‘Training floors’ and ‘training ceilings’: metonyms for understanding training trends

This article outlines a conceptual framework for mapping and understanding training trends. It uses the metonyms of floors and ceilings to distinguish between different types of training configurations. The argument is made that the ups and downs of employer reports of training activity are a crude basis on which to make judgements about the resilience or otherwise of training to the economic cycle. The article, therefore, suggests and demonstrates with qualitative evidence that using the metonyms of ‘floors and ceilings’ provides a more nuanced understanding of the multiple pathways that lead employers to increase, decrease or maintain training activity during an economic recession.


Introduction
A key challenge in recent times has been to improve our understanding of trends in employer-funded training in times of recession. The purpose of this article is to develop existing metonymsshorthand representations of conceptsand propose complementary ones in response. Based on a study which examined the impact of the 2008-2009 recession on workplace training, the article develops the notion of 'training floors'originally proposed by Felstead and Green (1994) and their corollary 'training ceilings'introduced in this article for the first time. These two complementary metonyms are advanced as a framework to deepen our understanding of training trends, especially in times of recession. To illustrate their analytical value, we use extracts taken from qualitative interviews carried out in the course of the research to investigate the meanings behind some of the trends.
Early on in the recession of 2008-2009 policy-makers in the UK were alarmed that training activity would be cut back as the economy weakened and moved into one of the longest and deepest recessions in living memory. This concern was expressed in general and specialist commentaries on the impact of the recession on training at the time (e.g. Charlton 2008;Eyre 2008;IoD 2009;Kingston 2009). In response, the UK Commission for Employment and Skills (UKCES) together with the Confederation of British Industry (CBI), Trades Union Congress and some of the UK's senior business leaders published an open letter which called on UK employers not to cut training in the recession (UKCES 2008).
These early concerns were fuelled by surveys such as the CBI Industrial Trends Survey which asked employers about their training intentions. The results from these 'barometer' surveys are often presented as indicators of business optimism on a range of matters, including training. The likely path for trainingas for other indicatorsis measured by the difference between the percentage stating they intend to increase training expenditure and those saying that they intend to reduce it. Such surveys are relatively quick to carry out and it is presumed they give an early indication of trends. These results have a wide audience and feed into 'decision-making processes at the Bank of England and the Treasury' (Bye, Wood, and Bush, 2008). According to this evidence, employers became exceedingly pessimistic about future training prospects in October 2008 and remained so for five quarters. The gap between those planning to increase and those planning to decrease training expenditure reached an historic high of 30 percentage points in April 2009. Similar levels of pessimism with respect to training were recorded in the British Chambers of Commerce which also carries out survey of its members' training activities (Nolan et al. 2012).
The concepts presented here were developed in the context of a study of the impact of the 2008-2009 recession on training in the UK which aimed to examine whether this level of pessimism was warranted. The study also examined the theoretical basis used by academics to understand and interpret the actions of employers in making training cutbacks in recessions (Brunello 2009;Jewson 2011, 2012). Our aim was to discover whether the economic downturn of 2008-2009 had resulted in cuts in training provision, and to explore the reasons that employers and managers gave for their decisions. Accordingly, we initially sought to discover whether organisations had increased, decreased or maintained training during the recession and its aftermath. In this respect, we followed the protocols used in the barometer surveys outlined above. However, we soon became aware that this conceptualisation was far too simplistic. Our qualitative interviews revealed the complexities that lay behind these apparently straightforward directions of travel. Organisations differed in their starting points, their pathways and their destinations. Some types of training proved more amenable to cuts than others. In some organisations, training was embedded in and elided with broader managerial strategies; in others, it was isolated within discrete silos. In some organisations, training was critical to maintaining occupational skills and integral to personal identities; in others, it was a cost, an interruption or a chore.
