Southampton UCU asked two interesting questions last week in a meeting with senior HR colleagues: What is appraisal moderation for? What are appraisal scores for?

We posed these questions because of growing concerns about how appraisals are being conducted and the use of moderation in ways that did not adhere to our negotiated agreements. We are especially concerned at the use of a bell curve (normal distribution) in moderation of appraisal scores – outside the agreed policy – despite a wealth of evidence that this practice is not only flawed but deeply counterproductive.

The original purpose of the agreed appraisal system

Members may recall that teams from UCU spent over two years in a negotiations process called the Pay and Reward (P&R) Project – looking at a range of policies for managing careers and promotion of our staff. At the heart of this negotiation was the new appraisal system. Both sides – the union and the managers – recognised that the old PPDR (personal performance development review) was broken. We needed a system that enabled us to help staff set objectives and develop, and which could enable managers to support them in ways that in turn helped the University achieve its ambitions. There was willingness from all parties to work hard to get it right.

Key to the roll out of the new appraisal policy [intranet link only] was a programme of staff development helping people to understand what good appraisal looked like. And while not everyone likes ‘active learning scenarios’ one aspect of this training – delivered by actor-facilitators – was powerful at reminding staff what could go wrong in these important conversations. Alongside this core training there was some online training to support the process (though never intended as a substitute for the interactive learning) and a new online appraisal system for storing annual documentation. The latter was not perfect, we all recognised that the software might need to be improved over time in response to feedback from users.

Southampton UCU had clear objectives in the negotiation of the new appraisal policy. We wanted a system that supported better management and development of our staff. We welcomed appraisal as an opportunity for staff to celebrate success and, where necessary identify areas for further support and development. Our vision of appraisal was a conversation – taking place regularly but formally documented annually – that moved away from managers telling staff what to do, towards shared and supportive listening, reflecting and planning. Local employers such as UHS NHS Trust have just such a system – indeed this particular example even has staff wellbeing at the heart of its appraisal process.

UHS Appraisal Process

At the very least, the system was meant to embody a future-focussed principle:

Emphasis on the Future – The majority of the appraisal will be future-focused, modelled on existing ‘Personal Best’ discussions. This will include:
1. Vision and intent
2. Processes rather than outcomes, including development of competencies.
3. Support, resources and training required.

The current distortion and misapplication of the appraisal system

Since its implementation in the 2015 academic year, it has become clear that the University of Southampton appraisal system is not compliant with the agreement we negotiated. The document that was ratified by UCU’s national panel still remains, but its visibility and intent have been obscured by the ratings scheme, added to the document late in the negotiations and only agreed with very specific limitations as to its use, and a moderation process and guidelines that were never part of the negotiated agreement. Nor is the process even being followed according to its own guidelines: rating criteria were meant to be defined locally, but this is inconsistent with the moderation process, and the moderation process lacks transparency and allows the moderators (possibly far removed from the team) to overturn a recommendation from line manager. Crucially, the moderation process breaks confidentiality of the appraisal conversation, and this is against one of the first principles in the agreed policy. The software does not even encourage the completion of the ‘thorough appraisal’ which the policy scopes:

1. Joint review and update of job description
2. Assessment of contribution to the University
i. Review achievement against past objectives
ii. expectations via future objectives
3. Behavioural competencies
4. Career aspirations
5. Development objectives

Alarmingly, training for appraisers has disappeared (apart from the online module), and since there has been no professional development training available for academic staff since 2016, there is no way to act on any development needs identified in appraisal conversations. All that has been achieved by the hundreds of hours spent on the project in its development is the creation of successive layers of documentation that must be processed by dozens of already overworked academics, the practical outcomes of which may not be acted upon constructively.

Our specific question, ‘What is moderation for?’ arises from this chaos: why is it vital that not too many colleagues are seen to be good at what they do? We might answer this positively – because while we would want to celebrate excellence wherever we find it, exceptional work must really be exceptional. When we agreed to the moderation of 4- and 5-rated scores, this was the rationale. But when the moderation process was introduced, it extended across the full range, turning the scores into the focus of an entirely different enterprise.   The moderation process now lacks transparency and allows the moderators (possibly far removed from the team) to overturn a recommendation from line manager.  Appraisal stopped being about celebration and development, and became all about sorting the wheat from the chaff. And because we cannot act constructively on any shortcomings identified in the process, those who are scored or moderated down to the bottom of the scale face only negative outcomes.

