Democracy Index?

Today's bit of silliness comes from The Economist in the form of its Intelligence Unit Democracy Index. Apparently, the United States is only the 17th most democratic nation in the world, behind Canada (???), Sweden (!!!) and FINLAND.

In fact, Sweden tops the list. Sweden, ladies and gentlemen, is meant to be the the model democracy. Yes, you read that right - the country that essentially has one-party rule, virtually legislates how many women and men can serve in parliament and has no judicial review is the example for the rest of us to live up to. It has hate speech laws and a film censorship board (the longest-functioning in the world, in fact), regulation of radio and TV. Nearly all major industries are owned by the government, and the taxation rate is among the higest in the world.

But of course, I'm confusing "freedom" with "democracy." The two are NOT the same thing (though they are highly correlated). I guess the US is pretty definitely a freer country than Sweden. But I believe it is also more democratic.

Let's take a look at some of the Economist's sleight-of-hand.

The methodology is as follows: in each of five categories (Electoral Process and Pluralism, Functioning of Government, Political Participation, Political Culture, Civil Liberties) there is a number of questions, usually 12-15. The question is scored either 1 or 0 depending on whether the nation in question meets the requirement or not. Presumably, therefore, it is possible to get a score higher than 10 in some categories, but this is only scored 10, thus biasing the entire survey upward - but critically more in some cateogires than others. A further complication here is that some questions arbitrarily have a 0.5 rating option available, so the questions also have the drawback of being "categorical" in an uneven fashion. Another obvious problem with this methodology is that it gives equal weight to disparate measures, something that is almost certainly inaccurate. To pick an example from the survey - "Do laws provide for broadly equal campaigning opportunities?" is deemed equally important as "Can citizens cast their vote free of significant threats to their security from state or non-state bodies?" So a nation in which gangs of ruffians routinely beat up people who do not vote the way they like loses ... one point. And a nation which does not provide public campaign financing loses ... one point. Each of which actually comes out to 0.2 points, as the overall final score is an average over the 5 categories. Yeah, this sounds like an informative survey.

Of course, it will be objected that any survey which tries to weight various things according to significance will have engaged in spin and/or bias. That is, it will have veered away from those things which Everyon Can Agree constitute a functioning "democracy" and instead prejudiced the results to favor its chosen candidates. And I suppose to some important extent this cannot be denied. But that rather speaks to the futility of trying to come to an "objective" analysis of what constitutes a functioning "democracy" in the first place. It would probably have been preferable to simply draw up one's own subjective list and then try to justify it. But even if we're committed to continuing the "objective" analysis in this way, it strikes me as being a better approach to simply rate the surveyed nations on tiers rather than with continuous scores. That is, come up with five "levels" of democracy or something, rather than attempting, on the basis of absurdly weighted all-or-nothing scores, to give a rating on a continuous scale of 1-10 of "relative democracy."

Now, I could really take two attitudes to this - one to say it's bad (or at least questionable) methodology, and another to say that the results have been deliberately skewed. Since this is the Economist, I opt for the latter.

Of course I have nothing but circumstantial evidence and my general displeasure with the Economist to go on. But a close look at some of the questions seems to render my suspicion that the test has been biased in favor of Sweden in particular at least plausible.

Consider this question:

10) Do opposition parties have a realistic chance of achieving government? 1: yes. 0:no.

And there's a 0.5 option "There is a dominant two-party system in which other political forces never have any effective chance of taking part in national government."

This question seems not only tailored to come out against the United States, but specifically tailored in such a way to make sure that Sweden still makes the grade. Sweden, you see, is in a situation where the ruling party has won every major election since 1932 save four - and until very recently those aberrations resulted in at most a 6-month stint for the opposition. So Sweden has something much worse than the two-party "monopoly" that the United States has: its politics are completely dominated by a single party and have been in living memory. However, there is no 0.5 rating for "Has an effective single-party system where the majority party has been known to lose an election or two once in a very blue moon, but usually isn't out of power for long." I suspect there's also some sleight-of-hand going on with the parliamentary system here. In nations like Sweden and Japan - both of which absurdly score higher than the US on "plurality" - even though both are de facto one-party states (Japan is much, much worse than Sweden here - the LDP has as good as never lost a national election - only one aberration in 1992 when the Socialists won, and that only happened due to LDP party-internal intrigue) - other parties are technically "represented" in parliament to a greater extent than, say, the Libertarians or the Socialist Worker Pary are in the US. But when your electorate has a habit of putting the same cronies into a majority every time, and when Parliament is the whole of the law (unlike in the US, where we have separation of powers), then the reality translates into something much less "plural," I'm afraid.

