Varoufakis and Elster, place your bets…


Understood through the lens of game theory, Athens and Europe are suspended between two alternatives: cooperation or defection. Greek finance minister Yanis Varoufakis discusses the insights, and errors, that emerge from a game theory analysis of the Greek crisis with the philosopher Jon Elster. Translated by David Broder; read the French version here.

Jon Elster is Emeritus Professor at New York’s Columbia University and the Collège de France. He is known for his both rigorous and cutting critique of the basic axioms of economic theory – actors’ interests and rationality. He is the author of Traité critique de l’homme économique.

Yanis Varoufakis is Economics Professor at the University of Athens, visiting professor at the University of Texas and a former consultant for video games developer Valve Corporation. He was named Finance Minister in Alexis Tsipras’s government after the January 2015 elections. A game theory specialist, against the dominant tendency, he thinks that cooperation is more rewarding in the long term than each-for-his-own. His work The Global Minotaur, a fascinating essay on the crisis of global capitalism, has just been translated into French by Éditions du Cercle.


As night falls on the Acropolis, I stride along the narrow streets of the Plaka district toward Syntagma Square, where the neoclassical Parliament building stands opposite the modern Finance Ministry tower. I have an appointment with the minister Yanis Varoufakis. A brilliant economist and a specialist in ‘game theory’, he has agreed to have a phone conversation with the economics philosopher Jon Elster in New York. The stigmata of the crisis are visible in Athens: in some parks young people have set up tents and are growing vegetable patches; on the café terraces old people try to sell packets of tissues to get a few euros; an installation at the entrance to the ministry bears the traces of a mobilisation by five hundred cleaning women who worked there until they were laid off under Troika pressure, before being rehired again. One of them, of Bulgarian origin, was even elected a Member of the European Parliament. Everyone agrees that everything seems as if, faced with the abyss – collective bankruptcy, euro exit, mass unemployment and generalised precarity – Greece is inventing a new way of doing politics, breaking with the norms of a caste that traded the people’s votes for privileges. But all Greeks are torn between the demand to reduce their debt and that of staying in Europe, between the need to put an end to the austerity that is strangling the population and the promise of restoring their dignity. Is it in the Greeks’ interests to cooperate with the Europeans? Do they have the means to do so? Or should they play the card of defection? And what about the Europeans: are they right to back the Greeks into a corner? After two hours of fascinating discussion, addressing history, morality and strategy just as much as economics, Yanis Varoufakis grasped my arm, saying ‘Given the situation, people will criticise me for wasting my time talking with philosophers. But I really like to do so. They see things more clearly’. Leaving the Finance Ministry, I reach the side of the Parthenon, where two and a half thousand years ago a certain Solon founded the first democratic institutions… and cancelled the debts of the popular classes. Will history repeat itself?

Yanis Varoufakis: I am not a professional politician. I have been an economics theorist my whole life, my speciality being game theory. I went into politics two and a half months ago. I define myself as an ‘erratic Marxist’ insofar as I think that in giving his theory the status of a science of society and history, Marx bears some responsibility for the authoritarianism of part of the Left. But I remain convinced that he provided one of the most penetrating visions of capitalism. Every time that I look out the window I am confronted with the contradictions that he identified.

Jon Elster: I would present myself more as a ‘faded’ Marxist. My Marxism even ‘faded before it could flower’. I grew up in Norway. My father was a spokesman for the Labour Party. When I came to Paris I wrote a thesis on Marx with Raymond Aron. I became a Marxologist. After that I reflected on rational choice, which supposes that reason and interest govern our choices. I developed a critique of this vision of the ‘rational, interested’ man, which does not correspond to the way that individuals behave in their lives, or to their motivations. That is why I am curious to know what motivated an academic like you to get involved in politics.

