Category Archives: Presidential power measures

New book series – Palgrave Studies in Presidential Politics: Robert Elgie and Gianluca Passarelli (series editors)

We are announcing a new book series, Palgrave Studies in Presidential Politics. The series is edited by Robert Elgie and Gianluca Passarelli and the books will be published by Palgrave Macmillan. The series will include books on all aspects of presidential politics. We are currently accepting proposals for books in the series. The first volume, authored Philipp Köker, will be published in 2017.

Palgrave Studies in Presidential Politics will include books on contemporary presidencies, including presidential powers, the administrative presidency, and presidential advisers, as well as the history of presidential offices, and presidential biographies. The series will also include books on presidential elections, including presidential party politics, and the media and presidential communication.

The series will focus on presidents throughout the world including the US, Latin America, Europe, Africa, and Asia, including both directly elected and indirectly elected presidents. The series will publish single-country and comparative studies of presidential politics. The series will also publish books on individual presidents. The series will focus primarily on empirical studies of presidential politics, but it could include volumes on conceptual or theoretical aspects, such as how to measure presidential power.

The series will publish books that look at the reform of presidential politics, e.g. the reform of presidential elections. However, it will not publish obviously partisan, clearly normative, or personally critical studies of presidents or presidential politics. The series will have a disinterested, academic focus.

The series will normally take the form of 80,000-word monographs, or edited volumes. However, shorter books, or Palgrave Pivots, will also be considered. To submit a proposal, you should complete a proposal form. These are available from Ambra Finotello (, or from the series editors.

For further information about the series and to submit a proposal for consideration, please contact Ambra Finotello ( at Palgrave, or the series editors, Robert Elgie (, and Gianluca Passarelli (

Feel free to send an informal e-mail to the series editors if you wish to discuss a book idea prior to the formal submission of a proposal. We look forward to hearing your ideas for books and to receiving your submissions.

Anna Fruhstorfer – Constitutions and Presidents. How formal rules constrain and empower

This is a guest post by Anna Fruhstorfer. It summarizes the main argument and findings of her PhD thesis, Constitutions and Presidents. How formal rules constrain and empower, which was Defended at Humboldt-Universität zu Berlin in July 2015 (PhD committee: Silvia von Steinsdorff, Zachary Elkins, Ellen Imoergut.



Although presidential power is a hot topic in political science research, little attention was paid to presidents in non-presidential systems with a powerful parliament and prime minister. So far, scholars have mostly been focusing on semi-presidential systems, whereas parliamentary systems with an indirectly elected president like in Germany or Estonia are hardly ever discussed (with the exception of Tavits 2009). Thus, in my research, I study presidential power in both parliamentary and semipresidential systems. In addition, little is known about the role of constitutions in presidential power. Treating constitutions as the explanatory variable is unusual. However, some influential studies have sought to explain the central role of constitutions in the conceptionalization of presidential power. From different angles, Amorim Neto and Strøm (2006), Tavits (2009), Schleiter and Morgan-Jones (2010) or more recently Bucur (2013) have significantly contributed to our understanding of presidential power. Nevertheless, because these studies are mostly concerned with explaining why a discrepancy between constitution and reality occurs, they do not pay much attention to how constitutions influence reality. In my thesis I answer the how question. I argue that specific characteristics of constitutional power make presidential institutions more sensitive to outside influences than others. Therefore, presidents that act within these institutions adapt their behavior accordingly. What emerges from these actions and what we can observe are different patterns of presidential leadership.

Empirically, that means that I conduct a comparative case study of 46 countries for a time frame of up to 75 years.[1] First, I compile an original dataset for these countries (CPS dataset)[2], which is based on a new measurement tool of constitutional presidential strength (CPS index). Secondly, using the dataset, a principal axis factor analysis is used to confront the unidimensional perspective of presidential power and form two dimensions of constitutional power. This two-dimensional perspective then lays the ground for the third step; a typology of presidential institutions with four types (CPS typology).

Measurement of constitutional presidential strength

In the course of the development of the stated argument and the conceptualization of power, it became clear that established measurement tools of presidential constitutional power do not always adequately describe the president’s role in parliamentary systems. Nevertheless, these facts were necessary for my research project. Hence, I have developed the index of constitutional presidential strength (CPS) for this study. It advances established tools to better capture the functional logic of parliamentary systems, to facilitate both low-level and high-level constitutional competences, and to enhance methodological and conceptual issues.

