Via a couple of sources, I see that the Treasury has downgraded its future growth forecasts in light of the impending Eurozone crackup. This will of course affect NZ's future economic state - and in particular its fiscal state, which depends rather heavily upon an as-yet-illusory post-recession surge of growth:
Prime Minister John Key has shrugged off Treasury's cut in the growth forecast saying it is ''not overly dramatic''.
Treasury has lowered its growth forecasts from 3.4 per cent to 3 per cent for 2013 and warned of cuts to government revenue as the European crisis rumbles on.
Dim-Post is fulminating against this apparent cock-up in a rather uncharacteristic fashion:
Right. So just to recap: Treasury made these very optimistic growth predictions in the 2011 budget, and everyone said they were delusional. But Treasury stood by their forecasts, and a month out from the election they repeated their insanely optimistic outlook in the PREFU.
The government and opposition parties based all of their fiscal policies on the Treasury forecasts. Both main parties claimed they’d get the country back into surplus in a few years, because of all the additional revenue from all that economic growth. Jobs weren’t a big topic of debate in this election, because Treasury promised us that all the growth would create 170,000 new jobs!
Now here we are, just over a week after the election and whaddya know? Treasury has announced they were too optimistic. Growth is going to be less than they said it would a month ago – even though nothing substantive has changed since then. The European crisis didn’t just emerge on Sunday the 27th.
This looks an awful lot like a government department conspiring to defraud the country during an election campaign. If it ain’t that then it’s gross incompetence.
Coincidentally, I just took a look at the Treasury's recent history of economic forecasting. I suspected, with moderately good reason, that their forecasting models aren't any good at predicting shocks to the economy. e.g. financial crises, which have been all the rage lately. My suspicion was that Treasury forecasts had run off the rails in recent years.
It was fairly simple to check this - I plugged forecasts from 2000 to 2011 into a spreadsheet along with GDP growth statistics from those years. (Figures from Stats NZ National Accounts tables and the Treasury. Underlying data available here in a Google Spreadsheet.) Played around with the numbers a bit, and got some interesting graphs (in hideous form, due to the fact that OpenOffice's graphing software was coded by Satan):
This graph is pretty straightforward and tells a quite interesting story. The black line is real GDP growth. As you can see, the purple line - Treasury's May estimates of GDP growth in the current year - is a fairly accurate forecast of growth. So Treasury can figure out what's happening at the present moment. Yay.
The dashed green line - which represents Treasury's forecasts for the following year - seems to be a reasonably good predictor of actual growth. It tends to get the overall momentum right but often predict the wrong level of growth. Meanwhile, the remaining two dotted lines - forecasts for the year after next, and the year after that - are laughably wrong. In 2006 and 2007, Treasury didn't predict the recession that would occur in 2009. (Looking behind the scenes, this might actually be a case of garbage-in-garbage-out. DSGE models tend to assume that the economy will continue functioning at a certain "equilibrium" level of growth unless perturbed by outside factors, and that it will return to the growth trend relatively quickly after any disturbances. I suspect that this means that it will almost always predict around 3% growth two years hence.)
Also of interest is the following scatter plot, which shows the relationship between predictions made in the May forecast and the December update:
In short, the two tend to agree closely on present-year growth estimates, as shown by the tight scatter around the trend line for the blue points. However, there's not a very strong relationship between the two forecasts of growth in the following year - look at the much wider variation around the trend line for the purple points.
And May and December forecasts for expected growth two years hence - i.e. this year's forecasts for growth in 2013 - are almost totally uncorrelated. There are often large variations between the two forecasts; this one was only noticed due to the potential for economic crisis. That is what is at work here - a forecasting model with a rather poor recent record of predictions beyond the following year - not a Treasury conspiracy. Dim-Post has missed the point, in spectacular fashion.
0 comments:
Post a Comment