YV: Isn’t a Marxist necessarily engaged in politics? My family is a result of the Civil War that tore Greece apart. My father grew up in Cairo, where he received a French education, in Voltaire, Rousseau, and his idol, Robespierre. He migrated to Greece in 1946, just before the Civil War broke out again. Following a conflict with the university management over his right to enrol, the secret services captured him and demanded that he sign a document denouncing communism. He got three and a half years in a camp for refusing to do so. When it closed, in 1951, he returned to university, where he met my mother, who had joined a fascist organisation. What was her job? Watching my father! So here I am! (laughs). With time they learnt to have rather more temperate relations, but when fights broke out my mother would shout at my father ‘damned communist’!’ and he would reply ‘dirty fascist!’ I mention all this just to say that I became politically conscious very early.

JE: But how did you end up with Syriza, when, like me, you were an academic at some distance from political life?

YV: In 2008, after the Wall Street crash, the Greek economy was the first to collapse. It was then that I came to believe that the plans to save it would lead to an impasse. Faced with the threat of the banks going bust, Europe made the biggest loan in history – on condition of austerity. That cannot work. If you put the greatest loan in history on your shoulders, in a context of falling incomes, you condemn yourself to repeat the experience of the 1930s. Very quickly, nations clash and the extremes surge. Since I considered that the answers being offered weren’t the right ones, I entered into politics.

JE: I take a close interest in the European crisis, but I am not an economist. To me the economics of finance seem like an inextricable tangle of economics, psychology and politics – which makes it impossible to understand.

YV: You’re claiming that you’re not an economist? Allow me to disagree! I have been reading you for a long time. I still have in my head the unanswerable critiques you formulated in Ulysses and the Sirens, on the inconsistency of the principle of maximum utility. Whoever wrote a text like that has no right to pretend not to be an economist, to claim not to understand anything about finance! 

JE: I am interested in the role that emotions play in individuals’ behaviour. I am struck by the influence that emotions linked to historical memory have in the Greek crisis. The Germans will never forget the hyperinflation of the 1920s, and the Greeks will never forget the atrocities that the Germans committed during the occupation. These painful memories have produced emotions, or even obsessions, that are very difficult to shake off in the interests of reaching a good agreement. Emotions induce individuals and peoples to think and behave in an irrational manner. They affect both our capacity to evaluate risks and our degree of risk aversion. 

YV: Nazism did not triumph in the 1930 because of hyperinflation, but because of mass unemployment. By 1931–32 the hyperinflation had been beaten, the Mark had been replaced, prices had stabilised, and the Nazi Party got no more than 10 percent support in the elections. Then Hjalmar Schacht implemented his austerity policy, which caused unemployment to explode, and it was then that the Nazis scored electoral victories. As for Greece, I don’t think we are obsessed by the occupation any more. What we are defending, faced with the Germans, is our dignity.

JE: But in your negotiations didn’t you put the question of the loan that the Germans imposed on the Greeks during the occupation back on the agenda?

YV: That was unique in the whole Second World War… there is a fascinating document in the Greek Central Bank, drawn up and signed by the German army. They took an enormous amount of money to finance the costs of the occupation. Before leaving, they left a signed contract committing to returning this money, and specifying at what rate they would do so – only Germans would do something like that! We still have this contract. I would not demand that the Germans pay it back in full. But if Mrs. Merkel wants to come to the Central Bank in Athens and give us a symbolic 1 euro in lieu of this debt, even that would be great.

JE: Here we have two moral fables, one saying that the Germans have to conform to their promise from the occupation, the other saying that the Greeks have to pay their debt back in full.

YV: Let’s exchange a euro for a euro. Or one for twenty. What counts is that there should be a settlement. When the German government insists that the war reparations were settled, I ask: when? By whom? The debt, they retort, was wiped out during German reunification. That is not true, legally speaking. In the 1950s and 1960s the Germans said ‘We will not pay until the two Germanies are reunified. There is no reason why the West should have to pay on behalf of the East’. In 1991, at the moment of reunification, Greece was excluded from the negotiations. And nothing was paid. 

JE: It’s important to avoid moralising what is at issue, here. 

YV: I am against moralising, particularly when it comes to debt. It is not ethical, and it’s not good economics either. But when you tell me that my debts are sacrosanct while other people’s have been eliminated, because they unilaterally decided that this should be so… that will not do.

JE: We also have to stop allowing the past to contaminate the present. 

