NPR ran a story Monday on a poll about proportional representation — replacing winner-take-all House districts with seats shared out by vote share. The headline: "Voters are sick of the 2-party system. A new poll shows they're open to big changes." Now paragraph two, where the poll's birth certificate lives: it was "commissioned by supporters of election reforms and shared with NPR." * Credit first, because NPR earned some: the disclosure is right there, high up, in plain words — the poll comes from the advocacy group Fix Our House and the think tank New America, and NPR labels the latter "liberal." Plenty of outlets would have laundered that into "a new survey finds." But disclosure is where the good news ends, because everything built on top of the poll treats it like neutral evidence anyway. The headline states the commissioners' conclusion as fact. "Shared with NPR" is the polling version of the soft launch — an interested party hands its own numbers to one sympathetic outlet. If that delivery route sounds familiar, it's the same one the White House vaccine fact sheet took through a friendly publication this very afternoon. Different living room, same doorbell. * Then there's the star witness. The poll's chief interpreter, quoted four times, is Lee Drutman — a senior fellow at New America, the think tank that commissioned the poll. His verdict on his own questionnaire: "It really did shock me how supportive people were." A man cannot be shocked by results he paid to generate; that's not analysis, that's a product demo. The only outside voice, Pew's Jocelyn Kiley, is quietly pumping the brakes in the basement — reforms "move from theory to practice" and "hit some resistance" — while the headline does 90 upstairs. * And the numbers deserve a second look, which the piece never gives them. Half of 1,000 respondents support the idea, "nearly a quarter" oppose, and a full third aren't sure — reasonable, for a concept the article itself calls "fairly obscure." The flashiest stat, "6 in 10 Trump voters who have an opinion" back it, walks through the classic trapdoor: delete everyone who shrugged, then poll the survivors. And nowhere — nowhere — does the story give a margin of error, field dates, or how the poll was even conducted. NPR knows how to do this; its own NPR/PBS News/Marist polls publish every one of those numbers like clockwork. The standards exist one desk over. They just weren't invited. * To be clear, proportional representation might be a perfectly good idea — 500 scholars wrote Congress saying so, and that letter is real news. The problem isn't the cause. It's the supply chain: when the commissioner is also the interpreter, and the headline is also the conclusion, what ran on Monday isn't coverage of public opinion. It's distribution of an advocacy product, with a disclosure sticker doing the work of a firewall. * We praised NPR twenty-four hours ago for fitting a whole uncomfortable truth into one sentence about private prisons — so consider this the other half of honest grading. Even a good desk can take a handout when it shows up pre-interpreted, pre-quoted, and pre-headlined. The poll didn't need to be smuggled past NPR's standards. It was walked in the front door, wearing them.