As Maryland's General Assembly considers proposals for new congressional districts this week, plenty of people are upset about how the new lines might affect election outcomes. Few, it seems, have noticed that by gerrymandering voting districts, politicians have also gerrymandered our free press and other forms of political speech.

The fact that incumbents use redistricting to make the world safe for incumbency is not surprising. Reporters, who know how the gerrymandering manipulates election results, may scoff at the suggestion that news coverage is also being manipulated. But the truth is that most election coverage is "horse race" coverage. It's about who's ahead and who's behind; who's surging and who's fading. It's also about making pundits look smart, so news outlets don't just report what the candidates are saying — they also predict how the races will end.

Unfortunately, those predictions often determine whether a particular race gets covered at all. And thanks to gerrymandering, the vast majority of congressional races across the country produce no horse race and therefore no coverage.

I write from experience. In 2010, I was the Libertarian challenger to U.S. Rep. Chris Van Hollen in Maryland's 8th District. Our district was custom-designed to go Democratic back in 2002, so reporters had every reason to consider the race uncompetitive. Indeed, The New York Times estimated Mr. Van Hollen's "probability" of re-election at 100 percent. (It's a good thing the statisticians at the Times don't make their living underwriting life insurance.) Even I expected to lose by a wide margin, and the voters validated that prediction.

I also expected to receive less press coverage than Mr. Van Hollen's Republican challenger. But there I was wrong. I received just as much coverage as the Republican candidate, because neither of us received any. I was not on the outside of a two-party race looking in; I was in a non-race, a race of which the press took almost no notice whatsoever.

In fact, not even Mr. Van Hollen received any coverage about the 8th District race. As the leader of the Democratic Congressional Campaign Committee, he was on television almost every day, fielding questions about the national horse race. He just wasn't required to field any questions about his own race or his own record. That race wasn't close, so that race wasn't covered.

The problem is not that the lack of coverage affected the outcome. The problem is rather that voters in the 8th District were entitled to a real public debate about the issues the country faces, and they didn't get one. Gerrymandered district lines succeeded not just in making the outcome a foregone conclusion but in suppressing the robust political discourse that actually shapes public opinion — and ultimately affects policy at least as much as the election results do.

Gerrymandered districts also discourage other kinds of political discourse. In Maryland's 2nd District, no community organization was willing to schedule a debate among the candidates for the seat held by Rep. C.A. Dutch Ruppersberger. But don't blame the League of Women Voters; blame gerrymandering. Debate-hosting groups like the league are normally organized along community lines, by city or county. But there is no city or county that is even mostly in Maryland's 2nd District. Instead, the district is cobbled together from relatively small pieces of four different political subdivisions (and Gov. Martin O'Malley now proposes to add a fifth). As a result, no community organization considers Mr. Ruppersberger "their" congressman during election season. He's the congressman from nowhere.

It may be that news coverage alone can't make gerrymandered districts competitive. But choosing a winner isn't the only function of an election. Election campaigns are the primary means by which reigning political orthodoxies can be challenged. New views, whether about foreign wars or gay marriage or drug legalization, generally need to appear somewhere as the views of a small minority before they can become majority views. And those challenges to the received wisdom become more and more important as the seats become "safer" and the incumbents become more entrenched. It is nearly always from small minorities raising new ideas that majorities learn how to govern better.

There are excellent alternatives to partisan gerrymandering. Some are completely objective, based entirely on mathematics and population density. Others involve judicial or nonpartisan commissions, which leaves room for human tinkering, for better or worse. These alternatives should be thoroughly explored by the legislature. Meanwhile, it's up to us to realize that we can deny the gerrymanderers much of their power if we collectively refuse to be corralled into the voting blocs they've pre-arranged for us. Paying more attention to all the races, regardless of their predicted outcome, would be a good start.

When incumbents gerrymander, they don't just sew up elections; they also shut down debate. If the press can't stop politicians from gaming the redistricting process, it can at least refuse to play along at election time.

Mark Grannis is a lawyer in Washington, D.C. and member of the executive board of the Libertarian Party of Maryland. He was the 2010 Libertarian candidate in Maryland's 8th congressional district. His email is mgrannis@gmail.com.