Reason访谈速读：统计数字怎么变成宣传

这期 The Reason Interview 是一场关于「数字为什么会骗人」的访谈。嘉宾 Aaron Brown 不是在说统计学没用，而是在拆一个更日常的链条：研究者做出一个脆弱但吸睛的数字，期刊和大学新闻办公室把它包装成标题，媒体再把它送进公共讨论，最后政策、诉讼和舆论都开始围着这个数字转。节目围绕 Brown 的新书 Wrong Number: How to Extract Truth From a Blizzard of Quantitative Disinformation 展开，适合想训练「看数据标题先多问两句」的读者。1

Nick Gillespie 开场介绍 Brown 的背景：他曾任 AQR Capital Management 董事总经理和首席风险官，做过衍生品交易，也长期写 Bloomberg 专栏和 Reason 的 Wrong Number 视频。Brown 的新书副标题是 How to Extract Truth From a Blizzard of Quantitative Disinformation，直译就是「如何从数量化虚假信息的暴风雪中提取真相」。1

这期节目的核心问题是：为什么很多明显站不住脚的数字，会在大学、期刊、媒体和政策讨论里一路通行？Brown 给出的解释很刺耳：有些研究「太无能，不像阴谋；又太像阴谋，不只是无能」。他形容这种数字往往是一杯混合饮料：一份无能，一份意识形态，再加上八盎司「没人关心」的气泡水。

这句话是整期的入口。Brown 不是把所有错误都归咎于造假，也不是说所有研究者都坏。他关心的是一个激励系统：研究要新奇，期刊要影响力，大学媒体部门要新闻点，记者要标题，公众要简单结论。于是，一个统计上很脆弱的数字，只要足够符合某种情绪，就能跑得很快。

Brown 说，科学研究本来像一块砖。砖只有放进建筑结构里才有意义，也就是要能被后续研究验证、修正、反驳和扩展。问题在于，有些论文不是用来建房子的，而是「用来砸窗户的砖」：它们不服务一个长期研究计划，只服务一次舆论冲击。

这个比喻解释了他为什么反复谈媒体。Brown 批评的不是普通读者没有统计训练，而是主流媒体的科学记者没有执行基本怀疑。他说，面对这些研究，记者至少该像买二手车时一样多问几句。这个标准很低，但很多报道没有做到。

节目里最反复出现的术语是「correlation is not causation」，相关不等于因果。Brown 的重点又比这句话多一层：很多研究甚至连相关都很弱，却通过选变量、换分母、换图表尺度、挑标题，把微弱或错误的关系做成了确定感。

第一个案例是 USAID。Brown 讨论了一篇发表在 The Lancet 的研究，该研究声称 USAID 在 21 世纪拯救了约 9000 万人的生命。Brown 说，最荒谬之处在于，按这项研究的口径，USAID 贡献了全球死亡率下降的 114%。这意味着其它援助项目、医学进步和 GDP 增长不仅没有贡献，反而好像在「杀人」，否则 USAID 不可能超过总下降量。1

这不是细节误差，而是量级上说不通。Brown 还指出，全球死亡率下降的大头发生在中国等几乎没有接受大量 USAID 援助的国家。如果一个研究提出这种结论，正常反应应该是要求极强证据。可他读到研究后发现，对方连「很差的证据」都没有。

为什么这种数字会被传播？Brown 的判断是，这类研究有顺风：有人希望它是真的，也有人觉得它「应该」是真的。它符合某个政治立场，媒体又需要一个大数字，最终就变成了容易复述的标题。读者看到 9000 万，只会觉得这是一个巨大的数字，很少有人停下来问：全球对应指标总共下降了多少？这个数字怎么可能超过 100%？

这也是 Brown 给读者的第一个实用提醒：看到大到离谱的数字，先别急着问立场对不对，先问分母是什么、总量是多少、它有没有违反常识边界。

气候变化部分最容易被误读。Brown 明确说，他不是气候变化否认者。他接受人类环境足迹很大，而且还在变大；他也认为这会带来长期问题。但他反对把渐进曲线加工成更吓人的图表，因为这种做法会让公众进入短期恐慌，转而支持冒险、仓促、可能适得其反的方案。

节目谈到「hockey stick graph」，也就是冰球杆图。它通常表现为前面长时间缓慢变化，最后突然陡升。Brown 批评某些气候传播者为了得到这种形状，会拼接不同数据、改用比例或选择更戏剧化的展示方法。他说，只要用高中数学课那种正常画法呈现同一批数据，结论仍然显示气候问题存在，但它不再像冰球杆。

一个具体例子是「创纪录高温日」和「创纪录低温日」。Brown 说，美国的创纪录高温日并没有像某些图表暗示的那样暴增，1930 年代甚至更多；真正显著变化的是创纪录低温日几乎消失。这本身很重要，意味着物种分布、农业和居住安排会受影响。问题在于，传播者有时把「高温纪录日 / 低温纪录日」做成比值。分母接近零时，比值会暴涨，图就会变成红色陡坡。1

这里的差别不只是技术细节。说「低温纪录日正在消失」和说「高温纪录日变成过去的 20 倍」，会触发完全不同的公众情绪。前者让人准备长期适应，后者让人以为灾难正在垂直逼近。Brown 和 Hank Green 的争论，也围绕这一点展开：Green 认为 Brown 用不那么吓人的图表是一种操纵；Brown 则认为，正常展示数据本来就不该被叫作操纵。

第三个案例是大麻和心脏事件风险。Brown 批评一类研究把「过去 30 天是否使用大麻」和「是否曾经发生过心脏事件」放在一起看，然后暗示大麻使用提高心脏风险。问题很简单：如果一个人心脏事件发生在很多年前，过去 30 天的大麻使用不可能是原因。

这听起来像低级错误，但它很常见。Brown 还提到，这类研究常依赖电话调查，让受访者回答大量医学问题。研究者声称这种数据可靠，并引用其它研究支持；Brown 去看那些引用，发现它们实际说的是这类数据很不可靠，比如自称有心脏事件的人里相当一部分并没有医学记录支持。

更深一层的问题是变量筛选。研究者手里可能有上百个问题、二十多个协变量。如果他们先看数据，再挑一个惊人、符合传播需求的变量发表，读者看到的就不是严谨检验，而是事后挑选。Brown 把这种做法放在枪支研究里也讲了一遍：有些变量影响更大，却没有进入报道；最后被拿出来的，是最符合媒体叙事的变量。

真正更好的做法是什么？Brown 说，要做长期面板研究。也就是跟踪一大群人在同一个医疗系统里的完整记录，看看 20 岁时的行为和 50 岁时的心脏结果是否相关。这样的研究成本更高，结论往往也更弱、更不刺激，所以不容易得到头条。

节目后半段有一个很具体的交通案例：Chinatown buses，也就是从纽约唐人街一类街边站点发车的低价城际巴士。Brown 说，这类巴士最早由移民小公司运营，票价低、发车灵活，很快威胁到传统长途客运公司的利润。后来媒体和监管讨论里出现一种说法：街边巴士比传统 Greyhound 等客运危险七倍。

Brown 拆这个数字时，问题一层比一层荒唐。首先，触发恐慌的重大车祸并不是街边巴士，而是一辆赌场巴士。其次，统计机构没有掌握乘客里程或乘客数，只用公司拥有的巴士数量做分母。更离谱的是，他们不是把所有街边公司合并计算，而是逐家公司算事故 / 车辆比例，再平均。于是，一个只有一辆车、发生过一次致命事故的公司，就贡献了 100% 的极端比例，拉高了整体结果。

Brown 给出的「数字警报器」很实用：如果某项服务真的危险七倍，你为什么很少听说它的致命事故？研究有没有给绝对风险？有没有合理分母？他补了一句：从纽约坐巴士到波士顿的风险，大约相当于过几条街。即使某个比值看起来大，绝对风险也可能很小。

这个案例提醒读者，比例不是越大越有意义。没有基线、没有绝对风险、没有分母来源，所谓「七倍」可能只是一个被设计出来的恐惧形状。

Brown 对 The Lancet 的批评很重。他称它曾经很有声望，但现在太政治化，不能简单信任。更广泛地说，他认为当前学术发表系统有结构性激励：政府资助研究，研究者付费发表，期刊再向机构和政府收费让人阅读；论文发表又是学术职业晋升的基础。

在这个系统里，就算认真学者也会被迫玩游戏。好研究不一定够多，职业却需要持续发表，于是大量质量一般甚至很差的论文进入管道。同行评审也不总是独立纠错，因为评审者和作者未来还会互相评价、互相影响。

Brown 倾向于「发表后评审」：先把研究放到公开平台上，让别人验证、反驳、复现、扩展。如果没有人愿意基于你的研究继续做，也没有人愿意认真反驳，那它很可能没有说出多少东西。

预注册假设也被提到。所谓预注册，就是研究者在看数据前先公开说明：我要检验什么假设，用什么方法，最后无论结果支持还是反对都报告。Brown 承认这有帮助，但他不认为这是根本解法。根本问题还是激励：如果新奇、惊人、政治上有用的结论更容易换来职位、经费和媒体，研究者总会找到绕过规则的方法。

