Hubert Heinelt Darmstadt University of Technology Institute of Political Science.

Transcription

1 Hubert Heinelt Darmstadt University of Technology Institute of Political Science What about the Council (and Councillors)? Considerations based on surveys with mayors and municipal chief executives Paper to be presented at Workshop 3 Local Political Leadership in a Changing Context ECPR Joint Sessions in Nicosia, April 2006* * I acknowledge the support of Björn Egner (Institute of Political Science at Darmstadt University of Technology) for helping me in struggling with the data, and of the Istanbul Policy Centre at the Sabanci University for providing me with the opportunity to work on this paper (and others) in a stimulating atmosphere during a stay there as a visiting fellow in March 2006.

2 2 1. Introduction Political leadership at the local level has attracted growing attention in recent years in parallel with reforms of local government (as the core of representative democracy at the local level) and of the municipal addministration (according to different versions and applications of the New Public Management approach) as well as the debate on a shift from government to governance (see e.g. Le Galés 2000; John 2001; Kersting/Vetter 2003; Haus et al. 2005; Reynaert et al. 2005). But this debate is mainly focused on single leaders, i.e. mayors or executive officers. Considering the power triangle of (i) the mayor, (ii) municipal administration (executive officers) and (iii) the council it is surprising that the latter has gain little interest so far. This paper will look at the role of councils in a comparative way by referring to a survey on municipal chief executives (the U.DI.T.E survey; see Klausen/Magnier 1998 and Mouritzen/Svara 2002) 1 and another survey on mayors (the POLLEADER survey; see Bäck et al. 2006) in several European countries. 2 The collected data do not only give answers to the question how municipal chief executives and mayors are perceiving and assessing the functions of councils in different European countries. These surveys gathered also a lot of information about the specific political context of the respondents like majority in the council, frequency of communication etc. 3 The paper will not only try to explain differences in the perceived and assessed functions of councils by referring to typologies of local government systems. To look at (country) specific contextual condition is without any doubt important to explain differences. But institutional arrangements are not independent variables which directly determine the role of the mayors, municipal executive officers and the councils. It is crucial how the given institutionally determined feasible set is perceived, how opportunities are used and institutional conditions are reshaped by actors. Such considerations are linked to a specific approach namely actorcentred (neo-)institutionalism and emphasise the impact of specific local actor constellations. They may be influenced by institutional settings (local government systems) and related power relations (e.g. characterised by a strong political or executive mayor). But how mayors, councillors and municipal chief executive officers are actually interacting (e.g. cooperatively or not) may be related to particular local circumstances as well. 1 From the U.DI.T.E survey just the countries have been included which are also covered by the POLLEADER study. 2 The survey on mayors was conducted in 2003 and It is based on a questionnaire sent out to all mayors of towns with more than 10,000 inhabitants in Sweden, Denmark, England, Ireland, France, Belgium, the Netherlands, Germany, Poland, the Czech Republic, Austria, Switzerland, Hungary, Portugal, Spain, Italy and Greece. A first publication of comparative results based on the whole sample is envisaged for 2006 (Bäck et al. 2006). The survey on chief executive officers, i.e. the highest ranging appointed officers, was conducted (with the exception of Germany; see below) from 1995 to This survey (the so-called U.Di.T.E. Leadership Study) covered Belgium, Denmark, England, Finland, France, Ireland, Italy, the Netherlands, Norway, Portugal, Spain, Sweden as well as Australia and the USA. Results of this study has been published in Klausen/Magnier 1998, Dahler-Larsen 2002 and Mouritzen/Svara Because the survey in Germany was undertaken (by Oscar W. Gabriel and Hubert Heinelt) quite late (in autumn 2000) the results for this country have not been included in these three publications. For the main German results see Haus/Heinelt For inter-country comparison data have been weighted to compensate for varying national response rates. This means that inferences are made to the total population of European mayors and municipal chief executives (restricted to the participating countries).

3 3 2 Defining the variables 2.1 The dependent variable: How is the council perceived? The role the council is playing under different settings in European cities is hardly to determine or to ascertain by policy outputs or outcomes. What is possible on the basis of the mentioned survey is to take the perceptions of mayors and municipal chief executive officers about the role councils are playing as an indication. Within the POLLEADER survey mayors were asked How would you rate the overall effectiveness of your city council? The answers to this question by countries are show in the following table ( excellent = 4, good = 3, average = 2 and poor = 1). Tables 1: The perception of mayors on the overall effectiveness of their city council COUNTRY means Standard deviation NETHERLANDS SWITZERLAND GERMANY ITALY BELGIUM SWEDEN FRANCE POLAND HUNGARY SPAIN CZECH REPUBLIC DENMARK PORTUGAL AUSTRIA GREECE ENGLAND IRELAND TOTAL Based on data from the POLLEADER survey. N=2,632 Furthermore, mayors were asked how they would define the actual contribution of the local assembly on the definition of the main goals of the municipal activity, the control of the municipal activity, the representation of requests and issues emerging from local community and the publicising of debate on local issues. The responses to these topics can be linked to different functions or notions of representation exercised by the council, namely (see Denters 2006) the internal or inward oriented role of taking decisions about the (main) goals of the municipality and controlling its activities, i.e. the combination of the two first topics, the external or outward oriented role of articulating demands from the local community and of making debates on local issues public, i.e. the combination of topics three and four.

