Let's not make a deal

Franklin D Roosevelt and Winston Churchill viewed making a deal with Stalin as an acceptable compromise in the fight against the greater evil of the Third Reich.

In a thought-provoking and sometimes troubling book about the morality of political compromise, the philosopher Avishai Margalit excavates the past but says little about the present. On Compromise and Rotten Compromises Avishai Margalit Princeton University Press Dh100 The idea that the Second World War was the definitive "good" war, in which radical evil was confronted and vanquished, has always run up against the problem of what happened on the Eastern Front. As the British historian Max Hastings has put it, the story of how Stalin's Red Army defeated Hitler's Third Reich is "not for anyone with a weak stomach". Stalin was utterly ruthless in his disregard for the lives of his own subjects, which he tossed away by the million. The only people who fared worse were his enemies (which included many of his own subjects) against whom he unleashed campaigns of unimaginable vindictiveness. His was a truly horrible regime. So what does it say about the moral integrity of the western democracies that they were only able to defeat someone as unspeakable as Hitler by throwing in their lot with someone as vile as Stalin?

This is the question that underpins Avishai Margalit's important and troubling new book about the nature of political compromise. Margalit keeps coming back to the great laboratory of wickedness that was the Second World War, which he describes as being to morality "what the supercollider is to physics: extreme moral experiences and observations emerged out of the high energy clashes". He thinks we need to have an answer to the question of why it is acceptable to choose Stalin over Hitler - or, as he puts it, why Munich was a "rotten" compromise, but siding with Stalin was a necessary one. The answer he provides is unashamedly grounded in morality. He believes that it is a mistake to try to distinguish between these two regimes in terms of how evil they were in degree (this invariably leads to the futile and miserable business of counting up their dead). Instead, he argues that Stalin's evil was of a different kind from Hitler's. The reason it is never acceptable to compromise with someone like Hitler is because Nazism negated the very idea of morality, by repudiating the notion of a shared humanity. The Nazis wanted to dismiss great swathes of the human race from moral consideration altogether. Stalin, by contrast, believed in a shared human future, even if his route for getting there was monstrously callous. So any compromise with Hitler is a rotten compromise, because it contaminates everything it touches. Getting into bed with Stalin, for all the squeamishness it provokes, belongs to a world in which morality at least remains a possibility.

This is an admirably forthright answer, but it is fraught with difficulties. It places a great deal of weight on good intentions rather than outcomes, which is always dangerous in that it allows even the nastiest political operators to say that underneath it all they meant well. The idea that at bottom Stalin meant well is more likely to make a mockery of morality than to leave the door open for its resuscitation. There is also the question of how long we should be prepared to wait. Margalit distinguishes between an ideology like Hitler's that destroys some races for the sake of a master race, and an ideology like Stalin's that destroys the present generation for the sake of a future one. The implication is that Stalinism at least leaves room for hope, whereas Nazism is the end of all hope. But waiting for a future that never arrives may simply be prolonging the agony. Hope can be deeply counterproductive if it serves to delay the moment when we say enough is enough.

Yet Margalit is surely right when he says we need some answer to the question, why Stalin but not Hitler. He quotes approvingly Winston Churchill's line: "If Hitler invaded Hell, I would make at least a favourable reference to the Devil in the House of Commons." In a subsequent radio broadcast, Churchill went on to say of the Nazi regime that it was "devoid of all theme and principle except appetite and racial domination". This suggests a different way of looking at the problem. The reason it may be impossible to compromise with someone like Hitler is that Hitler was so uncompromising himself - he was devoid of any principle that might hold him back, willing only and always to act in accordance with his own will. How can anyone do business with someone like that? Stalin, by contrast, was a more compromised figure, just as cruel, just as wishful, but constrained by ideology rather than being propelled by it. Margalit makes this point when he says that Stalin's saving grace was his obvious hypocrisy, which proved that Communism wasn't simply nihilism (a hypocrite has to have principles to betray). Hitler wasn't a hypocrite, which is what made him such a total nightmare. Margalit cites a remarkable survey from January 1939, in which the American public were asked who they would prefer to win if it ever came to a war between Nazi Germany and the Soviet Union. By an overwhelming majority (83 per cent to 17 per cent), they sided with Stalin over Hitler. It seems unlikely, in a country in which anti-Semitism was almost certainly more widespread than pro-communism, that this was a moral judgement. It was prudential one. A victorious Hitler was such a dangerous prospect for the United States because he would have been utterly impossible to deal with.

