The effects of rules on local political decision-making processes: How can rules facilitate participation?

It is often claimed that participation empowers local actors and that an inclusive decision-making process is crucial for rural development. We aim to investigate how formal and informal rules are set in local decision-making processes and how those rules may impact the actual level of participation by local actors. In a comparative case study, the rules-in-use for the planning of community projects in Thailand are examined. For our analysis, we use the Institutional Analysis and Development framework, which allows for more precise analysis of the impact of the rules. Fifty-three villages are served by four selected Tambon Administrative Organisations (TAO) which are either known for success in achieving participation or ranked as problematic in implementing the decentralization and local participation goals of the Thai government. The study is based on 60 semi-structured interviews with TAO staff, a survey of village leaders in 50 villages and a household survey of 104 villagers. We scrutinize seven types of rules and show some particular differences in terms of the impact from the rules-in-use. In the TAOs ranked as less participatory, the attendance rate in the meetings is found to be lower (boundary rule), villagers are informed about a meeting with a shorter notice (information rule) and more villagers mention that elites interfere in the project selection process (aggregation rule). A high level of fuzziness appeared in the position and authority rules. Further, we obtained information on the particular deontic logic, showing generally a high share of de facto may-statements in the implementation of the rules. We conclude that if the policy goal is enhancing participation, rule-setting offers good scope for intervention. From a practical perspective, information on administrative procedures has to be made more accessible and public administrators should receive procedural training.


Introduction
An important issue in political decision-making, rural development and local natural resource management is empowerment of local actors through participation. This participation should be ideally a form of engagement that values discussion, reflection and consideration over simply voting (Chambers 2003). Even though the issue has its controversies (Speer 2012), a vast portion of authors promotes citizen participation and engagement highlighting positive impacts thereof on numerous development outcomes (Irvin and Stansbury 2004;Gaventa and Barrett 2012). An example of such an impact can be an improved transparency in decision-making processes as well as strengthened perceived legitimacy of their outcomes (Birnbaum et al. 2015). Likewise, participation facilitates the break down of the complexity of problems and increases the participant's level of knowledge (Raisio and Vartiainen 2015). Research in the field of policy sciences has shown the positive impact of participation also on local political decision-making and development. Birnbaum et al. (2015), for instance, point out that the perceived legitimacy can be increased with participatory procedures, which is one factor of success not only in natural resource management projects but also in development projects. Thus, we have to ask how participation can be effectively achieved and more precisely, what the specific rules are that enable and structure initial deliberative processes leading to improved participation. Considerable work in political science and economics has examined how specific rules affect the incentives and outcomes for a wide diversity of situations (Ostrom 2014). Yet, in contrast to Araral (2014), we do not study how rules should be designed in order to reduce transaction costs, we rather want to see, in the first place, what rules we can detect and to what extent those are responsible for effective participation.
In order to understand the detailed rules-driven underlying mechanisms of community members' decision-making within the local development context in Thailand the article proceeds as follows. In ''Rules in participatory procedures'' section, we continue with explaining the ambitious decentralization program of the government structures in Thailand and the peculiarities of Thai communication habits. We state that rules, as exogenous factors, are responsible for the actually achieved level of participation, not without the account of the criticism on the distributional consequences of participatory processes per se. For our analysis, we use the Institutional Analysis and Development (IAD) framework, as described in ''The IAD framework: a rules framework'' section, which allows for more precise analysis of the impact of rules, and we conclude with classifying seven types of rules. ''Research area and data collection'' section describes the research area and the data collection. ''The action situation: local development planning'' section disentangles the Action Situations I and II of the planning process which are ''surveying villagers' needs'' and ''compiling villagers needs into a Tambon development plan.'' Further, participants and positions in both action situations are presented. ''Rules'' section provides detailed results on the rules and deontic logic discovered, structured along the seven types of rules and according to the action situations I and II. We conclude with findings on the differences between the more and less participatory TAOs as well as with some practical lessons that policy makers can actually address in order to increase public participation.