4 4 The results are shown in Table 2 (means of a scale ranging from 5 = very important contribution to 1 = very negative influence ). Tables 2: COUNTRY The contribution of city council on different aspects of representation (means by countries) Defining the main goals of the municipal activity Controlling the municipal activity Internal role Representing the requests and issues emerging from local community Publicising the debate on local issues External role AUSTRIA BELGIUM DENMARK ENGLAND FRANCE GERMANY GREECE HUNGARY IRELAND ITALY NETHERL POLAND PORTUGAL SPAIN SWEDEN SWITZERL CZECH R N 2,653 2, * 2,641 2, ** Based on data from the POLLEADER survey. * Standard deviation = 0, ** Standard deviation = 0, Tables 3: Responses of municipal chief executive offices on the statement that it is an advantage if the chief executive is of the same political opinion as the majority of council (means by countries) Country means standard deviation FRANCE PORTUGAL GERMANY BELGIUM SPAIN NETHERL ITALY ENGLAND SWEDEN DENMARK IRELAND TOTAL Based on data from the U.DI.T.E survey., N=3,176

5 5 The U.DI.T.E survey did not include such questions which address more or less directly the municipal chief executive officers perceptions of their council. However, the following questions was asked which tackle the importance of the council - or more specifically: the importance of the majority of the council. The executive officers were asked about their relations with politicians and whether they agree or disagree with the statement that it is an advantage if the chief executive is of the same political opinion as the majority of the local council. Table 3 gives an overview about the responses by countries (means of a scale ranging from 5 = strongly disagree to 1 = strongly agree ). In the U.DI.T.E survey municipal chief executive officers were also asked which actors do have influence (from great influence = 1 to non influence = 5) on decisions (i) on the budget and (ii) on economic development. Budget decisions can be regarded as the core domain of a representative body like the council. This maybe not apply for decisions on economic development but they are crucial for the economic wellbeing of the city or the citizens. Therefore, answers to these questions can shed light on the influence of a council. The influence of the council have been measured in relative terms. The mean value for the majority in the council (in this case explicitly not only of one party) has been related to those of (i) the mayor and (ii) the leading municipal executive officer. The results are shown in Table 4. Table 4: Country Influence of the majority in the council on decision about the budget and economic development in relation to the mayor and leading municipal executive officer by countries (differences of means; scale: 1 = big influence, 5= no influencer) decisions about the budget Power distance council majority mayor Power distance council majority leading municipal executive officer decisions about economic development Power distance council majority mayor Power distance council majority leading municipal executive officer PORTUGAL BELGIUM ITALY FRANCE SWEDEN DENMARK SPAIN GERMANY ENGLAND NETHERL IRELAND TOTAL Std. Dev. 1,253 1,378 1,149 1,262 N 3,070 3,079 2,094 2,097 Eta Based on data from the U.DI.T.E survey.

