Rethinking Local Governance : Hierarchies and Networks in Mexican Cities

In recent years, the literature of governance has  gained relevance in public administration studies.  One of its main contributions has been to propose  that the use of networks in policy processes, instead of hierarchies, has modified our understanding of contemporary government. Governance  literature has been produced mainly in AngloSaxon countries; nevertheless, this article argues  that it has some value as a tool to explain present  dynamisms in Mexican municipalities. Local  networks have become more important to policymaking in Mexico and, as a result, nongovernmental actors are now more influential in  some policy sectors. The article proposes that this  partial shift from government to governance has  increased bureaucratic fragmentation, the blurredness between the public and private spheres, and  non-governmental self-organization. Defined in  these broad categories, governance theoretical  frameworks are useful to understand Mexican  cities.  Resumen:  Reconsiderando la gobernanza local: Jerarquias y redes en ciudades mexicanas En los ultimos anos, el enfoque teorico de la gobernanza (governance) ha ganado terreno en los estudios de la administracion publica. La gobernanza  propone basicamente que al introducir redes (en  lugar de jerarquias) en los procesos de politicas  publicas, se modifica el significado tradicional de lo  que es el gobierno. Aunque la literatura de la gobernanza se ha generado sobre todo en paises anglosajones, este articulo sostiene que posee un  cierto valor para analizar los gobiernos locales de  Mexico. Las redes ahora son mas importantes en  los procesos de politicas publicas locales y, por  consiguiente, los actores no-gubernamentales son  mas influyentes en algunos sectores. El articulo  propone que este cambio parcial del gobierno a la  gobernanza ha incrementado la fragmentacion  burocratica, la borrosidad entre lo publico y lo  privado, y la auto-organizacion no-gubernamental.  Las proposiciones teoricas de la gobernanza, definidas de esta manera amplia, son utiles para entender las ciudades mexicanas.

than the use of traditional interventions based on hierarchic command, shows that in Mexican municipalities there has been a certain shift from government to governance.
The shift, however, has not been all pervading.This article argues that, although the shift to governance can be verified in some Mexican urban municipalities, it has occurred in an uneven and varied way.The great diversity of basic socio-political, economic, and even geographical conditions present in Mexico has generated a patchy picture in the use of networks for policy-making purposes.
When Mexican local governments seek to improve their public service provision, most of them do it by requesting larger amounts of federal transfers; deciding on policy priorities following administrative and operative criteria formulated in state capitals or federal ministries (Cabrero 2003b).This method to determine policy objectives suggests that most municipal policy communities are closed, formed almost exclusively by governmental actors and institutions.Comparative studies carried out by Cabrero (2003a), Guillén and Ziccardi (2004) and Ziccardi (1995) indicate that systematic policy-making substantially based on networks can be verified in some municipalities, mainly urban or semi-urban, that have experienced political alternation in office, and that maintain close-knit neighbourhood relations.Mayors and City Councils tend to use networks in programmes related to areas with high legitimizing potential, like street paving, the introduction of water pipelines, or education.Networks are rarely used in budgeting or in policy sectors that risk becoming too complex to handle (Flores 2005).
The Mexican case confirms that hierarchies and networks co-exist in local policy-making; nevertheless, the particular mix of policy areas using traditional hierarchical mechanisms and the ones using 'soft' network co-ordination varies according to contexts.The complexity of bureaucracies, the general political stance of the party in power, and the interdependence among non-governmental and governmental actors play a role in shaping governance patterns.By recognizing the importance of variables like these, recent academic literature has become capable of developing sophisticated and differentiated accounts of local governance in non-European environments.Emphasizing structural attributes rather than specific government-society forms of interaction is more useful to explain the Mexican context.The introduction of non-governmental actors in policy-making has made Mexican local politics more fragmented, blurred, and self-organized than in the past.In that measure, municipalities have made use of governance mechanisms.
In order to develop this argument, the article considers the contributions of the literature on governance that appear to be especially useful for analysing the Mexican case.After that, the three cases of study are compared according to the main lines established by the literature's analysis.Finally, the conclusion discusses their theoretical and empirical value.

