Friday, July 19, 2013

"If Men Were Angels..." Majority Rule is not Democracy!


We live in an age in which, increasingly, popular mandates (in the form of elections or referenda) are seen as necessary for legitimizing governance. This is in most respects a positive development. However, this has also come with a downside. The conflation of popular mandates with democracy leads to a sort of crude majoritarianism that devalues other important aspects of democracy such as the rule of law or the protection of individual liberties. In short, it leads to a “we are in the majority; we can do what we want!” type of mentality.

Recent examples abound. And I don't even need to bother discussing Russia, Venezuela, or other such dubious cases of democracy.

I conclude by drawing on political psychology to explain just why majoritarianism can be so dangerous. In short, human psychology has several tendencies that lead to undemocratic and dangerous actions, particularly in times of crisis.

First, some examples

Turkey: the causes and responses to the Gezi Park protests are a case study in crude majoritarianism (see here and here for my earlier posts). Broadly, the cause of the protests has been a series of government actions designed simultaneously to impose its own vision (in infrastructure, social policy, etc) while pursuing an intimidation campaign against the media and proposing sweeping constitutional reforms to centralize power in the hands of an elected president. The relatively small initial protests were met with a series of violent police crackdowns, leading to much larger protests against government repression. Prime Minister Erdogan’s response has been to dismiss the protestors' concerns while demanding that they cease their activities and seek relief at the ballot box. In the end, the protests were crushed by the police, with reports of at least five dead and thousands more injured.

Hungary: the European Commission (and many other experts) has criticized the government of Viktor Orban for its constitutional amendments designed to centralize power. Specifically, the amendments undermine judicial review, along with weakening media and academic freedom. Though the Orban government has backtracked on some issues in response to EU pressure, the reforms have the overall effect of centralizing power in the hands of the elected majority.

Czech Republic: the elected President Milos Zeman has created a recent political crisis by refusing to appoint the proposed Prime Minister of the governing coalition in the Parliament (Miroslava Nemcova), instead choosing his preferred candidate Jiri Rusnok. At the heart of the dispute seems to be his sense of having an electoral mandate, as he noted that he had received the votes of "nearly 3 million people, more than any political party".  The danger in this is that he could provoke a constitutional crisis by seeking to connect directly with voters, bypassing the parliamentary majority.

The common thread in each of these disputes is an attempt by a democratically elected majority leader or party to expand and centralize power at the expense of the minority. And this gets to the heart of the problem.

Political scientists agree that democracy requires much more than simply holding elections. And this is not particularly new. Robert Dahl made this point decades ago. Schmitter and Karl made this point forcefully two decades ago.

Three related problems emerge with associating democracy just with elections:

1. It ignores the fact that politics is fundamentally about competition and compromise over policy goals and means.

The point, of course, is that the majority should not get everything its way, because the minority still has defensible interests.

2. It negates and ignores the fundamental protection of individual liberties that are central to democratic theory.

Conflating majority rule with the “national will” makes it very easy to justify extreme policy measures, such as property confiscation or extra-judicial detentions, that violate individual rights.

3. It neglects the hugely important role of the rule of law in guaranteeing democracy and these rights.

The rule of law guarantees fair and equal treatment of all citizens and that government (and its officials) must operate in a stable and equitable legal environment. In other words, they cannot make arbitrary decisions, bypass necessary procedures, or otherwise override the rights of individuals or of political minorities.

Why does this happen? 

I believe that the inherent dangers in majoritarian democracy arise from human psychology. Humans are groupish by nature, and we tend very easily to overestimate unanimity and recoil from disagreement within our groups when we're in the majority. We very easily fall into the trap of thinking, like Turkish PM Erdogan apparently has, that the majority opinion is the 'national will'. This tendency creates a dangerous temptation to disregard the concerns of the minority.

Even more worrisome is the finding from recent research (paywalled) that ordinary people become more authoritarian when confronted with threat. (Even university students in far-away Arizona became more authoritarian in the days after the 9/11 terror attacks). People will support highly undemocratic policies when they feel threatened. Think about the Japanese-American internment during World War II or the extraordinary measures contained in the hastily-passed USA Patriot Act after the September 11 terrorist attacks. Needless to say, the target of these rising antidemocratic (and potentially violent) authoritarian responses will always be minorities and 'outsiders'.

In summary: in the best of times, we tend to overestimate the degree of shared opinion among fellow citizens (the 'national will'), many of us already have some tendency to support repression against those we don't agree with, and this tendency is exacerbated for most people when things get scary.

Given what we know about human psychology, a political system that allows the leaders of that majority to govern with few checks upon their power or guarantees of individual rights is very dangerous indeed.

