The rationale is simple. The ethnic minority voter group is a VOTER group. Illegal immigrants are not voting, so do not make up part of the voter base. Thus, they wouldn’t increase the size of the ethnic minority voter group. Only ethnic minorities who were citizens would be voting.
This may also apply to immigration & how easy it is to become a citizen via citizenship tests. The easier it is to become a citizen, the more rapidly the ethnic minority group increases. Having a high amount of lag on this policy may also be advised - say, it takes 24-32 turns for it to take full effect? I’m not sure.
Pure food for thought. Immigration should still definitely have an effect on population, unemployment, service usage, etc, just as it does right now.
But currently, as illegal immigration increases, so does the size of the ethnic minority voter group - which is odd when citizenship tests are strict. If they were easy & asylum rules were lax, this would make more sense.
tl;dr disconnect the immigration rate from the size of the ethnic minority voter group. Instead, have citizenship-related policies, combined with immigration, affect its size.
Ideas for future policies: dual citizenship (allowed, not allowed, or only for culturally-similar countries? This makes a BIG difference in citizenship rates) and citizenship inheritance (only if a parent is a citizen or if you’re born in that country or if you grew up there) might let players feel more in control of the ethnic minority voter group size.
An alternative to setting it up this way might be to set the lag time on ethnic minority group size from population to be ludicrously-huge; say, 120 turns. The lag would be lower or higher depending on how easy naturalization (IE citizenship tests) were.
Pure food for thought!