6 6 In these table the average power distance per country are given. The values result out of subtracting the values of the council from that of the mayor respectively the values of the council from that of the chief executive officer (each on a scale from 1 = big influence to 5 = no influence). To give examples: If the influence of the council is perceived as big (= 1) and the mayor is perceived as an actor with no influence (= 5) the result is 4 (i.e. 5 minus 1). If the influence of the mayor is perceived as big and the council is perceived as a body with no influence the result is 4 (i.e. 1 minus 5). Therefore, the values indicate the power distance seen from the council. Taking the results for England (see Table 4) in respect to budget decisions as an example, the council is just slightly more influential than the mayor (by 0.13) and a bit more influential than the chief executive officer (0.45). Remarkably (or not; see the following considerations on local government systems), the power distance between mayors and the councils in respect to budget decisions is just in two countries in favour of the council namely in England and Ireland. These two countries stand in contrast to Portugal, Italy, France and Belgium where the mayor is clearly indicated as more influential in budget decisions than the council. The pattern for the power distance between the council and the municipal chief executive officer in respect to decisions on the municipal budget points to strong chief executive officers (city managers) in Ireland and influential councils in Sweden, Denmark and England (which can be clearly related to the typologies on local government systems mentioned below) as well as in Germany and Spain. In respect to decisions on economic development the power distance between the councils and the mayors is in any country more in favour of the mayor than in relation to decisions on the budget. The same applies (except Denmark) for the power distances between the councils and the city managers (i.e. the latter are also more influential in decision on economic development than on budget decisions). An obvious interpretation for this is that councils are in general not so influential in operational (but nevertheless important) decisions like those on economic development than on budget decisions as a core function of representative political bodies. 2.2 The independent variables: How to conceptualise the relations between the council, the mayor and the administrative executive? There are different typologies of power relations between the council, the mayor and the administrative executive. In the follow just those will be considered which promise the most relevant insights Layman rule, political leadership and professionalism Mouritzen and Svara (2002) offer a typology of local government systems oriented explicitly to the power relations between the council, the mayor and the administrative executive. 4 Their considerations rely on the following hypothesis: The structural features of municipal government in any specific country reflect a balance or compromise among [ ] three organizing principles: layman rule, political leadership, and professionalism (Mouritzen and Svara 2002: 50-51). 5 4 The following two sub-section are strongly based on Heinelt and Hlepas 2006: Whereas the layman rule means that citizens elected for political office should be involved effectively and intensively in the making of decisions (Mouritzen and Svara 2002: 51), the notion of political leadership implies the concept of politicians promoting value choices and feeding energy and passion into policy systems (Mouritzen and Svara 2002: 52); finally, professionalism rests on the crucial distinction that as politicians respond to demands, professionals respond to and seek to address needs (Mouritzen and Svara 2002: 53; referring to Svara 1990).

7 7 Although Mouritzen and Svara consider all three elements, political leadership is the starting point for the development of a typology of government forms. The key issue is how political power is obtained, maintained, exercised, and shared. [ ] Political power is a function of the degree of control a political actor a person or a collective body has in two arenas. First, to what extent is the city council controlled by one or more political actors? The second arena is the executive, and the question is to what extent is control over the executive in the hands of one or more political actors. Formal structure is important to answering these questions, but so are informal institutional rules and norms (Mouritzen and Svara 2002: 53). Based on the these considerations Mouritzen and Svara distinguish four ideal types: The strong mayor form: The elected mayor controls the majority of the city council and is legally and in actuality in full charge of all executive functions (Mouritzen and Svara 2002: 55). The committee-leader form: One person is clearly the political leader of the municipality with or without the title of mayor. He may or may not control the council. Executive powers are shared. The political leader may have responsibility for some executive functions but others will rest with collegiate bodies, that is, standing committees composed of elected politicians, and with the CEO (Mouritzen and Svara 2002: 56). The collective form: The decision center is one collegiate body, the executive committee that is responsible for all executive functions. The executive committee consists of locally elected politicians and the mayor, who presides (Mouritzen and Svara 2002: 56). The council-manager form: All executive functions are in the hands of a professional administrative the city manager who is appointed by the city council, which has general authority over policy but is restricted from involvement in administrative matters. The council is [ ] headed by a mayor who formally has presiding and ceremonial functions only (Mouritzen and Svara 2002: 56). Although Mouritzen and Svara (2002: 53) reflect not only formal structures but also informal institutional rules and norms, their approach is oriented to ideal types of national local government systems applicable to particular countries. 6 Although Mouritzen/Svara s typology was published relatively recently it does not capture changes effected in the last few years and more importantly not all the countries included in the POLLEADER-study are considered (even Germany is missing). Therefore, the national teams involved in the POLLEADER-study from the countries not included in Mouritzen/Svara s typology subsumed their countries under the four types of this typology. The results of the information are summarised in Table 5 (see below) and lead to the grouping of 6 Depending on the research question, this can be an advantages in comparison to Bäck s typology (outlined in sub-section 2.2.3) because his distinction between majoritarian and consociational forms of democratic local decision-making is more focused on locally defined informal rules of the game (whereas a monistic or dualistic form of local government can be see as a given formal institutional structure), so that his typology is only partly related to country-specific institutional structures. Nevertheless, Bäck (2005: 82-83) acknowledges that his assembly government is very close to what Mouritzen and Svara term the committee leader form and that Mouritzen and Svara probably would classify also the parliamentary system as a committee leader form. Furthermore, he stated that semipresidentialism and the presidential system would be classified as strong mayor forms by Mouritzen and Svara. What is not covered by Bäck s typology is the council-manager form and the collective form of Mouritzen and Svara. But may be both could be subsumed under parliamentarism. See his classification of individual cities of countries which belong according to Mouritzen and Svara to the council-manager form and the collective form in Bäck 2005: 87.

