Are elected health boards an effective mechanism for public participation in health service governance?

Abstract

Background  There is growing interest in the idea of elected members on health service governing boards as a means to induce public participation in planning and decision making, yet studies of elected boards are limited. Whether elected boards are an effective mechanism for public participation remains unclear.

Setting and participants  This article discusses the experiences of New Zealand where, since 2001, there have been three sets of elections for District Health Boards. Information on candidates and election results is presented along with data gathered via post‐election voter surveys. The article also considers the broader regulatory context within which the elected boards must operate.

Discussion and conclusions  The New Zealand experience illustrates that elected health boards may not be an effective mechanism for public participation. Voter turnout has declined since the inaugural elections of 2001, and non‐voters form the majority. Reasons for not voting include failure to receive voting papers, a lack of interest, or no knowledge of elections. The elections have also failed to produce minority representation, while the capacity for elected members to represent their communities is subject to constraints. On the upside, elections have enabled public involvement in various dimensions of participation, including oversight and processes of governance. New Zealand’s mixed performance suggests that elected boards may need to be complemented with other participatory channels, if increased public participation is the goal.

Introduction

Various ideas sit behind the pursuit of locally elected health service governing boards. First, that installation of elected board members will contribute to aims of involving the public in a more meaningful manner in health‐care planning and decision making, in turn, helping to build deliberative processes. Second, that elected members will be able to more closely represent the wishes of their communities, resulting in improved and more acceptable policy and services provision. Third, that elected members will have greater legitimacy within the communities who have elected them and, as such, will be better able to hold health‐care providers to account, to explain decisions to the public and to receive feedback. Viewed this way, elected governing boards have the potential to facilitate several previously identified components of public participation, including ‘overseeing’ ¹ and involvement in processes, ² community consultation and partnership, ³ the building of a ‘policy’ perspective where elected representatives work with the values of the broader community in mind rather than for specific interest groups, ⁴ and the delivery of people who could be either ‘ordinary’ or ‘professional’ in terms of the expertise they might bring to the board table. ⁵ Of course, political leaders may have other motivations for creating elected boards. In Canada, during the 1990s, responsibility for budgetary cuts was partly behind devolution to regional boards. ⁶ , ⁷ Such moves may similarly be designed to offset public concerns about a broader restructuring agenda, or of poor central government performance and overcentralization.

Pursuing public participation requires traversing a range of complicated processes. As others note, the nature of participation needs to negotiated among the parties involved (providers, the public and their representatives), which can be complex as there are divergent views of how to represent public perspectives, what a legitimate viewpoint is, and when public perspectives should be incorporated into the decision‐making process. ⁸ , ⁹ As a channel for participation, elections are a notoriously imperfect tool with wide‐ranging debates over whether they facilitate participation and the quality of this delivered. ¹⁰ Indeed, electoral systems often fail to produce minority representation, those elected may not represent community interests and voter turnout, especially at local government level, is frequently low. For such reasons, theorists such as Barber argue electoral democracy actually undermines and is far from the best method for engendering community participation. ¹¹ Thus, there are grounds to question whether, in practice, elected health boards would be an effective mechanism for public participation.

There is a growing trend in selected developed world health systems toward embracing the idea of elected members on health service governing boards. The Scandinavian countries have long had locally elected representatives responsible for health care, but these responsibilities have often been coterminous with local government and welfare services and so not explicitly attached to health. ¹² , ¹³ , ¹⁴ Two Canadian provinces have had short‐lived encounters with elected boards, with present developments re‐centralizing power. ¹⁵ In the United Kingdom, a series of announcements and reports have pointed in the direction of locally elected boards. In a 2008 speech on the future of the British NHS, Prime Minister Gordon Brown implied that communities would need to look for new ways of increasing local accountability for decision making as the government seeks to ‘devolve more responsibilities to the local level’. ¹⁶ In a similar vein, in 2007, the UK Local Government Association created a Health Commission to examine ways to improve local accountability. The resulting report suggested adding local representatives to Primary Care Trusts. ¹⁷ From 2010, the Scottish government will pilot elected boards in two districts with a long‐term goal of elected members on all district health boards. In the words of Scotland’s Minister of Health: ‘Elected health boards will give power back to local people. They represent a major boost for democracy and accountability’. ¹⁸

