A written agreement spells out expectations (and hopes)
Each group knows what the other would like in terms of cooperation. The difference between a promise and a best effort is important information.
Examples: flaring conditions, results to be shared, possible special analysis.
A written agreement avoids misunderstanding
Examples: Issues like confidentiality and authorship
None of us want our results being presented by some other group without our knowledge.
We probably prefer not to have our names appear on a paper without our knowledge.
Authorship guidelines can be defined, e.g. the TeV telescopes all prefer to have authorship open to all team members.
A written agreement defines procedures
We need to know the process by which we cooperate.
Example: How do we maintain confidentiality?
Example: Who chooses the leads on a particular paper?
Example: Peter Michelson is the only one from LAT who can sign agreements. The rest of us offer advice.