Abstract
Participatory budgeting is a democratic process through which citizens make decisions about budgeting. It has received attention in public administration literature. Many scholars apply “participatory budgeting” too broadly. They conflate nominally participatory practices with those that emanated from the Global South and now expand throughout the United States. This article explores various applications of “participatory budgeting.” It also presents a heuristic framework, developed out of the work of Arnstein and Fung, which assists practitioners and researchers in identification and evaluation of participatory processes, including participatory budgeting. This framework is applied to four cases to illustrate determination of participatory budgeting practices.
Introduction
Participatory budgeting is on the rise. The practice originated more than 40 years ago in Porto Alegre, Brazil, and by 2010 had expanded to nearly 1,500 municipalities worldwide (Sintomer, Herzberg, Allegretti, & Röcke, 2010). It came to the United States in 2009. Since starting in Chicago’s 49th Ward, it has expanded to six U.S. cities. A basic working definition for participatory budgeting is “a democratic process in which community members directly decide how to spend part of a public budget” (Participatory Budgeting Project, 2017). Possibly because the practice is new to the United States, American public administration scholarship uses the concept in many ways, not all of which cohere with one another or with the Brazilian projects that first termed themselves participatory budgeting.
Some public administration scholars use participatory budgeting as a blanket term to refer to all modes of citizen participation, including surveys, simulations, citizen budget committees, focus groups, budget forums and meetings, direct citizen interaction, and neighborhood district meetings (Franklin, Ho, & Ebdon, 2009; Guo & Neshkova, 2013; Kim & Schachter, 2013; Zhang & Liao, 2011, 2013; Zhang & Yang, 2009). The use of “participatory budgeting” in this broad way creates confusion because it disconnects the concept from the specific practices that emanated from the Global South and are now growing throughout the United States. If this term is used too generally, there is no way to conceptually differentiate what is especially unique and valuable about those practices found in Chicago, New York, Boston, Vallejo, and other U.S. cities.
The term has also been used problematically when denoting a specific practice, such as communication patterns between citizens and administrators (Liao & Zhang, 2012), intermediary organizations (Justice & Dülger, 2009), and a university budget committee (Rossmann & Shanahan, 2012). Although each practice involves citizen participation, labeling them “participatory budgeting” loses sight of what is distinctive, valuable, and potentially transformative in the idea. One purpose of this article is to encourage focus in the research literature’s conceptualization of participatory budgeting while remaining attentive to and appreciative of its variations in implementation. This piece aims to show important similarities that persist in one strand of what has been called “participatory budgeting.”
That said, participatory budgeting takes a number of different forms; there is no one single practice uniquely called “participatory budgeting.” It is more accurate to refer to a family of practices that share key characteristics of scope, process, and outcomes of public participation. This article presents a heuristic framework called the “Modes of Participation” to help researchers and practitioners more easily identify and evaluate participatory possibilities in democratic processes. One useful application, which this article carries out in four test cases, is examining the participatory budgeting family’s range of practices. The framework builds on and revises Shelly Arnstein’s (1969) “Ladder of Participation” and Archon Fung’s (2006) “Democracy Cube.” The Modes of Participation preserves what is best in the models offered by Arnstein and Fung, but it has two clear advantages: First, it presents scales that enable easy evaluation of practices along continua of more and less robust forms of citizen participation. This stands in contrast to much public administration research, which frequently treats participatory mechanisms in terms of dichotomous, fixed variables. Second, each of the framework’s scales contains clear minimum standards for identifying instances of participatory budgeting. While these scales are used in this article to identify and distinguish varieties of participatory budgeting, they can be applied to evaluate many different forms of citizen participation in governance.
There are four Modes of Participation that can be usefully evaluated in all forms of citizen participation in budgeting: (a) participant selection, (b) communication, (c) decision making, and (d) authority. With the specific case of participatory budgeting, the practice ideally meets a minimum standard for each of these Modes. Participant selection, at a minimum, must be open to all adult residents and inclusive of underrepresented groups. Communication must have some form of deliberation. Decision making must be citizen-driven and collaborative: This means citizens set the agenda, create priorities, come up with projects, and collaborate with government and civil society organizations. Finally, citizens should have significant final authority over outcomes. The priorities they set or projects they select are consistently implemented. This article presents the Modes of Participation as a way to evaluate various forms of citizen participation in budgeting and then shows how one kind of citizen participation, called “participatory budgeting,” can be graded on the Modes. Such evaluation highlights participatory budgeting’s distinctive features when compared with other forms of citizen participation.
The article is organized as follows. The section “What Is Participatory Budgeting?” briefly describes participatory budgeting as a family of related practices. The section “Citizen Participation in Budgeting: Renewed Energy on an Old Issue” presents a short history of citizen participation in public budgeting. This leads to the section “Conceptual Confusion,” which describes in detail and then critiques the current breadth of application of “participatory budgeting” in American public administration literature. In the fifth section, a new heuristic framework, the “Modes of Participation,” is presented. The section “Some Examples” applies this framework to analyze four cases of citizen participation in budgeting. In the final section “Conclusion,” the merits of this particular framework are discussed for both researchers and practitioners.
What Is Participatory Budgeting?
Because participatory budgeting takes a variety of forms, there is no one accepted standard definition. Some are broad like Wampler’s (2007) “Participatory budgeting is a decision-making process through which citizens deliberate and negotiate over the distribution of public resources” (p. 21). This definition covers a wide array of practices and possibilities for citizen involvement. It is applicable to processes from budget advisory committees to participatory budgeting as first practiced. Other definitions are narrow and generally reflect only the original participatory budgeting experiments in Brazil. Consider Goldfrank’s (2007) definition:
Participatory budgeting is a process that is open to any citizen who wants to participate, combines direct and representative democracy, involves deliberation (not merely consultation), redistributes resources toward the poor, and is self-regulating, such that participants help define the rules governing the process, including the criteria by which resources are allocated. (p. 92)
Here, key conditions of inclusion, deliberation, and citizen-driven authority are basic to participatory budgeting, and it includes a social justice dimension. However, as Goldfrank also noted, defining participatory budgeting too narrowly excludes many vibrant participatory practices. Sintomer et al. (2010) estimated that, depending on the definition used, there were between 795 and 1,469 participatory budgeting sites in 2010. If there can be general agreement about central features of this practice, debates are then limited to narrower, marginal questions, such as the role of institutions and nongovernmental organizations.