Conceptually, the CPS index emphasizes the functional logic of parliamentary systems. This means that it treats the power distribution with the government’s survival placed in the parliament’s hands as its most important element. Methodologically, the CPS index builds on Fortin’s (2013) description of measurement shortcomings of other measurement tools. The CPS dataset, established by implementing this measurement tool, provides a unique data collection of presidential power. From a historical-comparative perspective, the measurement and the data display a picture of great diversity of presidential power within a system of checks and balances. As a result, I observe consistently higher values compared to the normalized overview proposed by Doyle and Elgie (2015), as well as to individual measurement tools. The main reason for these differences is that the CPS index provides a more pronounced portrayal of the presidential role in power-sharing modes. The differences in the constitutional power in Germany and India as parliamentary systems and Austria and Mozambique as semi-presidential systems stand in stark contrast to each other. Some directly elected presidents are therefore not even strong on paper, such as in Austria, Iceland, Ireland, Macedonia or Montenegro, while indirectly elected presidents are already strong on paper, e.g. Bangladesh (1972-1974; 1991-2013), Estonia, India, Latvia, or Slovakia.

From a large-N perspective, empirical evidence shows that the two groups of directly and indirectly elected presidents differ significantly from each other. Hence, statistically, direct election goes hand in hand with a higher level of constitutional power (as well as the president’s role in the cabinet and the command of the armed forces)

T-Test for differences between directly and indirectly elected presidents

Levene’s Test for Equality of Variances t-test for Equality of Means
F Sig. t df Sig. (2-tailed) Mean Difference Std. Error Difference 95% Confidence Interval of the Difference
Lower Upper
Constitutional Power Equal variances assumed 129.167 .000 -22.415 3285 .000 -6.12560 .27329 -6.66143 -5.58977

Patterns of presidential de-facto power

Based on this comparative perspective, I further stressed the effect of constitutions on how their power shapes and determines presidential behavior in decision-making. Whereas these effects are diverse, there are also clear patterns. The strong-weak spectrum of constitutional power cannot fully account for these different patterns of de-facto power. An exploratory factor analysis (PA) of the measurement data indicates that constitutional power is indeed a two-factor construct (see also Fortin 2013).

These 2 dimensions are the basis for a typology of presidential institutions with 4 types, which can be characterized by the differences in their level of discrepancy between constitution and reality. Semi-presidentialism research has already shown that countries have alternate patterns of leadership (Elgie 1999, 283). These patterns vary between “dominant pattern(s)” (ibid.), be it either a dominant president, a dominant executive, or “a shift from one dominant pattern of leadership to another” (ibid.). What I argue and provide evidence for on a case-based description, is that these patterns are driven by the constitutional structure.

Presidential institutions labeled the notary (e.g. Austria, Germany, Albania, Czech Republic), i.e. with little power on both dimensions, are not able to change their de facto role. Empirically neither a patrimonial leadership legacy combined with a problematic democratic development, such as in Albania, nor exceptional political situations, such as in 1999/2000 Austria, nor even the newly introduced direct election in Czech Republic allow for an increase of presidential de-facto power.

The same pattern can be observed for presidential institutions like the almighty (e.g. Georgia, Bangladesh, Ukraine). Presidencies with above-average competences on both dimensions are so powerful, that they do not (have to) vary their de facto role. These almighty presidencies are largely insensitive to outside influences. Their dominance in most power-sharing situations offers hardly any room for other political actors to establish an influential position. The strong negative correlation with the democracy level of this type is therefore no surprise.

For presidencies labeled custodian and firefighter (e.g. France, Slovenia, Moldova, Poland, and Ukraine), varying patterns of de facto power can be observed. These cases confirm expectations and established literature regarding the questions of why the de-facto power of presidents varies[3] – surprisingly both for directly and indirectly elected presidents within these two types. The Estonian case (and Moldova after 2000), for example, provides some evidence for different patterns of presidential de-facto power. Presidential involvement or ‘activism’ (Köker 2014) increases in times of cohabitation; Estonian Presidents for example veto legislation more often and with a higher frequency in this situation. In times of cohabitation, indirectly elected presidents do not have roots in the ruling party. In most cases, the constitution does not even stipulate a role in cabinet meetings. Presidents therefore lose any influence on the decision-making process within the cabinet and the parliamentary majority. Thus, it is no surprise that these presidents use their legislative veto power frequently.


