The impact of the central government’s intervention on local democracy

ABSTRACT

Does the central government’s intervention in the operation of failed local authorities affect the functioning of local democracy? We examined the theoretical and practical factors associated with the opportunities and risks inherent in an intervention policy in Israel called convened committees. We developed an index called the Effective Number of Council Parties by Seats at the Municipal Level (ENCPSML) and a number of measures that together express the degree of democratic representation, governability and voter turnout. We compare the functioning of democracy before and after the convened committees in 31 local authorities and a control group of 191 local authorities during four elections in 2003-2018. Contrary to what one might expect, convened committees did not hurt the exercise of democracy and in some cases, improved it. We discuss the practical and theoretical implications of these findings in the context of multi-level governance, local democracy, regulation and coping with local crises.

KEYWORDS:

• Central government’s intervention
• convened committees
• democratic deficit
• representation
• governability
• effective number of council parties at the local level (ENCPSML)
• local crisis

Introduction

Since the beginning of the twenty-first century, local governments have undergone a fundamental shift from local government to local governance. While local authorities were traditionally responsible for providing basic services, they are now expected to act as proactive entrepreneurs by offering diverse services and addressing economic, environmental and community development issues. In the wake of these growing responsibilities, local authorities must face new challenges, standards and complexities. From the municipal perspective, one of the side effects of the New Public Management (NPM) reforms introduced since the 1980s is the erosion of central and municipal control, which in turn led to the fragmentation of the delivery of public services (Bache, Bartle, and Flinders 2016). In the era of multi-level governance some local authorities developed new methods of governing as an alternative to traditional government. Examples include networks and regional alliances (Peters and Pierre 2004). However, the local authorities that were unable to adapt were faced with an assortment of functional, financial and democratic failures (Osborne 2010; Stoker 2004). Recently, the COVID-19 outbreak retriggered a major debate about the relationship between the central and local governments and the ability to manage local crises. Both sides claimed that they should have power, control and autonomy over such situations (Beeri 2021; Kuhlmann and Franzke 2021; Kuhlmann et al. 2021).

The ongoing demands for better performance have led some governments to introduce multi-level governance (Rosenau, 2004; Skelcher and Torfing 2010). The term is defined as the horizontal and vertical interdependent policies of nested networks of governments at several territorial tiers and levels such as the supranational, national, regional, and local – as a result of the extensive re-allocation of some centralised functions to the local level (Grimm 2011; Marks 1993). Multi-level governance is more about the process that promotes the public’s interests rather than its individual member institutions (Peters and Pierre 2004). Thus, multi-level governance also deals with the current deterritorialisation of politics, governance and the persistent seeking of local democratic legitimacy (Schmidt and Wood 2019; Skelcher and Torfing 2010). Indeed, the degree of local democracy is key to understanding the public’s satisfaction with the local government’s performance (Mendez and Grose, 2011; Lowande, Ritche, & Lauterbach, 2019).

In Israel, the central government has continued to intervene to help failing local authorities and enforce its changes. This practice raises questions about the degree to which these interventions impinge on the democratic functioning of the local municipalities (Cornforth and Paton 2004; Navarro and Velasco 2022). In this study, we investigate one of the interventions that took place in Israel, that is, neutralising the elected local leadership and replacing it with a convened committee. A convened committee is defined as a temporary, non-political management team deputised by the Interior Ministry to turn around failing local authorities (Reingewertz and Beeri 2018). Convened committees are installed to replace failing, sometimes corrupt, local leaderships and to deal with a combination of shrinking resources and poor management (Beeri 2013; Beeri and Navot 2013; Martin 2002) exacerbated by economic slowdowns.

Previous studies have examined whether the neutralisation and replacement approach is effective in turning around the quality of service and financial performance of failing local authorities (e.g., Beeri 2013; Beeri and Kval 2010; Reingewertz and Beeri 2018). One of the criticisms of this practice is that it harms democracy in that it deprives citizens of their representation and thus makes the ruling committee less accountable to the public (Bache, Bartle, and Flinders 2016). Thus, there is still room to expand the discourse and the long-standing theories on the relationship between the central and local governments and organisational turnaround.

Our goal is to extend the discussion in this field by identifying the theoretical and practical factors related to the possible effects and harm to local democracy that this practice might have. We assess the possible harm using three measures: representativeness, governability and voter turnout. We define representativeness as the relationship between council members and the electorate when the former governs in a manner responsive to the expressed interests of the latter. We measure representativeness as the number of council members and political parties relatively to the size of the council (Eulau and Prewitt 1973; Pitkin 1967). Governability is defined as the ability to govern, the capability to make decisions and implement policy, and to exercise the legitimate authority given to elected members by the electorate (Beeri and Zaidan 2021). Voter turnout is the ratio of the number of voters to the population that is eligible to vote in local elections.

Based on Haus and Sweeting’s (Haus and Sweeting 2006) typology of competing notions of local democracy, we consider the dangers posed by this solution to the effectiveness of democratic and governance mechanisms. Thus, our research question asks: How does the central government’s intervention in local municipalities in the form of convened committees affect local democracy?

To explore this question empirically, we use a before-and-after methodology with data from Israel. We compare the level of local representativeness, governability and voter turnout before and after the convened committee took office. Israel provides a good case study for two reasons. First, it is known for its centralist approach to governing (Beeri 2021). Second, because local governments have gone through several crises, the national government has made frequent use of this method for dealing with failed local authorities (Beeri 2013; Reingewertz and Beeri 2018). As a result, we have data on our three factors before and after the imposition of convened committees that we can use to conduct a comparison.

