Dr. Toby S. James, University of East Anglia*

Overview

Both established and emerging democracies often struggle to run elections smoothly.  Common problems with electoral management include low levels of electoral registration, miscounts, lost ballot papers and delays in the announcement of results.  A cause in some circumstances is the poor performance of electoral officials.  Officials managing the local implementation of elections, for example, might not have followed the guidance that their supervisors or international donors have provided or they might have not undertaken sufficient planning for an election.  Equally, systems might not be in place for electoral officials to share ideas for best practice between themselves so that they can improve their efficiency and the quality of delivery. 

Fortunately, there have been many developments in the tools that governments have used to increase the monitoring and performance of public sector workers.  One method is the use of performance benchmarking.  This involves defining standards of best practice and then measuring different units of an organisation against these standards.  The aim is to spread best practice and motivate better performance from local officials.  This article explains in more detail what it is and provides a case study of Britain where it has been useful for improving the quality of electoral management. The final section of the article makes recommendations for how it can be adapted for use in countries other than Britain and some of the dangers that need to be considered before doing so.

What is Performance Benching Marking?

Benchmarking is the measurement of performance across a range of similar units or teams within an organisation.  The idea first emerged from the private sector in the 1950s as a means to increase the quality and efficiency in the production process.  The rationale is that comparative measurement might allow best practices to be identified, shared and learnt.  It is a tool to motivate workers to improve their performance by either working harder or following the best practice established elsewhere.  It also provides a method of control within organisations.[1]  It has become commonly used worldwide in government since the 1980s. 

The UK Electoral Commission’s model

The UK has a decentralised system for implementing elections.  Although laws for elections are made in the national Parliament, the task of implementing them has always been done locally.  In the nineteenth century poor-law administrators played a key role in compiling the electoral register.  In the early part of the twentieth century the task of compiling the electoral register was given to employees of local government called electoral registration officers (hereafter ‘EROs’).  Meanwhile, officials called returning officers (hereafter ‘ROs’) have been responsible for running the poll and the count.  Often, these officials are the chief executives of local government.  The ERO and RO are often also the same person.  Either way, they have overall responsibility for managing a team who work within local government to deliver elections.  Different arrangements are in place in Scotland where Valuation Joint Boards, organisations who administer council tax, are responsible for the electoral register.   In Northern Ireland there is one central organisation called the Electoral Office for Northern Ireland which runs elections – but which is not considered in this case study.

Performance benchmarking has been widely used across the UK, especially within local government for over three decades.  It has not been publicly used in the organisation of elections, however, until an internationally innovative scheme was introduced in the mid-2000s.  This followed concerns, mostly raised by politicians, that the quality of electoral management was better in some parts of Britain than others.  The Labour government therefore included a provision in the Electoral Administration Act 2006 to give the Electoral Commission the power to implement a new performance management scheme.   The Commission was given responsibility for setting standards for electoral registration and the organisation of the poll.  They also were given powers to publish whether or not ROs and EROs met those standards. 

Following the passage of the 2006 Act the Electoral Commission undertook a consultation with stakeholders to decide upon the performance standards.  Ten standards for EROs were published in July 2008 and seven for ROs in March 2009.  These standards were processes (rather than targets for key performance indicators) which the Electoral Commission, in consultation with stakeholders, thought reflected best practice.  They are listed in Table 1.    For each standard, performance indicators were designed to measure whether each local authority was ‘not currently meeting the standard’, was ‘at the performance standard’ or was ‘above the performance standard’.   For example, the first ERO standard required them to use information sources to verify entries on the register.  To ‘meet the standard’ EROs needed to identify and use records such as the annual canvass, to verify and validate electors on the electoral register.  To be ‘above the standard’ they needed to go further and ‘identify and contact potential electors who may have moved into, or within, the local authority area’, using additional sources such as ‘council tax records to identify residents of newly occupied properties’.  To demonstrate that they had met this they needed to maintain records of which data sources were checked and when. Undertaking neither of these exercises would mean that an ERO was ‘below the standard.’[2] 

Local authority officials were asked to self-assess their performance and report it to the Electoral Commission. The Electoral Commission undertook a sample-based verification exercise to ensure that the self-assessment forms had been completed accurately.  The results were made available publicly via an online web-tool.[3] 

The Electoral Commission subsequently changed the standards in 2013 to facilitate the introduction of individual electoral registration.[4]  Different management systems were also in place for the Welsh Assembly and AV referendums in 2011 because the Commission has the power of direction in referendums. [5]  This article, however, focuses on the impact of the original indicators in use between May 2009 and May 2011. 

Table 1: Performance Standards defined by the UK Electoral Commission 2008 and 2009

Effects of the Quality of Electoral Management

What effect did the performance standards have?  Did they have the desired effect of bringing about change?  The author undertook an evaluation of the performance standard scheme.  The methodology involved qualitative interviews with all 74 electoral officials who were required to meet the standards.  The results of these were published in an academic article.[6]  The remainder of this ACE case study summarises these findings and considers whether the UK system could itself be rolled out as good practice elsewhere.