YV: But I don’t think that we need to get rid of historical memory. 

JE: Neither of the two sides should demand absolute respect for contracts from the past. 

YV: On that point, I’m with Hannah Arendt. Since at least one German died at Auschwitz, the German people cannot be considered collectively responsible for Nazism. Similarly, so long as there is one honest Greek, the Greek people cannot be stigmatised as a band of layabouts and fraudsters. I am prepared to distinguish who is responsible for what, to dissociate past from present. But my interlocutors never stop telling me: a debt is a debt. If that is the case, then a debt is a debt for everyone.

JE: In the 1920s Keynes wrote that to turn taxpayers into bondholders’ slaves is the beginning of the revolution. His statement is more relevant now than it’s ever been. 

YV: This is what I call ‘the commandments of economics’. When my interlocutors point the finger at us, I remind them that for every irresponsible borrower there is an irresponsible lender.

JE: The question of responsibility is an essential one. Who is to blame? Firstly, those who agreed to Greece entering the Eurozone before undertaking structural reforms. The functionaries in Brussels knew that the Greek economy was a bottomless pit. They integrated Greece even though they knew this, in the name of Europe and democracy. Papandreou’s Socialist government also knew that the country was full of holes, but there were clients to serve so it recklessly ploughed ahead anyway. During a visit to Athens my son had his leg bitten by a stray dog. We went to the emergency room, and I saw that the suspicion that Greece is a free-spending country was no myth! 

YV: Who is to blame? Let’s go back to the origins of the crisis. What happened in Europe in the 2000s? After having linked together their currencies, the European economies worked up a frenzy of improper loans. The French and German banks came to deficitary countries in Europe north and south, seeking firms and local governments who they could shower with liquidity. These loans were then cut up into small slices, wrapped up as nifty financial assets and resold. The idea was that if they were discharged a little bit everywhere, this would cancel out the risk. In 2008, that resulted in the subprime crisis. When these loans went bad, those who had profited from them were bankrupted. That’s what ‘saving Greece’ has mostly been about: of the €240bn borrowed by the Greek state from then on, 91 percent has been to save the banks. Those responsible for this frenzy of borrowing have been ‘saved’ by this new loan, while the new taxes are incumbent on citizens who had not themselves borrowed anything. Those who had never touched Pandora’s Box had to pay.

JE: Why is it Greece that has been sacrificed? 

YV: Let’s imagine that Greece hadn’t joined the Eurozone in 2000: the same thing would have happened to Ireland or Portugal. The problem comes from the fact of a badly configured monetary union between deficitary countries and other, self-sufficient ones. How does a union like that work? Germany sells cars to Greece, Greece buys cars from Germany, hence a negative balance of trade. The profits that BMW realises in Germany are invested in Greece. Since this latter has no car industry, the money goes into property, feeding a housing bubble. Then came the 2008 crisis. The deficitary countries found that they had debts that they could not pay. German banks forced German taxpayers to lend to the Greeks so that they could reimburse the German banks. Hence there emerged this word ‘debt’ [Schuld] which in German means ‘fault’, ‘sin’… and ‘debt’.

 JE: The problem is all the more impossible to solve because devaluation is ruled out in advance…

YV: … of course. Even the American libertarians agree with me on that point. Faced with a crisis of this type, the liberal solution is: devalue. But Greece does not have its own currency, so, yes, you can devalue – and that’s what we did: incomes fell by 46 percent and we lost close to 25 percent of our workforce to unemployment and mass emigration – but if your currency belongs to a foreign bank and if, at the same time, your debts are in that same currency, the process of internal devaluation is no help! Your incomes retreat, but not your debts! 

JE: Isn’t the reason why the consequences of this series of errors hit Greece also to do with the Greeks behaving irresponsibly? Didn’t the elites maintain the illusion of an artificial growth? 

YV: They truly believed that ‘private vices make public virtue’, to quote Mandeville’s Fable of the Bees [of 1714]. Rather than tax capital and manage the conflict between labour and capital, they thought that it was sufficient to let finance off the leash.