Our data come from the 2003, 2008, 2013 and 2018 local elections for 191 Israeli local authorities, 31 of which had convened committees installed. Based on the Effective Number of Parties index introduced by Laakso and Taagepera (1979), we developed an index called the Effective Number of Council Parties by Seats at the Municipal Level (ENCPSML). This index calculates the degree of representativeness and governability of the local council and voter turnout.

Previous studies have usually examined the tension between representativeness, efficiency, governability and accountability on the national parliamentary level (e.g., Anckar 2000). Our research contributes to the literature by exploring this tension on the local level. We also develop a tool that future studies can utilise in their explorations of this tension in other countries and contexts.

The paper is organised as follows. In the first part we describe the development of crises in local government in Israel and the use of convened committees as a central government intervention to cope with these crises. We then propose the research hypotheses based on the expected association between the appointment of a convened committee and the development of local democracy. In the second part we outline the research method and the findings. In the final part we discuss the findings in view of existing knowledge as well as their implications with respect to local crises, government regulation and the shaping of local democracy.

Literature review and development of the hypotheses

Local failures

The role of local government as the main interface between citizens and government is evident in extreme situations. For example, during emergencies and crises citizens expect the local authority to provide both personal security and community resilience. In effect, under such circumstances the definition of the role and ethos of the local authority expands, such that it is no longer limited to a commitment to provide high quality local services. Today, the local authority is also expected to take the initiative, adopt a more holistic view and thus develop and involve the local community, while promoting the economy and environment in the area of its jurisdiction and in the regional space. For example, in response to the COVID-19 crisis, municipalities were expected to play an autonomous and active role in fighting the social and economic effects of the pandemic (Navarro and Velasco 2022). As a result of these changes, local authorities are subject to increased demands and expectations on the part of national decision-makers and local stakeholders and residents. However, some local authorities respond less effectively than others and have difficulty meeting these demands, resulting in major management challenges (Cornforth and Paton 2004; Kuhlmann et al. 2021; Reingewertz and Beeri 2018).

In the last four decades, these demands have clashed with the local authorities’ declining revenues and often ineffective management and crisis management, leading to a widening gap between expectations and the performance of local government. Simultaneously, national governments have become less tolerant of local failures (Cornforth and Paton 2004). Such failures involve the inability of the local political leadership to govern effectively, the steep decline in the delivery of local services and the overall poor performance of the government (Balgobin and Pandit 2001; Beeri 2013). They also imply the inability to meet the economic and social needs of communities and residents. They threaten a locality’s future development, and if uncontained, can have repercussions for other localities and for the system as a whole (Boyne and Meier 2009).

One of the reflections of these failures is democratic deficits. These deficits are evident in apathetic voter turnout, the composition and functioning of the local council, and the mutual relationships between political parties and the voters. Failing local authorities may suffer from the erosion of local trust, political instability that stems from the over-representation of relatively small groups in the form of numerous small political parties, and the lack of participation of community members (Bessette 1980). During crises, coalitions in local authority councils tend to split into many parties, leading to a shaky, fragmented, and over-representative political system (Ghanem and Mustafa 2009; Paxton 2020). Shifts in values and changes in populations can lead to more small parties, which might be positive, in that they reflect the community’s diversity and are more representative of it. Nevertheless, we contend that in practical terms such a situation indicates a decline in political stability and local democracy. This fragmentation makes it more difficult to create a local council that is stable and can actually govern.

As a result, the relative size and power of the ruling party in these fragmented councils is small. Here again, there is a tension between representation and stability, both of which are important for local democracy. On one hand, an effective coalition may involve a range of small parties. On the other hand, the dominance of one or two parties may be stable but not particularly democratic. Indeed, there is no clear-cut rule for achieving both stability and representation. However, at least in the case of crisis and organisational turnaround, fragmented councils tend to form loose coalitions in which the price of political stability increases, and major compromises are required to please small parties. Consequently, this democratic deficit might deteriorate into the inability to achieve a turn around and govern, as well as poor performance in the areas of budgeting, accountability, effectiveness, leadership, quality of management and ethics (Beeri and Razin 2015; Ben-Bassat and Dahan 2009). Therefore, nations with varying government traditions are eager to find effective methods for dealing with poor performing local authorities (Shah 2006). Examples include the United States, which has utilised scrutiny of the mayor (Hambleton and Sweeting 2004), the UK, which imposed strict auditing, performance measurements and management strategies to try to turn around these local authorities (Boyne 2006), Germany, which modernised local administration through the New Steering Model (Kuhlmann, Bogumil, and Grohs 2008), Japan, which enforced local re-structuring through the self-reform of the local bureaucracy itself and by empowering citizens to pressure for outsourcing, privatisation and other forms of collaborations (Kudo 2003) and Israel, which neutralised and replaced underperforming local authorities (Reingewertz and Beeri 2018).

Most OECD countries use some type of enforcement mechanism to ensure discipline at the sub-national level, usually through fiscal responsibility laws that mandate fiscal accountability, transparency and stability and that impose sanctions on non-complying local authorities (Andrews et al. 2005; Joumard and Kongsrud 2003). Beyond fiscal control, other interrelated reasons justify intervention by the central government in cases of local failures. If left unresolved, long-term local failures can exacerbate the erosion of important administrative, democratic, political and social principles, including good governance, and corrode the relationship between the local and central government (Beeri and Yuval 2015; Cornforth and Paton 2004; Reingewertz and Beeri 2018).