访谈后半段转向 Brown 本人。他说自己是 libertarian，但更准确地说是怀疑主义者。他怀疑人们知道自己在做什么，怀疑政府大型自上而下项目能按承诺运转。他的哲学底线是：主动对无辜者施加暴力总是错的；如果能不用这种方式运行社会，就应该尽量不用。

他的家庭经历解释了这种敏感。Brown 的父亲在 1950 年代因反对大学忠诚宣誓一类做法，被卷入反共政治压力，原本在华盛顿大学的工作一度被撤回。这个经历让 Brown 从小看到多数派和制度压力如何伤害个人。

另一个底色是赌博。Brown 很早迷上数字、概率和模式，14 岁时就走进地下酒馆打扑克。他说自己在牌桌上的最大优势，不只是数学好，而是他上桌就是为了赢钱；很多人其实有别的动机。这个经历也影响他看预测市场和体育博彩：下注不是万能真理，但它让少数人检验自己是否真的能打败群体判断。能赢钱的人学到一课，不能赢钱的人也会学到另一课。

这和整期主题是连在一起的。Brown 不迷信专家，也不迷信群众。他关心的是反馈：你的判断有没有被真实世界检验？你的数字有没有被别人复现？你的图表换一种画法后，结论还站不站得住？

这期最值得带走的不是某个单独案例，而是一套读标题时的慢动作：

Brown 的底层建议很朴素：数字不是越精确越可信。很多时候，一个看起来严整的统计结论，真正需要的是常识先拦一下。

以下为本轮基于完整音频生成的 ASR 转录。音频没有可靠说话人分轨，因此统一标注为「未分轨讲话者」；时间轴按约 10 分钟一段保留。英文为 ASR 原文，中文为对照译文。原始节目页见 How Statistics Become Propaganda。

未分轨讲话者（英文原文）：Our guest today is Aaron Brown, who is the former managing director and chief risk manager at AQR Capital Management. He's a veterans derivatives trader, and he's the author of several books, the most recent of which is Wrong Number, in which he argues that many of our deepest assumptions about chance causation and prediction are misguided. The book is just out. The subtitle of Wrong Number is How to Extract Truth from a Blizzard of Quantitative Disinformation. Aaron writes a column for Bloomberg, and he also writes and appears regularly in a series of reason videos called Wrong Number. The most recent of which has over a quarter of a million views, I believe, and was the object of a long impassioned critique by Hank Green, a very popular, what used to be called a vlogger. And we're going to get into that in a bit. But Aaron,start off by explaining, in your words, why you wrote this book. Why Wrong Number and why now? I've been interested in wrong numbers since I was a kid. And the field got pretty crowded in the 21st century, and a lot of people started doing it. We had a Stanford professor, John Ioannidis, wrote a famous paper, Why Most Published Research Findings Are False. Now we have Retraction Watch and places like this, and they go after fraud and competence and whatever. I started doing the wrong numbers for a reason without a lot of expectation, and they proved extraordinarily popular. And I realized there's an intersection here that really interests me. Results that are too incompetent to be a conspiracy and too conspiratorial to be an incompetent. And the way I put it is you've got, you know,one shot of incompetence, one shot of conspiracy, and that eight ounces of nobody cares sparkling water. Why does nobody care? Why am I the only person writing this book? You know, these are the things. All these numbers we're seeing in newspaper headlines that are driving policies, making court decisions, complete nonsense. And you can tell. I mean, you just look at the number and tell that it's ridiculous.and nobody cares. Well,you can tell,right? Yeah,you can tell too. You can tell, believe me. Yeah, and obviously the book helps. I mean, it provides a framework. There's statistics in it,but it's mostly,you know, you say like just kind of become a critical thinker and think through certain things. But one of the ways that you talk about this, and then we're going to talk about a couple of specific examples that you can work through. But isn't part of it that more and more academics in particular are publishing more and more papers? They have better and better kind of media relations outfits that put–they take a study and they say this finding is amazing. And then journalists kind of lap it up uncritically and then it goes out and then it becomes part and parcel of everyday conversation. That's exactly it. And that's one of the threads. And the way I put it is, you know,a study, any scientific study is a brick, and it's only useful when it's part of a structure. But some people are writing papers that are bricks to throw through windows. And their university media departments are a big part of the blame. And the reporters, I mean, science reporters for a Wall Street Journal,a New York Times, a major media, these are smart people who should be critical about it, but they buy it like a late-night television mutual fund or something. They don't apply the basic skepticism you would to a used car. Let's go through a couple of examples that you use in the book. One is the oft-quoted claim that USAID saved something on the orders of 90 million lives. And so when funding for it was polled, It meant many, many more people died. Would you recap the main study or article that you were kind of rebutting and explain where it went wrong? Yeah.so this is published in The Lancet, which is the oldest and most prestigious medical journal in the world. They boast on their website about how they have this incredibly rigorous review process with statistical experts. They seek out people with contrary opinions and all this. But this paper claimed, I mean, the headline number here is they claim that USAID was responsible for 114% of the entire global decline in mortality in the 21st century. And USAID is only a fraction of total foreign aid. And apparently all these other foreign aid programs plus medical advances, GDP growth, are killing people because they have to for USAID to save them. And the other thing about this is the large majority of these declining global mortality was in countries, China, which doesn't get significant USAID aid. So you would think that people would just look at this and say, okay, you need extraordinary evidence for this. And then when you actually dig down to the study, you find out they don't even have bad evidence for it. So is it just that that's a story that people,you know, are at the very basic level? I mean, can you kind of go back to the researchers? Why would researchers put forward, you know, a study that just on its face doesn't make a lot of sense? Why would the university hype it? Why would journalists buy it and kind of circulate it? Yeah,well,okay, so this is a brick-through-the-window study,right? It wasn't part of any ongoing research project. Nobody tried to validate it. Nobody tried to refute it. It was just, you know, stuck out there. I don't know the researchers personally, but I assume their motivations were primarily political. They thought this was true. They thought it should be true. They were kind of uncritical, so kind of a combination of incompetence and...ideology, putting it out. And I don't really, okay,so somebody's going to do that, right? You're always going to have, if this would come out from an interest group,from a lobby group or something, you'd say,fine. But nobody put any, you know,the Lancet,all the, you know, was carried in Wall Street Journal, New York Times, AP,BBC, whatever.it and not one of these people. If you're a science reporter for one of these places,you should know, gee, global mortality only dropped to 79 million in the 21st century. You can't save 90 million lives. You know, I was an English major, so I'm trying to do the math in my head. I don't know. I was lost when you were talking about 114%,you know, before. I think most people read this and it's just big number. They could have said 9 million, they could have said 900 million, and it's just a big number. Well, before we go on to another case study, is there something that you can kind of–that predicts when an obviously dubious finding catches on? Is it–Is it just that it's not the study that deserves to be published, but it's the study we want? Or is there anything predictable about why something becomes viral? I mean,well, it's pretty clear there are tailwinds for something like this. So there are people who want this to be true. But not everything that has tailwinds gets there. In climate change, we had the 12 years to climate disaster. I'm sure that's back from 2018.So yeah. of the way through the period by now. But it just–that's a number. It catches on. Everybody,12 years, 12 years. Nobody really thinks much about what it means. Well, talk about that because you discuss that in the book. And we wrote a lot about that at Reason. One of the things that was fascinating, a number of people started saying in 2018 that,okay,well, we have 12 years.to completely decarbonize the global economy, otherwise everything dies. I mean, Alexandria Ocasio-Cortez was probably the best-known politician who was pushing that line. That idea is not even in the paper. Were you able to identify how that language or that phrasing–got into the kind of media bloodstream. Yeah, people needed that, and people remember the 12 years. They don't agree on what the 12 years is. They don't agree on when it started, right? So you had people still quoting it six years after the paper came out. And it wasn't,some people said it was six, you know, 12 years until we're all dead. Some people it's 12 years until we,you know, have to make these fixes. And the paper said something completely different. The authors of the paper, by the way, very quietly walked back some of the, not, they didn't walk back their paper. They criticized some of this coverage. That got no attention at all. Well, let's talk about the video that you did for a reason that raised the ire of Hank Green. It came out in February of this year,and it was titled, These Climate Change Charts Are Wrong. Here Are the Real Versions. Can you summarize what you were talking about in that video? Sure. So some of you are aware of the term hockey stick graph, which was popularized by Michael Mann. I'm sorry, he didn't popularize the term. He did the first hockey stick graph. And a hockey stick graph is slowly increasing that goes up like the blade on a hockey stick. I don't know why they picked hockey stick for that, but that is what they call them. And since then, it's a cottage industry among climate change activists to come out with these charts. And the data just don't support it. You know,climate change, if you're...It is a gradually increasing curve and it's scary if you extrapolate