Questions around whether elected health boards really do boost participation and accountability remain under‐researched as experiences are limited. Published studies have focused predominantly on board members themselves. Lewis et al. surveyed board members in the Canadian province of Saskatchewan, where elected boards existed from 1995–2002. They found boards constrained by provincial government directives, but also that elected and appointed members had similar viewpoints on core issues, raising questions over ‘why the government should bother with elected boards’. ⁶ In any event, Canada’s experiment with elected boards was over by 2003 when Alberta abandoned theirs. More recent surveys of New Zealand board members produced similar findings. ¹⁹

This article probes a research gap. In particular, it explores whether elected boards are an effective mechanism for involving the public in governance processes. The Canadian experience suggests not. This article corroborates and extends upon existing research, describing the case of New Zealand where elected District Health Boards were introduced in 2000. With three rounds of board elections to date, there is considerable experience to draw upon. The article discusses election data, results of a series of post‐election voter surveys, and the broader regulatory context of the elected boards. The conclusion is that elected boards alone may not engender public participation; other channels may also be required.

New Zealand’s health system and district health boards

There are similarities between New Zealand’s health system and the systems of other tax‐funded OECD‐member developed world countries. ²⁰ The national government is the primary funder, contributing almost 80% of total expenditure, and public facilities, free of patient charges, dominate hospital care. Private hospitals provide only elective services to paying patients. Primary medical care is almost entirely provided by private practitioners who receive considerable government subsidies.

New Zealand has a tradition of locally elected health boards, with such structures in place prior to the 1990s. The decade of the 1990s was something of a historical aberration with government‐appointed boards intended to bring a business focus to hospital governance. The present District Health Board (DHB) system was created following the 1999 election, largely through the belief of a new left‐leaning government that the business‐like governance structures lacked local presence and failed to allow ‘the people to have a say’. ²¹ In many ways, the DHB system reinstated the pre‐1990s structures. ²² The system consists of 21 separate DHBs funded by central government, each responsible for planning, prioritizing and purchasing health services from an appropriate range of providers for their respective regions’ residents. They are required to focus on and develop strategies to improve population health and do so in collaboration with the community and in keeping with central government policy. ²³

Each DHB has an 11‐member governing board. Seven are elected by popular vote, with the remaining four appointed by the government including the crucial positions of board chair and deputy chair (although these may be selected from among elected members). Members are paid in accordance with the government fees framework, and also receive meeting expenses. In recognition of the Treaty of Waitangi, the founding agreement signed in 1840 between the British Crown and indigenous New Zealand Maori, the government decided that two of the 11 members should be of Maori ethnicity (roughly reflecting the proportion of Maori in the population). Serving each DHB is a permanent staff. As noted, the government’s aims for the DHB system were to improve population health, but also increase public participation in the health system by devolving decision making to the local level and democratizing the governing board. ²¹

The three DHB elections

Background

New Zealand has had three rounds of DHB elections, in 2001, 2004 and 2007. These elections were held in conjunction with other local body elections including city and regional councils and community boards. Local elections are held every three years and conducted using the first‐past‐the‐post‐electoral system which was the method used for the inaugural 2001 DHB elections. In 2001, DHB regions were broken into wards to ensure representation across diverse geographic and demographic areas. DHB regions were typically divided into urban and rural wards containing anything from one to five seats depending on the size of the ward population. DHB regions (built around the pre‐existing hospital regions) and wards did not correspond with local and regional government boundaries.

Local government officials were given responsibility for conducting DHB elections. Voting was by post, as for other local government elections. Local government was also charged with educating the public about DHB elections, receiving nominations and facilitating polling. Around a month prior to the election closing date, voting papers were mailed to all registered voters along with other local government voting papers. Voting packs included booklets containing candidate profiles and photographs.

