When Getting it Wrong is Getting it Right
The world‚Äôs oldest democracies have fallen victim to a basic philosophical mistake. Without realising it, we have embraced the self-centred idea that ours is perpetually the most likely of possible worlds. That belief may seem innocuous, but it is badly distorting the way democracies make policy and evaluate their leaders.
In the United States, for example, lawmakers who opposed President Barack Obama‚Äôs stimulus plan hope to take advantage of this logical lapse. They know that if the plan succeeds, Obama will get the credit, whether Republicans supported him or not. But if the plan fails‚Äîas even its advocates concede it may‚Äîthe Republicans will be able to run in 2010 or 2012 on a platform of ‚ÄúWe told you so‚Äù. In strategising this way, the Republicans are counting on the American public to make a critical assumption: if the stimulus plan fails, then the sceptics must have been right about it.
On first thought, what could be more natural? If the stimulus plan fails to accomplish its goals‚Äîdespite costing nearly $1 trillion‚Äîthen almost by definition, it was a mistake. On second thought, however, this assumption is clearly false. You would only think the wisdom of the plan and its success automatically converge if you‚Äôve excised chance from the universe‚Äîthat is, if you believe that the most likely thing always happens to us.
Suppose, for example, that the stimulus has a 70% chance of success, as Vice President Biden apparently estimated. Then opposing the stimulus plan, without a superior alternative, would be a clear case of bad judgment. If America has the misfortune to inhabit one of the possible futures where the stimulus fails, the plan‚Äôs opponents may appear prescient, but their votes will remain as unwise as the day they were cast.
Still, many would take the failure of the President‚Äôs plan as a sign that the critics must have been right all along. Call this the ‚ÄúMost Likely World‚Äù assumption, since it rests on the belief that whatever happens was the most likely to happen. This basic mistake is remarkably widespread; the stimulus debate in America is but one example.
Of course, there is something to the assumption that undergirds this mistake. That is why it is so seductive. The most likely thing does not always happen, but it usually does‚Äîthat, after all, is what it means to be likely.
The trouble with this rule of thumb is clearest when we apply it to multiple judgments at once. Take two examples from the headlines in the United Kingdom. Prime Minister Gordon Brown might display sound judgment by supporting two unrelated courses of action‚Äîsay, a costly effort to rescue the nation‚Äôs banking system, and a proposal to create a permanent standing force for NATO in Europe‚Äîbecause in each case, the plan has a 70% chance of accomplishing its objectives. Despite the prudence of these decisions, it will actually be as likely as not that at least one of the plans will fail to achieve its goals.
If one of the two initiatives does fail, then, it would be a serious mistake to infer that Brown displayed bad judgment, or that the sceptics have earned a chance at the helm. But this is precisely what the Most Likely World assumption would lead us to conclude. What‚Äôs more, our governments are liable not for two high-stakes gambles, but for dozens.
In fact, as long as we continue to make this basic mistake, we are creating a perverse incentive for the minority party to avoid cooperating with the government‚Äîeven when they ought to agree. Because few of today‚Äôs problems come with surefire solutions, the opposition can simply fight the government on each difficult decision, and thereby guarantee itself a significant number of ‚Äútold you so‚Äù moments over time.
To remove this motive for insincere opposition, we must come to grasp that the sceptics might have been wrong to oppose a plan even if it didn‚Äôt work. But it is equally important for us to realise that the sceptics might have been right to oppose a plan, even if it did work.
Nowhere was this latter confusion clearer than in last summer‚Äôs tortured debate over the success of the American troop surge in Iraq. When George W. Bush called for a surge in Iraq at the end of 2006, Barack Obama vigorously opposed the proposal‚Äîpredicting that it would not subdue sectarian violence, but rather would ‚Äúdo the reverse‚Äù. By May 2008, however, both civilian deaths and US casualties were down 70% since the surge began. So in a series of interviews and debates, the media confronted Obama with what seemed like a natural question: ‚ÄúWas your judgment about the surge wrong, or is the surge not succeeding after all?‚Äù
The national conversation was doomed from the outset, however, because this question glaringly presupposes the Most Likely World assumption. In fact, Obama may have been right that the surge was unlikely to reduce violence; perhaps its partial success was a stroke of good luck. For example, many experts believe that the surge could only have gone as well as it did in concert with other events that were themselves unlikely or unpredictable, including the Sunni Awakening and the tragically effective Shiite ethnic cleansing campaign in Baghdad.
But because the Most Likely World assumption is ingrained in modern political culture, it was nearly impossible for Obama to concede that his fears about the surge had not been realised without appearing to admit that he had erred in opposing it. When he did acknowledge that ‚Äúthe surge has succeeded in ways that nobody anticipated‚Äù, he was accused of making excuses rather than owning up to his imagined error. One interviewer‚Äôs response: ‚ÄúRight! So why can’t you just say ‚Ä¶ ‚ÄòI was wrong about the surge?‚Äô‚Äù
It is urgent that the world‚Äôs democracies develop civic discourses sophisticated enough to reject the Most Likely World assumption‚Äîand thereby to recognise that Obama might not have been in error at all. Otherwise we will continue to encourage sophistry in our politics, and we will regularly punish our leaders for making the right decisions, while rewarding them for making the wrong ones.
Of course, sloppy decision-making is a familiar feature of democratic political systems. People fall short of perfect rationality even when they try their hardest‚Äîand few citizens take their democratic responsibilities that seriously. In a society where elections are often influenced by the physical attractiveness of the candidates, the Most Likely World assumption is perhaps the least of one‚Äôs problems.
This error merits special attention, however, because it may be especially susceptible to correction. Unlike many of our irrational tendencies, the Most Likely World assumption is a false premise that we can become consciously aware of, and thereby work to avoid. Perhaps civics courses in our secondary schools could spend less time on names and dates, and more time educating future voters to analyse political debates that appeal to past outcomes clearly.
We cannot know when the next likely success will fail, or when the next likely failure will succeed. But we can be certain that these dynamics will recur sooner or later. To respond appropriately, we must abandon the smug assurance that by making the right decisions we can guarantee good results, and that good results are a sure sign of good decisions. That will require jettisoning the implausible view of probability on which these assumptions rely.
Ben Eidelson is reading for a BPhil in Philosophy at Oriel College, Oxford.