Our qualitative research, then, revealed the complexities of the situation. For example, one of our case study organisations, which had spent no more than the legal minimum on training, could legitimately claim that it had not cut training during the recession. The statement was true, but was not indicative of a proactive and lively training culture. In contrast, a second organisation, that routinely spent a great deal on training, decided to weather the market downturn by temporarily delaying aspects of its extensive 'nice to have' programmes. This strategy might be taken as evidence of a negative attitude to training. However, even after retrenchment, the latter organisation was offering employees far more than the former. A third enterprise which had raised expenditure on training during the recession might be portrayed as a good news story. In reality, the increase represented no more than amends for past neglect, forced on management by the demands of new customers. As our research proceeded, therefore, we looked for ways to differentiate between different approaches to and strategies for managing training over time. A simple distinction between organisations which increased, decreased and held steady did not do justice to the situation.
In generating a more nuanced view, we developed the concepts outlined below. Rather than addressing our original research question, this article outlines a new conceptual framework that we derived in approaching our analysis. It is, then, a contribution to theoretical debates. In order to illustrate our argument, in the second half of the article we have incorporated some brief case studies of a few of the organisations that participated in our study. We hope they provide a glimpse of the rich narratives generated by our interviews and leaven the otherwise abstract nature of the text. However, this article is not a presentation of our empirical findings. Rather, it is an outline of an intellectual toolbox for analysing processes of change to training programmes. Although we present our analysis and case studies with specific reference to the 2008-2009 recession and its aftermath, the concepts and associated metonyms developed below may be deployed in identifying training configurations and tracking training trajectories over any phase of the business cycle, or indeed other time periods.
The article proceeds as follows. In Section 2, we provide a short account of our research project, and, in particular, the qualitative interviews from which the arguments presented here are derived. Section 3 then moves on to define and outline the metonyms of training floors and training ceilings as a shorthand way of referring to the conceptual distinction between what training organisations have to do and what is discretionary. In other words, the metonyms suggest a variety of training configurations. Section 4 underlines the empirical utility of these metonyms, by illustrating from our fieldwork some of the real-world scenarios and histories of training in recession. These are presented in a series of boxed vignettes. In Section 5, we map the ways in which training configurations may, in theory, change during economic downturns. By retaining the division between those that increase, decrease and maintain training as used in barometer surveys such as those carried out by the CBI, we elaborate on the different possibilities that may lie behind reported changes to training activity. Section 6 ends the article with a short summary of the value of the intellectual toolkit we provide.

Research study
The objective of our research project was to investigate the impact of the 2008-2009 recession on workplace training in the UK. It included both quantitative analysis of national data sets and qualitative interviews with a sample of employers. It is the latter which provides the basis of this article.
Interviews were conducted with a quota sample of organisations derived from the 2009 National Employer Skills Survey (NESS). This is a biennial survey of employers which began in 1999 (DfEE 1999, 90). In 2009, over 79,000 employers were interviewed (Davies et al. 2012). The qualitative interviews reported here were follow-ups to a selected number of respondents who took part in that survey. Their contact details were supplied to us by the UKCESthe current custodian of the NESS series. This sample was taken from respondents who agreed to be involved in 'further work on related issues in the future' and to be contacted by 'approved contractors'in this case, the research team for this project.
Respondents were selected to include establishments of various sizes, public and private sector, manufacturing and services, and those with and without apprenticeship schemes. Contact was established with personnel who had detailed knowledge of the training programmes on offer in their employing organisations. When investigating very large employers, we focused on one site or a cluster of related locations.
In the first phase of interviews, conducted in mid-2010, representatives of 105 organisations were interviewed by telephone (60 private sector and 45 public sector). Subsequently, in late 2011 and early 2012, 72 of these were interviewed for a second time, also by telephone (42 private sector and 30 public sector). The first interview addressed the effects of the recession in the previous two years (i.e. 2009-2010). The second interview asked respondents for an update on events in the previous 12 months and to look back over the whole period of the downturn and its aftermath.

