A compensatory mixed-member proportional representation system means that we decide on the number of compensation seats and constituency seats. To properly correct the distortion of constituency seats, there must be a sufficient “bank“ of compensation seats. This must be monitored at the national level and at each sub-level, whatever it may be.
Statistically, to correct a ratio of 60% constituency seats, 40% compensation seats are required. This affects only the compensation technique. In the event of arbitration and if the number of constituency seats needs to be increased and the number of compensation seats decreased—if the goal is to maintain the number of current seats—issues arise, and not only in terms of effectiveness. The statistical effectiveness of correction may still exist, except that fewer compensation seats means less opportunity to also correct the male/female ratio, despite the measures implemented such as alternating males and females on the compensation lists.
The mechanisms are easier to implement. We can implement alternation on the lists, but we can't implement it in constituencies. The distortion and maybe—we don't know—the population's voting patterns could result in a lower percentage of women elected. The votes by constituency are still cast based on the first past the post system. We must still expect there to be fewer women elected in a first past the post system. It's still a majority system in part. More women would be elected as a result of the compensatory aspect. The compensation list must be consistent overall on a regional and national level.