For the better part of the past year, I've been asking myself on what basis we do, and should, select leaders on the national level. I have been asking myself what it is that the Executive Branch does and should do in its daily operation. I finally have a half-answer.
The scope of the Executive
The Executive Branch is huge. Just, massively, mindbogglingly huge. Seriously, click through that list and start scrolling—and compare it to this list of Senate-confirmable positions. There's not even a number attached to the number of confirmable posts; only a range exists. Twelve to fourteen hundred seats are filled by Executive nomination and Senate confirmation. Anyone who has done hiring can confirm that there is no way to vet and organize that many people quickly, especially in the short amounts of time available to an incoming administration.
A (typical) new administration can set its broad goals, but there is no way—short of freezing time and spending a year between November and January reviewing staffing decisions—to optimally build an organization like that. This is where party machinery comes in.
I suspect that the two major parties maintain lists of generally qualified candidates, specific shortlists for certain positions, and additionally know who to call to identify talent. Government of this scale doesn't turn on overnight, but with the right networks it might seem to. But this machinery is in many ways a consequence of a long-standing two-party system. There is no easy to conduct a national campaign in a country of 300 million people without leveraging some machinery, and one way to reward key pieces of that machine is by handing out jobs. Patronage is nothing new, and we can safely assume that among those 1400 positions there are bound to be a few sinecures.
Without describing other possible configurations of the Executive Branch, it's worth pausing to consider if this is the government we want. By their rhetoric, you might suspect that there would be more difference in the configuration of successive governments. Nevertheless, small-government conservatives have traditionally needed to reward their benefactors in the same way that tax-and-spend liberals do—winning national elections is hard—so I suspect the main difference, historically, is in the number of "good faith appointments." That is to say, the person nominated for a seat as undersecretary of the department of we-don't-really-care-about-this could be anyone.
The New York Times posted an excellent visualization of what issues were most important to Americans at different times. Click through the link and play around for a bit; there are arrows that allow you to navigate through history and not just those surveys conducted immediately prior to elections.
Anyone with a cursory knowledge of American history can see that there's a clear correlation between the concerns of the day and the specific administrations' actions. The recession of the early 2000s is writ there, as is NCLB; by contrast, domestic issues are predictably overshadowed by "foreign policy" (i.e. terrorism) by October 2001.
An administration is like the Eye of Sauron: it can see anywhere, at any time, as long as it knows to look there. When priorities turn 180º and it comes time to bring in the back-benchers, an administration's true nature can be revealed.
To put it another way (that's mostly for my own benefit), a election is decided on who offers the best piecewise linear match to an unpredictable function, and damn the rest.
"How can you support him?"
So: how do we decide who should be elected to office? (And not: how should we decide?)
There are ugly bits, such as party selection machinery (or party non-support, in "unwinnable districts"), the primary system, and so forth, but when it comes to early November, my theory is this: we can easily forgive a multitude of sins, as long as the problem that a candidate offers to solve is sufficiently big that the sins seem small.
In short, voters act like this: "If I sincerely believe that X" (where X is killing babies, the war on terror, global warming, inequality) "is an existential threat, I would elect Pennywise the Clown to office if I think he can help fix it, despite knowing that kids seem to mysteriously go missing at all of his campaign stops."
The sins that the right and the left can ignore vary, but both sides ignore sins in roughly equal measure. There are a different ways to win elections in this kind of climate, but one thing is clear: it's a pretty good way to end up with a congress full of clowns.