According to my statistical analysis of past elections, the unemployment rate has been responsible for about a third of switched congressional seats. Further, when the unemployment rate is above about 7 percent, that contribution leaps to somewhere between half and two thirds.
When unemployment is high it is a huge factor in house seats being lost by the president's party. When unemployment is low, other factors come more into play.
Based on this theory, you could have predicted that Democrats would lose at least 40 seats in the House without looking at a single news story, poll result, party platform, or campaign ad.
But there's more to this story.