Research on presidential power points to multiple, interrelated causes for the varying de-facto power of presidents. I claim that constitutions structure the choices and thereby create a path for the direction of the power distribution. By defining how the game is played and laying the ground for the battle, they frame who gains and who loses power. This argument has been illustrated for several countries, but this can only be the starting point for further research. A large-N test of this may use for example the number of candidates in presidential elections as proposed by Cheibub and Chernykh (2008) as a dependent variable for a comparison of presidential de-facto power in different democratic and regional settings.


Amorim Neto, Octavio, and Kaare Strøm. 2006. “Breaking the Parliamentary Chain of Delegation: Presidents and Non-partisan Cabinet Members in European Democracies.” British Journal of Political Science 36 (4): 619–43.

Bucur, Cristina. 2013. “Who fires ministers? A principal-agent approach to ministerial deselection.” PhD Thesis, School of Law and Government, Dublin City University.

Cheibub, José A., and Svitlana Chernykh. 2008. “Constitutions and Democratic Performance in Semi-Presidential Democracies.” Japanese Journal of Political Science 9 (3): 269–303.

Doyle, David, and Robert Elgie. 2015. “Maximizing the reliability of cross-national measures of presidential power.” British Journal of Political Science.

Elgie, Robert. 1999. “Semi-Presidentialism and Comparative Institutional Engineering.” In Semi-Presidentialism in Europe, edited by Robert Elgie, 281–99. Comparative European Politics. Oxford [u.a.]: Oxford Univ. Press.

———, ed. 1999. Semi-Presidentialism in Europe. Comparative European Politics. Oxford [u.a.]: Oxford Univ. Press.

Fortin, Jessica. 2013. “Measuring presidential powers: Some pitfalls of aggregate measurement.” International Political Science Review 34 (1): 91–112.

Hicken, Allen, and Heather Stoll. 2013. “Are All Presidents Created Equal? Presidential Powers and the Shadow of Presidential Elections.” Comparative Political Studies 46 (3): 291–319. DOI: 10.1177/0010414012453694.

Köker, Philipp. 2014. “Czech Republic – A new government and the evolution of semi-presidentialism.” Accessed October 20, 2014.

Schleiter, Petra, and Edward Morgan-Jones. 2010. “Who’s in charge? Presidents, assemblies, and the political control of semipresidential cabinets.” Comparative Political Studies 43 (11): 1415–41.

Tavits, Margit. 2009. Presidents and Prime Minsters. Do Direct Elections Matter? Oxford [u.a.]: Oxford Univ. Press.


[1] For this I gratefully acknowledge the support of the Comparative Constitutions Project, in particular the University of Texas, Austin.

[2] I label this constitutional presidential strength (CPS) to avoid any possibility of confusion with the different conceptualizations of presidential power presented and with other measurement tools.

[3] This list of arguments explaining why presidential de-facto power varies under certain conditions is infinite. Every country holds its unique set of surprises. One important explanation that proves solid is the role of the president as de facto party leader (Bucur 2013). Additionally party cohesion in combination with cohabitation (Amorim Neto and Strøm 2006), (Elgie 1999), and the proximity of presidential and parliamentary elections (Hicken and Stoll 2013) explain the situation in a variety of countries.

Anna Fruhstorfer is a postdoctoral researcher at Humboldt University Berlin at the Department of Social Sciences. Her main area of research is Comparative Politics, with a regional emphasis in Eastern Europe. Her research concentrates on presidents, parliaments and the relation of law and politics, in particular constitutional politics. She is also affiliated with the University of Texas at Austin via the Comparative Constitutions Project.

Ryan E. Carlin and Shane P. Singh – Executive Power and Economic Accountability

This is a guest post by Ryan E. Carlin and Shane P. Singh from Georgia State University and the University of Georgia respectively

Accountability’s centrality to democratic theory makes it “one of the holy grails of empirical political science research” (Samuels and Hellwig 2010, 393). It refers to the public’s ability to “discern whether governments are acting in their interest and sanction them appropriately” (Manin, Przeworski, and Stokes 1999, 40) through elections, civil society, the media, protests, and public opinion. Are some executives held more accountable than others by design? Can they shape their own accountability by the decisions they make?