Rules in participatory procedures
In examining the participatory procedures in Thailand more closely, it became apparent that the actual participation of local communities can be hindered by both endogenous and exogenous institutions, which are mutually interlinked (Aoki 2007). Following Laswell's pledge for contextuality (1971, p. 3), decisions are always part of a larger social process and require to integrate the endogenous cultural factors, too. In Thailand, the endogenous factors such as cultural values and communication habits toward higher ranking officials are driven by collectivism, high-power distance and the cultural norm of Krengjai. 1 In that context, Nuttavuthisit et al. (2014) raise the question in how far formal rule-setting can help to achieve people's participation in local government despite the hindering culture of compliance to superiors. We would argue, however, that even if the endogenous factors were in line with what is needed for successful engagement, participation could still be hindered by the rules, according to which the actual decision-making process in local public administration and government is organized. But, independent of the endogenous factors that affect participatory success, exogenous factors can allow significant scope for intervention (Lowndes et al. 2006). Such exogenous factors include the rules-in-form and rules-in-use 2 within the administrative bodies that define the interaction with the local population. Rules are understood as being socially designed and being one key element in understanding participants' behaviors and interaction. To understand where the obstacles to villager participation might be reduced, it will be necessary to clearly identify these rules within the administrative bodies.
Such rules need to be analyzed for their distributional consequences as well. In that regard, there has been substantial criticism about the concept of participation. Some scholars argue that participatory approaches have often failed to achieve meaningful societal change owing to a failure to reflect issues of elite capture and politics (Bardhan 2002). Lortanavanit (2009, p. 176) concludes for Thailand that rural communities do not have enough weight to push through development projects that are better in line with their needs. Chaowarat (2010, p. 106, 109) likewise criticizes that the current process of decision-making in local development in Thailand is a one-way flow of information without genuine discussion or negotiation between local government and communities. To counteract this development, the understanding of rules is important, as administrative rules for interaction with the local population have to be constructed in a way that restricts the power holders from getting hold of the process and gaining disproportional benefits. Yet rules are often treated as a black box and how they matter has not been extensively studied (Araral 2014).
In contrast to the majority of studies in participation research, we are not studying the performance of participation efforts and the impact of stakeholder participation on policy making. Rather, we proceed from a ranking of well and badly performing regional administrative bodies in Thailand in terms of participation. We want to scrutinize the different formal and informal rules in local policy arenas, particularly in view of development planning processes, according to the achieved level of participation. We want to know the determinants of the actual level of stakeholder engagement that are influenced by precise rule-setting of an action arena.
We selected Thailand as a case study because it provides a clear evidence of how within the last two decades the decentralization process has brought a substantial change in the local administration with respect to delegating governance to local administrative bodies (Nagai et al. 2008, p. 29). The Thai case, even more, allows investigating the implications of rule making as a result of decentralization in a substantial detail. The success of decentralization is closely interlinked with the question whether or not decentralization gives space for villagers' participation in local government (Larson and Ribot 2004). Thus, we understand that decentralization and participation ought to gain momentum when promoted simultaneously. Nonetheless, participation in decentralization is a major administrative challenge, as will be shown. It does not take place in an institutional vacuum. It happens in a sociocultural environment as well as in a context of inherited rules and procedures. Within an ambitious decentralization program of government structures in the early 1990s, the Thai government has created the Tambon Administrative Organizations (TAOs) which represent the lowest administrative bodies and likewise the arenas in which deliberative stakeholder participation ought to take place (Dufhues et al. 2015). The TAOs are supposed to increase the role of the rural population at the community level and empower them in decision-making. In this context, various tasks of the central administration such as budget and personnel management have been transferred to the TAOs (Kanjina 2008;Krueathep 2004). The TAOs serve as our research units, or more specifically, the ''development plan making'' activities institutionalized within these organizations constitute the basis of our analysis. 3 The rules introduced since decentralization within the TAOs will be classified according to the IAD framework's typology of rules (Ostrom 1998, pp. 68-73). This analytical classification allows us to disentangle the detailed rules-in-use within the local planning process of community projects in Thai TAOs and to judge on the deliberative democratic qualities of the decision-making process. Already in 1971, Lasswell (1971, p. 2) hints to the idea that the norms of conduct are partly determined and made effective outside the machinery of legislation, administration and adjudication, which, once again, stresses the need to study the actual rules-in-use.
This leads us to our more abstract research question: What is the possible impact of certain rules in a decision-making process on the actually achieved level of participation?
We hypothesize that in the more successful TAOs the rules are structured and set up in a way to take the immanent constraints within administrative organizations (such as bureaucratic inertia and path dependency) into account and to be better compatible with the endogenous factors of the society. Thus, we aim to reach conclusions on preconditions, i.e., necessary up-front rule changes that would allow for better participation within the process of local development planning.