Three basic propositions
As Börzel (1998) argues, one of the main difficulties in attempting any consistent use of governance definitions is their 'Babylonian multiplicity'.Following Krahmann's (2003) considerations, I analysed governance understandings that meet three conditions.The first one is that they are theoretical frameworks.This means that regardless of the particular assessment about the consistency or accuracy of models, contributions considered in this article treat governance as an analytical tool that explains or helps to understand policy-making in conditions of increased socio-political fragmentation.The second is that they are conceptual frameworks designed basically for the national and sub-national levels of analysis.Finally, they examine the State's capabilities in contemporary polities and their implications on our traditional notions of government.Circumscribed like this, governance literature, especially what Marinetto (2003, 597) calls the 'Anglo governance school', can be seen as a cumulative construction that has gradually evolved towards more differentiated understandings based on national and local contexts.In my opinion, Anglo-governance literature has developed three basic propositions that are important to understand the Mexican case: 1. Governance and government are not a dichotomy.Government and governance can be seen as ideal types of policy-making and implementation (John 2001) and, as such, they are never fully verified in all their possible dimensions.These ideal types are useful, for they identify overall trends of co-ordination modalities and are effective in comparative analysis; but they can be misleading if are interpreted as mutually exclusive poles in a dichotomy.As Genschel (1997) argues, in real political systems what we can find is a blend of different governance modalities.Hierarchies continue to be used (and so are traditional government's intervention mechanisms), which still are a formidable source of legitimacy.In situations of governance, the usual is that policy-making processes based on networking (and characterized by fragmentation) accompany hierarchies.Equally important are policy arrangements based on the markets' participation in some policy sectors.
This combination produces political systems that are nearer to government or to governance (Pierre and Peters 2000).This evaluation, however, must be qualified according to the particular mix of countervailing tendencies present in sociopolitical systems.Governance cannot be defined as radically different to government, as some initial interpretations on the arguments of Rhodes (1997) appeared to imply.Another matter, of course, is how compatible (and problematic) is to have networks and hierarchies working together.

2.
Governance is characterized by increased fragmentation, blurredness, and selforganization.Governance is a 'flexible pattern of public decision-making, based on loose networks of individuals' (John 2001, 9).This definition assumes certain elements that are common to governance understandings, and that point at basic traits of local governance in Britain and other western industrialized countries.In the first place, governance as a policy-making paradigm, supposes fragmentation.This implies a move towards decentralization and power dispersal into a greater number of governmental or quasi-governmental agencies, many of which are composed by non-elected members from the three societal sectors (Andrew and Goldsmith 1998).This produces policy co-ordination problems that are counterbalanced by the development of strong mayoral leaderships and other alternative steering methods based on personal governing styles (Le Gales 2002).
In the second place, greater citizens' involvement in policy-making and the increase of the importance of networks blurs the Weberian distinction between the public and private spheres, producing private actors dealing with provision of public services, and governmental officials behaving like entrepreneurs (Stoker 1998).Finally, governance presupposes that self-organization, among networks and other non-governmental actors, has increased as the result of the redefinition of traditional governmental roles (Kooiman 1993).Fragmentation, blurredness, and selforganization, however, do not remain constant, but vary according to local and other factors.

Contexts shape local governance.
Local governance variations are moulded by specific contexts (John 2001, Le Gales 2002).In the studied Mexican municipalities, governance patterns are shaped by several factors of the local and national contexts, among which we can find some of the abovementioned variables.
The results found in the case studies adapt broadly to these theoretical propositions.The introduction of networks as a policy-making instrument has been limited to fundamentally the same policy sectors, but they have not been used to the same extent.The position of the governing party over the legitimacy of NGO representation plays a role in the kind of networks with which the Mayor interacts.The municipality with the higher needs in public service improvement is the one that has established larger planning and implementation systems based on networks.Finally, despite the increased self-organization among industrial and commerce chambers, governmental and non-governmental actors have been able to develop convergence of policies, although in varying degrees.The following sections of the article analyse these findings in more detail.

