George Tsebelis is the Anatol Rapoport Collegiate Professor of Political Science at the University of Michigan. He is famous for his theory of “veto players,” a technique for analyzing the steps necessary in a given political system to effect change. His 1995 article, “Decision Making in Political Systems,” introduced the concept, and his 2002 book “Veto Players” refined it. We talked on the phone Friday; a lightly edited transcript follows.
Dylan Matthews: For our readers who may be unfamiliar, walk me through the basic concept of veto players.
George Tsebelis: The basics of it are that in every political system, there is a group of actors that need to agree in order to change the status quo. The reason that people care about politics is precisely because of the implications on policy. The most important events that happen in everybody’s life, except for very personal things, are political events. So being able to change the status quo or not being able to change the status quo is of paramount importance.
Obviously the concept is not new, it dates back to concepts used in the Founding Fathers’ documents. I was trying to think how far back I wanted to go and, I had found it in the history of Rome. It’s not a new concept. What is new is that I am trying to have conclusions that are consistent with the framework. So that’s what I have been trying to do for many years, ten years since I came up with the book. The first time that I came up with this idea was 15 years ago. But as I said, it exists already for centuries among people who have been thinking about politics.
The basic idea is that the more veto players you have, the more difficult it is to change the status quo. And the more ideological distance the veto players have from each other, the more difficult it is to change the status quo. The more different the players are, the more checks and balances you have. So you can think about all political systems that way, whether they are presidential or parliamentary, and you can think about systems in the United States, which has three institutional veto players — the House, the Senate, and the president — and parliamentary systems like Italy, where there are multiple parties that are veto players and form the government coalition, and once you have multiple veto players it is very difficult to change the status quo.
The people who don’t like the status quo who will say they don’t like the system, and people who like the status quo, when they are in a system like Britain where things can change from one policy to the next very easily, will say that as well. The attitudes that people have depend on their predispositions toward the status quo. But on the other hand institutions have big inertia. You cannot change the system every time you don’t like what it produces.
Look at what’s happened with the filibuster in the Senate. On the one hand, Democrats want to reduce its impact. Obviously the easiest way to do that would be the nuclear option, but the Democrats don’t want to do it because they’re afraid that in the next round they’ll be in the minority and they want to be protected by the filibuster rules. You may not like the rules but they are going to stay there for a long time.
DM: You identified the House and Senate as veto players. Is it possible to think of individual legislators in those bodies as veto players themselves?
GT: Yes. First, in 2009, the Senate had 59 democrats and the president tried really hard to make it 60. Once you have sixty, then only one of them is required in order to allow the filibuster, and that’s why the 60 Democratic Senator Senate required unanimity among the Democrats. So somebody like Joe Lieberman, who comes from a state where the insurance companies have their bases, was a veto player. And then it was not possible to change the legislation without his approval, and in the later stages of the game, you couldn’t modify what the Senate had done, because it would require a new vote, and the number of Republicans had become 41 after Kennedy left. Then there were 60 veto players.
If you have a big majority of one of the political parties, then nobody in particular is a veto player. Of course, you can make a majority without him or her, but on the other hand the bigger the required majority, the more difficult it is to get Senate approval. Because remember: veto player is the individual or collective actor who’s required to change the status quo. In the United States, the collective actors are the Senate, and the House (the president I an individual veto player). If you open these collective veto players and you have more than the required majority, then none of their members in particular is a veto player (required to agree for a change of the status quo).
DM: So strong party discipline would reduce the number of veto players. You don’t think of, say, individual MPs in Britain threatening the passage of legislation.
GT: Indeed, if the US had strong party discipline the only veto players would be the President, the Speaker of the House and the Majority leader in the Senate (assuming filibuster proof majority). So the British prime minister usually, when you have the single party goverment he is the only veto player in Britain. That’s why change in the status quo in Britain is easy. Same thing in Spain, because again you have a strong majority and a single party majority, and historically in Greece, because again you have a single party majority.
