End Gerrymandering!
This should be of serious concern.
The fact that a great many districts are now 'safe' seats only assures us that Congress will become ever more permanent, isolated, arrogant, an indifferent to the voters.
There have been some efforts to reform the way districts are redefined, removing the process from politics (and from control of the 'winners'), but in all cases it has been suggested that it be turned over to some supposedly "neutral' body, such as a group of retired judges or businessmen.
Obviously, any proposal involving a set of humans will eventually fail for the same reasons that legislative redistricting fails.
It's the human element, stupid ! There may be reasons for it, but I fail to understand why no one has proposed a purely software-based system of honest-to-goodness random redistricting based only on population and the contiguity of population polygons which nest snugly with one another? No party, no race, no economic class, no tentacles, no nuthin', to bias the resulting voter distribution.
Computer experts have told me this would be perfectly feasible.
For example: Currently, a typical Congressional representative district has a population of roughly 650,000 people (= national pop of 280M / 435 districts).
Thus, if a state has 6,500,000 people, it will have 10 congressional districts.
This formula is fixed nation-wide.
Almost every state has its population distributed in a few heavily populated cities, with the rest of the people spread out in suburbs, towns, and in low density rural areas.
The computer program would start out by fitting the map of a 10 district state with 10 equal-area polygons (mostly square-like, with no tentacles) within the state's boundaries.
Those polygons, or cells, located over cities would immediately shrink, in place, to encompass a space populated by 650,00 people, creating a district.
The cells adjacent to that first district would pickup the spillover of people and shrink or expand in place accordingly, until they each met the target of 650,000 people, to form districts.
The algorithm would tend to optimize compactness.
This process would continue until all 10 districts shared equally the total population of the state, and all districts were roughly compact polygons, with no significant tentacles.
I am sure this feasible software-wise.
Why doesn't some truly independent organization take this idea and run with it?