Health Board Elections and Alternative Pilots: Final Report of the Statutory Evaluation

4 Costings

Elections

4.1 The Health Boards and Local Authorities reported that the pilot elections and associated publicity cost £473,850 in Fife and £299,406 in Dumfries and Galloway.

4.2 The main costs of an all-postal election come from printing and posting ballot papers and counting the returns. The cost of running a direct election is therefore proportional to the number of people eligible to vote and the turnout. In principle, we can therefore estimate the cost of running similar elections across Scotland by multiplying the average costs of the elections in the two pilot areas to correspond to the population of Scotland. The two pilot areas had 403,139 registered electors, 10.36% of the electors in Scotland. Simply multiplying the costs of the pilot elections (£773,256) to correspond with the number of electors in Scotland suggests a total cost of £7.46m per election.

4.3 However, this assumes that turnout in a Scotland-wide poll would be the same as turnout in the pilot elections. Returning Officers were instructed to prepare for the election on the perhaps optimistic assumption of 60% turnout, so as to reduce the chance of the count being delayed by unexpectedly large numbers of votes. This would presumably be repeated in any future elections. However, the postal cost would vary depending on the actual level of voter turnout. This should be factored into our estimate. Given the large differences in turnout between Dumfries and Galloway and Fife, it is difficult to predict the turnout rate across Scotland. However if one assumes equal turnout across all board areas then we estimate the most basic cost of running the elections at different levels of turnout would range from £7.44M at 15%, through £7.73M at 30% and up to £8.13M at 50%. However, such estimates of cost do not allow for important issues of accessibility and security which we believe would be considered desirable in future elections. We deal with these issues in sections 4.8 to 4.9. Other additional cost factors are discussed in sections 4.9 to 4.15

4.4 Estimates assume that elections in all Health Board areas would be contested. The pilots attracted many candidates and it seems reasonable to assume that at least the first elections would generate enough interest to require a vote.

4.5 These figures reflect the cost of the election process. We have not included the additional costs of paying and training larger cohorts of non-executive members if their numbers were increased as a result of the elections.

4.6 There are several important caveats to these figures, which suggest that the figures would be a low estimate of the cost of direct elections.

4.7 Firstly, the pilot elections did not have the security features normally required for postal voting. In a conventional election, only electors who have registered for a postal ballot receive one. When requesting a postal ballot those electors give their date of birth and a sample signature. When they return a completed ballot paper, they give another signature that is checked against the specimen to verify their identity. In the pilot elections, no signatures were required. As there is no national database of electors' signatures, in order to obtain a similar level of security Returning Officers would need to either create such a database or devise an alternative security check. While we saw no evidence of fraudulent voting in the pilot elections, if Parliament were to mandate a national rollout without security measures there would be an increased risk of fraudulent voting. We understand that when the Bill which led to these pilots was going through Parliament a representative of the Scottish Assessors estimated the cost of such measures at £1 per elector, which would imply an added cost of a little under £4m per election cycle. Thus this factor alone could increase the cost of an election to a total of somewhere around £11 to £12 Million. This and further factors articulated below make it likely that the cost of running these elections under normal conditions would significantly exceed the figures reached by a simple extrapolation from the pilot elections.

4.8 The second of these factors is translation. The ballot papers and candidate statements were delivered only in conventional print and in English. If elections were rolled out across Scotland there would need to be some mechanism for translation. There is no national database showing which electors do not understand English or have sight problems. Some mechanism would need to be put in place to ensure electors who needed translations or large print/tactile ballots and candidate statements would have access to them. This would inevitably add additional cost.

4.9 Thirdly, the cost for electronic counting is extrapolated from the outcome of negotiations between the Returning Officers and Opt2vote on this occasion. New payment structures would have to be negotiated were the elections to be rolled out nationally. There are only a few private companies specialising in the electronic counting of ballot papers that could fill this role in Scotland. The nature of their business involves bursts of high demand for personnel and equipment on election days. If elections occurred at points of high demand, Returning Officers might find it more difficult to obtain such competitive terms as they did for the pilots.