Countries like this can change the status quo easily, and countries like the United States and much, much more so the European Union cannot change the status quo. When we see that, we as Americans ask, “Why doesn’t the EU do what it needs to do?” Well, because there are 27 actors, 27 veto players. Because they decide by unanimity at the state level, they all need to agree, and it’s really very difficult.
DM: This is all quite intuitive at a theoretical level, but I take it there’s empirical evidence, concrete examples to back the theory up.
GT: There are. One example would be if you changed, say, the budget. Big budget changes, that’d be a big difference. If you make changes that affect many people, if you make legislation about the equality of women, that’d be significant legislation. There are cases when legislation is significant even when it doesn’t affect the budget or many people, like gay marriage. Empirical studies demonstrate the more veto players you have, the more difficult it is to change things. In my book there is a chapter that compares labor legislation across European countries. There I find that countries with many veto players have fewer significant labor bills than countries with few.
If you take Obamacare, it’s legislation that happens once every hundred years. Clinton tried to do it and failed, and the problem is that American institutions are not designed to permit big changes. Obama had an extraordinary situation, where he had a Democratic Senate and a Democratic House. Now that it’s done it’s really hard to undo it, because it requires three republican veto players, and if that doesn’t happen then you continue having the same legislation. If you have one Democratic veto player, he will veto any changes from the status quo.
DM: And you’ve been able to correlate the number of veto players with objective outcomes for various countries, such as economic growth, no?
GT: Some research has been done onbudgets, and they have found out that changing the composition of the budget is more difficult once you have a multiparty coalition government. Economists have found that over long periods of time countries with multiple veto players have higher levels of growth (their argument is that investment is higher because investros are not threatened by political change).
Other people have found the same thing with foreign direct investment. Among the countries that have appropriate legislation (property rights), countries with more veto players will attract more direct investment, because the investors won’t be afraid that the situation is going to change. If you don’t have the correct legislation, and you have many veto players, then you don’t attract investment, because the stitutation may change. the number of veto players determines how active the political system is going to be. If you have few players, it’ll be simple to change things at any given time.
DM: This seems relevant to the debate surrounding presidential and parliamentary systems. You have people like Juan Linz who argue that presidential systems are likelier to collapse into dictatorship.
GT: Yes, we have found out that presidential systems are more vulnerable to moving toward non-democratic solutions. Parliamentary systems have been adjusting by themselves. If the government is not doing whatever the population wants in a parliamentary system, then the government is going to fall. In a presidential system, this really does not exist. This is something that empirically has been demonstrated. The likelihood of survival of democracy is much greater in parliamentary systems than presidential systems. The United States obviously is the big exception to that.
DM: Under what circumstances do you see big institutional shifts in the number of veto players? Britain, for example, in the early 20th century dramatically reduced the power of the House of Lords. What prompts changes like that?
GT: Right now, I am looking at the European Union, where the problem is much more acute than in the United States. It requires 27 countries to agree to do anything. They were able to do a treaty, a fiscal pact among the 27 countries because it was extremely urgent — the Euro would collapse otherwise — and because there was essentially an agreement between [German chancellor Angela] Merkel and [then-President of France Nicolas Sarkozy] which then got introduced among the chiefs of the government, and it was approved despite objections. On the one hand, this isn’t democratic, but on the other hand, the Euro was at stake. They’ll try to have a much bigger debate down the road.
If it is extremely important and you have a deadline before you, the objections become less pronounced. That is the same thing with the senate, because you see that it is dysfunctional and there is thus pressure to change. You can politically try to do it, and whether you are successful obviously depends on the positions of the different actors and the severity of the problem. It depends on whether the Republicans and Democrats will be able to find an agreement.
In the United States, if this kind of sharp division and polarization that we are seeing now is one we’d want to try to resolve institutionally, then the major solution would be a change in the electoral system, and a transformation to Single Transferable Vote, which means that every voter would rank the different candidates. Essentially, that would eliminate the primaries, which are the cause of both Republican — I don’t know if “extremism” is the word — but Republican intransigence, but also in some cases Democratic intransigence.
What STV does is, even if you are an extremist of one side or the other, your second vote will go to the more moderate candidate on one side. You’ll have a house or senate where moderates prevail. Major institutional change like this could change the outcome.
DM: One objection that mathematicians sometimes make to Single Transferable vote, and its sibling Instant Runoff Voting, is that they don’t always pick the Condorcet winner, or the candidate who would win a one-on-one contest against all the other candidates.
GT: The only way it won’t select the Condorcet winner is if the winner gets eliminated in the beginning. If you have three candidates — right, center, and left — then unless the centrist candidate gets the lowest number of votes out of the three, he is going to be elected. The reason is because if you lose one of the two sides first, if it is the left candidate, then their votes will be added to the center one. and given that the other guy didn’t get a majority, then it means the majority will go with the center. So, the centrist is elecgted unless he got the lowest number of first votes. But even if the centrist is eliminated, then among the left and the right, his votes will go to the person who is closer to the center. So STV does affect the outcomes in favor of the center, and it eliminates polarization.
DM: You would also probably see less negative campaigning. Candidates would be competing both for #1 votes and for #2 and #3 votes, so they wouldn’t want to insult a candidate who other voters are ranking first.
GT: Yes, exactly. It affects both the voters and the candidates. It’s going to have a big impact. In the US, it has been used in municipal elections, and also in systems like Australia, Malta, Ireland, has been used at the national level. So it’s, in my opinion, not so foreign, and not so a priori excludable because of worries like “how would the people be able to do it?” If it becomes clear it would have this kind of effect, the moderate republicans would want to have such a change, because it would solve their problem in the Republican party. It is a politically feasible solution, in my opinion.
DM: It seems like the other main form of proportional representation, party list, would both increase party unity and increase polarization.
GT: Yes, and it doesn’t work in presidential systems, because then we would have complete deadlock most of the time.
DM: To circle back a bit, you identify the House, the Senate, and the president as the three veto players in the US. Where does the judiciary come in? Is that a veto player?
GT: Well, it depends on whether it’s going to make a judgment on statutory interpretation or constitutional interpretation. If it’s statutory then it isn’t. If it is a constiutitonal interpretation, then it is a veto player. But it depends on what its position is going to be.
Here, the American constitutional court could make a judgment that is constitutional, however the judgment is going to be on the basis of the median voter inside the court. This median voter is very unlikely to be politically on the other side of the american political system, because if there is a political decision, it requires the president, the House, and ⅗ of the Senate to pass in the first place. So it is very unlikely that the median voter of the court is going to be outside this political spectrum.
On the other hand, in Europe, the constitutional courts make only constitutional decisions. They are a separate institutional branch that deals only on constitutionality issues. The judges are selected on the basis of overwhelming majorities, and they have requirements by the laws to be either higher court judges or university professors, lawyers, and to have this kind of majority support, so these people are really in the middle of the road politically. They are not going to be like [failed Supreme Court nominee Robert] Bork in the United States.
They are unlikely to excercise a veto, except in the form of an expert judgment. They would say to the government, “If you want this to happen, you should not be doing it this way.” Then the government is going to take their opinion and modify its recipe, but not their political goal. While constitutional courts are veto players, they rarely exercise a political kind of veto.
Someone has written a paper analyzing the Italian constiutional court, because its constitution can be changed very easily. It required the support from the government and the Communist party. So whenever the government and the Communist party were closer to each other, the Italian constitutional court would not invalidate many pieces of legislation. When they were far apart and the constitution could not be changed, then the Italian constitutional court would interfere much more actively.
DM: Are there cases where some issues have a certain set of veto players and another category of issues have a different set? In the U.S. and France, say, the president is constrained by parliament or Congress on domestic policy but has fairly free reign on foreign affairs.
GT: Well, yes, there are, obviously. Let’s take the United States. Why does Obama have much more success in foreign affairs, and change the climate in the world from what had existed under George Bush, compared to how difficult it has been for him in the last two years to do things domestically? In foreign affairs he is the sole veto player, whereas in domestic affairs he is one of the three. And why is it that presidents in their second term do foreign affairs as opposed to domestic? Because they don’t have the same majorities. It’s a difference in the number of veto players.