Below are some of the concerns raised by members in direct contact with SUCU, some of whom have gone on to be represented by the branch. These are indicative, not exhaustive; some are more serious than others. Some represent potential discriminatory behaviour; some represent behaviour likely to instigate a grievance; some are simply breaches of the principles agreed in the policy. Some of the concerns have been raised with us simultaneously with a request for representation in a ‘protected conversation’, after a settlement agreement has been presented to the member. These issues were presented to HR via a JNC paper over six weeks ago, and we have not yet had a formal response. Informally, in the meeting last week, we were told by senior HR managers that they could not account for how appraisals were conducted at Faculty level.

1. Appraisal conducted as a perfunctory online process, not face-to-face
2. Appraisal focussed on outcomes, rather than based on development needs
3. Moderations not communicated to member
4. Moderations communicated to member without justification
   1. Line manager unable to explain moderation decisions
   2. Refusal to communicate the basis of a moderation decision leading to an as-yet-unresolved FOI request
5. Promotion/end of probation denied on the basis of student module evaluations
6. Research leave denied on the basis of moderated appraisal score
7. Appraisal outcome submitted by line manager without agreement of appraisee
8. Failure of manager(s) to conduct appraisal, potentially resulting in outcome of ‘2’ through no fault of the appraisee – particularly seen in cases where employee has two managers/is employed cross-faculty
9. Failure to return moderated scores, even as late as September 17, exacerbating and complicating the redundancy consultation that began earlier in the summer
10. Performance measures used outside the appraisee’s control (grant income; student evaluation)
11. Appraisal not conducted in line with member’s contract, leading to an improperly constituted capability procedure
12. Single element of appraisal framework becoming the basis of the final score (research, teaching)

Fundamental to all of these issues – and the problems, even crises – that they have wrought on members is the rating system, to which we reluctantly agreed, and the use of the moderation process and the benchmark ‘bellcurve’ distribution, to which we did not. The principles, which stressed a forward-looking, developmental conversation, have disintegrated into a numbers game, in which the only monitoring or measure of success reported to JNC has been completion rates. When we are all expected to exceed expectations without exceeding expectations (see the VC’s address of 2 June, streaming at 1:06:38: ‘we all need to appreciate that for most of us, we would expect to get a rating of meeting expectations in most years for the job we are doing – in some years I really hope we would all get exceeding expectations’), the scores become meaningless in terms of career development or performance enhancement. They are only there to pit our individual achievements against each other, rather than to encourage community. This gives us the answer to our second question ‘What are appraisal scores for?’ Their only purpose is as management information – poor, unreliable, and crude data that inform performance management of and by inadequately trained or prepared academic managers, and that might ultimately inform restructuring and redundancy consultations.

What can we do now?

The meeting with HR ended with a plan and a request. The plan was to set out a timetable for the long-awaited formal review of the Reward policies. The request, made directly to the HR Director who sits on UEB, was that she take our proposal to the rest of senior management for the rating system and the moderation process to be abolished as soon as possible. In its place, we proposed a report template based on the principles that we agreed in 2014.

Eventually the software should be reframed to reflect the agreed principles and scope of appraisal – and, given the traumatic experience of many members of the current system, we would recommend that wellbeing be added to the discussion. Proper training must be re-instated, for appraisers, appraisees, and HR partners – and the time for new and refresher training must be added to workload models (when they finally appear – but that’s another blog).

We are only too aware that the appraisal system, with all its faults, both in design and delivery, is a major cause of poor staff morale. While the branch will inevitably have its hands full with restructures, current and future, we do not forget our commitment to support the staff that remain employed here. We still want the University of Southampton to be a happy and thriving community of workers. We see a functioning, supportive appraisal system at the heart of such an aspiration, and we will continue to press management until we have achieved this.