What can be done to protect against crude majoritarianism? 

The answers are fairly obvious on paper, if not so easy to implement in practice. The benefits of institutional checks and balances, which limit the power of any arm of the state and give the other branches incentives to fight encroachments upon their powers, are fairly obvious. In most cases, consensus democratic institutions that promote power-sharing among parties and institutions are preferable to those that concentrate power.  Building a foundation of the rule of law and protection of individual rights is harder, but surely essential.

In short, a constitutional order that guarantees basic individual liberties, that enshrines the rule of law, and that disperses power among rival institutions and that facilitates the development of a multiparty system creates the necessary framework. That being said, the history of the conflicts in places like Cyprus remind us that institutional checks and veto powers must be proportional themselves. Majority rule does have a place in democracy; it is just not equivalent to democracy itself. James Madison understood this basic point over 200 years ago when he wrote, "If men were angels, no government would be necessary. If angels were to govern men, neither external nor internal controls on government would be necessary."

In future posts, I will elaborate on some particular institutional arrangements I think are especially problematic given what we know of human nature and democracy, as well as some that might be beneficial.

Sunday, July 14, 2013

Authoritarianism and Euroscepticism, an update

I wrote a commentary for the LSE EUROPP blog two weeks ago that summarizes the main claims and findings of my forthcoming article (paywalled) on authoritarianism and opposition to the EU.

Not surprisingly, I received some negative feedback from this commentary. One person on Twitter referred to me as an "academic with head up arse". Sadly, I haven't been that flexible in years. While much of it was predictable bluster (you can read the comments section of the LSE blog for a pretty good example), I think there were two objections (that I saw, anyway) with merit.

First, I didn't make clear my view that authoritarianism is one source of Euroscepticism. Of course, there are other causes of Euroscepticism, including political and economic concerns. There is a large academic literature on Euroscepticism that emphasizes these different sources. To give a couple examples, Kopecky and Mudde distinguish between those who oppose European integration altogether or on principle vs. those who may support the notion of European integration but oppose the EU in practice. Hooghe and Marks do a nice job of synthesizing the economic, political, and social identity sources of EU attitudes.

Authoritarians, I argue, are more likely to oppose the EU altogether (and not just because they think it doesn't function well, for example), and they are one (important) group of Eurosceptics. My conclusion, noting that "European leaders will need to find enough support from other citizens to overcome the opposition of authoritarians", was meant to highlight the point that non-authoritarians also might oppose the EU, but perhaps the point wasn't clear enough.

Second, a few comments seemed to suggest I was characterizing authoritarianism as a psychological pathology. I think this objection arises from two concerns. The first is that some academic studies (such as the original study, The Authoritarian Personality) do treat it as something like a disorder. But I don't agree with that view. Healthy societies surely benefit from having a mixture of individuals with different personality types. Authority, sanctity, and social cohesion are hugely important values for any society. Some people are predisposed to care about some of these values more than others. I see authoritarianism as describing one particular predisposition that has the consequence of making one more likely to oppose the EU in the present environment.

But my description of authoritarianism in the blog is also stark because I'm describing an ideal type. Authoritarianism is typically measured on a scale, and in the sample of Europeans I analyzed in the academic article it fairly resembled a normal distribution. Most people are somewhere near the middle, which implies that many Europeans are not as strongly authoritarian as my description. But, some are. And many citizens also have at least some authoritarian tendencies, which can be activated to oppose the EU.

Finally, none of this is to suggest that being Eurosceptical is a bad thing!

You can read the commentary I wrote for the LSE EUROPP blog at this link.

Sunday, July 7, 2013

The Trouble with Turkey's High Threshold

States using proportional representation (PR) electoral systems face a sort of dilemma. A major advantage of PR is that encourages a wider and more diverse range of parties to enter parliaments, as they don't face the high electoral barriers posed by plurality ("first past the post") systems. The downside is that too many small parties could mean ungovernability.

In an attempt to balance these concerns, many countries that use PR (or mixed PR/plurality systems, like Germany and New Zealand) employ a legal election threshold: a minimum share of the national vote that a party must win in order to gain seats in the parliament. Typically, this threshold is between 3-5% of the national vote. Sometimes, countries go a step further to protection parties representing particular regional or minority interests. In Sweden, for example, a party that does not meet the 4% national threshold can still gain seats if it wins 12% of the vote in a single electoral district. Other countries have exceptions for parties representing ethno-linguistic minorities. So, on the whole, it seems like a pretty good way to obtain a high level of representation without producing an overly fragmented, ungovernable parliament.

