Secondly, key escrow will impose considerable engineering costs. Most of these are described in a document entitled ``The Risks of Key Recovery, Key Escrow, and Trusted Third-Party Encryption''. This is also appended and is hereby incorporated. Government must consider, if deciding to impose a significant cost on trade and industry in the name of law enforcement, whether the greatest returns would come from key escrow, or from some other initiative more relevant to real policing concerns (such as the provision of low cost helicopters).
In addition to the arguments in that paper -- which I coauthored with ten other cryptologists as a submission to the US Senate -- I would like to add some specific UK experience gained in the course of consulting for a large range of organisations including various police forces, the British Medical Association, a number of systems houses and the UK insurance industry's Loss Prevention Council (though I speak here only on my own behalf).