What if? It’s one of the most powerful questions we ask ourselves about the past – and the future. It’s a common tool used by armchair historians, and actual ones too. In this video, I briefly discuss counterfactual analysis, a qualitative method that uses what if questions to draw conclusions about cause and effect.
First, what is a counterfactual?
At the most basic level, it is a what if statement. What if some historical event happened differently or didn’t happen? What if we implemented a different policy in the past or from here forward?
Just FYI, in addition, statistical analysis, we call the null hypothesis – the alternate explanation to what we think is true – a counterfactual. We can also generate counterfactuals (for example, the implications of different vaccine distribution policies) through computer simulations when we have a lot of data.
In qualitative analysis, though, we use these what if questions to talk about cause and effect by demonstrating that later events would or would not have happened if whatever you think the cause of an event changed.
We do this all the time. What if Hitler were never born? What if Gorbachev didn’t enact perestroika? Or, more recently, what could we have done to prevent 9/11?
But not all counterfactuals are created equal. To make this a useful tool for historical or policy analysis, it is best to stick to three principles when creating a counterfactual.
First, clarity. As with all research, you need to be very clear about what you think is a cause and what is an effect – what are the variables you are actually looking at. You also need to be really specific about what would need to change to make your counterfactual true, as opposed to what actually happened in history.
Second, your argument needs to have logical consistency. For example, you should not assume that two things that go together don’t happen together in your timeline. For example, if you wanted to think about what the impact of the development of railroads had on industrialization, you can’t assume that we would have just developed cars instead, when the technology for both is very similar.
Last, you need to maintain as much historical consistency as possible. This is called the “minimal rewrite rule” – try to just change one thing. On this, one example of a bad counterfactual is, What if Nixon were president during Cuban Missile Crisis? It’s a bad counterfactual because you have to assume a lot – that Nixon was elected, that he would have acted differently about the Bay of Pigs invasion, etc.
A good example of a counterfactual is, what would have happened if Al Gore had won the 2000 presidential election? Specifically, would the United States have gone to war in Afghanistan and Iraq. This is a good counterfactual because the results were really close – they ended up coming down to a recount of a small number of contested ballots in Florida. Gore could have won if Palm Beach County used a different type of ballot, or observers counted “hanging chads” differently, and so on.
What you then do is perform a thought experiment on necessary causes – if one cause were absent, what else would have been absent. (If no X, then no Y). We use the Bush-Gore counterfactual to talk about the role of leaders in U.S. foreign policy. If Bush were not president, 9/11 would almost certainly have happened and Gore would have gone to war in Afghanistan – it was a war that had wide international support and was considered to be in the U.S. national interest. But Gore probably would not have gone to war in Iraq – that was driven in part by Bush’s personal and ideological goals.
So that is counterfactuals – a simple concept, but often difficult to actually do well.