The tables show that there were around 132,000 workplaces that reported a dedicated H&S rep. Some of these may be 'distant' representatives therefore we should not assume one representative per workplace. Table 2 showed that 321,000 workplaces were covered by employee reps and Table 4A above showed that 15 per cent or 105,000 workplaces had reps onsite. This implies a ratio of 3.08 representatives per covered workplace. Applying this ratio to 132,000 workplaces gives an estimate of 43,000 dedicated free-standing health and safety representatives active in the WERS population. However, this is not a meaningful estimate of the population of dedicated health and safety reps, as it does not include representatives sitting on a dedicated health and safety committee. Furthermore, WERS only asked about free-standing reps in the absence of a committee structure.

We also want to account for those on a dedicated H&S committee, which are not concerned with matters in the workplace (therefore not captured by estimates in Table 1). WERS data shows that there were 56,000 workplaces that had such a structure. Unfortunately, it is not possible to estimate the number of people on these committees, nor how many of the members are concerned with other issues.

2. Benefits of Workplace Representatives
Estimating the benefits of employee representatives is somewhat less straightforward than estimating the costs. Representatives in the workplace facilitate the indirect, collective participation of employees in various aspects of their work. The 'pluralist' perspective of 'partnership' in industrial relations focuses on indirect forms of employee involvement via representatives. However, there are also direct methods of involving employees, such as team briefings, suggestion schemes, and quality circles (see Bryson, Forth, and Kirby (2005) for discussion). These can also bring benefits to the workplace. A hybrid framework recognises the importance of employee representation as well as direct participation. Through a 'mutual gains' framework, management and employees work together to ensure benefits for all concerned.

Here we do not focus on the distribution of benefits, which (as highlighted in the empirical literature) is not necessarily straightforward, and can be influenced by the nature of industrial relations, for example, productivity gains may lead to higher wages or increased profits.

Potential benefits are organised in the following categories: skills and training; exit rates, labour turnover and dispute resolution; worker safety; and productivity.

Skills and Training (Union Learning Representatives)
The relatively new concept of Union Learning Representatives (ULRs) (given statutory footing in 2002) has meant there is little empirical evidence on their impact. However, they bring benefits by helping individual employees develop their skills and thereby enhancing the contribution they can make at the workplace. The Department for Education and Employment (DfEE) estimated the benefits of ULRs to businesses (in the form of improved productivity) ranging from around £11m to £49m in the first year. This is under the assumption that 28,000-41,000 employees receive training as a result of access to a ULR.¹⁷

As reported above, reliable estimates from the DfES puts the number of ULRs trained since 1998 at around 15,000. This corroborates with data from the WERS survey in which managers in around 14,000 workplaces said that there was a presence of an on-site union rep with training and learning responsibility. Coverage extends to around 11 per cent or 2.4 million employees in these workplaces.

Although not all of these employees will have direct access to the ULR in their workplace, it is worth noting the evidence that ULRs are actively encouraging colleagues in the workplace to undertake training. An evaluation of the Union Learning Fund (ULF) between 2001 and 2005 conducted by DfES¹⁸ found that there had been over 153,000 learning opportunities delivered as a result of the ULF. 31 per cent of learners surveyed stated that they found out about learning opportunities through their ULR.

Based on information on the activities of ULRs, the DFEE estimated that around 12 employees per ULR would complete their training and enhance their productivity as a result of the intervention of the ULR. Applying this assumption to an estimate of around 14,000 active ULRs, this gives a mid-point estimate 168,000 employees directly benefiting from the activities of ULRs in the workplace.

How these productivity gains are accrued (wages, profits) will vary according to the structure of the labour and product market, as well as the climate of industrial relations. However, if we assume a 3-5 per cent premium on the average wage of those who received additional training,¹⁹ this gives annual benefits in 2004 of £94 - £156 million.²⁰

Exit rates, labour turnover and dispute resolution
The 'voice' mechanism that employee representatives can provide has been widely documented in the literature. Freeman & Medoff (1984)²¹ argued that representatives offer an alternative to 'exit' in voicing grievances. Kaufmann and Levine (2000) identify the benefits of employee representatives in so far as market imperfections cause firms to under-invest in dispute resolution, and opt for dismissal instead. However, the existence of negative externalities (involuntary unemployment, loss of skills) imply that this may not be the optimal outcome for society.