The IAD framework: a rules framework
The Institutional Analysis and Development (IAD) framework, developed by Elinor Ostrom and scholars associated with the Ostrom Workshop in Political Theory and Policy Analysis at Indiana University, contains a nested set of variables that scientists can use in efforts to understand human interactions and outcomes across diverse settings. In line with the broad prevalence of this framework in natural resource management (Ostrom 1998, p. 82), its application for participatory research is steadily growing (Denters and Klok 2003;Monnikhof et al. 2003;Koontz 2005;Smajgl et al. 2009;Lowndes et al. 2006;Van Damme and Brans 2012). 4 Speer (2012) even stresses its potential to study participatory processes. The framework is known for its ability to address the challenges of collective and deliberative decision-making. In one of his approaches to policy sciences, Lasswell (1971) uses ''an action arena'' or ''the situations in which interaction occurs' ' (1971, p. 25) in the center of his framework to point out that the actual situations are part of the equipment essentials to a policy scientist. In his model of social processes, same as in the IAD framework, Lasswell (1971, p. 24, 44) for instance conceptualizes participants, specializing here participation in decision-making, that take on various positions.
One of the main strengths of the IAD framework is its explicit attention to rules and rule-ordered relationships (Koontz 2005;Ostrom 2014). Thus, we use the framework to pay attention to the structure of the various rules-in-use that have an impact on our action situation, which is the planning process administered within the TAOs for new community development projects at the Tambon level. Applying the IAD framework does not necessarily mean an extensive analysis of all variable sets. Although we are very much aware of the context dependence-whether participation can be realized or not, the intention of this paper is to focus and elaborate on the impact of rules.
The center of the framework builds the action arena, a social space in which individuals interact, exchange goods and services, solve problems, dominate one another, or argue (Ostrom 1998, p. 82). The framework considers three sets of variables influencing the outcome of collective decision-making: (1) biophysical conditions, (2) attributes of the community and (3) rules-in-use that create incentives and constraints for certain actions (Ostrom 2005(Ostrom , 2010. The rules shape action situations and people's behavior differently depending on what behavior they permit, the specific institutional decision-making procedures used to devise them and how they are implemented or enforced (Ostrom 2010). All three sets of external variables impact on an action situation to generate patterns of interactions and outcomes that are evaluated by participants and shape the next action situation (Ostrom et al. 1994, p. 37;Ostrom 2010). Participants as defined in our action situation can hold various positions such as TAO staff, various committee members, village key persons and ordinary villagers (Fig. 1). The action situation filled with actual participants, their positions and the corresponding action, is what we refer to as a particular action arena. We define the action situation as the development planning process of local communities (Fig. 1). 4 Denters and Klok (2003) and Monnikhof et al. (2003) use it for analyzing public consultations; Koontz (2005, p. 459) investigates stakeholder participation in environmental policy by using the IAD framework. Lowndes et al. (2006) examine how the level and style of participation is shaped by locally distinctive rulesin-use. Smajgl et al. (2009) test how the IAD framework can be applied to the circumstances of Australia's outback regions. One case study touches upon community participation in land use decision determined by rules-in-use (Smajgl et al. 2009, p. 50, 74). Van Damme and Brans (2012) use it as a conceptual tool to study the openness of process rules in public consultations.
Looking more precisely into the building blocks of institutions, rules have to be differentiated from strategies and norms (Ostrom 2014). Following the Grammar of Institutions outlined by Crawford and Ostrom (1995, pp. 584-586) and aiming at a common and more precise language for institutional analysis, individual's internalized moral values and social mechanism of enforcement to determine why individuals do or do not follow an institution and thus specify a norm. Participants do acquire most norms in the context of the community, in which the individual interacts frequently, based on an internal value assigned to the norm-conforming action. Rules are similar to norms but rules carry an additional assigned sanction (or-else parameter) if forbidden actions are taken and observed by a monitor (Ostrom 2014). Conceptually, some kind of monitoring and sanctions must exist, yet often this is not known to the people involved. In contrast to Ostrom (2014), where she coded and described the change from a norm-based system to a rulebased system, we are not interested here in the evolution of rules, but rather in their existence and impact on participants in an action situation.
Many strategies, norms and rules are not formulated in an easy-to-detect way or often just exist implicitly (Ostrom 2005: 139). Knowing how difficult it is to detect particularly the or-else parameter empirically (Schlüter and Theesfeld 2010), we will rather give priority to the difference which actions are obliged, permitted or forbidden. Thus, we should be able to assign a deontic logic, i.e., may, must or must not to the formal and informal institutional statements which we outline in the following.
Further, it is important to note that the distinction between rules and norms does not correspond with the distinction between formal and informal institutions. Our cases reveal a high share of informal rules, where the sanctions are not outlined in written form, or a  Fig. 1 IAD framework to analyze the rule-ordered participation in local political decision-making-the community planning process. Note: Dashed line arrow -feedback, i.e. outcomes of a participatory process impact on the rules. Solid line arrow -relationships which we particularly investigate. Source: Adapted from Ostrom (2005, p. 33) punishment for breaking the rules is not set up clearly. Particularly, the distinction between informal community-level rules and norms is empirically difficult to specify (Schlüter and Theesfeld 2010). To cope with that, we follow Ostrom (2014), comprising both (rules and norms) as rules-in-use. Ostrom (2014) suggests, in order to do effective field work and as the best way to determine the rules-in-use, to let participants outline how they would explain the actions and processes to fellow villagers. Thus, what we name in the remainder of the text rules-in-use would be in the narrow sense of the Grammar either a norm or a rule.
Rules can be classified into seven broad types (see box at the left side of Fig. 1) which define in general: who is eligible to make decisions in some arena, which actions are allowed or constrained, which aggregation rules will be used, which procedures must be followed, which information must or must not be provided and which payoffs will be assigned to individuals dependent on their actions (Ostrom 2008;Schlüter and Theesfeld 2010;Crawford and Ostrom 1995). Ostrom et al. (1994, pp. 41-42) proposes a horizontal classification related to the aims of the rules. Transferring these aims from natural resource appropriation to a collective and deliberative participatory process in development community planning, we classified the rules accordingly 5 : 1. Position rules specify a set of positions and how many participants are to hold each position in the planning process; 2. Boundary rules define who is eligible as a participant (e.g., living in the village might be required in order to join a community development meeting), and how a certain person is selected for a position or can leave it. 3. Authority rules specify what a participant in a particular position must, must not or may do. They define the authority a person has in a particular position in a particular point in a decision process. 4. Scope rules define the functional scope and the geographic domain that can be affected by the item being discussed or decided-here a development plan. A permitted or required action is limited to one location or condition. In our case, this is predetermined, because the Tambon development plans apply to a certain geographical outreach. 5. Aggregation rules refer to decision-making procedures, more specifically to the participant's contribution to a final decision about action. This includes the arrangements to aggregate the preferences of villagers into the community development plan and in particular the decision practices within the TAOs in the final issuance of the plan. 6. Information rules define what information participants holding a certain position must, must not or may communicate to other participants holding a particular role at certain points in the decision process-i.e., communication among staff of TAO, among villagers and between TAOs and villagers. The form of how information is distributed is also defined by information rules. 7. Payoff rules refer to the incentives and disincentives in terms of resources (e.g., human resources and funding) available to all position holders to exercise their authority. It is about rewards and external sanctions linked with certain actions.
When studying the rich diversity of specific rules, it is helpful to turn back to the IAD framework and particularly to the action situation itself. Each of the seven types of rules affects an individual working part of an action situation (Fig. 2). This helps to distinguish between the various rules. It is useful to distinguish three levels of rules that affect the actions taken and the development projects selected: operational rules, collective-choice rules and constitutionalchoice rules (Ostrom 2005, pp. 58-59). Operational rules are nested in collective-choice rules and constitutional-choice rules. In this context, the operational rules affect the day-today decisions made by villagers and TAO officials concerning when, where and how to decide on development projects for the villages, what information to exchange or withheld, whom to invite to a meeting and how to structure the meeting. Collective-choice rules indirectly affect operational choices. The collective-choice rules are the rules that are used by village leaders or by higher ranking TAO officials, or external authorities to determine who has what role in making policies, e.g., the Guidelines for Development Planning (Ministry of Interior 2005, 2011. The use of the IAD framework provides important insights how these two levels interact (Smajgl et al. 2009, p. 3). To define the collectivechoice rules, it is important to understand TAOs' work as influenced by the Community Development Department under the Ministry of Interior. Constitutional-choice rules in turn affect the set of operational rules through their effects in determining who is eligible to craft the set of collective-choice rules. The constitutional-choice rules might be important in analyzing the political economy of creating the decentralized development plans in Thailand (Dufhues et al. 2015) or the dimension of how the social and political unrest of recent times influences administrative and political processes. 6 In the present research, we will primarily concentrate on the operational rules, i.e., the daily rules-in-use: (a) within 26 selected villages that prepare a community development plan and (b) within the procedures of four TAOs to finalize the Tambon development plan including budgetary guidelines.

