Sinclair et al 'Unfair Dismissal, Representation and Compensation'

BAILII is celebrating 24 years of free online access to the law! Would you consider making a contribution?
No donation is too small. If every visitor before 31 December gives just £1, it will have a significant impact on BAILII's ability to continue providing free access to the law.
Thank you very much for your support!

	[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]
	United Kingdom Journals
You are here: BAILII >> Databases >> United Kingdom Journals >> Sinclair et al 'Unfair Dismissal, Representation and Compensation' URL: http://www.bailii.org/uk/other/journals/WebJCLI/2000/issue5/sinclair5.html Cite as: Sinclair et al 'Unfair Dismissal, Representation and Compensation'

Unfair Dismissal, Representation and Compensation

Adele Sinclair, Neil Botten and Sandra Cahill*

Principal Lecturers and Research Assistant at Portsmouth University Business School

Copyright © 2000 Adele Sinclair, Neil Botten and Sandra Cahill
First Published in Web Journal of Current Legal Issues in association with Blackstone Press Ltd.

* The authors wish to thank Peter Charlton and staff and students from the Business School for their help with this research.

Summary

This article analyses data on unfair dismissal cases at employment tribunals, including primary statistics collected by the authors. It looks at the effect of legal representation on both the likelihood of success for the parties involved and on the size of compensation payments made to applicants. It concludes that legal representation appears to be a significant determinant of both success and compensation.

Introduction

On 18 May 2000 the Lord Chancellor announced that there would be a ' wide-ranging independent review of tribunals in England and Wales'. The terms of reference include consideration as to whether 'there are adequate arrangements for improving people's knowledge and understanding of their rights and responsibilities in relation to ... disputes, and that tribunals and other bodies function in a way which makes those rights and responsibilities a reality,' (Lord Chancellor's Department 2000). The review will include Employment Tribunals.⁽¹⁾

Employment Tribunals were established to allow lay people access to employment dispute resolution without the need to follow strict procedures or to become immersed in legal technicalities. They were intended to be "user friendly," where the chairmen and lay members should apply as much commonsense as law to the issues before them, taking the approach of an industrial jury. The Franks Committee 1957 sought to create tribunals which would have the virtues of 'cheapness, accessibility, freedom from technicality, expedition and expert knowledge' (Franks Committee 1957 para 38). Similarly the Donovan Commission approved facilities for the settlement of employment disputes which were' accessible, speedy, informal and inexpensive' (Donovan Report1968 paras 568 and 578).

The growth in work undertaken by employment tribunals both in terms of jurisdiction⁽²⁾and the numbers of cases heard each year is substantial. The decisions in some of these cases, notably involving discrimination or transfer of undertakings, clearly present complex legal issues, whereas others, such as claims for redundancy pay or unfair dismissal, appear to be relatively free of 'legal' debate, although it could be argued that the legal rules still play a significant subliminal role.⁽³⁾ In 1998/9 there were around 92,000 claims across all jurisdictions (Labour Market Trends 1999). Not surprisingly tribunals have been experiencing difficulties in coping with the workload.

In order to address this problem, the Employment Rights (Dispute Resolution) Act 1998 provides for an alternative method of dispute resolution in unfair dismissal cases. Under sections 7 and 8 of the Act, ACAS has produced proposals for an arbitration scheme which will operate as an alternative to employment tribunal hearings, where parties consent (ACAS 1998). It is intended that disputes should be resolved 'speedily, informally and in private'. The scheme is being designed for those cases which do not involve complex points of law but parties may still be legally represented (ACAS 1998, 9).⁽⁴⁾

A party to an employment dispute may choose their representative (s.6(1) of the Employment Tribunals Act 1996). Although tribunals may control the way a representative conducts a case they may not refuse to allow any person to act as a representative (Bache v Essex County Council [2000] IRLR 251).⁽⁵⁾ The Human Rights Act 1998 requires, by the incorporation into UK law of Article 6 of the European Convention on Human Rights, that every person should have the right to a fair trial and this may include representation in appropriate cases.⁽⁶⁾ It is therefore of considerable interest whether a person who is not legally represented in an employment tribunal or in the new arbitration scheme may be seen to be disadvantaged without legal representation.