It will, of course, be argued that if 56% of Sweden wants this one-party rule, then isn't it "democratic" to let them have it? Quite so - but then if a much greater percentage of the American population is satisfied with the current two-party system, then isn't that "democratic" as well? There's nothing specifically barring the Libertarians from Congress, it just happens that they're not popular.

Another bit of annoyance is that lots of questions seems to be weighted against systems that have had recent changes. For example:

Is the process of financing political parties transparent and generally accepted?

It's the generally accepted part that, again, seems biased. Political financing in the US is completely transparent, but new regulations were adopted recently and are still controversial (indeed, I am staunchly opposed to McCain-Feingold). Nations like Canada, Sweden and the Netherlands, which have long had draconian speech and finance regulations for elections, don't fall afoul of the "generally accepted" addendum because those issues were decided long ago. A less biased way of phrasing this would have been "followed according to the rule of law" or some such.

Extent to which the government invokes new risks and threats as an excuse for curbing civil liberties. 1: low 0.5: medium 0:high

This seems like an uncontroversial question, but what constitutes an "excuse" for curbing civil liberties? Somehow I imagine the US got a 0.5 on this because of the PATRIOT Act, even though I highly doubt that the 9/11 attacks were an "excuse" for curbing liberties. I don't personally agree with the PATRIOT Act and would vote against it if in Congress - certainly I'm against its renewal - but neither do I think it's based on an "excuse." I believe that most of the people who support it are sincere; it's not a trojan horse to dictatorship.

The use of torture by the state

This one only has yes/no options. There is no 0.5 option for "used, but only against enemy combatants." Again, one can't help but wonder whether the US got a 0 on this one because of Abu Ghraib, even though there is no policy of torturing citizens over domestic matters - i.e. Abu Ghraib has little to do with democracy here at home.

The question on free speech includes a clause "bar only generally accepted restrictions such as banning advocacy of violence." So again, it seems designed to prevent the US, with its much more liberal speech laws, from having an advantage over places like the Netherlands and Sweden, which have Holocaust Denial restrictions and heavy hate speech codes.

The tell-all example, though, is an opposition between the question on voter turnout/participation and the question on women in parliament. Under the question on voter turnout, there is an instruction to score 0 if the country has a mandatory voting policy (like Australia). Fair enough - if we're measuring "participation," I suppose it's actually "voluntary participation" that we're interested in. But then there's a question on the representation of women at the national level. There are three categories here too: score 1 if more than 20% of your legislature is female, 0.5 if it's 10-20%, and 0 if it's less than 10%. The US may or may not have made the 10% mark. The Senate is 13% female, the House 7%. The telling thing here, though, is that there is no corresponding instruction to exclude countries like Sweden where women are overrepresented in parliament due to quota laws. Apparently, forcing people to vote does not constitute "participation," but forcing people to vote for women is off the hook when we're studying sexism.

The real question, of course, is why we're measuring female participation in terms of outcomes in the first place. A more appropriate category would've been whether or not women can safely run for and hold office.

Anyway, I could pick at this all day. The reader is invited to have a look for himself and see what I mean. Let me just close with my personal take:

1. I am not so concerned with whether the US is at the top of any international ranking on how "democratic" any given nation is since I'm sure we're not number one no matter how you draw the charts. My beef here is with nations like Sweden and Canada outranking the US. I don't consider either one an exemplar of democracy. Certainly I'd rather be Canadian than Burkina Fasan, but Sweden and Canada are NOT model democracies, and they're certainly NOT more "democratic," on any sensible measure, than the US. Australia might be. New Zealand might be. The Netherlands might be. But Sweden and Canada? Nope.


2. Any survey that thinks Japan is a strong democracy is smoking crack.


3. It is probably futile to try to rank nations on the basis of how "democratic" they are - and it's ultimately useless. I don't think this can be boiled down to a single term. Even so, if we must do a survey, I would prefer categorial results rather than a continuous scale.


4. Democracy is less important than freedom (civil liberties). I'm happy choose to live in an undemocratic oligarchy that guarantees my rights over a democracy that might not.

That is all.