1. The Policy Case for Metropolitan Transport Authorities

One ambiguity in this second notion regards the fact that organisations can be considered systems of rules structuring the individual actions of people within administrations. In this case, they overlap with institutions intended as rules of behaviour. The apparent confusion is easily resolved by North, who argues that the critical distinction is between organisations as players and organisations as rules (North 1990). Hence, individuals within organisations will be subjected to both institutional and organisational rules, while organisations as players will pursue their goals within a given institutional setting.

This can be done following the syntax of the ‘grammar of institutions’ developed by Crawford and Ostrom. In this view, institutions are made of five components: Attributes, Deontic, Aim, Conditions, and Or Else. Attributes distinguish as to whom the institutional statement applies (e.g., female, degree holders, and so on). Deontic indicates the modal verb specified by the institutional rule, indicating if it permits and obliges or forbids a certain action. Aim is a holder for the specific action or outcomes implied by the Deontic. Conditions specify when, where, how and to what extent a certain Aim is permitted, whether it is obligatory or forbidden. Or Else is a label that indicate sanctions imposed for non-compliant behaviours. The syntax permits the ability to distinguish institutions intended as rules from two alternative views: institutions as equilibria and institutions as norms (Crawford and Ostrom 1995). Institutions as equilibria (Riker 1980) come out of shared strategies among individuals. They maintain stability because they result as the optimal response of the interaction among rational actors. In the syntax, they are identified by an attribute, an aim and certain conditions, but have no deontic and externally enforced sanctions. Institutions as norms are intended as social conventions and are the typical object of study of sociological research. With respect to institutions as equilibria, norms have a connected deontic, but are not provided with external sanction. Individual agents respect norms because these are internalised into a code of conduct and as a way of social acceptance.

Van de Velde (2005) distinguishes four components of ‘good-practice’ integration. The basic level discusses the integration of information concerning routes (i.e., having a common map), timetables and fares, plus an effort in providing unitary communication. The second is the typical function of the metropolitan agencies, including the development and management of ticket and fare integration. A more advanced form of integration concerns network integration. Supposedly, this would permit the optimisation of systems by coordinating investments in infrastructures and interchanges in the planning of services and in their operation. Finally, a wider level of integration is reached by coordinating with non-public services (such as private cars or bicycles) or by pursuing inter-policy coordination (e.g., with urban planning, environmental and social policies). In the same volume, Viegas (2005) revolves the concept on the demand side and distinguishes three dimensions: physical integration, tariff integration, logical integration (information and perception of the system).

This is no place for discussing whether and how far integration is a good principle per se. Clearly, the different forms of service integration discussed in note 3 have beneficial effects on users. In comparing different organisational arrangements with an application to transport, Bendor has shown the importance of parallel non-integrated systems in enhancing system reliability (i.e., avoiding system failures) and enhancing innovation (Bendor 1985). This is an appealing treatment of the importance of overlaps and redundancy, commonly dismissed as bureaucratic disorders.

In Barcelona, some tariff integration existed as an agreement between service providers before the establishment of the ATM. In London, the first regulatory body for transport dates back to 1933. In 1908 the first common map of the transport system had been made, common signals had been agreed, negotiations on tariff integration were ongoing, and works to increase and improve exchange stations had begun. All these came out of simple agreements among independent partners.

The general design principles mentioned are revisability, robustness, sensitivity to motivational complexity, publicity and variability. Revisability means that institutions should be prepared to change, since trials and errors will be the normal way by which fallible individuals will approach social problems. Robustness counterbalance revisability, in that institutions should resist to minor changes to provide stability. Sensitivity to motivational complexity means that institutions should be open to diversity of preferences and avoid institutional colonisations and monopolies. The publicity principle is a test for most institutions in that they should be defensible in public. Finally, fallibility and the need for institutional experimentation explain the variability principle, typically found in plural settings, such as federal countries (Goodin 1998).

As mentioned, several intervening variables may affect the relation between institutions and policy outcomes. However, other problems may downplay the results of institutional change. In situations of ambiguity and uncertainty, for instance, structural choices, such as the establishment of a new bureaucratic agency, may be due to mere symbolic moves, only motivated by showing interest to a problem. In this respect, the institutional change will be at best a first step towards the solution of the problem, but all effects will depend on the subsequent actions taken by the established agency. As I will show in the next chapter, understanding the preferences of institutional designer is fundamental to predict the likelihood of such changes to produce results.

Game theory elaborations were made using the Gambit software (McKelvey et al. 2007).

When played simultaneously, games of pure coordination such as the battle of the sexes are telling of the importance of information. The battle of the sexes is a two-player coordination game in which the two players prefer to coordinate on their decisions with respect to all non-coordinated outcomes, but have nonetheless different preferences over the final choice. The example is given by a couple deciding on their evening plan, the husband willing to go to a football match, the wife to the opera. Both prefer to stay together, though their best solution is going to the show they prefer. If there is no communication and moves are played simultaneously, there is no way to secure the maximization of their utility. They could actually end up with the worst possible result, the husband to the opera and the wife to the football match.

The story behind the game is that of two criminals arrested and put in separated jails with no possibility to communicate. Since the police has not enough evidence to charge them with the principal crime, the officers decide to charge them with a minor offence, corresponding to 1-year imprisonment. However, the police offers both prisoners to betray their partner and confess. If both betray, they will get 2 years each. If one betrays and the other stay silent, the betrayer will be set free, while the other will get 3 years. Given the structure of the payoffs, the absence of communication and trust, and the non-repeated character of the game, the dominant strategy for both players is to betray, in so getting more years than if both had cooperated. In acknowledging the diffusion of the game, Poundstone reports several literary examples, included the one between Tosca and Scarpia in Puccini’s opera (Poundstone 2011).