Tutor’s comments
- Very well written. Strong structure.
- There can be good reasons, methodologically, to exclude outliers from a regression. This criticism in itself is quite weak. However, it is fair to be more sceptical about the results because of it. You also want to consider if those exclusions produce a conservative test or not (e.g. does including the UK with Northern Irish terrorism reinforce the relationship instead of weakening it?).
- Good critical distinction of the two dimensions.
- Stronger far-right under PR: the dissatisfaction causal mechanism is debatable, it would need to be supported by better evidence, vis-a-vis alternatives.
- Very good and comprehensive essay. Balanced conclusions, well supported.
Essay
There is not sufficient empirical evidence or theoretical explanation to support the claim that consensus democracies are “gentler” than majoritarian democracies in terms of their level of civil violence and extent of political polarisation. However, we can legitimately say that consensus democracies tend to be less efficient in policymaking, where “efficiency” is understood to mean the speed and ease with which legislation is passed. It must be noted that the quality of policies thus produced is a separate (and partially normative) matter in which the evidence indicates no substantial difference between consensus and majoritarian systems. In this essay, I will first briefly define what features constitute a consensus and majoritarian democracy, before exploring empirical findings on whether differences in gentleness and policymaking exist between the two types of systems (and if so, whether these consistently lie in the same direction).
The consensus/majoritarian model provides a more nuanced way to classify democracies than merely sorting them based on whether they are presidential, semi-presidential, or parliamentary. How decision-shaping power is distributed within a system determines its classification: simply speaking, countries where power is concentrated with the majority are majoritarian, whereas those where power is spread out as much as possible are consensus (Lijphart 2012, 2). Further detail can be introduced by decomposing this univariate spectrum into two sub-components which separately capture the effects on distribution of power from constitutional design and from actual political outcomes (the federal-unity and executive-parties dimensions, respectively). As Bogaards (2017) notes, the classification of individual countries as consensus or majoritarian is generally uncontroversial, with critics of Lijphart focussing on the relevance and generalisability of his approach when explaining or predicting political and policy outcomes. Therefore, I will save discussion of methodological objections to the taxonomy until after we have explored its practical usefulness.
Lijphart’s central claim is that consensus democracies are “kinder” and “gentler” than their majoritarian counterparts, as a result of them involving minority groups within decision-marking as a matter of course (Lijphart 2012, 274). One way this can be seen, according to him, is in lower levels of violence and instability. Lijphart references the result in Powell (1982) that democracies which emphasise consensus in decision-making are better at controlling levels of violence, as well as presenting his own analysis finding that moving in the consensus direction along the executive-parties dimension has a positive effect on political stability and negative effect on deaths from domestic terrorism (2020, Table 15.2).1 At a first pass, these data seem to clearly vindicate Lijphart’s position. However, the picture is more complex than the one he paints for two reasons.
First, his own dataset has a number of dubious coding and selection decisions. For instance, Lijphart appears to systematically over-record the number of terrorist incidents in majoritarian countries and under-record the number in consensus ones. Once these There is a response, The Logic of Domestic Terrorism Revisited. , there is no longer any statistically significant difference between the two (Borgaads 2020). This sort of (perhaps unintentional) bias is also found in Lijphart’s removal from the dataset of countries he calls “extreme outliers”. This group included the UK, justified merely by brief appeal to the “special problem of Northern Ireland”. It seems reasonable to believe that had these countries fit with the trend that Lijphart expected to find, they would not have been excluded from the regression, and so our confidence in the robustness of his findings should be reduced. Lijphart is not the only academic to have found associations between consensus democracy and reduced political violence, but similar statistical issues are also found in others’ work (Clark et al. 2012, 800).
