What can urban policies and planning really learn from John Rawls? A multi-strata view of institutional action and a canvas conception of the just city

The basic structure of society and background (procedural) justice

In reconsidering Rawls’s own perspective, two main notions are crucial: (i) the idea of the “basic structure of society” and (ii) the idea of “background procedural justice”. These two notions are the core of Rawls’s view. Even if they are interconnected, each of them has its own distinctive features and implications.

Let us start by considering the idea of the basic structure.

If the question concerns what the subject of a theory of justice could be, there are three main alternatives (Bedau, 1978): (i) persons (e.g. their characters and/or their acts); (ii) patterns of possession (in terms of things, goods, resources, distributed among persons); (iii) institutions (especially public institutions).

Rawls defends a particular version of the third option, and this makes him uniquely innovative in a certain tradition of political philosophy (Barry, 1995: 214; Audard, 2007: 60-61). According to Rawls, the primary and specific subject of justice is actually the “basic structure of society”: that is, the way in which the major social institutions assign fundamental rights and duties, regulating the division of advantages that arises from social cooperation over time (Rawls, 1971: 7, 1993: 257-288 and 2001: 10-12, 52-57). The basic structure constitutes the social framework within which individuals’ and groups’ activities occur. ⁴

Rawls clearly states that his two principles must be applied primarily to the basic structure of society. In other words, they are not intended to guide everyday policies directly (e.g. urban policies and plans) but to design the basic institutional framework. The two principles of justice assess the basic structure by considering how it regulates accessibility to primary goods: that is, the all-purpose means necessary to enable citizens to pursue their different specific conceptions of good. Principles for the basic structure constrain and limit, but do not determine, the suitable principles for post-constitutional choices. In short, there is no attempt to define first principles that apply equally to all subjects and situations (Rawls, 1993: 258). To simplify, we can say that Rawls’s theory of justice is primarily a theory of constitutional design (Moore, 1979). In conclusion, “the difference principle is not a principle of public policy” (Brown, 2012: 628); it is a super-rule, a meta-normative principle. ⁵

Let us now consider the notion of background (procedural) justice

According to Rawls, distributive shares are a matter of procedural justice. “The intuitive idea is to design the social system so that the outcome is just whatever it happens to be” (Rawls, 1971: 84-85). In other words, procedural justice obtains when there is no independent criterion for a valid result: “instead, there is a correct or fair procedure such that the outcome is likewise correct or fair, whatever it is, provided that the procedure has been properly followed” (Rawls, 1971: 86). This point is stressed in many other subsequent works. For instance, in Rawls (2001: 54):

“When everyone follows the publicly recognised rules of cooperation, the particular distribution that results is acceptable as just whatever that distribution turns out to be […]. This allows us to abstract from the enormous complexities of the innumerable transactions of daily life and frees us from having to keep track of the changing relative positions of particular individuals”. ⁶

According to the idea of procedural justice, no distribution can therefore be assessed in isolation from the background system of which it is the actual outcome, and from what citizens have done in good faith in the light of established expectations (Rawls, 1971: 88). In this case, any state of affairs is acceptable if (and only if): (i) the state in question is the result of a procedure that is fair, (ii) the procedure has been actually carried out (Dupuy, 1988: 69). Note the importance of the latter clause. It would be illegitimate to claim that a particular state of affairs, which is to our advantage, is the just one because it could have been obtained by adopting a fair procedure. It is an unavoidable condition that the procedure had actually been carried out (Dupuy, 1988: 69).

The Rawlsian distinction between “procedural” justice and, so to speak, “non-procedural” justice is therefore different from the one in normal usage (Ceva, 2012; Meshelski, 2016), which is typical also in planning theory and urban studies (see e.g. the debate on “procedural justice” deemed typical of communicative planning approaches: Nylund, 2014; Fainstein, 2017). In planning theory, the distinction is usually interpreted as a distinction between formal justice of processes (i.e. who takes part in voting procedures and public and participatory debates, and how they do so) and substantive justice of outcomes (i.e. who finally obtains what). Fainstein (2005, 2009) highlights how we should have both. Innes and Booher (2015) observe how we cannot but have both: processes and outcomes in this sense are always in some way connected in each normative approach to planning.