Local governance in Mexican municipalities
The municipalities used for the analysis were chosen following three main criteria.The first one was that their main socio-economic indicators (including quality of life) were similar.The second was their referential character: their respective governing parties considered them as models for municipal programmes applied elsewhere.This condition was decisive for choosing León, even though this meant having to compare two medium-size municipalities with a large one.Finally, the three local governments kept archives with public access.In this way, I selected León (Guanajuato), Orizaba (Veracruz), and Zacatecas (Zacatecas state).León is governed by the National Action Party (PAN), Orizaba by the Institutional Revolutionary Party (PRI), and Zacatecas by the Democratic Revolution Party (PRD), 2 which are the main national parties in the country.
The three cities display many of the legacies of the hegemonic party system that treated municipalities as part of the PRI political machinery, which allowed that party to retain the presidency for seven decades.The official party's weight would normally show itself in two ways: the usage of local government's resources to reinforce PRI's influence over voters, and the establishment of clientelistic networks of citizens, by which public services and other benefits were provided in exchange of political support (Camp 1996).In this context, administrative and technical requirements for the efficient functioning of municipalities occupied a subordinated place, unless their neglect caused problems over social governability.Nonetheless, there have been two factors that have been instrumental in modifying the pre-eminence of politics over administration: partisan alternation and new federal policies on transfers.
The municipality was the first level of government to experience partisan alternation in Mexico.In 1946, the first of a series of significant post-electoral municipal conflicts took place as a result of credible accusations of electoral fraud; but the first non-PRI municipality was recognized by the federal government in the state of San Luis Potosí until 1958 (Martínez-Assad andZiccardi 1988).The total number of opposition municipalities in the period between that year and the late 1980s remained unimportant.The presence of this small number of opposition municipalities, however, was helpful in legitimating the PRI's vast majority of local governments, for it was usually mentioned as proof of the democratic character of the Mexican political system (Ward 1998).The new Federal Code of Electoral Institutions and Procedures (COFIPE) of 1990, and the resulting reforms of 1996, relocated the supervision of elections from the Interior Ministry to the Federal Electoral Institute (IFE) and the corresponding Institutes in the states.These institutes are public-funded, autonomous, and their decisions rest ultimately on a-partisan bodies, attributes that were influential in the rapid increase of political alternation in City Councils. 3Table 1 shows the considerable dispersal of partisan control in the last decade.The other factor instrumental in changing the national context for municipalities was the 1998 Law of Fiscal Coordination (LFC), which was the result of the 1997-2000 Chamber of Deputies' composition.In an effort to reduce local clientelism and make federal transfers to the states more transparent, the opposition deputies (who had majority over the PRI for the first time) passed a LFC that mandated mu-nicipalities to organize citizens' consultations in order to receive Branch 33 funds (Ramo 33).Branch 33 funds are calculated following publicized formulae that take in account poverty and tax collection levels, among other variables, and are earmarked for financing municipal public services.The LFC requires local governments to inform residents about the public works funded with the Branch 33, their cost, their objectives and beneficiaries (Estados Unidos Mexicanos 2007, art.33 I).
Mayors also have to include benefited citizens in the planning, execution, control, and final assessment of the funds' expenses (art.33 II-III); all this supervised by Local Congresses.As a result of these national changes, the levels of sociopolitical fragmentation, blurredness, and self-organization in many municipalities began to increase.

Fragmentation
Fragmentation in the literature of sub-national governance is typically understood as the proliferation of autonomous governmental agencies.This usually implies the dispersal of the places of power within public administrative structures in order to gain in flexibility, diversity, and quickness of governmental responses; but it also considers the associated problems of lack of co-ordination and policy-coherence among governmental departments.The main problem that bureaucratic fragmentation tries to tackle is rigidity because, it is argued, it increases the probability of excluding citizens from access to public services when their particular situations do not conform to the administrative or procedural boundaries of governmental action (Andrew and Goldsmith 1998).Governments, trying to avoid these legitimacydamaging situations, have reacted generating an approach based on the specialization of policy sectors, expanding the scope of governmental responsibility into subjects that were unthinkable thirty years ago, but concentrating the agencies' interventions on 'rather narrow boundaries of particular 'sub-arena[s]' or 'specialized policy arena[s]' (Campbell, Baskin, Baumgartner and Halpern 1989, 86).
