✦   ✦   ✦

breaking news

News, read through The Primal Race
← 全部评论 · all commentary

中期选举不是投票,是共谋的续约The Midterms Are Not a Vote, They're a Lease Renewal on Explanation

国际 元暴力 · 结构层 · 文化层 The Guardian ↗ 2026-11-01 § 链接
新闻摘要美国中期选举于11月3日举行,将改选众议院全部435席、参议院100席中的35席以及39个州长职位。共和党在众议院仅占微弱多数,民主党只需净增4席即可夺回控制权;参议院方面民主党需增4席。此次选举被视为对特朗普第二任期前两年的检验。
The newsUS midterm elections on 3 November will decide all 435 House seats, 35 of 100 Senate seats and 39 governorships. Republicans hold a narrow House majority, with Democrats needing a net gain of four seats for control; Democrats need four Senate seats too. The vote is seen as a referendum on Trump's second term.
两党争夺的从来不是方向,是解释权的租约
Both parties fight over the lease on explanation, not over direction.

这篇报道把美国中期选举写成了一场赛马:谁领先几个百分点,哪几个席位是战场,民主党赢了会怎样、共和党守住会怎样。赛马报道的默认前提是,跑道上的人都在为选民办事,只是办法不同。这个前提是假的。

真正在动的是元层。众议院435席全改选,参议院100席里改选35席,共和党53席对民主党与独立派合计47席,民主党要净增四席才能拿多数;众议院民主党要净增四席。这些数字决定的是未来两年谁握着调查权、传票权、听证会的排期权——也就是谁有权决定什么算“事实”、什么算“丑闻”、什么该被追问、什么该被忘掉。解释权的租约每两年续一次,这才是选举的实质。

报道里有一句很诚实的话:过去20次中期选举,总统所属政党有18次丢失众议院席位,1946年以来的90%。这不是什么民主的自我纠错机制,这是结构性的疲劳周期。选民惩罚在任者,不是因为换上来的人更好,是因为在任者已经把能许的愿许完了。2002年共和党逆势增席,靠的是9·11之后的战争叙事;2010年奥巴马被“shellacking”,2018年特朗普丢了众议院。同一台机器,换不同的操作员。

德州那场值得单独看。共和党为丑闻缠身的州检察长Ken Paxton砸了约2.1亿美元(来源:The Guardian),而自1994年以来没有民主党人赢过德州的全州选举,2024年特朗普在德州赢了近14个百分点。花两亿多在一个稳赢的州,这笔钱买的不是席位,是“谁配代表德州”这个叙事的独家代理权。钱在这里的功能从来不是说服摇摆选民,是给党内各方定价:谁出得起,谁就有资格定义党的脸。

报道问“选举会自由公正吗”,然后列了一串:司法部起诉各州要未删节的选民名册,拿去和联邦移民数据库交叉比对;一位联邦法官裁定该政策违法,理由包括权限越界和数据不准的风险;NAACP就计划部署的联邦选举监督员提起诉讼,认为在场本身就会吓退选民;还有邮寄投票规则的改动、对计票认证环节施压的可能。这些不是选举的“周边问题”,这就是选举正在被武器化的过程——从认知入口下手,先制造“选举不可信”这个前提,再让任何不利于自己的结果都落进这个前提里。

报道写到了代价,但写得很轻。ICE的强硬做法被民调显示“走得太远”,可能抹掉特朗普2024年在拉丁裔选民中的收获;生活成本常年居民调首位,民主党主打食品、住房、燃油、医疗的可负担性;伊朗战争给共和党添了麻烦。这些是人在承受的东西。但报道把它们排成议题清单,和“数据中心、AI、关税、贸易、堕胎权、投票规则、民主的未来”并列,仿佛它们是同一类东西。

真正缺席的是谁。这篇报道里没有一个普通选民的名字,没有一个被移民执法带走的人,没有一个因为新身份证件法而投不了票的人,没有一个在伊朗战争中失去家人的人。有的是候选人、金额、百分点、历史概率。这不是报道的疏忽,这是赛马框架的必然产物:跑道上只有马和骑师,看台上的人不需要名字。

