Why Social Scientists need to learn math

Obvious statement: math is useful for a lot of things.

Yesterday morning I had an infamous lecture of Public Policy. It's interesting in the sense that you can predict the kind of socialist ideas that'll come out of it;

1. Thatcher was bad, 
2. Critique of Neoliberalism, 
3. The horrible effects of market economy

Yesterday was a special day in that the third was arguably not even included. However, the Professor did give a good start, critizising neoliberalism ideology after a mere 9 minutes of speaking. Even better, I think he set some kind of record; 4 minutes into the class, his first severe charge against Thatcher was laid. Impressive.

Not to say that this example means anything in particular, but I suppose it's fair to assume that Public Policy at University of Glasgow foster a rather strong hatred towards anything non-socalist, especially if it can be connected to Thatcher & Co.

Enough, what I wanted to convey today was the surprising lack of math insights and how that cripples the social scientists. This was his example:

Amount of UK Labour force in Manufacturing employment:
1978      28,5%
2009      10%

Conclusion; only a third of the people who worked in manufacturing 30 years ago, remain in that industry today.

Wait. Read that again.

Ok, well, no. First of all, people change jobs. They can change occupation/exit or enter industry, especially over a 30 year period. Second, a fair share of the people employed in 1978 will have moved on to enjoying a life of retirement by now. But 10/28,5 = 0,35, so the math still works out, doesn't it?

Math issue: between 1978 and 2009 the population (and the distribution of people in working age) have changed significantly. Thus, if inputs in a %-figure changes, the output changes. That is, 10% of the Labour force in 2009 is most likely not 1/3 of 28,5% of the Labour force in 1978.

Additionally, I've even stopped being suprised of the cases where people misuse Percentage for Percentage Point, ending up with completely flawed conclusions. My point here is that social scientists tend to have a limited understanding of math and what affects the numbers (especially regarding %). And, as they use statistics and math in describing their field, they inevitable end up with flawed or unsupported conclusions. That's a serious issue, heavily diminish the value of their study, should this be a theme for a larger part of social scientists. Let's hope it isn't.

If you're to use math or statistics, you need to understand what comprice and influence them. If you don't, you'll end up misinterpret data. Some kind of "know-thy-limitations" argument, I guess.

____

This lecturer also gave me some idea to why there are so many socialists in University environment - and what happens to them afterwards. Well, in his case, some of them seem to remain at Uni, teaching Public Policy (or some other fluffy subject like sociology).