Research area and data collection
Our research took place in the province of Khon Kaen in Northeast Thailand. As mentioned earlier, we focussed on the TAOs, the most widespread decentralized administrative bodies, and thus the arenas where the deliberation and participation ought to take place. 7 In 2009, there were 5767 TAOs in Thailand (Department of Local Administration 2010). There were 158 TAOs in Khon Kaen province, on average eight to ten TAOs per district. We worked in the Nom Pong, the Sri-Chompoo and the Manchachiri Districts.
The Khon Kaen Provincial Department of Local Administration assesses the administrative and management capabilities of each TAO on an annual basis. One part of the evaluation is to identify in how far public participation of villagers in local planning is promoted. A committee from the Department of Local Administration reviews the TAO documents, reports and conducts interviews. 8 While over 80% of the TAOs received the 6 We observe since 2006 the political turbulences in the center of Thailand culminating in the military coup on May 2, 2014. They do have, however, a minor impact on the work of the decentralized administration. First, the decentralization has started 20 years ago and produced stable and well-rooted administrative bodies. Second, the conflict between the ''red and yellow shirts'' is carried out in Bangkok with minor incidents in the rural areas. Third, the issue of decentralization has been low on the agenda of the warring political parties. 7 The TAOs received an autonomous status in 1994 through the Tambon Organization Act. full score in 2010, two were rated with the lowest score (Department of Local Administration 2010).
The TAO assessment was important for the selection of our four TAOs for the case study. We chose the two worst-scored TAOs (names not disclosed but known to authors). From the top TAOs, we selected two particularly well-performing TAOs, the choice was substantiated by an award the TAOs received for more participation. We did so in order to be less dependent on the provincial ranking for our case study selection. We At the first stage of our sampling procedure, this conscious selection, instead of a random selection, provided us with anticipated most contrasting cases. Thus, the first two TAOs are supposed to be less participatory in their involvement of villagers into their planning processes (subsequently named less participative TAOs) in contrast to the latter two, which are supposed to be more participatory (subsequently named more participative TAOs). Some basic characteristics of the four TAOs are shown in Table 1.
The livelihoods in all four Tambons are very similar and mainly based on smallholder agriculture with a focus on paddy rice and sugarcane production. Moreover, many households receive remittances from relatives in Bangkok or abroad. In terms of infrastructure connection, the TAO B. stands out from the other three in its remoteness. But the other less participatory Tambon K. is closest to the provincial capital. The local tax collection serves as an indicator for the local economy and underlines the similarity in all four Tambons. 9 We assume that the way in which the operational rules are set differently in the four TAOs under investigation has an influence on the success of the participation process and the level of local engagement in political decision-making (Pretty 1994  alike in all groups, the exogenous factors-here the rules-determine the difference in success. ''Obtaining information about rules-in-use and their deontic statements (must, must not, or may) requires spending time at a site and learning how to ask nonthreatening, contextspecific questions about rule configurations,' ' Ostrom (1999, p. 53). Coding forms 10 for the IAD framework to analyze natural resource management had to be reframed for this context of villagers' participation in public administration and local political decisionmaking. Obviously, the topic was not about appropriators withdrawing a resource unit, but rather stakeholders taking decisions on development projects. For the operational rules, we rely on inside views from the participants, explaining their understanding of processes and implicit rules that guide decisions. To conduct the analysis, we observed community meetings and conducted 60 semi-structured interviews with the staff members of the four selected TAOs in 2011. In total, these four TAOs serve 53 villages. Second, we conducted a survey with 50 village leaders using a standardized questionnaire. This corresponds to one village leader per village, except three villages which we could not reach. Questions related to the boundary rules were, for example: ''Who in the village participates and who not and why?'', ''Do you usually participate, if not why not?'', ''Why did people not participate?'', ''Are some people actively excluded (who and why)?'', ''Who decides who can participate?'', ''Are you satisfied with these rules?'', ''Are other people in the village not satisfied with these rules?'' In 2012, to validate the information gained from position holders, we ran an additional household survey with 104 ordinary villagers, which were randomly selected in 26 out of the 53 villages belonging to the respective TAOs. The survey contained questions in relation to rule implementation. Regarding the boundary rules, questions were for instance: ''Who can join the community development plan meeting?'', ''Do you usually attend the community plan meeting, and if not, why not?'', ''Are there any people in the village who never take part in the community plan meeting, and if yes, what kind of people and why do they not take part?'' As far as possible, we asked for a clarification if the actors may, must or must not conduct a certain action and for explanation of the sanctions in case of non-compliance. Yet, if people were not aware of the actual rule-setting, we needed to draw conclusions on the obligation, prohibition or permission status of the rule-in-use, based on the usual routines the majority of people described. Further, it turned out that most of the rules we discovered were informal and not written down. Those informal rules-in-use can only be detected if asked to explain a process to a fellow. It is even more difficult to infer on nonmaterial sanctions assigned to this statement, which can derive, e.g., from individual's internalized moral values, like ''I do not feel good, if I do not participate in the village meeting.'' Thus, there is a very fine line between a norm and a rule in a narrow sense.
All answers have been checked for differences between the affiliation to the less and more participative TAOs. If not stated otherwise, all reported answers are statistically significant at least at the 0.05 level.

