I've been seeing variations of this question floating around for a while, and of course, Trump floats it as his standard defense against the Russia thing. Trump won, Clinton lost, and it's Hillary's fault. If Clinton had done X, Y or Z, she would have won, so it is Clinton's fault that Trump won. Now stop asking about Russia. She didn't campaign in Wisconsin, she didn't have the right platform, she didn't "connect" with the right voters (meaning white people), etc.
There are two interesting political science aspects to this question. First, the normative political theory question. Should elections turn on what candidates do? Second, would the outcome have been different had Clinton behaved differently? This may take a while, so this may turn into a long series. We'll see. I'll start with the normative side.
Trump views everything as a zero-sum game of winners and losers. I have addressed that before, as have other commentators, and while it is false for economics, including trade, an election really is zero-sum. So, our Russian-loving President is right:
Kurgan-as-Trump works on so many levels...
Anyway, as a game theorist, I study how candidates interact strategically in order to win elections. Normative political theorists, though, address issues like how elections should work, and actually, my first book was more about that. There is a normative component to what we call "spatial theory." In spatial theory, two candidates strategically select locations along a policy dimension (or, if we want to get messy, in a multidimensional policy space), and voters vote for the candidate who offers them the most "utility." In a unidimensional model, there are actually social benefits to having the candidates converge to the location of the median voter. It is, in some sense, representative, or at least, more representative than other possible outcomes.
We can base normative assessments of election outcomes on other factors. In the "Assessing democracy..." series from December and January, I wrote a lot about "valence" and how to combine the valence dimension with the policy space, and there are social benefits to having the higher valence candidate win. Remember that the "valence" dimension is the dimension that captures candidates' traits that voters just intrinsically value, like competence and honesty. Stop laughing. Or crying. Whichever you are doing right now. There are social benefits to having competent and honest officials rather than incompetent and corrupt officials. It's a hard lesson for some people to learn, and many will never learn it, but let's move on.
Policy and valence: one of these factors is based on strategy, and the other is... less so. Your policy platform can be selected strategically. You can move left or right to appeal to certain segments of the electorate. There are tradeoffs, obviously, and that's the dilemma. From a Democrat's perspective, you can move to the center to try to win more of the independents, but you risk having some of those fucking Bernie people who don't understand why capitalism beat communism whine about how they won't feel special enough anymore, and sit out the election, thereby letting the most unfit candidate in history become president. Or, a Democrat can move left to try to give warm and fuzzy feelings to those precious little snowflakes who voted for Bernie in the primaries, but lose the center.
Of course, Republicans have the same general dilemma, and that's the game. A Republican can move to the center and risk having the racist, redneck yokels sit out the election because he accidentally admitted that the planet is more than 6000 years old, or he can move right to satisfy the toothless brigade and lose the center. That's the game of spatial positioning, as Anthony Downs wrote about it in An Economic Theory of Democracy in 1957. He was just slightly less snarky about it.
According to conventional spatial theory, the answer is that both parties should converge to the center unless we add really messy complications. Mathematically, it is surprisingly hard to make it rational for candidates to not adopt the location of the median voter, and we have about 50 years of scholarship where people jigger the models to generate those results. Still, platform selection is strategy.
Valence? Not strategy, or at least, not to the same degree. In principle, one could choose to read, as a matter of strategy, but if Donald Trump wanted to make himself smarter, he couldn't. He is a stupid person, and stupid people cannot choose to be smart. He couldn't choose to be competent. Competence relies on intelligence. See previous sentence. He can't choose to be honest. He doesn't know how. I'm not sure he understands the concept of objective truth, or has the capacity to think beyond himself in the way that a true civil servant must. He is what he has always been because he can never choose to be otherwise. A psychopath cannot choose to be a good person as a matter of strategy.
Now, I started this post with a pair of questions: should elections turn on what candidates do, and could Clinton have changed the election by behaving differently? If we believe that valence should play a determinative role, then it follows that strategy should not because valence is not strategically selected. If we believe that elections should be more about policy, then elections should turn on candidate strategy. And that's where I'll pick things up tomorrow...