In terms of practice, each variant of participatory budgeting is implemented within a particular context (formal and informal political system, legal framework, budgeting environment, political structure, civic culture, policy framework, institutional capacity, etc.). These concrete differences make abstract definition difficult. Moreover, there is a great deal of variety among different forms of participatory budgeting.
Of course, at a high enough level of generality, many of the practices called “participatory budgeting” are identical: Residents work together to allocate public money. On closer examination, however, the differences in these implementations are striking, with possible variance in each of the process’s components. After the United Nations, World Bank, and Inter-American Development Bank named participatory budgeting a “best practice” for local governance, significant worldwide expansion occurred. In some cases, existing budgeting practices, such as participatory planning, rebranded themselves as participatory budgeting. In others, there was a diffusion of innovation, sometimes adopting the Porto Alegre model, sometimes adapting this practice for their particular circumstances.
With such a variety of programs bearing the name “participatory budgeting,” one might think there is no clear way or even value in trying to give it a definition. There are theoretical resources available, however, that encourage thinking about meanings in a more holistic and practice-driven way. Such a reorientation is useful when grappling with “participatory budgeting.”
The Austrian philosopher Ludwig Wittgenstein suggests in his Philosophical Investigations that meanings are found by seeing how words are used in language-games, the actual practices of deploying, interpreting, and responding to words. As an illustration, Wittgenstein asks readers to consider the many activities understood as “games,” noting that examination of them all—from solitaire to Monopoly to baseball—will not reveal a common feature shared by all, except the feature of being called a game. Such an evaluation does not reveal an essence or a determinate meaning, but it also does not mean that no definition whatsoever can be offered. Rather, “we see a complicated network of similarities overlapping and crisscrossing: sometimes overall similarities, sometimes similarities of detail” (Wittgenstein, 1953/2009, §66). The term he chooses to pick out these similarities is family resemblances. In the same way, various descendants of the same parents and grandparents share some but not all features of personality and appearance, yet are still identifiably one family, a word’s many usages form a family of meanings. The boundary between what is included and excluded by some concept is always drawn for some purpose (Wittgenstein, 1953/2009, §68-69).
Because “participatory budgeting” describes one such family of practices, descended from Porto Alegre, this Wittgensteinian way of talking about it can be helpful. The purpose guiding this article’s definitional work is helping researchers and practitioners distinguish different forms of citizen participation in budgeting. It is a mistake to look for a single common attribute (e.g., citizen participation in budgeting) to define the concept. Rather, we want to highlight what is unique in participatory budgeting. To do this, we introduce the Modes of Participation.
Citizen Participation in Budgeting: Renewed Energy on an Old Issue
Today, there is widespread agreement among researchers and practitioners that citizen participation in local budgeting is desirable (Berner, Amos, & Morse, 2011). These discussions, however, are nothing new. Berner has noted that “the question of how citizens should be involved in government is as old as government itself” (Berner, 2004, pp. 410-411). This long debate has alternated between periods that encouraged citizen participation and ones that prioritized expert administration. Key moments in these debates include the progressive reforms in the early 1900s, the Great Society programs of the 1960s, and trends in New Public Management (NPM) in the 1980s and 1990s.
In the past 10 years, there has been renewed interest in citizen participation in political decision making, particularly in the area of public budgeting. The years 2005 to 2014 showed an upward trajectory in the number of articles referencing “participatory budgeting.” Such articles were searched for in the top 10 budgeting and finance journals identified by Bernick and Krueger (2010). No articles on the topic were published in these journals prior to 2005. After searching “participatory budgeting” as a keyword, a total of 41 articles were identified, but only 12 of them provided elaboration and not simply identification of the phenomenon. Of these, only two pieces were published between 2005 and 2008, while 2009 and 2011 each saw three. One additional article was published in each of 2012 and 2013, and 2014 delivered two.
Much attention in the public administration literature focuses on citizen participation in formal budgeting. Despite this renewed emphasis, however, governments are struggling to find ways of including citizens in traditional forms of budgetary decision making (Ebdon & Franklin, 2006). In the United States, participatory budgeting is a quasi-legislative practice and therefore has resonance with new attention that has been shown to the concept of governance (Boyte, 2005). Bingham, Nabatchi, and O’Leary (2005) define “governance” as a process that “seeks to share power in decision making, encourage citizen autonomy and independence, and provide a process for developing the common good through civic engagement” (p. 547). Instead of focusing on traditional decision-making hierarchies, governance-theorists highlight horizontal networking as a new structure of decision making consisting of public, private, and nonprofit organizations. Bingham et al. envision this as a “new face” of governance, in which citizens “engage in activities that range from the legislative or quasi-legislative to judicial or quasi-judicial” (p. 547). Still, questions about the relationship between traditional budgeting and governance remain and signal uncertainty in the public administration literature on the topic of participatory budgeting.
Conceptual Confusion
American public administration research literature applies the concept of participatory budgeting in a broad way that can lead to confusion. Why? One possible reason is that there was a momentary confluence when public budgeting literature was swinging toward interest in citizen participation while a new family of practices called “participatory budgeting” was spreading from the Global South to Asia, Europe, and then finally North America. These two trends may have conceptually commingled. Another reason might just be definitional. As stated above, there is no one accepted definition.
