A memorial repost

Friday, March 2, 2018

Explaining the principal-agent problem

I thought I posted this years ago.

The Butler and the Maid from The Carol Burnett Show

One more note on the Uber fleet -- it makes surge pricing go away*

[* realized what a bad title this was. Uber can still jack up prices at rush hour; it just won't put any more cars on the road.]

I'm nervous about using economic language here -- I'll probably get some of the terms mixed up -- but one of the the most attractive parts of the original Uber business model was surge pricing and the resulting elasticity of supply.

In order for this to work you need to satisfy a couple of essential conditions. First you have to have to have a big surplus of supply most of the time. It's the classic problem of staffing to the spikes. Second, you have to find a way to avoid paying for people and equipment when you're not using them.

Uber and Lyft actually managed to solve those two problems. There are a huge number of underemployed people out there with underutilized vehicles, and when they aren't carrying passengers, they cost the companies nothing.

Conventional wisdom recently has been that both Uber and Lyft need to jump on the self-driving bandwagon and acquire a fleet of autonomous vehicles as soon as we hit level five technology (something that's probably at least a good decade  away, but we'll put that aside for the moment). In addition to the previously mentioned flaws in this plan, both companies lose this tremendous latent supply. If they want to have a million cars available in a metro area for Friday afternoon rush hour, they will have to buy a million cars, and it's fair to assume a large majority of those cars will be sitting idle a large majority of the time.

It still feels weird to say this, but Elon Musk gets this one right.

Well, that pretty much just lays it out in the open

From a characteristically good LA Times article by Sam Dean

“Profitability in the first three-to-five years is not the focus,” said Daniel Ives, an analyst at Wedbush. “The focus is on doubling down on growth and further expanding this Uber economy over coming years.”

The company is just getting started, the thinking goes, and its core ride-hailing service, Uber Eats food delivery business and Uber Freight shipping logistics division are poised to take over many times their current market across the world.

“As someone who’s covered technology for 20 years, I could count on one hand the stories that are potentially transformational on the consumer enterprise side,” Ives said. “Uber has the blueprint to be what I view as the Amazon in transportation.”

Once Uber reaches that world-eating scale, the believers say, it will have such reach into drivers’ and riders’ lives that it can start tightening the economics, and introduce more profitable subscription models or, eventually, self-driving cars, and fend off any competition from other deep-pocketed tech giants.

Ives sees Lyft, which only operates in North America and hews closer to its core ride-hailing product than the expanding Uber, as a less enticing investment precisely because it has said it plans to work to reduce losses.

“A major strategic mistake that Lyft made was putting their back against the wall talking about the path to profitability in the next few years,” Ives said. “Ultimately with Uber, either you believe or you don’t.”

A few points.

1. We're talking about a world-wide monopoly. Among the other difficulties facing this plan, it requires a certain degree of buy-in from the various governments involved. This is a big stumbling block in Europe and an almost insurmountable obstacle in China.

2. Putting aside the unlikeliness of pulling this off, the phrase "tightening the economics, and introduce more profitable subscription models" is a pretty way of saying "take advantage of the monopoly and start gouging."

3. It's a bit off topic but there are few meaningful parallels between Amazon and Uber business models. The name is evoked here strictly for its magical power.

4. And perhaps my favorite part. Lyft's mistake was even thinking about paths to profitability. That shows a lack of faith, and for all magical thinking, faith is essential. If it works for the Great Pumpkin...

What if self-driving cars will actually hurt Uber and Lyft?

Two reasons for me to be nervous about this argument: first, it is very much in the minority and second, it puts position of defending the business logic of Elon Musk, at least in relative terms.

Of course, the level of autonomy required for viable driverless taxis is still a long ways away. That's not to say that AVs won't start having a big impact on transportation before then (my money's on long-haul trucking), but  go anywhere with no human on board functionality still faces big challenges.

That said, if you could get the technology safe enough and reliable enough, and you put aside Musk's typically exaggerated promises (the cars will not be paying for themselves in a year and a half), the basic idea of setting up a ride share company using privately owned Teslas is not necessarily a bad one.

By comparison, the argument, now absolutely entrenched in conventional wisdom, that Uber and Lyft will greatly benefit from the advent of fully autonomous cars continues to strike me as deeply flawed.

