Does distinguishing amongst regimes based on whether they are presidential, semi-presidential or parliamentary tell us much about political outcomes?

First published: ; Last modified:

This content was originally written in .
Footnotes were present in the original; sidenotes are used for comments.

Essay

No, it does not tell us much about political outcomes. Using the presidential/semi-presidential/parliamentary trichotomy

The threeway distinction seems to originate from Alvarez, M., Cheibub, J. A., Limongi, F., & Przewroski, A. (1996). “Classifying political regimes.” Studies in Comparative International Development, 31(2), 3–36
to distinguish between regimes does give us some information about likely outcomes, but other, more nuanced, classifications of legislative-executive relations provide far greater explanatory power. In particular, whilst there are empirical differences in government formation, cabinet composition, and democratic survival between presidential and parliamentary regimes, the magnitude of within-type variation is so great that regime type alone tells us only a little about political outcomes. Additional factors – such as electoral system, number of veto players, and concentration of political power – play a decisive role in determining political outcomes. Importantly, there are only plausible accounts of why institutional arrangements would cause political outcomes to differ as they are observed to once such factors are brought into the picture. In this essay, I will first outline the standard definitions of the three regime types and introduce objections to this taxonomy, before exploring empirical findings on what success the approach has in explaining the variations in political outcomes that we observe.

One attraction of the presidential/semi-presidential/parliamentary taxonomy is its simplicity. Presidential regimes are those in which in which the government does not require the support of a legislative majority to exist; if it does, the regime is either semi-presidential or parliamentary – the former only if there is also a head of state elected directly by voters for a fixed term (Cheibub et al. 2010). Because the boundaries between these categories are binary and crisp

Hmm, though this paper suggests that there is debate around definitions of categories too. The conclusion here seems to be that there were disagreements over where to put countries, but now there is more conceptual clarity, and so there are no longer (e.g. they’re able to say whether classifications are “wrong” or not).
, modern criticisms of the taxonomy focus on challenging the utility and coherence of its classifications, rather than claiming that specific countries are incorrectly placed (Elgie 2020). Such objections are the most forceful – and have existed the longest – in the case of the semi-presidential category, whose members significantly differ from each other even in terms of de jure constitutional arrangements. Many scholars argue that the group should be subdivided into premier-presidential systems, where the government is accountable only to the legislature, and president-parliamentary ones, where the government is also accountable to the president (Elgie 2020). Others, such as Lijphart (2012, 108) suggest that the semi-presidential group can simply be split up into and subsumed by
Drawing on Clark et al. (2012, 504): Premier-presidential regimes are very similar to parliamentary ones. President-parliamentary regimes are more distinctive, and their characteristics depend on the balance of power between at a particular time. However, the fact that a president-parliamentary country is either presidential-like or parliamentary-like at a time and alternates between the two suggests that this designation does not add much on its own – we simply need to know what the electoral situation is and then treat it as one of the pure types in practice. On the other hand, you could argue that this capacity for oscillation is a unique and important feature of president-parliamentary systems, justifying making the distinction.
 the other two. Either way, given that there is around legislative-executive relations in the semi-presidential group, it seems unjustified to suggest that merely knowing that a country falls into that category could tell us much about political outcomes there.

Therefore, let us focus on the two remaining categories of parliamentary and presidential regimes, and investigate what distinguishing between countries along these lines tells us about political outcomes. There are three main areas where differences in average outcomes can be found: government formation, cabinet composition, and democratic longevity. I now examine each in turn, identifying the key empirical results, and, crucially, analysing whether the presidential/parliamentary distinction alone can convincingly explain the differences found.