Level of compliance

It is important to note that the Electoral Commission had no formal powers to fine or otherwise punish those local authorities who did not comply with the performance standards.  Given this, the level of adoption was high.  In the first year of the use 19.8% of RO standards were ‘below standard’ but this dropped to 9.1% in the second year.  Meanwhile, only 3.3% of EROs standards were not met by 2010.[7] 

Table 2: Trends in local authority performance results according to the Electoral Commissions performance standards 2008-2010

Explaining compliance

Why did local electoral officials comply with the standards if they had no legal or financial reason to do so?  The reasons for meeting the standards varied.  The standards made officials aware of new ways of working or gave them the confidence to introduce reforms that they had heard about elsewhere.  Sometimes the standards prompted either formal or informal reviews of ways of working – practices which had otherwise been unquestioned for a long period of time.  Often they were adopted because they were associated with professionalism – it just was ‘the right thing to do’.  They also provided a template for organising elections in periods of change such as authority mergers or the appointment of new members of staff. 

However, the most common reason for why the standards were adopted was that individuals or organisations felt that they would suffer reputational loss if the standards were not met.  Middle managers commonly implemented standards because the reputation of the Chief Executive (who is often also the ERO and RO) was perceived to be at stake.  However, often the Returning Officer took action to ensure that changes had been made.  One junior official reported that she was ‘roasted’ by her RO (who was also the Chief Executive of the authority) because the authority did not meet the standards and this reflected ‘badly on her’.  ROs frequently knew their peers at other authorities very well and are part of a closed knit network.  The results from the standards were made available publicly via an online web-tool and they would check how they fared against their comparators.  Where individuals felt that their own reputation was not affected by the standards, they were less likely to act.   

Explaining non-compliance

Those officials who did not meet the standards explained their reasons for not doing so.  The most important was that they questioned the legitimacy or efficacy of the scheme.  Some thought that compliance would not actually improve elections in their area and resources would be better placed elsewhere.  Others stressed that they were already accountable to the law and courts.  The Commission’s scheme therefore held little motivation for them.  There was also evidence that some officials deliberately marked themselves as not meeting the standards initially, even if they thought that they were doing enough to meet them.  They could then subsequently claim improvements in future years.

The positive effects of compliance

There are some strong reasons to think that the standards had very little effect on electoral management.  Indeed, many officials stated meeting the standards was mostly a ‘box ticking’ exercise which didn’t affect the way that they ran their services.  A common theme from the interviews was that meeting the standards required them to document existing procedures but this did not change how they worked.  Some authorities even copied and pasted plans from officials at other authorities and occasionally even forgot to change the name of the authority on the plan.  Others, who were initially above the standards, reported that the standards encouraged them to drop their performance to being at the standard.

However, while many officials reported that there was no substantial change, many others said that it had a positive effect (see table 3).  Importantly, these were not always a consequence of what the standards were but a consequence of the presence of a set of standards.  One key benefit was that having externally defined standards increased confidence in procedures amongst local politicians and other elite stakeholders.  This is significant because other research shows how the public, knowing little about election administration themselves, take cues from politicians about the quality of election administration.  The presence of performance standards can therefore be important for improving waning confidence in the administration of elections.

Lessons for best practice

In conclusion, the performance benchmarking scheme used in Britain can improve electoral management and offers a useful template that other electoral management boards worldwide might want to consider if they want to improve the implementation of elections.  Benchmarking provides an effective way of identifying best practices, measuring the extent to which local officials are using them, and incentivising the wider adoption of these best practices.  The Electoral Commission did not produce any official estimates of how expensive it was to run.  However, it seems that costs were low since it involved only a small team of officials to administer on a part-time basis.

Those seeking to adopt this system might wish to consider the following recommendations:

• It is likely to be of greater use within decentralised electoral management boards where policy makers are concerned about variation in performance across the whole country.
• They are very likely to be more useful during periods of rapid change or when staff lack experience.
• A well-functioning EMB website is needed.  Performance data will need to be accessible, visible and updated on a regular basis.
• Standards should be regularly reviewed.  Once compliance for ‘base-level’ standards has been achieved, then there is scope for setting more challenging standards.
• To work well the scheme needs publicity.  Explaining the scheme to politicians and the media is important so that they can hold officials to account for non-compliance and motivate them to change.
• Identifying ‘best in the class’ local electoral officials is important so that their experience and ideas can be passed up and across rather than just down.  This will also increase the credibility of the standards amongst those who have to meet them. 
• There is a risk that the scheme can also undermine staff morale by ‘naming and shaming’ officials who are dedicated and working hard.  ‘League tables’ can put teams in competition against each other and this might have negative side effects.  The non-adoption of best practice might be because of a variety of factors including insufficient resources.  A supportive environment should be encouraged to help all officials achieve the performance standards to begin with.

Efforts to improve electoral management often focus on the legal-framework.  This performance management scheme, however, shows that the creative use of different managerial systems could have a substantial effect.  It also points to the need for further research on the policies that can be used to improve the delivery of elections.