Haskell Productivity

[Steinitz, Dominic J]

The Haskell website claims that

"Ericsson measured an improvement factor of between 9 and 25 in one set of experiments on telephony software".

Presumably this is with Erlang not with Haskell. I have searched for the reference that substantiates this claim but I've only been able to find:

http://set.gmd.de/~ap/femsys/wiger.html 

which talks about a productivity factor of 4

and

http://www.md.chalmers.se/Cs/Research/Functional/Fudgets/haskell-vs-ada-abstract.html

which suggests that Haskell is about 2-3 times as productive as imperative languages.

Can someone point me at some more references? Especially the one that talks about a productivity improvement of 9-25?

Thanks, Dominic.

-------------------------------------------------------------------------------------------------
21st century air travel     http://www.britishairways.com