Training floors and training ceilings
Central to our analysis are two metonymsone based on previous research, the other developed in the course of the research reported here. The latter is the 'training ceiling'. This is a shorthand way of referring to the totality of all training offered within an organisation. Training ceilings include all types of training opportunities available to staff by virtue of their participation in employing organisations. A greater or lesser proportion of the elements that make up training ceilings are legally required and functionally indispensable programmes. These non-discretionary forms of training are referred to using the metonym of 'training floors'. Training ceilings, then, include training floors but may, to a greater or lesser extent, incorporate additional programmes at the discretion of the employer. That portion of an organisation's training ceiling which is discretionary is that which occupies the space, or gap, between the floor and the ceiling. It comprises training activities that are not legally mandatory or operationally essential.
Training ceilings typically, then, include some provisions that are non-discretionary, indispensable and relatively inflexible, as well as some that are discretionary and relatively malleable. Many of our respondents referred to the former as the 'must have' and the latter as the 'nice to have'. It is the 'nice to have' portion of the total training package offered within organisations that is most immediately vulnerable in times of economic hardship. Thus, the distinction between discretionary and non-discretionary training figured prominently in our analysis of the impact of the 2008-2009 recession.
The metonym of training floors was developed by Felstead and Green (1994) in their analysis of the recession of the early 1990s. They have two main sources. First, training that is defined as mandatory by legislation, statutes, regulations and codes of practice. Our respondents commonly called this 'compliance training' or 'mandatory training'. The sources of such requirements are varied, including, inter alia, central and local government, professional associations, industry codes, corporate bodies, quango regulations, international treaties and EU regulations. Second, training that is essential in order for employees to participate in operational procedures that are integral to the production of goods and services. These entail mastering the basic physical, intellectual or social skills entailed in doing a job. Not only are training floors non-discretionary, they also frequently identify specific channels through which certified training can be obtained. In some cases, only accredited providers of particular qualifications are eligible to provide such training, which is often formal in character and assessed. This makes it difficultthough not impossiblequickly to devise alternative and cheaper ways of discharging these obligations. However, training floors are far from uniform in extent and intensity. Some organisations and economic sectors are confronted by more demanding training floors than others.
In some organisations, training floors comprise a high proportion of total training programmes and expenditure. In these circumstances, training ceilings are almost wholly accounted for by non-discretionary 'must have' provisions. In others, training floors constitute a relatively small proportion of total training. Here, substantial resources are devoted to programmes above and beyond the legally prescribed or operationally necessary.

Using floors and ceilings to understand training trends
The metonyms of training floors and ceilings allow us to map contrasting patterns of training within organisations. They also make it possible for us to trace directions of change over any given time period, such as a recession and its aftermath. The particular juxtaposition of training floor and training ceiling within an organisation we designate as its training configuration. We can identify a range of possible training configurations.
Processes of change in training configurations we refer to as training trajectories. Drawing on the metonyms of floors and ceilings, we are able to scope out a wide range of potential trajectories taken by organisations during recessions. Each of these pathways will have distinctive consequences for economic productivity, employee development and the lived experience of participation in the labour process. Our approach allows us to spell out a richer and more nuanced account than that offered by simply dividing organisations into those that increase, decrease and maintain training.
Some organisations respond to economic downturn by lowering their training ceilings; i.e. reducing the total amount of training offered to employees. These organisations we call 'cutters'. However, cutters vary considerably in their starting points, trajectories and destinations. Most cutters lower their training ceilings by reducing their discretionary training, while leaving their non-discretionary training floors unchanged (Table 1, position 1.1). This is the most common trajectory followed by cutters. However, theoretically there are other possibilities, although Table 1. Training trajectories characteristic of cutters.