8 8 France, Germany (without the federate state/land of Hesse), Austria (with six of its nine Länder), Greece, Hungary, Italy, Poland, Portugal, Spain and the English cases with a directly elected mayor under the strong mayor form, Denmark, the English cases of alternative arrangements 7 and Sweden (as well as the remaining three Austrian Länder) under the committee leader form, Belgium, the Czech Republic, the Netherlands, Switzerland, and the English leadercabinet model cases, (as well as the German Land of Hesse) under the collective form, and Ireland as the only country of the council-manager form of local government systems 8. These groupings will be used in the following chapters when the relations between the mayor, the council and the municipal administration are addressed by referring to the Mouritzen and Svara typology Horizontal power relations in the municipality and the functions and competencies of the local government level In examining the role of mayors in a particular local government system, it is not only their relations to the council and the municipal administration that have to be considered. The vertical distribution of functions and competencies between the local level and upper-levels of government also have to be reflected. This is of particular significance for determining whether a mayor is called upon to represent and/or to lead a municipality with a broad or a quite restricted spectrum of competencies and responsibilities (e.g. in the field of service provision) as well as with limited or wider fiscal and financial discretion. Furthermore, considering the task of a mayor together with the municipal s legal and financial capacity to govern local affairs is of notable importance against the background of the often cited shift from government to governance and the (possible) ensuing challenge for urban leadership in (newly) evolving local governance arrangements (see Borraz and John 2004; Haus et al. 2005; Heinelt et al. 2005). Therefore, within a typology of local political leader developed by the POLLEADER project vertical and horizontal dimension are combined by considering the Hesse and Sharpe typology (1991) for local-central government relations 9 and that of Mouritzen and Svara (2002) for power relations between the mayor, the council and the municipal administration. The individual countries covered by the POLLEADER project are placed in different boxes of Table 5 either on the classification of the cited authors or on the information of the partners from the respective countries involved in the project For the changes in the English local government system see Leach and Wilson 2004 and the remarks below. 8 One English council also has this form of decision-making. They did not respond to the survey and are not included in this analysis. 9 For the strengths of the Hesse and Sharpe typology in comparison with those of Bennett (1989; 1993a; 1993b), Page and Goldsmith (1987), Page (1991), Goldsmith (1993) and later John (2001) as well as Loughlin (2001); see Heinelt and Hlepas (2006: 21-29). 10 As shown in Table 5 the three Eastern European countries covered by the POLEADER study, i.e. Hungary, Poland and the Czech Republic, have not been subsumed under one of the three types the Hesse and Sharpe typology, but rather classified as a separate group: the Central East European type. Although local-central relations in Hungary, Poland and the Czech Republic have some features in common with the North and Middle European group with respect to local competencies and fiscal or financial discretion of local governments, they have been considered as a distinct group because their historical background and, in particular, the quite recent (radical) decentralisation in these countries needs to be taken into account in discussion of vertical power relations. These power relations are (as the different schools of neo-institutionalism emphasise) not just characterised by certain formal (legal) rules for the distribution of

9 9 Table 5: forms of local government systems according to horizontal power relations (Mouritzen and Svara 2002) (i) (ii) (iii) (iv) (v) (vi) Political leadership types according to different local government systems Strong mayor form committee leader form collective form types of local government systems according to (vertical) local-central relations (Hesse and Sharpe 1991) Franco type Anglo type North-Middle European type political mayor France, Greece, Italy, Portugal, Spain collegial leader Belgium councilmanager form In the cases with a directly elected mayor. executive mayor England (i) collegial leader England (iv) collegial leader; England (vi) ceremonial mayor Ireland executive mayor Germany, (ii) Austria (iii) collegial leader Denmark, Sweden, (Austria) (v) collegial leader Netherlands, Switzerland, (Germany) Without the Bundesland Hesse which is subsumed under the group of collegial leaders. In six of its nine Bundesländer. In the alternative arrangements cases. In three of its nine Bundesländer. In the cases with a leader-cabinet model. Central-East European type executive mayor Poland, Hungary collegial leader Czech Rep. The strong mayor type (according to Mouritzen and Svara ) is found in countries of the Franco and the North-Middle European (according to Hesse and Sharpe) as well as in the Central-East European type of local government systems. In addition, the following distinctions can be drawn between mayors of these types of local government systems: Due to the fact that mayors in the North-Middle and Central-East European types of local government systems are not only formally the heads of municipal administrations which hold responsibility for a broad spectrum of public provision but are also in full charge of their administrations, these mayors will here be called executive mayors. This applies to mayors in Hungary, Poland and Germany (with the exception of the Land Hesse; see below) and most of the Austrian mayors (i.e. the directly elected mayors in Austria). Because strong mayors in the Franco type of local government systems lead a municipal administration that is responsible for a relatively limited scope of state functions, but are nevertheless clearly the political representative (and agent) for the local community, they will be called political mayors. France, Greece, Italy, Portugal and Spain are included in this type. competencies and resources, but also by particular meaning systems of what is perceived as appropriate or inappropriate.