Here's an example. Imagine you own one of two delivery services in a town. Both you and your competitor have roughly the same number of trucks but you have invested a great deal of money upgrading and making sure that your vehicles are as energy-efficient as possible. So far, the cost of the upgrade has been balanced out by your savings on diesel so that you are able to charge roughly the same rate as your competitor. A drop in fuel prices will reduce your operating cost. Normally that would be a good thing, but the cost for your competitor will drop by even more so that he will be able to undercut you on prices.

The Uber business model is based on the fact that there are a huge number of underemployed people who own underutilized cars (virtually all private vehicles are underutilized). Since car and driver are already just more or less sitting there most of the time, Uber is able to offer rides at a rate that would not otherwise be sufficient to cover all the assorted cost.

(Technically Uber doesn't offer the rides, but you get my drift.)

{And, yes, there are people who buy cars just to drive for Uber. There are also people who buy commemorative plates as a hedge against inflation.}

If you take drivers out of the equation, suddenly it becomes unclear what advantage Uber has over taxicab companies, car rental services, car dealerships or any business that maintains a large fleet of cars. Let's consider the Hertz example here in Southern California. Currently you have locations spread around LA and Orange counties, with each lot having to maintain a minimum stock. With truly driverless cars, you can get awfully close to 100% utilization for much of the day. Just have your extra vehicles prowl for fares and make deliveries, then send them to whatever location needs them next. Add to that maintenance facilities, purchasing power, a late model fleet and countless economies of scale.

You can imagine similar scenarios for any number of other businesses and in each of those scenarios, Uber and Lyft get screwed over by large, new, well-positioned competitors.

All of this leads us to the dirty little secret of the ride sharing industry. Though it was made possible by technological innovation (specifically the smart phone), the stability of the business model depends not on sustained disruption and transformation but on things remaining basically the same.

This does not bode well

As many have noted, the competition for economic development has a record of corrupting and/or sharply lowering the intellects of state and local officials. The hyperloop and other Musk proposals have a history of doing the same thing.

Now imagine what is likely to happen when you combine the two.

Only three months ago, another Virgin Hyperloop executive, Assistant General Counsel Nathan Roth, said Texas was "basically ... in the lead" for a hyperloop route because transportation officials in the Dallas area had started a federal environmental impact study — something he said no other area had done.

Such an impact study is different than what Missouri completed and was recently lauded by Walder. Missouri's study released in October was a nine-month feasibility report conducted by Black & Veatch and Olsson Associates. It focused on social impact, potential station locations, route alignments, regulator issues and rights-of-way access.

The process of securing a hyperloop route is indeed complex — so much so that the U.S. Department of Transportation in March launched an organization to help new transportation technologies such as hyperloop come to fruition quicker.

Virgin Hyperloop isn't the only company pushing the technology made popular by tech visionary Elon Musk. Earlier this year, it was reported that Transonic Transportation LLC, a Louisiana-based startup, is working on a hyperloop route in Texas that would transport freight. The company originally investigated a route that would move people.

"There's just no way we can do passenger transport in Texas in the next 20 years, so we ended up refocusing on freight," co-founder Josh Manriquez said in January.

The freight route he now envisions would run from Laredo at the Mexico border to San Antonio.

[Side note. I may not have mentioned it recently, but maglev vactrains are even more problematic for carrying freight than they are for carrying people.]

It must suck to have a hundred million dollar production come in second to happy little trees

One of the fundamental rationales of the content bubble, a multi-billion dollar explosion in television and movie production, is that the value of older content will drop to next to nothing as the newer shows come online.

Content accumulates and, as we've said before, if older shows hold their appeal, let alone find new audiences, then it's not clear how the market (even allowing for healthy international expansion) can absorb a really big surge, and big doesn't begin to describe what we've been seeing.

From Forbes:

Calculations of viewer's habits have uncovered some interesting trends, one of which pointed out that older content seems to appeal to consumers of the on-demand audience.

As an indication of this, Hulu reports that viewers watched more than 1000 million hours of TGIF content in 2018. The ‘TGIF brand’ includes the series Full House, Family Matters, Sabrina:The Teenage Witch, Step By Step, Perfect Strangers, Boy Meets World, and Hangin’ With Mr. Cooper. All of these series premiered in the 1990s. Over 1200 episodes of this content is available for viewing.

Viewers were also draw to take in nearly 1 million hours of Bob Ross’ The Joy of Painting.

This series originally aired from 1983 until 1994, amassing 403 episodes. Star/On-Screen Instructor Ross passed away at age 52 in 1995.

