Political Leadership: A Pragmatic Institutionalist Approach

Political Leadership: A Pragmatic Institutionalist Approach
Robert Elgie
Palgrave Macmillan, 2018

This book provides a philosophically informed, institutionalist account of political leadership. It is rooted in a Peircean version of the American pragmatist philosophical tradition and privileges the study of institutions as a cause of leadership outcomes. The study includes identifying the psychological effects of presidentialism and parliamentarism on leader behavior, a study of the impact of institutions on electoral accountability for economic performance, studies of president/cabinet conflict in Europe, presidential control over cabinet composition in France, and constitutional choice in France and Romania. It adopts a multi-method approach, including a lab experiment, large-n statistical tests, and Qualitative Comparative Analysis, as well as two in-depth process-tracing case studies. The aim is to show that an institutional account has the potential to generate well-settled beliefs about the causes of leadership outcomes.

In this post, we outline the work in one chapter. In this chapter, we re-examine Hellwig and Samuels’ (2007) article on economic voting and the clarity of institutional responsibility. Like Hellwig and Samuels, we are interested in the relative effect of parliamentary and semi-presidential institutions on electoral accountablility for economic performance. We are also interested in exploring the effect of variation in presidential power on economic voting in this context. In short, we are interested in whether institutions condition the extent to which presidents and prime ministers are rewarded/blamed for good/bad economic performance.

To address this issue, we update Hellwig and Samuels dataset, noting certain revisions to the way in which they record the vote at elections with the aim of maximising the reliability of the values in the dataset. We then use exactly the same estimation technique as Hellwig and Samuels.

There is insufficient room here to go through the results in depth. (Which is just an ill-disguised invitation to buy the book). There is also no space to describe how the variables have been operationalised. Again, all that material is in the book. Here, we just wish to provide a flavour of the results.

We find support for Hellwig and Samuels’ basic finding that electoral accountability for economic performance is greater under high-clarity elections, i.e. where there is a single-party government, than low-clarity elections where there is not.

More interestingly, our results also show support for Hellwig and Samuels’ finding that the electoral accountability of the president’s party for economic performance is significantly greater during periods of unified government relative to cohabitation. Figure 1 reports the basic results of our models in the same way that Hellwig and Samuels present them in their paper.

Figure 1    The conditional effect of cohabitation in semi-presidential regimes on economic accountability

However, there are some differences between Hellwig and Samuels’ results and ours. Perhaps most notably, we find that electoral accountability for economic performance is significantly greater at presidential elections than legislative elections. This makes sense. At presidential elections, the clarity of responsibility is likely to be clearer because voters can hold a single person/party responsible for the state of the economy. This is the result that Hellwig and Samuels expected to find in their work, but which was not returned. Using the updated version of their dataset, we now find support for their intuition. (See Figure 2.)

Figure 2         The conditional effect of the type of election on economic accountability

While we are concerned with re-testing Hellwig and Samuels’ thesis, we are really interested in exploring how presidential power shapes the clarity of responsibility for economic voting. Hellwig and Samuels do not follow up on this issue in their article. So, we are trying to build on their work by integrating presidential power into their analysis.

We find that presidential power does help us to understand how institutions shape electoral accountability for economic performance. For example, when we include presidential power in the model we find that there is significantly greater economic voting at presidential elections with strong presidents. Again, this makes sense. When there is a strong president, the clarity of responsibility should be higher. Voters know better whom to reward or blame. By contrast, when there is a weak, non-executive presidency, we would not necessarily expect the incumbent president or their party to be held accountable for economic performance. (See Figure 3 relative to Figure 2).

Figure 3        The conditional effect of presidential power and type of election on economic accountability

In addition, we also find that electoral accountability for economic performance is conditional upon presidential power during cohabitation. In these periods, there is significantly greater economic voting during periods of unified government when there is a strong president. (See Figure 4 relative to Figure 1). In other words, the combination of unified government and presidential power shapes economic voting at elections under semi-presidentialism.

Figure 4         The conditional effect of presidential power and cohabitation in semi-presidential regimes on economic accountability

These are only a flavour of the results in the chapter. Spoiler alert, not all results are as expected. Most, though, are.

We would like to thank Hellwig and Samuels for supplying their dataset for replication purposes. Obviously, all results presented here and in the book are the author’s responsibility alone.


Hellwig, Timothy, and David Samuels (2007), ‘Electoral Accountability and the Variety of Democratic Regimes’, British Journal of Political Science, 38: 65-90.

Leave a Reply

Your email address will not be published. Required fields are marked *