10 10 Under the Anglo type of local government systems the council-manager type is found in Ireland. In this case mayors exercise a mainly ceremonial function/role, while the municipal administration is directed by a professional manager. Therefore, Irish mayors are here termed ceremonial mayors. The North-Middle European type of local government systems covers not only the strong mayor and the council-manager form, but also the committee leader and the collective type according to the Mouritzen/Svara typology. In several North- and Middle European countries mayors or elected local leaders of the municipality without the official title of a Mayor (in Sweden, where no Mayor exists, and in England, where most mayors see below have restricted ceremonial functions) are required to co-operate collegially with other powerful actors or bodies. Therefore, these institutional settings offer room only for collegial leaders. This type of mayor can be found in Denmark, Sweden and partly in Austria (i.e. in the three Länder where the mayors are not directly elected) as well as in the Netherlands, Switzerland, the Czech Republic and, due to its Magistratsverfassung, in the German Land Hesse. Although local competencies for service provision in Belgium are different from the above mentioned countries of the North-Middle European type, Belgian mayors can also be subsumed under the collegial leaders because of the collective form of local government in this country. In England the Local Government Act (Hambleton and Sweeting 2004; Sweeting 2003) offers councils four options. 11 Because the leader-cabinet model closest to the collective form is currently the most widespread, collegial leaders are also dominant in England. The distinction between executive mayors, political mayors, ceremonial mayors and collegial leaders will be used in addition to the Mouritzen/Svara typology (i.e. the horizontal power relation) to see whether or not institutional structures including the allocation of functions and competencies at the local level make a difference in the mayors perception of the role and performance of the council Dualistic and monistic local government systems and consociationalism and majoritarianism A common general distinction reflects the legally defined distributions of tasks between legislative and executive functions. Whereas in dualistic systems the two functions are separate, they are merged in monistic systems. Or more precisely: In the monistic type of local government the elected local council is regarded as the (sole) supreme decision-making body, while the local administration, including its head/chief executive, acts under the instruction and scrutiny of the council without any autonomous executive decision-making power of its own. In a dualistic system the elected council is recognized as the prime decisionmaking body of local government, but the head/chief executive of the local administration is seen as possessing some ( executive ) decision-making powers of his/her own that are not derived from the local council (Wollmann 2004: 151). This typology can be further refined to explore additional distinctions: (i) the role of the mayor exercising or not the executive 11 These are: (i) directly elected mayor and cabinet (the mayor-cabinet model adopted by 10 councils); (ii) directly elected mayor and council manager (the mayor-council manager model which has been adopted in one case not included in the POLLEADER data set); (iii) leader and cabinet (the leader-cabinet model closest to the collective form, 316 councils opted for this model); (iv) alternative arrangements (available as an option only for councils with less than 85,000 inhabitants, 59 councils opted for this model, which is closest to the committee leader form).

11 11 function, (ii) exercising the executive function alone or together with a collective or collegiate body, and (iii) the form through which the mayor is elected (see Wollmann 2004: ). 12 Drawing on traditional ideal types in comparative constitutional analysis (of national government) and combining the distinction between (i) a monistic versus dualistic organisation with (ii) a majoritarian versus consociational form of democratic local decisionmaking Bäck (2005: 82-83) proposed the following typology: Assembly Government: represented by situations where executive power is in the hands of a proportionally composed committee of the council, i.e. monism in combination with consociationalism. Parliamentarism: the combination of monism with majoritarianism, i.e. situations featuring a collective executive appointed by the council and headed by a leader (sometimes but called the mayor; see below) not using proportional techniques but some variation on the majority principle. Presidentialism: a separately elected mayor, appointing his/her own cabinet of deputies (and taking decisions) without consideration of the party-political composition of the council. In this form the dualistic and majoritarian principles are combined. Semi-presidentialism: the mayor is surrounded by a council-appointed collective executive. Here dualism is combined with consociationalism or majoritarianism depending on how the collective executive is appointed by the council and how decisions taken by the mayor together with the collective executive are reflecting positions in/of the council. It has to be emphasised that majoritarian and consociational forms of decision-making do not only depend on formal (legal) rules defined in municipal codes and/or the electoral systems. Rather, they rely on a socially determined and locally embedded logic of appropriateness (March and Olsen 1989) or even a pragmatically driven political logic of consequentiality, taking certain actual local (power) constellations into account. Acknowledgement of such locally determined patterns of policy-making is precisely the background on which Bäck (2005) has developed the mentioned typology. Accordingly, for the purpose of this paper the typology proposed by Bäck will be used for a comparisons of majoritarian and consociational forms of interactions between the mayor and the council possibly cross-cutting the monistic or dualistic distributions of tasks. For reflecting the legally given monistic or dualistic distributions of tasks a distinction will be made between cities (countries) where the mayor is directly elected (or designed) 13 and cities (countries) where this is not the case. For detecting majoritarian and consociational forms of interactions on the basis of the POLEADER-survey the communication of the mayor with different members of the council will be taken into account: 14 We will consider 12 However, the difference between monistic and dualistic systems is not always as clear in practice as it would appear from a legal point of view. This is especially the case when in a dual system (as in some of the German Länder), councils can intervene in administrative matters and thereby genuinely exercise executive functions (through majoritarian power, whereby a majority in the council formed by one party or a coalition is able to control the administration). 13 This includes mayors who are designated (i) by direct election or (ii) as the official leader of a majority formed by election like in the cases of France, Spain and Portugal. 14 The mayors were asked: How often do you normally communicate (oral communication only, not written, including meetings, telephone calls, etc., ) with the following persons/groups of persons? They could mark their communication with (i) majority leaders in the local assembly, (ii) other members of the majority in the local assembly, (iii) opposition leaders in the local assembly, and (iv) Other members of the Opposition in the Local Assembly as daily, 2-4 times a week, once a week, 1-3 times a month and seldom.