Maybe we just need to give Clarke a bit more time

A couple of years ago, we ran a thread on this video from Arthur C. Clarke speculating on the world of 2000. In particular, we discussed suggestion that what we would now call telecommuting would make cities obsolete.

Here was the main reason we suggested for the failure of telecommuting to live up to its promise.

But I think a third factor may well have been bigger than either of those two. The early 60s was an anxious but optimistic time. The sense was that if we didn't destroy ourselves, we were on the verge of great things. The 60s was also the last time that there was anything approaching a balance of power between workers and employers.


This was particularly true with mental work. At least in part because of the space race, companies like Texas Instruments were eager to find smart capable people. As a result, employers were extremely flexible about qualifications (a humanities PhD could actually get you a job) and they were willing to make concessions to attract and keep talented workers.


Telecommuting (as compared to off shoring, a distinction will need to get into in a later post) offers almost all of its advantages to the worker. The only benefit to the employer is the ability to land an otherwise unavailable prospect. From the perspective of 1964, that would have seemed like a good trade, but those days are long past.

Of course, that was 2017. With two more years of an improving job market, we are getting closer to those post war levels, which makes this perhaps a bit less surprising.

The final stages of TEDification include your own personal Netflix special

Ken Levine gives us another example of the researcher -> TED Talker -> celebrity progression.

So I come upon a special by someone named Brene Brown. Who? Well, she must be famous if she has a Netflix Special. So I click on and it starts like every other stand up special – the performer backstage (basically a waste of the first three minutes), and then this attractive middle-aged woman steps out onto the stage. It’s a big theatre with balconies. You can’t do a Netflix special without balconies. And she immediately gets a standing ovation. Have I been marooned on a desert island for five years? Who is this person getting a standing O? She starts off with a few mild jokes that are getting screams. And then I start to realize she’s not actually a comedienne, she’s a self-help guru. But she’s one for Millennials because every sentence was peppered with “So I’m like… and then he’s like… and I’m like… and like they’re like…”

...

Finally, she makes reference to a TED talk she once did. So I decided to turn off the special and seek the TED talk.
...
I guess Oprah or somebody discovered it and the TED talk went viral. And suddenly Brene Brown is a social media star. She now has a bunch of books (I assume with covers that she is allowed to approve), a top draw lecturer, and Netflix Special-er.

Her message sounds sound and every few years another self-help guru comes along (where is Susan Powter when we need her?), but to me the most interesting thing about Brene Brown is her transformation from academic lecturer to zeitgeist celebrity. She’s now got the new hair, new wardrobe, new zippy patter, new Millennial-speak. Someone should really study that phenomenon. Hey, maybe there’s a Hulu Special in your future.

Another reminder that the stakes for Mr. Musk are very, very high

From CNBC:

Tesla said Thursday it plans to raise up to $2 billion, with $1.35 billion coming from convertible notes and $650 million from new equity, including a big purchase from CEO Elon Musk.

In a filing, Musk signaled his intent to buy about $10 million of the electric auto maker’s stock in the new offering. The total equity offering is for 2.7 million shares of Tesla. Musk’s purchase would be 41,896 shares. Before the offering, Musk owned about 20% of Tesla’s outstanding shares, worth about $12.6 billion, according to FactSet.

Forbes lists the next worth of Elon Musk as $20.3 billion as of 5/4/19. $10 million is not by any stretch of the imagination a "big purchase" for him. $12.6 billion, on the other hand, is quite a bit, particularly given the tendency of Forbes billionaires/"billionaires" to exaggerate their wealth (and Musk's tendency to exaggerate everything).

The collapse of Tesla would cost Elon Musk most of his fortune, not to mention his reputation and all those perks of celebrity. Maintaining his persona as messianic visionary and disruptor is absolutely essential for keeping the funding flowing in.

We will probably never know how sincere Musk is when he talks about Mars colonies, or maglev vactrains or telepathy brain chips, but it is important to remember that the buzz those futuristic "ideas" generate buzz that is worth considerably more than $10 billion dollars to the "real life Tony Stark."

[And it appears to be working. ]

Drinking from the wrong pipe -- Bill Barr edition

For those who just got here, we have a long running thread about the implications of Straussianism in the conservative movement. One aspect we've been particularly focused on was instability of a system operating within a relatively free society (as compared with cases where the state controls the media) that depends on channeling censored information and sometimes outright disinformation to the base while keeping the leadership well informed.