Governments in presidential regimes are more likely to be in a legislative minority and less likely to rule as a coalition than those in parliamentary regimes. Looking at almost all democracies from 1946-99,

demonstrate that in situations where no party had an absolute majority, a minority government resulted 68% of the time in presidential regimes, but only 38% in parliamentary ones. The same study also showed that such situations lead to coalition government around 30% less frequently in presidential systems. The structure of legislative-executive relations gives us a good reason to expect these differences: a parliamentary government must as a minimum command implicit majority support in the legislature to avoid a vote of no confidence, whereas this condition is not required in presidential regimes. A similar logic explains why coalitions are less likely: presidents have no need to offer ministerial posts to opposing parties as a means of buying their support in the legislature, because the government does not have to be backed by a legislative majority. However, the extent to which this is true depends on regime-specific factors such as how much the president is motivated by purely office-seeking aims, and their ability to effect policies via decree – the less true these both are, the more likely it is that majority and coalition governments result in presidential regimes (Clark et al. 2012, 511).1 So, knowing whether a regime is parliamentary or presidential tells us a little about likely government formation, with the caveat that the actual outcomes depend on the aims of the individual actors and the scope of legislative powers
Doing a veto players analysis here might tell us more, though Colomer is against that (2006, 225) for reasons I don’t understand (his argument seems to be that the Tsebelis’s approach is too focussed on how institutions work in practice so isn’t useful for constitutional design?). More saliently, it doesn’t distinguish between autocratic one-party and democratic one-party governments. Looking at concentration of power instead, as Colomer does, seems pretty closely related, but perhaps better.
 held by the executive.

The ability of presidential governments to exist without majority legislative support also contributes to differences in cabinet composition compared to parliamentary ones. Presidential cabinets tend to have more non-partisan ministers and a lower proportionality in the allocation of ministerial posts (Samuels 2007, Table 29.1 & 29.2). As above, this is because a minority president does not need to buy legislative backing from other parties in the same way that a minority prime minister would. However, there is an enormous amount of within-group variation here,2 again explainable by the particular president’s priorities and their capacity to directly legislate by decree (Neto 2006). Therefore, whilst one can say that a randomly-chosen presidential cabinet will typically have more non-partisanship and less proportionality these features than a randomly-chosen parliamentary one, the level of within-group variation means that knowing a particular regime is presidential, on its own, tells you very little about what to expect

On the basis of their Tobit regressions (I couldn’t figure out what these are), they say there is a statistically significant difference between presidential and parliamentary systems. But that doesn’t mean that you get much information about a regime from being told that it is one or the other!
of its cabinet composition.

Finally, presidential systems appear to be associated with shorter-lived democracies

There’s lots more discussion and statistics on this in Clark et al. (2012, 810-823), paraphrasing Mainwaring’s argument: “legislative fragmentation increases the likelihood of cabinet instability in parliamentary systems, whereas it increases the likelihood of democratic instability in presidential systems”. See also Samuels (2007, 706-7), who says that Przeworski et al. (2000) “reconfirm that presidentialism is more fragile than parliamentarism, but question the connection between party system attributes and presidential regime fragility.”
. Amongst democratic non-OECD countries from 1973-89, the democratic survival rate for parliamentary regimes was 60%, triple that of presidential ones. Similar statistics exist for robustness against coups and level of democracy, even when adjusting for (Stepan & Skach 1993, 10). Causal conclusions must not be drawn too quickly, though. As Colomer (2006, 230) notes,3 young democracies – those most likely to have lapsed into undemocratic government – are almost exclusively presidential in South America and pluralist parliamentary in Eastern Europe. This tight association between region and regime type strongly suggests that the kind of institutional arrangements decided on in a new democracy is determined endogenously by factors such as pre-existing political culture (Elgie 2020). If so, we should be wary of attributing differences in political outcomes to regime type when both are actually causally downstream from common variables not explicitly in our model. Further cause for caution is provided by the fact that parliamentary majoritarian regimes are generally more unstable than the typical presidential regime, while parliamentary pluralist regimes are the most peaceful of all (Colomer 2006, 230). Just like with government formation and cabinet composition, the distinction between presidential, semi-presidential, and parliamentary regimes is about democratic stability. Indeed, as Cheibub et al. (2014) show, a country’s regime classification is less useful than simply knowing when and where its constitution was written in predicting institutional features
See Table 2 in that paper for the details.
 such as whether the executive has powers to make decrees and vetoes. It would thus be strange to suppose that the taxonomy can give us helpful information about political outcomes in a country, given its inability to predict even basic facts about executive-legislative relations which are upstream of those terminal outcomes.

