Wednesday 3 September 2014

Counterfactual Conditionals and the Fundamental Flaw in an Otherwise Powerful Argument

Sexy title, huh?  I bet you can't get enough counterfactual conditionals in your blog posts!

I want to explain why an argument - which looks at first sight enormously powerful - is actually fundamentally flawed.  The argument is one used by George Monbiot in this very powerful piece which argues in favour of independence.  I've always thought that the central argument here is a powerful one: "If it was the other way round, and Scotland was independent, would it vote to join the UK?"  It's a powerful argument because it looks like the answer is obviously no, but it ain't so obvious.  In fact it's so fundamentally flawed, all we can say is "well, nobody knows one way or the other."

To see why it's so flawed, we'll need a bit (quite a lot) of history and a bit (much less) about counterfactual conditionals.   We'll do that second bit first.

Looks fun, huh? 
I have a little bit of a history with counterfactuals.  My late father taught me what they were: they are an if-then statement where the if-part is not true (counter-factual) and so expresses what might have been or what could be.  E.g. "If my father was alive today he would be wearing a monocle".  My father also taught me that the truth of counterfactuals is essentially unknowable: since we don't live in a world where my father is alive, how do we know he would wear a monocle - even though he did regularly when alive.

Yes, my father he was a monocle-wearing logic-imparting kind of father and he was exactly as awesome as he sounds.  My favourite quote from him: "When you've discovered the furthest known object in the universe, the rest of your career is a a bit of an anticlimax".  It was QSO B0642+449, since you ask.

Years later - and now years ago, my first academic paper was about the logic of counterfactual conditionals. The picture shows an excerpt from it to add to your life's store of entertainment.

Now onto the history.

Monbiot writes as if no state has ever been crazy enough to form a union with another state.

Absolutely states have formed unions in their best interests. We are not talking about minor little countries here and there. We're talking about the USA, Germany, and Italy. The idea of German and Italian nations existed for centuries - in fact millennia - before the unions that formed the countries were formed. (In the 'Social Wars' of about 90BC, the rebellious Roman Allies hopefully called their capital "Italia", but no such country existed as a state until 1861).  The USA did not exist as an idea for centuries, but was formed as a federation of thirteen separate colonies in 1776.  Thirteen colonies fighting for independence and forming a union to help them be the greatest they could be.

Oh but obviously that is all ancient history, right?  Well, if you regard 1990 as ancient history when the two Germanies unified.

The example of East and West Germany illustrates the other problem and the fundamental flaw with Monbiot's argument.

The argument is flawed because it has a hidden false (well actually unknowable) premise. The argument is actually:

If Scotland was independent, and was basically exactly the same as it is now, would it vote to join the UK?

If Scotland was independent, it would not be the same as it is now.

What would it be like? Obviously nobody knows: that's the nature of counterfactual conditionals. But the case is strong that Scotland would be exactly the kind of place - like East Germany in 1990 - that would love to join its larger neighbour.

If you are interested only in Scotland's economy, it obviously benefitted massively - I mean absolutely ludicrously massively - from being in a Union with England and Wales. Economically, being in one country where the industrial revolution was taking place was exactly where you wanted to be. And then as that finished exploding, it wasn't half bad for your economy being in the same country as had a massive empire with things like India as parts of it. Scots were pretty good at empire too: for example it was basically the Scots who colonised Canada (I'm sitting writing this in Cupar, Fife: the only other Cupar in the world is in Canada.)

Without the Union, it's clear that Scotland could have been massively poorer than it was at least until oil kicked in about 30 years ago.  Of course the rest of the UK would have been poorer too. Everybody would have lost.

What about since oil? Well, if Scotland had magically gone independent on the right day, then fair enough, it would have made a lot more money from oil than it did. Also we wouldn't have had Thatcher. But don't fantasise about a beautiful liberal social democracy with a Norway style oil fund - or if you like fantasise about it, because it would have been a fantasy. At exactly the same time - early 80s - was the peak of the hard-left labour party. Hard left labour have been in total control in Scotland and would absolutely not have been interested in investing billions in the stock markets. Certainly there would have been no Thatcherism in Scotland, and while that would have been much better in many ways, not necessarily in all of them. How much of Scotland's oil money would have been spent on loss making steel, coal, and ship production? Yes, people in those industries would not have lost their jobs. But would the economy be better than it is now?

So basically, my version of the counterfactual history is that Scotland would now be a deeply impoverished country if it had never been in a union, or at best (independence in about 1980) throwing its oil money away on deeply uncompetitive industries.

So my counterfactual conditional I offer in response to Monbiot is:

If Scotland was independent, we'd have grabbed at union as eagerly as East Germans did in 1990.

Yes, my counterfactual is as deeply and fundamentally flawed as Monbiot's.  But now I hope you can see why his piece - while undeniably powerful -  is a powerful piece built on foundations of air.

I leave you with a joke, and I think it may be the best joke I know in an academic paper.  In his paper on counterfactuals, Matt Ginsberg says that counterfactuals are sometimes used to indicate precisely no linkage between the if and the then. He gives an example:
"Even if I was free for dinner tonight, I still wouldn't go out with you."
and then he says
I am indebted to a former Miss Texas for this example. 


  1. I am surprised that you left this in: "How much of Scotland's oil money would have been spent on loss making steel, coal, and ship production?" Look at South Korea & POSCO as an example of a state "spending money on loss making steel". This lead to Hyundai Heavy Industries, South Korean again, the largest ship builder in the world. In an alternative counterfactual argument Scotland would have invested its oil revenue in the same way as South Korea and had its own POSCO, Hyundai and LG. These are all state funded enterprises. Ian, I think you have been economical with your counterfactuals.

    1. Certainly I can't rule that out. My main point though is not that my counterfactual version of history is correct. It's that Monbiot has a hidden assumption in his - that everything would be the same - which is not justifiable.

  2. ... and the joke reminds me of another: Jean-Paul Sartre is sitting at a French cafe, revising his draft of Being and Nothingness. He says to the waitress, "I'd like a cup of coffee, please, with no cream." The waitress replies, "I'm sorry, Monsieur, but we're out of cream. How about with no milk?"

    1. That's another good one :)

    2. Another hugely successful English/Scottish partnership also had fun with this theme...

  3. This is fascinating stuff. I'm afraid that honesty, logic and rational thought is much missing in the debate. Or it is from my perspective in North Wales. Which make my views irrelevant even though the result could have a massive impact on my life.