The action situation: local development planning
The action situation is the central element in the IAD framework. It provides the institutional context that actual participants in an action arena have to face (Ostrom 2010). For analytical purposes and to carefully consider the various participants' interactions, we divided the local development planning process into two phases: Phase 1-''Surveying villagers' needs''-results in the community development plan as an output and represents the first action situation. Phase 2-''Compiling villagers' needs''-represents the second action situation, where needs are integrated into the Tambon development plan prepared within the respective TAO, and finally built together with the budget the guideline for implementing different projects within this plan. We identified two action situations in this way because the institutional context changes and the possible positions to be held by participants are different. Figure 3 presents a summary of the two phases of Tambon development planning and the various collective units.

Action situation I and II in the planning process
The window of opportunity during which direct deliberation between TAO officials and villagers can take place is in the course of the annual community development plan meetings. In these community meetings, the development projects of the forthcoming Tambon development plan should be prioritized by the villagers themselves.
As regards the first action situation, no formal rules exist on how to conduct the community development meetings. All rules concerning those meetings can be seen as informal and rather ''may-statements'' without formal material sanctions in case of nonfulfillment, but most likely addressing internalized moral values or social mechanisms of Policy Sci (2017) 50:675-696 685 enforcement. TAO officials may join the community development meetings. We found that they participate in about 80% of the villages. Some of those villages (40%) conduct two development meetings (the first one without TAO officials and the second one with TAO officials) (Path 1 and 2 in Fig. 3). Action situation II outlines the usual routines following the obligation to condense the individual village project wish list to a workable and budget-conforming TAO project list. Action situation II specifies how the selection of village projects, which are finally implemented in each TAO, is made. After the yearly updates of the community development plans are collected, a draft of the Tambon development plan has to be prepared by the promotion committee and then submitted to the public Tambon civic forum. After revisions are made, it should be submitted via the promotion committee to the development committee and then checked by the president of the TAO who also may change the plan. Finally, it has to be approved by the TAO council meeting. Further, an extra budgetary Planning process phase one -Surveying villagers' needs Planning process phase two -Compiling villagers' needs The president of the TAO, two vice presidents of the TAO, three TAO representatives, three local experts (appointed by the TAO president), three representatives from the government sector (appointed by the TAO president), three to six members of the Tambon civic forum, chief administrator of TAO and the plan and policy analyst. Source: Own figure based on Ministry of Interior (2005) meeting has to be held and the TAO council has to decide this time on the final budget. 11 Although this rough formal procedure is an obligation outlined by the Ministry of Interior (2005), empirical evidence shows that the detailed application of the law is what matters. We found that, in practice, those meetings are reordered, combined or replaced by other meetings with or without village representatives. The deviance from the official meeting requirements (which would be, in the narrow sense, only a norm and no rule as no sanctioning had been reported for a dropped meeting) is higher in the less participatory TAOs. Likewise, the extent to which officials disregard the original village plan is greater.