Central government also launched a public education drive to announce that candidate nominations were open and elections imminent. This included television, newspaper and movie theatre advertisements, creation of an election telephone info‐line, and household letter box mail outs. Government information consisted of overviews of the DHB system, highlighting the aim of public participation in health‐care governance. Particular note was made of the fact that the elections offered individuals the opportunity to participate as candidates and voters.

Arrangements for the 2004 and 2007 DHB elections were largely the same as 2001, with a key difference that the single transferable voting system (STV) was mandatory. STV was optional for other local body polls, so intermittently employed. Also to be used in Scotland’s 2010 health board elections, STV requires that voters rank candidates in order of preference. Following calculations based on the number of voters and candidates, a quota for the number of votes required to be elected is set. Once a candidate has reached the quota and been elected, additional votes for that candidate are discarded and ‘surplus’ votes shifted to voters’ next ranked candidates until the next most preferred candidate reaches the quota. ²⁴ The use of STV meant abolition of localized wards in favour of pan‐DHB region electorates.

Candidates

As illustrated in Table 1, there was considerable candidate interest in the inaugural 2001 elections. 1084 candidates contested 146 out of 147 seats (one seat had only one candidate), meaning there were 7.4 candidates per contested seat. Some wards attracted high candidate numbers. For instance, voters in the Waitakere ward, one of three composing the Waitemata DHB, had a choice of 50 candidates contending three seats. In the Christchurch ward, 75 candidates contested five seats. The 1084 candidates were 55% male. 54 candidates (5%) were DHB employees, 73 (7%) were incumbent board appointees, and 127 (12%) were of Maori ethnicity. Perhaps predictably, given the different objectives and potential for greater politicization inherent within the DHB system, only around half of the pre‐2001 appointed incumbents decided to stand for election.

Table 1.

 DHB candidate information

	2001	2004	2007
Total candidates/seats contested	1084/146	518/147	428/147
Candidates per seat	7.4	3.5	2.9
Male (%)	55	57	58
Incumbents as % of all candidates/as % of contestable seats	7/52	24/84	28/82
Maori (%)	12	13	11

In 2004 and 2007, the number of candidates progressively dropped. Reasons for this remain unclear, but could correspond with the diffusion of knowledge about the contextual constraints on DHBs discussed below. The gender split in all three elections was roughly similar, as was the number of Maori candidates. There was a substantial increase from 2004 in the proportion of incumbent candidates (over 80% of incumbents stood for election in both polls). In some DHB electorates high candidate numbers remained. The Capital Coast and Counties Manukau DHBs had 40 and 41 candidates respectively, and several had over 30. However, the at‐large electorates meant candidates were vying for seven seats.

Election outcomes

Table 2 contains the selected outcomes of the three elections. Voter turnout was 50% in 2001, with a reduction in the subsequent two polls. The 2004 and 2007 results were further undermined by the fact that a sizeable number of voting papers were returned either blank or incorrectly filled out, rendering them invalid. There was an increase in the percentage of seats filled by incumbents in 2004 and 2007, although the proportion of incumbent candidates was also higher (Table 1).

Table 2.

 DHB election outcomes

	2001	2004	2007
Voter turnout (%)	50	42	43
Males elected (%)	55	57	54
Incumbents elected % of all contestable seats	36	56	66
Maori elected (%)	3	8	8
Blank or invalid votes (%)	6	15	17

Candidates from a range of professions and backgrounds were elected in 2001. For instance, 37% had experience in the ‘health professions’ including medicine, nursing, midwifery and pharmacy, 31% had worked in business or law or had company director/analysis experience, and at least 11% had backgrounds in community work and advocacy. In 2004, 12% of those elected were employed by the DHBs they were elected to. More than half had prior experience in local government. By 2007, over 70% had such experience but the proportion of DHB employees elected dropped to 6%.