Possibilities
Training floor Training ceiling Difference between the floor and the ceiling empirically they are quite rare. Cutters may lower their training ceilings by reducing their non-discretionary training floors, while maintaining their discretionary training (Table 1, position 1.2). This is inherently difficult because floors, by definition cannot be breached, and so such a situation is a rarity. However, it is sometimes achieved by changing the mode of delivery of non-discretionary training. Another possibility is that cutters may lower their training ceilings by reducing both discretionary training and non-discretionary training floors (Table 1, position 1.3). For example, organisations teetering on the brink of insolvency might cease training of all kinds, at least temporarily. Other cutters lower their training ceilings by either increasing their training by an amount less than the cuts they make to their training floors (Table 1, position 1.4); or by increasing their training floors by an amount less than the cuts they make to discretionary training (Table 1, position 1.5). An example of the former would be an organisation that found itself in a position to make savings on the provision of non-discretionary training, part of which it diverted into discretionary programmes. An example of the latter would be an organisation facing rising requirements to provide non-discretionary training that responded by slashing discretionary provisions.
In contrast to cutters, some organisations raise their training ceilings during a recession, thereby increasing the total amount of training they offer employees. These organisations we call 'boosters'. They, too, vary considerably in their starting points, trajectories and destinations. The most common trajectory taken by boosters is to raise their training ceilings by expanding their discretionary 'nice to have' training, while training floors remain unchanged (Table 2, position 2.1). Boosters adopting this trajectory may include organisations with markedly different training configurations. Some boosters raise their training ceilings by increasing their nondiscretionary training floors, while leaving their discretionary training unchanged (Table 2, position 2.2). For example, during the 2008-2009 recession, some legal and financial enterprises found that statutory regulations surrounding money laundering were enhanced, requiring additional staff training. Other boosters raise their training ceilings by increasing both discretionary and non-discretionary training (Table 2, position 2.3). Another group of boosters may raise their training ceilings either by increasing their discretionary training by an amount greater than the cuts they make to their training floors (Table 2, position 2.4); or by increasing their training floors by an amount greater than the cuts they make to discretionary training (Table 2, position 2.5). An example of the former would be an organisation that opted to increase its overall training spend at a time when non-discretionary training regulations were diminishing. An example of the latter would be an organisation Table 2. Training trajectories characteristic of boosters.

Possibilities
Training floor Training ceiling Difference between the floor and the ceiling facing a steep rise in non-discretionary training requirements which it met partly, but not wholly, by transferring funds from discretionary programmes. Some organisations respond to economic downturn by maintaining their existing training ceilings; i.e. the total amount of training they offer to employees. These we call 'stickers'. However, stickers, too, vary considerably in their starting points and destinations. Most stickers preserve their training ceilings by maintaining a steady state for both discretionary and non-discretionary training (Table 3, position 3.1). Organisations following this trajectory may include those with different training configurations. Some stickers maintain their total expenditure on training either by increasing training floors and decreasing discretionary training by equivalent amounts (Table 3, position 3.2) or by reducing training floors and increasing discretionary training by equivalent amounts (Table 3, position 3.3). The steady state of their training ceilings is a product of shifts in the boundaries between training floors and discretionary training. An example of the former would be an organisation with a fixed training budget that compensates for a rise in non-discretionary training requirements by reducing discretionary programmes. An example of the latter would be an organisation that encounters a fall in non-discretionary training regulations and diverts released funds into discretionary training. In both scenarios, stickers following this trajectory may include those with markedly different training configurations.

Illustrations of training metonyms at work
Our conceptual analysis indicates, then, that cutters, boosters and stickers are a mixed bunch. It should not be assumed that all cutters, boosters and stickers undergo the same processes and confront the same outcomes. Each category includes organisations with divergent training configurations and trajectories. It is not possible to examine all the possibilities here; indeed, their full investigation constitutes a research programme in itself. We have, however, selected a number of salient scenarios, revealed by our qualitative interviews, in order to illustrate the dynamics of some of the complex empirical situations highlighted by our conceptual approach.