The focus of this article is on legal representation in unfair dismissal cases. The authors review statistical findings concerning the value of legal representation on a 'win and lose' basis, and, from their own research, a study of employment tribunal decisions undertaken by Portsmouth Business School (PBS) in 1992-1995, attempt to assess whether legal representation contributes not only to the success of a case but also to the amount of compensation awarded to a successful applicant. If legal representation is seen to have a significant impact on the outcome of cases then it could be argued, in the absence of legal aid,⁽⁷⁾ or a satisfactory alternative, that adequate arrangements are not in place for the pursuance of rights, and that tribunal proceedings with an imbalance of legal representation could infringe principles of human rights.

1. Statistical Sources

The authors have drawn on three main sources of statistical information in order to present their own research. The first is the Department of Trade and Industry (DTI)⁽⁸⁾ statistics which cover all cases passing through employment tribunals. This data is published approximately eighteen months in arrears. The DTI statistics involve no sampling, hence avoiding any sampling error, but only have a limited breakdown of the data. These official statistics show how the outcome of a case depends on respondents' and applicants' representation and on the case's jurisdiction. In addition, for unfair dismissal cases only, a histogram of compensation amounts is available, as are median compensation values. The DTI statistics are less detailed than the other sources referred to but, as they include annual national statistics on both outcomes and compensation, they provide useful comparative material.

The other reports referred to are: 'The Effectiveness of Representations at Industrial Tribunals' by Genn and Genn and 'The 1992 Survey of Industrial Tribunal Applications' by Tremlett and Banerji.⁽⁹⁾ These reports are of particular interest as they directly address the issue of representation in some detail for the tribunals surveyed and were undertaken at comparable dates. The Genn report studied four types of tribunal, including the analysis of 1000 employment tribunal cases in 1988. Their information was collected in three ways, by studying tribunal files, by observing cases and from interviews with applicants to tribunals. The report looked not only at representation at tribunals, but also at the source of advice before a hearing. The information of greatest help in this report was an analysis of how the outcome of a case depended on the applicants' and respondents' representation at four different employment tribunals. The findings covered a number of disputes and did not isolate unfair dismissal from them. The Tremlett and Banerji survey is comparable to Genn's analysis, and looked at 2000 cases at employment tribunals between 1990 and 1992. Their survey included telephone and personal interviews with applicants and respondents. Concentration was on unfair dismissal and wages disputes.

The PBS data is drawn from the results of a study of several thousand cases for all jurisdictions at the Bristol and Southampton tribunals, covering the period from 1992-1995. The study represented almost a full census for this period for these tribunals and included many cases which were withdrawn or settled. Data was collected in two formats. All cases were categorised by jurisdiction and result and then some 700 unfair dismissal cases were revisited to record the detailed attributes for each case.⁽¹⁰⁾ These statistics allow analysis of success rates according to representation, location and a number of other factors. This article is concerned with these cases and a smaller group of some 250 cases, where unfair dismissal was found and compensation awarded. As compensation figures were available for this sub-set of cases, together with some twenty attributes for each case, the effect of a number of these attributes on the size of the compensation figure could also be assessed.

2. Success Rates for Unfair Dismissal Cases

Most employees who have worked for an employer for one year or more have a right to present a claim of unfair dismissal to an employment tribunal. This right was introduced by the Industrial Relations Act 1971 and is now contained in Part X of the Employment Rights Act 1996 (ERA). Once an employee has proved that he or she is qualified to claim and has been dismissed,⁽¹¹⁾ the employee must give the reason for dismissal and the tribunal will then decide whether that reason reflects the truth of the situation and whether it falls within those categories of reasons specified in s.98 ERA which are capable of being fair, namely: capability, conduct, redundancy, some statutory restriction or some other substantial reason. The tribunal determines in the light of this finding whether the dismissal was fair both in substance, i.e. the reasonableness of the decision to dismiss, and, procedurally, regarding the manner and procedures surrounding the dismissal.

