On proportional versus majoritarian representation

I'm not a specialist in voting systems. Because of my interest in improving institutions, however, I get asked about my thoughts on it with some frequency. Considering how I wish more people would share their views on parliamentarism, specialists or not, I thought it was only fair to expect of me to share my views on this topic too, so here it goes.

In short, I think voting systems might make a large difference, though probably not as large as parliamentarism, and I think that implementing such a reform would be at least as hard as promoting parliamentarism, if not harder.

Why I think the difference is probably not as large

I'm quite happy with the structure I applied to evaluate the parliamentary question in my book. Have it sanity checked, examine the theory, examine the direct evidence, examine auxilirary evidence. In all of these matters, I don't find the case for proportional representation (or the case for majoritarian representation, for that matter), as convincing as a transformative measure (but I'm not confident that it isn't either, as will make clear). 

Let's take the sanity check. Is it true that countries with proportional representation do that much better? Yes, to some extent. Most countries in Europe, the most developed region in the world, are not only parliamentary, but they also adopt proportional representation. If we examine presidential countries, the ones which have adopted proportional representation (concentrated in Latin America) seem to be doing better overall than the ones which adopt majoritarian elections (concentrated in Africa).  Parliamentary countries which adopt majoritarian systems can be very successful (like Canada, Japan, or Australia), but also not as successful as most European countries (like some Caribbean islands).

When we look at history, do we see that proportional representation was as important as parliamentarism for the great enrichment? I'm not so sure (but do correct me). When we look at the history of PR, we see that its adoption is relatively recent, when the advancements in institutions in Europe and its offshoots were already well underway. 

The success of countries which stuck to the majoritarian model also seems to point that PR is not that decisive. Yes, there are successful presidential countries too, but as I point out in the book, those seem to be exactly the ones that have the strongest parliaments, so they seem like parliamentarized presidential countries. But does it make sense to say that Canada or Japan are proportionalized majoritarian countries? Doesn't seem like it. 

When we look at theory, I also don't find the rationale as compelling. Yes, there is good theory in favor of PR, I quite like Huey Li's "Dividing the Rulers". But that theory seems to be particularly reliant on the rational voter assumption - that people vote their interests and their representatives defend those interests in parliaments. As I argue in the book, however, I see many reasons not to give too much weigh to the capacity of voters to hold their representatives accountable. Their main (and crucial) role seems to be in selecting reasonably competent and virtuous people which will be responsible for definig policy in a negotiated instead of a personalized way. And majoritarian elections can accomplish that.

What about the empirical evidence? I know one important study by Knutsen (feel free to point to others) which does find very large benefits from PR. Importantly, that study finds PR to be more important and more robust than parliamentarism itself, so that should count. But I fear that the restriction to democracies may introduce an effect of selection on the outcome, if there's something to the hypothesis that PR+presidentialism is a difficult combination. In other words, if we systematically drop the countries that have PR and presidentialism because they rarely become stable democracies, we will have more parliamentary+PR and presidential+majoritarian in the sample, so that the effect of parliamentarism might be attributed to PR. 

When it comes to auxiliary evidence, it doesn't seem nearly as clear as with respect to parliamentarism. First, I don't think that we have good studies showing differences in local government performance depending on PR or majoritarian in the same way as we have for "parliamentary" and "presidential" municipalities in the US. Why not? Most importantly, it seems that PR is very rarely used in companies, which can be achieved through the "cumulative voting" method which tries to ensure minority representation. Boards of directors are overwhelmingly majoritarian. To paraphrase the old saying, if PR is so good, why aren't people using it to get rich?

[Edited to add argument some thought was worth including] The point on the  the relative importance of parliamentarism or proportional representation may be clearer if we flip the argument. I think presidentialism is more harmful than majoritarian elections, if only  because presidential elections are the least proportional possible. We assign power to a single person. And in practice, in a separation of powers country, control of the executive seems more determinant of a country's policies than control of the legislative. 

Why I think it is at least as hard to implement as parliamentarism

Set aside the current relative popularity of parliamentarism and PR as causes. Consider what entails, in practice, switching from majoritarian to PR (or the other way around, as many propose for Latin American countries). The whole logic of who gets elected would change. That is the point of the reform. But if a country already adopts a system, we should expect that its politicians have a comparative advantage at getting elected under that same system. So the people responsible for implementing that reform would likely be those which stand to lose the most from it. 

What does parliamentarism entail for politicians in a presidential regime? For the vast majority, that they would be empowered. Of course, the current president of any country would not like to see their power curtailed. But such a reform could be designed in such a way that it would only be implemented after the president had left the office, in which case their opposition could be expected to greatly diminish. In fact, as I argue in the book, it can be a lot safer to be a former leader of a parliamentary country than to be the former leader of a highly personalized country. In his most recent book, Bruce Bueno de Mesquita shows that this was true of monarchs when creating and conceding powers to parliaments. I believe the logic still holds for presidents. 

One may object, however, that it is obvious that the relative popularity of parliamentarism and PR is not the same. The opinion in favor of PR is much more widespread than the opinion in favor of parliamentarism. I would agree, but point out that the relative neglectedness of a cause which, if given support, could be feasibly implemented, is an argument in favor of pushing for it, not against.

All in all, I support PR and hope it gets implemented. But I think that parliamentarism is still a better cause for anyone to push for, and particularly for me.