未分轨讲话者（中文译文）：今天的嘉宾是 Aaron Brown。他曾任 AQR Capital Management 董事总经理和首席风险官，是资深衍生品交易员，也写过多本书。最新一本是《Wrong Number》，他在书里说，我们对概率、因果和预测的很多基本假设都是错的。这本书刚出版，副标题是「如何从数量化虚假信息的暴风雪中提取真相」。Aaron 给 Bloomberg 写专栏，也经常参与 Reason 的 Wrong Number 视频系列。最近一期视频大概有 25 万以上播放，还引来了 Hank Green 很长、很激烈的批评。我们稍后会聊到这件事。Aaron，先用你自己的话说说，为什么写这本书？为什么是 Wrong Number，为什么是现在？Brown 说，他从小就对错误数字感兴趣。进入 21 世纪后，这个领域变得拥挤起来，很多人都开始研究。斯坦福教授 John Ioannidis 写过著名论文《为什么大多数已发表研究发现都是假的》，现在还有 Retraction Watch 这样追踪造假、无能等问题的机构。他开始给 Reason 做 wrong numbers，本来没抱太大期待，结果非常受欢迎。他意识到自己感兴趣的是一个交叉地带：有些结果太无能，不像阴谋；又太像阴谋，不只是无能。他打了个比方：一份无能，一份阴谋，再加八盎司「没人关心」的气泡水。为什么没人关心？为什么只有我在写这本书？那些出现在新闻标题里的数字正在推动政策、影响法院判决，却完全是胡扯，而且你一看数字就能看出荒谬。主持人说，这本书提供了框架，里面有统计学，但更多是在训练批判性思考。Brown 接着解释，问题之一是学术界发表越来越多论文，大学媒体部门也越来越擅长把研究包装成「惊人发现」；记者不加批判地接收，再把它变成日常谈资。Brown 说正是如此。任何科学研究都是一块砖，只有放进建筑结构里才有用。但有些论文是拿来砸窗户的砖。大学媒体部门对此负有很大责任；《华尔街日报》《纽约时报》这类媒体的科学记者很聪明，本该怀疑，却像半夜电视购物里买共同基金一样照单全收，连买二手车时会用的基本怀疑都没有。随后他们谈到 USAID 拯救约 9000 万生命的说法。Brown 说，这篇文章发表在《柳叶刀》上，而《柳叶刀》自称有严谨评审、统计专家和反方意见。但这篇论文的头号数字声称，USAID 对 21 世纪全球死亡率下降贡献了 114%。USAID 只是全球援助的一部分；如果它能贡献 114%，那其它援助、医学进步和 GDP 增长好像都在杀人，因为只有这样 USAID 才能「拯救」那么多人。更何况，全球死亡率下降的大部分发生在中国等几乎没有获得大量 USAID 援助的国家。正常人应该要求极强证据，但读完研究后会发现，对方连很差的证据都没有。主持人问，为什么研究者会提出一个表面上就说不通的研究，大学为什么会宣传，记者为什么会传播？Brown 说，这就是砸窗户的砖，不属于任何持续研究项目，没有人试图验证，也没有人试图反驳，只是被扔出来。他不认识作者，但猜测动机主要是政治性的：他们认为这是真的，也认为它应该是真的，于是不够批判，把无能和意识形态混在一起。Brown 说，如果这是利益团体或游说组织发布的东西，也就罢了；问题是《柳叶刀》和《华尔街日报》《纽约时报》、美联社、BBC 等媒体都带着它跑，而没有一个科学记者指出，21 世纪全球死亡率只下降了 7900 万，你不可能拯救 9000 万。主持人自嘲自己是英文专业，听到 114% 已经算不动了。Brown 说，多数人看到的只是「一个大数字」，9 百万、9 千万、9 亿都只是大数字。之后他们转到「可疑发现为什么会传播」。Brown 说，像 USAID 这种说法有顺风，因为有人希望它为真。但有顺风并不保证传播。在气候变化话题里，「12 年后气候灾难」也是类似数字。大家记住了「12 年」，但没人同意 12 年到底指什么：有人说 12 年后大家都死，有人说 12 年内必须完成修复，而原论文说的是别的东西。论文作者后来很安静地批评了部分报道，但没人注意。然后他们谈到 Brown 给 Reason 做的气候图表视频，题为「这些气候变化图表是错的，真正版本在这里」。Brown 解释了「冰球杆图」：前面缓慢上升，最后像冰球杆的杆面一样陡升。Michael Mann 做了第一张这类图。后来气候活动者形成了一门小产业，不断做这种图，但数据并不支持。气候变化如果按世纪外推，是逐渐上升的曲线，也确实令人担心；我们要么希望它改变，要么主动改变，否则 100 年后会有严重问题。但传播者想要冰球杆形状，仿佛一切到 1970 年前都还好，之后突然爆表，12 年后就完了。Brown 说他们用各种方法改变数据。Mann 原图最大的问题之一，是把一条使用完全不同数据的红线贴在末端，形成「杆面」，像拼接怪物图。Brown 在视频里取了三张近期冰球杆图，拿同一批数据按高中数学课里的正常方式画，结果显示确实有变化，但不是冰球杆。

未分轨讲话者（英文原文）：this out a century. We have to do something. We have to hope it changes or we have to change it or in 100 years there's going to be some serious problems. But they want a hockey stick. Everything was fine up until 1970 and suddenly we're off the charts and in 12 years we're going to be. We're just cooked, literally. And so they do all kinds of things to change the data, to do it what Michael Mann did in the original graph. There was a lot of problems with it, but the biggest one was he had this slowly increasing graph, and then he took this red line with completely different data, unrelated to the chart, and he pasted it on the end to form the blade. It was a Frankenstein chart. Anyway, so I took some three of the more recent, hockey stick charts and just showed, okay, just take the same data and graph it the way you would,you know, in high school,you know, math class. And you see there's something there. We're not saying things are getting cooler, but we are saying it's not a hockey stick. Well, and you talked about how the way that you present data, if I'm remembering correctly, it was that in one of the studies or one of the charts that you were talking about, they used the ratio of hot days to cool days in order to create a kind of progression that looks scarier when that's not the right way to present the data. Could you explain that? Yeah, this is actually one some of you may have read Stephen Kunin's book. Sorry. Oh, Unsettled was the name of it. And he actually talks about the same chart. So you have the number of record hot days are basically been pretty steady over time. We are not seeing more record hot days and we're seeing fewer record hot days than we did in the 1930s in the United States. Globally, we have a few more. What we are seeing, and this is the really worrying thing, if you want to worry about climate change, worry about the fact Record cold days are practically disappearing. There are almost no record cold days. But that's not very scary to people. You're out shoveling snow and you're thinking, well, I wouldn't mind if it were five degrees warmer. What they do is they show the ratio of record hot days to record cold days, and if you don't want to scare anybody, mathphobes out there,whatever, but if you have a very small denominator, you can get very big ratios. And so you have this scary looking chart with these big red lines going up, and it suggests to people there are 20 times as many record hot days as there used to be. And they're not, they're about the same. So this is serious. It is a serious problem that record cold days are disappearing. We should worry about it. Why should we worry about that? Well, it is a major environmental change. It's gonna change species distribution, it's gonna change agriculture, it's gonna mean people have to move, things like that. I don't mean we should worry like it's gonna kill us all, it's an asteroid coming to kill us all, but it is something we should think about and either prepare for or do something about if we can. And for the record, you are not a climate change denialist. You're not particularly fond of terms like climate change or global warming and things like that, but you are not, things have never been better and the future appears bright no matter what. I am a holistic. I say,look, human environmental footprint is big and getting bigger and it can't get If it keeps doing that, we're going to have a lot of problems, but we want a holistic solution. It's a 50-year problem. It's not something to geoengineer. It's not something to ban coal or ban meat or ban air travel. It's something to let's rethink the power grid. Let's cut down some of the subsidies and things like that. I read the IPCC reports. I don't know how many people do that, but they're really good. You have nothing going on in your life. You gamble, you talk about betting a lot, and then you read IPCC reports. Yeah, well,it's like a novel, you know. But if you quote those to people, they'll call you a denialist. If you actually read what the science is saying. So,well, Hank Green called you, or he referred to your video, which has about a quarter of a million views or 300,000 views. Hank Green's rebuttal has something like 1.7 million, and he called it a master class in manipulation. And then you kindly wrote a rebuttal to it at Reason.com. What is, is he wrong to say that you are manipulating people? Yeah. Well, I want to say a few good things about Hank Green. He quoted me full context, right? He didn't take the little sound bites out of context. He had reasons for everything he said. And basically what he said is, okay, here's how they graphed it. And with a hockey stick, here's how Aaron graphed it. And Aaron's graph isn't as scary. And so far we're 100%agreement. That's the whole point. Only he took the view that that was manipulation. It was manipulation to show it in the normal way. And again, he has this, and I think this is a very common attitude among climate change activists, is it's okay to do this because you have to get people excited. My feeling is people are already too panicked about it,and they're pursuing risky, short-term, counterproductive strategies as a result. Yeah,and you are, I mean, one of the things that's great about your rebuttal, or when you talk about it, is that you also,and I mean, this is part of what wrong number is about,but the series,the book, the videos,everything is, you should be aware, we should all become more critical consumers of media. Because you do make choices when you present data. Hank Green is making choices. And,you know, we need to be able to think for ourselves. Yeah. And actually, one of the things in the book, I criticized a marijuana study that Marty Makary,who wrote,who was, stepped down from head of a...He stepped down,was pushed,whatever. But, well, you know...He was the FDA commissioner. But he wrote Blind Spots, which is a similar kind of book specifically on medicine, and he praises the study. So, and he's,you know, his book is good. He's also doing the same kind of thing. So talk, yeah, talk a bit about the marijuana and cardiac risk, because this is something that you've discussed kind of at length. There's, you know, a series of studies that have come out over the years that say, you know,if you smoke marijuana, you're much more likely to have cardiac events. And you essentially argue that the evidence isn't very good for that. The evidence isn't very good, but this study is particularly bad. So,again, a couple of problems. They rely on a survey that's done by the U.S.government that calls people up and asks them hundreds and hundreds of medical questions. And one of the questions was, have you used marijuana in the last 30 days? And another question was, have you ever had a cardiac event? So they found a correlation, which turns out that was wrong. There really wasn't a correlation. But even if there had been,you know, how can using marijuana in the last 30 days have anything to do with whether you ever had a cardiac event? Right.problem is,and this is-Well, wait, but isn't the argument that if you smoke a ton of weed, that could screw up your heart, right? Well, not the weed you smoked 30 days ago. Okay. Okay, the general rule here is correlation is not causation. But we know for a fact if the marijuana is after the heart attack that it couldn't possibly have caused it. You're so naive. But there's... The other issue here,though, is here is a very common thing in studies. So they say,you know, telephone studies might not be accurate, right? You're asking people about their drug use. You're asking people about their health. But we found that this is accurate, and they quoted three studies that say this is accurate. And then you go and look at the studies, and all three studies said it's worthless. But people say 50%of the...people who alleged they had cardiac events never had them. So the data is pretty useless. That seems a very odd thing to brag about. You know, it's not like the number of sexual partners. Yeah,maybe,but like,oh,yeah, I had that. I think it's, you know, because the question is something like, has a health professional ever told you that you had a heart issue? And so you're thinking,oh, yeah,my doctor once said, you fall in love too easily or something. I don't know. Or you remember vaguely somebody once said,yeah, I had some chest pains. And so you just say yes. And remember, you're answering hundreds of these questions. Or, I mean,if you had this experience, you go to a doctor's office and there's 40 things you have to check if you ever had this,this,this,this, this. And if you're like me, just go,you know,go down, know with everything without thinking about it. Or maybe I'm just lucky. Is there a way around that with certain, I mean, this is partly what you're talking about is the way things are framed. Part of it is like when you look at how data is actually collected, it's just. You know, the phrase repeats in the book, garbage in, garbage out. With something like studies of drug use and it's illegal or illicit, as the federal government insists on calling drugs like marijuana, is there a way to get good data about that? Or is it just like always going to be a kind of forest of vape? No, there's a great way to get data. Unfortunately, it requires national health systems, so you've got a downside there. But yeah, they're panel studies. So you take a big group of people,thousands, hundreds of thousands of people who all use the same health system, who get all their diagnosis in the same place, and you have complete medical records on them, and you study the same people over 40 years. So then you find,gee,