In a forthcoming article in the Journal of Politics, we posit that both presidents’ legislative powers and how frequently they resort to legislation by decree shape the extent to which the public holds them accountable for economic outcomes. Two forms of responsibility attributions connect presidential powers to accountability. “Role responsibility” concerns the obligations citizens expects presidents to fulfill and the consequences for failing to do so. Presidents endowed with vast lawmaking powers accrue great role responsibility for outcomes on their watch. “Causal responsibility” is doled out when presidents’ intentional actions are seen as proximate causes of economic outcomes. In proportion to the decree powers presidents use to legislate unilaterally, they will incur causal responsibility for the outcomes.

To test our expectations, we gathered data from the Latinobarometer on presidential approval, sociotropic economic retrospections, and several common controls from 94,861 respondents in 120 surveys from 18 Latin American presidential systems over the years 2002-2009. We tested whether the degree to which socioeconomic retrospections were related to presidential approval was a function of two measures of presidential power: an index of the legislative powers afforded to the president by the constitution and the total number of decrees and decree laws the president issued in the previous year, as measured with the Global Legal Information Database.

As predicted, economic evaluations influence the approval of powerful presidents more than their weaker counterparts. That is, economic accountability is calibrated to account for presidents’ legislative powers and prerogatives—their role responsibility for outcomes. Accountability also keeps pace with decree usage, a finding consistent with a mechanism of causal responsibility attribution. We can see this graphically through the marginal effects of economic evaluations on the probability of approving of the president according to legislative powers and the use of decrees as captured in Figure 1.

Screen Shot 2015-07-22 at 15.35.23

To put some flesh on these results, imagine an individual whose president has Legislative Powers that register in the 10th percentile of our sample—Honduras since 1982. Results illustrated in the left-hand panel of the figure suggest that a unit increase in retrospective economic evaluations corresponds with, on average, roughly a 12.5 percentage-point increase in the likelihood that she will approve of the president. If this same individual were transplanted to a country where the president’s Legislative Powers ranked in the 90th percentile—Brazil since 2001—a unit increase in economic retrospections should boost the likelihood that she will approve of the president by around 16.4 percentage points.

Turning to the link with decrees, illustrated in the right-hand panel of the figure, for a citizen whose president falls in the 10th percentile for decree issuance, a unit increase in economic evaluations corresponds with, on average, about a 13.5 percentage-point increase in the likelihood that she will approve of the president. If that citizen’s president falls in the 90th percentile for issuing decrees, the same increase precipitates around a 15.9 percentage-point rise in the chances of approving of the president.

So presidents who enjoy strong legislative power and who use it extensively invite greater economic accountability. Not only do these measures of role responsibility and causal responsibility influence economic accountability independently but, as Figure 2 shows, their effects are even greater when analyzed in interaction.

Screen Shot 2015-07-22 at 15.36.07

The public, thus, holds presidents to account for economic outcomes to a greater degree where their governing role responsibilities are vast and their causal responsibility is made more obvious by frequent use of such power. In other words, accountability for economic outcomes is strongest where presidents are strong legislators both de jure and de facto.

In presidential systems, executive power can be a blessing or a curse. When the public perceives a humming economy, strong presidents who use their legislative powers habitually can expect their popularity to grow multiplicatively. But when they feel the economy slumping, vast presidential powers compound the problem, sending support in a downward spiral. This may seem unfair, given the difficulty even strong presidents face in advancing their agendas (Saiegh 2011) and delivering outcomes citizens prefer (Mainwaring and Shugart 1997)—and the fact that congress may ask the president to act unilaterally in order to get things done (Carey and Shugart 1998). Nevertheless, we conclude that the more lawmaking power constitutions grant presidents, and the more they use these powers, the more responsibility they assume for economic policy outcomes in the public’s eye.

Our results weigh in on second-generation debates about presidentialism and democracy. In a thoughtful essay on why the “difficult combination” of presidentialism and multipartyism (Mainwaring 1993) has outstripped scholars’ predictions for durability and governability, Chaisty, Cheeseman, and Power (2014) credit presidents’ mastery of various power tools to build and manage multiparty coalitions. Yet the authors are quick to warn:

“The very same presidential tools that enhance governability may also undermine accountability – a tradeoff that has been largely overlooked even as analysts celebrate the survival of multiparty presidential democracy.” (86).