Fragmentation is also defined as the augmented influence of public-private policy networks, and the resultant hybridization of the instruments employed by governments to steer.This approach understands fragmentation as the process by which heterogeneous societal actors and 'spheres of action' mix together generating 'hybrid frameworks of participation' (Mlinar 1995, 145).It assumes proliferation of agencies or interest groups, their increased autonomy, and a set of resulting non-hierarchical interactions, but on a public-private basis.The resulting fragmented units are made up by both governmental and non-governmental institutions and individuals.Rhodes (1997) proposes a governance understanding (that of the management of inter-organizational networks) that is practically indistinguishable from fragmentation defined in this way.
If the proliferation of governmental agencies (both traditional and publicprivate) is considered, León is the most fragmented among the case studies, followed by Orizaba and Zacatecas.León has experienced a continuous rise in the number of departments and autonomous agencies since 1989.In the first PANista administration, the municipality had eleven departments and 2,900 employees.They grew to fourteen departments and 3,100 employees in the following administration.In the period between 1995 and 1997, the Mayor established another de-partment, although he maintained basically the same number of employees (Cabrero 1999).In the studied administration (2000)(2001)(2002)(2003), the total number of relevant agencies was 28, including fourteen departments, six autonomous agencies 4 , three directorates, and five public-private planning councils that were noticeable for their influence in public opinion.The municipality now has 4,054 employees (H.Ayuntamiento de León 2005).
Some of these agencies deal with policy sectors that are atypical in Mexican municipalities, like Public Health, Housing, and Ecology.In this case, the process of fragmentation consisted of both the atomization of agencies and the introduction of new ones, like the autonomous agencies of the System of Potable Waters and Sewers of León (SAPAL), the Municipal Institute for Housing, the Municipal Institute of Women, and most notably, the Municipal Institute for Planning (IMPLAN) (see table 2).
Orizaba is the second most fragmented municipality of the three cases.During the studied administration, Orizaba kept a structure with an agency for each important policy sector -in total fourteen, including a planning citizens' council.Among these, two are especially significant.The first one is the Department of Governance, which has preponderantly political functions.It is in charge of the directorate of the spokesperson and the relations with the local media, but it also deals with emergent political problems (Velázquez 1999).The second noteworthy agency is the Office of Commerce, which in Orizaba has a departmental status while in León it is a lower-level body.This suggests a confirmation of the importance that traditionally PRI governments have given to the relations with street vendors and the commercial businesses in general, for both Departments of Governance and Commerce deal with what is considered the most conflictive policy sector after urbanization services in Orizaba.In any case, Orizaba has fewer departments than León; it has not established any autonomous body, and the sole public-private agency with significant influence on overall policy is the Council of Municipal Development (see table 2).Zacatecas is the less fragmented municipality of the three cases if the number of municipal agencies and their levels of autonomy are considered.Zacatecas has a very compact administrative structure, with only seven main departments (none of which is autonomous) and one citizens' council relevant for decision-making on services funded by federal transfers (the Council of Municipal Development).This Council decides over the priorities in the maintenance and introduction of basic urban services paid with Branch 33 funds, and its decisions must be implemented by both the Mayor and the City Council (Juárez 2002).However, the actual fragmentation of policy produced by this change of the place of power is limited, given that the Mayor, the Aldermen, and the directors of key departments are still able to influence the decisions of this Council using their expertise and political prestige.
León is more fragmented given its larger size and therefore, the complexity of its urban problems.Nevertheless, it is important to note that the party in office is also relevant in this particular case.The PAN is known for its preference to establish direct citizen-to-governmental agency exchanges, as a means to circumvent neighbourhood associations and unions that traditionally have been co-opted by the PRI.By fragmenting its structure, León is in a better position to respond to the specific needs of its interest groups and to bypass traditional social brokers.

Blurredness
Governance understandings argue that the dynamisms introduced by networks in policy-making blur the distinction among the public, private, and third sector actors.Governance is characterized by the 'blurring of boundaries' between the traditional sectors, not only because the increased societal complexity has redefined some of their usual functions, making this tripartite classification too rigid to analyse hybrid or border dynamisms (Pierre 2000), but also because it is becoming increasingly evident that complex policy issues require an inter-sector and multilevel approach.This approach focuses on resources and places of power dispersed among a great diversity of actors; regardless of the sectors in which they have been routinely classified (Kooiman 1993).