Public budgeting scholars in the United States tend to rely on broad definitions and do not draw on actual participatory budgeting practices globally (an important exception is Pinnington, Lerner, & Schugurensky, 2009). For Zhang and Yang (2009), “participatory budgeting refers to a situation in which the government invites citizens’ input during the budget process and allows citizens to influence budgetary decisions” (p. 289). This definition includes components of input and influence, but it does not give any sense of quality or degree. To what extent do citizens actually influence decisions? What is the form of input? In addition, their definition neglects key issues of authority, inclusion, and deliberation. Most importantly, scholars such as Zhang and Liao (2013), Kim and Schachter (2013), Zhang and Liao (2011), Franklin et al. (2009), and Zhang and Yang all use participatory budgeting as a blanket term that covers many and possibly all modes of citizen participation. A good example is Zhang and Liao’s statement:
The mechanisms of participatory budgeting include public hearings, citizen surveys, advisory boards, and forums or workshops open to citizens, as well as efforts aimed at encouraging citizens to speak out about the budget at regular meetings and posting budget materials on the Internet. (pp. 282-283)
Franklin, Ho, and Ebdon’s (2009) article “Participatory Budgeting in Midwestern States: Democratic Connection or Citizen Disconnection?” never defines “participatory budgeting,” despite its presence in the title. Instead, they analyze the extent to which municipalities use and city officials value participatory mechanisms such as public hearings, special budget meetings, direct citizen interaction, neighborhood meetings, surveys, televised hearings with call-ins, focus groups, citizen budget committees, and simulations. Although the particulars vary, the general trend in the literature is to claim that these mechanisms encompass participatory budgeting. While these individual practices may be part of a participatory budgeting process, including all of them under the participatory budgeting umbrella obscures the concept’s distinctive meaning and the range of practices it usefully covers.
In some other cases, the term participatory budgeting is used in relation to a specific practices, such as communication patterns between citizens and administrators (Liao & Zhang, 2012), intermediary organizations (Justice & Dülger, 2009), and a university budget committee (Rossmann & Shanahan, 2012). Justice and Dülger (2009) is a good example of the conceptual moves that may lead to the term’s overly-broad application. They analyzed how intermediary organizations facilitate citizen participation in budgeting. Their goal is to expand Ebdon and Franklin’s (2006) logic model for citizen participation by including effective transparency and fiscal information intermediation. Indeed, these two important variables should be added to any logic model of citizen participation. However, Justice and Dülger mistakenly claimed that Ebdon and Franklin presented a model for participatory budgeting in the United States. Ebdon and Franklin never used the term, preferring the more precise “citizen participation in budgeting.” Justice and Dülger then faulted “contemporary U.S. participatory budgeting literature” for not including transparency and information intermediation (p. 268). As a corrective, they offer their own logic model: “A Selective Model of the Democratic Logic of Participatory Budgeting/PPEM” (p. 262). This is a clear case of conceptual slippage. The authors take a general article about citizen participation in budgeting and then use the term participatory budgeting to reference it. This conceptual shift is magnified when they then make a larger claim about an entire field of literature. The consequence is that the features identified in Ebdon and Franklin, with a modification, are then seen as constituting the field of “participatory budgeting.” This is an example of the conceptual confusion this article hopes to clarify.
So what should be done? Do we simply need a better definition? Because participatory budgeting takes a variety of forms, a tighter definition might actually be counterproductive. With any strong and narrow definition, exclusion is a consistent risk. A variety of definitions have been offered for “participatory budgeting.” Many of these overlap. Some seem to be drawn too broadly, including practices which are participatory in only a trivial sense. Other definitions appear too narrow, ruling out any practice that is not identical with the paradigmatic Porto Alegre participatory budgeting progenitor. This situation leads to confusion in use by researchers and practitioners. In response, some theorists might attempt to put forward a new definition, redrawing the conceptual lines to express more clearly just what participatory budgeting does and does not involve. The consistent failure of theorists to delimit this concept successfully suggests the task’s difficulty. Indeed, there are practical as well as theoretical reasons to believe that as attempted thus far, this kind of defining is not a worthwhile enterprise. Part of what is unique in participatory budgeting is its history as it developed, diffused, and adapted into a global phenomenon. This uniqueness cannot be reduced to a single feature, so it is better to identify minimum standards within four overlapping Modes of Participation.
Modes of Participation
This section presents a heuristic framework called the Modes of Participation. It is not a full-fledged research model that can be operationalized; rather, it is a conceptual framework designed to help researchers and practitioners distinguish truly participatory practices from other forms of citizen participation that merely appear participatory. It is composed of four scales, evaluative criteria for each, and minimum standards that serve to identify distinctive instances of participatory budgeting. Any form of citizen participation in budgeting can be evaluated on the Modes of Participation to disclose its degree of participatory robustness.
This framework builds on Sherry Arnstein’s (1969) “A Ladder of Citizen Participation” and Archon Fung’s (2006) “Varieties of Participation in Complex Governance.” Fung’s schema promised more precision than Arnstein’s for examining different dimensions of participation and also gave a visualization of interacting dimensions. However, his model obscures as much is it illuminates. Therefore, the Modes of Participation generally follow Fung but make four important modifications: (a) decoupling modes of decision and communication, (b) returning to independent scales, (c) offering minimum standards, and (d) focusing exclusively on lay participation.
Arnstein’s “ladder” is a one-dimensional hierarchical representation with eight steps: manipulation, therapy, informing, consultation, placation, partnership, delegated power, and citizen control. Her ladder focuses on citizen power, examining the extent to which citizens can actually affect procedural outcomes. This framework and conceptual vocabulary was helpful to distinguish the ways in which citizens’ power might be undermined under the rubric and language of participation. She claimed that “[t]here is a critical difference between going through the empty ritual of participation and having the real power needed to affect the outcome of the process” (Arnstein, 1969, p. 216). Most importantly, she provided a conceptual tool to expose that many Great Society programs included citizens as tokens despite claiming “maximum feasible participation.” Her typology allowed theorists, practitioners, and ordinary people to understand the relationships and disconnects to be found between nominal citizen participation and effective citizen control.
