This  paper examines optimal mechanisms for dealing  a global commons problem with a dynamic  ``usage externality."  A leading example is carbon consumption.  At each date,  each country derives simultaneous    benefit both from the use  and the aggregate   conservation   of an open access   resource.   Countries benefit from   conservation because it allows them to  avoid the  environmental costs of resource   consumption.    The  relative benefits of consumption  compared to    conservation     are summarized by a privately observed parameter  --- the country's resource  type --- which evolves   stochastically  each period. We  characterize the constrained-efficient resource quota  when   neither  truthful disclosure nor  compliant resource use is presumed.  Not surprisingly, with complete information the optimal quota allocates more of the resource each period to countries with high value of consumption (and low value for  conservation). However, under incomplete information, the optimal quota is invariant  to the country's resource type in any period. In the case of $\mbox{CO}_2$,  this means that  countries with higher than average carbon requirements receive the same allocation as those with lower than average requirements.  We refer to this as the property of {\em extreme quota compression}, and show that  extreme compression  is robust to the distributional process  on shocks.