Governance theoretical frameworks argue that, from the government's perspective, networks in policy-making perform two main functions: on the one hand, they are instrumental in legitimizing policy; on the other, they supply governmental institutions with resources.The first function is carried out by policy networks and communities through the interactions among their members, which permit authorities to know prima facie the citizens' demands, make the necessary policy adjustments and definitions, and gather resources in order to sustain a given governmental programme or augment the acceptance to it (Rhodes and Marsh 1992).The second function performed by networks is to obtain resources that are not easily available when using traditional hierarchical methods of control.Usually, most of these resources are informational, and relate to citizens' participation in the detection of problems, the design of policies, and the evaluation of their outcome, all of which has, ultimately, legitimizing purposes (Bardach 2001).However, there is an increasing trend to use networks' resources in portions of the policy process that in the past were exclusive to government, such as the actual implementation of programmes, or the pooling of funds in order to finance public works.
The three cases of study use networks for policy-making purposes in different ways, but they are especially noticeable in their Neighbourhood Committees (NCs).NCs are citizens' policy communities established and supervised by municipal officials, with the objective of including users in the decisions on services directly affecting them.NCs have different names in the three municipalities, but they basically perform the same functions: they must a) consult neighbours about public services' priorities, b) propose public works to municipal departments, c) evaluate the quality of public services, d) participate in citizens' councils, and e) inform neighbours, in public meetings, about the public works done in the neighbourhood.
In general, establishing NCs has been relatively easy for the three municipalities, regardless of the size of neighbourhoods or the administrative structures necessary to manage them.The proportion of neighbourhoods with an NC is high in the three cases, ranging from 60.32 per cent (León) to 75.25 per cent (Zacatecas).León is the municipality with the highest number of NC officials (2,250), although proportionately, Zacatecas has more (745, or 0.60 per cent of its population).In total, León has established 450 NCs, while Orizaba and Zacatecas have 80 and 149 respectively (see table 3).The high percentage of NCs suggests that, at least initially, citizens' participation is facilitated by the expectation of obtaining improvements in public services.Participation is potentially higher in neighbourhoods of lower socio-economic strata, where is more probable to find deficient public services' coverage.This general impression was confirmed by key interviewees, who complained about the difficulties experienced to establish NCs in neighbourhoods with good-quality services, and get their members to attend City Hall meetings (Alvarado 2001, Conde 2002, and Ortíz 2003).As one interviewee put it, 'affluent people do not participate in Neighbourhood Committees' (Ramírez 2002) for, among other reasons, they usually have formal jobs (that is, not in the informal economy) and cannot get involved in activities that normally take place during office-hours (Muñoz 2001).
Given the limitations of space it is not possible to analyse participation rates in NCs and the involved policy issues in detail.It suffices to say that, as Desai and Imrie (1998) have pointed out for other places, León, Orizaba, and Zacatecas have introduced citizens' participation in non-strategic areas, for the public service improvement carried out with the NCs is marginal.The bulk of public service provision is still planned, implemented and evaluated in traditional ways.In the three cases, the City Councils (Cabildos) retain the power to establish the initial budgets that are the limits to petitions from NCs.They can also veto particular public works resulting from the priorities defined by public-private councils if they are not considered technically viable.
All the same, there has been a significant involvement of citizens in NCs, which has been instrumental in the improvement of water, electricity, and sewers supply in poor areas.In general, policy networks have become more relevant for local policy in the relatively narrow sector of basic urban services, either by micro problem-solving implemented through the NCs, or the macro prioritization made in public-private planning councils.As Bevir, Rhodes, and Weller (2003, 10) argue, the shift from government to governance has taken the form of governments giving a 'reduced and shared role in problem solving' to citizens.In the three cases, this role has been given to inhabitants of neighbourhoods with problems in the quality or supply of basic services.The participation of citizens, however, is not apparent in other important policy sectors, such as the planning of local economic development, the regulation of commerce, sanitation, overall public safety, or even internal administrative services such as the Office of the Municipal Comptroller, or the Municipal Treasurer.