未分轨讲话者（中文译文）：如果把这条趋势外推一个世纪，我们确实需要做点什么。要么希望它自己改变，要么我们主动改变，否则 100 年后会有严重问题。但传播者想要冰球杆效果：一切到 1970 年都正常，突然就爆表，12 年后我们就彻底完了。于是他们用各种方式改造数据。Michael Mann 的原始图有很多问题，最大的问题是他有一条缓慢上升的曲线，然后把一条完全不同、与图表无关的数据红线贴到末尾，做成「杆面」。Brown 说那是一张弗兰肯斯坦图。于是他拿近期三张冰球杆图，用同一批数据按高中数学课里的正常方式重画。结果不是说气候变冷，而是说它不是冰球杆。主持人提到，Brown 在视频里说过，有些研究或图表用「高温日与低温日的比率」制造更吓人的上升趋势，而这不是正确呈现方式。Brown 说 Stephen Koonin 的《Unsettled》也谈过同一张图。美国创纪录高温日的数量基本稳定，甚至少于 1930 年代；全球范围内略有增加。真正令人担心的是创纪录低温日几乎消失。可这对公众不够吓人，因为一个人在铲雪时可能会想，暖五度也不错。于是有人展示「创纪录高温日 / 创纪录低温日」的比率。分母很小时，比率就会很大，红色柱状图看起来吓人，好像创纪录高温日变成过去的 20 倍。但高温日本身大约差不多。Brown 说，低温纪录日消失是严重问题，因为它会改变物种分布、农业和人口迁移；但这不是小行星撞地球式的恐慌，而是需要准备或处理的长期环境变化。主持人确认，Brown 并不是气候变化否认者。Brown 说自己是整体主义者：人类环境足迹很大，而且越来越大，如果持续下去会有很多问题。但解法也应该整体化。这是 50 年问题，不是靠地球工程、禁止煤炭、禁止肉类或禁止航空旅行解决的，而是要重新思考电网，减少一些补贴。他说自己读 IPCC 报告，报告很好，但如果把报告内容原样引用给别人，有些人反而会叫你否认者。接着他们谈到 Hank Green 对 Brown 视频的批评。Green 的反驳有约 170 万播放，并称 Brown 的视频是「操纵大师课」。Brown 先肯定 Green：Green 没有断章取义，引用了完整上下文，每个说法都有理由。Green 的意思是：原图是冰球杆，Aaron 的图不那么吓人。Brown 说到这里他们完全一致，问题就在这里；Green 把正常画法视为操纵。Brown 认为很多气候活动者有类似态度：为了让公众兴奋或警觉，可以这样做。但他的感觉是，人们已经太恐慌，并因此追求高风险、短期、适得其反的策略。主持人说，Wrong Number 系列的重点就是让大家更会消费媒体，因为展示数据时一定会做选择，Hank Green 在做选择，Brown 也在做选择，读者必须自己想。Brown 接着提到书里批评过一项大麻研究，而 Marty Makary 的《Blind Spots》虽然是一本不错的医学批判书，却赞扬了那项研究。关于大麻和心脏风险，Brown 说，很多研究声称吸大麻会显著增加心脏事件风险，他认为证据并不好，其中某项研究尤其糟。它依赖美国政府电话调查，给人打电话问几百个医学问题，其中一个问题是「过去 30 天是否使用过大麻」，另一个问题是「是否曾经有过心脏事件」。研究者发现一个相关性，而 Brown 说相关性其实也不成立。即使存在相关，过去 30 天的大麻使用怎么能解释一个人是否曾经有过心脏事件？主持人追问，大量吸大麻是否可能损害心脏。Brown 说，不可能是 30 天前吸的大麻导致了更早的心脏事件。一般规则是相关不等于因果，但如果大麻使用发生在心脏事件之后，那它就不可能是原因。另一个常见问题是数据质量。研究者承认电话调查可能不准，因为它询问吸毒和健康状况，但他们引用三项研究声称数据可靠。Brown 去看后发现，这三项研究实际都说数据没用：很多自称有心脏事件的人并没有真实记录。主持人说，这种自夸很奇怪。Brown 猜测，问题可能来自提问方式：比如「是否有医务人员告诉过你有心脏问题」，受访者可能想起医生说过胸痛，或模糊记得什么，于是答是。人们还要回答几百个问题，就像去医生办公室勾选 40 个病史项目，很容易不认真看。主持人问，研究非法或联邦称为 illicit 的大麻使用，有没有更好的数据办法，还是永远雾蒙蒙？Brown 说有好办法，但需要全国健康系统之类的条件。面板研究会跟踪几千、几十万使用同一个健康系统的人，他们所有诊断都在同一系统，有完整医疗记录，再研究同一批人 40 年。这样才能看出 20 岁吸大麻的人 50 岁是否更容易心脏病发作。