The participants and positions in the action situations
As stated earlier, the three important internal working parts of an action situation are the participants, the positions they are respectively assigned to and the set of actions that participants can take (Ostrom 2005, p. 38). While in Phase 1 of the development planning process ordinary villagers can influence the outcome (as participants of a community development meeting), during the second phase ordinary villagers no longer occupy a position in the planning process (Fig. 3). In Phase 2, all positions are held by elected representatives, assigned key persons, politicians and civil servants, who of course may have originally been ordinary villagers. The TAO consists of two units-the TAO council and the TAO committee. We can confirm the findings of Badenoch (2006, p. 45), who stated that besides such responsibilities as drafting the budget, implementing plans or monitoring TAO committee's activities, the main responsibility of the TAO council is approving the Tambon development plan and the corresponding budget. The responsibilities of the TAO committee are to manage development activities, propose development and budget plans to the council and to report progress to the council.
The TAO council has to be composed of, among others, two directly elected representatives from every village. The council is the immediate source of local accountability. The chairman of the TAO council, the vice-chairman of TAO council and the secretary of the TAO council must be elected by the TAO council members from their membership. But, in addition the post of secretary of the TAO council may be filled by a non-council member, who is usually the Chief Administrator of the TAO, a paid staff member called the Palad. Typically, the Palad ought to have a good knowledge of administrative procedure, and it was found to be the reason why all the TAO councils in our sample filled this position with the Palad.
The TAO representatives are a new sort of leadership in the village due to their role in accessing the Tambon resources through drafting the Tambon development plan (Badenoch 2006: 88). Village committee members (the village head, due to his position, and other elected villagers) must in turn elect members into the Tambon civic forum.
TAO civil servants also influence the plan making process. Those staffers include the Palad, the plan and policy analyst, the head of the financial division, and the head of the public works division. Local experts, such as school directors, heads of public health centers or hospitals, may also be assigned to various subcommittees preparing the Tambon development plan, as are representatives of the Tambon civic forum and committee. The plan and policy analyst, in particular, has a central role in the process. He/she, for instance, shall record the proposed projects from the village community meetings. He/she shall draft the Tambon development plan several times. The first draft may include a budget and a time frame for individual projects suggested at the community meetings. This should be an input for the promotion committee whose members may adjust the budget and time frame of each project and may make a pre-selection. The development committee must decide subsequently whether or not to keep those projects in the plan. That means in fact, if the development and promotion committees are not holding a meeting-as in one of our less participatory TAOs-then the initial proposal drafted by the plan and policy analyst is what goes directly to the TAO council. As we already found here slight exceptions from the obligation-without reported formal consequences-we define these rules according to their deontic logic as may-statements, i.e., permissions or optional provisions.

Rules
In the following, we will present the rules defining an action situation along the lines of the seven types of rules (Ostrom 2005(Ostrom , 2010 and as applied to participation in development community planning as outlined in ''Rules in participatory procedures'' section.

Collective-choice rules: action situations I and II
The Department of Local Administration, which is part of the Ministry of Interior, has set up the guideline for the creation of the Tambon development plan and the development budget. 12 These formal requirements determine that the budget is the final output of the planning process and must contain the projects which will be actually implemented and their corresponding budgets (Office of the State Council 1998). The Tambon development plan and the development budget must be reviewed annually to evaluate the development strategies, methods and measures that had been chosen earlier and to assess whether they are still effective. We understand these guidelines as collective-choice rules. They define the participants' room for maneuvering.

Operational rules: action situation I-surveying villagers' needs
As there are no formal rules for the way the community plan meeting has to be organized, we will depict the rules-in-use, which we found most frequently across the 50 villages studied. It is difficult to differentiate such rules by the type and seriousness of sanctions associated with them, as particularly informal sanctions relying on individual or social pressure are very hidden from a researcher (Schlüter and Theesfeld 2010). Table 2 depicts the descriptive statistics of the collected data underlying the selected rules-in-use.
Boundary rules The participation in the community development meetings is voluntary; every person from the age of 15 on may participate. Village leaders such as the village head and TAO representatives usually take part, which indicates an obligation. In some cases, the TAO has explicitly invited village leaders or certain villagers. Some people never join and some of them state that they do not care about such meetings. Another decisive factor is timing. About a quarter of the villagers do not care when the meeting takes place, but some state they are unable to join the meeting during daytime due to their working hours. This especially concerns poorer households where adults have to work often outside the village. Still, a meeting in the evening is preferred only by about 40%. In some cases, the community development meetings are held jointly for several villages. Participation of more remote villagers is low in those joint meetings, because of travel distances.
The representativeness of the community development meetings varies greatly between the villages, but on average the number of meetings is lower and, as shown in Table 2, the attendance rate is indeed lower in the less participatory TAOs. Likewise, data show with statistical significance that more villagers in the problematic TAOs state that some people never participate in any community development meeting (74% compared to 39%).
Generally, however, among boundary rules specifying timing, merging of meetings or eligibility attributes of participants, no exclusionary impact can be found toward particular village groups. When considered together with the information rules on timing and outreach of the announcement (presented in the following), the boundary rules do have an impact on the attendance rate.