In 2006, at the peak of the democratic and economic crisis in Israeli local authorities, 76% of them operated under deficits, 21% held back the wages of thousands of employees for months and over 60% activated recovery plans. The economic resilience of some of these local authorities deteriorated to the point of economic bankruptcy. Although strict supervision by the central government gradually reduced the level of debt relative to income, the per capita expenses of the local authorities also rose during those years (Beeri 2021; Ivanyna and Shah 2013).

Convened committees as an intervention of the central government

Stoker (2004) identified four major policies adopted by central governments to deal with failing local authorities: enabling, promotion, intervention and neutralisation. These policies reflect a spectrum of strategies, beginning with compromise and moving upward towards tighter levels of control. According to the enabling approach, when a poor performer moves away from accepted norms, the central government sets aside formal rules in favour of compromise and negotiation, while avoiding direct involvement. In the promotion approach, both the central and local authorities are motivated by the spirit of engagement rather than by intervention. This engagement may entail using a consensual set of fiscal rules to rebalance the local budget, combined with imposing financial penalties on the local government. The intervention approach may involve offering conditional government grants or removing certain powers from local hands. Alternatively, intervention may take the form of more insensitive actions, such as appointing an external accountant to keep the central government informed and to approve all local expenditures or forcing the local authority to implement a recovery plan that specifies goals in accordance with national priorities (Beeri and Yuval 2015).

Neutralisation constitutes the most severe response to poor performers and is employed as a last resort. While other interventions take some powers away from local councils but leave the local leadership structures in place, under neutralisation the central government (the Ministry of the Interior or its equivalent) removes all or part of the current local leadership. In doing so, it severely limits local elections, democracy and autonomy. The government then replaces the local leadership with a temporary management team in the form of a convened committee that usually serves for three to seven years. In other words, by replacing the local leaders with a convened committee, the government intervenes at the local level and imposes a new leadership that is responsible for turning around local performance (Reingewertz and Beeri 2018).

The appointment of a convened committee reflects the agency model perspective, which emphasises the notion that the local government is the executive arm of the central government – a local branch from which the central government draws its power. Objections to appointing a convened committee reflect the opposite perspective of the partnership model, according to which a democratic regime should be a dual regime in which the two branches of government – local and central – are equal in power. This approach stresses the local government’s right to independence and its obligations to the residents from whom it draws its strength rather than its dependence upon the central government. This approach deprives the central government of the right to intervene in the local government in any drastic manner. According to this approach, imposing a centrally appointed local leadership is unacceptable (Chandler 2001).

The perspective that the local leadership of a failing local authority must be replaced to recover is based on the assumption that the failure stems primarily from the deterioration of internal organisational resources, meaning, those leaders elected and bureaucratically appointed (Mellahi, Jackson, and Sparks 2002). Support for this assumption means that the recovery program should not and indeed cannot be implemented by those currently serving because they lack the administrative, leadership and political skills required for this task. Thus, the turnaround literature agrees that the dispersion of governance across multiple players who perform multiple functions is more likely to be able to solve problems, and is more efficient, specialised and professional than leaving the governing to one player (Beeri 2012; Hooghe and Marks 2004; Peters and Pierre 2004). In contrast, local authority officials tend to claim that responsibility for the failure is almost completely tied to outside factors and that the elected leadership should remain in place and be strengthened in order to insure governmental continuity (Peters and Pierre 2004).

The relationship between the appointment of convened committees and local democracy

Normally, the convened committee is not part of the local government framework, which makes its appointment a drastic choice that raises serious questions. Its intervention in local affairs introduces complexities and temporary imperfect solutions that most likely will result in new difficulties. The imposition of the convened committee reflects a clash between the traditions of the local and the central governments, between democracy and efficiency (Kersting and Vetter 2003). Indeed, the appointment of a convened committee goes hand-in-hand with the liberal democracy associated with New Public Management reforms. These reforms focus on the sovereignty of representative institutions that distinguish between the national state and society (Hupe and Edwards 2012). According to this view, when a local authority has failed to represent the public’s choices and has not upheld the high standards of good governance, there is justification for replacing local representatives with a temporary managerial team that, although not democratically elected, potentially represents the public will in a better way (Peters and Pierre 2004).

Opposition to the appointment of a convened committee is consistent with the call for interaction between the state and civil society, and for the enhancement of plurality by dispersing and fragmenting political identities and cultures (Sørensen and Torfing 2005). This approach views local, community and network actors as sub-elites capable of challenging the dominant elites by focusing on the plurality of interlinked democracy. It promotes this plurality by obscuring the boundaries between the state, society and local communities (Torfing, 2005). Those advocating this view will strive to keep local sovereignty in local hands, not in the hands of the central institutions, even at the cost of a lower standard of living.

From a more concrete perspective, the convened committee is liable to bias certain intrinsic policy outputs and preferences. Restricting the political and administrative powers of sub-national governments subverts fundamental concepts of local government, such as localism and decentralisation (Hooghe and Marks 2004; Pratchett 2004; Stoker 2004). More importantly, the neutralisation of the local representative democracy means that the essence of democratic local elections is called into question. The rights to vote and to be elected are diminished, at least in the short run, thus weakening the legitimacy, representativeness and governability of the local democracy (Hupe and Edwards 2012).