So, to conclude – no, distinguishing amongst regimes based on whether they are presidential, semi-presidential or parliamentary does not tell us much about political outcomes. Although differences in outcomes between in the average presidential regime and the average parliamentary regime are borne out by the data, the extent of within-type variation means that this does not imply that we can draw useful inferences about the political outcomes in a particular country on the basis of its regime type. The substantial differences between premier-presidential and president-parliamentary regimes mean that the semi-presidential classification is particularly unhelpful. Precise predictions about political outcomes can only be made once additional features, such as position on the majoritarian/consensus axis, are incorporated into the model. Therefore, as I have argued

It sounds like Przeworski (e.g. 2003) criticises the distinctions a lot. Could be worth looking at him.
, the presidential/semi-presidential/parliamentary taxonomy is not informative about political outcomes on its own, but can become valuable when used as part of a multidimensional classification.

References

Cheibub, J.A., Gandhi, J. & Vreeland, J.R. (2010). Democracy and dictatorship revisited. Public Choice 143, 67–101. https://doi.org/10.1007/s11127-009-9491-2

Cheibub, J.A., Elkins, Z. & Ginsburg, T. (2014). Beyond Presidentialism and Parliamentarism. British Journal of Political Science, 44(3), 515-544. https://doi.org/10.1017/S000712341300032X

Clark, W.R., Golder, M. & Golder, S.N. (2012), “Chapter 12: Parliamentary, Presidential, and Semi-Presidential Democracies: Making and Breaking Governments”, in Principles of Comparative Politics (2nd ed.). London: CQ Press.

Colomer, J.M. (2006). “Comparative Constitutions”, in The Oxford Handbook of Political Institutions (ed. R. A. W. Rhodes, Sarah A. Binder, and Bert A. Rockman). Oxford: Oxford University Press.

Elgie, R. (2020). An Intellectual History of the Concepts of Premier-Presidentialism and President-Parliamentarism. Political Studies Review, 18(1), 12–29. https://doi.org/10.1177/1478929919864770

Lijphart, A. (2012). “Chapter 7: Executive–Legislative Relations: Patterns of Dominance and Balance of Power” in Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries (2^nd^ ed.). New Haven: Yale University Press.

Neto, O. A. (2006). The Presidential Calculus: Executive Policy Making and Cabinet Formation in the Americas. Comparative Political Studies, 39(4), 415-440. https://doi.org./10.1177/0010414005282381

Rom, M.C., Hidaka, M. & Walker, R.B. (2022). “Semi-Presidential Regimes” in Introduction to Political Science. Houston, TX: OpenStax.

Samuels, David (2007), “Separation of Powers”, in The Oxford Handbook of Political Institutions (ed. R. A. W. Rhodes, Sarah A. Binder, and Bert A. Rockman). Oxford: Oxford University Press.

Stepan, A., & Skach, C. (1993). Constitutional Frameworks and Democratic Consolidation: Parliamentarianism versus Presidentialism. World Politics, 46(1), 1–22. https://doi.org/10.2307/2950664

Tutor’s comments

  1. “Governments in presidential democracies look more like those in parliamentary democracies if the president is weak.” ↩︎

  2. For example, the average proportion of non-partisan ministers in parliamentary republics is 3%, but with a standard deviation of 10%; the corresponding figures for presidential republics are 29% and 29%. ↩︎

  3. “Almost no new democracy established in the world during the broad “third wave” of democratization starting in 1974 has adopted the British-style constitutional model of parliamentary regime with a two-party system and majoritarian electoral rules” ↩︎