Players and utilities

A developer $D$ wants to construct a project in one of the districts. A politician $P$ , who determines the project approval process, has concern over the externality experienced by local residents in each district but is uncertain about their magnitudes, represented by ${\beta _I}$ (with $I \in \left\{ {A,B} \right\}$ ). Each district also contains an opponent ${O_I}$ , so $P$ can learn about local costs by creating participatory institutions to allow opposition. However, imposing delay is costly not only for ${O_I}$ but also for $P$ given that a project is pursued.

The politician $P$ chooses among three possible development regimes: unilateral (denoted with a superscript $\mathbb U$ ), participatory (denoted with a superscript $\mathbb P$ ), and decentralized (denoted with a superscript $\mathbb D$ ). In the unilateral regime, $P$ sets a development fee ${\phi _I}$ in each district. In the participatory regime, $P$ observes levels of opposition and then sets the development fee. And in the decentralized regime, each district $I$ contains a median voter ${V_I}$ , who is uncertain about ${V_J}$ ’s externality and who chooses ${\phi _I}$ . Following the procedure of each development regime, $D$ decides whether and where to build.

Utilities for each player are as follows:

$${U_P}\left( {{\phi _A},{\phi _B}} \right) = {\pi _A}{\rm{\;}}\left( {{\psi _A}{\rm{\;}}\left( { - {\beta _A}{\rm{\;}}b} \right) + \left( {1 - {t_A}} \right){\rm{\;}}{\phi _A}} \right) + {\pi _B}{\rm{\;}}\left( {{\psi _B}{\rm{\;}}\left( { - {\beta _B}{\rm{\;}}b} \right) + \left( {1 - {t_B}} \right){\rm{\;}}{\phi _B}} \right),$$

$${U_{{O_I}}}\left( {{t_I}} \right) = {\pi _I}\;\left( { - \gamma \;{\beta _I}\;b} \right) - {\kappa \over 2}\;{t_I}^2,$$

$${U_{{V_I}}}\left( {{\phi _I}} \right) = {\pi _I}{\rm{\;}}\left( { - {\beta _I}{\rm{\;}}b + {\phi _I}} \right),$$

$${U_D}\left( {{\pi _A},{\pi _B}} \right) = {\pi _A}{\rm{\;}}\left( {b - {\phi _A}} \right) + {\pi _B}{\rm{\;}}\left( {b - {\phi _B}} \right),$$

where ${\pi _I}$ indicates that the project is built in district $I$ , ${\psi _I}$ is $P$ ’s level of concern for $I$ , ${\beta _I}$ is $I$ ’s externality coefficient, $b$ is the magnitude of the project, ${t_I}$ is transaction costs imposed by ${O_I}$ , and ${\phi _I}$ is the fee in district $I$ . The parameter $\gamma $ is ${O_I}$ ’s level of concern about the project. Next, $\kappa $ is each ${O_I}$ ’s level of opposition costs. $D$ may select at most one ${\pi _I} = 1$ .

Order of moves

The order of moves in the game is as follows:

1. 1. The type ${\beta _I}$ of each opponent ${O_I}$ is drawn independently and revealed only to ${O_I}$ , with each ${\beta _I}\sim U\left[ {0,\bar \beta } \right]$ .

2. 2. The politician $P$ decides whether to delegate decision-making to the districts.

3. 3. If $P$ delegated, each ${V_I}$ simultaneously chooses ${\phi _I}$ .

4. 4. If $P$ did not delegate:

   1. (a) $P$ chooses the level of opposition costs $\kappa \in \left\{ {\tilde \kappa, \infty } \right\}$ , with $0 \lt \tilde \kappa \lt \infty $ .

   2. (b) Each ${O_I}$ selects transaction costs ${t_I}$ to impose contingent on a project being built in district $I$ .

   3. (c) $P$ observes each ${t_I}$ and sets each district-specific application fee ${\phi _I}$ .

5. 5. The developer $D$ decides whether and where to pay the fee and build the project.

6. 6. The game ends and payoffs are realized.