未分轨讲话者（英文原文）：the people who smoke marijuana at 20 have a heart attack at 50.Now we've got some real evidence. And people do that. And when they do that, they find very weak associations. I'm not going to say there's none. I'm not going to tell you smoking marijuana is good for you. But they don't get headlines. They don't get the kind of headline results that this study got. And then you talk about also how many, you know,people who say they smoke marijuana, they may have also been longtime cigarette smokers or,you know, there are so many confounding variables that it's just hard to tease out what's actually going on. And that's true. But what's worse is it's easy to manufacture what you want. And that's precisely what they did. They have,you know, 20 different covariates and they chose the ones that Could you explain a little bit, because that's also something that comes up. I was saying to somebody before, I took, God, like a year's worth of statistics in college, which if it degrades at 50%a year, I'm at that 114%,90 million, whatever. But you talk about like these covariance so that,you know, researchers will list 20 different possible outcomes. And they don't necessarily, when they're reporting on something, they don't say like, Thank you.had the biggest effect. Let's talk about that. They just kind of sift through for the one that they want. Yes. So what they do is you take–so you've got this study. You've got hundreds of questions. Yes.and you say, what are the biggest effects? What are the things, the biggest effects that cause heart attacks? And you've got a list. And the biggest one is something you don't care about. Maybe it's lack of exercise or something. And maybe the next biggest one is overweight or something. And you go down the list and you want to find something that's surprising. You can't get it published unless it's surprising,right? And also that fits in, may fit in with whatever you're doing. push. So they skipped a better example of that, particularly when there's a gun control study like that, where they go through and there were literally 10 things that were more important than having a gun. And this was the study in Connecticut,was it, or Rhode Island that showed if you have a gun in the house or the way it was reported, you were much more likely to be the victim of homicide. Right, right. This was actually in the southeast. But,yes, so you're more likely to be a victim of a homicide if you have a gun in the house. But it's far bigger effect if you rent your apartment, if you live in a gated community. If you ever got in trouble at work from drinking, you have 20 times the risk. So none of these were in the paper, right? They just go down. Oh, if you own a rifle or a shotgun, you have much less risk.none of that got reported. They just picked out, you know,guns are the big one. And again, many more problems with the study, but that is one. Why,you know, you mentioned the Lancet is, you know, the most–it's the oldest and most prestigious journal. It comes up in almost every chapter,it seems, for publishing crap studies. What is going–I mean, is there a breakdown in whether it's peer review or just,you know, kind of research periodicals because –Do most of these claims have to be laundered through some kind of scientific gatekeeper that we still trust? Yeah,well, the Lancet is the scientific American of the medical world, I'm afraid. That it used to be good, has become entirely political. You just can't trust it anymore. And I think because it was the oldest and most prestigious, it was the most attractive But I mean, is it more broadly that journals are just not,you know, I mean, as libertarians, we kind of like living in a world beyond gatekeepers or a world where you choose your own gatekeepers. But is it just everybody's asleep at the switch? No,no,no, no. It's a huge business. It's a multi -billion dollar business. You know, the way scientific research works, the scientific journal complex, whatever you want to call it, the government gives people a lot of money to do research. Those people pay journals a lot of money to publish the research, and then the journals charge the government a lot of money so you can read the research that the government paid for in the first place. Many billions of dollars, and it's a foundation of academic careers. And even good academic. And I know a lot of good academics who do good work, but they can't do enough good work for their career. So they have to publish a bunch of junk. And they need journals that will publish the junk. And they peer review each other's work. It really is something we just have to get rid of. What might replace peer review or the current system? I'm a big believer in consumer choice, that you publish it in a forum. Most of the review should be post-publication. So if you're publishing this stuff, you can put it online. All the real publication is online, by the way. So people put their work online, it comes out two and a half years later in a journal, and nobody actually reads it from that. The real peer review is afterwards. Can people validate your work? Do they try to refute it and fail? If nobody's interested enough to either build on your work or try to refute it, then it probably wasn't saying much in the first place. Can you talk at various points about kind of like listing your hypotheses ahead of time? or,you know, what does that mean and why would that help kind of limit insane findings that are really not related to reality? Yeah, that's the mom and apple pie thing. That's the more John and I, and I just think, I'm not, I don't have a great faith in that, but it is true. What you should do is you should pre-register your hypothesis. Say, I'm going to look at these data, I'm going to test this hypothesis, and I'm going to report back true or false whether the data supported or went against it. Instead, what people do is they look at the data, they pick a good hypothesis, and they publish it as if that's what they went in there looking for. So that would be a help, but it's not the fundamental problem. The fundamental problem is incentive. And if you don't fix the incentive, people are going to find a way around it. And is the incentive, I mean,part of the incentive is, you know, academics. And I guess it starts with academics, but they need to publish in order to advance their careers. The journals do not have an interest in publishing findings that aren't exciting or different or kind of,you know, throwing the brick through the window. And is it also that,you know, peers...aren't really interested in calling bullshit on the same people who might be reviewing their papers later. Yeah, and they are interested in pushing away the papers of people who aren't on their side. So,yeah, you've got these. And then there's the media problem. The media problem is separate, the media takes us. But the fundamental thing, this is big science. When the government is funding this stuff, people are going to find ways to get the money. And many of these people, and I know plenty of scientists and they're honest,hardworking, sincere people who want to learn the truth and help humanity. But they still have to play this game. And it is a huge, huge game. It's part of a bigger university. Universities have become a sort of way to soak up taxpayer money for administrators. And it's not, so it's the professors, the students, all these people are getting shafted. The people who rely on research are getting shafted. But there's no, you know, It's not like there's any one person doing this, it's people are kind of forced into this. It's the game you have to play. Is there, could you talk a bit about how, are there different incentives in private industry, either to do a kind of basic research in the first place, Or I knew somebody who used to work at Procter&Gamble and they told me that Procter&Gamble sponsors as much research as virtually all the universities put together. They spend a ton of money doing research. Some of it, most of it is applied to household products or things that they're selling. They're trying to figure out how to make Tide even tidier and all of that kind of stuff. But is there a public good to having basic research done? And is the private sector capable of doing that? Or is it just not? Well, I think,okay, so there are problems with private sector research as well. And the biggest thing is if you look at something like a pharmaceutical company,right? They're very much interested in, and nothing wrong with this is what they should do. They're interested in making a profit, but they're not going to necessarily do the research that a public, you know, greatest good of the greatest number public health person would do. I would say the case for public good for basic research is stronger than almost everything else. So it would be one of the last things I would slash when I'm drowning government in a bathtub. There's a lot of private foundations, there's a lot of charity for this, there's a lot of good basic research done by private companies. We certainly don't need anything like the size of government and we certainly don't need much of the work that is being done is not for any clear public good. Is there,and this might be beyond your, area of interest or expertise, but the university system is changing

未分轨讲话者（中文译文）：如果研究的是 20 岁吸大麻的人 50 岁是否心脏病发作，这才有真实证据。有人确实做这类研究，结果通常只发现很弱的关联。Brown 不说完全没有关联，也不说吸大麻对身体好；只是这类研究得不到刺激标题，也得不到那项糟糕研究那样的传播效果。主持人补充，很多自称吸大麻的人可能长期吸烟，混杂变量很多。Brown 同意，但说更糟的是，研究者很容易制造自己想要的东西。他们有二十个协变量，就挑自己想要的。主持人说，自己大学学过一年统计，现在可能已经衰减到听不懂 114% 和 9000 万的程度了；研究者列出二十个可能结果，报道时未必说明哪个影响最大，而是在里面筛出想要的东西。Brown 解释，一个研究里有几百个问题。你先看什么因素对心脏病影响最大，得到一个列表。最大的可能是缺乏运动，下一个可能是超重。你一路往下看，想找一个惊人的，因为不惊人就发不出来，也可能要符合你正在推动的议题。他举枪支研究作类比：某项研究声称家里有枪更容易成为凶杀受害者，但里面有十个因素比有枪更重要，比如租房、住在封闭社区、曾因饮酒在工作中惹麻烦等，后者风险高 20 倍。拥有步枪或霰弹枪甚至对应更低风险，但这些没有进入报道；报道只挑出「枪」这个变量。之后他们谈到《柳叶刀》。主持人说，《柳叶刀》作为最古老、最有声望的医学期刊，却几乎每章都在发表糟糕研究。Brown 说，《柳叶刀》像医学界的 Scientific American：过去很好，如今完全政治化，已经不能信任。主持人追问，是不是期刊和同行评审整体失灵？Brown 说这不是大家睡着了，而是一门数十亿美元的大生意。政府给研究者很多钱做研究；研究者付很多钱给期刊发表；期刊再向政府和机构收费，让人阅读政府资助的研究。这是数十亿美元产业，也是学术职业的基础。Brown 认识很多好学者，他们做认真工作，但仅靠好工作不够支撑职业，所以还得发表一堆垃圾；他们需要愿意发表垃圾的期刊，然后彼此评审。他认为这个系统必须被替换。替代方案是什么？Brown 倾向消费者选择和发表后评审。研究可以先放在公开平台上，真正的发表反正都在线上；很多工作先上线，两年半后才出现在期刊里，也没人从期刊版本开始读。真正评审发生在之后：别人能不能验证你的工作？有没有人试图反驳却失败？如果没人有兴趣继续基于你的工作研究，也没人愿意反驳，那它可能本来就没说出多少东西。主持人问到预注册假设。Brown 说，这是人人都赞成的事，但他并没有巨大信心。正确做法是先登记假设：我要看这些数据，检验这个假设，无论数据支持还是反对都报告。现实是，很多人先看数据，挑一个好假设，再像自己一开始就要检验它那样发表。这会有帮助，但不是根本问题。根本问题是激励。如果不修激励，人们会找到绕开的办法。主持人总结，学者需要发表来晋升，期刊不想发不刺激、不新奇、不能「砸窗户」的发现，同行也不愿意批评未来可能评审自己论文的人。Brown 补充，同行还会排斥不属于自己阵营的论文。媒体问题是另一层，但根本是 big science：政府资助这些东西，人们就会想办法拿钱。很多科学家诚实、勤奋、真诚，想寻找真相、帮助人类，但他们也必须玩这个游戏。大学也成了一种吸收纳税人资金、供养行政人员的方式；教授、学生和依赖研究的人都被这个系统伤害。没人单独设计这一切，人们是被迫进入这个游戏。主持人问私营部门是否有不同激励，比如宝洁投入大量研究。Brown 说，私营研究也有问题。药企追求利润，这没有错，但它不一定会做公共卫生角度「最大多数人最大利益」需要的研究。他认为基础研究作为公共品的理由，比多数政府项目都强。即使他想把政府规模大幅缩小，基础研究也会是最后砍的项目之一。私人基金会、慈善机构和公司也能做很多好研究，但现有政府规模并不必要，许多工作也没有明确公共利益。最后主持人问大学系统正在变化，比如入学人数峰值、终身教职减少，这会不会修复问题。Brown 开始回答说，这可能会按比例集中问题、让问题更糟，但大学和期刊的重要性在下降，网络课程和在线论文会带来替代方案。