Our findings, however, suggest no tradeoff between the powers that facilitate governability and the vertical accountability mechanism of public opinion between elections. In fact, citizens appear able to calibrate responsibility attributions according to presidents’ possession and use (or even abuse) of lawmaking power. Thus, our preliminary diagnosis is that presidential powers pose no barrier to accountability at the level of mass politics.

But viewing these findings in light of our related research, blogged at Presidential Power earlier this year, renders a less sanguine conclusion. Democratic attitudes, the basis of regime legitimacy, are most positive under a middling—neither too high nor too low—degree of presidential powers (Singh and Carlin 2015). In other words, crafting an institutional framework to optimize both of these core democratic ideals under presidentialism may, therefore, prove rather challenging.


Carey, John M., and Shugart, Matthew Soberg, eds. 1998. Executive Decree Authority. New York: Cambridge University Press.

Chaisty, Paul, Cheeseman, Nic, and Power, Timothy. 2014. Rethinking the ‘Presidentialism Debate’: Conceptualizing Coalitional Politics in Cross-Regional Perspective. Democratization 21 (1):72-94.

Mainwaring, Scott. 1993. Presidentialism, Multipartism, and Democracy: The Difficult Combination. Comparative Political Studies 26 (2):198-228.

Mainwaring, Scott, and Shugart, Matthew Soberg, eds. 1997. Presidentialism and Democracy in Latin America. New York: Cambridge University Press.

Manin, Bernard, Przeworski, Adam, and Stokes, Susan Carol. 1999. Elections and Representation. In Democracy, Accountability, and Representation, eds. Przeworski, Adam, Stokes, Susan Carol and Manin, Bernard. Cambridge, U.K.; New York: Cambridge University Press, 29-54.

Saiegh, Sebastian M. 2011. Ruling by Statute: How Uncertainty and Vote-Buying Shape Lawmaking. New York: Cambridge University Press.

Samuels, David, and Hellwig, Timothy. 2010. Electoral Accountability and the Clarity of Responsibility: A Conceptual and Empirical Reassessment. Journal of Elections, Public Opinion, and Parties 20 (4):393-414.

Singh, Shane P., and Carlin, Ryan E. 2015. Happy Medium, Happy Citizens: Presidential Power and Democratic Regime Support. Political Research Quarterly 68 (1):3-17.

Ryan Carlinpolitical scienceRyan E. Carlin is an Associate Professor of Political Science and Director of the Center for Human Rights and Democracy at Georgia State University. He received his Ph.D. in 2008 from the University of North Carolina at Chapel Hill. His research interests are comparative political behavior and public opinion, with a regional emphasis on Latin America. His work has been funded by the National Science Foundation, USAID, the Mellon and Ford Foundations, and the Latin American Studies Association. He is co-editor of The Latin American Voter (University of Michigan Press, 2015) and his work has appeared in numerous academic journals and volumes. His website is found at

SinghShane P. Singh is an Associate Professor in the School of Public and International Affairs at the University of Georgia. He received his Ph.D. in 2009 from Michigan State University. His research focuses on the institutional and contextual foundations of political behavior and attitudes. His work has been funded by the Social Sciences and Humanities Research Council of Canada, and his research has appeared in many academic journals and edited volumes. His website is found at

Shane P. Singh and Ryan E. Carlin – Happy Medium, Happy Citizens

This is a guest post by Shane P. Singh and Ryan E. Carlin from the University of Georgia and Georgia State University respectively

As democracy’s third wave gathered steam, constitutional engineers argued fervently that “getting the institutions right” could help inoculate fledgling democracies from breakdown. Linz famously warned that reverting to presidential models of democracy in Latin America would perpetuate the cycle of democratic instability in the region. He cautioned of a “dual mandate” problem: since both executive and legislature could claim to speak for the voters, they were destined to clash. Fixed terms and an aura of sovereignty would only make the matter worse and lead to praetorian politics. Such arguments notwithstanding, newly (re-)emerged democracies throughout Latin America maintained presidentialism but reformed its institutional framework by improving its flexibility and stability in various ways. This broad institutional experimentation begs the question: are certain models of presidentialism more likely to produce consolidated democracy?