In Rawls (1971: 517; 2001: 50), the distinction between non-procedural and procedural justice is, by contrast, a more interesting and original distinction not so much between substantive outcomes and formal processes as between allocative justice and background justice.

Avoiding Nozick’s challenge: beyond simplistic dichotomies between processes and outcomes

Rawls’s ideas of the basic structure of society and that of background procedural justice are crucial. As regard the former notion, Barry (1995: 214) write: “If Rawls had achieved nothing else, he would be important for having taken seriously the idea that the subject of justice is […] the ‘basic structure of society’”. ⁹ The thesis that certain basic institutions are the main and primary subject of social justice assessments seems convincing (Frankena, 1962; Barry, 1989; Höffe, 1987; Miller, 2001; James, 2005). Moreover, focusing on background justice (rather than on allocative justice) gives importance to both processes and outcomes in a way that appears appropriate to a constitutional democracy. ¹⁰

Combining the ideas of the basic structure and that of background justice enables Rawls ¹¹ – and any approach of a similar kind – to avoid Robert Nozick’s (1974) challenging criticisms of end-state patterned principles of justice ¹² (exemplified by the famous Wilt Chamberlain example ¹³ ). Moreover, taking Rawls seriously shows how Nozick’s arguments are not against redistributive approaches in themselves (i.e. against any redistributive approach, as Nozick seems to believe) but only against certain redistributive approaches (e.g. allocative conceptions of justice). As Lovett (2011: 63) notes, the conviction that Nozick’s arguments can be levelled against Rawls is based on a confusion (that also this article underscores):

“The confusion lies in assuming that the difference principle is meant to be applied directly to distributions of social and economic goods. This is not the case. The principles of justice, as Rawls stresses many times, are meant to apply to the basic structure of society – that is, to the design of society’s major social institutions and practices”.

In conclusion, Rawls is actually a kind of egalitarian, but not a time-slice egalitarian; the pattern that he intends to weave into society is “a pattern of equal status”, to be applied not so much to a distribution as to “an ongoing relationship” (Schmidtz, 2006: 201; see also Pogge, 1989: 28ff.). ¹⁴

First point: A multi-strata conception of institutional action

Like that of Rawls, many philosophical theories of justice are above all theories regarding the basic structure of society (Barry, 1989; Höffe, 1987). Therefore, their application to single urban policies and plans, at the local level, is neither straightforward nor immediate (Campbell, 2006; Campbell and Marshall, 2006). This is not due to the gap between theory and practice; rather, it is due to the intrinsic nature of many theories of justice. Among the few authors who have clearly recognised this in the planning literature, with specific reference to the Rawlsian perspective, is Klosterman (1978: 45): “Rawls does not suggest that these principles (i.e. the two principles of justice) be used to evaluate specific governmental actions and policies but, rather, limits their use to more fundamental decisions concerning the nature of major social institutions”. Along similar lines, Campbell (2006: 93) writes: Rawls (as well as Habermas) does not intend “his constitutional-level conceptualisations of justice to be applied to the everyday situated judgments with which planning and many other areas of policymaking are concerned”. As regards transportation planning, Nahmias-Biran et al. (2017: 199) observe: “Rawls’s theory of justice deals with the basic structure of society; therefore, it would be a mistake to simply adopt one of Rawls’s principles of justice to the domain of transportation, as has been done in the transportation literature”.

An article in which Howe (1990: 140) discusses ethics in planning contains this comment: “Cleveland’s Rawlsian comprehensive planning effort ultimately failed because as a society we do not accept the difference principle, or the absolute centrality of justice”. In the perspective propounded here, the problems faced in Cleveland were more connected to the questionable attempt to apply a constitutional-level principle to each everyday urban policy.