Put another way, the neutralisation approach may lead to a crack in the ‘loop model’ of democracy identified by Fox and Miller (1996). According to this model, democracy begins with citizens’ needs and preferences, which in turn influence the policies of candidates and parties. Successful candidates and coalitions can govern. They pass legislation, based on which the citizens judge their performance and accordingly reward or punish them in the next electoral round. However, appointed convened committees take away the local elected leaderships’ responsibility and accountability. They also take away the local electorate’s opportunity to judge local processes and performance and to reward or punish local politicians. Thus, under neutralisation, the local community – meaning, both voters and elected officials – does not rule but rather is ruled (Reingewertz and Beeri 2018).

The relationship between convened committees and local democracy can also be observed through the concept of local democracy introduced by Haus and Sweeting (2006). They identified four competing concepts of local democracy: representative, participatory, network and market democracy. Based upon this typology, the neutralisation approach and convened committees can be seen as outcomes of market democracy in that they marketise political relationships and stimulate competition over the production of local services with the aim of promoting governability, customer satisfaction and good governance. Inspired by public choice theory, market democracy is highly critical of traditional representative democracy on the grounds that the latter gives undue weight to the interests of bureaucrats, trade unions and politicians while not assigning enough weight to citizens’ preferences and the welfare of the community (Buček and Smith 2000). By disregarding the voices of both politicians and voters, and abrogating communal autonomy and the right to elect or be elected, convened committees contravene the notions of local democracy as representative, participatory and networked.

However, in contrast, one can also argue that the use of the convened committee mechanism is not totally anti-democratic, at least for the intermediate and long term. For local authorities facing crises, measures that are less severe, such as external scrutiny through auditing and inspection, may not be adequate to ensure recovery, especially where poor performance is judged to stem from intrinsic factors such as poor political leadership, weak management and executive failure (Ashworth 2003; Beeri 2012). Indeed, retaining local power in incapable hands runs the risk of eroding the residents’ will even more, creating further losses in efficiency and imperilling redistributive policies, such that no amount of auditing or inspection can guarantee a full recovery (Hooghe and Marks 2004; Joumard and Kongsrud 2003; Peters and Pierre 2004).

Consequently, the very features that make the central government’s intervention through convened committees effective in turning around failing local authorities (e.g., Reingewertz and Beeri 2018) may also make it suitable for use within a democratic context for several reasons. First, neutralisation creates the potential for re-establishing the relationship between the central and local governments on a more cooperative footing, with the convened committee more open to inspection and oversight that in turn may lead to a more transparent climate. Improvements in this relationship may also lead local government agencies to be more accessible and responsive than under the displaced elected leadership. Second, given that the convened committee is a non-political, non-partisan body, this form of neutralisation removes the distorting features of local politics from the decision-making environment, leaving room for rational and professional policies (Papadopoulos 2005). Such policies may re-establish and repair the ‘loop model’ (Fox and Miller 1996) in that local preferences, needs and demands are more likely to be represented and responded to, but without the price tag of over-representativeness that results in political fragmentation and instability during the process of turning the organisation around. Furthermore, repairing the ‘loop model’ may increase the residents’ trust in the local authority and local elections. As a result, they may participate more and reward the local officials by re-electing them. Finally, neutralisation may lead to the establishment of a cooperative and rational relationship between the convened committee, the citizens and the local institutions within a supportive culture.

To sum, we stress that convened committees have the real potential to strengthen rather than weaken local democracy. One possible way to assess the impact of the convened committees on local democracy is to look at the multidimensional nature of the concept of representation as formalistic, descriptive, substantive and symbolic (Pitkin 1967). Formalistic representation denotes the legal authority to act for another. Unlike the failing local leadership, convened committees are not formally elected, so they are less representative of the public. However, they are appointed by the central government and thus answer to the Interior Ministry, which is considered a higher level of government than local representatives. Descriptive representation denotes the representatives of a group who share characteristics with the represented electorates. Here again, unlike the failing leadership, a convened committee does not stem from the local community and does not share the same characteristics, in terms of local identity, ethnicity, minority group and culture. Nonetheless, convened committees are more likely than the failing leadership to commit themselves to the principles of good governance and managerialism. Given that convened committees are appointed by a higher authority that sees them as able to bring about quality management, they are expected to be better able to keep politics out of administration. In this sense, these practical democratic mechanisms share common values with the electorate who can identify with them, particularly during a local crisis. Examples of the public values, democratic mechanisms and characteristics that the represented electorates and the convened committee may share include a commitment to good governance, managerialism, the rule of law, transparency and equity (Beeri 2013; Beeri and Yuval 2015; Reingewertz and Beeri 2018).

Descriptive representation may lead to increased feelings of substantive and symbolic representation. Substantive representation includes the representatives’ opinions and actions that reflect the wishes, needs and interests of the people they represent. Given the evidence that convened committees are effective in helping municipalities stabilise the municipal budget and improve the quality of services provided (Beeri 2013; Reingewertz and Beeri 2018), they are likely to increase the public’s feelings that its wishes, needs and interests are heard, represented and acted upon. Symbolic representation refers to the representatives’ presence and the meaning they have for those they represent. Convened committees usually operate in neglected, peripheral, low socio-economic local authorities that have suffered an economic crisis. Therefore, the residents of these municipalities may regard the central government’s appointment of a committee to turn the local authority around as an indication that the central government cares about them. Placing a convened committee in charge of the local authority may be symbolic of the central government’s attempt to compensate these residents for the poor service they have received to date. They may see it as an effort to bring justice to a broken system that has harmed them. In such a case, imposing a convened committee on a local authority might improve the residents’ faith in the municipality and its legitimacy, reduce feelings of alienation and increase their participation in local democracy (Sanchez and Morin 2011).