Four decades later, Fung argued that Arnstein’s single scale is incomplete and imprecise in capturing the multifaceted nature of citizen participation. This is correct: A single ladder is too reductive. Its attractive simplicity blunts its analytic possibilities. A full measurement of participation requires consideration of many interplaying features. He therefore extended Arnstein’s ladder with a multidimensional “Democracy Cube.” In place of the single scale from less to more citizen power, Fung presented three dimensions of participation: Participant Selection, Communication and Decision, and Authority and Power. Each dimension has an evaluative scale. His innovation was to place these three scales into an interacting institutional design space. This setup allows more complex analyses of both individual dimensions and their interactions.

Archon Fung’s democracy cube (Fung 2006, p. 71).
However, this three-dimensional model is problematic in four major ways. First, Fung combined the modes of communication and decision into one scale. For Fung,
These six modes of communication (first three) and decision making (second three) can be arrayed on a single dimension that ranges from least intensive to most intensive, where intensity indicates roughly the level of investment, knowledge, and commitment required of participants. (Fung, 2006, p. 69)
Communication, for Fung, is less intense than decision making because public discussions may have no effect. Participants at city council meetings might express views in public comment sessions, but there is little chance their input will change any public policies or actions. However, the unified Communication and Decision scale, if evaluated according to Fung’s definition of intensity, is incoherent. There is little doubt that participating in an intensive deliberative forum, which exemplifies both communication and decision-making features, such as British Columbia’s Citizens’ Assembly on Electoral Reform, requires more investment, knowledge, and commitment than voting in a referendum (Warren & Pearce, 2008). Here, a group of randomly selected representative participants worked intensely for weeks to recraft the province’s electoral system, though they did not have final implementation authority. Fung likely wanted to use only three axes due to the difficulty of visually presenting a fourth. And yet, both communication and decision making can be more and less intense (with respect to knowledge and buy-in) as well as more and less participatory. These factors vary independently, so collapsing communication and decision making into one category is mistaken.
Second, Fung’s combination of his axes into a three-dimensional figure raises issues. The construction allows for more finely grained, interactive evaluations of participatory projects than Arnstein’s unilinear ladder. Fung’s model promises to analyze how these axes interact. However, as constructed, this model is confusing, especially for practitioners. Most importantly, it is not clear how Fung selected each of his axes’ orientations. While “communication and decision” and “impact of public participation” are both placed with their most robust elements (“technical expertise” and “direct authority”) nearest to the graph’s origin point, “participant selection” begins with its least inclusive element. That two scales go from most to least robust and the third goes from least to most inclusive makes interpretation difficult.
Third, while Fung’s model effectively displays how values on three axes coordinate to slice out a cube-section of the available territory, it is less effective when used for evaluative purposes. It is not immediately apparent whether a larger figure is preferable, whether figures farther from the graph-space’s origin point indicate more desirable (viz. more democratic) practices, or even how these cube-figures are to be compared against one another.
Fourth, Fung’s model evaluates practices’ intensity of participation regardless of who is participating. Consideration of who participates is limited to one axis, participant selection. Fung is concerned to offer a broadly applicable schema for examining all forms of political participation. This has its uses, but it obscures the value and presence of lay citizen participation, especially by those individuals who are ordinarily marginalized.
In light of these problems, modifications seem appropriate. Yes, participatory practices cannot be gauged purely in terms of a single scale of more and less, in the manner of Arnstein, but too much combination and interplay between scales can muddy the waters, making it unclear which practices should be encouraged. This can be especially difficult for practitioners. Therefore, the Modes of Participation offers four distinct heuristic ladders, focused on the participatory possibilities of lay citizens. Each of these includes at its bottom political practices and elements that are least encouraging of robust participatory involvement. At the ladders’ tops are those programs and features that most encourage this involvement. Each scale includes a minimum standard for participatory budgeting. Practices that meet each minimum clearly are participatory budgeting; practices that meet none of the standards clearly are not. Those that meet some standards but not others, or meet a standard in only a limited phase of their process, should be scrutinized carefully, perhaps with further empirical investigation. This new evaluative tool allows analyses that are both simple in the sense of intuitiveness and complex in the sense of multifacetedness.

The modes of participation.
The Mode of Participant Selection
The first scale of this article’s framework, participant selection, largely replicates Fung. The least inclusive kinds of governance are those in which a single official or a small group of administrators make decisions directly without outside input. An example is executive agency rulemaking. Second on this scale are practices that invite only a narrow group of stakeholders. While they incorporate a larger share of a population, perhaps lay representatives rather than only government officials, included perspectives are still limited. Neighborhood advisory groups that bring together government officials and citizens are an obvious case. A third level of inclusivity is found in those practices that are open to anyone who cares to join. Public comment sessions frequently display this sort of inclusiveness: Showing up at the right time and signing up for a turn are the only official bars to entry, but such methods may be intimidating for an average citizen (Berner, 2004). The most inclusive practices on this scale are open to everyone and include targeted recruiting efforts. This step is a slight modification to Fung’s account: While self-selecting participation seems more inclusive than targeted sorts, without intentional efforts to increase participation by populations that might otherwise not get involved (minorities and less affluent people, especially), political processes wind up primarily limited to “the usual suspects”—those who are White and better off. Evidence of this danger can be found in the homogeneity of participating populations in the early years of Chicago Ward 49 participatory budgeting (Stewart, Miller, Hildreth, & Wright-Phillips, 2014). Since that time, recruitment efforts have been successful in raising the participation of minority and poor residents (Crum, Salinas, & Weber, 2013). Thus, open access itself is not enough to guarantee inclusivity of process.