Here's how we put it in 2017:

The initial purpose of this "noble lie" approach was to use the propaganda to keep the base sending money and showing up for the polls through of a combination of rage and fear. As with all Straussian systems, it was assumed that those in power would be in on the joke while the people who believed the lies would simply serve as electoral cannon fodder.

At some point though (I suspect inevitably), a couple of things happen. First, the believers become leaders. This is become blindingly obvious with Trump, but the children of Fox News have been in control of the party since at least 2010 and the roots go back further. Remember how Dick Cheney insisted while traveling that all hotel televisions be tuned to Fox News?

The second, and possibly more dangerous problem is that a propaganda-fed base has no capacity to self correct, rather it continues follow unsustainable paths that only gain momentum, often exacerbated by ratcheting mechanisms. Soon you reach a point where, even if the leaders accurately perceive the situation and realized the best solution, they can no longer reconcile that reasonable course of action with what the vast majority of their supporters have been told to believe for decades.

We referred to leaders being caught up in the propaganda as "drinking from the wrong pipe."

Keeping all of that in mind, check out this insightful piece of analysis from Josh Marshall:

It’s a common refrain among non-Republicans that Fox News and the rest of the conservative media superstructure have essentially brainwashed 30 percent or 40 percent of the population over the last couple decades. But implicit in that belief is that it’s those people, voters, for lack of a better word the audience of national politics. Elites or high level appointees or operatives may cynically participate in this flimflam. But somehow they’re not part of the process, they not stewing in the same cauldron. They’re cynical, amoral, pick your description.

This is a major blindspot. Bill Barr is another Republican guy in his late 60s who’s been living, as Miller puts it, in that Fox News/GOP legal circles cocoon for two decades. Why would he be any different from your birther uncle you avoid at holiday dinners?

More to the point, why would we be in the current situation if the bacillus of Foxism or rightwing authoritarianism (whatever you want to call it) wasn’t as pervasive with the Bill Barrs of the GOP as the ordinary Joes you see at the Trump rallies? More articulate, yes. But different? Not really. And why would it be?

Beyond the stonewalling and outrageous comments from Barr yesterday, one thing that struck me is that more than a few times he didn’t seem familiar with basic facts of the case or the Report. I don’t mean points in dispute between pro and anti-Trump commentators. I mean, basic factual details. It wasn’t clear to me he’d actually read the Report itself. At least some of his arguments seemed based on Republican commentaries rather than the actual document. Much the same applies to his comments about 2016 “spying”. This isn’t to excuse any of Barr’s lawless and now, in at least certain cases, criminal behavior. But it’s not clear to me he’s even sweating the details on behalf of his authoritarian aims.

"Her Pilgrim Soul"

With all of the attention going to the new Twilight Zone (haven't gotten around to it -- the queue is long -- but I'm big Jordan Peele fan), I thought it would be a good time to look back at the uneven but often very good 1985 series. It was the show that brought J. Michael Straczynski to the grown-up table and started his surprising long collaboration with the notoriously difficult Harlan Ellison.

The show also revealed a surprisingly sensitive side to Wes Craven, who directed a number of episodes. Here is perhaps the best.




Nice to hear Charles Aidman's voice again.

Stephen Repetski actually read that enormous environmental assessment from Musk's Boring Company

And here are some of the highlights.

Here's the Boring Company's draft DC-Baltimore "Loop" Environmental Assessment: https://t.co/pTkpugoidw

— Stephen Repetski (@srepetsk) April 18, 2019

The "purpose and need" of the Boring Co's "Loop" is that we need "high-speed passenger transportation between Washington, D.C. and Baltimore."

Isn't that called Amtrak? Or MARC? pic.twitter.com/of6BVdfraE

— Stephen Repetski (@srepetsk) April 18, 2019

"the above-ground infrastructure associated with Ventilation Shafts
would typically be behind tree-lines and therefore screened from view from B-W Parkway"

Yes, because I would hate to see some ugly buildings when driving along the Parkway in my car.

