NPV's original wording (2006, not recommended): NPV wanted the compact-states to award all their electoral votes to the national Popular Vote winner. This was to take effect only once an electoral-vote-majority worth of states joined the compact.
Our growth & joining-motivation analysis: If a state joins the compact, it may be forced to vote against its own interests. By not joining, a state's popular vote is still taken into account by the compact states (as part of the National Popular Vote) and hence still acts to swing the compact. Hence by not joining, a state gains all the benefits of joining, without the disadvantage.
Conclusion: if enough states act in their own interests, then the compact will never grow large enough to trigger. Hence the entire NPV plan is doomed to have no effect. Both Hawaii and California in their governor-vetos of the NPV compact, raised as their primary point that joining could force that state to vote against its own interests. As time goes by this will get even clearer. So the whole current wording of the NPV plan is a dead end. It is doomed. It cannot succeed.
Our suggested revised wording: The compact-states will award all their electoral votes to the Popular Vote winner of the compact-states only. This takes effect immediately no matter how many or how few compact-member-states exist.
Our joining-motivation analysis: If a state joins the compact, it may be forced to vote against its own interests. But it also gets a chance to swing the compact's bloc vote, thus magnifying its voting power greatly (and whenever that happens it always is in the right direction, i.e. in favor of its own interests). By not joining, a state's popular vote is ignored by the compact. A state not part of the compact is likely to be ignored by presidential candidates as unimportant. The huge compact bloc-vote is all-important and decisive for them. Getting ignored by presidents is undesirable. So that is more motivation to join the compact.
So in summary, joining is a gamble for a state which sometimes can pay off with a big otherwise-unachievable impact, but other times can actually hurt the state's interests. Not-joining is a sure thing which will lead to that state being ignored by presidents for the rest of eternity. If joining is a winning gamble, then the state is motivated to join. Is joining a "winning gamble"? We claim the answer is "yes" if the compact is suffciently "balanced." A state which joins a balanced compact has a good chance of swinging the compact its way, thus hugely magnifying its power. Even a mere 25% chance for that huge positive far outweighs the negative that the state might not be able to swing the compact (which even then is only a "negative" if the compact happens to vote against that state).
Our growth analysis: Our new rules automatically naturally cause rebalancing forces. Let us explain.
(I) If the compact states had about a 50-50 split of R and D voters, that highly motivates a new state to join the compact to swing it over their way. (E.g. if an R-state joins, then it swings the whole compact to R-voting. That's huge power, very desirable for that state.)
(IIa) But then, we have a slightly unbalanced R-favoring compact. That highly motivates a D-state to join the compact to flip it the other way (with good probability) – again giving it huge power and very desirable. (Example: If 50-50 chance of causing flip-over versus a 50-50 chance of not quite, hence having your state's vote thrown in the garbage can, then it is very worth it to accept that gamble, since it gives your state huge expected power gain.) This joining would rebalance the compact.
(IIb) But consider instead an R-state joining an already-R-favoring unbalanced compact. That move does not give that state any extra power because it does not change anything. No election result would change thanks to it joining.
Summary: the motivational forces tend to encourage rebalancing the compact more than they favor further-debalancing it; hence it should grow in a balanced fashion.
Worries: Those rebalancing forces were all very nice, but maybe not nice enough. There is noise and randomness too. If via bad luck we do reach a highly unbalanced R-favoring compact, then a D-state will not want to join it (will not be enough to swing the compact, hence simply will be electoral suicide for that state) whereas an R-state will be neither motivated nor unmotivated to join it. This status would also be a dead-end, then. That's bad, since we want the compact to keep growing.To avert that dead-end, further wording could be put into the compact to artificially force balance or at least to create extra forces favoring balance. For example, this wording could say "you are not allowed to join unless your state's popular vote last election was opposite-direction to the pre-joining compact's vote."
The initial start-up of our compact might be random, or might not. (Extra wording could try to get a good start-up.) Bad start statuses do not necessarily really matter because at first, every state is highly motivated to join the compact to swing it its way, since when the compact is small, the swing probability will be large often enough that there should always be states motivated to join it.
Contrast: The original wording of the NPV compact is a dead-end immediately for sure, while our improved wording only reaches a dead-end with some probability. That's better.
Dead is not dead: But looking deeper, suppose the compact-done-our-way still does fall into a dead-end unbalanced state. Fine. In that case, there still is motivation to start up a second, co-existing, rival compact. It will keep growing until it either gets big or falls into its own unbalanced dead-end. If the latter, then there will be motivation to start a third. Note in this "battling rival compacts" picture, the states in balanced compacts hold all the power since only they have the power to change an election result. Presidents will only pay attention to these "live" compacts. All this (creation of more live compacts; motivation by compacts to try to stay live) action is good in my view since it represents progress.
There could be further wording allowing two rival compacts to merge if they agreed to do so by majority vote. This would allow two unbalanced "dead" compacts to, by merging, both become larger and more powerful still, and also to reclaim live-status. Also there could be further wording saying if two rival compacts both reach some size-threshhold, then they must merge.
My speculation: If the rival compacts scenario did happen without merger, then there would be a lot of popular disgust at the situation – which would force congress to act to either invalidate the compacts or to forcibly merge them into a super-compact (the latter would require a constitutional amendment; actually hopefully the public would be presented with a choice of 2 rival constitutional amendments: one to get rid of these compacts, the other to eliminate the electoral college). In the invalidation case, fine, we've made zero but not negative progress, and we've called a lot of attention to voting reform issues, so: net positive. In the amendment case: better, total success.