The Mode of Communication
The second element of the Modes of Participation is Communication. This scale shows a continuum that draws on Nabatchi (2012) to indicate that public participation processes can move from one level to another, making use of one-way, two-way, or deliberative communication modes. One-way communication is a “unidirectional flow” or transfer of information from the sender (e.g., a public administrator) to the receiver (e.g., a citizen). This type of communication provides little opportunity for response. Some common tools used for one-way communication are political websites, direct mail, and media briefings. Conversely, citizens can send information to the public administrator through client satisfaction surveys (Nabatchi, 2012). This mode of communication is also visible in some public meetings where participants are largely spectators and passively receive information or opinions about public policy. During the public comment period, citizens express their views which officials passively receive. Fung suggests that “[a]t most public hearings, for example, officials commit to no more than receiving the testimony of participants and considering their views in their own subsequent deliberations” (Fung, 2006, p. 68). Communication like this limits discussions and prevents opportunities for negotiation (Nabatchi, 2012).
Two-way communication is the “bidirectional flow of information.” This is communication flowing back and forth, where individuals are able to serve the role of both senders and receivers. Citizens, for example, may inquire through email about policy or positions from public administrators who may provide direct responses (Nabatchi, 2012). Zhang and Liao (2011) provide other examples like visiting civil groups, meeting with individual residents, and holding public budget presentations (with follow-up question-and-answer sessions) as two-way mechanisms. Arnstein (1969) gives an example that appears to show two-way communication, but which actually exemplifies its failures: A Model Cities planning meeting in Providence, Rhode Island, met to discuss the placement of six “tot-lots,” small playgrounds for children. Noting that the neighborhood had a 50% divide among both Black and White residents, Black representatives questioned why only two of the tot-lots were proposed for the Black neighborhood and four for the White neighborhood. City officials responded with a technical explanation that was beyond residents’ understanding. Intimidated by the legal jargon and prestige of the officials, citizens accepted the information provided and supported the proposal to locate four lots in the White district. Here, there was only an exchange of ideas, not mutual understanding. This exchange only aped the form of two-way communication.
Deliberative communication is a more structured process and is aimed toward solving problems. Participants are engaged with one another as equals and contribute to rational decisions about public problems (Fung, 2006). Specifically, this mode of communication requires that the process is open and accessible, that all participants are treated with respect and given an opportunity to speak. This means participants must listen attentively and consider others’ contributions. Deliberative communication is constituted by reasoned discussions that involve presentation of alternative solutions as well as consideration of the strengths and weaknesses of each approach. Importantly, deliberative communication may require greater structure and guidance than typical municipal forums for citizen participation. In many cases, trained facilitators, structured questions, and ground-rules constrain discourse in ways that make it possible for all participants to be treated as equals.
More recent renditions of deliberative democracy have looked beyond only reasoned arguments to include fellow deliberators’ emotions, values, and testimonies (Nabatchi, 2012). Fung (2006) provided an example of how the interest and support of the mobilizing public influenced the direction and discussion of the 9/11 Commission (National Commission on Terrorist Attacks upon the United States) recommendations. The Deliberative Polling method developed by Fishkin, Luskin, & Jowell (2000) is also clearly a deliberative participation process: Citizens are given information about policy issues and then encouraged to discuss this material with other participants. Additional examples of deliberative participation processes include citizen juries, the National Issues Forum, and 21st Century Town Meetings.
The Mode of Decision Making
This third evaluative mode refers to how public decisions are made. This scale largely follows Fung with one important exception. He measured decision making in terms of the relative intensity of the investment, knowledge, and commitment of participants. A key problem with Fung’s scale, however, is that the most intensive practice is deploying expertise. While this may be intensive for experts and policy makers, lay citizens are largely absent from this process. The Mode of Decision focuses on lay citizen participation and, therefore, in reversal of Fung, expert decision making the first and least intense step. Its remaining stages, in order of participatory robustness, are decisions by elected leaders, aggregation, bargaining, and finally collaboration. The first two stages—decisions made by experts or elected officials—represent indirect methods of citizen participation.
In the first level of the Mode of Decision-Making, appointed expert administrators, regulatory commissions, and staff conduct research, offer policy recommendations, and write regulations. These experts are not directly accountable to the public. The only chance for citizen input is during public comment periods for proposed regulations, which are rarely effective. The second, related, step is when decisions are made by elected leaders. This is a more open process as citizens can lobby their officials on particular issues as well as vote candidates in or out of office.
Aggregation, the third step, covers special elections and referenda where citizens vote directly on matters of public policy. The most prevalent examples are tax levies, such as California’s Proposition 13, which placed a limit on the amount of taxes local governments could collect. Though citizens at large have ultimate authority in deciding these matters, the intensity of their commitment (voting) is minimal. Though some may take time to learn about the particular ballot measure, the majority do not.
Bargaining is the fourth step. Here, select citizens try to influence governmental decisions by working within the terms of governmental decision-making processes. Citizens may lobby officials directly or join interest groups to achieve their objectives. Bargaining is framed in adversarial terms; the goal is to win (Nabatchi, 2012, p. 701). Groups are pitted against each other to carve out portions of the budgetary pie. If one project receives more funds, others will receive less. There might be compromise and give and take between groups.
The final and highest step is collaboration. Rather than trying to gain resources for a particular group at the expense of others, collaboration involves interaction between citizens, interests groups, and government focused on a particular problem or issue. “Winning” means arriving at an outcome that all parties can agree to. Collaboration typically involves deliberation and shared authority—features identified in other modes. Thus, there is interaction between the different modes of participation. They can be understood distinctly, however. While the British Columbia Citizens’ Assembly was collaborative and deliberative, it did not result in shared authority over outcomes.
The Mode of Citizen Authority
The final scale of the Modes of Participation is Citizen Authority. Where the Mode of Decision-Making notes how citizens participate, Citizen Authority identifies how effective their participation is. It asks to what degree citizens affect their political situation. The scale’s first stage is tokenism. Arnstein’s critique of putatively democratic processes was basically right: Participation without redistribution of power is an empty and frustrating process of powerlessness. Practices that merely inform or consult lay citizens tend also to be mapped near the bottom of the communication and decision mode-scales, though with their mass-politics focus, they often fare well in terms of inclusion. Exemplary practices here are campaign ads, form-letter responses to impassioned letters or emails, and many public comment sessions.