We say that $P$ chooses the unilateral regime when not delegating along with setting $\kappa = \infty $ (because participation is infinitely costly), the participatory regime when not delegating along with setting $\kappa = \tilde \kappa $ , and the decentralized regime when delegating.

Assumptions

Our first assumption is without loss of generality:

This assumption captures inequality across neighborhoods. For the politician $P$ to care unequally about the welfare of either neighborhood is a reduced form way of representing inequalities of political power or resources in urban politics. Next, we assume that externalities are small enough so that no matter what $P$ learns, the project is always placed somewhere rather than nowhere. This corresponds to the following assumption:

Thus, while opposition may inform in which district the project should go, it does not entirely preclude projects. While local opposition obviously prevents many projects from being constructed, this assumption has two advantages. First, it allows us to focus on those projects that go forward but are misallocated across districts. Second, it eases the analysis of a signaling game in which the private types of two different players are drawn randomly.Footnote ⁹

Equilibrium

As this is a sequential game of imperfect information, the natural equilibrium concept is perfect Bayesian equilibrium (PBE). We focus on pure-strategy equilibria that survive the D1 refinement. A pure strategy for $P$ is a function ${{\mathbb S}_P}:{\mathbb R}_ + ^2 \to \left\{ {0,1} \right\} \times \left\{ {\tilde \kappa, \infty } \right\} \times {\mathbb R}_ + ^2$ that maps each district’s choice of transaction costs to a delegation decision, a level of opposition costs to impose, and the application fees in each district given no delegation. Given the order of moves, the delegation decision and level of opposition costs are constant in each district’s choice of transaction costs. A pure strategy for each ${O_I}$ is a function ${{\mathbb S}_{{O_I}}}:\left\{ {\tilde \kappa, \infty } \right\} \times \left[ {0,\bar \beta } \right] \to {{\mathbb R}_ + }$ that maps $P$ ’s choice of opposition costs and $I$ ’s externality level to a choice of transaction costs. A pure strategy for each ${V_I}$ is a function ${{\mathbb S}_{{V_I}}}:\left[ {0,\bar \beta } \right] \to {{\mathbb R}_ + }$ that maps $I$ ’s externality level to an application fee given delegation. A pure strategy for $D$ is a function ${{\mathbb S}_D}:{\mathbb R}_ + ^2 \to \left\{ {\emptyset, A,B} \right\}$ that maps application fees in each district to a choice of which district (if any) in which to pay the fee and build the project. $P$ ’s beliefs about $\left( {{\beta _A},{\beta _B}} \right)$ are denoted by the function $\mu \left( {{t_A},{t_B}} \right)$ . A PBE is a strategy profile and system of beliefs such that players’ strategies are sequentially rational and beliefs are consistent with Bayes’s rule on the path of play.

Summary

To summarize, for $I \in \left\{ {A,B} \right\}$ , the exogenous parameters are ${\psi _I}$ , $b$ , $\gamma $ , $\tilde \kappa $ , and $\bar \beta $ . The parameter ${\beta _I}$ is a random variable drawn from a uniform distribution with support $[ {0,\bar \beta } ]$ . The endogenous choices are ${\phi _I}$ , ${\pi _I}$ , ${t_I}$ , and $\kappa $ . The equilibrium concept is PBE. We focus on pure-strategy equilibria that survive the D1 refinement.

The policymaking environment

The starting point for our approach to land use regulation is an assumption that residents experience local costs from new development, but they weigh these against potential economic benefits. This reflects an extensive empirical literature on the costs that residents perceive from new housing (Hankinson Reference Hankinson2018; Marble and Nall Reference Marble and Nall2021; Monkkonen and Livesley-O’Neill Reference Monkkonen and Livesley-O’Neill2017; Trounstine Reference Trounstine2023), as well as resident responsiveness to economic benefits when present (Gerber and Phillips Reference Gerber and Phillips2003, Reference Gerber and Phillips2004; Alster Reference Alster2023). These economic benefits might arise as a result of bargaining between residents and developers in the form of community benefits agreements (Been Reference Been2010; King and Rich Reference King and Ashlin Rich2022).