两党都在共谋。民主党靠“反特朗普”这个负号集结,共和党靠“选举被偷”这个前提集结,双方都需要对方继续存在,因为对方的威胁是各自最有效的募款文案。报道结尾说尘埃落定后马上转向2028,没有喘息——“the land of the endless campaign”。永久的竞选,就是永久的解释权争夺,就是永久没有人真的为那8.2个百分点的领先后面那些具体的人负责。

11月4日早上特朗普还是总统。这句话报道开头就写了,写得对。它没写的是:不管谁赢,那个让90%的中期选举变成惩罚在任者、让两亿多美元买一个稳赢州的叙事、让选民名册变成武器的东西,还在原地。

This piece covers the US midterms as a horse race: who leads by how many points, which seats are battlegrounds, what happens if Democrats win and what happens if Republicans hold. The default premise of horse-race coverage is that everyone on the track is working for the voters, just by different means. That premise is false.

What is actually moving is the meta layer. All 435 House seats are up; 35 Senate seats are up, with Republicans holding 53 against 47 for Democrats and independents who caucus with them; Democrats need a net gain of four seats for a Senate majority, and four for the House. Those numbers decide who holds investigation power, subpoena power, the hearing calendar for the next two years — that is, who gets to decide what counts as a fact, what counts as a scandal, what gets pursued and what gets buried. The lease on explanation is renewed every two years. That is the substance of the election.

One honest line in the report: in 18 of the past 20 midterms, the president's party has lost House seats, a 90% rate stretching back to 1946. This is not democracy self-correcting. It is a structural fatigue cycle. Voters punish the incumbent not because the replacement is better, but because the incumbent has already spent every promise. Republicans gained seats in 2002 on the post-9/11 war narrative; Obama's Democrats took a "shellacking" in 2010; Trump lost the House in 2018. Same machine, different operators.

Texas deserves its own look. Republicans have spent about $210m to help the scandal-plagued attorney general Ken Paxton (source: The Guardian), in a state where no Democrat has won a statewide election since 1994 and where Trump won by nearly 14 points in 2024. Spending two hundred million in a safe state does not buy a seat. It buys the exclusive franchise on the narrative of who gets to represent Texas. Money here never functions to persuade swing voters; it prices the party's factions: whoever can pay gets to define the party's face.

The report asks whether the elections will be free and fair, then lists: the justice department suing states for unredacted voter rolls to cross-check against a federal immigration database; a federal judge ruling that policy unlawful, citing limits on its authority and the risks of inaccurate data; the NAACP suing over planned federal election monitors, arguing their presence could intimidate voters; changes to mail-voting rules; pressure on local officials to certify results. These are not side issues. This is the weaponization of the election in progress — entering through the cognitive gate, manufacturing the premise that elections cannot be trusted, so that any unfavorable result falls neatly inside that premise.

The report does note costs, but lightly. Polls suggest many voters feel ICE's hardline approach has gone too far, which could wipe out Trump's 2024 gains with Latino voters; cost of living tops opinion polls, with Democrats campaigning on food, housing, fuel and healthcare affordability; the war in Iran complicates things for Republicans. These are things people are carrying. But the report arranges them as an issues list, side by side with "datacentres, AI, tariffs, trade, abortion rights, voting rules and the future of democracy," as if they were the same kind of thing.

What is truly absent is who. Not one ordinary voter is named in this piece. Not one person taken by immigration enforcement, not one person blocked from voting by a new ID law, not one person who lost family in the Iran war. There are candidates, dollar figures, percentage points, historical probabilities. This is not an oversight. It is the necessary product of the horse-race frame: only horses and jockeys on the track; the stands do not need names.

Both parties are complicit. Democrats rally around the negative of Trump; Republicans rally around the premise that elections are stolen. Each needs the other to keep existing, because the other's threat is the most effective fundraising copy either has. The report closes by saying there will be no respite once the dust settles — attention turns straight to 2028, "the land of the endless campaign." The endless campaign is the endless fight over explanation, and the permanent absence of anyone actually accountable for the specific people behind that 8.2-point lead.

On the morning of 4 November, Trump will still be president. The report says so at the top, and it is correct. What it does not say: whoever wins, the thing that turns 90% of midterms into punishment of the incumbent, that makes $210m buy a safe state's narrative, that turns voter rolls into a weapon — that thing is still standing exactly where it was.