It might be anticipated that organisations that spend little on training and give it a low priority become cutters in a recession. However, in practice, their scope for cutting is very limited. Such organisations often adopt a minimalist strategy, taking care of their legal obligations and ensuring their employees have sufficient training to carry out their work tasks, but with very little expenditure beyond this baseline. As a result, total expenditure on training of all kinds is restricted, training ceilings are low, and 'must have' non-discretionary training comprises a high proportion of the small amount on offer. Hence, it is difficult if not impossible to reduce training expenditure. Costs might be reduced by innovating in mode of delivery or sourcing elsewhere. However, this option is constrained and, in any case, is unlikely to yield large sums. Savings are more readily found in other directions; for example, by reducing staff numbers. Organisations with this type of training configuration, then, are unlikely to become cutters. In our research interviews, routine manufacturing companies and organisations under the threat of transfer of ownership were among those adopting this training configuration. Food Processing Company 1 is an example of the former.
Food Processing Company 1 employed a core of permanent employees and larger group of temporary staff, depending on seasonal demands. Training of shop floor workers was confined to basic food hygiene and health and safety. Team leaders were selected haphazardly and received little preparation for their roles. Managers were only offered add-on versions of shop floor training. There were no training plans, training timetables, mentoring schemes, coaching arrangements or apprenticeships. The company had been struggling to remain profitable for several years and the recession of 2008-2009 intensified pressures considerably. It responded by switching from producing upmarket items to mass market, lowprice commodities. Ruthless cost cutting and waste reduction in all aspects of the business, which had been the norm for several years, was intensified. Substantial redundancies were declared. Although training expenditure was low, further economies were made by shifting from professional external trainers to ad hoc in-house provision by co-workers. The basic system of training remained in place because there was no alternative. Having entered the recession with low training floors and ceilings, the company attempted to weather the downturn by deploying more of the same.
Logically, it is organisations with high levels of discretionary training that can most readily cut training in a recession. Such a situation is typical of organisations that invest heavily in the wider personal and professional development of their employees. Some of these also support 'must have' training floors that consume considerable resources. These are organisations with high training floors and ceilings. They have the scope, potentially, to cut training expenditures, either marginally or substantially, in response to an economic downturn. Although they cannot avoid maintaining a programme of non-discretionary training, they have the option of reining in discretionary provisions. Furthermore, because their training bill is substantial, shifts in the mode or frequency of delivery of discretionary training may generate significant savings. In our sample, Financial Services Company 1 followed this trajectory, becoming a cutter.

Financial Services Company 1 is a large organisation with a number of divisions.
Our interviews focused on a cluster of sites, employing 500 staff, engaged in customer-facing administration of financial products. High training floors reflected the requirements of Financial Services Authority regulations and government legislation. Annual obligatory training included updates on specific products, client needs and customer service priorities. Prior to the 2008-2009 recession, staff were offered continuous professional development of various kinds, including the opportunity to study for advanced professional qualifications, degrees, NVQ Levels 4 and 5, and specialist skills. Following the 2008-2009 recession, natural wastage, redundancies and a recruitment freeze reduced staff numbers and hence the training bill. In addition, increased work intensity among the depleted workforce reduced their availability to attend courses. Expenditure on discretionary training gradually declined until attention was solely focused on essential provisions. At the same time, high and rising training floors consumed resources. The 'nice to have' melted away. In an attempt to reduce costs and increase flexibility, the company turned to e-learning to replace much classroom provision. As discretionary training was squeezed, training delivery was centralised and training contents standardised. The company went from a high training ceiling, high training floor organisation to one where ceilings had been trimmed.
Although high levels of discretionary training are relatively vulnerable during economic downturns, it should not be assumed that cuts are inevitable. For example, some of our respondents with high training floors and ceilings elected to become stickers or even boosters. In our research, these included organisations offering legal, financial, medical, scientific and educational services. The Legal Services Company is one example of an organisation maintaining training activity despite relatively high levels of discretionary training.