12 12 interactions as majoritarian if interactions of the mayor are stronger (more frequent) with councillors of the (his/her) majority than with those of the opposition, and as consociational if the mayors interacts with councillors of the opposition as much as with those from the majority. 15 This means that we will talk about assembly Government if the mayor is not directly elected (or designed) and interacts with councillors of the opposition as much as with those from the majority, parliamentarism if the mayor is not directly elected and interacts more strongly with councillors of the (his/her) majority than with those of the opposition, presidentialism if the mayor is directly elected and interactions with councillors are limited (compared to the average), consociational semi-presidentialism if the mayor is directly elected and interacts with councillors of the opposition as much as with those from the majority, and majoritarian semi-presidentialism if the mayor is directly elected and interacts more strongly with councillors of the (his/her) majority than with those of the opposition. Tables 6 gives an overview about the distinction which has been done on the basis of Bäck s typology and how this distinctions has been made. Table 6: communicatio n/interaction with councilors Local government systems according to dualistic and monistic as well as and consociationalism and majoritarianism direct elected yes no of own party = other consociational semi- assembly government parties of own party < or > other parties lower than the average of elected mayors presidentialism majoritarian semipresidentialism presidentialism parliamentarism This means that in cases where the mayor is not directly elected the frequency of communication/interaction with councilors will not be taken into account. It will be just considered whether or not the mayors interacts with councillors of the opposition as much as with those from the majority (assembly government) or not (parliamentarism). In cases where the mayor is directly elected those with an overall low frequency of communication/interaction 16 are put under presidentialism, the rest of the elected mayors are again divided between those interacting with councillors of the opposition as much as with those from the majority (consociational semi-presidentialism) and the others interacting more 15 For this purpose, the communication frequency with majority and minority representatives in the council (both leaders and single councillors) were merged into a ratio of concordance. If the communication between mayor and majority on one side and mayor and minority on the other side ranges between 0.8 and 1.2 points, his communication behaviour is defined as consociational. If his communication ratio either falls below 0.8 (which means that he talks much more with the opposition than with the majority) or exceeds 1.2 (which means that he talks much more with the majority than with the opposition), he acts in a majoritarian way. 16 A mayor is showing a low frequency of communication/interaction with councillors, if he is communicating less than one time a week with all of the four actor groups (leaders and single councillors of majority and opposition both).

13 13 strongly with councillors of the (his/her) majority than with those of the opposition (majoritarian semi-presidentialism). Table 7: Distribution of local government systems according to Bäck s typology by countries (in per cent) consociational semipresidentialism majoritarian semipresidentialism presidentialism assembly government parliamentarism SWITZERL FRANCE GREECE ITALY SPAIN PORTUGAL NETHERL DENMARK SWEDEN BELGIUM GERMANY HUNGARY POLAND IRELAND ENGLAND CZECH R AUSTRIA TOTAL Based on data from the POLLEADER survey. N=2,698 There are clear country-related patterns (see Table 7) pointing to the relevance of the monistic resp. dualistic institutional structure of local government, but there are also differences highlighting the importance of locally rooted forms of governing urban affairs by interaction which are not determined by the formal institutional structure of a national local government system. The latter is most obvious in case of the countries listed in the lower part of Table 7 (perhaps with England as an exception because the Local Government Act allows for different solutions; as pointed out above). In the upper part of Table 7 countries with more clear patterns are listed. This is the case with Switzerland with consociational semi-presidentialism, most of the countries belonging to the Franco type or the strong mayor type of local government systems (according to the typology of Mouritzen/Svara) or having a political mayor (according to the POLLEADER typology), 17 i.e. France, Greece, Italy and Spain, with majoritarian semi-presidentialism, Portugal with presidentialism, 17 Overviews about the combination between the types of the Bäck typology and those of the Mouritzen/ Svara and POLLEADER typology is given in Table II and III in the annex.

