I am virtually certain it would go to this piece by Phillip O’Connor from the University of Auckland. I mean, I thought this is the sort of award that would always go to NZPA, as they have to quickly release something that sometimes misses the point. However, a Senior Lecturer from Auckland has managed to illustrate to me how bad analysis can be.
Now in case other people can’t see some of the issues with his “analysis” I will discuss some things under the flap. And to be honest, I’m writing this in sheer shock – I have never met an economics lecture this off track. Maybe it is because I went to Victoria University – where the lecturing is top rate 😉
The “fact” that higher interest rates have led to higher house prices
In one of the largest, most ridiculous, methodological and statistical mistakes I have seen in a very long time O’Connor states the following:
Higher mortgage rates have lead to higher house prices. In fact, a mathematical measure of co-movement, the correlation coefficient, where +1 is perfect co-movement and -1 is perfect opposite movement, has a value of 0.79. This indicates that house prices and mortgage rates are strongly moving together.
And then goes to state that it is A FACT that higher mortgage rates have led to higher house prices.
Now, the first thing to remember is that you can only analyse data when you have an implicit model – his implicit model is that only interest rates can impact on house prices, and that house prices have no impact on interest rates. This is a bad model, a terrible way of trying to figure out causation (which requires some theory and maybe knowledge), and is A RIDICULOUS application of ceteris paribus – as it is painfully obvious that other variables that are involved in the relationship have changed, and we need to correct for that.
So his implicit model is incomplete, and I would say wrong, and his data analysis is weak and biased.
So what is missing:
- Causation – trying to figure one is driving the other,
- Other variables that could be moving at the same time, and causing both to move up,
- Lucas critique – does he think that he is capturing the deep parameters of the model, or could a change of policy (which he might hint at) going to throw out his relationship.
An implicit model, and some actual econometrics, would help to clear this up. I mean what f’ing mechanism does he think leads from interest rates to house prices – if he doesn’t tell us, and then incorporate that into his analysis the whole thing is a load of crap.
Aren’t you being a bit harsh
I have no doubt that my tone is harsh – very harsh. But this man is a Senior Economics Lecturer at what some people view our top university. And he has made a set of mistakes that would probably see a first year student at Vic fail.
At the start of his article he says we can’t assume all else is equal all the time – then he goes on to assume it, in an inappropriate case, mix up his causation, and then dare to mention the word “fact” (which is a VERY bold word to use in economics).
I am being so harsh because I am embarrased – people in this position should be telling me the multitude of methodological and statistical errors I make, not embarrasing the discipline.
Trust me – these are the sort of errors that we have to work so hard to convince the public we aren’t making. We do a LOT of work trying to distill truth out of messy numbers, and trying to make explicit the different assumptions we are making. The sort of economics shown in that article is not welcome.
But what if he is right, and house prices fall!
OMG, of course house prices are going to fall (have fallen) – falling job security, tightening credit markets, and a realisation that house price gains were unsustainable will do that for you. Looking at the movement in these variables MAKES MORE SENSE than looking at the interest rate – it is called thinking before making an empirical model.
Think about it yourself – if interest rates were lower, would you now sell your house for less? If your mortgage was costing you less, would you get more desperate to sell your house?
I’m going to go have a stiff drink …
Update: Tom M further criticises the piece on his blog Defective equilibrium. Glad to hear that he is a student at Victoria University of Wellington – as it illustrates to me that my old uni is still teaching top quality analysis 😉
Update 2: As people have been repeatedly trying to point out to me – the lecturer is from accounting/finance. As a result, it is not a black mark on the economics department – it is a black mark on whichever of those disciplines he teaches. As a result, the reputation of Auckland, and New Zealand, economists is saved, hooray!!!