The Legal Services Company had high training floors as well as ceilings, both of which were sustained throughout the 2008-2009 recession and its aftermath. For legal staff, training floors (determined by legal, statutory, regulatory, professional and operational sources) were high. Mandatory training, involving formal courses and informal mentoring, was continuous, wide ranging and could lead to external qualifications. Discretionary training was equally extensive. The social, presentational and networking skills of employees were regarded as a business asset, reflected in the provision of personal development training. Advanced courses in fields indirectly related to job performance were funded by the company. For support staff, training floors and ceilings were lower but not neglected. Their mandatory and discretionary training was of high quality. The recession of 2008-2009 posed difficult challenges to the firm. There were a large number of redundancies and a recruitment freeze, leading to a fall in numbers requiring training. Nevertheless, ways were found to maintain training activity. More creative and cost efficient channels of delivery were devised. The lull in business was taken as an opportunity to up-skill remaining staff. As a result, the quality, range and effectiveness of training were maintained.
Other organisations in our interview sample were characterised by low training floors accompanied by high training ceilings, which they chose to preserve despite financial pressures. Hospitality Company is a case in point. Typically, organisations of this kind have a culture that values and nourishes workplace learning. Training opportunities may be perceived by management and workers to be part of the bargain struck between effort and rewards. Engaged and committed employees are regarded as a key plank of high performance managerial strategies. High-quality training may be recognised as a vital contributor to productivity and, therefore, worthy of protection even when staff numbers fall due to redundancies. In public sector organisations, the provision of high-quality training may be embedded within an ethos that celebrates a duty of care to employees and clients (John and Johnson 2008;Le Grand 2010;Jewson, Felstead, and Green 2014). Some private sector companies viewed discretionary training as crucial to integrating ethnically diverse labour forces and sustaining a shared organisational culture.
Despite economic pressures, Hospitality Company did not waver in sustaining the training programme in its global chain of up-market hotels throughout the 2008-2009 recession. Our interviews focused on a cluster of UK establishments, employing several hundred staff. Jobs included kitchen work, housekeeping, waiting tables, reception and sales. With few exceptions, mandatory and operational training floors were relatively low. However, the company provided an elaborate, formalised, high-quality programme of training for all employees, depending on their grade. The lower rungs of this training ladder focused on high standards in basic skills. Supervisors and managers attended in-house courses on HR issues, time management, leadership, 'wow' factors, personal effectiveness and company culture. More senior staff were eligible for intense motivational courses in exotic locations. Participation in training was tracked, assessed and rewarded from head office. Although some local input was permitted, materials and timetables emanated from company HQ and were strictly standardised. Although the 2008-2009 economic downturn reduced customer demand and led to a small number of redundancies, training continued unabated. Senior executives believed that the discretionary training system fostered high performance standards while preserving a uniform company culture and managerial strategy throughout the estate.
It should not be assumed, on the other hand, that organisations which spend a great deal on meeting the demands of high training floors will also invest yet more resources on sustaining high training ceilings. The considerable costs in time and money of servicing 'must have' training floors may make it difficult to justify yet further expenditure on the 'nice to have'. Employers may take the view that by bringing staff up to the standards demanded by high training floors they are, in any case, providing them with well paid, prestigious and intellectu-ally interesting careers. On economic grounds, therefore, they may choose to restrict training to 'must have' activities, thereby minimising that which is discretionary. In these circumstances, the scope for taking the cutter trajectory is circumscribed. Even though many of these organisations are very different from routine manufacturing operationsexhibiting the low floor/low ceiling scenario discussed abovethey are subject to analogous constraints. In both cases, minimal difference between training floors and ceilings prohibits cuts to training. In our research interviews, organisations of this kind included some legal and financial services enterprises. Financial Services Organisation 2 is an example of those facing 'must have' training demands that are substantial, unavoidable and continually changing, while at the same time minimising commitment to discretionary training.
Financial Services Company 2 provided financial advice and services to a niche market. Most recruits joined with high levels of educational qualifications but without job relevant experience. The training provided by the company for newcomers was intense and took several years. Compliance training was very extensive, detailed and audited by outside regulators. Continuous change made it essential for the company to provide employees with on-going up-dating training. Non-discretionary training floors were high, rising, costly and demanding. Prior to the recession a small amount of additional discretionary training was available but was always strictly limited. The difference between the training floor and the ceiling, then, was low. The recession had a major impact on the company, leading to staff reductions and a recruitment freeze. Training spend was henceforth strictly focused on the 'must have'. Anything other than free discretionary training was cut back to zero. Expenditure on compliance training for key operational staff squeezed out other training for support staff. The already low scope for discretionary training fell still further as training floors continued to rise.