Thus, we hypothesise that:

H1:

Within failing local authorities, after the intervention of the convened committee

local democracy will function better than before the intervention, such that:

H1a:

Within failing local authorities, after the intervention of the convened committee there will be more representativeness than before the intervention.

H1b:

Within failing local authorities, after the intervention of the convened committee the level of governability will be better than before the intervention.

H1c:

Within failing local authorities, after the intervention of the convened committee voter turnout will be higher than before the intervention.

Methods

Sampling local authorities and research population

To test these hypotheses, we used data from the 191 relevant Israeli cities and local councils.¹ We examined the variables dealing with democracy, representativeness, governability and voter turnout that emerged from the voting patterns in the local authority elections. We conducted pre- and post-analyses in all of the local authorities whose governments were replaced with convened committees between 2003 and 2018 (N = 31). These analyses compared attributes of local democracy before the convened committee was appointed to those after the convened committee completed its term of duty. In addition, we compared local authorities with convened committees to all of the other local authorities in Israel that served as a control group. The data were collected from the election results published on the Internet sites of the Ministry of the Interior and the Central Bureau of Statistics.

Research variables

There are many measures used to count parties and assess the electoral system with regard to democratic deficits. Some of these measures have been widely criticised (e.g., Dunleavy and Boucek 2003; Golosov 2010). Following Blau (2008), we weighted political parties alongside other methods that are relevant to the representativeness and governability of the mayor and his/her coalition. These measures enabled us to deepen our empirical examination of how the central government’s intervention using convened committees affects local democracy.

Effective Number of Council Parties by Seats at the Municipal Level (ENCPSML)

Extending the popular Effective Number of Parties index introduced by Laakso and Taagepera (1979), we developed the ENCPSML index (Effective Number of Council Parties by Seats at the Municipal Level) to measure representativeness at the local political level. The unit of analysis of the ENCPSML is local councils, meaning local parliaments in a given local election. The idea behind this measure is to count the number of political parties that won sufficient votes to gain a seat in the local parliament and weight this count according to the relative strength of these parties in their local councils.² By relative strength, we mean a party’s share of seats on the local council. For example, using this measure, we would say that a local council with 20 seats held by 5 parties is more representative than a local council with 20 seats held by 2 parties. Similarly, a local council with 20 seats held by 5 parties is more representative than a local council with 40 seats held by 5 parties. Finally, we are interested in the average score of a number of local councils on the ENCPML, not just a single council.

We calculated the ENCPSML using the following formula:

$\mathit{ENCPML}=\frac{1}{\bar{\chi }\mathit{p}{\mathrm{i}}^2}$

where pi is the proportion of political parties that won i seats on the local council and $\bar{x}$ is the average of a group of local councils’ pi proportions. Thus, a low score on the ENCPML indicates that the local councils are very representative and vice versa.

One should bear in mind that representativeness is a multidimensional concept (Pitkin 1967). In our research, it mirrors the unsolved tradeoff between giving voice to many groups, often at the expense of political stability and efficacy. Using the metric of the number of parties that won seats on the local council has advantages and disadvantages. The effective number of parties is a frequent means of operationalising the representation and fragmentation of a party system (Carey and Hix 2011). Numerous studies stress that voting might be more effective if coalitions included fewer, compact, coherent parties (Carey and Hix 2011; Tuttnauer and Friedman 2020; Warwick 2000). Our ENCPSML tool represents the operationalisation of the potential democratic deficits and tensions in a local council. Therefore, an extremely low score may indicate a relatively fragmented local government political system in which there are numerous political parties, designed to give everyone a voice. In contrast, an extremely high score reflects relatively limited representativeness that might be more efficient, accountable and can actually govern. We will discuss this apparent conundrum later.

Additional measures of local democratic representativeness, governability and voter turnout

Since a single measure cannot be used to explore a complex phenomenon such as local democracy (Kotzé and Steenekamp 2009), we added several more measures³:

Representativeness

• The number of candidates running for mayor, where a high value indicates that local politicians exercised the right to representation to a large extent.

• The number of parties running in the local elections, where a high value indicates that interest groups and sub-communities exercised the right to claim representation to a large extent.

• The number of parties that won seats on the local council, where a high value indicates that parties, interest groups and sub-communities exercised the right of representation to a large extent.

Governability

• The rate of support for the elected mayor, where a high value, calculated by the number of votes in favour of the chosen candidate divided by the number of valid ballots, indicates that the mayor will have more power to govern.

• The rate of support for the party of the elected mayor, where a high value, calculated by the number of votes cast in support of the elected mayor’s party divided by the number of valid ballots, indicates that the mayor’s party will have more power to govern.

• The rate of support for the largest party, where a high value, calculated by the number of votes cast in support of the largest party divided by the number of valid ballots, indicates that the largest political party will have more power to govern.

• The number of cases of runoff elections for mayor, where a high value, calculated by the number of runoff elections divided by the total number of local authorities that hold elections, indicates a political stalemate that will result in a lack of governability.