The high candidate volume in many wards in the 2001 poll meant that most DHB members were elected with a small percentage of total votes. For example, in the Tauranga ward (33 candidates; three seats) 87 485 votes were received from 45% of eligible voters, each of whom was able to vote for up to three candidates. The three successful candidates received between them 23 160 votes (26.47% of the total). The situation in 2004 and 2007 elections is difficult to compare owing to the introduction of the STV voting system and at‐large electorates. However, only a small proportion of the low numbers of participating voters backed each of the seven successful candidates in each DHB electorate.

Voter behaviour

How voters made their choices is an important question, given the numbers of candidates in each of the three elections. In 2001, an added difficulty was the small number of incumbents and fact that most candidates lacked experience in DHB governance. To investigate voter behaviour, a fixed‐response telephone survey of 500 voters, randomly selected from telephone directories, was conducted by a research company for the author immediately following each election (the Appendix contains the full survey). In 2001, 100 respondents were sampled from each of five wards with large candidate numbers, representing rural and urban areas in the North and South Island of New Zealand. In 2004 and 2007, five DHB electorates were targeted with a similar sampling approach. The surveys had a margin of error of 4.3%.

In both the 2001 and 2004 surveys, 65% of respondents had voted in the DHB elections, suggesting the survey sample was not wholly representative of the general voting population. In 2007, 53% had voted which was closer to the election turnout, but still not representative. Those who had not voted were asked why. The most frequently cited reasons are listed in Table 3. Notable is the increase in 2004 and 2007 in those who did not know why, or who did not know about the elections. Notable also is the proportion in all three polls who did not receive voting papers.

Table 3.

 Why people did not vote¹

	2001, n = 173	2004, n = 183	2007, n = 237
Don’t know (%)	20	28	35
Didn’t know about elections (%)	15	16	19
Didn’t receive voting papers (%)	12	18	12
No interest in elections (%)	27	30	17

¹Highest scoring categories.

Respondents were asked how they made their choices from among the multiple candidates. As shown in Table 4, the candidate profiles supplied with voting papers proved a useful information source for many. A proportion looked for candidates they knew. A small number resorted to guess work.

Table 4.

 How voters chose candidates¹

	2001, n = 327	2004, n = 345	2007, n = 263
Used candidate profiles (%)	61	53	64
Looked for someone I knew (%)	26	35	27
Took a guess (%)	4	3	3

¹Highest scoring categories.

Respondents were also asked about the main candidate qualities they looked for in making their choices. The results are shown in Table 5. Responses for all three elections were similar, with a low preference for candidates experienced in management and finance. Finally, in the 2004 and 2007 surveys, respondents were asked whether they found the STV system confusing. One third agreed that it was.

Table 5.

 Main candidate qualities sought¹

	2001, n = 327	2004, n = 345	2007, n = 264
Experience in health service (%)	57	61	56
Experience in community work (%)	20	22	23
Experience in management/financial matters (%)	7	7	7

¹Highest scoring categories.

Appointed board members

Following the 2001 elections, the government released its list of DHB chairs and appointed members. Appointees were selected in accordance with a number of criteria, including the desire to ensure a complement of skills and experience around the board table. Over two‐thirds of appointees were pre‐election incumbents and more than 60% fulfilled the criteria of being Maori. A number of appointees had stood as candidates but failed to win seats. In 2004 and 2007, the vast majority of sitting chairs were re‐appointed to provide, as suggested by the Minister of Health, ‘continuity and stability’. Contrasting with 2001, there were few incumbents among appointees, but a similarly high percentage self‐identified as Maori.

Contextual constraints on the boards and participation

The context of DHB boards is complex as they must carry out governance duties in an environment where government funding is restricted yet responsibility for how that funding is spent is devolved to localities. ²⁵ Several additional tensions surround the DHB boards. ¹⁹ , ²⁶ , ²⁷ , ²⁸ Evident in other health systems with local governance boards, ⁷ , ²⁹ these are perhaps accentuated due to the elected nature of DHBs.

