The sub-project ‘Agency and the Polity’ focuses on the ‘macro’ level of socio-political institutions, and the degree of agency – of political rights – that have been acquired there. It is well known that in the very long run there is a strong correlation between the quality of political institutions and economic development, but it is still unclear which is causing which. One school argues in the footsteps of Douglass North that constraining the executive – via democratic institutions – is a pre condition for economic growth; the Glorious Revolution of 1688 is classic example of such a link (North and Weingast 1989; more recently North, Wallis and Weingast 2009) Another, perhaps equally influential tradition, argues on the other hand that democratisation will induce the state to introduce redistributive policies which may harm economic growth (Acemoglu and Robinson 2006). Lindert (2004) on the other hand demonstrated that there is indeed such a link – democracy tends to lead to more social transfers – but their effects on economic growth are mixed, and social transfers on balance do not lead to a deceleration of growth. Economists therefore do not agree about the long term impact of democratic institutions on growth. There is also no consensus about the question what exactly causes the relationship between GDP per capita and democracy. The most promising contributions however focus on the role played by human capital formation – this appears to be the key variable explaining why improvements in institutions are realized (Glaeser et.al. 2004). In a related contribution, Glaeser et.al. (2007) have demonstrated that the main causal connection is that high levels of education create a class of people with clear interests in high quality political institutions. Increased education thus leads to democracy and enhances the stability of such institutions.
Within the context of this debate, the contribution of this sub-project is first to produce the global datasets that will make it possible to develop a better understanding of the changes since 1850 in the quality of institutions and the degree of agency enjoyed by the people involved. Secondly, we will develop and test new ideas about the links between gender relationships, patterns of household formation and changes in the quality of institutions. The key hypothesis that we will explore is that there are links between the character of institutions at the micro level – in particular the family and the household – and the capacity of societies to develop democratic institutions. Already Todd (1987) put forward the hypothesis that deep-rooted family norms determined a society’s conception of equality (determined by inheritance rules) and authority (determined by co-residence patterns), and thus the foundation for democratic institutions. There is a rapidly growing literature arguing for similar links between gender relations and ‘good governance’; Branisa, Klasen and Ziegler (2009) have in a recent review of this literature argued that ‘social institutions related to high gender inequality inhibit the building blocks of good governance. In societies with social institutions favoring gender inequality political systems will be less responsive and less open to the citizens, so that voice and accountability will be reduced’. By including family-system indicators into the equation (for instance as a control or instrument variable), we will get more grip on the complex interactions between democracy and economic development.
This analysis will be combined with a study of the links between the quality of political institutions and human capital formation (in its turn, as we saw already, affected by family systems). Again, it is very clear that there exists a strong correlation between levels of education and democracy, and that the causal links are complex. By analysing transitions from high levels of authority to low levels, they show the impact of education on democratic development. We will test this hypothesis by extending the time period studied backwards to 1850, and by analysing interactions with the microvariables mentioned already.
An important part of the project will be focused on measuring different dimensions or components of political ‘agency’. We will use the already mentioned Polity IV dataset as a starting point, but the quality of that dataset for this kind of research has to be tested, and additional data have to be entered. For example, the Polity IV dataset does not include colonies, which means that large parts of the 19th-century world are not covered. Moreover, additional data on actual political participation (active and passive voting behaviour), preferably of men and of women, have to be collected and linked to quality of institutions. Not only the ‘inputs’ into the political decision making process are relevant here, the analysis should also include its ‘output’, in particular government policy towards education and human capital formation in general (building on the important work already done in this field by a.o. Peter Lindert 2004).
This research is being conducted by Selin Dilli at Utrecht University.