未分轨讲话者（英文原文）：dramatically. By most accounts, we have peak college attendance in absolute numbers around 2010, 2011.Nobody is predicting there's going to be a huge boom in that. Things like tenure track positions are declining generally in terms of annual offerings. Will that fix this problem or will it have no effect? I would say, and this is just a guess here, that it concentrates the problem and makes it worse on a ratio basis, but because universities are declining and journals are declining in importance... and everybody can take classes on the net,can read papers on the internet, and so on, that we're going to get alternatives. And either, I don't know if universities will fix themselves,which they could do, or they'll die off. I don't know. Before we continue with the Reason interview with Nick Gillespie, I want to tell you about an incredible event Reason is sponsoring. Argentina has become the world's most closely watched experiment in free market reform. Inflation is falling. Economic growth is returning. And policymakers around the world are watching closely to see whether a country long associated with economic dysfunction can now chart a course toward prosperity. You have the chance to hear about this remarkable transformation from the very people who are making it happen. Reason is bringing together leading thinkers, reformers,policymakers,journalists, and advocates of liberty for a special conference in Buenos Aires. It's called Winds of Change, Advancing Liberty in Argentina and the Americas. From August 31st through September 3rd,conference participants, including me, Nick Gillespie, I will be there, will gather in Argentina's capital city to hear directly from the people shaping one of the great political and economic stories of our lifetimes. If you'd like to join Reason and me in Buenos Aires and be part of this incredible event, visit Reason.org slash Argentina 2026 for information and registration details. Again, that's Reason.org slash Argentina 2026. I hope to see you in Buenos Aires on August 31st through September 3rd for Winds of Change, Advancing Liberty in Argentina and the Americas. And now, back to the Reason interview. Let's talk about curbside buses or Chinatown buses. And this might have been–was this the first thing that you wrote about for Reason or did a video for? I wrote about it many years ago. I have not done a video on this one. Okay. Then I think one of our–mutual colleagues did. But this was about Chinatown, or I mean,in the popular kind of parlance, it was Chinatown buses, curbside buses. So these are things that are not the exalted brands of Greyhound and Peter Pan and Trailways, which I guess merge with Greyhound or... whatever,but those were good, publicly registered and regulated bus services that were supposedly good and safe and beautiful and wonderful. And then there were Chinatown buses, which were buses that just pulled up in weird parts of town and would transport people from New York to Boston or from New York to DC or places in between at a much cheaper and a more irregular way. And there was a flurry of reports saying that going on one of these buses was like, you might as well live in a gated community with a bunch of marijuana. You were going to die if you went on these buses. Explain where the phobia and the fear of curbside bus service came and why it was so widely propagated by the media. Yeah, I'm not an expert in this. So it's Fung Wah was the first one back in 1996 in Chinatown. And my students would all say,you know, it was$10.You got from New York to Boston. It was$25 in Greyhound. But Greyhound, you had to go to Port Authority. You had to wait in line. And they never had seats at the times you wanted to go, because they had a fixed schedule. And then Thanksgiving, it was full. Feng Hua, you went under the Manhattan Bridge. A bus pulled up. If the bus got full, they brought another bus. Which is amazing, right? Just the idea,right? So now,and the interesting thing, the ironic thing about curbside is now all the major carriers run curbside services too. So the Chinatown got out. I suspect there was some racism. There was some just fear that this...The guy who founded Feng Hua was a noodle delivery guy, and he saved up a little money and chartered a bus, and history was made. So that was certainly part of it. But there's also the fact that this was the fastest growing segment of bus transportation. It was threatening profits,whatever, and there was a real push. Greyhound wanted to get out and do their own. So, curbside was sort of invented as a politically correct term for this, but they really went after all the companies they shut down. The 29 companies they shut down were all Chinatown, were all immigrant-run small companies, not big regulated carriers. So, it really had nothing to do with going through terminals or picking up at curbside. But there were a couple of high-profile bus accidents, and then the news started carrying reports that you were,what, seven times more likely to die on a curbside bus than on a great American Greyhound bus? Being a passenger on a bus is incredibly safe. There are people killed, but they're pedestrians or they're other drivers. If a bus hits a car, it's not the bus passengers who get killed. No, there was a horrific crash by a casino bus. It was not a curbside bus. It was a casino bus that ferried people from New York back to a casino. And I think 15 people were killed. And that was through Mohegan Sun,was it? Yes, Mohegan Sun,right. Right. AC, you know,keep going to AC, but maybe Connecticut, not so much. And it was a horrific crash, and it was clearly negligent that, you know,the driver was asleep, had been working 20 hours. I mean, there were all kinds of things wrong with it. But it wasn't a curbside bus. The curbside bus had excellent safety records. So the NTSB, the National Transportation Safety Board, did a study. It was Chuck Schumer and Nadia Valdezquez requested it. And in the request, they dictated the conclusion. And so in order to get the seven times figure, they took all the Greyhound fatal accidents and they stuffed them into the curbside ballot box. And then they refused a Freedom of Information Act request to say which companies they had considered curbside So it's a think about this there the NTSB says there are bus companies out there Seven times as dangerous as other companies, but we're we're not going to tell you which it's a secret who they are You can't know And Jim Epstein who is here somewhere Finally figured it out and and and got that but even without that I There were all kinds of problems with this study. Can you explain a little bit what some of those problems were? Sure, they didn't have data on passenger miles traveled or passengers. So all they knew is the number of buses a company owned. They took the number of buses and they divided by, they took the number of fatal accidents divided by the number of buses. And this next one's a little technical, but what they didn't, they didn't take all the fatal accidents from curbside companies and divide by all the curbside buses. They did it company by company. And there happened to be one company called Skyhorse Travel that owned one bus that managed to kill somebody, a pedestrian. And so that's a 100%ratio. And when you average a 100%ratio in with 71 other companies, you get this incredibly high ratio. How, when you encounter stories like this in the press,what,you know, what are the tells to be like, okay,you know,and I,you know, here's, here's like the basic Aaron Brown checklist of,you know,what, what sets off your bullshit detector? Yeah.first thing is this is pretty incredible. You know, you would think that if one bus company were seven times as dangerous, or a whole group of companies were seven times as dangerous, you would have heard about it. You know, how come you've never read about a curbside bus fatal accident? And then you look at the report. So the NTSB report,it's 70 pages or something, of which maybe two have any data or argument. It's a long list of how great the NTSB was, how much work they did for this, color pictures, whatever. This kind of study, you just know, was not done by statisticians to communicate. information. It also, what's kind of interesting about it is there's no baseline, there's no absolute risk. It turns out that riding a bus to Boston is about as risky as crossing four streets. So even if it is seven times as dangerous, if you have to cross a few more streets to get to the terminal, it's break even. But people read, you know,seven times as dangerous, and they assume that means it's a big number. Do you feel like people are getting, you know,we're forced to, we're not forced to, but I think the way things are now, more people consume more media from more different places. Does that make us better consumers, or does it just kind of dull us to believing whatever,you know, the most memorable headline tends to be? I don't know. I mean, I know personally, you know,

