Slate Star Codex has a well-argued post asking conservatives to vote for Clinton in swing states or voting for Clinton or Johnson in non-swing states. If you are interested in Trump vs Clinton arguments, I recommend reading it.
I find some of Scott’s arguments more compelling than others, especially his point on variance, a similar argument I made in my Trump piece. Trump is a completely unknown factor; he could allow his advisers to make most policy or he could take complete control and make policy himself. On the campaign trail he has been quite capricious, changing his mind on every issue. Hillary Clinton is a known entity, and what we know about her is pretty bad (I’ll get into policy in a different post), but the worst Trump presidency is absolutely more terrible than the worst Clinton presidency. Given the House is likely to remain in Republican control, I’m also willing to bet that the House will oppose most of Clinton’s policies, meaning she’ll be less able to implement drastic policy changes. Whether and to what extent the House will stop Trump from implementing bad policy is another unknown. Whether you vote for Clinton or Trump could come down to the level of risk you are willing to take.
I’d say I’m somewhat risk averse, and Scott identifies similarly. The world today is healthier, wealthier, less corrupt, more democratic, and more free than ever before in human history, so totally dismantling institutions is quite a risky endeavor. However, if you support Trump because you aren’t risk averse, then you shouldn’t be worried about voting for a third party anyway, and I’ve already made a long case as to why Gary Johnson deserves your vote over Trump.
However, I’d like to point out that even in swing states, voting for a major candidate (like Clinton) may not be as appealing as it sounds. I pointed out previously that the chance of your vote making a difference, even in a swing state, is incredibly low. Scott brings up that the federal budget is so large even if the president only has control over some discretionary spending, and even if one candidate is only marginally better at allocating the budget to important things, the difference in payoff is in the hundreds of billions of dollars. Thus, the expected value of voting in a swing state with a one in 10 million chance to determine the outcome is still hundreds of thousands of dollars.
However, while the worst of Trump is definitely worse than the worst of Clinton, the expected value of the Trump budget may be very similar to the expected value of a Clinton budget, especially for someone whose political outlook is similar to that of this blog, it’s just that the expected value of Trump’s budget has a much higher standard deviation. Additionally, as of early October, markets don’t have this election as particularly competitive. There is plenty of time for that to change, but if the election were held today, even voting in Florida would likely not be a “swing state” in the normal sense. Moreover, the vast majority of people do not live in swing states. According to the 538’s latest forecasts, there’s about a 50% chance of the “tipping-point state” (the state which provides the 270th electoral vote) being either Florida, Pennsylvania, Michigan, or North Carolina. Florida is the most likely at around 17%. Most voters do not live in these states, and while I think even people casting ballots in them could probably vote for a third party without affecting the outcome, everyone voting outside of those states will have no impact on the election.
Finally, the benefits of voting for a third party remain: ballot access is a serious problem and there will never be viable third party candidates if it’s impossible to get on the ballot. This year would have been a wonderful year to have an alternative right-of-center party for Never Trumpers to move to, but there’s just no way for a candidate to get on the ballot in most states late in the process. Additionally, if any third party hits 5% of the national vote (Gary Johnson is the only candidate close to this), their party qualifies for federal funding in the next election. Votes for a third party this year set the groundwork for more parties in the next election cycle which can affect much more than presidential elections. Votes for the main parties are completely wasted in non-swing states, and even in competitive states, the chances of affecting change with the two parties is incredibly low.