The second step is that of indirect influence, where citizens interact in a way that only may have influence, including those practices where direct authority is understood to be held in trust or obtained secondhand. Those who make the decisions are in some minimal way responsible to others in the broader situation, perhaps through election or worries about popularity polling, but these elements evidence only slight authority by the vast majority of those individuals participating.
The third stage is de facto authority. Here, decision makers regularly implement projects or other proposals that originate from lay citizens, though they are not required to do so. Mechanisms used to determine citizen preference vary from issues-based polling to voting to consensus processes. Chicago’s participatory budgeting, in which aldermen have ultimate authority but consistently carry out much of the results of the participatory budgeting process, is one sort of de facto authority. Participants in processes at this level can be described as having binding but not final authority. They are frequently thought of as co-governance or quasi-legislative processes.
The most developed level on this scale is de jure authority, that which Fung calls “direct authority” and Arnstein terms “citizen control.” Arnstein’s description is clear: This level displays
that degree of power (or control) which guarantees that participants or residents can govern a program or an institution, be in full charge of policy and managerial aspects, and be able to negotiate the conditions under which ‘outsiders’ may change them. (Arnstein, 1969, p. 223)
The line between this category and the one prior is fluid. One way of parsing the difference is by reference to statutory authority: At the third level, citizen desires may regularly be implemented by custom, but the fourth level expresses them by law. This characterization allows a useful distinction between most contemporary instances of de facto authority in participatory budgeting and fully realized direct democracy.
Combining the Modes
Each ladder represents a distinct mode of participation. While each ladder is important for analysis, assessing the participatory robustness of a particular practice requires consideration of all ladders together. The standards offered, individually and collectively, serve as tools to identify participatory budgeting and as aspirational guides. There is one important caveat: The minimal standards are not absolute. Indeed, many practices simultaneously occupy different steps on a single ladder. With respect to the Mode of Decision, Chicago’s participatory budgeting implementation is guided by initial collaboration, but the final moment of project selection is aggregative. Because of the presence of multiple forms of actions, criteria on the Modes should be seen as on a continuum. So rather than rigid markers that determine what practices are in or out, these standards serve as guides to move practice in more participatory directions.
Some Examples
The important question, however, is where to draw the line between genuine participation and not. To show one way in which this distinction might occur, participatory budgeting is a useful test case. Why, as will be shown, does Rossman and Shanahan’s university budgeting committee not belong in the family of participatory budgeting? Why does Guelph, whose final projects are also determined by a committee, belong? There are no a priori criteria for inclusion and exclusion. Rather, cases should be looked at in their particular contexts with a mind toward the aim of distinguishing purposefully. The purpose assumed in this article is preserving what is unique about the family of participatory budgeting practices.
Four examples show how the Modes of Participation can be applied and what value it brings to public administration literature. Each of these examples has appeared in the literature as a participatory budgeting implementation, though their differences are significant and demonstrate the variety of practices that fall under that term as well as how it is sometimes applied problematically.
1. 1989 Porto Alegre’s “Orçamento Participativo” was the original instance of participatory budgeting. It operated in two stages: First, neighborhood assemblies met to debate local budget prioritization, to propose desirable projects, and to elect representatives. Second, these delegates assembled with local governmental and civil society representatives to decide among priorities and projects, affecting the whole city (Wampler, 2007). As social justice was an aim of this participatory budgeting process, funds were distributed according to a “Quality of Life Index,” with more funds directed to low-income and low-infrastructure districts, including those that participated at a higher rate in initial meetings. This funding prioritization scheme ensured social justice outcomes. On each of the Modes of Participation, Porto Alegre ranks highly, and thus it is clearly an instance of participatory budgeting.
Participant selection: Porto Alegre participatory budgeting’s initial neighborhood meetings are open to any residents who care to participate, sometimes drawing thousands of individuals (Baiocchi, 2001). The quality of life index, especially its participation-measure, coupled with the continual and intentional focus on social justice outcomes guarantees the Porto Alegre process is open selection with targeted recruitment. To support this, see that Baiocchi noted better-off residents were significantly underrepresented in a typical year, 1998 (Baiocchi, 2001, p. 52). Communication: At each stage, from initial assemblies to the delegates’ final meetings and decisions, a deliberative framework is operative. There are opportunities for sharing ideas, exchanging perspectives, and reciprocally engaging with differing parties. One- and two-way methods of exchange are used only for organizing the meetings and reporting outcomes. Decision: The process in Porto Alegre has multiple stages and multiple modes of decision. The initial regional assemblies are collaborative, with lay citizens and civil society groups work together to generate ideas. Neighborhood delegates are elected to work with elected officials to learn more about the technicalities of projects, budgeting and establish priorities. Again, this is a collaborative process. Delegates from these meetings attend a Second Plenary, where projects are determined decisions though bargaining and aggregation. Authority: Participatory budgeting in Porto Alegre has no legal recognition, but the municipal government consistently passes its decisions without modification, fearing political repercussions if they were not to do so (de Sousa Santos, 1998). This is de facto authority.
2. In Guelph, Canada, participatory budgeting is administered through a civil society organization called the Neighborhood Support Coalition (Pinnington et al., 2009). Although the City of Guelph contributes money to the process—initially via a community grant, now through a budget line-item—the Coalition also raises funds for participatory budgeting from “various foundations and city, county, and provincial agencies, accumulating a bigger pot of money than the city alone could have provided” (Baiocchi & Lerner, 2007, p. 11). The process was begun because city employees noticed “some neighbourhoods were over-resourced and others were under-resourced.” To remedy this, a city manager “suggested that funding would be more equitable if the neighbourhood groups deliberated their needs and priorities together” (Lerner & Van Wagner, 2006). Guelph participatory budgeting includes some funds that are restricted to capital improvement and some that may be spent more freely. As in Chicago and Porto Alegre, an early stage of the process is neighborhood meetings, but these generate lists of both “needs” and “wants” along with electing representatives. Wholly unlike Chicago, and somewhat differently from Porto Alegre, ultimate authority for deciding allocations is left to a 12-person finance committee made up of some but not all neighborhood delegates. This committee negotiates until it reaches consensus.