Nonetheless, Margalit is suspicious of this kind of prudential reasoning, because he thinks it risks pragmatic justifications for rotten compromises. Though he accepts the West had no choice but to side with Stalin against Hitler, he still thinks some of the deals that were done to cement this alliance were rotten, in particular the Yalta agreement of February 1945, which allowed the forced repatriation of Soviet citizens, most of whom faced a dreadful fate on their return. He writes of Anthony Eden, who had opposed the Munich agreement because he did not trust anything Hitler said, that at Yalta "Eden believed he could deal with Stalin, not because Stalin was moral, but because he was prudent". But this, for Margalit, was a kind of self-deception, and he thinks it would have been better trying to influence Stalin by refusing to compromise with him than by playing him at his own game. The fact that Stalin was more tractable than Hitler doesn't mean that he was in any sense to be relied on. Sometimes a compromise that is intended to draw a bad regime into the world of negotiated agreement simply ends up corrupting everyone involved.

What this shows, though, is that the question of when to compromise is not simply a moral one. It always turns on judgements about who is really compromising, and who is simply deceiving themselves. That is certainly true for the other main category of political compromise that Margalit discusses - deals over land for the sake of peace. Here it is clear that the template is the Israeli-Palestinian conflict, though Margalit is careful to abstract away from the details. His main concern is to show that sectarianism, revanchism and ideological purity are the death of compromise, but also that compromise can be the death of this kind of inward-looking inflexibility. Simply by recognising the possibility of talking with the other side we take an important step "in humanizing the enemy and in acknowledging the enemy as holding legitimate concerns". But again, it rather depends. Convinced and thoroughly inward-looking ideologues may persuade themselves that it's possible to do a deal precisely because they think their core purpose will be untouched by it: the other side might be compromising, they will think, but we are just furthering our cause. This kind of self-justification can exist on the left and on the right. Early 20th-century revolutionary socialists compromised by taking part in democratic elections only because they believed the revolution was coming anyway. Twenty-first century neo-conservatives have acquiesced in torture only because they think democracy is bound to triumph in the end. (It is probably no coincidence that many neo-cons began life as Trotskyites.)

Margalit's short book touches on a remarkably wide range of complex moral and political questions, from slavery (he thinks the accommodation with slavery in the US constitution made it a rotten compromise) through to the use of nuclear weapons (any thought of compromise here is inherently dangerous). But perhaps surprisingly he doesn't say much about the most obvious contemporary instance of the central problem he discusses: terrorism. Should we ever be willing to compromise with terrorists? Should the West be prepared to do a deal with the Taliban, or with al Qa'eda? One of the distinctions Margalit draws between Stalin's and Hitler's regimes concerns terror: Stalin's rule, he argues, entirely depended on it, as shown by his readiness to terrorise anyone, including party loyalists; Hitler, by contrast, was careful not to treat everyone the same, and loyal Germans got off relatively lightly. Again, for Margalit, this makes Hitler worse, or at least different in kind, because he deliberately discriminated in his use of terror, sparing some categories of people and so casting his victims utterly adrift. One of the problems with this analysis is that it risks turning Stalin into the punchline of a tasteless old joke - you couldn't call him prejudiced, since he was willing to kill absolutely anyone. But it does also raise some unusual questions about contemporary terrorism.

Does the indiscriminate nature of so much terrorist activity - evidenced by a willingness to target almost anyone - mean that on Margalit's account the terrorists are not to be placed with Hitler, beyond the pale of political compromise, but closer to Stalin, somewhere at its edges? Certainly, Margalit intensely dislikes blithe Hitler analogies of the kind that have peppered the war on terror. Part of the point of his book is to get us to stop thinking every wicked leader is another Hitler, and to ask whether they might not be another Stalin instead. Though he doesn't say it, Saddam Hussein would clearly fit his Stalin model: not someone to cosy up to, but also not someone it would be impossible to live with (so the US got it wrong on both counts, first by selling him weapons, which showed too much readiness to compromise, and then by getting rid of him, which showed far too little readiness to compromise). But bin Laden? At this point, it is hard to see how Margalit's analysis is going to be of much help. If we ever reach the point of seeking an accommodation with al Qa'eda, or any other group of terrorists, it will not be because we have reached a fine moral judgement about the nature of their particular brand of cruelty. It will be a political judgement, and it will rest on a prudential calculation about the least worst option, with all the hazards of self-deception that go along with that. Margalit would like to help us to know in advance when political compromise is an option. But we can't always know in advance. His book isn't able to answer that question. What it can, and does do, though, is make us think.

David Runciman teaches politics at the University of Cambridge. His most recent book is Political Hypocrisy: The Mask of Power, from Hobbes to Orwell.