• The rates at which incumbent mayors ran for and won re-election, where high values indicate a high level of governability because the re-election of incumbents preserves continuity and stability.

Voter turnout

• Voter turnout in the local elections, where a high value, calculated by the number of valid ballots divided by the number of eligible voters, indicates that voters exercised their right to vote to a large extent.

Findings

The structural attributes of the local government authorities that participated in the study are depicted in Table 1.

Table 1. Summary statistics, Israeli municipalities, 2003–2018.

			Mean	St. dev	Min	Max
All local authorities (N = 191):
Socioeconomic cluster			4.7	2.4	1	10
Peripheral cluster			5	1.9	1	10
Number of council members			12.8	5.1	7	31
Number of residents (K)			38.2	81.4	1.2	883
Number of eligible voters (K)			31.4	67	0.9	638
Local authorities in which a convened committee was appointed (N = 31)
Socioeconomic cluster			3.3	1.3	1	8
Peripheral cluster			4.1	1.3	2	7
Number of council members			11.3	2.9	7	19
Number of residents (K)			16.6	14.3	1.4	74
Number of eligible voters (K)			11.6	11.1	1.3	57

Representation

Effective Number of Council Parties by Seats at the Municipal Level (ENCPSML)

H1a posited that within failing local authorities, after the intervention of the convened committee, the level of representativeness would be greater than before the intervention. To test this hypothesis, we employed bootstrap-assisted paired sample t-tests (5000 samples) that revealed a significant difference (t = 3.53; p < .01; CILL = 1.32, CIUP = 4.31) before and five years after the convened committee’s term. Five years after the convened committee’s term, the ENCPSML score (mean = 1.69, SD=.77) was significantly lower than before the appointment of the convened committee (mean = 4.37, SD = 4.54), reflecting greater representativeness. In a similar vein, we found a significant difference (t = 3.55; p < .001; CILL = 1.08, CIUP = 3.36) between the two election cycles that took place after the convened committee’s term. Five years after the convened committee’s term, the ENCPSML score (mean = 1.69, SD=.77) was significantly lower than right after the convened committee’s term ended (mean = 3.77, SD = 3.80), reflecting greater representativeness. Finally, there was no significant difference in the ENCPSML score before the convened committee’s term (mean = 4.37, SD = 4.54) and right after it ended (mean = 3.77, SD = 3.91) (t = 1.19; p=NS; CILL=−.31, CIUP = 1.50). Figure 1 illustrates the between-group comparison of the ENCPSML levels over time.

Figure 1. Effective number of council parties by seats (ENCPML).

Figure 1 also indicates that the ENCPSML scores of the local authorities in which a convened committee was appointed were significantly lower (mean = 4.37, SD = 4.54) than those in which a convened committee was never appointed (mean = 7.00, SD = 10.31) (t = 2.21; p < .05; CILL=.33, CIUP = 4.93), reflecting much greater representativeness.

Additional measures of representativeness

Figure 2 reveals mild fluctuations in the number of mayoral candidates in authorities in which a convened committee was appointed. Before the appointment of such committees, the average number of candidates was 3.9. After the convened committee’s term, this number rose slightly to 4.1, while in the elections five years later the number decreased to 3.5. In other words, the presence of the committee did not hurt the degree to which local politicians and leaders exercised their right to represent the public. The average number of candidates running in all of the local authorities was similar: 3.9, 4.1, 3.8 and 3.5 in 2003, 2008, 2013 and 2018, respectively.

Figure 2. Number of mayoral candidates.

As Figure 3 indicates, the number of parties running in the local authorities in which a convened committee had been appointed generally remained stable. Before the appointment of such committees, the average number of parties was 10.8. After the convened committee’s term, this number rose slightly, to 11.4, but in the elections five years later, it dropped again to 10.6. In other words, the presence of the committee did not hurt the degree to which interest groups and sub-communities claimed their right to represent the public through parties. The average number of parties running in all of the local authorities was similar: 10.6, 10, 9.9 and 97 in 2003, 2008, 2013 and 2018, respectively.

Figure 3. Number of parties running.

Figure 4 shows that among the local authorities in which a convened committee was appointed, the number of parties that exceeded the threshold and gained representation on the local council remained stable and even rose a bit. Before the appointment of the convened committee, the average number of parties was 6.9. After the convened committee’s term, this average rose a bit, to 7.2, and in the elections five years later it dropped slightly, to 7. In other words, the presence of the committee did not hurt the degree to which political parties, interest groups and sub-communities exercised their right to representation. The average number of parties that exceeded the threshold and gained representation on the council in all of the local authorities was a bit lower: 6.6, 6.6, 6.5 and 6.7 in 2003, 2008, 2013 and 2018, respectively.

Figure 4. Number of parties that won seats.

Governability

H1b posited that within failing local authorities, after the intervention of the convened committee, the level of governability would be greater than before the intervention. As Figure 5 illustrates, among the local authorities in which a convened committee was appointed, the rate of support for the elected mayor generally remained stable and even rose slightly. Before the convened committee’s appointment, the average rate of support was 46%. In the elections immediately after the convened committee’s term, this figure dropped slightly to 44%, but rose by 3% to 49% in the elections five years after the convened committee’s term. In other words, the ability of the mayor to govern appeared to increase after the convened committee’s term. The rates of support for the elected mayor were somewhat higher in all of the local authorities: 51%, 52%, 53% and 53% in 2003, 2008, 2013 and 2018, respectively.