Our formal framework, which builds upon Foster and Warren (Reference Foster and Warren2022), reflects these possibilities. Players $P$ or ${V_I}$ set a fee $\phi $ that captures the compensation for residents that developers must provide.Footnote ¹⁰ Notably, $t$ operates as a source of inefficiency that reduces potential bargains between $D$ and $P$ , so it does not matter which player is assumed to pay $t$ directly.Footnote ¹¹ Each project has some value $b$ (assumed to be positive) representing the net social value of the project. While there are social benefits to the project, there are also local costs, represented by ${\beta _I}$ .

Unlike the previous model, however, which focuses on voter attitudes toward development in light of obstruction by project opponents, the current paper asks why politicians (or voters) establish such institutions to facilitate obstruction in the first place. For this to be plausible, participatory institutions must provide a benefit over the cost to the institutional designer. In keeping with this, empirical evidence indicates that politicians obtain useful information from public participation. For instance, Einstein et al. (Reference Einstein, Glick and Palmer2020, 121) observe that “resident comments can change board members’ minds.” Correspondingly, information gathering is useful to politicians in our model due to uncertainty over the local cost of projects ${\beta _I}$ . Starting from this uncertainty allows us to investigate when and why politicians decide to establish participatory institutions.

The opponent ${O_I}$ represents the local residents most opposed to a proposed project. As a committed opponent, ${O_I}$ experiences no benefits from the project. Empirically, individuals who attend neighborhood meetings tend to be unrepresentative (Einstein et al. Reference Einstein, Palmer and Glick2019), and even a handful of opponents can substantially delay a project and increase regulatory expenses (Einstein et al. Reference Einstein, Glick and Palmer2017). Even though ${O_I}$ ’s preferences need not reproduce those of other local residents, ${O_I}$ is assumed to have preferences at least connected to those of district $I$ residents more broadly. Simply put, the worse a project is for typical district $I$ residents, the worse it is for ${O_I}$ , even while ${O_I}$ represents those residents who are most opposed. This link is controlled by the parameter $\gamma $ , with the above interpretation implying that $\gamma \gt 1$ (though this need not be assumed in the model analysis).

The politician $P$ ’s choice of protest costs is discrete. If $P$ chooses the unilateral regime, opposition is infinitely costly (i.e., prohibited), while if $P$ chooses the participatory regime, there is a positive but finite level of protest costs. Were $P$ allowed to choose any strictly positive and finite value of $\kappa $ , then $P$ would have no best strategy. That is because in equilibrium, all information is revealed for any such value of $\kappa $ , while the costs imposed on $P$ by each ${O_i}$ are decreasing in $\kappa $ . Of course, in reality, $P$ may find it harder and harder to distinguish different levels of opposition costs (and thus types) the closer that they get, as occurs the larger that $\kappa $ becomes. Explicitly introducing such noise would complicate characterization of the equilibrium, but our assumption about $P$ ’s choice set can be seen as representing the possibility that opposition needs to be sufficiently strong to communicate information.

The signaling mechanism

Conditional on the establishment of participatory institutions, the model bears closest formal resemblance to an all-pay reverse auction. Similarly, Cotton (Reference Cotton2016) studies an environment in which two lobbies exert costly effort to persuade a politician to review their policy proposals. The politician then selects a proposal for which to investigate the level of quality. Cotton shows that this “contest for attention” fully reveals the quality of both proposals even if the politician reviews only one of them. As in that model, the purpose of the all-pay auction is not necessarily to maximize revenue but rather to elicit information. Our participatory institutions, which allow residents to send a costly signal of their preferences to politicians, thus resemble informational models of protest in authoritarian regimes (e.g., Lorentzen Reference Lorentzen2013).