In their study on elected officials in urban policymaking (which draws on interviews with around 100 elected officials in 12 cities), Schumaker and Kelly (2013: 93) find that “officials have some allegiance to the idea that they should focus on the poor, as suggested by the justice principles of John Rawls, but making a Rawlsian argument is unlikely to be effective because few officials think they should always focus on the poor. Perhaps officials doubt that every policy decision has to improve the conditions of the disadvantaged” (see also Schumaker and Kelly, 2012). Whatever we might think about the motivations generating the answers of the local officials, it is not so strange to find the conviction that we cannot adopt the difference principle for every local public decision; this is simply impossible and far from Rawls’s own view.

Clearly, we can try to imagine how to develop Rawls’s general view more adequately in order to conceive an approach suited to local institutions: that is, a substantive and operative idea of the just city. What this article stresses is that, if we want to follow this line (i.e. build a more convincing Rawlsian approach to urban issues, or any other justice-based approach to this kind of issues), we have to keep in mind that this endeavour will have to fulfil three conditions.

First (let us call this the no-flatness clause), it has to incorporate the idea that this should take place by imagining an appropriate sequence of institutional levels. The institutional environment is not flat; rather, it is organized in layers. Local policies and plans cannot but be understood as developed within – and conditioned by – a broader institutional framework (Wolman and Goldsmith, 1992; Sellers, 2002; McCabe and Feiock, 2005). ¹⁶ In particular, when considering issues of justice (in the planning field too: Moroni, 2018 and 2019), a clear distinction between the constitutional and the post-constitutional level of discourse is necessary: that is, a distinction between discourses concerning the basic structure of society and discourses concerning everyday policy decisions. The constitutional level is the level at which we can imagine a group of constitutional delegates conceiving and designing a constitution. Many questions that are crucial for land use issues are constitutional-level questions (e.g. the problem of expropriation for public utility, taking issues, exclusionary zoning: Siegan, 1997; land-value taxation as a pre-distributive mechanism: Kerr, 2016; issues concerning evictions: Strauss and Liebenberg, 2014; the possibility to work on the street: Meneses-Reyes and Caballero-Juárez, 2014; the possibility to build and frequent any type of place of worship: Chiodelli and Moroni, 2017; land grabbing: Montilla Fernández and Schwarze, 2013). The post-constitutional level is the level of the policy activities (e.g. regional and local planning) that occur within the accepted constitutional constraints. These two levels can be understood as analytical levels: that is, as standpoints that anyone can – at any time and even merely hypothetically – assume in order to consider specific issues at the appropriate level and deal with them by acknowledging the argumentative requirements suited to that level. In the case of constitutional democracy, they correspond to real-world situations; that is, to real deliberative stages. At the constitutional level the “veil of ignorance” is thick, while at the post-constitutional level it is partially lifted; partially because some empirical facts are known at the latter level (e.g. the contingent status of a particular metropolitan area), but an impartial view on society (i.e. on individuals comprised in that area) is still required for any public decision. (We can interpret the veil hypothesis as a “thought experiment”, but we can also imagine inserting actual “veil rules” at different decision-making stages; in this case, the “veil approach” becomes an institutional-design strategy that may be implemented, at various institutional levels and through various tactics, in order to impede the adoption of non-impartial views by decision-makers: Vermeule, 2001 and 2007). This theoretical approach has, so to speak, a “stage-wise” structure. In other words, it is an approach that recommends thinking sequentially, from the basic constitutional framework to the development – and subsequent implementation – of increasingly focused laws and policies. Therefore, this view rejects the idea that it is possible to identify a single principle of justice – or a single set of principles – that applies to any level in an undifferentiated manner. Principles suited to the different situations should instead be established step-by-step in an appropriate sequence. Note that this is another way to criticise traditional utilitarianism – that so strongly influenced planning (see e.g. Allison, 1975 and Sillince, 1986). In fact, in the case of utilitarianism, there is only one principle – i.e. the maximisation of collective utility – that is valid for every situation.