The first stage followed by the authors after data collection was to compare the results found in their survey at Bristol and Southampton with the national statistics published by the DTI for all employment tribunals for the years of the PBS survey (Employment Gazette 1994 and Labour Market Trends 1997).

In the official tribunal statistics for the years 1993/5, there were a total of 25803 unfair dismissal cases heard at tribunal, of these 10781, or 41.8% were successful for the applicant, i.e. unfair dismissal was found. In the PBS statistics, 41.1% were successful for the applicant. A chi square test for the goodness of fit for the expected and observed figures gave a chi square figure of 0.258, with one degree of freedom, representing a significance of 61%. This indicates that the data from the PBS study is in almost perfect agreement with the national figures. This match is most evident from the chart below.

Chart 1

The Tremlett and Banerji report was also used as a comparator for the PBS data. In their report 41% of unfair dismissal cases were concluded in favour of the applicant,⁽¹²⁾ this agrees with both the DTI statistics and the PBS data. From these good matches it seems reasonable to assume that further conclusions drawn from the PBS data would be valid nationally.

3. The Effect of Representation on the Chances of Success for the Applicant

The PBS cases were broken down according to the four possible representation combinations for the applicant and the respondent. For each combination the proportion of cases producing an 'unfair dismissal' outcome were calculated. The table below shows the results and the 'value' of the lawyer, i.e. the change in percentage success when comparing legal representation with 'other' representation (for these purposes 'other' included self-representation). If the lawyer is valuable for the applicant, his or her chance of success should rise; if the respondent's lawyer is valuable, the chance of the applicant succeeding should fall.

	Applicant's Representation
Respondent's Representation		Lawyer	Non-Lawyer	Value of Lawyer
	Lawyer	58%	41%	17%
	Non-Lawyer	64%	66%	-2%
	Value of Lawyer	6%	26%

'Lawyer' for these purposes covers instructed barristers or solicitors, 'Non-Lawyer' included any other type of representation, such as company representatives, trade union officials, consultants, representatives from Law Centres and Citizens Advice Bureaux, friends and 'litigants in person'. It should be noted however that those who were not recorded as 'barristers' or 'solicitors' on the decision details may in fact be legally qualified or considerably experienced in conducting tribunal cases.

As can be seen from Table 1, the lawyer showed positive value in three of the four combinations. Two of the results were particularly significant. The respondent's lawyer increased the respondent's chance of success by 25% if the applicant had no lawyer and the applicant's lawyer improved his or her chances by 17% if the respondent also had a lawyer. Where the respondent was not represented the applicant's lawyer appears to have had a negative effect, reducing his or her chances by 2%.

Similar analysis was carried out using the Genns' and Tremlett and Banerji's results. Genn and Genn summarised their information about representation in several ways, but did not separate unfair dismissal from other jurisdictions. Representation was broken down into three categories for both applicant and respondent - lawyer, other and self.⁽¹³⁾ For each of the nine representation combinations they show how many cases were dismissed, how many allowed and how many settled. For the purposes of comparison 'other' and 'self representation' have been considered as 'non-legal representation'. Those cases with a 'settled' outcome have been ignored.

Their results consistently showed a positive value for the lawyer ranging from 10%-20%. Tremlett and Banerji's showed the value of a lawyer to be around 5% for the respondent, but surprisingly the applicant's lawyer showed a 5% drop in success.⁽¹⁴⁾ This adverse result from the applicant's perspective was replicated in the PBS figures where the respondent was not represented, but not otherwise.

Looking at all three sources, and taking a crude average of the results, the respondents' lawyer gives an improvement of 15% for the respondent, and the applicants' lawyer a gain of 6.5% for the applicant, but as recorded above these averages conceal quite significant variations.