A small number of organisations in our study were boosters, responding to the recession by increasing their investment in training and thereby raising their training ceilings. Many of these had been badly affected by the downturn but had responded to its impact with enhanced training strategies. In our admittedly small sample, these tended to be enterprises characterised by low, or at best intermediate, training ceilings and similarly low training floors, rather than those with already high training ceilings. Some were driven by pressures from their customers and those who retailed their products, such as supermarkets; for example, Food Processing Company 2. Others were facing disaster and had developed survival strategies that entailed innovative training programmes; for example, the Heavy Industry Company.
Food Processing Company 2 is a large organisation located in multiple sites across the UK. Our interviews focused on one of its biggest plants. The company entered the recession as a prime example of an organisation with both a low training floor and a low training ceiling. Most staff were employed as production line operatives. Engineering, maintenance, management and administrative support made up the rest. Prior to the 2008-2009 recession, training for all staff was limited and underdeveloped. However, during the downturn the quality, coverage and delivery of training floors was greatly improved. Initially, health and safety, food safety, manual handling and related courses were enhanced, as well as training in operational procedures. Later an ambitious programme of discretionary leadership and managerial training was rolled out for team leaders, supervisors and managers. This promoted problem solving, reflexive learning, empowered decision-making, leadership, open communications and striving for excellence. Innovations followed in the provision of NVQs, mentoring and apprenticeships. Expenditure on training increased dramatically. Training ceilings soared as discretionary training ballooned, while training floors continued to be met. These developments were driven by new customers, acquired as a result of the recession. The recession increased sales of the type of low-priced processed foods produced by the plant. Supermarkets and other large retailers wanted these items but also demanded, and audited, high standards in a range of aspects of the business, including training. Hence, the 2008-2009 recession created new market opportunities and new training needs for the company, turning it into a booster.
The Heavy Industry Company is a medium-sized enterprise, located in Victorian premises in an urban area. Prior to the recession, most training was focused on traditional intermediate training floors around health and safety and plant operations. Provision of additional discretionary training was very limited. During the 2008-2009 downturn business plummeted, a third of the workforce was made redundant and closure was only narrowly avoided. However, a revival strategy was devised, generated by innovative senior managers, which targeted the production of high-quality products. It entailed up-skilling the workforce, investing in machinery and fostering new working methods. As a result, by 2012 the company was enjoying strong sales in highly competitive European and Asian markets. The company was saved partly by technological upgrading but also by new ways of working and managing that were more reflexive, efficient and flexible. The recession created down time, enabling the company to expand a variety of training opportunities and to raise training floors. Key personnel were sent on courses in lean manufacturing, high performance working and technical quality. A programme of NVQs was rolled out for all workers. Some employees attended seminars, conferences and short courses, networking with peers and leading industry figures. Staff morale lifted, worker engagement grew, technical skills increased and potential customers were impressed. At the end of our research period, discretionary training had grown relative to 'must have' activities as the training ceiling rose while the training floor remained unchanged.
One of the benefits of our conceptual framework is that it enables us to track the direction of travel of organisations that cope with recession by adopting a sequence of contrasting responses. For some, the experience of recession and its aftermath generated training trajectories that were varied and subject to countervailing pressures. They became cutters, boosters and stickers in turn, depending upon circumstances. Our concepts facilitate mapping the twists and turns of mixed training strategies that take organisations such as the Furniture Manufacturing Company on switch back rides through different training configurations.
