« The Bank celebrates stupidity | Main | The economics of Pop Idol »

October 31, 2006


Dave Petterson

I think that although the politicians could follow the simple rules there would be undue attention paid to additional, political, issues that would occasionally screw the results.

In the 1980s I installed a forecasting module into our ordering system. It forecast our orders based on seasons and history. On the bench it kept our inventory down to a minimum with no shortages and when it went live our inventory only dropped a few %. After a few weeks with no real inventory decrease we examined the figures and discovered that the warehouse manager was taking the figures from the system and 'fixing' them prior to order because he was concerned about running out and didn't trust the computer. Once he was 'fixed' our inventory dropped so much we sold the warehousing site and kept a small warehouse on the manufacturing site.

Human intuition and experience impacts on the systems and some of us have hidden agendas which to those on the outside means it does not make sense.

Mrs Trellis

Could just be that your Taylor rule is close to the MPC benchmark - to operate it will have evolved a benchmark which balances output and inflation deviations, and probably based on a rather simplistic expectations formation model. A behavioual model of the power plays between MPC factions might be sufficient to explain the deviations - or it could just be that the data has been revised since then.

The comments to this entry are closed.

blogs I like

Blog powered by Typepad