JE: I would like to ask you about game theory, which you are a specialist in. It is a formidable toolbox for understanding the structure of social interaction, but a posteriori, once action has already taken place. But to me it seems dangerous to use it to predict how actors will behave. 

YV: Your work of critique has really counted a lot for me. After all, game theory reveals the limits to the rationality that economists ascribe to economic actors. We are social beings: so-called second degree beliefs – taking into account the other person’s view of me – are of decisive importance. From the moment that you start developing your thinking on what your interlocutor is thinking, things can get carried away: you try to divine the beliefs that the other person is forming with regard to how you are going to act – Greece tries to divine what Germany thinks that Greece is going to do – but still further, the others’ moral beliefs on what you have to do – Greece tries to divine the Europeans’ moral expectations regarding the debt. Once you’re considering these second-degree beliefs, you can come to depend on this game of mutual anticipations. 

JE: I don’t think that we ought to push the calculation of anticipations and our reflections on belief too far, since this can paralyse action. If we had to refer to one of the scenarios that game theory has formulated, which one would best apply to Greece? Some have mentioned the case of the ‘prisoner’s dilemma’, and other people the ‘chicken’ game [see box below].

YV: These are scenarios of non-cooperation, whereas Europe is meant to function on the basis of cooperation. In a cooperative game we have the possibility of reaching binding agreements, of speaking freely. In a non-cooperative game like the prisoner’s dilemma, we do not speak to each other: we can only choose between cooperation and defection. In Europe, I try to convince my interlocutors that we are in a cooperative game and not a closed one. We no longer have any right to bluff. When I say that we will end up leaving the euro if we have to accept new austerity measures, that is no bluff. 

JE: What do you think of the comparison between the negotiations on the debt and the so-called ‘chicken’ game, which is essentially about bluffing?

YV: In this scenario two motorists – let’s say the Troika and my government – are heading towards each other. The first to deviate does benefit – he avoids the accident – but he wins less than he who had the courage to take on the risk of an accident. For us, deviating would mean accepting new austerity measures, humiliating ourselves by throwing away our anti-austerity programme. For the Troika, deviating would mean letting us get away without imposing more austerity. I really like to cite Jean Monnet. According to him, if we sit across from one another at a table, staring at the whites of each others’ eyes and reciting already-prepared arguments, the chance of reaching an agreement is a thin one indeed. Whereas if we sit at the same side of the table and place the problem on the other side, the chances of reaching an agreement are greater. In the game, the players have one and the same motive: but it is wholly in the interest of democracy and the union among democracies that motives are malleable. Reason and argument have to be able to shape people’s motives in order that they can align, rather than them being treated as if they were immovable. That is why I prefer to leave game theory to one side.

JE: There are two typical obstacles to the application of game theory. Indeterminacy and irrationality. Now it seems to me that actors are less rational than they would like to admit. 

YV: I would add that there could even be an interest in feigning irrationality. If an actor pretends to be irrational for long enough, in a convincing way, there is a good chance that the other person will see him as irrational. So we could say that it is not irrational to pretend to be irrational…

JE: With Europe, I get the impression that everyone is continuing to bluff a bit. Or, let’s say, that each side is acting in bad faith… the Europeans keep telling the Greeks that they have to stick to austerity, even though they know that it does not work. And the Greeks are acting as if they hardly have any time left. So they are lying to themselves. 

YV: Europeans don’t all think the same. Most have come to understand that austerity is a medicine worse than the illness itself, and that what has happened to Greece risks happening to them too. And they no longer have confidence in their governments. The Greeks are not lying to themselves, they are saying: ‘we will try to pay back as much as possible, but in exchange for that we demand a restructuring of the debt’.

JE: If we move on to the terrain of solutions, I believe the best way out would be for Greece to agree to default for five years, the time needed to mount some structural reforms, like the deregulation of part of the labour market. 

YV: I prefer to speak of restructuring rather than wiping out the debt. That is something that banks do all day long. Historically capitalism would not have experienced such rapid growth without limited liability companies. If your business goes bankrupt, you pay what you can out of its funds, but your children still have a roof over their heads.

