HIGH PERFORMANCE WORK SYSTEMS AND COMPANY PERFORMANCE

IN SMALL FIRMS- AN EMPIRICAL ANALYSIS

ABSTRACT

The study of the impact of Human Resource Management (HRM) on small firm performance is a neglected area of research in the literature. This paper seeks to help to fill this gap by examining the impact of High Performance Work Systems (HPWSs) on a number of indices of company performance. The empirical study reported, carried out on a sample of 96 companies with between 10 and 100 employees, shows that the implementation of HPWSs is positively related to company profitability, but not to productivity and labour turnover. These findings make clear the importance of size of firm in the analysis of the impact of HPWSs on company performance.

Key Words – human resource management, high performance work systems, small firm, company performance

  1. Introduction

In the last two decades there has been a significant increase in interest in HRM due principally to the large body of empirical evidence which has displayed the positive impact of the effective management of human resources on company performance ( Schuler and Jackson,2005). The majority of the empirical research has sought to study the effect on company performance of the so-called high commitment or high performance work practices (Datta et al., 2005). These consist of a group of consistently applied human resource practices designed to achieve greater employee commitment to the work process as well as higher levels of productivity, as a result of which a company´s labour force can become a source of sustainable competitive advantage (Pfeffer, 1995).

In the literature it is difficult to find a precise definition of HPWSs(Datta et al.,2005). Nonetheless most empirical and theoretical work in the area usually include as high performance practices comprehensive recruitment policies, careful selection, formal performance evaluation, group incentives, multi-skilling, the incorporation of flexibility in job design and focused communication practices with employees (Datta et al., 2005; Guthrie, 2001; Huselid, 1995;Pfeffer, 1995).

There is a considerable body of research on the impact of HPWSs providing substantial evidence of their positive effect on company performance. In particular they have been shown to be a powerful tool for improving labour force retention as well as for increasing productivity and profitability (Combs et al., 2006). However, with the odd exception (e.g. Sels et al., 2006; Way,2002), most research has focused on the impact of HPWSs in large or medium-sized companies, almost completely neglecting small firms (the definition of small firm will be dealt with in section 3).

This gap is a reflection of the limited interest displayed by researchers in the management of human resources in small firms (Heneman et al., 2000). The little research that exists tends to be descriptive, largely based on case studies. In general these suggest the dominance in small firms of informal practices in the management of personnel. (Bacon and Hoque,2005; Heneman et al.,2000).

The lack of research on the role and impact of HPWSs in small firms is of some significance at both a theoretical and practical level.

At a theoretical level it prevents the wider application of findings on the impact of HPWSs generated in larger firms. So, by examining the impact of practices in small firms, one is helping to define the heuristic limits of the theoretical claims which have been developed to date about the role and impact of HPWSs. At the same time such research makes a contribution to the continuing debate about how different small firms are and whether the management principles tried and tested in large companies can be applied to small firms (Torrés y Julien,2005).

At a practical level , the management of human resources in small firms is a key issue, representing as it does, one of the main concerns of small firm managers (Flash Barometer,2007).In addition, the level of investment needed can be a strong argument against introducing HPWSs in small firms , if their benefits for company performance are not clear. As a consequence research which establishes if these practices can help improve small firm performance could be of considerable use at a practical level, particularly bearing in mind that, in most western economies, small firms not only constitute the largest category of company numerically but also make the largest contribution to GNP and employ a high proportion of the labour force (Faems et al., 2005). For example, 99%of companies in the European Union (EU) are small and medium sized enterprises (SMEs), and they employ two thirds of the labour force in the private sector (Eurostat, 2006).

Accordingly our objective in this study is to examine the impact of HPWSs on different measures of small firm performance, in the specific context of the United Kingdom. The study is presented in three parts. First there is a consideration in more detail of relevant literature and the development of associated hypotheses, secondly there is a presentation of the empirical analysis and finally there is a discussion of the results obtained and indication of the most important conclusions.