These 'win and lose' results must also be seen in the context that many of those who appear to have no representation have already received considerable help in the way of legal advice and assistance from lawyers, prior to making the decision to file an application, on the likely consequences of taking such action and how to proceed. ACAS plays a considerable role in promoting settlements and this may also impact on these statistics.⁽¹⁵⁾

4. Compensation in Employment Tribunals

Since legal representation has a marked effect on the success of unfair dismissal cases, the authors wished to discover whether representation also affected the amount of compensation given to a successful applicant. The problem with such research is the broad range of circumstances which occur in unfair dismissal cases which are likely to affect the outcome.⁽¹⁶⁾ Cases were recorded in as much detail as the written decisions and time allowed. In some cases full details were not available as the records kept by employment tribunals are often just a summary of the case.

Compensation for unfair dismissal

The remedies available for unfair dismissal to a successful applicant include reinstatement and re-engagement but these are seldom used in practice.⁽¹⁷⁾ Compensation is assessed under two headings: the basic award and the compensatory award. The basic award is set according to a formula relating to age and length of service and the amount is not within the discretion of the tribunal.

The compensatory award may cover loss of earnings to the date of the hearing, loss of an accrual of statutory rights, loss of pension contributions, and future loss, but not injury to feelings.⁽¹⁸⁾ Although loss of earnings between the dismissal and the tribunal hearing may not be in dispute, the loss of future earnings is more a matter for the tribunal's discretion and this is when tribunals may be susceptible to persuasion from those experienced in presenting arguments concerning compensation. In the past, the compensatory award has been limited by modest statutory caps and the average awards have been limited.⁽¹⁹⁾ However, the general threshold was raised from £12,000 to £50,000 in October 1999 (s.34 Employment Relations Act1999).⁽²⁰⁾ This increase should allow more scope for the assessment of future loss.

The Size of Compensatory Awards

The DTI national statistics (Employment Gazette 1994 and Labour Market Trends 1997) include a summary of compensatory award figures in the form of a table, showing how many cases fall in a certain range, for example, from £400 - £499 or from £7000 - £7999. A sub-set of cases from the PBS study, where compensation was awarded for unfair dismissal, allowed a second comparison to be made with these DTI statistics. The individual compensation figures from the PBS database were broken down into these same ranges, and the percentages falling into each range for both the DTI and the PBS figures can be compared in the chart below.

A chi-square test was carried out to see how the actual numbers of cases falling into each compensation range in the PBS database compared with what would be expected if the national proportions applied, a chi square statistic of 19.71 resulted, with 18 degrees of freedom, representing a significance percentage of 34.9%. As with the last test, this implies that the PBS data fits very well with the national statistics.

5. Legal Representation and the Compensatory Award

PBS's simplest analysis, attempting to predict the effect of representation alone, looked at average compensation figures for different combinations of representation for the applicant and respondent. For simplicity representation was categorised as 'lawyer' or 'others'. The table below shows mean and median values for each combination.

Representation Combination Applicant / Respondent	*Mean Compensation*	*Median Compensation*
Lawyer / Lawyer	£7340	£6162
Lawyer / Other	£4771	£4686
Other / Lawyer	£3664	£2534
Other / Other	£3316	£2243
All Cases	£3789	£2536

The average compensation is highest if both parties use lawyers. This was possibly because the parties to a particular case might have anticipated that high compensation could be awarded and consequently employed lawyers to improve their position. A similar argument could explain why the average compensation was lowest, at £3316, if neither party had a lawyer. If only the applicant were represented, the average compensation rose by £1455 to £4771, seemingly making representation by a lawyer worthwhile for the applicant.

It might have been anticipated that if the respondent alone were represented, then compensation to the applicant would drop. In fact it rose by £348. A possible explanation is that tribunals are sympathetic to unrepresented applicants who have successfully presented their cases against a represented respondent (Macmillan 1999, at 40). Another explanation might be that the skills which the applicant required to win the case were also utilised to present the claim for compensation.