However, our model differs from these earlier works in several key respects. First, $P$ is conceived of as a biased social welfare maximizer, such that $P$ does not necessarily purchase the object from the lowest bidder. Although this appears directly in $P$ ’s utility function, it corresponds to the effects of racial, economic, or other forms of inequality. With greater resources, particular kinds of residents are able to exert greater electoral pressure on politicians, thus inducing the preferences that we represent in reduced form. We interpret the relative values of ${\psi _A}$ and ${\psi _B}$ to correspond to social inequalities that determine relative neighborhood power.Footnote ¹²

Second, residents’ effort is costly not only to themselves but also to the politician contingent on building the project. Consistent with a typical signaling model, imposing delay is costly on the player imposing it, which may allow for separation of different types and thus learning by the politician. But a key aspect of this sort of institution is that signaling takes the form of imposing delays and thus costs on other actors. Though the literature has other such examples (e.g., Patty Reference Patty2016), in our model there is a direct utility effect of the signal (i.e., unrelated to beliefs) contingent on which action the receiver selects. This corresponds to the particular environment that we study, in which institutions allow residents to impose costs and delays conditional on moving forward with a potential public project.

Third, we allow the politician to choose between establishing such institutions and pursuing other alternatives. The fact that signals are costly to receive creates a distinctive tradeoff for $P$ in the model. The politician $P$ must weigh the benefit of eliciting information from residents against the costs of complicating the project approval process. Depending on $P$ ’s relative concern for different districts – corresponding to districts’ relative ability to pressure the politician – the politician may choose instead to make decisions unilaterally or to delegate decision-making to the districts. In analyzing the receiver’s calculation of whether to establish such institutions, we contribute to research on the relationship between a politician’s goals and tradeoffs inherent in alternative information-gathering institutions (e.g., Gailmard and Patty Reference Gailmard and Patty2013).

$P$ ’s optimal behavior

$P$ observes each district’s level of opposition and decides the district in which it prefers the project to go. Recall that $P$ determines the stringency of land use regulation in each district (i.e., $P$ selects ${\phi _A}$ and ${\phi _B}$ ). Then if $P$ prefers the project to go in $A$ , for example, $P$ sets ${\phi _A} = b$ and ${\phi _B} \gt b$ (we assume that when indifferent, $D$ builds the project).

Upon observing the levels of opposition, $P$ learns ${\beta _A}$ and ${\beta _B}$ . Given $P$ ’s utility, $P$ then prefers to place the project in $A$ when

$${\psi _A}{\rm{\;}}\left( { - {\beta _A}{\rm{\;}}b} \right) + \left( {1 - {\omega _A}{\rm{\;}}{\beta _A}} \right){\rm{\;}}b \ge {\psi _B}{\rm{\;}}\left( { - {\beta _{\rm{B}}}{\rm{\;}}b} \right) + \left( {1 - {\omega _B}{\rm{\;}}{\beta _B}} \right){\rm{\;}}b \Leftrightarrow $$

(1) $${\beta _A}{\rm{\;}}\left( {{\psi _A} + {\omega _A}} \right) \le {\beta _B}{\rm{\;}}\left( {{\psi _B} + {\omega _B}} \right).$$

Notice then that two different factors determine the district in which $P$ prefers the project to go. The first is which district faces a lower externality, accounting for $P$ ’s weights determining its relative concern for each district. The presence of participatory institutions allows $P$ to learn the type of each ${O_I}$ and therefore which district the project “should” go in. However, the fact itself that $P$ cares more about one district may lead $P$ to place the project in the other district, even if it faces a higher externality. Second, $P$ ’s decision is additionally contingent on each opponent’s level of opposition, which not only informs $P$ but is also directly payoff-relevant.

Optimal opposition choices

As conjectured, we find a semiseparating equilibrium in which the intensity of each player’s opposition is a linear function of its type (up to a maximum value). The following lemma summarizes the equilibrium coefficients on the externality experienced that determine the amount that ${O_A}$ and ${O_B}$ oppose:

Lemma 1 Optimal choices of ${\omega _A}$ and ${\omega _B}$ , respectively, are

$${\omega _A}^{\rm{*}} = \sqrt {{{\gamma \;b} \over {\overline \beta \;\tilde \kappa }} \cdot {{{\psi _A}} \over {{\psi _B}}}}, $$

$${\omega _B}^{\rm{*}} = \sqrt {{{\gamma \;b} \over {\overline \beta \;\tilde \kappa }} \cdot {{{\psi _B}} \over {{\psi _A}}}} .$$

Proof. All proofs are in Appendix A unless otherwise noted.□

We see that for each ${O_I}$ , the level of opposition conditional on ${\beta _I}$ is increasing in $\gamma $ , $b$ , and ${\psi _I}$ , and decreasing in ${\psi _J}$ , $\overline \beta $ , and $\tilde \kappa $ . And when opponents’ concern $\gamma $ or project magnitude $b$ equals zero, obviously ${\omega _A}^{\rm{*}} = {\omega _B}^{\rm{*}} = 0$ .

