According to O’Donnell, in the functioning of contemporary societies political institutions ensure a fundamental level of communication and aggregation between different social groups and between structural factors and individuals. The institutional level affects, in a fundamental way, the model of the organisation of the society: some participants in the political process become its representatives, others are excluded from it. The author observes the costs of institutionalisation in the form of excluding some participants of the political process or bureaucratisation. He finds an alternative in subordinating social and political life to the principles of the prisoner’s dilemma (the institutions aim at turning cooperation into rational choice, and the prisoner’s dilemma appears when the rejection of others’ decisions and refusal to cooperate become rational) (Marczewska-Rytko, 1996: 92-93). Noninstitutionalised democracy is characterised by a limited scope, weakness, and limited number of political institutions. Their place is occupied by the practice of clientelism, mono-party system and corruption.
Delegative democracy is based on the principle granting full liberty to the winner of the presidential elections. He is only constrained by the interrelations between different powers and the constitutionally defined term of office. The president is treated here as the embodiment of the will of the nation and the principal guardian of its interests. His political base is a political movement eliminating all factions and conflicts usually associated with the functioning of political parties. In this kind of democracy, the president appears as a person who is above all divisions resulting from party conflicts or rivalry of different interest groups.
Delegative democracy is largely a majority democracy. This means that a majority elevating an individual to power constitutes itself in the elections. The elected individual becomes the embodiment of the highest national interest for the duration of his term of office. The question arises as to the difference between delegative and representative democracy. Representative democracy entails two kinds of accountability: vertical and horizontal. Vertical accountability consists in the accountability of the elected representatives towards their electorate. Horizontal accountability denotes the functioning of a network of different, independent institutions which may question the actions of the rulers, or even punish them. In delegative democracy we are faced with vertical accountability and the lack or limited functioning of horizontal accountability.
Both types of democracy are also differentiated by the criterion of political decisionmaking. In representative democracy, policy is conducted by different, relatively autonomous forces. Thus, on the one hand, the process of decision-making is prolonged. On the other hand, however, this situation provides protection against serious errors and allows the division of accountability among different forces. In delegative democracy, the president enjoys the privilege of bearing practically no horizontal accountability. He can make quick political decisions, although at the price of a higher probability of committing a serious mistake. In this situation accountability is associated with one person. The president, as it were, cuts himself off the majority of institutions and organised interest groups, assuming full accountability for the result of his policy. Consequently, the popularity of the president in such a system is constantly oscillating between enthusiastic support and total rejection. These plebiscitary tendencies of delegative democracy can be observed in many Latin American, Asian, African, and post-communist countries (Kubicek 443). This kind of government has been analysed within the framework of studies on authoritarianism and given the name of Caesarism, Bonapartism, or populism (Marczewska-Rytko, 1992). According to O’Donnell. delegative democracy does not contribute to the constructing and strengthening of political institutions, and the economic policy conducted in such a system does not have to be perceived as wrong (O’Donnell, 1993: 1355-1369).2