Second, apart from data quality concerns regarding positive results, the picture is further complicated by the existence of empirical studies which have found that moving in the consensus direction is associated with an increase in violence. Moreover, just as Lijphart theorises that their representation of is causally responsible for consensus democracies’ being gentler, the authors of these opposing studies also propose plausible mechanisms by which features of consensus democracy might increase violence. Take federalism: perhaps it increases communal violence Brancati (2006) “argues that political decentralization reduces ethnic conflict when regional parties are weak but that it can increase ethnic conflict when regional parties are strong”. Clark et al. (2012, 805): “it appears that incongruent and asymmetric federalism can help reduce ethnic conflict and stabilize democracy in countries with geographically concentrated ethnic groups.” . A key plank in Lijphart’s argument was that by guaranteeing representation for minorities through distributed decision-making power, consociationalism helps prevent civil unrest. Yet several studies have shown that federalism in fact increases the amount of protest that minorities engage in (Clark et al. 2012, 802). We are able to explain these observed outcomes more convincingly when we perform comparative analysis in terms of whether the veto players are present in the same institution or separate ones (“collective” as opposed to “competitive”). Systems with collective veto points are better able to cope in ethnically divided societies, because they tend to produce more compromise and less instability-inducing deadlock (Bogaards 2017). Although this veto player approach might sound similar to the consensus/majoritarian distinction, it cuts across Lijphart’s characteristics: federalism is associated with competitive veto points whereas proportional representation (PR) is associated with collective ones. This suggests that the executive-parties and unitary-federal dimensions of Lijphart’s classification can come apart, and therefore that we should not necessarily expect there to be consistent patterns within consensus or majoritarian systems.
Nevertheless, there is evidence that the proportional representation found in consensus democracies can, in some senses, lead to gentler politics. Because plurality electoral methods like first-past-the-post (FPTP) distort the relationship between number of votes and seats in the legislature, PR systems produce governments which tend to be closer to the median voter’s opinion (Boorman 2010). In addition, the larger district size in PR systems means that there are fewer knife-edge cases where political parties are strongly incentivised to use underhand tactics such as funnelling money into marginal constituencies to gain victory in a winner-takes-all contest (Clark et al. 2012, 784). So, consensus systems lead to more representative legislatures, and less destructively adversarial pre-election behaviour from parties, which we could reasonably call a gentler result.
This brings us on neatly to questions of policy. Consensus democracies are less decisive at policymaking, because of the larger number of veto players in their legislatures. As Tsebelis (1995) notes, the coalition style of government usually found in proportional systems means that many different agents (i.e. parties) must agree to proposals before changes to the status quo can gain legislative approval. This is not true in majoritarian systems, where there are fewer institutional and partisan veto players who must consent to reforms. As a result, consensus democracies have much stickier, or resolute, policies than majoritarian ones. Research by Kreppel (1997) into Italian government since WW2 bears this out: she found an inverse correlation between the number of parties in coalition at a time and the number of laws passed. The intermediary factor of greater policy stability provides an explanation for why there is greater support for far right-wing parties in consensus democracies (Andeweg 2001): the lack of reform fuels dissatisfaction with the existing political parties and As a result, we would expect to see greater government (and, in the case of presidential systems where there are no constitutional means to remove a government or legislature mired in deadlock) regime instability. . If one believes that this outcome is less gentle or otherwise normatively undesirable, then this is one disadvantage of consensus democracies and their proportional electoral systems.
If “efficient” policymaking is merely about speed and decisiveness, it is undoubtable that consensus democracies do worse. However, Lijphart’s argument stresses that the quality of legislation in those countries is superior, pointing in particular to the positive association of consensus democracy with low inflation and better environmental outcomes (2012, Tables 15.1 and 16.2). As with their gentleness, though, the reality is more complex. In the case of The picture is a bit different for fiscal policy: PR systems appear to be associated with more public goods, larger and more redistributive social welfare programs, and a larger overall size of government than majoritarian systems (Clark et al. 2012, 778). This gets attributed to the election of leftwing governments, and also party fragmentation & coalitions leading to less budgetary discipline (Ibid, 786); Persson, Roland, and Tabellini (2007). Note that Tsebelis disagrees and says this is just for historical reasons. Also, it’s a normative question as to whether this is better policy. , the decisive components in producing the positive association are central bank independence and corporatist interest groups (Bogaards 2017 and Armingeon 2002). However, the former of these components is now found in most advanced economies rather than just consensus ones, and the latter is less a political feature than an economic one (Roller 2005, 229). This suggests that this association is incidental to Lijphart’s sample of 36 countries, and not due to genuine differences between consensus and majoritarian political systems in general.