Focusing on attitudinal consolidation, in an article recently published in Political Research Quarterly, we reason that executive powers are associated with opinions about democracy. On the one hand, voters expect a degree of control over presidents, and a president that supersedes his or her mandate could spur disillusionment with the principles and performance of democratic regimes. On the other hand, an enfeebled president who lacks power could end up in perpetual conflict with the legislature, leading voters to become unhappy with policy immobilism and, ultimately, a deficit in representation. Thus, from a citizen’s perspective, the sweet spot should lie in the middle. Presidents who are either very weak or very strong will foster discouragement with democracy. But presidents who have balanced powers, and thus must engage in some give and take with the legislature, will cultivate public support for democratic regimes.

To test our expectations, we gather surveys spanning 1995-2005 from 18 Latin American countries, all with presidential models of democracy. Our dataset includes over 100,000 individuals, each of which was asked a.) whether they believe democracy is the best form of government, and b.) whether they are satisfied with the way democracy works in their country. We also gathered three measures of presidential power: an index of the legislative powers afforded to the president by the constitution, a measure of the degree to which the president is constrained by the legislature, and a measure of the president’s realized legislative success, or “box score,” calculated as the percentage of bills initiated by the president that were approved by the legislature.

As expected, we find that each measure is associated with attitudes toward democracy in a curvilinear fashion. That is, as presidential power and prerogative initially rise, individuals become more positive toward democracy, but at a certain point, further increases in presidential strength harm attitudes toward democracy. Our findings are illustrated in the figure, which makes it clear that each index of presidential power has an inverse-U shaped relationship with democratic support.

Screen Shot 2015-03-10 at 17.40.18


















As an example, the bottom-right panel depicts the relationship between legislative success and the probability that a citizen is “very”or “fairly” satisfied with democracy. Consider a president whose passage rate is just 33 percent, like Colombian president Ernesto Samper in 1996. Starting from this low success rate, a standard deviation increase is associated with about a 7-percentage point increase in the likelihood that a citizen is satisfied with democracy. Citizens value presidents who are not completely ineffectual. Compare this to a president with twice the passage rate (66 percent), such as Honduran president Carlos Roberto Reina in 1996. Starting from this middling success rate, a standard deviation increase is associated with about a 10-percentage point decrease in the likelihood of a respondent expressing satisfaction with democracy. Citizens also value presidents who are not completely dominant.

Interestingly, this curvilinear relationship is not conditional on economic performance or trust in the president. In other words, citizens do not desire very strong powers for presidents who preside over booming economies, and even citizens who fully trust the president become less positive toward democracy when he or she wields vast authority.

Our findings suggest that, when it comes to fomenting mass regime support, not all forms of presidentialism are created equal. Presidential democracies are most legitimate in the eyes of the citizens where presidents enjoy a happy medium of lawmaking power. And although presidents with vast formal powers and/or broad legislative majorities can avoid gridlock and fluidly implement their agendas, such behavior can fuel discontent with democracy and undermine its legitimacy. The same is true if the president’s every attempt to effect legislation is thwarted. Of course, the idea that a balance of power is good for democracy is nothing new. As James Madison wrote in Federalist No. 51, “Ambition must be made to counteract ambition.”

Note: a slightly different version of this post appeared at The Quantitative Peace in June of 2014.



Shane P. Singh is an Assistant Professor in the School of Public and International Affairs at the University of Georgia. He received his Ph.D. in 2009 from Michigan State University. His research focuses on the institutional and contextual foundations of political behavior and attitudes. His work has been funded by the Social Sciences and Humanities Research Council of Canada, and his research has appeared in many academic journals and edited volumes. His website is found at


Ryan Carlin political scienceRyan E. Carlin is an Associate Professor of Political Science at Georgia State University. He received his Ph.D. in 2008 from the University of North Carolina at Chapel Hill. He is a Faculty Affiliate of GSU’s Center Latin American and Latino Studies and the Center for Human Rights and Democracy, and an Affiliated Researcher of the Latin American Public Opinion Project at Vanderbilt University. His research interests are comparative political behavior and public opinion, with a regional emphasis on Latin America. His work has been funded by the National Science Foundation, USAID, the Mellon and Ford Foundations, and the Latin American Studies Association. He is co-editor of The Latin American Voter (University of Michigan Press, forthcoming) and his work has appeared in many academic journals. His website is found at