Figure 5. Votes for Mayor.

Figure 6 shows that among the local authorities in which a convened committee was appointed, the rate of support for the party of the elected mayor generally remained stable and in some cases even rose. Before the convened committee’s appointment, the average rate of support was 14%. After the convened committee’s term, this rate dropped slightly, to 13%, but rose by 3% in the elections five years after the convened committee’s term, to 16%. In other words, the potential power to govern given to the elected mayor’s party was slightly strengthened after the convened committee’s term. In all of the local authorities, the rates of support for the party of the elected mayor were higher, indicating somewhat higher rates of governability: 19%, 22%, 22% and 23% in 2003, 2008, 2013 and 2018, respectively.

Figure 6. Votes for Mayor’s party.

As Figure 7 indicates, the rate of support for the largest party in the council remained stable or even rose slightly among the local authorities in which a convened committee was appointed. Before the convened committee’s appointment, the average rate of support was 24%. This rate dropped slightly to 22% after the convened committee’s term, but rose by 3% five years after the convened committee term, to 25%. In other words, the potential power to govern given to the largest party on the council increased slightly. In all of the local authorities, the rates of support for the largest party demonstrated greater fluctuation, reflecting a somewhat higher degree of governability: 27%, 29%, 13% and 30% in 2003, 2008, 2013 and 2018, respectively.

Figure 7. Votes for the largest party.

Figure 8 illustrates that among the local councils in which a convened committee was appointed, the rate of runoff elections (indicating a clear lack of decision and hence a crack in governability) decreased drastically. Before the convened committee’s appointment, the average runoff election rate was 31%. After the convened committee’s term, this rate remained 31%, but in the election five years after the convened committee’s term, this rate dropped by 8%, to a total of 23%. In other words, the number of tied elections in which there was no clear winner dropped after the convened committee’s term, while the potential governability rose. In all of the local authorities, the runoff rates demonstrated greater fluctuation. In general, the rate was lower, pointing to greater governability: 28%, 19%, 20% and 25% in 2003, 2008, 2013 and 2018, respectively.

Figure 8. Rates of Mayoral runoff elections.

Figure 9 indicates that among the local authorities in which a convened committee was appointed, the percentage of incumbent mayors who ran for and won an additional term rose significantly after the convened committee’s term. Immediately after the convened committee’s term, only 38% of the mayors who had served prior to the convened committee’s term ran for an additional term and only 12% of them won. In contrast, in the elections held five years after the convened committee’s term, 77% of the incumbent mayors ran for an additional term and 50% of them won. In other words, increased continuity and stability point to increased trust in the local leadership and its power to govern. The rate of candidates who won the elections five years after the convened committee’s term was similar in all of the local authorities: 72% ran and 46% won in 2008, 65% ran and 60% won in 2013 and 80% ran and 52% won in 2018.

Figure 9. Rates of incumbent Mayors who ran for and won an additional term.

Voter turnout

H1c posited that within failing local authorities, after the intervention of the convened committee, the level of voter turnout would be higher than before the intervention. As Figure 10 indicates, the degree of political participation in local authorities in which convened committees were appointed generally remained stable after the appointment of the convened committee. Before the committee was appointed, the average voter turnout was 83%, while after the convened committee’s term it dropped a bit, to 81%. Yet five years after the convened committee’s term, voter turnout returned to its original level of 84%. In other words, the presence of the committee did not hurt people’s democratic right to vote. Indeed, political participation in all local elections dropped during 2003, 2008 and 2013, but rose in 2018: 69%, 69%, 51% and 73%, respectively.⁴

Figure 10. Voter turnout in local elections.

In summarising the findings in this section, one can claim that among the local authorities in which a convened committee was appointed, democratic deficits did not increase. The convened committee’s term did not damage the degree to which democracy functioned in these municipalities when measured by representativeness, governability and voter turnout. Indeed, five years after the convened committee’s term ended, these measures were actually slightly higher. Compared with all local authorities, representativeness and voter turnout in municipalities in which convened committees were appointed were a bit higher and governability was lower.

Discussion and conclusions

In this research we examined how top-down central government intervention in local crises by neutralising the local leadership affects deficits in local democracy. We did so to answer the call to explain ‘democratic disenchantment’, evident in civic disengagement and apathy, and declines in voter turnout, party membership, and participation in community activities and public affairs (Skelcher and Torfing 2010). To achieve this goal, we developed the ENCPSML index to quantify the degree of representativeness, governability and voter turnout in local authorities and used additional measures to assess the impact of neutralising the local leadership on democracy. Hence, future research can use this index to measure local democracy and representativeness, compare local government systems in different countries or over time, or while controlling for numerous intervening and structural variables.

Our second contribution is in demonstrating empirically that the governability of the local authorities in which a convened committee was appointed was significantly worse than in other local authorities as a whole. In retrospect, this finding to some extent explains the functional and economic crises that these authorities faced, which in some cases apparently stemmed from the limitations and weakness of local democracy. Both before and after the convened committee’s term, there was less support for the elected mayor, the mayor’s party and the largest party relative to other municipalities. Therefore, in many cases the first round of elections led to no clear decision, necessitating more runoff elections. In neutralised authorities, such a situation was probably inevitable, leading to more fragile coalitions comprised of a large number of parties that might felt less obligated to the coalition. In general, while the rates of voter turnout in the neutralised authorities were higher, indicative of political participation, the degree of representativeness was much higher, pointing to greater fragmentation. The number of parties that won enough votes to have a seat on the council – according to the ENCPSML – was much higher, indicating more small parties vying for political power. Hence, our empirical findings offer a cautious support for the central government’s intervention policy and the neutralisation approach adopted in this case.