Second (no-isomorphism clause), local issues will probably imply something quite different from principles introduced primarily in order to judge the basic structure of society, such as the two Rawlsian principles. This means that principles X employed to design the constitution could imply constitutional principles Y, which in turn imply, at the regional and local level, principles W. To provide an example, in certain countries of the Global South poverty problems are also created by the fact that women do not have the right to own land (Tamale, 2004; Kalabamu, 2006; Mitchell et al., 2008; Linonge-Fontebo, 2018): a principle to design or revise a constitution, focused on creating the best condition for the worst-off, could in this case suggest a constitution granting the right to own land to all, and post-constitutional policies favouring and fostering this right. ¹⁷ Another example may be this: certain zoning policies favour the rich and create problems for the poor (e.g. large minimum-lot zoning, that is, a form of zoning reducing housing affordability: Atash, 1990; Trambley, 2020; Karki, 2015): a principle with which to assess a constitution focused on creating the best conditions for the least advantaged could in this case entail, through the various stages, post-constitutional local policies that avoid certain forms of zoning. In other words, intrinsically “redistributive” principles for the design or revision of a constitution do not directly and necessarily imply merely “redistributive” constitutional and post-constitutional strategies. (Clearly, also the reverse might be true: non-intrinsically redistributive principles for the design or revision of a constitution could entail constitutional and post-constitutional redistributive strategies).

Third (no-exhaustiveness clause), although justice makes stringent moral demands upon institutions (including local institutions), this does not mean that there are no other moral demands to which people – as professionals or ordinary citizens, for example – are subject, for instance in urban contexts; simply, the latter are not demands of (social) justice (Porter, 2009). In other words, justice does not cover people’s entire normative world and apply only to specific institutional issues (Galston, 1980: 108; Forst, 2012: 9). Trying to include any moral question into just city issues simply creates theoretical confusion and ineffective local policies and plans. To avoid misunderstandings, a fundamental point should be specified here. Each theory of justice, and a theory of the just city as well, must accommodate different and multiple values (for example, freedom, security, welfare): various substantive theories propound different hierarchical schemes and trade-offs in this regard. The point is that this does not – and could not – cover the entire range of values that may be important for a contemporary urban society. Loyalty towards friends is, for example, one of them. However, the latter do not generally concern a normative theory of institutions, that is, a theory of social justice.

Second point: A canvas conception of the just city

Moreover, I suggest that any kind of theoretical development at any institutional level will have to maintain the idea of background justice. Especially in the case of a constitutional democracy, any theory of justice cannot but be a theory concerning the justness of the (public) framework institutions, both of the major institutions of the basic structure and of the lower-level institutions (down to the local level), once an appropriate sequence is imagined to consider them as well. This implies that a theory of the just city has to fulfil three further conditions.

First (anti-contingency clause), contingent allocative distributions – involving specific individuals with known preferences – are not an issue for the local public institutions once (national and) local background justice has been satisfied; that is, once just (national and) local public institutions have been granted. We cannot look at the concrete allocation of goods or resources in an urban area at a single moment, and in itself, and ask ourselves if it is just or unjust. To provide an example: if a particular business sector in city A, or a particular local industry plant in city B, makes extra profits in a certain period, while respecting all the public rules, this cannot be an issue of justice. And here is another example: if certain specific buildings are vacant in city C at time T (a situation occurring without the breach of any general public rule), this cannot be an issue of justice merely because certain specific individuals in that same city are simultaneously unable to find a house to rent (Moroni et al., 2020; De Franco, 2022). In short, social justice in the city is not an issue concerning particular individuals in specific time slices; rather, it concerns “whatever citizens”, even “potential citizens”, in an ongoing common endeavour. How otherwise could it be possible to take into consideration also the future inhabitants of a city (i.e. future generations)? In conclusion, we need a dynamic approach to the just city and not a static one (based on sections in time). The just city is not the collection of successive desirable “snapshots”; it is instead the desirable “institutional script”.