— Stephen Repetski (@srepetsk) April 18, 2019

Excuse me but wasn't part of what Boring Co was doing different from everybody else is reusing the dirt removed from the tunnels and turning it into bricks? But now they're going back to disposing it? pic.twitter.com/f7SFeh30iM

— Stephen Repetski (@srepetsk) April 18, 2019

"DUE TO LIMITED SIZE OF THE WASHINGTON, D.C. LOOP STATION LOCATION, INITIAL OPERATION OF THE LOOP SYSTEM WOULD BE LIMITED TO 1,000 PASSENGERS PER DIRECTION PER DAY." pic.twitter.com/n45iug1Bzh

— Stephen Repetski (@srepetsk) April 18, 2019

The daily initial operating capacity of the Boring Co's DC-Baltimore system is equivalent to that of a single WMATA 8-car train.

Except WMATA's trains run more than once per day.https://t.co/x8mK78UG3M

— Stephen Repetski (@srepetsk) April 18, 2019

Due diligence in Boring Co's EA means they list their competitor transit modes between DC and Baltimore, which they say include regularly-scheduled bus and rail service that are "inexpensive, reliable" and "frequent" for a cost starting at $8. pic.twitter.com/dnslTrkiJY

— Stephen Repetski (@srepetsk) April 18, 2019

TBC DC-Baltimore Loop would pass through several aquifiers. The tunnels themselves would be fairly shallow, 30 to 44 feet down. pic.twitter.com/vwZGXthtne

— Stephen Repetski (@srepetsk) April 18, 2019

I've seen nothing in this document how TBC plans to deal with water infiltration and outflow pumping infrastructure.

— Stephen Repetski (@srepetsk) April 18, 2019

Repost: Netflix's social media problem

Twitter was pretty much wall-to-wall Games of Thrones this weekend. Though the show does still have a marketing budget, it's safe to say the buzz is now mostly self-sustaining, unlike most of its competitors. Walking Dead did that. As did the Voice. Possibly the Americans.



GoT's impressive social media presence got me to thinking about this post from a few years ago.

Thursday, June 20, 2013

Curiously, agressively anti-social

As previously mentioned Certain business models limit you to certain marketing approaches. For example, the standard model for scripted cable series is to run weekly for about three months usually following long story arcs with start dates varying from show to show. This model lends itself to promotion through blogs and social media and it may not be a coincidence that original, scripted shows have increased greatly in popularity and influence over the past dozen years along with social media. When it works well, these shows can create a powerful weekly cycle of buzz and feedback starting with Twitter traffic during the actual broadcast and building from there.

The sheer volume of tweets, posts and podcasts we're talking about is astounding and it's made even more valuable because it bypasses our normal anti-advertising filters. These are people we know recommending a show. What's more, there's a tremendous social norming aspect. Watching the show become part of what's expected.

Keeping that in mind, think about the Netflix direct-to-binge model. Social media thrives on having a critical mass of people sharing a common experience.  With a shows like House of Cards, the kind sustained build-up you see with a Game of Thrones is impossible and even an ordinary discussion requires you to find a group who are at same point in the viewing.

Ted Sarandos, Chief Content Officer of Netflix, responds to this concern with a truly extraordinary statement:

“No one has ever watched anything on Netflix that they couldn’t watch all at once,” Sarandos said. There was no interest in changing that model for a new group of originals. But that not only meant changing consumer behavior, it also meant dealing with the realities of today’s social network environment.

Sarandos called it a “different style of watercooler etiquette.” Rather than having to deal with the weekly conversation that is produced, viewers need to ask each other which episodes they’re watching and dealing [sic] with that. Still, the strategy seems to be paying off, as viewers are continuing to tune in.

(quick aside: "paying off" implies improvement over what would have happened otherwise. By this standard you could argue that having disgusting bathrooms "pays off" for a filling station as long as someone still buys gas there.)

Sarandos is saying that part of the company's strategy is to get viewers to engage in less word of mouth promotion. That's an amazing position, hoping that people will refrain from conversation until everyone has had a chance to catch up on all thirteen hours of a show. Of course, by the time that happens (assuming it ever does), the show will be an old topic for the people who watched it when it first came out.

In an age where social media is generally considered the inevitable wave of the future, Netflix is launching a programming model based on people talking less about their shows. It's possible that there's some method to the madness here.

Of course, it's also possible these people haven't thought this through.

"The notion of singularity ... is a religion"

This essay from Joi Ito makes a lot of points that will feel familiar to regular readers (mistaking S-curves for exponential when discussing technological progress, the rise of cult-like thinking -- what we've been calling magical heuristics). Lots of other good stuff as well, which is pretty much what you'd expect from the director of MIT’s Media Lab.

