Abstract: Standard theoretical arguments suggest that republics ought to grow faster than monarchies and experience lower transitional costs following reforms. We employ a panel of 27 countries observed from 1820‐2000 to explore whether regime types and institutional reforms have differential growth effects in monarchies and republics. A set of Barro‐type regressions show that there are no significant growth differences between the two regime types and that the effects of incremental reforms do not differ between them, but that those of large‐scale reforms do. Specifically, we find a strong “valley‐of‐tears” effect of large reforms in republics while monarchies benefit from such reforms in the ten‐year perspective adopted here. We offer some tentative thoughts on the underlying mechanisms responsible for the results.
Since the 1950s, the study of different political institutions has been part of political economy and political science under the joint heading of “public choice” and more recently of the ʺnew institutional economics” and the “new public economics”, and these lines of research have produced immense amounts of theoretical and empirical results regarding the possible consequences of political institutions. Nonetheless, only very few papers have been devoted to the study of monarchical arrangements. As noted in the introduction, many scholars seem to view monarchy as an essentially outdated institution not even worthy of academic interest. Yet, the study of monarchy as a political institution deserves wider attention: it has been around almost as long as mankind and historically has been the dominant regime type, and it arguably can have real social, political and economic consequences, even in modern times, where a non‐trivial number of countries continue to be monarchical in some form. In the following we shall offer some relatively simple, theoretical considerations that hopefully may enlighten our understanding of the potential differences between monarchies and republics on some central issues of political economy.
We have in this paper explored the long‐term growth rates in monarchies and republics as well as the growth effects in the medium and long run of institutional reforms in the two types of systems, and in particular of large reform. While economic growth and growth effects of reforms have also been the objects of several previous papers, the innovation of this paper lies in focusing specifically on republics and monarchies and in distinguishing between reforms in the two.
The results of regressing decadal growth rates in the period 1820‐2000 in 27 Western countries on institutional quality and reforms, as measured by the Polity IV index, reveal that large‐scale reforms induce a sizeable reduction in growth in republics, but a growth spurt in monarchies. As such, we find evidence of a “valley of tears” effect of substantial transitional costs in the former but not in the latter countries. This result is robust to excluding potential outliers and does not depend on differences in the timing of the reforms. Furthermore, monarchies have on average been able to introduce more sweeping reforms than republics when deciding on large‐scale institutional reorganization.
As such the results must be seen as falsifications of the two hypotheses tentatively suggested here, namely that republics will grow faster than monarchies and that the former also will recover faster from the initial adverse economic effects of larger reforms. The findings therefore give rise to a set of questions that we dare not attempt to answer in detail at present but nonetheless believe are necessary to raise in the context of our empirical results. For example, have monarchs exerted real influence over policy decisions? Have they instead exerted an indirect influence as political veto players or otherwise constrained governments’ discretionary power? Or do monarchs instead perform a socially stabilizing role, which insures countries against large social conflicts and economic disruptions following institutional reform episodes? Our results give rise to these questions but we cannot provide any answer to them. However, we may speculate about the reasons and suggest some possible interpretations.