P’s location decision

Now that we know the equilibrium strategies of ${O_A}$ and ${O_B}$ , we can give the (subgame) equilibrium condition under which the project goes in $A$ . Recalling from Inequality 1 that we require ${\beta _A}{\rm{\;}}\left( {{\psi _A} + {\omega _A}} \right) \le {\beta _B}{\rm{\;}}\left( {{\psi _B} + {\omega _B}} \right)$ , we see now that this implies

(2) $${\beta _B} \ge {{{\psi _A}} \over {{\psi _B}}}\;{\beta _A}.$$

In $\left( {{\beta _A},{\beta _B}} \right)$ -space, the equation corresponding to Inequality 2 represents a ray emanating from $\left( {0,0} \right)$ that separates the region in which the project goes in $A$ (top left) from that in which it goes in $B$ (bottom right). Figure 1 illustrates $P$ ’s alternative project location decisions for different values of ${\beta _A}$ and ${\beta _B}$ .

Figure 1. In the white region, where $P$ builds in $A$ , and the light-gray region, where $P$ builds in $B$ , $P$ ’s choice is socially optimal. But in the dark-gray region, $P$ builds in $A$ despite it being socially optimal to build in $B$ . For a given realization of the externality ${\beta _B}$ (denoted by ${\hat \beta _B}$ and shown by horizontal line), an increase in $P$ ’s concern for $B$ , corresponding to ${\psi _B}$ increasing to ${\psi _B}{\rm{'}} \gt {\psi _B}$ , shrinks the region in which $P$ locates the project in district $B$ .

The fact that Inequality 2 depends upon ${\psi _A}$ and ${\psi _B}$ represents the benefit of participatory institutions for $P$ . When $P$ is equally concerned about each district ( ${\psi _A} = {\psi _B}$ ), then Inequality 2 reduces to ${\beta _B} \ge {\beta _A}$ , and $P$ places the project in the lower-externality district. However, when ${\psi _B} \gt {\psi _A}$ , there are times when ${\beta _A} \gt {\beta _B}$ but the project goes in $A$ regardless. In this case, the disadvantaged district $A$ bears the negative externalities of the project despite the fact that those externalities would be smaller if the project were located in $B$ . But $P$ nonetheless chooses to locate the project in $A$ in order to protect the residents of the advantaged district $B$ .

Using the above, we construct $P$ ’s expected utility from establishing participatory institutions. This is

$$\begin{gathered}\mathbb{E}U_P^\mathbb{P} = \frac{1}{{{{\bar \beta }^2}}}\int_0^{\bar \beta } {\left( {\int_0^{\frac{{\psi A}}{{{\psi _B}}}{\beta _A}} {{\psi _B}} \;\left( { - {\beta _B}\;b} \right) + \left( {1 - {\omega _B}^{\text{*}}\;{\beta _B}} \right)\;b\;d{\beta _B}} \right.} \; \\\left. { + \int_{\frac{{{\psi _A}}}{{{\psi _B}}}{\beta _A}}^{\bar \beta } {{\psi _A}\left( { - {\beta _A}b} \right) + \left( {1 - {\omega _A}^{\text{*}}\;{\beta _A}} \right)\;b\;\;d{\beta _B}} } \right)d{\beta _A} \\ \end{gathered}$$

$$\!= {b \over {6{\psi _B}^2}}\left( \!{{\psi _A}\sqrt {{{\overline \beta \;\gamma \;b\;{\psi _A}\;{\psi _B}} \over {\tilde \kappa }}} + \!{\psi _B}\left( {\overline \beta \;{\psi _A}\left( {{\psi _A} - 3{\psi _B}} \right) + \!6{\psi _B} - 3\sqrt {{{\overline \beta \;\gamma \;b\;{\psi _A}\;{\psi _B}} \over {\tilde \kappa }}} } \right)} \right).$$

D’s problem

As before, the developer $D$ chooses to build the project in the district charging the lowest fee, given that any district charges less than the entire value of the project.