The notion of singularity – which includes the idea that AI will supercede humans with its exponential growth, making everything we humans have done and will do insignificant – is a religion created mostly by people who have designed and successfully deployed computation to solve problems previously considered impossibly complex for machines.

They have found a perfect partner in digital computation, a seemingly knowable, controllable, machine-based system of thinking and creating that is rapidly increasing in its ability to harness and process complexity and, in the process, bestowing wealth and power on those who have mastered it.

In Silicon Valley, the combination of groupthink and the financial success of this cult of technology has created a feedback loop, lacking in self-regulation (although #techwontbuild, #metoo and #timesup are forcing some reflection).

On an S-curve or a bell curve, the beginning of the slope looks a lot like an exponential curve. According to systems-dynamics people, however, an exponential curve shows a positive feedback curve without limits, self-reinforcing and dangerous.

In exponential curves, Singularitarians see super-intelligence and abundance. Most people outside the Singularity bubble believe that natural systems behave like S-curves, where systems respond and self-regulate. When a pandemic has run its course, for example, its spread slows and the world settles into a new equilibrium. The world may not be in the same state as before the pandemic or other runaway change, but the notion of singularity – especially as some sort of saviour or judgment day that will allow us to transcend the messy, mortal suffering of our human existence – is fundamentally a flawed one.

Revisiting the New York Magazine pseudo-science thread -- now with consequences

And pretty damned horrifying ones at that.

As of this week, there have been 695 cases of measles in the U.S. across more than 20 states this year—the highest annual toll seen since the disease was declared extinguished in the U.S. in 2000, according to the Centers for Disease Control and Prevention. Given that it’s only April and we’ve already broken a yearly record, it’s worth wondering: Just how much worse could things get?

Measles is a highly contagious virus, capable of infecting someone through airborne droplets left behind by someone else, even hours after they’re no longer present. But measles’ one major weakness is humanity itself. Humans are the only natural host the virus uses to reproduce and spread. That means if you can fully stop the chain of transmission between people—by vaccinating practically everyone who could be exposed to it, for instance—you can eradicate measles completely.

In the U.S., the eradication of measles was formally declared in 2000, thanks to a tremendous public health effort and a mandatory vaccination program. But since there are still parts of the world where measles happens regularly, even with vaccination, travelers have continued to catch measles somewhere else and bring it to the U.S. Because most Americans continue to be vaccinated against it at an early age, though, outbreaks and cases of measles since 2000 have largely been isolated.

The anti-vaccination movement, however, has provided the kindling for this resurgence in measles, according to Peter Pitts, former associate commissioner for external relations at the Food and Drug Administration and president and co-founder of the Center for Medicine in the Public Interest.

“This measles epidemic is a perfect storm of vaccine denialism, stupidity, and groupthink,” he told Gizmodo.

As has been widely reported, one of the major epicenters of the anti-vaxx movement was the pricier sections of the west side of LA, which brings us to...

Tuesday, July 18, 2017

David Wallace-Wells, autism and bad science

David Wallace-Wells has been catching a lot of flack (most of it richly deserved) for his recent New York Magazine article on climate change. It is a hugely troubling sign when the very scientists you were claiming to represent push back against your article.

This controversy illustrates a larger problem with science reporting at the magazine. We already have a post in the queue discussing the neutral-to-credulous coverage of topics ranging from homeopathy to magic crystals to Gwyneth Paltrow's goop empire. The Wallace-Wells piece takes things to another level and goes in a very different but arguably worse direction. Rather than giving bad science a pass, he takes good science and presents it so ineptly has to do it a disservice.

I am not going to delve into that science myself. The topic has been well covered by numerous expert and knowledgeable writers [see here and here]. The best I could offer would be a recap. There are some journalistic points I may hit later and I do want to highlight a minor detail in the article that has slipped past most critics, but which is perfectly representative of the dangerous way Wallace-Wells combines sensationalism with a weak grasp of science.

Other stuff in the hotter air is even scarier, with small increases in pollution capable of shortening life spans by ten years. The warmer the planet gets, the more ozone forms, and by mid-century, Americans will likely suffer a 70 percent increase in unhealthy ozone smog, the National Center for Atmospheric Research has projected. By 2090, as many as 2 billion people globally will be breathing air above the WHO “safe” level; one paper last month showed that, among other effects, a pregnant mother’s exposure to ozone raises the child’s risk of autism (as much as tenfold, combined with other environmental factors). Which does make you think again about the autism epidemic in West Hollywood.