14 14 the Netherlands, Denmark and Sweden, i.e. countries belonging either to the collective form or the committee leader form (according to the Mouritzen/ Svara typology) and having a collegial leader (according to the POLLEADER typology) with assembly government, and Belgium with parliamentarism Policy-making style: Cooperative or non-cooperative The mayor s perception of the council may also depend just on her/his personal or on a specific locally embedded policy-making style, i.e. without any direct impact of institutional conditions like the local government system. To consider this possibility we will just take the index on the interaction of mayors with council members (mentioned right before) into account. We will talk about a non-cooperative policy-making style if interactions of the mayor are stronger (more frequent) with councillors of the (his/her) majority than with those of the opposition, and of cooperative policy-making style if the mayors interacts with councillors of the opposition as much as with those from the majority. 18 Table 8:. Policy-making styles by countries (in per cent) Country cooperative N FRANCE ITALY BELGIUM GREECE SPAIN AUSTRIA PORTUGAL CZECH REPUBLIC HUNGARY ENGLAND IRELAND POLAND GERMANY SWEDEN DENMARK NETHERLANDS SWITZERLAND TOTAL Although policy-making styles should be here considered as personal characteristics or even local peculiarities without relating them to the institutional settings given by country-specific 18 Like before for the considerations on dualistic and monistic local government systems in combination with consociationalism and majoritarianism, the communication frequency with majority and minority representatives in the council (both leaders and single councillors) were merged into a ratio of concordance. If the communication between mayor and majority on one side and mayor and minority on the other side either is between 0.8 and 1.2 points, her/his policy-making style is defined as cooperative l. If her/his communication ratio either falls below 0.8 (which means that he talks much more with the opposition than with the majority) or exceeds 1.2 (which means that she/he talks much more with the majority than with the opposition), she/he acts in a non- cooperative style.

15 15 local government in Table 8 the distribution of a cooperative and non-cooperative policymaking styles is presented by countries. It turns out that all Swiss mayors seem to act according to a cooperative policy-making style. But also the percentage of mayors with such a style is significantly high in the Netherlands, the Scandinavian countries of Denmark and Sweden as well as in Germany. On the other side one finds the Mediterranean countries of France, Italy, Greece and Spain as well as Belgium where the proportion of mayors with cooperative policy-making style is quite low, i.e. where a non-cooperative style is clearly dominant among mayors. These pattern may fit nicely to typologies of local government systems (in particular to those of Hesse/Sharpe with their distinction between the southern so-called Franco type and the North-Middle European type although they consider primarily central-local relations) and to country-specific pattern of consociational and majoritarian democracy reflected in Bäck s typology. However, we should leave it open for the analysis if policy-making styles operationalised in the mentioned way alone, i.e. without relating them to institutional settings, contribute to the explanation of differences in the dependent variables Council majority or not The perceived influence and performance of a council may also depend on other contextual aspects. The most evident one could be a majority of one party in the council because this (what ever the institutional setting looks like) is fostering its influence at least in comparison to unclear majorities. A fostered influence of the council can have an impact on its performance at least in the perception of councillors and/or other actors. In the POLLEADER survey mayors were asked Does any party presently have an overall majority of seats in the local council? And they had the options to answer Yes, No, but there has been a stable majority from the beginning and No, there is no stable majority. This question implied, of course, that majorities in a council formed by a coalition could not be captured. The responses by countries are shown in Table 9. In the U.DI.T.E survey municipal chief executive officers were just asked whether or not one party had an absolute majority. The results are shown in Table 9 as well. The discrepancies between countries in respect to a clear majority in the council (based on the POLLEADER data; see columns 2-4 in Table 9) are striking. Denmark has an extremely high proportion (of more than 60 %, see row 4) of councils with no stable majority. This figure is also quite high in Hungary, Germany, Ireland, the Netherlands, the Czech Republic and Poland. But in Poland the proportion of councils with a stable majority is high as well due to the lowest proportion of councils where a stable majority got lost since the last election (see row 3). This proportion is in the European average (or those of the European countries included in the POLLEADER survey) nearly as high as that of councils with a stable majority (i.e. about 42 %). However, big differences are covered by these mean values. There are on the one hand countries (Austria, Portugal and Poland) where a stable majority is a usual feature (the same is true for Spain with a value of about 51 %). On the other hand there is a big group of countries (comprising the Netherlands, Italy, Hungary, Belgium, the Czech Republic and France) where in more than 50 % of the councils a stable majority got lost during the last term. Is seems to be interesting to see whether or not these differences in the stability of a council majority matters in respect to its perceived influence and performance.