JE: I have a Chinese friend, who is now a renowned professor in his country, who has written a dissertation for the University of Chicago where he explains that capitalism like socialism is founded on soft budgetary constraints, first among them the principle of bankruptcy…

YV: Take the two versions of Faust, Marlowe’s and Goethe’s. What’s the difference between the two? In Goethe’s, Faust’s debt [to Mephistopheles] can be cancelled. What happened in between the two? Protestantism, which made capitalism and the erasure of debts possible. Even if the Germans don’t like being reminded of it, their economic miracle would never have taken place if in 1953 a benevolent ‘dictator’ called the United States hadn’t wiped off half of their debt! Were the Americans being kind? No, they understood that for their postwar project to work there had to be a pillar of industry at the centre of Europe. 

JE: But is wiping it away sufficient? 

YV: No, we need two things: erase the debts that we can’t repay and find a way to attract investment, like Roosevelt did with the New Deal. I have officially called for the implementation of a plan of this kind at the European level, which could be called the ‘Merkel plan’. 

JE: Some think it’s necessary to go further and envisage Greece leaving the Eurozone. But the effect of a ‘Grexit’ is wholly uncertain. So the question is: how do you decide in a situation of uncertainty? I think that I am incapable of responding. But I believe that this does have to be addressed. 

YV: My interlocutors tell me: Europe has done its introspection, the Greek problem has been circumscribed, so we will avoid contagion; we can even envisage it leaving the euro without fearing for the system itself. My response is a dual one: firstly, that this means that the purpose of the salvation package was indeed to save the French and German banks, and not Greece. Now that they have been bailed out they can kick us out of the system. So ‘Grexit’ is a possibility that’s very much still to be determined. As Hegel said, the laws of indeterminacy are themselves indeterminate. Whoever claims to know what would happen if one day we were pushed off the cliff edge is talking nonsense, and working against Europe. 

JE: What is difficult, politically, is that the solution does not seem like an inevitable one, but one that has been chosen. Europe tends to consider that it has the right to exercise a benevolent dictatorship over Greece.

YV: That would be fine if it were truly benevolent. The problem is that the austerity medicine is not therapeutic but toxic. Beyond the economic lesson, there is a political lesson to draw from this failure. From the instant that the dictatorship emerges – be that the dictatorship of the Troika or the dictatorship of the proletariat – you find yourselves in a situation where the (dictatorial) means affect the (eventually beneficial) end that you are pursuing.

JE: Let’s return to these beneficial therapies. Some object that while the so-called ‘haircut’ solution – wiping away 50 percent of the Greek debt – would be an effective move, they oppose it on account of the ‘moral hazard’ that it would entail. In taking responsibility for the damage that someone or some institution has caused, you encourage everyone else to behave in the same risky manner, feeling that they have their backs covered. Like a motorist being encouraged to take risks because his insurance covers whatever damage he might cause. In this case, if Greece benefited from a debt write-off, this would incite other states not to balance their books in a virtuous manner. That said, I think that it is necessary to run this risk. 

YV: It is not a question of giving the Greek state a deal that would be refused to anyone else, but of proposing a structural solution that would be to everyone’s benefit. In 1993 the European Commission president of the day, Jacques Delors, was conscious of the difficulty of a monetary union without a political union. He drew up a white paper proposing the creation of Union Bonds. One of my friends was present when Delors submitted this proposal to Mitterrand. After listening to him Mitterrand told him, ‘Jacques, I am in agreement with your analysis and the necessity of your remedy, but Helmut [Kohl] and I do not have enough political authority to go beyond monetary union. But when, in ten or twenty years time, there is a very deep crisis, the success of our enterprise will depend on your remedy – or else the whole European enterprise will collapse’. And here we are.

JE: There is another actor who we still haven’t mentioned: the Greek people. Now, the Greek government is faced with a dilemma. The question is whether there is some possible level of concessions that is sufficiently high to satisfy the European Union and low enough to satisfy the Greek electorate. If that is not possible, the logical outcome would be new elections and perhaps the victory of the extreme Right. The only hope is that there can be an opening to reforms capable of satisfying these two constraints.   