No, David, no it doesn't.

I want to be painstakingly careful at this point. These are complex and extraordinarily important issues and it is essential that we do not lose sight of certain basic facts: by any reasonable standard, man-made climate change is one of the two or three most important issues facing our country; the effect of various pollutants on children's mental and physical development should be a major concern for all of us; high ozone levels are a really bad thing.

But the suggestion that ozone levels are causing an autism epidemic in West Hollywood is both dangerous and scientifically illiterate. You'll notice that I did not say that suggesting ozone levels cause autism is irresponsible. Though the study in question is outside of my field, the hypothesis seems reasonable and I do not see any red flags associated with the research. If Wallace-Wells had stopped before adding that last sentence, he would've been on solid ground, but he didn't.

Autism is frightening, mysterious, tragic. This has caused people, particularly parents facing one of the worst moments imaginable, to clean desperately to any explanation that might make sense of their situation. As a result, autism has become a focal point for bad science, culminating with the rise of the anti-vaccination movement. There is no field where groundless speculation and fear-mongering are less welcome.

So, if ozone and other pollutants may contribute to autism, what's so bad about the West Hollywood claim? For that, you need to do some rudimentary causal reasoning, starting with a quick look at ozone pollution in Southern California.

Here are some pertinent facts from a 2015 LA Times article:

EPA Administrator Gina McCarthy selected a limit of 70 parts per billion, which is more stringent than the 75 parts-per-billion standard adopted in 2008 but short of the 60-ppb endorsed by environmentalists and health advocacy groups including the American Lung Assn. The agency’s science advisors had recommended a limit lower than 70 -- and as low as 60.

...


About one-third of California residents live in communities with pollution that exceeds federal standards, according to estimates by the state Air Resources Board.


Air quality is worst in inland valleys, where pollution from vehicles and factories cook in sunlight to form ozone, which is blown and trapped against the mountains.


The South Coast air basin, which includes Los Angeles, Orange, Riverside and San Bernardino counties, violated the current 75-ppb ozone standard on 92 days in 2014. The highest ozone levels in the nation are in San Bernardino County, which reported a 2012-2014 average of 102 parts per billion.

Now let's look at some ozone levels around the region. West Hollywood, it should be noted, is not great.

But just over the Hollywood Hills, the situation is even worse.

Go further inland to San Dimas and the level is even higher…

Higher still in Riverside ...

Though still far short of what we find in San Bernardino.

If you look at autism rates by school district and compare them to ozone levels, it is difficult to see much of a relationship. Does this mean that ozone does not contribute to autism? Absolutely not. What it shows is that, as with many developmental and learning disabilities, the wealthy are overdiagnosed while poor are underdiagnosed. It is no coincidence that a place like Santa Monica/Maibu (a notorious anti-vaxxer hotspot) has more than double the diagnosis rate of San Bernardino.

The there's this from the very LA Times article by Alan Zarembo that Wallace-Wells cites [emphasis added]:

 Irva Hertz-Picciotto, an epidemiologist at UC Davis, suspects that environmental triggers such as exposure to chemicals during pregnancy play a role. In a 2009 study, she started with a tantalizing lead — several autism clusters, mostly in Southern California, that her team had identified from disability and birth records.

But the hot spots could not be linked to chemical plants, waste dumps or any other obvious environmental hazards. Instead, the cases were concentrated in places where parents were highly educated and had easy access to treatment.

Peter Bearman, a sociologist at Columbia University, has demonstrated how such social forces are driving autism rates.

Analyzing state data, he identified a 386-square-mile area centered in West Hollywood that consistently produced three times as many autism cases as would be expected from birth rates.

Affluence helped set the area apart. But delving deeper, Bearman detected a more surprising pattern that existed across the state: Rich or poor, children living near somebody with autism were more likely to have the diagnosis themselves.
Living within 250 meters boosted the chances by 42%, compared to living between 500 and 1,000 meters away.

The reason, his analysis suggested, was simple: People talk.
They talk about how to recognize autism, which doctors to see, how to navigate the bureaucracies to secure services. They talk more if they live next door or visit the same parks, or if their children go to the same preschool.

The influence of neighbors alone accounts for 16% of the growth of autism cases in the state developmental system between 2000 and 2005, Bearman estimated.

In other words, autism is not contagious, but the diagnosis is.