The voters’ problem

When delegation occurs, each voter must choose a fee. This setup is effectively a sealed first-price reverse auction with independent private costs in which the good for sale is the right to develop in a district and the price is the fee.Footnote ¹³ In search of a symmetric equilibrium, let us analyze ${V_I}$ ’s problem. ${V_I}$ ’s cost is ${\beta _I} \, b$ , and it believes that ${V_J}$ ’s cost is uniformly distributed over $[ {0,\overline \beta {\rm{\;}}b} ]$ . Suppose that ${V_I}$ believes that ${V_J}$ uses an affine strategy, i.e. ${\phi _J}\left( {{\beta _J}{\rm{\;}}b} \right) = \alpha + \eta {\rm{\;}}{\beta _J}{\rm{\;}}b$ , with $\alpha $ and $\eta $ not a function of ${\beta _J}$ . Then ${V_I}$ ’s expected payoff from choosing some ${\phi _I}$ is

$${\mathbb E}U_{{V_I}}^{\mathbb D}\left( {{\phi _I}} \right) = \left( {1 - {{{\phi _I} - \alpha } \over {\overline \beta {\rm{\;}}\eta {\rm{\;}}b}}} \right)\left( {{\phi _I} - {\beta _I}{\rm{\;}}b} \right).$$

Then the first-order condition implies

$${\phi _I}^{\rm{*}} = {{\alpha + \overline \beta \;\eta \;b} \over 2} + {1 \over 2}{\beta _I}\;b.$$

with the second-order condition satisfied. Looking back to the components of an affine strategy, we would therefore have $\alpha = {{\alpha + \overline \beta \;\eta \;b} \over 2}$ and $\eta = {1 \over 2}$ . Solving simultaneously gives $\alpha = {{\overline \beta \;b} \over 2}$ and $\eta = {1 \over 2}$ . That is, a consistent strategy for all ${V_I}$ is indeed affine, with

$${\phi _I}^{\rm{*}} = {{{\beta _I} + \overline \beta } \over 2}\;b.$$

We see then that $D$ always builds the project, it always goes in the place with the lower externality, and in expectation each voter is able to capture some of the surplus, as is $D$ . Importantly, this contrasts with the participatory regime. That regime exhibits an asymmetry (and thus kinks in many of the expressions) due to the politician favoring one of the districts. But the asymmetry vanishes under the decentralized development regime, because the politician is eliminated and the districts deal directly with the developer, who only cares about how much money is spent on compensation.

P’s expected utility

We assume that $P$ has utility over the project’s social value even though $P$ no longer directly collects the fee. $P$ ’s expected utility from decentralization is therefore as follows:

$${\mathbb E}U_P^{\mathbb D} = {1 \over {{{\overline \beta }^2}}}\mathop \int \nolimits_0^{\overline \beta } \left( {\mathop \int \nolimits_0^{{\beta _A}} {\psi _B}{\rm{\;}}\left( { - {\beta _B}{\rm{\;}}b} \right) + b{\rm{\;\;}}d{\beta _B} + \mathop \int \nolimits_{{\beta _A}}^{\overline \beta } {\psi _A}{\rm{\;}}\left( { - {\beta _A}{\rm{\;}}b} \right) + b{\rm{\;\;}}d{\beta _B}} \right)d{\beta _{\rm{A}}}$$

$$ = b\left( {1 - {{\left( {{\psi _A} + {\psi _B}} \right)\;\overline \beta /2} \over 3}} \right).$$

Atlanta

After desegregation and the Voting Rights Act, Maynard Jackson was elected mayor in 1973 with strong Black support, while Atlanta’s business elite were divided. Jackson pledged greater inclusivity, using a metaphor of a feast in which no groups are excluded. Jackson also strongly supported neighborhood groups, many of which had recently mobilized around highway construction and other issues (Stone Reference Stone1989). At the same time, Jackson sought positive relations with the downtown business elite and pursued business development with “enthusiastic support” (Reed Reference Reed1999, 168).Yet this represented a substantial shift from earlier years in which elite-led redevelopment displaced as much as one-fifth of Atlanta’s population, corresponding to the unilateral development regime (Stone Reference Stone1989). In terms of the model, neighborhoods are increasingly equal, though still unequal, within the politician’s utility.