A new dataset of presidential power scores

This is a post by David Doyle and Robert Elgie

Since the beginning of the 1990s there has been ongoing debate about the relative effect of different regime types, specifically presidentialism, semi-presidentialism and parliamentarism. From the very beginning of this debate, there has been an acknowledgement of variation in presidential power not just across these three regime types, but also within each regime type. This within-type variation causes a problem for cross-national studies. If the within-type variation is large, then estimating the effect of regime types themselves may lead to spurious results. For this reason, Crisp, Desposato and Kantha (p. 447) have stated that there is a “need for more explicit theoretical depictions of the institutional variation in the class of regimes referred to as presidential as well as the need for a systematic empirical exploration of the impact of that diversity on regime performance” (1).

In this context, some observers have preferred to estimate the effect of presidential power on outcomes rather than the effect of regime type more broadly. For example, Hicken and Stoll (p. 1114) note that the power of the Colombian president has varied over time as a result of constitutional amendments, even though Colombia has maintained a presidential regime throughout. As a result, they prefer to measure variation in presidential power over time and estimate the effect of such variation. They note: “our overall index of presidential powers reveals variation within each type of regime that the simple trichotomy [of presidentialism, parliamentarism, and semi-presidentialism] obscures. It is this greater level of precision that leads us to prefer the index”.

However, while there is now a well-established literature demonstrating that variation in presidential power has consequences across a range of political and economic outcomes, there are many different measures of presidential power. In fact, we have identified 19 separate and original measures of presidential power, plus a further 16 studies that used one of these measures with different countries, time periods, and/or scores from the original study.

This range of measures raises a number of issues. Individual measures are sometimes poorly correlated with each other, meaning that findings are sensitive to the particular measure of presidential power that is used. There is also a considerable loss of information across the set of measures as a whole as countries are included in some measures but not others and then for only certain time periods. More generally, as Jessica Fortin has recently shown, there are no theoretical priors to tell us which indicators of presidential power we should choose or how the scores for the individual indicators should be aggregated (3).

We agree with Fortin’s analysis. However, we draw a different conclusion from her. She concludes very skeptically, effectively questioning whether any measure of presidential power is likely to be valid. By contrast, we assume that most social science concepts, such as voter turnout, social equality, corruption, and so on, suffer from equivalent problems of construct validity. Therefore, we should not give up on the effort to generate a dataset of presidential power scores. Instead, we should focus on the reliability of the data that underpins the concept we are trying to capture.

We wish to generate a time-series cross-sectional dataset of presidential power scores with country years as the units of observation. To do so, we choose not to construct a new measure of presidential power from scratch. Instead, we draw upon the comparative and local knowledge already embedded in the existing measures of presidential power that we identified. To maximize the reliability of the scores we derive them solely from measures that are based on institutional indicators of presidential power and on the basis of a method that accounts for potential idiosyncrasies of country scores in the existing measures. In addition, we report the standard errors and the confidence intervals for all the country years in our measures, providing information with which scholars can make an informed choice about whether or not a particular country should be included in an estimation and which of our measures might best be used in comparative studies.

The paper outlining our full methodology will soon appear in British Journal of Political Science and is available in advance here. In the meantime, we are making available the full set of presidential power scores, including standard errors and confidence intervals for each country time period, in a separate page at the header of this blog. We also provide more detail about the scores.

Overall, we encourage people to keep developing new measures of presidential power and to update existing measures for as many countries and as long a time period as possible. One of the advantages of our approach is that new country scores can be easily incorporated into the method we have used, creating the potential for country coverage to be further extended, for existing country scores to be updated, and for cross-national measures to become even more reliable.

(1) Crisp, Brian F., Scott W. Desposato, and Kristin Kanthak. 2011. Legislative Pivots, Presidential Powers, and Policy Stability.” Journal of Law, Economics and Organization 27 (2): 426-452.

(2) Hicken, Allen, and Heather Stoll. 2008. “Electoral Rules and the Size of the Prize: How Political Institutions Shape Presidential Party Systems.” Journal of Politics 70 (4): 1109-1127.

(3) Fortin, Jessica. 2013. “Measuring presidential powers: Some pitfalls of aggregate measurement.” International Political Science Review 34 (1): 91-112.