As for the environment, whilst there appear to be differences in the sorts of green policies put into law by consensus and majoritarian democracies, neither is unambiguously more successful than the other. Poloni-Staudinger (2008) demonstrates that consensus democracies typically promote “mundane” initiatives like recycling, whereas majoritarian democracies have a heavier focus on conservation, but they are indistinguishable when it comes to the arguably more important policy areas of environmental taxation and development of nuclear energy. Therefore, in the case of the environment, drawing a link between consensus democracy and more effective policy also looks tenuous.
So, to conclude, the evidence currently available does not justify the claim that consensus democracies are gentler, or that they produce higher-quality policy, compared to majoritarian ones. There are plausible theoretical explanations which explain why they could be, but these are not supported by sufficiently strong empirical findings to That said, Roller (2005, 264) argues: “If one only examines the statistically significant coefficients, then constitutional negotiation democracies are superior to constitutional majoritarian democracies inasmuch as they show lower inflation rates and a higher national income, but inferior with regard to having more violent crimes, worse environmental policies, and an above‐average consumption of natural resources.” , particularly given that equally plausible-sounding theories exist arguing for the opposite trend. If we construe policymaking efficiency in a narrow sense of the volume of legislation passed, we can conclude that consensus democracies are less efficient than majoritarian ones, but that is not a particularly meaningful conclusion to draw, given that the content of those laws is what matters for citizens’ wellbeing. Whilst it is certainly a matter of historical fact that many ethnically divided European countries which adopted consociational models have now integrated their minorities effectively, that is not grounds to claim either (a) that the consensus model caused this, or the much stronger assertion (b) that currently divided societies would similarly A paper on the effects of negotiation democracy concludes that, whilst it is not kinder and gentler, “consociational democracy is a form of government for segmented societies [with] certain advantages in the integration of large minorities” . The evidence of a positive association between consensus democracy and desirable outcomes is far too mixed to conclude that significant differences If I’d had more space and time, I might’ve argued that an approach which focuses on the number and types of veto players present in different systems may do better at explaining the variations in outcomes that we observe. do indeed exist, particularly in the absence of a control group and given the presence of confounding cultural and economic differences between the countries under investigation.
Bibliography
Armingeon, K. (2002). The effects of negotiation democracy: A comparative analysis. European Journal of Political Research, 41(1), 81–105. https://doi.org/10.1111/1475-6765.00004
Andeweg, R. B. (2001). The Cons of Consensus Democracy in Homogeneous Societies. Acta Politica, 36(2), 117–128. Retrieved from https://hdl.handle.net/1887/3450795
Bogaards, M. (2017). Comparative Political Regimes: Consensus and Majoritarian Democracy. In Oxford Research Encyclopedia of Politics. Oxford University Press. https://doi.org/10.1093/acrefore/9780190228637.013.65
Bogaards, M. (2020). Kinder, Gentler, Safer? A Re-Examination of the Relationship between Consensus Democracy and Domestic Terrorism. Studies in Conflict & Terrorism, 43(10), 886–903. https://doi.org/10.1080/1057610x.2018.1507312
Bormann, N.-C. (2010). Patterns of Democracy and Its Critics. Living Reviews in Democracy 2.
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.
Lijphart, A. (2012), Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries, 2nd ed. New Haven: Yale University Press.
Poloni-Staudinger, L. (2008). Are consensus democracies more environmentally effective? Environmental Politics, 17(3), 410–430. https://doi.org/10.1080/09644010802055634
Powell, G.B., Jr. (1982), Contemporary Democracies: Participation, Stability, and Violence. Cambridge, MA: Harvard University Press.
Roller, E. (2005). The performance of democracies. In Oxford University Press eBooks. https://doi.org/10.1093/0199286426.001.0001
Tsebelis, G. (1995). Decision making in political systems: veto players in presidentialism, parliamentarism, multicameralism and multipartyism. British Journal of Political Science, 25(3), 289–325. https://doi.org/10.1017/s0007123400007225
-
Significant at the 1% and 5% level respectively. ↩︎