Compiled from [e68445]. Using FreeCol rules.
Unfortunately, now that I only use 64-bit versions of Windows 8.1, I can no longer use the original Colonization to test, so this is based on my recollection.
In the original, once you have a set number of Rebel sympathizers (1 in the case of the screenshots), that sympathizer will stay a sympathizer after more characters are added to that town, even though the percentage changes.
Normally, this is not a problem, but because the Bad Government is based on the number (not percentage) of Royalists, adding a new Colonist will upset the percentage and can change a Rebel to a Royalist.
In the example screenshotted, you will see that the Colony (New Amsterdam) has 1 Rebel and 7 Royalists and based on the Very Easy settings, there is a 8 Royalist cap before you get a Bad Government efficiency rating.
My intention was to move my new Elder Statesman to the Town Hall, but when I did I got a Bad Government efficiency notice. I figured out that when I added the Elder Statesman, not only did the percentage changed, but the Rebel sympathizer disappeared becoming 0 Rebels and 9 Royalists.
Saved game.
Interestingly enough after, I let a turn elapse (with the Elder Statesman outside the town), I tried adding the Elder Statesman to the Town Hall and this time, the percentage went from 13% (as it was originally) to 12% (instead of 11% as before) and the single Rebel remained.
This can happen. An unattended town hall still produces some bells. 11% is just too few to produce 1 rebel out of 9, but 12% is enough (11 < 100 / 9 < 12). It is just possible you were right on a critical amount.
However the main point is your claim that rebels do not reduce. I think this is fairly unlikely given the amount of work that has gone into the rebel calculation from people (unlike me) who did play Col1. I think the appropriate thing to do here is to move this report to the "Pending Features" tracker since it is an outright Col1-incompatibility issue, and add an item to the WWC1D wiki page in attempt to gather evidence.
Ticket moved from /p/freecol/bugs/2765/
Only one test case so far, but with a Col1 game on Dutch/Explorer, I had a colony with 1 Rebel and 4 Royalists and I added a colonist and the Rebel stayed with a percentage of 16% and 84% with the exact 1/6th percentage being 16 2/3rds.