  1. Theoretical Framework and Hypotheses

Currently available empirical evidence would indicate that the introduction of HPWSs could be particularly beneficial for small firms given the type of HR problems many seem to experience. For example in a recent EU survey (Flash Barometer, 2007) 28% of small firms indicated that they face problems in recruiting and retaining appropriately qualified staff. Similar results can be found in North American research during the last decade ( Deshpande and Golhar,1994; Hornsby and Kurtatko,1990). Given that research has shown that HPWSs can be an important tool for improving labour turnover rates, small firms could benefit considerably from their introduction (Way, 2002). Similarly evidence suggests that there is a high failure rate among small firms and that they have less than satisfactory levels of productivity. (Faens et al., 2005; Mole et al., 2004). HPWSs could help here as well as research has shown a strong relation between high performance practices and productivity and company profitability (Combs et al., 2006).

As previously indicated there has been little research on the role and impact of HPWSs in small firms. The two studies published to date have been located in the United States (Way, 2002) and Belgium (Sels et al., 2006). In a US context HPWSs appear to help to reduce labour turnover but have no impact on productivity (Way,2002) while the Belgian study, on the contrary, identified a positive relation between HPWSs and productivity but no impact on labour turnover (Sels et al.,2006).

In the interests of adding to the limited research in the area and facilitating international comparison, this study addresses the issue in the UK, one of the European countries in which HPWSs have become most established (Gooderham et al., 1999). The results of our analysis can also be compared with those obtained from studies of the impact of HPWSs in larger firms in the UK (e.g. Guest et al., 2003).

Although it would be wrong to consider the UK economy as one dominated by small firms (particularly when compared with Southern Europe), their role is nonetheless important. In practice, firms with less than 250 workers make up 94% of the total in the UK and they employ47% of the private sector labour force. In addition they are responsible for 47% of private sector turnover (Small Business Service, 2005)

Most studies of HPWSPs have measured their impact on company performance. For example Dyer and Reeves (1995) have identified three types of variable likely to be affected by HPWSs; human resource variables such as labour turnover, organizational variables such as productivity and, finally, financial variables such as R.O.E . Most research up to now has focused on some or all of these variables (Combs et al., 2006) and as a result they are the ones used in this study.

We go on to develop the hypotheses of the study:

2.1HPWPs and Labour Turnover

Labour turnover can be a problem for any company, irrespective of size. It involves both direct costs (e.g. recruiting new staff and training them) and indirect costs (e.g. the loss of clients) (Dess and Shaw, 2001).However evidence suggests that labour turnover can be particularly damaging for small firms, in part because of their limited financial resources (Sels et al., 2006) but also because of the difficulties they face in attracting appropriately qualified labour.

According to their advocates, HPWSs can be instrumental in creating a company environment which increases worker satisfaction, motivation and commitment (Appelbaum et al., 2000; Walton, 1985) resulting in a stronger ,positive culture and therefore a lower level of labour turnover (Guthrie,2001; Huselid, 1995). It is true that this particular impact of HPWSs has never been demonstrated empirically (Ramsay et al., 2000). However most studies of larger firms tend to suggest a negative relationship between HPWSs and labour turnover (e.g. Guest et al., 2003; Guthire, 2001; Huselid,1995).?????

In the light of this evidence it would seem relevant to examine whether HPWSs also have a positive effect on labour turnover in small firms. Thus:

Hypothesis I . In small firms there is a negative relationship between the implementation of HPWSs and labour turnover..

2.2 HPWSs and Productivity

Although small firms are responsible for a considerable proportion of the GDP in most European

countries (Eurostat,2006) , in general, their level of productivityis considered unsatisfactory, particularly in comparison with larger companies (Mole et al., 2004).There are a number of reasons for low productivity in small firms ranging from low levels of capital investment to their difficulties in achieving economies of scale (Idson and Oi, 1999). As a result of this problem and recognizing the importance of small firms for their national economies, different governments in the EU have introduced programmes to help firms improve their levels of productivity indirectamente. The majority of these programmes have had fiscal implications. For example, in the UK, the main approach has been to seek to reduce the tax burden on small firms (Storey, 1997). To date, the measures adopted have not always been effective and the productivity problem of small firms continues to be area meriting further research (Mole et al., 2004)

HPWSS can play an important role in improving productivity. Most studies have found a largely positive relationship between HPWSs and productivity, irrespective of sector (services v manufacturing), of the productivity measure used (objective v subjective), of the unit of analysis (workplace v company) and of the high performance practices included (Datta et al., 2005; Guthrie, 2001; Huselid, 1995).