Analysis of the variables which may affect compensation

The difficulty of looking at simple averages to analyse the effect of representation, as above, is that there is no scope for looking at the interaction of variables - for example, a male applicant might gain a greater increase in compensation than a female applicant if both have lawyers. In order to explore this interaction a statistical tool called the Analysis of Variance was used. This analysis attempted to show how applicants' compensation depended upon a number of factors associated with their cases. The factors considered were: the representation types for the applicant and respondent, the sex of the applicant, the location of the case, the reason discovered by the tribunal for the dismissal, the contributory fault of the employee and the amount of the basic award. In the event contributory conduct proved to be an insignificant factor in the estimation of the total award and was ignored.⁽²¹⁾

Two difficulties stood in the way of including the basic award in the analysis. First, the value of the basic award was only available for a proportion of the cases, since the amount was omitted from some case details,⁽²²⁾ secondly, the maximum amount of the basic award is relatively low.⁽²³⁾ However, since the amount of basic award reflects age, length of service and to some extent the salary of each applicant, it seems likely that it would have an effect on total compensation. When the basic award was added to the other variables no meaningful model could be found using either the additive or multiplicative concepts, discussed below. The only conclusion coming from this process was that the basic award was indeed a significant factor in the calculation of total compensation, but that it could not be analysed in conjunction with other variables because of the limited data available for this particular factor.

As a result the basic award was then considered on its own as a predictor of total compensation. A linear regression model showed a poor fit but the best predictor of total compensation was a fixed sum of £2700 added to a fixed proportion (85%) of the basic award. It is not clear why the fit was poor, but it is likely that the low upper limit on the basic award prevented the linear model from accurately predicting any cases with reasonably high total compensation. For this reason, the analysis shown below excluded this factor as a contributory variable in the calculation of total compensation.

The Analysis of Variance

The Analysis of Variance Model (Anova) predicts compensation by calculating a base figure, with values either added or subtracted from the base depending upon the value of each independent variable. For example, an amount might be added if the applicant has a lawyer, but a further amount added or subtracted if the applicant is male. Anova finds the 'best' amount to attribute to each of the variables' possible values so as to most accurately match predicted and actual compensation for all of the cases considered. Such a model is called additive - all the predicted individual factors are added to estimate compensation. By contrast, in a multiplicative Anova model, base compensation is again calculated, but instead of adding values for each variable, the base figure is multiplied by a succession of factors to give overall predicted compensation. In summary with an additive model the effect of an applicant being of male sex might add £2000 to the predicted compensation, with a multiplicative model, the effect could be to multiply compensation by 1.2, or to give a 20% increase in compensation.

There are two important factors required to make the Anova models meaningful. The 'R Squared' figure gives an indication of how accurately a model predicts compensation. The value of R Squared will always fall between zero and one. The closer the value is to one, the better the predictions. (If R squared equals one, all actual and predicted compensations would match exactly.) The second factor is the significance values for each variable. If a particular variable has a low significance value (normally below about 5%), this implies that the calculated effect of that variable is a real one, and most unlikely to have occurred by chance. The objective of the modelling process was to find a combination of variables which gave a high R Squared value while at the same time keeping the significance of each individual variable below around 5%.

It proved impossible with the additive model to find a set of significant variables which gave an R Squared value above 12% so the multiplicative model was then tested. Throughout the modelling process multiplicative models proved to be consistently better predictors than additive models so the conclusions are drawn from this second type of model. The Anova results shown below produced an R Squared statistic of around 17% with all variables having significance values below 6.4%. By no means a perfect predictor, but the best that could be found using the PBS data.⁽²⁴⁾

The model showed that compensation in a given case depended upon the following factors: applicants' and respondents' representation, tribunal location, applicants' sex and the reason discovered by the tribunal for the dismissal. Some factors had significance on their own, whereas others had an effect when combined with a second factor. The top section of the table below shows that all of the factors considered were significant. The base value for compensation chosen by the model before introducing the multiplying factors is £2454. The effect of each variable is described in the text below; note that a factor of 1.4 would raise compensation by 40%, a factor of 0.7 would reduce it by 30%.

