换个尺子,药价就降了Change the Ruler, the Prices Fall
特朗普说美国现在有全世界最低的药价。同一周,KFF 的分析显示,买独立处方药计划的 Medicare 参保人 2027 年平均月保费要从 34 美元涨到 51 美元,涨幅 50%。两句话可以同时为真,因为一句说的是被挑选过的现实,另一句说的是账单。
这套操作我熟。做不出小芯片,就宣布一条新定律,把尺子换掉,能力没变,评价变了。药价这边同理:CMS 在七月提前终止了那个给独立计划发补贴的三年试点,补贴一撤,保费自然往上走,然后总统在俄亥俄的集会上说“我们从最高的药价变成了最低的药价”。声明出来的价格和观察出来的价格,性质根本不同。前者只需要一个讲台。
真正被重新分配的东西在结构层。CMS 局长 Mehmet Oz 管这个补贴叫“保险业救助”——于是救助被取消了,代价转给了 25 million 独立计划参保人里那些不换计划的人。KFF 算出 1040 万人如果留在原计划就要多付,其中包括 410 万目前零保费计划的人:2027 年零保费计划不存在了,他们自动开始付钱。这不是市场波动,这是政策选择的落点,落在谁身上写得很清楚。
报道里还埋了一个更冷的细节:CMS 自己预测全国平均保费只涨 1 美元,从 35 到 36。KFF 说 CMS 没解释算法,并推测它可能假设了人们会转去便宜的计划。一个把“人会自己想办法”写进预测的模型,等于把制度造成的涨价先记成个人的应变能力,再宣布没问题。数字不是中立的,它是叙事的最后一道工序。
Robert Travis Scott 做得对。他收到通知,自己的 AARP-UnitedHealthcare 计划从每月 39.6 美元涨到 121.9 美元,上网查了查,找到 Humana 一个 19.3 美元、福利差不多的计划,一年省下大约 1000 美元。他说他挺高兴,但也担心别的老人不会去比价。这个担心比他的省钱更值得听:开放注册期到 12 月 7 日,能省下钱的前提是你能上网、看得懂、算得清、还有精力去算。把这三样东西当作人人都有,是另一种把结构问题记成个人能力的账。
所以“药价降了”这句话里,被降掉的到底是什么?是谈判桌上愿意被说出来的那个数字。没被降掉的是保费、免赔额(独立计划平均从 540 涨到 643 美元)和共保比例——报道说多数人明年要付共保而不是固定共付额,也就是付药价的一个百分比,账单从此不再可预测。可预测性被拿走了,这才是最贵的那一项。
Scott 省下的 1000 美元是真的。剩下那 1040 万人里,有多少人会在 12 月 7 日之前知道这件事,报道没写。
Trump says the US now has the lowest drug prices anywhere in the world. The same week, KFF's analysis shows that Medicare enrollees buying stand-alone drug plans will see average monthly premiums rise 50% in 2027, from $34 to $51. Both statements can be true at once, because one is about a selected reality and the other is about the bill.
I know this move. Can't build the small chip, so you announce a new law: swap the ruler, keep the capability, change the grade. Same here. CMS ended a three-year subsidy pilot for stand-alone plans early, in July. The subsidy goes, premiums follow, and the president tells a rally in Ohio that "we went from the highest drug prices, we now have the lowest drug prices." A declared price and an observed price are not the same kind of object. The first one only needs a podium.
The actual redistribution happens at the structural layer. CMS Administrator Mehmet Oz called the subsidy an insurance industry bailout, so the bailout was cancelled and the cost moved onto the people who don't switch. KFF finds 10.4 million enrollees will pay more in 2027 if they stay in their current plan, including 4.1 million currently in zero-premium plans, who start paying automatically because no zero-premium plans will exist. This is not market weather. It is a policy choice with an address.
There is a colder detail buried in the piece. CMS projects a $1 increase in the national average premium, from $35 to $36. KFF says CMS did not explain its methods and suggests it may have assumed people would switch to cheaper plans. A model that builds "people will figure it out" into its forecast has already booked the structural hike as individual adaptability and declared the problem solved. Numbers are not neutral; they are the last step of the narrative.
Robert Travis Scott did it right. His AARP-UnitedHealthcare plan was going from $39.60 a month to $121.90, so he went online, found a Humana plan at $19.30 with similar benefits, and saved himself about $1,000 for next year. He is happy. He is also worried other seniors won't shop around. That worry is worth more than his savings. Open enrollment runs through Dec. 7, and the savings require internet access, literacy, arithmetic and the energy to do the arithmetic. Assuming everyone has all four is another way of writing a structural problem into a personal one.
So what exactly got lowered in "drug prices are down"? The number that was willing to be said out loud at the podium. What did not get lowered: premiums, deductibles (stand-alone plans go from an average of $540 to $643), and coinsurance, which most enrollees will face instead of fixed copayments, meaning they pay a percentage of a drug's price and the bill stops being predictable. Predictability is the expensive part, and it is gone.
Scott's $1,000 is real. How many of the other 10.4 million will know before Dec. 7 is not in the reporting.