Information rules
The village head has to inform the villagers about the community development meeting. We can reproduce the findings of Fung (2003) who states that the number of participants in community events depends most of all on the ability of organizers to mobilize individuals. One rule deals with the way the information is spread. Villagers are usually informed via public loudspeakers, which are located at a central place in the village. In almost all of the 26 villages which have been surveyed more intensively, the interviewees were informed about the date and the time of the community meeting via village loudspeakers. A small share of people did not take part in the meetings because the loudspeaker is not loud enough to reach their houses. In a few cases, villagers are also personally invited by an invitation letter (about 5-6%), which shows that this permissionto invite in written form-exists as well. About 8% of the village heads stated that they have to ensure that every villager is invited, besides using the loudspeaker. Interestingly, this rule has been only mentioned in the more participatory TAOs. In total, 96% of the villagers were content with the way the meetings are announced.
Yet, meetings are announced on a rather short notice (on average 1 day before). There is statistically significant empirical evidence that villagers living in the more participatory TAOs are informed earlier (Table 2). Timing seems to be a crucial issue as a quarter of villagers mentioned that they would like to be informed earlier. A share of 30% of the villagers mentioned this in the less participatory TAOs, compared to 19% in the other villages. This particular difference could not be shown with statistical significance.
Another important information rule defines how and when villagers may start discussing development project proposals which shall be ranked. While over 80% of the village leaders stated that villagers talked about the development projects before the meeting, only 20% of the interviewed villagers stated the same. With statistical significance, we can show that the share of villagers making this statement is higher in the more participatory TAOs. We infer from that, that in those villages, people are permitted to know about the project proposals beforehand.
Payoff rules An additional payoff refers to incentives and rewards in terms of, e.g., funding or side-payments available to a position holder. One aspect showing an immaterial social reward is that those people who regularly take part in the community development meetings are more highly respected by the other villagers. In only three villages, which interestingly belong to the less participatory TAOs, it was mentioned that this would not be the case.
In the more participatory TAOs, the attendance rate is almost identical across the villages despite the show-up bonus, of 50 Baht (1.4 US Dollar), that was paid in one of the studied villages. Apart from that, in rare cases, candy or coffee is served, representing the only material payoff.
Aggregation rules An important aggregation rule is represented by the widespread voting procedure for projects: an open ballot, by raising hands. It is worth scrutinizing this operational rule. We would like to provide two cases to exemplify the possible options in such ranking and voting rules.
In one village located in a less participatory TAO, the project proposals were listed on the wall by order in which they were proposed. Then, the TAO official asked: ''What project do you like to have as number one?'' A villager would tell the name of a project and the moderator would ask whether all villagers agree to have it as number one. The villagers would vote. If a simple majority was found, this project would be accepted with that rank. That means the person who mentions a project quickly has the highest chance to have that project get a high rank. Furthermore, projects cannot be ranked against each other within such a procedure. In comparison, in a village located in a more participatory TAO, all proposals were listed, at first. Only thereafter, the question was: ''Who votes for project No. 1, 2, 3, …?'' Everybody had three votes and the proposal with the most votes would win. Unfortunately, it was not possible to collect such observational data at a larger number of meetings to allow generalization on these observations. Yet, it shows what impact such little procedural rule differences could have.
Another aggregation rule defines that elites can partly preselect development projects. According to villagers such a pre-selection happens in 10% of the meetings. However, in the less participatory TAOs this was mentioned more often. Yet the difference of 16% compared to 9% is not statistically significant. A corresponding aggregation rule is that elites may erase proposals from the list. A share of 9% of the villagers noted that during the meetings village leaders strongly interfere in the process, e.g., by deleting projects they dislike from the plan. Quite revealing here is that in the more participatory TAOs significantly less villagers mentioned this (4%) compared to (27%) in the other villages. This can be shown with a statistically significant group difference. When we combine both forms of interference (pre-selection of projects and direct interference at the meetings), the share of villagers who note such customs reaches 50% in the less participatory TAOs as compared to 9% in the more participatory TAOs. The survey of 50 village leaders confirms such customs at large.
In sum, 29% of the villagers in the less participative TAOs stated that they could influence the village development plan only with great difficulties or not at all, compared to only 18% who stated such difficulties in the more successful TAOs. However, based on our sample, we cannot substantiate this with an appropriate significance level.

Operational rules: action situation II -Compiling villagers needs into the Tambon development plan
Concerning the operational rules within the TAO administration, we focus on the rules specifying how a person can gain certain positions, as this has been regarded as nontransparent. Further, we focus on the aggregation rules, the most challenging to be revealed empirically and likewise the ones exemplifying the detailed decision-making procedure within the TAOs.

Boundary and position rules
In both the more-and less participatory TAOs, there is a lot of confusion among the villagers as to their functions in a particular meeting, and how long they may fill a certain position. One issue is that the various committee meetings are often combined, postponed or replaced by a different purpose meeting. There is empirical evidence that in the less participatory TAOs more stages of committee meetings are omitted. In that regard, one rule-in-use in a less participatory TAO is instead of letting the individual participants sign the participants' list, the Palad may officially approve that all members were present. The knowledge gap about the existence of various positions, unclear membership in certain committees and concerns with inconsistency in the names listed on committee meeting protocols are particularly tangible in the less participatory TAOs.

Authority and aggregation rules in the promotion and development committees
The promotion committee has to draft the development plan, including time frame and budget of each single project proposal. The committee then has to select the first three to five projects from each community plan (authority rules). The aggregation rule to reach an agreement in the more participatory TAOs is that after a village has received a big project, the village has to wait a year or two until they receive another one. In cases where two villages compete for funds, the promotion committee may consider those projects first which will create benefits for more villagers. In addition, representatives from government agencies in the committee may propose their own projects and add them to the draft plan.
In the less participatory TAOs, village members of the promotion committees reported that they redo the plan during this meeting, without much reference to the original community plans. Some state that they keep only those projects which they roughly kept in mind, without considering the written records. No sanctions have been reported for such actions, which lets us draw the conclusion that this permission to redo the list is the rule-inuse.
TAO officials could not outline the main features of development committee meetings in terms of objectives and procedures, referring to rather no strong obligations (maydeontic). The only exception is that most of them state that in both committees changes in the procedures must be accepted by a simple majority vote. Thus, we interpret the latter as an obligation (must-deontic).