The Furniture Manufacturing Company employed shop floor operatives whose intermediate training floors included mastery of specialist technical equipment. At the beginning of the downturn their training was increased in order to hoard specialist labour and keep a semi-idle workforce occupied until business picked up. Refresher and advanced skills training raised training ceilings and therefore widened the gap between the floor and ceiling. As time went on, however, declining markets and financial pressures lead to lay-offs, redundancies and a squeeze on all budgets. Training for both shop floor operatives and managers was reduced close to the unavoidable non-discretionary essentials. At this point, the training ceiling fell. By the beginning of 2012, yet another picture was emerging. The fortunes of the company were modestly reviving and with them a new perspective on training was emerging. Skills shortages in the local labour market and an ageing workforce encouraged management to adopt a longer term and more strategic approach to both recruitment and training. Investment in training was driven up by succession planning, thereby raising training ceiling and opening the gap between the floor and the ceiling once again. Over our research period, the company was booster, cutter and then booster again.
The length and intensity of a recession may be a key factor here, with different trajectories emerging at different phases in the economic cycle (Brunello 2009). In the early stages of brief and shallow recessions, employers may opt to hoard valuable skilled labour. Slack time may be accommodated by the introduction of additional training, thereby increasing discretionary training. However, if the downturn is deep and long, it may become imperative to make cost savings. As a result, discretionary training is cut back and training overall contracts. When, finally, economic growth picks up again, there may be yet another phase of bounce back in training ceilings.

Conclusion
This article seeks to outline, and illustrate the utility of, a new conceptual framework for understanding patterns and processes of change in training during periods of economic downturn. It argues that a distinction between the metonyms of training ceilings and training floors makes it possible to map a range of different training configurations and track multiple training trajectories. Equally, the principles of the framework suggested here could be applied to the wider concept of workplace learning goes beyond formal training activities to include doing, watching, listening as well as seeking the advice of others.
The newly developed metonym of training ceilings directs attention to changes in the totality of training offered by an organisation. Training ceilings along with training floorsa previously developed metonymhighlight boundaries and fault lines within the total training packaged provided by an employer. Together, these metonyms allow us to move beyond simple measures of increase, decrease and stability in training. They make it possible to specify pathways of change shaped by shifts in both the total training offered by an organisation and the balance between different types of training. They enable us to specify points of departure, travel and arrival in training trajectories. Understanding the many pathways taken by cutters, booster and stickers enables us to recognise some of the hidden or latent implications of change and to describe complex processes of development. It is striking that national quantitative data-sets indicate that up to three-quarters of employers were stickers during the 2008-2009 recession Jewson 2011, 2012). This might be taken as evidence that training has earned a recognised and defended place in the British economy. However, our analysis suggests that caution should be exercised in interpreting these figures. What proportion of stickers were characterised by low levels of discretionary training? To what extent is the maintenance of training expenditure during a downturn a measure of the low priority accorded to anything other than non-discretionary training by many UK enterprises? These remain unanswered questions, but the conceptual framework proposed here provides the intellectual tools to address them more thoroughly in future research.
We have seen that organisations with low training levels of discretionary training have little scope to respond to economic hardship by cutting training and are constrained to adopt a conservative strategy. They have little option but to become stickers, or even boosters. Not surprisingly, this is true of organisations which spend little on training overall but is also true of some with big training budgets devoted to high training floors. Our conceptualisation raises the possibility that marginal downward adjustments to high training ceilings may not be indicative of a lack of belief in the value of training nor a source of significant damage to training provisions. In short, cutters may not be quitters. Similarly, the analysis warns against complacency in situations where large numbers of employers report that their overall training levels have changed little. Stickers include those with different ratios of discretionary and non-discretionary provisions. So, it is quite possible for an unchanging training ceiling to be accompanied by a rising training floor with the result that discretionary training provision is squeezed. The concepts and metonyms outlined here prompt us to ask such questions and therefore provide future insights into the nature of training activity in the UK. They also highlight the need to better delineate the nature of training provided by employers into its non-discretionary and discretionary components as well as the role the government and other institutions can play in raising training floors, if not training ceilings.