This story by Dan Balz and Scott Clement about some midterm election poll results makes several errors in discussing some results as causes rather than effects. I believe this confusion of cause and effect is a result of the tendency for political reporters to view elections as sporting events, but a historical voting trend is very different from a batting average.
The first example is in the second paragraph:
Midterm elections generally favor the party that does not hold the White House, which gives the GOP a head start this year.
It is true that the party that doesn't hold the White House generally does better in midterm elections, especially if the President is in his second term. However, this historical fact doesn't "give the GOP a head start". The GOP's projected advantage lines up with this historical trend, but isn't caused by it. Both the historical trend and the GOP's projected advantage in 2014 are effects with common causes: inevitable dissatisfaction with whoever has been running the country recently.
In the next example, I will bold the confusion:
The poll shows broad dissatisfaction with Washington politicians. Just 22 percent say they are inclined to reelect their representatives in Congress. Almost seven in 10 Americans (68 percent) say they are inclined to look around for someone new this fall, the highest percentage recorded in a Post-ABC poll.
That does not mean the fall elections will mean defeat for significant numbers of House members, given the high reelection rates for incumbents and the polarized voting patterns of recent years.
As in the first example, high reelection rates for incumbents is a historical trend that is likely to continue in 2014, but the trend doesn't cause itself. The trend is an effect of "polarized voting patterns" as well as the human tendency to stick with "the devil you know".
With President Obama and Congress at loggerheads on major issues and little prospect for legislative action on major initiatives, the president's approval ratings have shown little change since earlier this year.
Here, it's not entirely clear if the authors are implying a cause-and-effect, or which way it's going. By my observation, it appears that the more President Obama "achieves" the less popular he becomes. That lowered popularity is partly the cause of the lack of legislative action, not the effect of it. If Obama were widely popular, he would have more success pressuring Congress.
All but about two-dozen House districts are occupied by someone from the same party as the presidential candidate who carried the district in 2012, which makes it harder for the opposing party to pick them off.
The fact that all but two-dozen House districts voted for a Presidential candidate of the same party as the Representative they elected doesn't "make" it harder to pick the Representative off. That the prior election and the upcoming election are likely to have similar outcomes is an effect of the voting preferences of the district.
Historical voting patterns are not like batting averages, and trends do not self-perpetuate in a causal fashion. Voting instances are reflections of underlying beliefs at a point in time.