Authority and aggregation rules in the TAO council
Interviews with TAO representatives revealed that the TAO council meeting, which has to decide on the Tambon development plan (authority rule), almost always follows the recommendation of the president of the TAO and the promotion and development committee (aggregation rule). Thus, the decision-making power at the TAO council can be regarded as a pro forma one. As pointed out by Walker (2008, p. 88), the increased financial budget autonomy of the TAOs has aggravated budgetary competition between the ten to fifteen villages in a single TAO. Empirical observation revealed that TAO representatives simply follow recommendations of the higher authority-supposedly to avoid conflict responsibility over the distribution of development budget among different villages (aggregation rule). This kind of social pressure (non-material sanction) can be regarded as the determinant why people follow this informal rule-in-use of simply letting the proposed plan pass.

Conclusions
The integrative approach of the IAD framework and its central concept of the action situation were very valuable in dividing the decentralized and participatory development planning process in rural Thailand into two phases. Namely, first, the village prioritizing process for development projects and second, the TAO decision-making processes toward a final development plan. It turned out to be extremely insightful to disentangle the respective participants and their involvement in this kind of local political decision-making process according to various types of rules (Ostrom et al. 1994). The study of rules by use of the IAD framework required first reordering of the rich empirical material according to rule types, rather than chronologically along the lines of the participatory procedure.
We reached conclusions on preconditions, i.e., necessary up-front rule changes that would allow for better stakeholder participation within the process of local development planning. Drawing concrete policy lessons for the Thai case, we stress that spreading information about aim and scope of certain committees and about possibilities to take and abandon positions in such committees is crucial.
The more participatory TAOs-administrative bodies which are known for success in achieving participation-follow in general more standardized rules. In contrast to the literature on interactive policy making, Van Damme and Brans (2012) have concluded that it is not particularly the openness of process rules in public consultation processes that are able to deliver good overall outcomes. They found that clear and detailed rules on who needs to be consulted, about what, and where seemed to be important for the performance of a consultation process. The trade-off is that the managers of the process lose autonomy in making the participatory process fit to the situation at hand. Yet, better designed formal rules can only control or steer that part of haphazard participatory processes which is due to unawareness while the intentional deviation from the rule to secure personal advantages remains more difficult to control. Our research further substantiates the advantages of some pre-defined obligations as well as a clearer distinction between obligations and permission. As shown here, openness of rules, which otherwise can be seen as flexibility from the managers' perspective, can easily turn into fuzziness and lack of clarity from the perspective of the local population. In the more participatory TAOs, the rules are perceived as more transparent and can be recapitulated by interviewees. The actual rules-in-use enabled a higher number of participants and thus a larger range of opinions in the committee meetings, restraining personalized decision-making.
Disentangling the various rules and participant involvement in the planning processes, it became evident that not all rules are equally important in supporting participation of local communities in development planning. As we have seen from the detailed studying of rules in the two phases of the development planning process, the aggregation rules within the higher authority, in particular-specifying how to transform actions into a final outcome, are decisive for letting the locals' propositions come into being.
We could highlight an impact of both formal and informal rules of the decision-making process on the actually achieved level of participation. Although most rules are rather informal, we can draw some conclusion on their deontic logic, differentiating obligations, prohibitions or permissions, which means the de facto must or may interpretation of the rules. Based on the interviewees' explanations of the routines, we identified a high share of may-statements and the severity of external and internal sanctions expressed.
We can further show particular differences in rules, some of them being statistically significant, depending on whether a TAO is ranked as more or less participatory. For instance, in the less participatory TAOs the attendance rate in the community meetings is lower (boundary rule), villagers are informed about a planned meeting with a shorter notice (information rule) and more villagers mention that elites interfere in the project selection process (aggregation rule). This clearly shows the importance of the exogenous rule settings in determining the deliberative democratic process, compared to the endogenous cultural factors which are almost equal across all 50 villages studied.
Besides the Thai TAO case-based output, this study has wider significance for the analysis of participatory processes in general. Rules can be used as an analytical tool to find a window of opportunity to intervene and facilitate participation. We can thus provide better policy recommendations of where and how to support the participatory efforts, instead of discussing specific habits of individual administrative bodies and committees. It seems fitting that the IAD framework, which originally grew out of municipal public administration studies in the 1970s before it was elaborated and widely applied to natural resource management, can now be applied fruitfully once again to deliberative democratic processes.
The application of the framework allows us to zero in on key factors that policy makers can actually address in order to increase public participation. Local-level policy makers can do little to change the sociocultural setting, even less to change the communication habits of the respective societies, yet they might be able to re-shape the collective-choice and operational-choice-level rules that prevail in participation measures in local decisionmaking. Further applied lessons encompass that information on administrative procedures have to be made more accessible and public administrators should receive procedural trainings (e.g., in inviting to meetings, writing protocols and managing open votes). Both steps are likely to bring more clarity to the decision-making processes and thus improve deliberation and public participation.