ANOVA RESULTS - MULTIPLICATIVE MODEL

VARIABLE	SIGNIFICANCE
Base Value	0.0%
Tribunal Location	6.4%
Applicant's Representation	0.0%
Respondent's representation with Applicant's sex	3.1%
Applicant's sex with Reason Discovered	4.1%
R Squared = 16.8%
BASE VALUE	£2,454.00

MULTIPLIER ESTIMATES:
VARIABLE	MULTIPLIER
Applicant's Representation: Lawyer	2.54
Applicant's Representation: Self	1.00
Applicant's Representation: Other	1.23
Location Bristol	1.40
Location Southampton	1.00
Respondent's representation: Lawyer, Applicant Female	0.74
Respondent's representation: Lawyer, Applicant Male	0.59
Respondent's representation: Self, Applicant Female	0. 47
Respondent's representation: Self, Applicant Male	0.59
Respondent's representation: Other, Applicant Female	0.56
Respondent's representation: Other, Applicant Male	1.00
Applicant Female, Redundant	0.47
Applicant Female, Unaccepted	1.00
Applicant Female, Other reason discovered	1.00
Applicant Male, Redundant	0.87
Applicant Male, Unaccepted	1.00
Applicant Male, Other reason discovered	0.81

As the table above shows, the applicants' representation had a significant effect on compensation. Compared to a base where the applicant represented him or herself, legal representation raised compensation by 150%, and 'other' representation raised it by 23%. In contrast, the respondents' representation alone did not prove significant. However location had a strong effect, with applicants whose cases were heard in Bristol having a predicted compensation 40% higher than those in Southampton. The survey noted that 35% of Bristol cases as compared to 18% of cases at Southampton were heard within 150 days of dismissal; whereas 27% of cases were withdrawn through settlement at Bristol compared with almost 36% at Southampton. The significance of location may therefore be explained by the longer waiting periods at Southampton, during the dates of the study, which allowed time for employers to consider settlement where high compensation might be awarded.

The other factors, besides applicants' representation and location, which emerged from the model arose from a combination of variables. The first combination in the table is that of the applicants' sex and respondents' representation. For males, the respondents' legal representation had no effect on compensation when compared to self-representation by the respondent. However, for female applicants the effect was marked; compared to self-representation by the respondent, legal representation yielded a 56% increase in compensation to the applicant. This result is surprising: it would be expected that if a respondent had a lawyer, then compensation awarded to the applicant would fall, as the respondent is using a professional to make their case and this, if anything, should work against the applicant. A similar argument could apply to the 'other' category of representation. There was further evidence for this reverse effect earlier in the paper when both Tremlett and Banerji and the PBS data showed that the applicant's lawyer could have a negative effect.

The second combination is that of the applicant's sex and the reason discovered for dismissal. The database records the reason for dismissal as one of three categories: 'redundancy', 'unaccepted' or 'other'. 'Redundancy' reflected the statutory definition, 'unaccepted' covered those cases where the tribunal did not accept the employer's reason or found constructive dismissal, and 'other' covered reasons capable of being fair (excluding redundancy) namely misconduct, capacity and some other substantial reason. For five of the six combinations of sex and reason discovered, the factors were close to 1, showing little or no effect. However, if the applicant was female, and she had been made redundant, compensation was more than halved with a 53% reduction in the predicted figure. This was interesting since it showed that, although more women win their cases⁽²⁵⁾ than men, the average compensation was less than for men, clearly reflecting the considerable pay gap between the sexes in the years under consideration.⁽²⁶⁾

Further work needs to be undertaken to address some of the anomalous results. However, overall the effect of a lawyer for both applicant and respondent as far as 'win and lose' is concerned is positive. Regarding compensation, the applicant's lawyer has a very strong positive effect.

6. The Parties' Perceptions of Representation

This article has been concerned with the evaluation of legal representation in terms of result, but it is recognised that the reasons for employing a representative (legally qualified or not) extend beyond the outcomes of a case. An employer does not instigate unfair dismissal proceedings (other than by the act of dismissing) and therefore has either to settle or defend an action if the applicant is determined to proceed. An employer may not have sufficient staff to absorb the extra work that preparation for a tribunal will involve and it may be cost effective to engage an expert rather than a manager who would be better employed with more familiar work. In addition, not everyone is temperamentally suited to the dispassionate approach necessary for the conduct of legal proceedings where the demeanour of the parties may be an important factor in the outcome of the dispute.