The modes of participation—Porto Alegre.
Guelph’s process is also clearly an instance of participatory budgeting. It reaches the standards on three of the four Modes of Participation: (a) Participant selection: The process’s motivation was redistributive, but because it is neighborhood group delegates rather than individuals who deliberate and decide, Guelph misses the mark here, ranking as only open access. (b) Communication: Decisions are made through collaborative meetings that aim toward eventual consensus, and delegates are seen as equals, so this is a case of deliberative communication. (c) Decision: While representatives are elected, the funding decisions are made by consensus. (d) Authority: The funds allocated are civil society, not governmental, so they can be allocated directly. As such, the process has absolute, de jure, authority over them.
3. Chicago’s 49th Ward was the United States’ first participatory budgeting implementation. It began in 2009. Each year since, its process has been modified in response to prior years’ limitations and challenges. Like other participatory budgeting processes, the 49th Ward’s is carried out in multiple stages. These stages vary with respect to each of the variables considered by the Modes of Participation. Thus, this example shows both the complexity of participatory budgeting as well as how the Modes can help clarify this complexity.

The modes of participation—Guelph.
The yearly participatory budgeting cycle has four general stages. It begins with open informational meetings that introduce residents to the process and show what kinds of projects may be selected. All program funds originate from Chicago’s Capital Improvement budget, so they are restricted for use on only concrete and discrete projects that improve public buildings or land. Previously selected projects have included a dog park, underpass murals, and street repaving. These initial meetings also allow residents to brainstorm and suggest project ideas, and then to volunteer to serve on one of the committees that determine project viability and cost. These committees are the process’ second stage. They also put together a final ballot, which lists a variety of projects and their associated costs. At a later election, which is the third stage, Ward residents vote to select from among the proposals. In a final step, the most popular selections are implemented.
a. Participant selection: The first and third stages of the process are the most inclusive. In Chicago’s first years of participatory budgeting, these two stages were open to all residents over the age of 16. More recently, in response to disproportionately low minority participation, the alderman’s office has implemented targeted recruitment strategies. These include hiring a bilingual participatory budgeting coordinator, holding a project fair to allow residents to learn more about proposed projects, conducting at least one meeting in Spanish, offering child care at meetings, and going door-to-door in less affluent sections of the Ward. Because of their improvement throughout the years, these stages of Chicago’s process now exemplify the highest level of inclusivity.
The second and fourth steps, committee meetings and implementation, are carried out by significantly smaller groups. As noted, the meetings are open to those people who volunteer. Whatever targeted recruitment occurs for the prior stage would have bleed-over to those included in the second one, but because there is no further specific mechanism for targeting, it ranks as only open selection. Implementation is harder to rank, because participants vary so much from project to project. Because road-improvement projects are implemented through the city’s capital works program, participating workers are doing their job rather than participating in a political process. Other projects, though, such as overpass murals, intentionally select artists from the community, but only a handful of individuals do the work. Ideally, any resident would be able to work on selected projects, but this has not yet occurred.
b. Communication: All three levels of communication are on display in stages of Chicago’s process, though the results of each stage are passed to the next via one-way communication mechanisms. Initial ward meetings begin with one-way communication: Aldermanic staff give presentations on process and past projects, passing information from presenter to recipient. The brainstorming portion of these meetings invert the room’s information flow, with residents reporting needs and ideas to the presenters. Project feasibility committees are deliberative, though their communicative subject matter is narrow. Project selection voting and implementation are both prototypical one-way communication: Decisions are made, reported, and then carried out. The hope is that though no one stage of the process is fully deliberative, the process itself is. In recent years, Chicago has worked toward this goal by implementing project expositions, meetings in which residents can come together to learn about and discuss the merits of various proposals. Chicago’s participatory budgeting process is on the border between two-way and deliberative communication. This ranking example, along with the one following, shows the complexity of identifying and ranking specific participatory criteria in concrete political processes, a point that will be discussed below.
c. Decision: Chicago participatory budgeting evinces multiple kinds of decision making. The initial brainstorming sessions may include collaborative moments, because residents openly suggest and then build upon one another’s identification of problems and solutions. Final project selection is made by ballot and is thus aggregative. The final step is where the ultimate decision is made, and so Chicago’s process is aggregative. Yet, the aggregation occurs after multiple steps of collaborative winnowing. This difference matters, because it indicates that a residue of collaboration may inform and survive the final selection.
d. Authority: The money used in participatory budgeting in Chicago’s 49th Ward is discretionary funds budgeted to the alderman. In most other wards, the alderman spends this money without substantial resident guidance. Though he is not obliged to do so, the 49th Ward’s alderman has consistently carried out resident priorities as expressed through the participatory budgeting process. As such, Chicago participatory budgeting rates as de facto authority on this second scale.
As noted above, the modes of decision and communication are difficult to rank in this example. Chicago’s process is a months-long affair, and each of its stages utilizes different methods of sharing information and collecting opinions. The key question is whether a complete process can maintain a higher level of participatory robustness even if it includes lower-ranked stages. There remain unknown elements that would be helpful in making these identifications: Do participants work together and build upon ideas collaboratively in brainstorming meetings? Though the movement from stage to stage is consistently one-way communication, do the desires expressed earlier stages influence the direction of later stages? That is, ultimately, to what degree do final choices reflect influence of other people’s needs and concerns? Answering these kinds of questions would be productive avenues for future empirical and theoretical research.
4. Rossmann and Shanahan (2012) study a university budgeting committee as a form of participatory budgeting. They do not present a formal definition, but instead contrast participatory budgeting with zero-based, program-based, and performance-based budgeting. The purpose of their article is interpretive and normative—to explore how participants understand and define democratic values through the process. However, this particular committee is a problematic example of participatory budgeting.