León has a greater number of policy sectors open to citizens' participation, for this municipality has established public-private bodies on subjects such as population growth and the prevention of addictions, as it indicated in table 2. León maintains the most complex participatory system of the three cases, in part, because it is the municipality with greater needs of basic urbanization services.Orizaba and Zacatecas, on the other hand, do not rely as much on NCs members, developing a more traditional approach to policy.León is closer to governance than Orizaba and Zacatecas, for it has a more blurred framework of public-private interaction.

Self-organization
Self-organization can be defined as self-government and autonomy from the State (Rhodes 1997).Self-organization plays an important role in governance given that, in conditions of increased social-political complexity, governments have experienced a reduction in their capacities to exert control over societal actors.The governmental capacities to effectively control and organize, especially nongovernmental actors, has decreased as the result of insufficient levels of acceptance of policy, increased societal fragmentation, and the complications resulting from more intricate policy-making processes (Kooiman 1993).
From the governmental perspective, self-organization is more noticeable in the sectors where networks are incorporated into the policy-making process, especially when they are inter-organizational and operate in a highly fragmented environment (Rhodes 1997).Self-organization among non-governmental actors produces polycentrism, for the networks of autonomous actors can accumulate a considerable amount of resources of many kinds, which can place them in a position to compete with the State (Jessop 1998).Early governance understandings considered that self-organization, although not necessarily incompatible with public interests, obstructed traditional governmental interventions.Nevertheless, more recent contributions have pointed out that, in spite of the general increase of self-organization among non-governmental actors and inter-organizational policy networks, the 'total sum of the State capabilities [remains] largely unchanged' (Pierre and Peters 2000, 92-93).The State continues to be the most influential actor over societal affairs, albeit sometimes in manners and policy sectors that were unusual some decades ago.Accordingly, self-organization is not automatically considered as opposite to governmental activity as a whole but, rather, as a condition incompatible with the traditional authoritative means to exert control and to ensure accountability to the State.
In the three studied municipalities, self-organization is especially evident in their relations with commerce and industrial chambers.These associations' greater influence in policy-making, if compared to other NGOs, is partly explained by the economic importance of their members.The fact that most chambers' affiliates belong to the middle classes gives them more power over urban issues, particularly in matters related to the use of land.Typically, the members of the local branches of the National Chamber of Commerce (CANACO) and the National Chamber of the Transformation Industry (CANACINTRA) are middle-aged men, with university degrees, cars, and incomes above the average (Guerrero 2001, Vega 2001, Velasco 2001).It is to meet the needs of these middle classes that the most expensive public works are usually done, including the changes in urban composition that the larger numbers of cars entail.This, and the traditional closeness between the public and business sectors in Mexican politics, has given commercial and industrial chambers a prominent place in local policy definition.
In the case of León, CANACO and the municipality have worked closely in programmes to activate the establishment of small and medium-sized businesses as a strategy to promote employment.The most publicized programme assigns grants and soft credits to starting companies, financed by the CANACO and the local government in equal parts (Gómez 2002c).The joint programme was considered of high priority by both the municipality and the chamber of commerce for, days before the official presentation of the programme, the CANACO had threatened to stop paying taxes if the government's policies on commerce did not improve (Gómez 2002b).In spite of this dispute, the joint programme was implemented.In addition, CANACO has agreements with the Mexican Social Security Institute (IMSS) and the Ministry of Economy to operate as an alternative assistance centre, giving advice on taxes, IMSS fees, and ways to avoid fines.
CANACO has also implemented some programmes to help the municipality to have an updated register of street vendors, usually asking its affiliates to report them if they try to sell their merchandise in unauthorized areas.The constant complaints of this chamber to the Directorate of Commerce have been instrumental in defining the administration's policies on street vendors, to the point of causing organized protests of vendors against CANACO itself (Gómez 2002a).CANACINTRA of León works in a similar way, helping the municipality with its Housing programmes.Using the experience of previous collaborative schemes with INFONAVIT (the Institute of the National Fund for the Workers' Housing), this chamber has given constant advice to the Municipal Institute for Housing.In addition, CANACINTRA members with social security rights can obtain public credits to buy a new or used house, build one, build an extension to an existing property, and pay debts related to these matters using CANACINTRA as under-writer for the credit or debt (CANACINTRA 2003).
