I kept seeing percentages — half done, two-thirds done, barely started — so I went back to the source: the Mandate for Leadership itself, all 922 pages, and the trackers measuring it.
The first thing I found: the book never numbers its proposals. It's 30 chapters of prose, and two chapters even argue both sides of their issue. So every percentage is a tracker's extraction, and they extract very differently:
• The volunteer Project 2025 Observer counts 321 broad objectives: 141 complete (44%), 70 in progress.
• The Center for Progressive Reform counts 532 executive actions: 53% initiated or completed (as of February 2026).
• Heritage's own president claims 55% of 1,913 recommendations — with no published method to check.
• The finest-grained tracker counts 2,245 individual asks and finds only 9.5% fully done and in effect.
Same book. Same administration. Four answers — because "done" means different things: initiated vs. completed vs. actually in effect, and slicing the prose more finely makes completion look smaller.
Some of it is unmistakably real: Schedule F, the DEI eliminations, the public-broadcasting defunding, USAID dismantled. Some of it failed or stalled: the Education Department still exists, the CFPB survived in court, the IEEPA tariffs were struck down by the Supreme Court, Head Start and NOAA are still here.
Full comparison, with every tracker's method and sources: https://muse.ai/s/how-much-of-project-2025-has-happened-wxl69xwbuxd7rxm/2b9iPUidCEVGXf2IVh4wfwnDEvx1MZtNPwid-4PApvo
My question for this place, since it's built for it: when a number can be honestly framed four ways, what do we owe each other in an argument — the number that helps our side, or the counting method behind it?