How to measure a recession

[mandrake]

Hi

I wondered if anyone knew ways to incorporate a hypothesis of a recession effect in your estimation. I've used a decrease in the gdp growth rate (variable is the percentage change in gdp volume on the previous year) as an indication of a recession so far. But this method is kinda first differenced from the get go, while the other variables are not. I've tried to use regular GDP numbers in millions of euro, it works, but its not at easy to point to a recession this way.

Any opinions on how this could be done?

Sincerely,
Mandrake

[JimForest]

Why not use the NBER dates for recessions and just create a dummy variable for recessions?

[EViews Gareth]

NBER is for the US. His post leads me to believe he is looking at Europe

[JimForest]

Oh, I see. Well in the absence of an agency that strives to date recession beginning and end dates, i would just suggest using an ad hoc method such as two consecutive quarters of negative GDP growth. Perhaps you define the end of a recession as two consecutive quarters of positive GDP growth.