The existence of this relationship is normally explained by the capacity of HPWSs to attract, motivate and develop a labour force capable of achieving higher levels of performance (Becker and Huselid,1998).

It therefore would seem logical to expect that HPWSs could play a useful role in helping small firms to improve their productivity. Thus:

Hypothesis 2. In small firms there is a positive relationship between the implementation of HPWSs and labour force productivity.

2.3. HPWSs and Profitability.

Research on medium sized and large companies has established a positive relationship between the implementation of HPWSs and profitability, irrespective of whether the latter is measured by accounting or market indices.???? (Combs et al., 2006)

HPWSs can increase profitability is different ways. Appelbaum et al. (2000) suggest that there are three principal avenues via which profitability can be increased. The first - and most obvious – is through an increase in productivity. The second is through cost reduction. HPWSs help to reduce the number of employees, particularly of supervisory and support staff, because a better-trained and more involved workforce will also be more able to work independently and flexibly. The reduction in the number of employees in turn reduces labour costs. In addition a more efficient labour force will reduce raw material costs and the need to carry stock which, in turn, will reduce the size of locale needed for production and therefore rental and insurance costs. The third way of increasing profitability is through increasing employee efficiency and so reducing costly interruptions to the production process, enabling delivery schedules to be met.

Bearing in mind the importance of small firms for most western economies, it is surprising there is so little research examining the impact of HPWSs on small firm profitability. This is particularly so given that one of the main difficulties which small firms face is that of raising finance (Flash Barometer,2007) which increasing their profitability would facilitate (Sels et al., 2006). It would therefore seem particularly important to understand the impact of HPWSs on small firm profitability and find out whether the implementation of HPWSPs can increase the profitability of small firms, as in larger companies. Thus:

Hypothesis 3. In small firms there is a positive relationship between the implementation of HPWSs and company profitability

  1. Methodology of the Study.

3.1The Population studied

Usually, in both academic and policy oriented literature, the number of employees is the main criteria used to distinguish firms by size (Way, 2002).It is certainly the case that some studies use criteria based on market share, management structure or ownership but an important advantage of using number of employees is that it enables comprisons to be made between sectors. However there does not exist a consensus on the labour force size of a small firm. Institutions such as the EU define small firms as those with more than 10 and less than 50 employees (firms with less than 10 employees are categorized as micro-firms)(Eurostat,2006) while in some academic studies firms with no more than 500 employees have been defined as small (Golhar and Deshpande,1997; Zheng et al.,2006).

In previous studies on HPWSs in small firms, some researchers have focused on companies with between 20 and 100 employees (e.g. Way, 2002) while others have concentrated on those with between 10 and 100 employees (e.g. Sels et al., 2006). This paper adopts the latter approach, focusing on companies with more than 10 and less than 100 employees. This is the group of firms which has received least attention in studies of the impact of HPWSs which have tended to concentrate on firms with more than 100 employees (Faems et al., 2005).?????

3.2. Sample and Data Collection

This study uses data from the Workplace Employment Relations Survey, 2004 (WERS, 2004), a government-funded national survey whose objective is to provide representative data about a wide range of employment practices in all sectors of the UK economy. As previously pointed out, most research on human resource management in small firms has been descriptive based on the case study method. Studies of this type can certainly help to reflect the complexity of the management process in small firms but it is impossible to generalize from their results. The use of data from WERS 2004 enables this problem to be overcome.

The unit of analysis employed in the WERS survey is the workplace, defined as the activities of an employer in one location e.g. the branch of a bank, a factory, a head office. The WERS Cross Section Management Questionnaire is completed through face to face interviews with the manager responsible for human resources at the workplace. The interviews took place between February 2004 and April 2005. A separate WERS questionnaire made available data on the financial results of the workplaces surveyed for 2003. Via these two questionnaires WERS provided the data necessary to test our hypotheses.

The data of the Cross Section Management Survey enabled us to identify 292 private sector firms with more than 10 and less than 100 employees, excluding workplaces which were part of a larger organization ( and which could benefit from the economies of scale associated with the implementation of HPWSs (Ferris et al., 1998)).??? However not all of the 292 firms identified replied to the questionnaire on financial performance, reflecting the general difficulties of obtaining data from small firms (Heneman et al., 2000; Wijewardena and Tibbits, 1999). As a result we could only use 96 companies, 32.9% of the initial sample of small firms and the study had a sampling error of 10%. Response rate

3.3. High Performance Work Systems

Previous researchers on HPWSs have recommended using a unitary index of high performance practices (Becker and Huselid,1998) and, given the lack of agreement about which practices to include (Becker and Gerhart,1996), have argued that one should be guided by the approach taken in previous studies (Becker and Huselid,1998). Following this advice we constructed a unitary index of 9 practices based on taking an average of those used in the literature. These practices are: careful selection (Datta et al., 2005; Huselid, 1995; Sels et al., 2006; Way, 2002), formal performance appraisal (Datta et al., 2005; Huselid, 1995), performance related pay (Huselid, 1995; Sels et al., 2006), group incentives (Datta et al., 2005; Guthrie, 2001; Huselid, 1995; Sels et al., 2006; Way, 2002), multi-skilling (Datta et al., 2005; Guthrie, 2001; Pfeffer, 1995), job rotation(Datta et al., 2005; Guthrie, 2001; Pfeffer, 1995; Way, 2002), quality circles (Datta et al., 2005; Guthrie, 2001; Huselid, 1995; Way, 2002), team working (Datta et al., 2005; Guthrie, 2001; Pfeffer, 1995; Way, 2002) and disclosure of information (Datta et al., 2005; Guthrie, 2001; Huselid, 1995; Pfeffer, 1995) (α = 0,63). The number of practices used is taken to determine the degree to which a HPWS has been implemented in each firm. The values of the variable HPWS go from 0-when none of the practices have been used- to 1 – when all 9 practices have been implemented in relation to all staff in he company (see Table 1 for the details of each practice).

3.4 Dependent variables

The dependent variables used in the study are: labour turnover, productivity and profitability (Table 1):

Labour Turnover

The study focuses on voluntary turnover where employees leave a company of their free will (and therefore where the company loses comeptent workers who could still be useful to it). This variable has been measured in terms of the percentage of workers who have left the company voluntarily during the previous year (Sels et al., 2006; Way, 2002)

Productivity

The productivity of a company reflects its ability to produce goods and services given a specific amount of labour, capital, materials, time, space, and knowledge or any combination of these factors.(Lindsay,2004). As in the case of previous studies (e.g. Datta et al., 2005;Guthrie, 2001; Huselid, 1995) this variable has been measured as the logarithm of the ratio of firm sales to number of employees.

Profitability

The ratio used for profitability is that of gross profits to fixed assets. This ratio compares profits with capital employed???

3.5 Control Variables

Five control variables were included in the analyses: capital intensity, trade union presence, age of firm, firm size, and sector (see Table 1).

Capital intensity. This was introduced because the degree of capital intensity can have an important influence on productivity (Datt et., 2005).This variableis measured as the logarithm of the ratio of fixed assets to number of employees (Huselid,1995)

Trade union presence. Union presence has been included as a control variable because trade unions can affect our dependent variables (Freeman and Medoff,1984).This variable has been measured via the creation of dummy codes representing two categories. The value zero is for those firms where there are no union members or, if there are members, the union is not recognized for collective bargaining. The value 1 is for those firms where a union is recognized for collective bargaining. This form of measurement assumes that only in those cases where unions are recognized for negotiating purposes are they likely to be able to exercise influence over company policy (Kelly, 1996).