In the case of Orizaba, the Mayor of the 1998-2000 administration lobbied members of CANACO in order to get land donations.The most important public work of the period was the Orizaba Avenue, a two-lane street parallel to the railroad tracks, which crosses the city on its West-East axis.Owners of the adjacent properties were against the project, for the Avenue required the use of some private land.The municipality, on the other hand, could not afford to pay the land's price.In the end, the Mayor asked the most important industrialists of the city, some of which had estates in the area, to donate a part of them to the municipality so the Avenue could be constructed without affecting minor private owners.The estate portions were given to the local government and, after a minor direction change, the Avenue was built (Dujowich 2001).CANACO has also participated in the design of the commercial policies for the city: it is a non-spoken rule that its president has a seat in the public-private planning council, regardless of how the other voting members are elected.In previous administrations CANACO was in contact with the municipal Department of Commerce on a daily basis (Kuri 2001).
In the studied administration, this commercial chamber and the municipality introduced a policy of zero-tolerance of illegal street vendors.Normally, the Department of Commerce assigns concessions to particular vendors or approves ad hoc agreements with unions using different criteria.In any case, the licence for selling in pedestrian areas is territorially limited and is temporary, awarded usually for the Fairs of June or the festivities between December and January (Orizaba en Red 2002b).As a long-term strategy, the municipality has tried to relocate street vendors in controlled street markets or in the conventional ones (Orizaba en Red 2002a).CANACO has similar collaborative schemes to those of León, asking its members to report any illegal selling, and helping them to obtain housing credits (Kuri 2001).
Finally, the chambers of Zacatecas have put into practice similar programmes to those implemented in León and Orizaba.CANACO has assisted the municipality in updating the street vendor records, asking chamber members to report possible violations of municipal regulations.In Zacatecas the negotiations with unions took longer than in Orizaba, but they achieved the conversion of permanent street markets into temporary ones.In this issue, CANACO, the local CANACINTRA, and the municipality converged over policy given the conditions put forward by the UNESCO World Heritage List, of which Zacatecas is part (Ponce 1999).
The experience of commercial and industrial chambers in the three cases suggests that convergence is relatively easy to reach when three conditions are met: a) governmental lack of resources, b) the private actor's willingness to perform activities that are usually exclusive of government, and c) an agreement on operational issues.When collaboration in joint programmes takes place, both municipal officials and chamber presidents interpret it as an expression of complementary policies.Interviewed CANACO and CANACINTRA officials argue that the obtained benefits are greater than the costs, especially in terms of saved time.Governmental bureaucracy, they argue, is not as efficient as private sector companies.Municipal officials, on the other hand, do not usually oppose collaborative programmes as long as regulations are followed, and a high level of supervision can be guaranteed.In the three cases, interdependence among governmental and non-governmental actors has limited the negative effects of self-organization, which is higher than in the past.

Conclusions: Anglo-governance from a Latin American context
The previous sections have shown a picture of local governance in three Mexican municipalities governed by different parties.In general, the evidence discussed here confirms Anglo-governance theoretical propositions, provided that they are articulated in structural terms and not in specific modalities of government-society interaction.As a conclusion, I comment the initial three propositions in the light of the experiences of León, Orizaba and Zacatecas: 1. Government and governance are not a dichotomy.Confirming the basic argument of John (2001), Krahmann (2003) and Pierre and Peters (2000), Orizaba, León, and Zacatecas cannot be classified as either cases of government or governance.The patterns described above show that given the increased fragmentation of policy sectors, and their corresponding dealings with society, it is necessary to develop detailed accounts of governance differentiating by concrete policy areas, administrative periods, and political orientation of the City Council (among other variables).In the three case studies there are policy areas, like the internal administrative services, sanitation, local economic development, and public security, where citizens' networks have not been incorporated in a substantial way.The three cases display patterns of strong Mayors, employing traditional hierarchical command in some areas, co-existing with policy sectors where networks have been introduced as an ordinary means of policy design.As a whole, and considering the number and quality of policy areas in which networks are used, León is closer to governance than Orizaba and Zacatecas, which display more traditional patterns of public administration.