16 16 Table 9: Existence of one-party majorities in the councils (in per cent) Yes* No, but there has been a stable majority from the beginning* No, there is no stable majority* Majority of one party** NETHERLANDS ,2 HUNGARY CZECH REP ITALY ,1 IRELAND ,3 DENMARK ,7 BELGIUM ,0 GERMANY ,7 FRANCE ,9 SPAIN ,1 POLAND PORTUGAL ,9 AUSTRIA * Based on data from the POLLEADER survey. ** Based on data from the U.DI.T.E survey. Surprisingly (but just at a first look), the figures about on a majority of one party differ for some countries remarkably between columns 2 and 5 from the left in Table 9: For Ireland, Denmark, Germany and Portugal the percentage of councils with a one-party majority is indicated higher in column 2 and for the rest the figure in column 5 is higher: How to explain this? The reason is that the figures column 2 are based on the POLLEADER and those in column 5 on the U.DI.T.E survey. That these two surveys lead to such different figures could be related to their different coverage: The POLLEADER survey covered just municipalities with more than 10,000 inhabitants and the U.DI.T.E survey municipalities with more than 1,000 inhabitants. But is seems to be more sensible that the different years these two surveys had been carried out (i.e. the mid 1990s for the U.DI.T.E and 2003/04 for the POLLEADER survey) and the changes which have taken place in the meantime are the reason for this pattern especially change in the structure of the party systems and the dominance of individual parties in countries like Italy, Spain, France and Belgium. 19 And changes in the party system leading to instabilities of council majorities should be kept in mind for further considerations on shifted power relations between the council, the mayor and the municipal administration Other contextual aspects and personal factors Another contextual aspect which could have an impact on the perceived influence and performance of a council is perhaps the size of the city (in respect of the number of inhabitants). Or to put it as a question: Do councillors in smaller municipalities have a better chance to scrutinize and control the executive (and especially the mayor), and to represent requests and issues emerging from local community and to publicise debates on local issues? It can be argued in this way because the executive (administration) is smaller and can be be overseen by the councillors more easily and they are nearer to the people. But it can also be argued that councillors can fulfil these roles better in bigger cities because in these cases they have (in most countries) more resources at their disposal to fulfil their tasks. 19 Also changes in the local election systems should be considered here.

17 17 Finally, it should at least be tested whether or not personal factors such as age, gender, education, experience (years) in office, full-time mayoralty (time spent on mayoral office per week) have an impact on the perceived influence and performance of a council. 3. What already has been done (with the data) Data from the survey on mayors (the POLLEADER survey) have already been analysed by Bas Denters (2006) regarding the relation between the mayor and the council. He charted not only how mayors evaluate the performance of the council both in its internal and its external role (Denters 2006: 271). He used also responses of mayors on changes in influence that have occurred in the last decade among the various actors in local affairs (see Table I in the annex). Finally, he had a look at institutional variations (having a directly elected mayor or not, and having a committee-leader form of local government or not) and if these matter in respect to the self-perception of influence by mayors and their evaluation of the role of the council. On the basis of his analysis Denters (2006: 283) concluded that mayors across Europe report a change in the balance of power in local government, in which the political executive and especially the mayor have gained in political leverage over the council. As far as mayoral perceptions are accurate this suggests that the strengthening of the political executive as one of the major desiderata in the NPM-doctrine is now very much a reality in many local governments across the continent. We have also found that the apparent widespread adoption of New Public Management maxims (e.g. let managers manage ) has not resulted in a decreased political control of the local bureaucracy. In the perception of mayors, there has rather been a move in the opposite direction: according to the mayors recent years have witnessed an increased influence of local politicians on local administrators. [And finally Denters stated:] The mayors consistently report that of all major local political actors, the councils are least influential. Reflecting different power relations between the mayor and the council based on the Mouritzen/Svara typology Denters concluded that the power distance between the two offices was much larger in strong mayor systems than in collective body systems (Denters 2006: 283) leaving the committee leader form more or less in-between (see Denters 2006: 275 f.). This suggests that the introduction of a directly elected mayor further tilts the balance of power between the mayor and the council in favour of the former as Denters (2006: 283) summarized. Distinguishing between the internal role of the council, i.e. scrutiny and control of the executive, and the external one, i.e. representing the requests and issues emerging from local community and publicising the debate on local issues, Denters (2006: 283 f.) came to the result: On the whole mayors are rather positive in their evaluations of the performance of the councils. This conclusion, however, needs some differentiation. In strong mayor systems, mayoral assessments are generally somewhat less positive then in collective executive body systems. This might very well reflect the more or less natural competition between two offices (the mayor and the council) that both have an independent democratic legitimation. Furthermore, Denters was interested in the impact of institutional variables namely the direct election of mayor and the presence of a committee leader system, i.e. the two factors