Second (open-endedness clause), many different physical urban configurations, as well as many different distributions of urban goods and resources, may effectively satisfy a normative idea of (local background) justice; there is no way to judge them and order them from, for instance, the most fair or equitable to the least fair or equitable, once principles of local background justice have been fulfilled. Let us for instance assume that the rules of justice in city D do not impede the distributive patterns X, Y, and Z (nor any of the urban agents’ actions potentially producing them): in this case, no occurrence of one or the other of these situations can be considered unjust. In short, the public rules of justice are the software requiring and granting certain standards, but many unpredictable operations – and combinations of operations, as well as their physical outputs – may be compatible with them.

Third (anti-adhocness clause), local public authorities cannot intervene in an ongoing manner – and independently from local background justice – to obtain particular results. This would make the justness or unjustness of urban situations intrinsically transient, and it would frustrate the expectations and long-term plans of urban agents. In other words, urban agents who have developed legitimate expectations, in compliance with the public framework, at time T1 (e.g. a shopkeeper who has just opened a new business, a developer who has started constructing a building, a house tenant who has entered into a new lease agreement) cannot find these expectations thwarted at time T2 simply because local authorities are reacting in an unpredictable way to contingent urban situations (Brown, 2012: 618). Public institutions – and local institutions also – have to grant a certain degree of general predictability (Moroni, 2007) and avoid generating normative uncertainty (i.e. a situation of unpredictability of upcoming measures). Here the point is rediscovering a model of policy-making based on principles, in place of an extemporary model based on spot-detections and spot-reactions (Landis, 1992; Barnett, 1998; Buitelaar et al., 2021). As Waldron (1989: 88) notes, public action should be guided by accepted, recognizable principles and “not merely by orientation towards certain aims or goals”. The point is that, “to plan our lives, we may need to know, reasonably far in advance, how the state can be relied on to behave in certain circumstances”; however, “such knowledge is in principle unavailable whenever state behavior depends on some official’s discretionary estimation of social costs and benefits at the time” (Waldron, 1989: 88). From this perspective, local authorities should not, for instance, frequently revise local taxes, fees and charges, or negotiate exactions with developers in a discretionary fashion (Selmi, 2011; Hills and Schleicher, 2015), merely to increase local revenues at any occurrence.

In short, we should not expect or require that all cities that satisfy certain background principles of justice exhibit any specific common allocative pattern. Background (local) justice is a sort of “canvas” that cities should share even if local contingent individuals’ and groups’ interactions create different concrete “paintings” over time. In other words, the local canvas should be the common granted justice platform starting from which urban life can variously develop. Without the canvas, “painting” by citizens cannot appropriately start; but, at the same time, it is the “painting” by citizens that create the various concrete cities. A theory of the just city therefore defines the framework and leaves to society – individuals and groups – to work out the details. Consequently, two cities may be equally just whilst the way of life of their citizens may be significantly different. Note that a canvas conception of the just city is perfectly compatible with recognition that a city is a highly complex system characterized by emergence and unpredictability (Moroni and Cozzolino, 2019).

To summarise, the (just) city is a (fair) system of social cooperation over time: countless present and future citizens intentionally and unintentionally cooperate to produce the social-spatial resources on which their claims are advanced. The central question for a (canvas conception of the) just city is therefore the following: How can we improve the chances that unknown urban agents have of successfully pursuing their equally unknown and constantly changing purposes in a complex, open-ended socio-economic environment?

In conclusion, what I call the “canvas conception of the just city” is clearly nothing more than the meta-theoretical skeleton of an approach to the just city, but I hope it may at least suggest – in the limited space of an article – a fresh line of discussion and indications for further research. ¹⁸