Confessions of a forecaster

7 Sep

An interesting piece in this weekend’s FT magazine on forecasting by the undercover economist (and FT contributor) Tim Harford (see http://goo.gl/wazg8F if you have an FT subscription; it’s in the magazine in hard copy).

My interest comes from the odd desire I had when in my twenties to do econometrics, which led, given I was an economist in government at the time, to spending five years in HM Treasury as a macro economist and member of their forecasting team.
It was an interesting time to be there (from 1980 to 1985) as it was just after the arrival of the first Thatcher government. And I arrived as GDP was falling through the floor – 1979-83 was the worst post war recession until the 2008 crisis, with a peak to trough drop in GDP of 6%.
So did the forecasters foresee this? Not really, although there was a significant “change of view” at some point in 1980 when realisation dawned that it was going to be pretty awful. And the forecasters stuck to their guns in saying that any recovery would be slow – and were quite right in doing so.
The worst of the forecasting exercises was the “Medium Term Assessment” (MTA), which in practice was simply a Panglossian exercise in saying that we would get back to trend growth in the medium term. Adequate commentary on this was given by my undersecretary saying at some point in 1981 (the depths of the recession) “but when we did the MTA in 1975, this was the year it was all going to come right”.

So what do my ramblings have to do with Tim Harford’s piece? Well, it focuses on the work of the Good Judgement Project (https://www.goodjudgmentproject.com) which has uncovered some interesting insights into why forecasting does or does not work and how it can be improved. Key to it is that many forecasters are not, surprisingly enough, interested in being accurate – they are instead more focused in advancing a political cause (shades of the MTA again). If you want to be a good forecaster, forget that (an obvious source of bias). Instead, the Good Judgement Project recommends teamwork, keeping score (so error correction) and open – mindedness.

So on this basis setting up the (newish) Office of Budget Responsibility was a good move, and its forecasts should be better than the old, politically influenced, HMT ones. Well, there’s a Popperian falsifiable proposition for you!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: