Breaking Ground
Construction
In the New York Times, Ezra Klein describes the state of construction in America as ‘disturbing’, fretting over the distressing implications for our economy. Klein is often considered an Explainer, distilling complex topics for the average Joe, but this piece seems to ask more questions than it answers. He claims the US is ‘worse’ at construction than it was fifty years ago, but admits no one - especially not economists - seems to know exactly why, or how.
The economists Klein cites use productivity metrics - how much construction is produced by each worker:
A construction worker in 2020 produced less than a construction worker in 1970, at least according to the official statistics. Contrast that with the economy overall, where labor productivity rose by 290 percent between 1950 and 2020, or to the manufacturing sector, which saw a stunning ninefold increase in productivity.
I can’t think of any differences in the sort of things being built in 1970 and 2020. The former construction executive Klein speaks to does mention all the new building and safety regulations which drive up costs and decrease per-worker productivity - noting it’s ‘well worth it’ that workers may avoid crippling injuries - but, again, a clean answer evades him.
Spending on construction has not decreased, because people need new places to live, shop, eat, and work, and the American economy runs on them having places to do so. So, humoring Klein’s overly reductive question for a moment, what is preventing us from building the modern utopia economists are clamoring for?
One thing a good city needs is transportation. In the past, people could trade the quiet of rural and suburban living for the convenience of city trains and subways. The problem is, much of the country’s transit architecture is decaying or outright broken - the NYC subway system hasn’t kept pace with its peers overseas.
Attempts to modernize the city’s subways have made headlines - topping out at $3.5 billion for a single mile of track. Hidden inside the numbers are a myriad of reasons why cities struggle to maintain and build new infrastructure. An NYU study sheds light on why it costs so much and takes so long to build anything:
The results of NYU’s Transit Costs Project are a little distressing because there isn’t one graspable, fixable issue; everyone with whom the MTA interacts, from contractors to other agencies to its own internal processes, adds to the cash-draining effect.
The MTA is building stations far bigger than necessary, because city departments tasked with maintaining the subways insist on having their own maintenance rooms. Different stations use different vendors for escalators, elevators, and other critical parts - meaning maintenance is more time consuming and costly. Various city services departments sidle up to the trough to use subway funds to fix water pipes or ‘sell’ the agency unused land to use. Politicians kowtow to their donors and local interest groups, demanding extra stations or features for their residents.
Then there are the consultants:
“The MTA used to have a capital-management arm that designed 90 percent of their projects, and I don’t want to say it was a golden age, but they had 1,600 people on staff. In 2011, when MTA was building Second Avenue and the 7 extension and East Side Access and Fulton Street, MTA Capital Construction had 124 employees.”
As the city outsourced design and procurement, everything became less efficient, because of course it did. Consultants are incentivized to drag projects out and stuff them full of pork, because it creates problems only they can solve. The result of decades of decay in public sector effectiveness, political intransigence, and the rise of the consultant class make it impossible to build anything useful in a timely or cost-effective manner. Every big infrastructure project is designed to monetarily and politically enrich everyone involved - social good has become an afterthought.
What about private development? NYC - admittedly an extreme example - is so starved for space developers have resorted to buying air rights to build luxury skyscrapers. In addition to being extremely expensive, these buildings require extreme engineering on the edge of what is physically possible - the book Supertall is an excellent read if such things interest you. Additionally, new residential construction in New York is a fraught process - even the most community-minded developer may have to endure years of legal and political challenges as a project winds its way through onerous approval processes. In high-priced urban markets, this means only deep-pocketed developers can get anything built - typically expensive commercial and residential space - which does little to help people without the spare change to buy 8-figure condos a thousand feet in the air.
The nature of what kinds of things America is building, the complex and politically fraught approval processes, and the outsourcing of every bit of construction to specialized firms all contribute to how productive our construction industry is. Consultants, lawyers, developers, and politicians all benefit from the current system, which explains why no one is in any rush to change it. And, let’s be honest, construction workers are often paid by the hour or job, so if things take longer to get built, they and their unions have no reasons to complain.
Parking
We can’t talk about America’s ridiculous building codes and city planning without discussing cars. The country’s century-long obsession with automobiles has shaped our cities, our transportation systems - why doesn’t anyone fret about how expensive and time-consuming it is to construct giant highways? - and contributes to the languid pace at which we build.
Nearly 93% of American households own a car. Most states average two cars per family. But we have eight parking spaces per car in this country. Insane! In most of the country, parking is a requirement for residential and commercial construction. Suburban and rural areas may have the space to build sprawling lots, but in a city parking means garages:
A 2016 study found that it cost roughly US$24,000 to $34,000 to build every single new parking space in a garage – an amount that has surely grown.
They aren’t going to get any cheaper, either, as American cars get bigger every few years. Garages built underground can be even more expensive and - you guessed it - slow down the building process further.
Parking adds to construction time, complexity, and ultimately cost. In cities without parking reform, this leads developers to stuff buildings with dead space - all those amenity rooms you may have seen in a nice apartment or condo complex aren’t charity, they’re cost savings for the developer. Which means less living space. Which means more buildings.
Our country’s obsession with cars makes everything bigger, dumber, slower, and worse for the environment. But! What if it wasn’t quite so awful? What if we could take parking lots and suburban sprawl and use it to help alleviate our dependence on dirty fossil fuels to power the country?
High Country News estimates that if we put solar panels on Walmart roofs and installed solar canopies above existing parking lots, we could power millions of homes. Rather than tearing up undisturbed land to build massive solar farms, we could simply stick them atop all the awful eyesores we currently have. Think of all the construction productivity building millions of solar panels would create!
It is easy to poke fun at pundits trying to puzzle out why we Just Don’t Build Things Anymore, but the truth is we build tons of things, and a lot of it is dumb and wrong or needlessly complicated by a hundred middlemen and profiteers. Using economist metrics to gauge societal progress is often the wrong place to start. The question we should ask is: how do we make our living spaces better, less damaging to the planet, and more socially valuable. Until we change the perverse incentives shaping development and construction, builders will keep erecting awkward edifices no one can afford to live in, and economists will continue to ask why they aren’t building them faster.
AI
It’s been a tough couple weeks for chatbots. Last week, Google’s launch of its Bard AI sent its stock tumbling as the bot got basic facts wrong in response to questions during a demo. Microsoft, riding high on its investment in OpenAI, released a beta version of Bing with the GPT-3 AI engine attached. Thus far the bot has proven combative, defensive, and emotional - in addition to simply getting (searchable!) facts wrong.
What gives? Insider’s Adam Rogers provides one explanation:
These chatbots, in other words, are not actually intelligent. They are lying dumbasses.
His larger point is one I’ve made before - training AI on the information available on the Internet is going to result in a lot of bad information. Grifters, SEO sneaks, and propagandists have spent decades stuffing search engines full of believable bullshit, and researchers have no choice but to train their AIs on it - compiling a dataset of only ‘good’ information, vetted by researchers or whatever, would be time consuming if not impossible. Machine learning models require huge amounts of source information to triangulate responses to all our silly or esoteric questions.
Unfortunately, as Rogers points out, humans are very prone to believing exactly the sort of misinformation AIs are prone to give us:
The AI-generated messages, it turns out, were just as convincing as the human ones. But the wild part is why. When the researchers debriefed their human subjects, those who preferred the chatbot articles said the artisanal, human-made messages relied too much on anecdote and imagery. GPT-3 was more evidence-based and well reasoned. The very quality that made the robot less human made humans more likely to believe it.
A core appeal of ChatGPT is its convincing sounding answers, even when they are wrong. A chatbot is programmed to provide straightforward, concise responses, and isn’t prone to human bloviating or distraction. Human brains are wired to find these answers credible. No wonder academics and ethicists are worried.
Financial markets and the tech press want to believe the future of AI chatbots is now, but mere months after the release of the most successful AI to date, the wealthiest, most powerful tech companies find themselves unable to get their bots to distinguish fact from fiction.
In a bit of grim irony, tech firms are cutting what little support they had to police misinformation on their platforms, leaving social and search networks even more prone to dangerous bullshit. Which, of course, is the bullshit they’ll be training their AIs on. It’s hard to fathom a world where we can’t trust anything we find online, but tech firms seem to be charging headlong in that direction.
Don’t Steal Bitcoins From The IRS
But while Larry [Harmon] was cooperating with the government, his younger brother – who was formerly employed by one of Larry’s companies, Coin Ninja – was stealing from it, pilfering crypto from Larry’s forfeited wallets.
According to a newly released memorandum filed by prosecutors in August 2021, Gary Harmon used recovery seed words to recreate numerous wallets that belonged to his brother. In April 2020, he made a series of eight transfers from Larry’s forfeited wallets – Trezor wallets being held in an IRS storage locker – to his own wallets. In total, the younger Harmon brother purloined a total of 712 bitcoins from the IRS – at the time, worth $5.4 million.
Prosecutors said that Gary initially vehemently denied draining his brother’s wallets, even when presented with evidence that he had done so.
Bro! Come on. We haven’t trotted out the Crime Do's and Dont's in awhile, but stealing crypto from wallets under government control feels like it should be on the list. Gary managed to violate another cardinal rule while he was at it:
Photos recovered from his cell phone show a lavish evening out at a club: Gary bathing in a tub full of cash, a blissful smile on his face as scantily clad dancers mingle behind him. A text reveals that Gary paid $122,232 for the privilege of swimming in the pool of 100,000 one-dollar bills surrounded by the dancers.
There really is a service catering to every rich idiot’s whim. Want to swim in a pool full of money? They’ve got you covered. Anyhow, the IRS caught Gary a year after the theft, and he is facing up to forty years in prison, so I hope he enjoyed himself.
Short Cons
Forbes - ““I don’t know that [Slync] was a business so much as it was a kleptocracy,” a former employee told Forbes in July. “Chris Kirchner was ultimately using sports to buy access to things he wouldn't have had as a regular guy.””
Grist - “After months of tense negotiation, a half-dozen states have reached an agreement to drastically cut their water usage and stabilize the drought-stricken Colorado River — as long as California doesn’t blow up the deal.”
Platformer - “…after Musk threatened to fire his remaining engineers, they built a system designed to ensure that Musk — and Musk alone — benefits from previously unheard-of promotion of his tweets to the entire user base.”
Know someone thinking of writing a NYT op-ed about the American construction industry? Send them this newsletter!