A recent ACAS Research paper sheds some light on the perception of parties to unfair dismissal disputes with regard to legal representation (Lewis and Legard 1998). The paper notes the complex pattern of representation which may range from initial advice or partial representation, for example in drafting documents, to representation at the hearing. The advocate may be an 'in house' lawyer, a lawyer employed from the start, or one brought in at a late stage to conduct the case at the hearing (Lewis and Legard 1998, 36-37). Both parties interviewed for ACAS identified the cost of representation and the wish to maintain control of their own cases as factors gravitating against the use of lawyers. However, when asked about the ease with which an unrepresented litigant could conduct a tribunal case the report states that -

The ACAS report found that unrepresented parties whose opponents had legal representation felt particularly disadvantaged both during the course of the case and at the hearing (Lewis and Legard 1998, 40). Views on the performance of representatives predictably appeared closely linked with the outcome of the case (Lewis and Legard 1998, 43).⁽²⁷⁾

Conclusion

Whether the perceptions of the parties to unfair dismissal hearings are sufficient to give an individual the right to legal representation is doubtful but the statistical evidence gives a more concrete base for such a claim. The PBS study shows that not only does legal representation have a significant effect on a 'win or lose' basis but that it may also affect the size of compensation awarded to the applicant. Employers may well be able to meet the lawyers' fees necessary to give them the best chance of winning or minimising compensation (although this is not always the outcome), but dismissed employees, by the very nature of their status, are often unemployed or on low wages and therefore unable to afford legal representation. As EU and, indeed, human rights law make aspects of unfair dismissal claims more complex and the hearing process subject to scrutiny, access to representation will need to be addressed to ensure that the right to be compensated for unfair dismissal can be pursued in a way that dismissed employees' rights become a 'reality' in financial as well as declaratory terms.

Bibliography

ACAS Consultation Paper (1998) 'ACAS Arbitration Scheme for the Resolution of Unfair Dismissal Disputes' (ACAS)

Donovan Report (1968) The Royal Commission on Trade Unions and Employers' Associations 1965-1968 (London HMSO) Cmnd 3623.

Employment Gazette (1994) 'Employment Tribunal and Employment Appeal Tribunal Statistics 1992-93 and 1993-94', October, 368-371, (London: Employment Department).

Franks Committee(1957) Report on the Committee on Administrative Tribunals (London HMSO) Cmnd 218.

Genn, H and Genn, Y (1989) 'The Effectiveness of Representations at Industrial Tribunals', Report to the Lord Chancellor (London: Lord Chancellor's Department)

Haley, C (2000) 'Public funding for representation at tribunals' Legal Action August, 17

Labour Market Trends (1997) 'Employment Tribunal and Employment Appeal Tribunal Statistics 1994-95 and 1995-96', April, 152-156, (London: ONS)

Labour Market Trends (1999) 'Employment Tribunal and Employment Appeal Tribunal Statistics 1997-98 and 1998-99', September, 494-497, (London: ONS)

Lewis, J and Legard, R (1998) 'ACAS individual conciliation: a qualitative evaluation of the service provided in industrial tribunal cases' ACAS Research Paper 1

Macmillan, J (1999) 'Employment Tribunals: Philosophies and Practicalities' Industrial Law Journal 28(1), 33.

Times (2000) 'ACAS plans tribunals in hotels', Christine Buckley, Monday October 9, 2000 (London: Times)

Roe, T (2000) 'Bache, Storer and the Human Rights Act 1998', 10 Employment Law Journal, 21

Tremlett, N and Banerji, N (1994) 'The 1992 Survey of Industrial Tribunal Applications' Employment Department Research Series No.22 (London: ED)

(1) Industrial Tribunals were established by the Industrial Training Act 1964 and were renamed Employment Tribunals by s.1 of the Employment Rights (Dispute Resolution) Act 1998.