未分轨讲话者（中文译文）：大学系统正在急剧变化。主持人说，按多数说法，美国大学入学绝对人数大约在 2010、2011 年达到峰值，没人预测会再大幅增长；终身教职岗位也在减少。这会修复研究发表问题吗？Brown 猜测，它可能会把问题集中起来，让比例上更糟；但大学和期刊的重要性下降，人们可以在网上上课、读论文，所以会出现替代方案。大学要么自我修复，要么衰落，他不知道会是哪一种。节目中间插入 Reason 在阿根廷布宜诺斯艾利斯举办会议的广告，主题是阿根廷自由市场改革、通胀下降、经济增长回归，活动名为 Winds of Change: Advancing Liberty in Argentina and the Americas，时间是 8 月 31 日到 9 月 3 日。广告结束后回到访谈，主持人提出街边巴士或 Chinatown buses 案例。Brown 多年前写过这个话题，但没做过视频。主持人解释，Greyhound、Peter Pan、Trailways 等传统品牌被认为是注册、监管、良好、安全、漂亮的长途客运，而 Chinatown buses 是在城里某些不太正式的地方停车，以更便宜、更灵活的方式把乘客从纽约送到波士顿、华盛顿等地。后来有很多报道把乘坐这些巴士说得极其危险。Brown 说自己不是这个领域专家。Fung Wah 是 1996 年在唐人街出现的第一家。学生们告诉他，纽约到波士顿只要 10 美元，而 Greyhound 要 25 美元；Greyhound 要去 Port Authority 排队，班次固定，感恩节还满员。Fung Wah 则是在曼哈顿大桥下上车，车满了就再调一辆车。Brown 说，这个想法很了不起。讽刺的是，现在大公司也都做街边发车。Brown 怀疑其中有种族因素，也有对移民小公司的恐惧。Fung Wah 创始人原来是送面的，攒了点钱包了一辆巴士，改变了历史。但也有商业因素：这是长途巴士增长最快的细分市场，威胁到利润，Greyhound 也想进入。于是「curbside」成了一个政治正确的名称，但监管真正打击的是唐人街公司。被关停的 29 家公司都是唐人街或移民经营的小公司，而不是大型监管承运商。所以这件事其实不在于是否经过车站或街边上车。随后出现了几起高关注车祸，新闻开始报道坐街边巴士死亡风险是 Greyhound 的七倍。Brown 说，巴士乘客本身非常安全；致死者通常是行人或其它车辆驾驶者。确实有一场可怕的赌场巴士事故，从纽约往赌场接送乘客，死了约 15 人，好像是 Mohegan Sun 相关。司机睡着了，已经工作 20 小时，显然有严重疏忽。但那不是街边巴士。街边巴士安全记录很好。国家运输安全委员会 NTSB 应 Chuck Schumer 和 Nydia Velazquez 要求做了一项研究，Brown 说请求本身就规定了结论。为了得到「七倍」数字，他们把 Greyhound 的致命事故塞进了街边巴士类别，还拒绝信息自由法请求，不说明哪些公司被算作街边巴士。Brown 说，想想看：NTSB 说有些巴士公司危险七倍，但不告诉你是谁，这是秘密。Jim Epstein 最后弄清了名单。即便不看名单，研究也有很多问题。它没有乘客里程或乘客数量数据，只知道每家公司有多少辆车，于是用致命事故数除以巴士数量。更技术的一点是，它没有把所有街边公司的事故加总再除以全部街边巴士数，而是逐家公司算比例再平均。一个叫 Skyhorse Travel 的公司只有一辆巴士，却撞死了一名行人，于是比例是 100%；把这个 100% 和其它 71 家公司一起平均，整体比例就被极端拉高。主持人问，看到这类媒体故事时，Brown 的警报器是什么？Brown 说，第一反应是「这太惊人了」。如果某家公司或一类公司真的危险七倍，你应该早就听说过。为什么你从没读到街边巴士致命事故？再看报告，NTSB 报告约 70 页，可能只有两页有数据或论证，其它是在说 NTSB 多么出色、做了多少工作、配彩色图片。这样的研究一看就不是统计学家为了传递信息写的。更有意思的是，它没有基线，没有绝对风险。结果表明，坐巴士去波士顿的风险大约相当于过四条街。即使真危险七倍，如果你为了去传统车站要多过几条街，风险也抵消了。但人们看到「危险七倍」就以为那是个大数。主持人最后问，现在人们从更多地方消费更多媒体，这会让我们变成更好的消费者，还是让我们更容易相信最容易记住的标题？Brown 说他不知道。

未分轨讲话者（英文原文）：the fact that when somebody says studies prove, I can click on the link and sometimes there's a study, sometimes there isn't. And sometimes it's good. I can send it to Claude. I can say, hey, Claude,do studies really prove this? It makes it so much easier to check these things out. But I'm not sure people are doing that. Let's talk a little bit about where you come from, literally and figuratively. You mention in the book that you are libertarian. Can you define what–that can mean a lot of things. Really? I think so. Yeah, as I've discovered over the past 30 years, being libertarian can mean lots of different things. What do you mean when you say you're libertarian? Really,if I'm being totally honest, I'm really more skeptical than anything else. I'm skeptical that people know what they're doing, that the government is here to help you, that big expensive top-down programs work. But the way I define it philosophically is I truly, really believe that it's always wrong to initiate violence against an innocent, and if there's any possible way to do that,run things without that, we should do it. I'm not, I don't really spend a lot of time worrying about whether it's possible to completely eliminate that. Whether public health or children's welfare or something like that requires us sometimes to violate strict libertarian principles. I just look around, I see so many egregious, low-hanging fruit, obvious things we could do to reduce violence, to reduce infringements on rights.that I would be perfectly satisfied if we could get rid of those and then we can argue about the rest. You know, just as an aside, you mentioned the–or you referenced the Reagan quote,you know, about how the worst words to hear are, I'm here from the government or, you know, and I'm here to help. What's interesting about that recently,which I, you know, I've heard and I'm sure I've written in many articles over the years, the actual speech in which he gave that was he was bailing out farmers. So it's like,it's always interesting. Like, I mean, and this is part of what your book is about, is about finding a larger context for things. And that can include trying to extend trend lines backwards and forwards as much as possible, but also just reading quotes in context. It's terrifying that,you know, Ronald Reagan in saying that was actually, you know, giving a bailout to farmers. And he literally says,like,you know, in a free market, we wouldn't be bailing you out, but we don't have a free market, so we're going to make it, you know, whatever. It's similar also to the big New York headline,you know, a Ford to city drop dead. Ford actually was committing in that speech, he was committing to bailing out New York. So it's always kind of curious that way. I want to go back to you grew up in various places, but you spent a lot of time in Seattle or in the University of Washington area. And you mentioned that your father was a semi-communist. And part of your life story kind of explains why you're interested in kind of.contrarian thinking and individualism and freedom. Can you talk a little bit, how was your father a semi-communist and what was his experience that stuck with you where being independent of thought and also having some protection from majoritarianism seems to be very important to you? Yeah,well, he was editor of the Chicago Maroon as an undergraduate, and he wrote some editorials that, you know,very mainstream liberal saying, you know, universities shouldn't enforce loyalty oaths. This was during the 50s. And so he ran afoul of the Red Hunters and when we,you know, he drove his family to University of Washington to get his first job and the university withdrew it for this. And there was one lawyer in Seattle, a young Jewish lawyer who was just starting out, who was willing to take his case and won eventually and he got the job and so it's all. Good news, but it really was a lasting experience on him. He was not a communist in the sense, in a sort of political sense. He was more of a sort of puritanical leftist sort of, he had no respect for people who made money without working. And working was kind of,didn't have to be making steel, but you had to be teaching school or doing good research, something like that. Very suspicious of finance,of bureaucracy, of anybody,you know,and you know, somebody who was a middle manager who didn't really seem to be doing anything. And I was very interested in gambling. I just,from a very young age, I used to look-Which is not generally considered a Marxist virtue, right? No, no. He was very upset about that because it was making money without working. I was just fascinated with the numbers, the patterns I could find. I used to read a newspaper from the back because there were the farm reports, the box scores,the stock tables, and they had all these numbers and the patterns.and do it. But he was very proud of me that I could do this, that I could actually make money of this stuff. So he was torn. And I've always felt that sort of growing up, that I had this thing I was interested. I understood it deeply cut against his feelings, but I also understood that he was proud that I could do it. And I think that was a formative experience. Talk more about your gambling past. At various points, you said you essentially supported yourself by gambling, either playing poker. You bet a lot on baseball,it seems, on professional sports. I did sports, but my real money was poker. And this was a very different. So this is not online. This is not casino poker. This is private games. I was a shy kid and the hardest thing I did in my life perhaps was to walk into a tavern basement when I was 14, sit down at a poker game with a bunch of adults, walk out with all their money. And it was amazing. They let me do it and they liked me. And this was amazing to me. And it gave me a real feeling of security. I have sort of a Jewish, you know, money in the bank can't make you secure because they can take it away. You want something, you can cross the border at midnight, broken friendless, and still get something. And poker was it. Anywhere in the world, you could find a game. Is that mostly –you can always find a poker game and then you–I mean,do you play, I don't want to say conservatively, but are you an unemotional poker player where you're always just playing the odds at any given moment? I am at the table to win money. And my biggest advantage is most people who are at the table are not there to win money. They have some other motivation. And,yeah, I'm good at math and I play and I take it seriously and I'm doing–and I'm just being there to make money is the biggest advantage. What do you think about current discourse about gambling? Because one of the–there are many amazing changes in American life over the past half century. Right.them is certainly the rise of fully legalized sports betting, where you can not only bet legally on things, but you can bet on individual serves in tennis games now, individual pitches and things like that. Most people...other than the people who are actually doing this, seem to see this as one more sign of the coming apocalypse. Do you worry about that, like the over-gamification or the over -gamblification of everyday life? No, I don't. I would say two things. For most people,it's entertainment, and it's not obviously worse entertainment than watching sports or doing this as participatory. You know, you're thinking, you're acting. But the biggest thing is 1%of the people doing it actually have an advantage. And if you're one of these people, there's no other way to find out that you have,you know, you can beat the wisdom of crowds. And it's an incredibly valuable lesson if you're one of the 1%.And if you're not one of the 1%, it's an equally valuable lesson that you should listen to the crowd. And you can't find it out anywhere else. You have to go in and see, can I make money betting in these prediction markets? Can I make money with the sports bookies? Or prediction markets as good as a lot of people claim they are. You know, and this is,we,you know, at least in the 21st century, we kind of started out on a wave and there was a bestselling book called The Wisdom of Crowds. Now we mostly worry about online moms. And online lynchings and things like that. So we're constantly kind of going between those extremes. Do betting markets, and when people do this for election results and things like that, are they actually good at predicting what happens or is that being oversold? Well,okay, every form of prediction has its pluses and minuses. So prediction markets have some very positive things, but they also have some negative things. Expert judgment has positives and negatives, statistical techniques for doing it. The combination of all of these is valuable. And I think your mobs, you need wisdom of crowds. It is very important for things, but you need a few people who can beat it to kind of their, what do you want to call this? Sort of the things that keep it from getting too far off track. And people are willing to put up their own money and bet against it and make a living that way. We're going to go to audience questions in a minute, but I guess for a final thing, because this one,you know, I've been following your work for years and everything, but the thing that kind of took me by

