Much of what has been said about the 2016 election was that the outcome came down to demographic changes in the key three states where Trump pulled off narrow victories unexpectedly (Wisconsin, Michigan, and Pennsylvania). Experts underestimated turnout by older, white voters—especially men—who voted for Trump. Less remarked upon is the fact that several thousand voters in these three states had more influence on the outcome of the election than millions spread across the remaining 47 and the District of Columbia. The reason for this is because these three states were among the four closest in the nation. In the method we use to our elect our president, the votes of most consequence are the ones cast in states for the person who wins by the slightest of margins.
Why? The “winner-take-all” rule, followed by 48 out of the 50 states plus DC. By state law, every electoral vote goes to the candidate who wins the most votes cast in that state. One could win a state by two-tenths of a percent, as President Trump did in Michigan, and still add all of that state's electors to his or her scorecard for the presidency. On the other hand, winning by millions in a state, as Clinton did in California and Trump did in Texas, offers candidates zero added advantage. Not only is winning a state by several million votes meaningless, so is coming in a close second. Hillary Clinton bested Barack Obama’s 2012 tally by a million more votes in California, half a million in Texas, and over a 100,000 more votes each in Georgia, Arizona, Florida, and New York. That advantage was wiped clean by her underperforming in Michigan, Wisconsin, and Pennsylvania and losing those three states by less than 7/10 of a point. Despite a difference of only "22,748 votes out of 2.9 million in Wisconsin, 44,292 out of 6.1 million in Pennsylvania.... [and in Michigan] 10,704 votes out of 4.8 million," President Trump won 100 percent of those states' combined 46 electoral votes while Hillary Clinton won zero.
With these three states statistically a toss-up, consider how a reversal might happen in the next election. President Trump could make great gains in the American Southeast, Texas, Ohio, and Florida in 2020 and still lose the presidency if Pennsylvania, Wisconsin, and Michigan slip even in the slightest to the Democratic nominee. In 2004, such an event nearly took place in Ohio when President Bush came within 60,000 votes of losing the presidency even though he was 3 million votes ahead nationally.
Demographics change, and as 2016 showed, they can swing an election. The way that they do this would be less extraordinary if the bulk of their impact were felt at the national level. If victory in the election were tied to the winning of the national popular vote, instead of exclusively to winner-take-all, demographic shifts in just a few states could not change the outcome of the election as a whole.
Changing the rules to give primacy to the popular vote would give all citizens, as well as all states, a voice. States would NOT, contrary to a popular counter-argument for reform, be giving up their say over who becomes president; they would in fact be enhancing it. A vote in Texas would always count as much as one in Michigan.
If nothing is done, the situation will likely become worse. More Americans are moving to states where the outcomes are less close. What demographic shifts can also do, unfortunately, is polarize the electorate and reinforce the boundaries of red and blue states. Again, if the winner-take-all selection method is allowed to continue as is, polarizing will only shrink the number of states which end up mattering. Whatever impact demographic changes such as migration and aging of the population create, the scenario if we stick with the current method will continue to result in an un-representative outcome for the next president, since it does not take into account the wishes of the majority of voters —whether they vote Democrat or Republican!