(2) For example, race, sex and disability discrimination, working time and wages cases.

(3) See further the issues raised in cases such as the Iceland Frozen Foods v Jones [1982] IRLR 439, Haddon v Van Den Burgh Foods [1999] IRLR 672, Wilson v Ethicon [2000] IRLR 4, and The Post Office v Foley; HSBC Bank plc v Madden, CA, 31.7.00 (discussed in IDS Brief 672,13).

(4) Note that the introduction of this scheme has been delayed due to the pressure of work on the DTI legal services (www.acas.org.uk/10073-version=1.htm). There are also human rights concerns regarding the proposals e.g. the private hearing and lack of appeal. Employers' organisations have expressed concern at the delay of the 'fast track' scheme (Times 2000).

(5) In Bache the Court of Appeal stated that the right under s.6(1)(c) is not overridden by the tribunal's general power under rule 9 to Schedule 1 to the Employment Tribunals (Constitution and Rules of Procedure) Regulations 1993 which permits tribunals to control the way in which a party or his representative conducts a case. Regulation 9 does not give the tribunal the power to dismiss the representative.

(6) The general right to a fair trial in Article 6(1) may include a right to free representation, but this right is thought to be very limited in civil cases, (Airey v Republic of Ireland (1979) EHRR 305). See further Roe (2000, 23) and Haley (2000,17).

(7) There is no legal aid available in employment tribunals although it is available for appeals to the Employment Appeal Tribunal. There are various bodies offering free representation, such as the Free Representation Unit of the Bar and conditional fees are also a possibility. See The Conditional Fee Agreement Regulations 2000 (SI 2000/692) See also, for example, Haley's discussion of s.6(8) of the Access to Justice Act 1999 (Haley 2000) and Roe's discussion of human rights issues in employment tribunals (Roe 2000).

(9) This was the second periodic survey of employment tribunal applications; a third periodic survey was due to be undertaken for 1997/98 but has not yet been published.

(10) Some cases were missing details and others were withdrawn and settled following the finding of unfair dismissal.

(11) Either actually or constructively (where the employer is in breach of a fundamental term of the employment contract).

(16) The age of the applicant, the position they held in the company, the size of the company. the likelihood of finding similarly remunerated work, contributory fault, failure to mitigate.

(17) The gap between the dismissal and the hearing, quite apart from the breakdown of the employment relationship, often makes these remedies impractical.

(18) Unlike in discrimination cases where injury to feelings is a head of compensation. It is recognised that on exceptional facts 'stigma' damages may be available for breach of contract. See Malik v BBCI SA [1997] ICR 606 cf. Johnson v Unisys Ltd [1999] ICR 809.

(20) If the reason for the dismissal is whistleblowing or a health and safety complaint, there is no limit and employment tribunals may award any sums which are 'just and equitable' on the facts of a particular case.

(21) Contributory fault reduces the total amount paid to the applicant. The issue under examination was whether representation lessened the likelihood of contributory fault being found or in some other way had an effect on compensation prior to the percentage deduction.

(22) Presumably this information is missing because it was not in dispute; since an employer will often have already paid the equivalent of the basic award as a redundancy payment to the dismissed employee.

(24) The low value of R-squared is caused because, when fitting a complex model using a number of variable each at several levels, the resulting cross tabulation has small cell sizes even when formed from a large data set. The estimates of the effects therefore have large sampling variances (wide confidence intervals) and hence the R-squared value is small. This does not invalidate the analysis or the conclusions but introduces a necessary note of caution.

(25) The PBS data showed that fewer women pursued cases for unfair dismissal, but that, of those that did, a higher percentage were successful than the male applicants.

(27) Macmillan, in contrast, takes the view that 'No one in an employment tribunal is going to be penalised for not pleading their case properly provided that they make it clear in broad terms what their case is. No one is going to find favour with the tribunal if they try and score procedural and debating points..' (Macmillan 1999, 40).

United Kingdom Journals