YV: We are not interested in making agreements with our European partners just for the sake of increasing our chances of being re-elected. Our red line is clear: we want the agreement, if there is one, to satisfy three criteria. First, stop the cycle of deflation – to put it another way, stop putting more people on the streets. Next, restore our dignity – that is, fulfil our reform projects without us being invaded by an army of technocrats. And finally, put an end to the ‘kleptocracy’ that dominates Greece. We can make major compromises, but not to the point of renouncing these three necessities. 

JE: As well as the ‘kleptocracy’ there is the corporatist structure of numerous Greek professions, from pharmacists to solicitors and tax drivers. Shouldn’t you be fighting against this corporatism? 

YV: Rather than target the little people, I would prefer to target the oil cartels or the TV channels that don’t pay a penny to the Greek state to use the national frequencies. If you lived here, Jon, you would not be pointing the finger at the pharmacists or the taxi drivers. 

JE: And what about the cliché that Greece is incapable of collecting its taxes? Do you think that there need to be fiscal or political reforms?

YV: Absolutely. We have a vast programme of reforms and we would really like to be able to implement it. And believe me, for two months I’ve been working on that day and night.

JE: To conclude, I’d like to tell you what a pleasure it’s been to talk to you, Finance Minister: I had feared that I would be treading on eggshells, but it was a real pleasure. All my career, I’ve had the impression that I am a semi-professional in what I’ve achieved. And this evening I’m speaking to the Greek Finance Minister, a professional… even if once again, it’s as a semi-professional. Doubtless that’s my fate…

YV: Except that you are speaking to someone who holds you in even greater esteem than you do him. In the preface to one of your books you write that you are standing ‘at the intersection of a number of faultlines’. It’s a formula that I’ve often used in my speeches.

JE: I’ll gladly let you have that: it perfectly suits your situation.


Game theory according to Jon Elster

Game theory is a way of modelling social interaction, designed to understand how in a given situation an agent’s optimal action depends on his anticipations of someone else’s behaviour. There are four main scenarios:

1.  Battle of the sexes

A man and a woman want to go out together, but they’ve forgotten what activity they were going to do together, and they cannot communicate. The man prefers football, the woman the opera. What should they do? Either each chooses their own preferred activity and they will not be together (0/0) or one chooses the other’s activity so that they are together (1/2 or 2/1) but without maximising her/his own preferences. 

2. The prisoner’s dilemma

Two prisoners who cannot communicate have to pick between three choices: A denounces B (and A is freed while B gets a ten-year sentence); A and B each denounce each other (and get five years each); A and B refuse to denounce each other (and each get five months). So while they have an interest in cooperating, each will tend to betray the other for fear of being betrayed. This is the only game where there is only one balance, that of the double defection (1/1).

3. The chicken game

Two cars are driving toward each other on a one-lane road. The one who deviates avoids the accident but gains less than the one who took the risk and is saved by the other (3/1 and 1/3). 

4. The assurance game

Two states are thinking of vaccinating their citizens. If one country doesn’t do so, then it is better that the other doesn’t either, given the costs of vaccination. So are the two states encouraged to vaccinate their populations?

Normal-form representations

These represent the gains that each makes in function of his/her choices and the choices that the other makes. The higher the figure, the more preferable this agent’s situation. A balanced pair (the numbers in blue) is made up of the optimal pair of choices. A great teaching of game theory is that the optimal outcome for two people (and thus for the collectivity) often does not coincide with what is rational for the single individual. Another is that in a game with many different balances, the agents’ rational choice can be indeterminate. 

C: cooperation/D: defection

In each case the first number is the gain for the agent choosing vertically, and the second the gain for the one choosing horizontally. So in the prisoner’s dilemma it is in the interest of each agent to choose D, whatever the other does. In the assurance game, it is in the interest of each of them to choose C. In the coward game it’s in each of their interests to choose C, if and only if the other does not do so. In the battle of the sexes there can be several cooperative strategies.

- Jon Elster is a contributor to The Politics of Human Rights and his work is addressed in Marcus RobertsAnalytical Marxism: A Critique

- More on Greece and the eurozone crisis here.