Gerrymandering -- jiggering district lines to get a desired result -- used to be a quite primitive process involving a fair amount of guesswork about which neighborhoods could be counted on to vote in what way. Sure, it often worked, but it was crude. Today election software makes it easy to use histories and the demographic characteristics of voters to lump us into groups whose leanings predict which party and even what sort of candidate will win in any set of boundaries. Just about the only way such gerrymandered districts change their partisan leanings is when people move in and out. Most elections just ratify the status quo rather than reflecting voter opinions.
Though Republicans were widely successful in drawing favorable Congressional districts for themselves in 2010 (because they had just won many state legislative contests), some of what they did was extreme enough so it is being challenged in court -- because effectively a well done gerrymander disenfranchises people whose votes can never count for a winning candidate. Federal courts have intervened repeatedly in North Carolina where gerrymandering reduced the effective power of Black voters. The Supreme Court is considering a Wisconsin case in which Republicans had managed to draw lines that yielded them 60 of the 99 seats in the Wisconsin Assembly despite winning only 48.6% of the two-party state-wide vote. The outcome of the case is uncertain; courts don't want to get into the job of examining the fairness of district boundaries because judges fear they'd be inundated with hard cases. And unless they can come up with unusually clear standards, that's almost certainly true.
Michigan voters are proposing to take the line drawing away from the politicians.
As a Californian, I've seen this in action. The Congressional lines drawn here after the 2000 census amounted to another kind of gerrymander: an incumbent protection plan. Sitting Congresspeople and legislators avoided a fight by drawing boundaries that tended to keep them in office, regardless of party. This worked fine for the politicians. In 2004, just three of the 53 districts were won with less than a 60 percent majority. Only one Congressional seat changed party during that decade!
Significant numbers of California voters felt disenfranchised, so we passed Prop. 11 in 2008 followed by Prop. 20 in 2010, giving responsibility for reapportionment to a Citizens Redistricting Commission. The result was a significant shakeup among Congressmembers; some members retired after losing their safe districts while quite a few seats became more competitive. Both political parties hate losing their chance to draw their own seats, but we probably have somewhat more competent and attractive politicians among the new crop. So far, this electoral gimmick seems to work for more representative governance.
California currently elects 39 Democrats and 14 Republicans to Congress -- not an unreasonable split given the partisan lean of state voters. Democrats hope to win even more Congressional races in 2018 since seven districts currently held by Republicans voted for Clinton in 2016. Republicans are targeting at least one highly competitive seat they hope to flip. When districts are reasonably fairly drawn, such changes become possible.
Michigan has 14 Congressmembers, currently divided 9 Republican and 5 Democratic. None of the incumbents had less than a 12 percent margin of victory in 2016 -- that is, none of the seats was competitive between the parties. Yet the state as a whole could hardly be more competitive. Donald Trump won Michigan with 47.50 percent of the vote to Clinton's 47.27. It seems very likely that a non-partisan redistricting commission could provide more fairness to the choices that are offered to Michigan voters. Voters Not Politicians collected their initiative signatures with volunteers hardly any support from established political players! This effort has the feel of a movement.