Third, a series of scholars (e.g., Bache, Bartle, and Flinders 2016; Hupe and Edwards 2012; Pratchett 2004; Papadopoulos 2005; Sørensen and Torfing 2005; Stoker 2004) have expressed concern that multi-level governance and the neutralisation approach would trigger democratic deficits, complexity, dispersion and fragmentation. However, our results seem to indicate that such is not necessarily the case. Indeed, the opposite is closer to the truth. Given that the convened committees are more specialised and professional than any individual player, and have no political aspirations (Hooghe and Marks 2004; Peters and Pierre 2004), they can help turn around the municipality without causing undue damage to local democracy. Our empirical findings indicate that the degree of representativeness was not damaged after the term of an appointed convened committee. On the contrary, and particularly in the five years after the convened committee’s term, the number of represented parties according to the ENCPSML index increased. In tandem, there was probably a more fragmented political system with numerous parties trying to give everyone a voice. In a similar vein, convened committees neither hurt the representativeness of the candidates and political parties who competed for or won seats, nor the ability of the elected mayor, the mayor’s party and the largest party to govern, all of which increased slightly. One might claim that, even though the convened committees ensured that the public was still represented, they did not create the political stability that is potentially inherent in a smaller number of large political parties. However, since the core turnaround process had already been accomplished, it might be reasonable to prefer representation to stability. At the same time, the number of cases requiring runoff elections dropped and the continuity of the government increased, as did political participation, evident in voter turnout.

In other words, a few years after replacing the elected leadership with a convened committee, we see a stable, functioning local democracy, slightly better signs of greater representativeness in local democratic institutions, a leadership that is better able to govern and a local community that is more politically involved. On the other hand, we also agree with Papadopoulos (2005) that the appointment of convened committees may result in side effects and bring about new complexities, such as increased political fragmentation in the form of numerous small parties. Hence, similar to Carey and Hix (2011), we contend that such fragmented electoral systems might result in unmanageable coalitions of relatively weak players. Hence, the central government’s intervention in local affairs is imperfect. It is still struggling to find a comprehensive solution to the gaps between different local authorities in economic, functional and democratic performance, while the latter seek more authority and autonomy.

From the ideological perspective, this research lends a certain degree of support to the appointment of convened committees and the intervention of the central government in the case of local functional and economic crises. Given that we found no evidence that convened committees weaken local democracy to any great extent, the central government could use the neutralisation approach to deal with other local and urban crises related to climate change, hunger, social gaps, immigration, pandemics and sustainability. Thus, we can conclude that, at least in the Israeli case and probably other Western democratic countries, the democratic cost of central government intervention by appointing convened committees is not unreasonable. They offer local authorities the ability to turn around, without hurting local democracy.

As in every research study, this study has its limitations. First, it is rooted in the context of the relationship between the central and local governments in Israel, which is traditionally highly centralised. The extensive use of convened committees and the circumstances that allowed the use of this government intervention tool narrow the ability to generalise the findings to other cases. However, it can still be seen as an example of the intervention of the central government in failing municipalities.

Another limitation is the duality of the measures. An increase or decrease in the different measures does not necessarily point to a clear trend that unequivocally affects the elected or appointed leadership. In addition, the ENCPSML index may be sensitive to external factors and trends. As is the case with wicked problems, the interpretation of the empirical findings can also be a double-edged sword. For instance, while the ENCPSML index points to the fragmentation of small parties as a sign of democratic deficit, at the same time it may reflect substantive and symbolic representation that in some circumstances lead to political stability and democratisation. Thus, the findings and conclusions should be viewed with caution and are probably more relevant to the process of turning around the management of an organisation.

The final limitation of the measures is that they are based on one specific electoral system. This system includes a set of rules that might not apply in other situations. For example, in Israel the mayor must win 40% of the popular vote, and there are two tiers of government with no regional elections.

Hence, future research may take several directions. First, it should consider the neutralisation approach and the appointment of convened committees as independent variables in a research design that measures the functional, economic and democratic influences of these variables as well. Second, future research can compare several other regulations and central government intervention tools to turn around municipal crises. Moreover, our research design and tools should be investigated in other countries and types of local and urban crises, including environmental ones. Last but not least, future studies can expand the research question and the study’s design to explore the effect of the central government’s intervention in the case of a local crisis in a field other than local democracy, such as the local economy, urbanism or local identity.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. We omitted regional councils because they have a different political system that consists of two layers based on the polling areas of the towns’ boundaries. Elections for regional councils usually take place on different dates than those for cities and local councils.

2. The measure normalises the size of the local councils in terms of the number of residents and number of seats.

3. In Israel, local elections are held every five years. In these elections, voters cast one vote for the party list, which is allocated seats on the council relative to the amount of support it receives, and one vote for the mayoral candidate. To be elected mayor, a candidate must receive 40% of all the valid ballots. If no candidate receives 40% of the votes, a runoff election between the two candidates with the most votes is held two weeks later.

4. The relatively large rise in political participation in 2018 can also be attributed to the fact that Election Day for the local elections became a statutory holiday for the first time that year.