2. Governance is characterized by increased fragmentation, blurredness, and selforganization.León, Orizaba, and Zacatecas have more governmental agencies than in the past; a fact that, especially in León, has increased fragmentation and the dispersal of decision-making places.As a result of the incorporation of participatory devices, and cooperation with autonomous local chambers, there has been an increase of blurredness between public and private functions.
Nevertheless, it is important to note that some of the results found, especially in the case of León, may owe themselves as much to the size of municipalities as to the internal dynamics of the cities.In general, larger municipalities require a more comprehensive and intricate bureaucratic structure given the complexity of their problems.That would explain the considerable difference between León and the other two municipalities regarding the number of local policy sectors.The degree of non-governmental involvement, on the other hand, seems to be related to the particular needs in those policy sectors and the political project of the City Council.
3. Contexts shape local governance.This theoretical proposition is especially important given that early governance understandings proposed specific patterns or means of interaction as the defining attribute of governance.The influential model of Rhodes (1997) argues that governance is characterized by the governmental management of inter-organizational networks, which can override the State's capacities if they become autonomous enough.This article proposes that in spite of the dynamics of governance present in the three case studies, there is not evidence of the formation of inter-organizational policy networks with the characteristics mentioned by Rhodes.There is, however, evidence of increased fragmentation, blurredness and self-organization in some policy sectors.
This fact suggests that the theoretical propositions of Anglo-governance might be useful for examining Mexican cities if they are formulated as general attributes of political systems, rather than as specific arrangements of government-society interactions.In the studied municipalities, these attributes are shaped by local conditions and the federal policies on transfers.Orizaba has been governed by the PRI (including its antecedent parties) in all post-1929 administrations, except in the periods between 1995-2000 and 2005-2007.The analysis focused on the 2001-2004 administration, which is a 'New PRI' one (because it employs some NPM criteria combined with more traditional politics).The PRD has governed Zacatecas since 1998, although it was the 1998-2001 administration that was studied.3. The increase of alternation is accentuated by the Constitutional prohibition for immediate reelection.The Mayor is elected, with his slate of Aldermen, in direct local elections.In most of the 2,439 municipalities of the country, the winner party keeps the Mayor and the majority of City Council seats and the rest are assigned according to proportional representation.Once the administrative period is over (usually three years) no member of the City Council can run for re-election, unless another period is allowed to pass.This, in addition to the fact that most municipalities lack a formal civil service, creates very high rotation levels and serious difficulties to maintain governmental programmes for more than a triennium.4. Autonomous agencies are funded by the municipality but their ultimate decision-making is made in their own Governing Trust (Junta de Gobierno).Mayors and Aldermen are usually members of this Trust.

Francisco
Porras holds a PhD in Politics and International Studies from the University of Warwick (UK), where he did research on local governance.He is currently a full-time Research Fellow at the Instituto Mora (Mexico City), and coordinator of the project 'Policy Networks and Fragmentation in Metropolitan Municipalities', which won funding from the National Council of Science and Technology of Mexico (CONACYT) in a national competition.Recent publications include 'Teorías de la Gobernanza y Estudios Regionales', Secuencia: Revista de Historia y Ciencias Sociales, No. 69, September-December, 2007.<fporras@mora.edu.mx>Notes 1.This article draws on material from more extensive research in Porras, F. (2005) Broadening Understandings of Governance: the Case of Mexican Local Government, unpublished PhD dissertation, University of Warwick (UK).2. The PAN has governed León since 1989, but the analysis focused on the 2000-2003 administration.

Table 1 .
Mexican Municipalities according to PoliticalParty, 1994Party,  -2007 (percentages)    (percentages) Source: My calculations based onE-local (2007).'Other' includes minor national and state parties.'Municipal Councils' are local governments directly appointed by Local Congresses due to unresolved post-electoral conflicts.'Usages and Customs' (Usos y Costumbres) apply in municipalities with indigenous population, whose traditional (non-electoral) mechanisms for appointing Mayors are recognized by the law.

Table 3 .
Number of Neighbourhood Committees (NCs) Compared to Total Number of Neighbourhoods and Population Source: my calculations and data from Goytia, 2000; H. Ayuntamiento de Orizaba, 2001; H. Ayuntamiento de León, 2002, and personal interviews.