I have been looking at this research report from ABARE on their forecasts for different scenarios of Global Warming effects and the cost Australia in terms of GDP. As someone trained in science one thing that I find striking about this is that whereas scientific estimates include some estimate of uncertainty, eg the IPCC’s estimate of the effect on AGW to 2001 of 2C and 4.5C. The forward estimates of GDP do not.

This strikes me as bizarre that no attempt to quantify the error band for GDP estimates is made particularly over such a long time frame. Even over the time period of one year forward we could expect a standard deviation of perhaps +/- 0.5% on the GDP estimates and that is probably generous. Over the 50 year time frame this error (assuming they are independent random errors) would scale to around 3.5%. Given that the differences in the outcomes between scenarios for all but the most punative carbon tax are mostly smaller than this it would tend to suggest that the model isn’t really accurate enough to distinguish between the different scenarios and instead they are quoting false precision.

Perhaps I have missed something and there is some footnote explaining this, or some standard assumption but I can see it and if so can someone please point out the relevant information. More likely I think they don’t want to admit inability of their models to clearly distinguish between different courses of action.

Advertisements

Like this:

LikeLoading...

Related

This entry was posted on Tuesday, October 31st, 2006 at 9:59 pm and is filed under Economics, Probability. You can follow any responses to this entry through the RSS 2.0 feed.
You can leave a response, or trackback from your own site.

5 Responses to Macro economic prediction and errors

Yes it suprises me because econometrics does this sort of analysis in great detail, test for signifgnace etc but I don’t see it in forward macro modelling predictions.

The model that ABARE use in the paer to which you linked is MEGABARE. It is an example of a computable general equilibrium (CGE) economic model. It seems to me that there are a few different types of error bands that you might obtain in CGE models. For example, you might derive confidence intervals for any predictions that are based on the statistical properties of the parameter estimates in the model. This would require you to estimate the parameters of the model jointly. I suspect that in some (and perhaps in many) cases, the parameters that are employed in CGE models are not statistically estimated within the model itself. Some of the parameters are probably based on other studies, while others might be based on model calibration. As such, an alternative method for calculating error bands might involve simulations to test the sensitivity of the model results to alternative combinations of parameter estimates. However, I am not a specialist CGE modeller, so I could be wrong.

I do not know why error bands are not reported in the ABARE study. Nor do I know whether or not they are commonly reported in CGE analyses. However, Peter Wilcoxen has a useful overview of CGE modelling on his website that includes a link to a paper that calculates confidence intervals for CGE model predictions. His website can be found here:

(Disclosure: I know Dr Wilcoxen through my graduate studies, although I was not taught by him.)

(I saw this post earlier, but did not post a response. However, when I saw your comment over at Catallaxy, I decided I would post a response, even if it is not as informative as you would like!!! I’m not registered at Catallaxy yet, so I can’t comment over there. As such, I am commenting here instead!!!)

Good point Steve, Having done econometrics at uni i would expect their models should have an error band. (an Rsquared might be a bit to ask)

Yes it suprises me because econometrics does this sort of analysis in great detail, test for signifgnace etc but I don’t see it in forward macro modelling predictions.

Steve,

The model that ABARE use in the paer to which you linked is MEGABARE. It is an example of a computable general equilibrium (CGE) economic model. It seems to me that there are a few different types of error bands that you might obtain in CGE models. For example, you might derive confidence intervals for any predictions that are based on the statistical properties of the parameter estimates in the model. This would require you to estimate the parameters of the model jointly. I suspect that in some (and perhaps in many) cases, the parameters that are employed in CGE models are not statistically estimated within the model itself. Some of the parameters are probably based on other studies, while others might be based on model calibration. As such, an alternative method for calculating error bands might involve simulations to test the sensitivity of the model results to alternative combinations of parameter estimates. However, I am not a specialist CGE modeller, so I could be wrong.

I do not know why error bands are not reported in the ABARE study. Nor do I know whether or not they are commonly reported in CGE analyses. However, Peter Wilcoxen has a useful overview of CGE modelling on his website that includes a link to a paper that calculates confidence intervals for CGE model predictions. His website can be found here:

http://wilcoxen.cp.maxwell.syr.edu/ .

The overview of CGE modelling at Dr Wilcoxen’s website can be found here:

http://wilcoxen.cp.maxwell.syr.edu/pages/371.html .

(Disclosure: I know Dr Wilcoxen through my graduate studies, although I was not taught by him.)

(I saw this post earlier, but did not post a response. However, when I saw your comment over at Catallaxy, I decided I would post a response, even if it is not as informative as you would like!!! I’m not registered at Catallaxy yet, so I can’t comment over there. As such, I am commenting here instead!!!)

Regards,

Damien.

Steve,

I need to make a slight correction to my earlier comment. The model that ABARE use in the paper to which you linked is GTEM, not MEGABARE.

Regards,

Damien.

Thanks Damien.