There are compelling reasons for adopting Balanced Approval Voting (BAV). It is a simple and easily understood voting system that measures voter opinion accurately. It completely avoids the spoiler effect and it makes possible elections with viable participation by many candidates, giving voters the freedom of choice they surely deserve in a democracy. BAV could be adopted for elections for local and state offices, for Congressional races and for the Senate. BAV would be an especially good improvement for our presidential primary elections where in fact, even today, there are often a great many candidates. And BAV can accommodate proportional representation.
But because of the Electoral College, presidential elections present some special difficulties. Because plurality voting is used by the Electoral College, the spoiler effect remains a problem and that problem is further complicated by Amendment XII of the Constitution. In the absence of a majority in the Electoral College, this amendment throws the selection of the President and Vice President into the House of Representatives. And these difficulties are greatly complicated should there be success with the National Popular Vote (NPV) effort.
If NPV becomes a reality, a majority of the Electoral College election will be committed to vote for the candidates who won the popular vote nationally, but it may not be clear what that actually means. So long as all states are using plurality voting (PV), the meaning of this is pretty clear and probably this would be the case for any other system of voting provided that one system were adopted by every state. But if there is a mixture of different systems of voting it becomes unclear how to compute the national popular vote. With ranked voting the situation is particularly difficult, but even with BAV used in some states but plurality voting still used in others, how to define a nation-wide count of votes for a candidate seems quite unclear. As we have pointed out in earlier articles, once we depart from the context of PV, the very meaning of "vote" can become ambiguous.
With BAV, each voter submits a ballot that evaluates, either explicitly or implicitly, each of the candidates, in effect giving each a score of -1, 0 or +1. Those scores are added for each candidate to compute a net-vote and the candidate with the largest net-vote wins election. Using the net-vote as a state's popular vote-count would perhaps be reasonable if every state used BAV, but the vote tally from another state using PV would mean something entirely different. Adding the two together would make as much sense as adding the number of watermelons from one state to the number of peanuts from another to gauge the combined agricultural production. This difficulty is not peculiar to BAV, it a problem whenever states might adopt different voting systems. In particular, it would be a very thorny problem if IRV were used in presidential elections by one state while PV is used in other states.
There probably is no right way to make these kinds of adjustments for adjusting between different voting systems. Rather there will be rationals invented for a variety of schemes and no doubt endless squabbling about how unfair any particular scheme would be.
Perhaps after the states perform their due diligence as laboratories of democracy, the states can all reach agreement on some single voting system (other than PV) to use in all states as well as in the Electoral College. But switching to that system for presidential elections would surely require amendments to the Constitution.
Presidential elections seem destined to using PV for quite some time. However, as we noted before, presidential primaries (if they still seem appropriate) could be changed to use BAV or any other system with no particular problem. These elections are not mentioned at all in the Constitution.
The insistence in Amendment XII for a majority in the Electoral College vote seems to reflect a hope for what can only be the illusion of a consensus. In a democracy, hope for a consensus is surely commonplace. No doubt this motivated the idea of eliminating candidates repeatedly until only two remain and in that last contest one of them will get a majority vote. But surely that only creates an illusion of consensus. It is a consensus in which many voters are, at the end, pressured to choose between two candidates who each seem a poor choice and quite possibly equally unsatisfactory.