That is, until you take a look at Turkey. In Turkey, the threshold is 10%. Why? Officially, for all the good reasons of promoting stable, strong governments. If you're more conspiratorial, you might suspect that it had something to do with keeping Kurdish parties out of parliament. Whatever the reason, it produces a pretty bad end result.

Why is this a bad idea?

The usual understanding is that the downside of SMD systems is that it hinders the formation of third parties. Because SMD systems are "winner takes all" at the district level (meaning that the winner gets the seat and everyone else gets nothing), voters are reluctant to "waste" their votes on small party candidates. This tends to favor the development of a two-party system--a pattern known as Duverger's Law. (I'm oversimplifying a lot here, but this is the basic idea...).

To think about it differently, elections are (among other things) a problem of coordination among party leaders, candidates, and voters: given a choice between two otherwise equal parties, you want to vote for the party with the better chance of winning. Prospective candidates, donors, and so on would also rather support a competitive party rather than an uncompetitive one. Duverger's Law is an empirical observation that, over time, parties and voters manage to (more or less) solve this coordination dilemma by settling on a pair of dominant, rival parties--to the exclusion of most smaller parties. While SMD has other problems, one upside is that it tends to produce a stable party system anchored by two big parties that usually take turns in power.

PR, on the other hand, "solves" the coordination problem by reducing the need for a solution. PR is supposed to allow for the easy entry of new or small parties into parliament. Under PR, one should be able to vote "sincerely" without having to worry about that party's electability. The problem is that an electoral threshold works against this electoral "permissiveness" by creating a barrier to entry, but without the corresponding incentives toward electoral coordination.

In PR systems with low thresholds, this isn't a terribly big deal. The level of electoral coordination needed to surmount a 4% threshold is not that great. Nor is it a terrible failure of representation if a couple of parties with 3.9% of the vote don't gain seats.

But when the threshold is 10%, as it is in Turkey? Then you've got a potential problem. Here's the 2002 general election result:

2002 Turkish General Election Results
Party
Vote %
Seat %
AKP
34.3%
66%
CHP
19.4%
32.4%
DYP
9.6%
0
MHP
8.3%
0
GP
7.3%
0
DHP
6.2%
0
AP
5.1%
0
Others
8.6%
1.6%

The 2002 election represented a failure of electoral coordination on a pretty grand scale. About 45% of those Turks who cast a ballot went unrepresented, because they made the mistake of voting for one of the parties that failed to clear the 10% threshold. Five parties finished with more than 5% but less than 10% of the vote...and no seats in the new parliament. 

This coordination failure produced an incredibly disproportional electoral result. With less than 35% of the vote, the AKP (Justice and Welfare Party) of Tayyip Erdogan controlled two-thirds of the seats. Only one other party, the CHP (Republican People's Party), was even in parliament.

What would it have looked like had Turkey had a more reasonable threshold of 5%? Here's my best approximation, given that Turkey uses the d'Hondt system and that it has quite a lot of very small electoral districts--both of which tend to benefit larger parties. (I'm just using the national vote totals to estimate the seat share, but this should be close enough--particularly considering that the AKP does best in rural areas with low district magnitudes).

Simulated Results with 5% Threshold
Party
Vote %
Seat %
AKP
34.3%
48.3%
CHP
19.4%
24.4%
DYP
9.6%
9.3%
MHP
8.3%
6.6%
GP
7.3%
5.5%
DHP
6.2%
2.2%
AP
5.1%
2.2%
Others
8.6%
0

(Since I don't quite know about the handful of independents who won seats in 2002, I leave them out of the equation here. It wouldn't greatly change things if I included them).

Note the differences. First, the AKP does not hold a parliamentary majority. Its share of the seats is reduced by nearly 18 percentage points. The CHP loses seats as well. Second, all of the parties over 5% of the vote get at least a few seats. While the combination of d'Hondt and many districts with low magnitudes means that the results still benefit the larger parties, the results are far more proportional!

What would this have meant for Turkey? Instead of an AKP government controlling two-thirds of the parliament, Turkey would have had either a coalition (with DYP being the most logical for various reasons) or a minority cabinet. Either way, AKP would have had to share power during that five-year period. In the present time, with many complaints about the ineffectiveness of the CHP as an opposition party, the barrier to entry for a new liberal or social democratic party would be considerably lower (and it's hard to dispute this would be a good thing for Turkey right now).

The long story short: the 10% threshold matters greatly for Turkish politics, and largely its effect has been damaging to the goal of representing citizens effectively. Reducing this threshold to 5% or less should be a central demand of all parties in Turkey today.