Hello, Andy Steingruebl here:
There was a great piece on NPR (a link to some of it
here: http://www.npr.org/blogs/thetwo-way/2010/04/safety_agency_to_use_ignored_p.html)
the other day about the MSHA (Mine Safety and Health Administration) and how
their enforcement against unsafe mines has been sorely lacking. One key
piece of the story is that essentially the statutory authority granted to the
MSHA is to close a mine, or a section of it. They don’t have the power to
levy fines for example, or prosecute individuals doing unsafe things.
As a result, their main weapon is essentially the nuclear
option against a mine operator. They are therefore reluctant to use that
authority, and so they have instead created a system of
letters-of-pending-action or somesuch. Under these they indicate to a
mine operator that they are getting close to a breach of their requirements,
and they give them a cure timeframe. Because mine operators can appeal
any judgments, this ends up being a real obstacle to credible
enforcement. Historically we’ve seen the same things from the FAA.
They are loathe to write up formal findings because they have the possibility
of causing major harm to an airline, rather than just getting them to modify
the offending behavior.
ICANN has a similar problem with policing their
registrars. historically the only penalty they can impose is removing a registrar’s
accreditation. This is the nuclear option. They have typically been
loathe to fully remove a registrar, and so we end up with bad registrars in the
system. We've seen them step up the enforcement recently, but I think these are probably still enforcements against only the most egregious offenders.
The lesson for us in working to craft internet governance
and regulations is that we need to understand this dynamic between regulator
and regulated-entity, and create flexible schemes that allow the regulator to
effectively police, without each policing action being a nuclear option, and
thus causing regulated entities to fight every judgment action in court, etc.
Comments