First, board members, particularly those elected, sit in a difficult position of having dual accountabilities to central government as the funder, as well as to their communities. In anticipation of this, the legislative framework, the New Zealand Public Health and Disability Act 2000, states that DHB members are responsible first and foremost to the government. DHB boards and individual members who fail to produce a ‘satisfactory performance’ can face various sanctions from funding withdrawal and government supervision through to removal. As such, there is limited scope for those seeking to challenge government policy or to advocate on behalf of voters. It has not been uncommon for DHB members to reiterate this in public comments, even to suggest that they are simply government messengers, ²⁷ fuelling public doubts about the potential for board members to represent their interests. Despite this, a survey by Barnett et al. found 70% of elected members felt ‘able to handle’ the dual accountability. ¹⁹

Second, DHBs differ from other local governing bodies, such as city councils, in that they have a limited role in formulating policy. They are largely responsible for implementing central government health policy and restricted to developing local methods for this. While this does permit DHBs to produce innovative localized plans, these must be designed to achieve government goals and endorsed by the government as the funder. ²⁶

Third, with a fixed budget yet ever‐increasing service demands, combined with a requirement to consult the public in prioritizing services, DHBs are in a position of having to seek community input into choices between different services and patient needs. ³⁰ That said, as a general rule, there has been minimal public interest in the deliberations of DHBs. ¹⁹ While all board meetings are open to the public, attendance is often only by local newspaper reporters. Boards are, however, required to consult the public in their planning and so will issue discussion documents for comment or approach known interested parties. Often, this has led to revised plans and decisions.

Discussion

Through the period of the three elections outlined above, some key shifts occurred. First, the number of members of the public presenting themselves as candidates progressively dropped. There are various possible explanations for this. Public interest in participation as elected representatives may simply have declined after initial excitement, but also in recognition of the fact that, as noted above, representatives are there primarily to serve the government’s, not their electors’, interests. Of course, this undermines the notion of local representation and the scope for elected members to advocate on behalf of electors. If measured against Arnstein’s participation ladder, ³¹ the result might be somewhere akin to a board, once elected, seeking ‘tokenistic’ public involvement or, at best, ‘consulting’ the public periodically during the decision‐making process. The drop could be a function of the shift to the STV voting system and eradication of local wards, reducing the capacity that candidates might envisage for representing specific communities. It could be that the increasing proportion of incumbents standing for and gaining election has had a crowding out effect on other potential candidates as observed elsewhere. ³² This said, the ratio of candidates to seats at all three of New Zealand’s health board elections could be seen as a measure of community support for, and therefore a positive outcome of, the move to elected boards.

Second, public interest in the electoral process diminished as measured by voter turnout. New Zealand could exemplify arguments of democratic theorists that when there is a failure to boost participation this is because people have limited opportunity to participate. ¹⁰ , ¹¹ As noted, capacity for representation via elected members is restricted. However, the low turnouts were in keeping with New Zealand local government elections, which generally attract under 50% of voters, but above the 36% in Britain’s 2008 local body elections and 10% for Saskatchewan’s health boards in 1999. ⁶ Research suggests that postal voting increases participation, particularly in local government polls. ³³ The DHB election turnouts may therefore have been worse with a traditional method requiring voters’ physical presence. Yet the turnouts raise questions over the extent to which the broader public are interested in health‐care governance and even whether they should be expected to vote. As Lomas et al. suggest, albeit in the context of appointed Canadian boards, ‘most citizens wish only to be consulted and they expect and prefer that “the experts” take responsibility for actually making the decisions’. ³⁴ If so, participatory mechanisms other than elected representatives may be more appropriate for garnering community input.

Like the results from any public opinion poll the survey data reported in this article needs to be treated with caution, but does provide additional insight into the low voting levels. A proportion of voters claim not to have received voting papers (it is possible these were received, misplaced and forgotten about), denying them an opportunity to participate, while a sizeable number had no interest or did not know about the elections. It is possible, given the restrictive operating environment of DHBs outlined in this article, that some voters saw little point in voting as they envisaged limited scope for elected members to have any influence. This context is not unique to New Zealand. Other centrally funded health systems could face similar challenges in creating operating frameworks for locally elected boards. Such challenges were certainly borne out in the Canadian case. ⁶

Third, the percentage of blank and invalid votes expanded. As discussed, the DHB elections were in conjunction with other local body elections which mostly use first past the post while others, such as DHBs, used STV from 2004. Voters therefore received papers containing a mix of voting methods which could well have been confusing. ³⁵ The post‐election surveys found that STV alone confused a third of voters.

