The second CBO procedural change was to discontinue the use of the “alternative fiscal scenario” in the long-term projections
There is more at the link.
I think that CBO modeling is way over-rated and biased toward interventionist policies. I would take them out of the modeling business almost entirely.
For budget projections, you need to do modeling. Budget projections are numbers, after all.
But models are subject to all sorts of errors. GDP growth could turn out to be higher or lower than expected. Interest rates could be higher or lower than expected. Your estimates of the cost of programs could be off, because people may respond to incentives differently than what you expect. Laws may change.
Given these sources of error, scenario analysis is a must. The CBO should be doing more scenario analysis, not less. I am increasingly convinced that the CBO as it currently operates is performing a huge disservice to public policy. Congress should make major reforms to the mission and operations of the CBO.