in different ways but it is deliberately used by firms, particularly those following the 'in-house training' strategy, to review recruitment decisions. The chances of a new recruit being retained also depend on the length of time firms allow workers to achieve the performance levels required of them, with regard to both quality and volume of work.
It is also important to emphasise that we observed considerable differences between firms in respect of their recruitment strategies, even for similar kinds of jobs and activities. Further analysis is required of the circumstances under which firms adopt one or other strategy. Besides the regional availability of suitably qualified workers, another important factor in this respect would seem to be the extent to which firmspecific knowledge is required and what costs are incurred during the skill adaptation training period.

Summary and conclusions
Our objective in this paper has been to make the case for a differentiated perception and analysis of low-skill jobs and to give some insight into firms' recruitment strategies. Our analysis has shown that the assumption that many firms experience difficulties in recruiting for low-skill jobs is only partially supported by the available survey evidence. Indeed, in our own company case studies on recruitment for 'low-skill' service jobs, the overwhelming majority of HR managers even emphasised that, on the contrary, they had to deal with an excessive number of applications. With regard to the requirements applicants for low-skill jobs are expected to meet, our 'search for clues' in the relevant surveys and studies revealed that the judgement that equates lowskill jobs with low requirements is too sweeping. In fact, the requirements profiles differ considerably, which makes any general classification as 'undemanding' jobs seem questionable. Over and above any such general classification, the decisive factor is the recruitment policies firms adopt, which result in different barriers to entry even when the requirements profiles are the same. Some firms, on the other hand, try to avoid incurring any costs for initial skill adaptation, either by stipulating the competences job applicants are expected to have on recruitment or making use of employment forms such as temporary work, marginal part-time jobs or work placements that tend to be classified as 'precarious' as a sort of 'trial period' prior to recruitment. Whether applicants with low levels of formal qualifications have a chance of obtaining such jobs is a question that cannot be answered in general terms; their chances depend, rather, on the selection criteria companies adopt. However, we did find some evidence that especially in care and retail firms quite often seek to recruit workers with formal qualifications for low-skill jobs, a strategy that is probably not unconnected with the high unemployment rate in Germany.
Even firms that adopt the '(self-)selection' strategy stipulate certain requirements for job applicants. These may include, firstly, high efficiency or productivity targets, fulfilment of which requires the learning of routines; even with the aid of such routines, not everyone can achieve these targets. Secondly, particular recruitment paths may be used, even for jobs with the lowest productivity and skill requirements. For example, if a firm seeking to recruit permanent staff gives preference to temporal and seasonal workers or to holiday replacement staff, this means that potential applicants must first have worked for the firm on contracts that do not generally provide a living wage and must have 'proved' themselves during this period of insecure employment to be worthy of a permanent job.
In our view, the results of our analysis show that it is time to bid farewell to the simplistic formulas that have been put forward as solutions to Germany's employment problems. In the interests of improved effectiveness, labour market policymakers would do well to identify targeted ways of supporting unemployed people with low levels of qualifications in order that they do not lose access to potential employment opportunities. Recruitment to a permanent position following a period as a 'probationary employee', a common occurrence with PSA initiative, seems to be successful in several companies. However, there is as yet little information on how such periods of 'precarious employment' impact on subsequent career development and future employment prospects. A recent analysis suggests that between 20 and 40% of all lowskill employees in Germany are working in mini-jobs, which do not offer the opportunity to earn more than 400 € per month (Reinberg/Hummel 2005). A further recent study shows that upward mobility among low earners has deteriorated significantly in the last two decades (Rhein et al. 2005). This may be an indication that the current labour market reforms, which are intended to increase the number of low-skill jobs with reduced social protection and low earnings should be accompanied by training programmes deliberately designed to help those in poorly paid jobs to be upwardly mobi-le. Otherwise, such precarious employment forms may turn out to be dead ends. Against this background, the sharp decline in the provision of further vocational training programmes in recent years would seem to be extremely problematic, despite the fact that there has been much talk elsewhere of 'lifelong learning'. Not only in the light of demographic developments, a change of direction in this respect seems to be urgently required.