Rossman and Shanahan’s qualitative analysis highlights participants’ skepticism with the process, noting how it was more representative than participatory and that representation was skewed toward academic and administrative powers (p. 61). Participants embraced their ethical democratic responsibilities, even though they thought the committee as a whole failed to meet normative ideals of openness and inclusiveness (p. 62). While this article illuminates citizens’ experiences in a budgetary innovation, it should have been framed more precisely, particularly by avoiding participatory budgeting language. What they discuss is a model of representative budgeting that is partially open to the public as opposed to typical university budgeting processes where the President or Chancellor, with input from senior administrators, makes executive budgeting decisions behind closed doors. While most meetings were open to the general public, voting members were either senior administrators or representatives of key stakeholders (e.g., chair of faculty senate, professional council chair). The president represented students and selected community representatives.
The university budgeting committee meets the de facto standard for authority. The president adopted budgets developed by the committee (even though he has final authority). It does not meet standards for inclusion and decision, however, and it is unclear whether meetings were sufficiently deliberative. The committee was largely composed of “elites” (upper administrators) with elected or selected representation from key stakeholders. Most importantly, one of the major “citizens” of the university—students—were excluded. Thus, it is not a case of ordinary citizens (with special recruitment of underrepresented groups) participating in budgetary decisions. In terms of decision making, the agenda was set by and information was controlled by the provost’s office. Although any member could add agenda items, participants suggest this process was administration-driven. This led many to reflect that they were merely rubber-stamping decisions. In the words of one participant: “Most of the decisions have been made before they come to the committee. . . . They come to the committee as recommendations and the committee validates with some decisions” (Rossmann & Shanahan, 2012, p. 62). While it appears there were discussions, their content and mode reflected power asymmetries. Budgetary information was often presented at the time of meetings, which privileges those who understand both budgetary complexities and jargon.
This university budgeting committee clearly does not meet participatory budgeting standards on two of the Modes of Participation: (a) Participation selection: While some meetings were open to ordinary citizens, they had no seat at the table. Moreover, there were no efforts to include marginalized populations. (b) Authority: Ordinary citizens did not have authority over decisions. Rather, authority was exercised by elites who represented select stakeholders. The President’s representing students was especially problematic given that he had final authority over all budgetary decisions.
With respect to the other two Modes, this example may meet the proposed standards: (c) Communication: The committee’s meetings putatively carried out deliberative discussions, but participants reported structural constraints that revealed two-way and perhaps even one-way elements. Particularly, because meetings were held in administrative buildings, and in a too-small room, participants felt intimidated. And the complexity of budgeting jargon and spreadsheets made it difficult for nonelite participants to deliberate effectively. d. Decision. The meetings were collaborative, but if the concerns in the communication mode are correct, then deploying expertise worries arise.

University budget committee—The modes of participation.
It is safe to conclude that this is not an example of participatory budgeting. This is not to say, of course, that such a committee could not move in the direction of participatory budgeting with careful modification.
One way in which the university budgeting committee could shift closer to being identifiably participatory budgeting is to adopt some of Guelph’s practices. It is especially laudable for being administered by a nongovernmental organization, for having multiple stages at which representatives discuss neighborhood and city-wide needs and wants with their constituencies, and for requiring consensus on final decisions. Guelph’s participatory budgeting meets or exceeds the minimum standards on three of the four Modes of Participation. It shows how budgetary decisions by committee may be enriched with additional deliberation, collaboration, and inclusivity.
Conclusion
With four ladders and four minimum standards, readers may be tempted to use the Modes of Participation in this application to form a new definition of participatory budgeting, based wholly on process: Participatory budgeting is a budgeting practice with open access (and targeted recruitment), active deliberation, collaborative decision making, and de facto authority. Because participatory budgeting is an emergent and context-dependent practice, it is not advisable to construct a final definition that is absolute, excluding practices that do not meet all standards. Indeed, only a few practices meet all these high standards, others nearly meet them, and others still only meet one or two. This complexity can be represented more fairly on the Modes of Participation’s continua rather than with simple binary variables. Such an advance is why the Modes are especially helpful in understanding a process’s participatory robustness. The Modes of Participation should not be understood as the rigid definitional standard for including and excluding. That is precisely what this article has been striving not to do. Rather, the modes present both a tool for identification and an ideal toward which participatory practices should develop. There are two key ways in which the modes of participation are useful.
First, this standard-driven conceptual model is presented as a heuristic framework. It is a visual and analytical model for researchers and practitioners to identify which practices belong in a family, such as that of participatory budgeting. Thus, it provides a basis to exclude practices that are far from the ideal, unrecognizable on its terms, and thus obfuscate distinctive usefulness. Clearly, there is no one essence or determinate meaning of participatory budgeting. Rather, there is a complicated network of similarities of detail and similarities overall. Yet, it is still possible to identify what belongs in a family of resemblances. The boundary between what is included and excluded is drawn based on the historical development of this family of practices. Perhaps those public administration researchers who identify public comment sessions with participatory budgeting are not yet fully familiar with this practice that developed in the global south and has been spreading worldwide. The Modes of Participation is a guide for scholars and practitioners to identify crucial practices and outcomes of participatory budgeting. Without this guide, budgeting practices that are participatory in the most superficial of ways are blended together with those that are robustly participatory, and what is of value may wind up lost.
Second, the Modes of Participation present an aspirational ideal: They show how multiple dimensions of participation may be maximized. Rather than a static definition, the Modes offer researchers and practitioners a vehicle to identify deficiencies and areas to improve. The goal is to move practices toward more participatory robustness. For instance, even though the university budgeting committee does not meet enough standards to count as participatory budgeting, it is a promising practice that can be developed or combined with other efforts to encourage participatory development. The way forward for contemporary participatory processes, including participatory budgeting, is to be found through learning about and then trying out what has worked in prior participatory processes. It is the aspirational ideal, realized variously and partially in participatory implementations worldwide, that makes participatory budgeting a site of possibility that is worthy of researchers’ study and organizers’ efforts.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