By reducing exit rates and labour turnover, employee representation can bring substantial benefits by avoiding costs such as support during job search (unemployment benefit); firm recruitment costs and a reduction in labour productivity (as a result of loss of job skills).

It is difficult to put a number on the extent to which the activities of representatives may lower exit rates and labour turnover. Fernie and Metcalf (1995) using WIRS 1990 data found unions were associated with a lower quit rate. Guest and Peccei (2001) found that direct and representative participation had a positive impact on employee attitudes and behaviour, which in turn improved labour retention and reduced absenteeism.²²

The WERS 2004 survey found the average dismissal rate (in the last 12 months) for workplaces with 5+ employees per 100 employees was 1.66.²³ Using the average, the estimated number of dismissals in the previous 12 months for WERS workplaces is therefore around 374,000.²⁴

The argument is that, in the absence of representative voice, the number of dismissals would be higher. Indeed, although finding no significant association between compliance with the right to be accompanied (in either disciplinary or grievance hearings) and the dismissal rate, Antcliffe and Saundry (2006) using the WERS 2004 data found union density had a significant negative effect on dismissal rates. They note that this suggests the day-to-day representation offered by union representatives could be more important than undertaking the role of the companion on a 'one off'
basis.

According to WERS 2004, the dismissal rate in workplaces with some kind of representative voice was 1.53, compared with 1.96 for workplaces with no representative voice.²⁵ To illustrate potential benefits, if we assume that representation contributed to reducing the dismissal rate in covered workplaces by 5-10 per cent (or, in the absence of reps the dismissal rate in these workplaces would be 1.74-1.83), then the benefits translate to around 13,000- 25,000 fewer dismissals in these workplaces. If we take the average cost of recruitment per vacancy as around £4,200,²⁶ this equates to over £107-£213 million to firms as a result of fewer dismissals.

In line with the Freeman and Medoff argument, representatives can also reduce the voluntary exit rate in a workplace. Voluntary job separation is in some ways symptomatic of a healthy labour market, as workers move around to find better matching jobs. However, in some cases where representative 'voice' can play a part in mediating potential disputes, the avoided costs can be substantial for both the employer and the employee.

_________________________

¹⁷ Regulatory Impact Assessment for Providing Statutory Backing for Union Learning Representatives,
Department for Education and Skills.
¹⁸ http://www.dfes.gov.uk/research/data/uploadfiles/RB789.pdf
¹⁹ Evidence on effects on wage premium are mixed. Booth, Francesconi and Zoega give a wage premium for male workers of 13.3% . A. L. Booth, M. Francesconi and G. Zoega, Unions, Work-Related Training, and Wages: Evidence for British Men. Industrial and Labour Relations Review, 2003. Vol. 57, Issue 1, Article 4.
²⁰ Assumes wage premium gains from training accrue in same year as training (2004). Average gross annual earnings of all employees = £22,266 (Annual Survey of Hours and Earnings, ONS, 2004).
²¹ R. Freeman and J. Medoff, What do Unions Do? 1984. New York: Basic Books.
²² Improvements in employee attitudes and behaviour improved the composite measure of labour retention (5 point score) by approximately 50%; and lowered absenteeism score by around one third.
²³ DTI estimates using WERS 2004 full dataset. Other estimates using this data include forthcoming paper by V. Antcliffe and R. Saundry 'Employee Representation in Grievance and Discipline - Making a Difference?' to be published as part of the DTI WERS 2004 Small Grants Fund. Other estimates of involuntary job separation rate - Heap (ONS Labour Market Trends, 'Job Separations in the UK') using the Labour Force Survey estimates 1.2 per 100 employees in 2004.
²⁴ (Calculation: number of employees in workplaces 5+ employees = 22,500,000/100)*1.66) rounded to nearest 1000.
²⁵ To infer causality, one would need to control for other factors. However, the empirical evidence to date is strong, and therefore the analysis is built on assumption that 'voice' and representation have a  negative impact on dismissal.
²⁶ Source: CIPD surveys.

Next page >