As Jackson took office in 1974, the city council established twenty-four Neighborhood Planning Units (NPUs) with the goal of facilitating neighborhood-level public participation in land use planning. Jackson supported the NPUs with administrative staff who assisted neighborhood residents in developing neighborhood-specific plans. These neighborhood proposals would then contribute to Atlanta’s comprehensive land use plan (Hutcheson Jr. Reference Hutcheson1984). Hence, the NPUs provide a forum for neighborhood residents to voice their concerns to city officials and potentially delay development proposals. As the model shows, politicians prefer participatory procedures along these lines when they want to protect disadvantaged groups from the worst projects, but still override local preferences in some cases on behalf of more powerful neighborhoods.

New York

In New York City, official neighborhood institutions were first established in 1963 as advisory bodies by Mayor Robert Wagner Jr. to counterbalance the power of the Democratic political machine (Angotti Reference Angotti1997). Nonetheless, as in the unilateral regime of the model, opportunities for neighborhood-level public involvement in planning remained limited. John Lindsay’s election in 1965 “represented a historic turning point,” with Lindsay’s “strong record on civil rights” and outreach to Black and Puerto Rican voters meaning that he relied on a wider coalition of voters than did Wagner (Viteritti Reference Viteritti and Viteritti2014, 15–16). Lindsay likewise supported the neighborhood-focused community groups that formed in the context of the War on Poverty.Footnote ¹⁷ Thus, Lindsay represents a politician in the model with more equal preferences across neighborhoods than prior politicians.

As mayor, Lindsay attempted to expand the power of neighborhood participatory institutions in his unsuccessful effort to create “little city halls” (Marcuse Reference Marcuse1987). Although Lindsay faced opposition from the city council (Viteritti Reference Viteritti and Viteritti2014), a formal role for these institutions in land use planning was established in New York’s 1975 charter revision. The votes of neighborhood planning institutions remained advisory – the city council still approved or rejected the community plans (Angotti Reference Angotti2008). Nonetheless, as in the model, increased concern by politicians for previously marginalized groups led them to establish participatory institutions, expanding opportunities for these groups to influence the planning process.

Portland

In the mid-twentieth century, Portland’s land use policy reflected the unilateral regime of the model. For instance, in the early 1960s, the Portland Development Commission displaced over 300 families (mainly Italian and Jewish immigrants) from the low-income South Auditorium neighborhood, ignoring protests by local residents (Wollner et al. Reference Wollner, Provo and Schablitsky2001). Starting in the mid-1960s, the Model Cities program provided funding for low-income, predominately African-American neighborhoods to organize. Other neighborhoods, such as in wealthier Southeast Portland, organized to demand funding for themselves. In 1970, Neil Goldschmitt was elected to the city council and established a reputation as a “neighborhood guy” (MacGillivray Reference MacGillivray2016). With support from neighborhood activists, Goldschmitt was elected mayor in 1972. We interpret Goldschmitt’s election to represent increasing equality within the politician’s utility across neighborhoods.

In 1974, Goldschmitt’s administration established the Office of Neighborhood Associations, which provided a budget and administrative support for neighborhood associations (Wollner et al. Reference Wollner, Provo and Schablitsky2001). Now formally established within the city government, neighborhood associations enabled local residents to collaborate with city staff on neighborhood land use planning, and at the same time, to lobby the city government on issues of neighborhood concern (De Morris and Leistner Reference De Morris and Leistner2009). Hence, greater political equality across neighborhoods led to reforms creating formal mechanisms for citizen participation, but these neighborhood-level institutions still lacked the power to approve or veto projects.