Fourth, the number of indigenous minority Maori candidates has been steady while the number elected has grown but remained low, invoking use of the appointments system to ensure representation. New Zealand’s mixed elected‐appointed board solution may be a useful means for countering the minority under‐representation that is a common function of electoral systems. ²⁴ This said, all Maori appointees are selected in accordance with several criteria and bring multiple skills and experience in community and health‐care service. ³⁶

These shifts have implications for the question of how effective elections are as a method for inducing public participation in health‐care governance processes. As noted in the introduction, electoral systems often fail to deliver on aims of democracy and participation, which could well have been the case with DHBs. To be fair, New Zealand’s elected boards have permitted members of the community to present themselves to the public as prospective members, in contrast with the select few appointed by government with the preceding governance system. In this sense, elections have formalized the involvement of public representatives at the highest local decision‐making level in two important dimensions of participation identified in prior studies: oversight ¹ and the processes ² of planning services configuration, funding and delivery.

The elections have delivered board members with a mix of backgrounds, perhaps satisfying the dual demands identified by Learmonth et al. for those in governing positions to represent ‘ordinary’ lay‐people, as well as bring ‘professional’ capabilities to the board table. ⁵ That said, the gradual increase since 2001 in the dominance on boards of incumbents – 70% of whom had prior experience in local government or health care – is perhaps indicative of a diminishing scope for Learmonth et al.’s ‘ordinary’ people to be elected. Moreover, the characteristics of New Zealand’s elected members have never mirrored those of society. In this sense, research into DHB governance processes concluded that ‘the architects of the reform overestimated the extent to which elected members could contribute unique community perspectives’. ³⁷ The source of tension that the presence of health‐care employees created in Canadian boards ⁶ does not appear to have arisen in New Zealand’s DHBs. ¹⁹ However, the concentration of DHB members with health‐care interests (as service providers, health‐care business owners, or DHB employees) has meant a requirement that all potential conflicts of interest are publicly notified.

To summarize, New Zealand’s experience with elections suggests the constraints on boards (which may be unique to centrally funded health systems), declining public interest in DHB elections, diminishing opportunities for lay‐people to be elected, and the failure of the elections to produce adequate Maori minority representation have meant restricted scope for delivering public involvement in governance processes. With this in mind, and the fact that the DHB system was intended to facilitate participation, the importance of other regulatory requirements increases. As discussed in this article, these include considering the needs and preferences of the entire community in the planning process, and promoting community consultation in planning and decision making. These requirements and the corresponding focus of boards, which studies suggest have shifted the orientation of New Zealand’s health system towards public health improvement, ²⁵ are in keeping with Charles and DiMaio’s notion of the building of a policy perspective that is concerned with the community rather than representation of narrow interests or concerns. ⁴ However, as this article has implied, whether the mechanism of elections is a component that contributes to this process remains questionable.

Conclusion

Those advocating elected health boards can learn from the New Zealand experience which indicates that electoral mechanisms may play only a limited role in promoting participation, and could possibly counter public involvement. Designing electoral systems and processes of involving the public in health planning and decision making are, as widely acknowledged, complicated, ² , ⁸ , ²⁴ and participation efforts often fail to involve more than a few, often unrepresentative, members of the public and only in select components of processes. ³ , ³⁸ , ³⁹ For these reasons, an elected board may be but one of multiple, parallel methods for public participation – such as Britain’s Local Involvement Networks, ⁴⁰ or New Zealand’s statutory requirements that health planners consult with the public – that might be implemented.

Supporting information

Appendix S1 Survey excluding background questions

Supporting info item