4.10 Fourth, in the pilots no attempt was made to create a register of 15-year-old attainers. This led to a large number of 16-year-olds whose birthdays fell on the wrong side of the 29^th of November being ineligible to vote in the pilot elections on the 10^th of June. If all 16-year-olds were to be eligible to vote in any future Health Board elections, then a register of 15-year-olds who would become 16 in the next year would need to be added to the annual canvass. Stringent privacy laws for under-16s may well imply a need to keep a separate register for those attainers, and there might be costs associated with this.

4.11 Fifth, ballots were printed in alphabetical order. There has been some speculation that this advantaged certain candidates over others because of the very large numbers of candidates who stood in the two pilot elections. We have no evidence to support this, but would advise that Parliament consider making some provision for the order of candidate names on ballot papers to be randomised if elections are rolled out. This would probably increase the cost or complexity of printing ballot papers, as more print runs would be required (see 6.6 for more on this point).

4.12 Sixth, the low levels of voter awareness suggest that extra spending on publicity might be beneficial.

4.13 There would also be some unquantifiable costs in staff time to both the Returning Officers and the Health Boards. By their nature the elections introduced a large number of relatively inexperienced non-executives at one time, which created pressure on pilot Board officials to provide extra support.

4.14 The pilot elections increased the number of non-executive members on the Health Boards, leading to an added pay cost (and an unpredictable figure for extra expenses). The number of non-executives allocated to each Board is a matter for Parliament and the Scottish Government, but each additional non-executive would cost £8000 per year plus expenses, assuming that no pay was reclaimed from executives losing their Board status. If territorial Boards gained six non-executives on average (as in Dumfries and Galloway) the payroll cost would be £672,000 per year^[16].

4.15 Note that we have not considered the possibility of paying members of Health Board Committees, a possibility raised by the Health Boards (Membership and Elections) (Scotland) Bill. We have not considered the possibility of extra legal costs were the result of a Health Board election ever to be challenged.

Alternative Pilots

4.16 The alternative pilots added only two new members to each Health Board. However, even taking into account the fact that the elections produced more new members the alternative pilots still had much lower costs.

Grampian

4.17 The major cost difference between the Grampian pilot and a conventional public appointment was an increase in advertising cost of £8022 for the Health Board area. The pilot also added two extra non-executives who are paid £8000 per year and can claim expenses; because of the NHS pay protection policy, the salaries of two executives who could otherwise have had Board status were not reduced.

4.18 The cost of the extra non-executives rolling out a similar advertising programme across Scotland would vary depending on the advertising techniques used and negotiations with individual media outlets. The cost of advertising will of course vary depending on the characteristics of the Board area: a square inch in a small local newspaper costs less than a square inch of a newspaper with a much larger circulation. On the other hand, newspaper and radio coverage does not necessarily map well onto the boundaries of territorial Health Boards, and media based in the major population centres within which advertising is expensive dominates many rural or suburban areas. It is therefore difficult to predict the relative costs of comparable advertising campaigns in different areas precisely.

4.19 Given these variables, a reasonable estimate for the extra cost is simply to multiply the additional costs for the pilot by the number of territorial Health Boards in Scotland, giving a figure of £224,000 per year, and £112,308 per appointment round, at 2010 prices.

4.20 There would also be significant but unquantifiable opportunity costs in staff time to the Health Boards. For example, staff would need to spend time providing advice and encouragement to potential applicants when they could potentially have been working on other tasks.

Lothian

4.21 The costs of the open recruitment round in Lothian were similar. Recruiting a member from among the existing PPF office-bearers actually saved money, as there was no need to publicly advertise the vacancy; the Board simply wrote to them. Otherwise, the costs of selection in OCPAS assessors' time etc. would have been comparable with a conventional public appointment.