未分轨讲话者（中文译文）：Brown 说，对他个人来说，现在别人说「研究证明」时，他可以点开链接。有时真有研究，有时没有；有时研究还不错。他还可以把研究发给 Claude，问它「这些研究真的证明了这件事吗？」核查变容易了，但他不确定人们是否真的会这么做。主持人随后转向 Brown 的个人背景。书里提到他是 libertarian，自由意志主义者。Brown 说，如果诚实说，他更像怀疑主义者。他怀疑人们知道自己在做什么，怀疑政府真是来帮忙的，也怀疑昂贵的大型自上而下项目能发挥作用。哲学上，他真心相信，主动对无辜者施加暴力总是错的；如果有任何办法不靠这种方式运转社会，就应该这么做。他不太纠结是否能完全消除这类暴力，也不急着争论公共卫生或儿童福利是否偶尔需要违反严格自由意志主义原则。他环顾现实，只看到太多明显、低垂、容易先改的东西，可以减少暴力、减少权利侵犯；如果能先清掉这些，他就很满足，剩下的再争。主持人顺带提到 Reagan 那句「最可怕的话是，我来自政府，我是来帮忙的」。有趣的是，Reagan 说这句话的那场演讲其实是在救助农民。他甚至说，在自由市场里我们不会救助你们，但我们没有自由市场，所以要这么做。主持人说，这也说明 Brown 书里的一个主题：要把事物放回更大上下文，延长趋势线，也要把引语放回原文。类似地，《纽约每日新闻》标题「Ford to City: Drop Dead」常被理解为福特让纽约自生自灭，但福特那场演讲其实是在承诺救助纽约。之后主持人问 Brown 的父亲为什么被他说成「半共产主义者」，以及那段经历如何让 Brown 重视独立思考、个人主义和免受多数派压迫。Brown 说，他父亲本科时是 Chicago Maroon 编辑，1950 年代写过很主流自由派的社论，反对大学强制忠诚宣誓，因此得罪了反共猎手。后来父亲开车带全家去华盛顿大学上任第一份工作，学校因此撤回职位。西雅图有一位刚起步的年轻犹太律师愿意接案，最终打赢，父亲拿到工作。结局是好的，但这个经历对父亲影响很深。Brown 说父亲不是政治意义上的共产主义者，更像清教徒式左派，看不起不工作就赚钱的人。工作不一定是炼钢，也可以是教书、做好研究，但必须是在做实事。他很怀疑金融、官僚机构和看起来没干什么的中层经理。Brown 自己从很小就对赌博感兴趣。主持人打趣说这通常不算马克思主义美德。Brown 说父亲对此很不安，因为这是不工作赚钱；但他自己迷恋数字和模式，读报纸时从后面开始读，因为那里有农产品价格、比赛统计、股票表格，全是数字和模式。父亲一方面不喜欢，另一方面又为他能靠这个赚钱而骄傲。Brown 觉得，这种矛盾是成长中的重要经历。主持人追问他的赌博经历。Brown 说自己赌过体育，但真正赚钱的是扑克。那不是线上扑克，也不是赌场扑克，而是私人牌局。他小时候很害羞，人生最难的事之一，是 14 岁时走进酒馆地下室，和一群成年人坐下打牌，然后赢走他们的钱。神奇的是，他们允许他这么做，而且喜欢他。这给了他很强的安全感。他说，银行里的钱不能让人真正安全，因为别人可以拿走；你需要一种本事，让你午夜越境、身无分文、没有朋友时仍能谋生。扑克就是这个本事，世界任何地方都能找到牌局。主持人问，他是不是冷静保守、只按概率打牌。Brown 说，他坐上牌桌就是为了赢钱。最大优势是，多数人上桌并不是为了赢钱，他们有其它动机。他数学好，认真打牌，但「上桌就是为了赚钱」本身就是最大优势。接着谈到美国合法体育博彩和生活游戏化。Brown 说他不担心。对多数人来说，这是娱乐，不明显比看体育更糟，甚至更有参与性：你在思考、行动。更重要的是，约 1% 的人确实有优势。如果你属于这 1%，这是发现自己能打败群体智慧的唯一方式之一，是极有价值的一课；如果你不属于 1%，同样会学到应该听从群体。你必须亲自进去看：我能不能在预测市场赚钱？能不能从博彩公司赚钱？主持人问预测市场是否像人们说的那样好。Brown 说，每种预测方式都有优缺点。预测市场有好处，也有坏处；专家判断有优缺点，统计方法也一样。几种方法结合才有价值。群体智慧很重要，但也需要少数能打败群体的人，像校正装置一样防止它偏得太远。这些人愿意拿自己的钱下注，并靠反向判断谋生。

未分轨讲话者（英文原文）：surprise was your argument that Jim Thompson is the greatest American writer and certainly the greatest American writer of finance and kind of basic economics. Explain who Jim Thompson is and why you have such a high regard for him. Okay, those of you who aren't noir fans may not recognize him. So Jim Thompson was a mid-20th century writer of pulp and novels, and they're crime novels or a similar genre. But what nobody seems to talk much about is he puts long sections in his books about finance and accounting. So here he was an actual...He was not a semi-communist, he was an actual communist. And yet he had a very keen understanding of economics. And the particular book I focus on most in Wrong Number is one called The Getaway. It was made into three different movies. And if you've seen one of the movies, they get rid of his ending. And they don't spend a lot of time on the banking either. They don't spend a lot of time. They rob banks, but they don't go into the theory of banking. Actually, they rob a bank before the movie begins, really. And then banks are never mentioned. The whole book is about taking money out of one bank and putting it in another and thereby moving from earth to hell. And it is a brilliant, brilliant book if you read it for that. And all of his books have this in them. But what do you know? Was there something in his biography that was the you know, did he aspire to be a bank clerk and just couldn't? And so he became a well-regarded but lower tier fiction writer. I thought a lot about that. There's a great biography of Jim Thompson by Robert Palito. And and what I gained from that is his father. had a miserable life, and he had a pretty miserable life himself. And the only comfortable positions his father ever had were fell apart because of bad bookkeeping. And he kind of learned from this, I think,that yeah, bookkeeping,you know, if you can get your account straight, if you can clean up your desk, you can have a nice,safe, happy life. And if you can't do it, nothing else matters. All right. Aaron Brown,author of Wrong Number, How to Extract Truth from a Blizzard of Quantitative Disinformation. Thank you for talking to Reason. Thank you.

未分轨讲话者（中文译文）：最后，主持人提到一个让他意外的观点：Brown 认为 Jim Thompson 是最伟大的美国作家，至少是最伟大的金融和基础经济学美国作家。Brown 解释，不读黑色小说的人可能不认识他。Jim Thompson 是 20 世纪中期的通俗小说和犯罪小说作家。很少有人谈到的是，他在书里写了大量关于金融和会计的段落。他不是「半共产主义者」，而是真正的共产主义者，但对经济学有敏锐理解。Brown 在《Wrong Number》里重点谈的一本书是《The Getaway》，它被改编成三部电影。如果看过电影，会发现电影删掉了原作结尾，也没有花太多时间讲银行。电影里是抢银行，但不讲银行理论；实际上银行劫案在电影开始前就已经发生，之后也很少再提银行。而原书整本都是关于把钱从一家银行取出、放进另一家银行，并由此从人间走向地狱。如果从这个角度读，那是一本非常出色的书。Brown 说 Thompson 的每本书都有这种东西。主持人问，他是不是生平中有什么经历，比如想当银行职员却没当成，于是成了受尊敬但层级较低的小说家。Brown 说他想过很多。Robert Polito 写过一本很好的 Jim Thompson 传记。Brown 从中得到的印象是，Thompson 的父亲过得很惨，他自己也过得很惨；父亲唯一舒适的职位都因为糟糕账目而崩塌。Brown 认为 Thompson 从中学到的是：记账很重要，如果你能把账理清，把桌面整理干净，就能过上安全、舒适、幸福的生活；如果做不到，其它都不重要。主持人最后说，Aaron Brown 是《Wrong Number: How to Extract Truth from a Blizzard of Quantitative Disinformation》的作者，感谢他接受 Reason 采访。Brown 回答谢谢。