Jeremy Côté

Adult Spectacles in Youth Sports

Tue, 16 Dec 2025 00:00:00 +0000

Here’s a heartwarming memory from my summer:

Two men shouting invectives at each other in a parking lot, the specifics unfit for print but with a generous helping of violent threats. Me—and likely many onlookers—wondering if this would evolve to blows. One of the men backing away towards his vehicle (smart), but continuing to hurl insults with the same ferocity (not smart). And the real lovely bit: the young boy listening as he trotted beside the man, presumably his father.

What had precipitated this spectacle between two adults on a Sunday afternoon? A semifinal soccer game. Between 14 year-olds.

And what better place to do it then a public parking lot! A busy tournament schedule meant adults, teenagers, and young children milled about, all within earshot of these two men. I was nearby preparing to coach my own team. Regardless of our wishes, we all secured tickets to this impromptu spectacle.

Their shouting match hadn’t emerged from nothing, of course. I had seen how the civility of some supporters frayed as the game marched towards its shootout climax. I had noticed how the flow of supporters from both teams converged as they exited the field after the game, a recipe for barbed comments and post-game fireworks. The spectacle I witnessed that afternoon felt like the inevitable culmination of many small moments.¹

I hope we can agree that such a spectacle has no place in sports, let alone youth sports. And yet, I’ve observed similar spectacles. I’ve seen clusters of parents from opposing teams get into shouting matches during a game. I’ve seen parents and coaches yell at referees for “bad” calls. I’ve seen coaches berate their players. All within an environment whose main purpose is supposedly for kids to have fun.

I doubt anyone wants these spectacles to occur. So why do they? I think it’s because sport creates a cocktail of stakes and emotion for coaches, spectators, and athletes. Every touch of the ball, swing of a racquet, step, throw, or jump is a moment in a tug-of-war between competitors. Everyone wants to win. But sport is abundant with challenges: tough opponents, silly mistakes, and calls that make us want to bring a referee to an optometrist.. Mixed together, this cocktail triggers spectacles.

I’m not surprised when athletes struggle to maintain composure in the face of this cocktail.² I see it all the time in my own adolescent athletes (and have continued to see it while writing this essay). Trying to perform your best while also not wanting to fail and look like a fool to their peers isn’t easy. External frustrations such as referee calls and opponent behavior only compound the difficulty of handling the stressors of sport with grace. This doesn’t mean I give my athletes a pass on their actions, but it does invite compassion from me as their coach.

The standard must be different for adults. My raison d’être in youth sports is simple: support the development of our kids and create an environment for them to have fun competing. These are the two pillars on which youth sports should stand. As a coach, I place them above winning. Don’t get me wrong: I love winning and strive to do so every day. But winning is a byproduct of competing, a much more useful habit that I try to instill in my athletes. You don’t have full control over winning, but you always have full control over the effort you give in every moment. In other words, you control how hard you compete.

What undermines this development? Adult spectacles in youth sports. When the two men made a scene after the soccer game, they sent a message to all kids within earshot: this is how we act after a game. When the coach of the losing team vented to a referee about a missed call following a game-tying goal, he sent a message to his athletes: it’s okay to waste precious mental energy on complaining to a referee who will never overturn the goal. What messages do we want to send our athletes?

Our actions set the standard for kids in youth sports. As spectators, parents, or coaches, we can support our athletes’ development or we can undermine it. When I’m coaching, I ride a rollercoaster of emotions like anyone else. But expressing it helps nobody. Worse, I set a bad example for my athletes and distract them from their task. My job as a coach is to help them focus on the game, improve themselves, and have fun. Nowhere does my job description say, “complain to the referee and/or scream from the sidelines throughout a game.” I’m not perfect, but this is the standard I strive towards.

Sports offer wonderful opportunities for growth and development through challenges and camaraderie. If I want to provide these for my athletes, I can’t get sucked into the emotional pull of sports. That means no yelling at referees, no emotional explosions to the athletes, and always a focus on their objective. The same applies to parents.

Every adult is a role model in youth sports, especially parents and coaches. We must take that duty seriously. If not, youth sports is simply entertainment for adults. That’s unacceptable to me. We should never find ourselves shouting violent threats to each other after a game between young kids. We can—and must—do better.

I don’t know these two men, and chances are I’ll never see them again. I also don’t have the entire context of their conversation, of what transformed tension into hostility. To give them credit, shouting was better than transitioning to physical violence. ↩
As per the time-honored tradition for many young Québécois boys, I played hockey. I developed a habit familiar to many athletes: questioning the calls of referees. By “question”, I of course mean I would yell at the referees. It didn’t matter if the call was good or bad; it just had to affect my team negatively. Have you seen sports trading cards for professional athletes? One year, my team made similar cards for all of the players. On mine, my coaches immortalized what they saw as my most frequent catchphrase: “C’mon ref!” While it’s a funny story now, I also cringe inwardly at my earlier self. ↩

An Almanac, Not An Index

Tue, 11 Jun 2024 00:00:00 +0000

In 1921, Lewis Fry Richardson wasn’t satisfied with the current state of weather forecasting. His criticism of the paradigm was fundamental: it relied on pattern-matching instead of modeling the physics of the atmosphere. In the preface of his book Weather Prediction By Numerical Process, which would go on to revolutionize how meteorologists forecast the weather, he described the dominant approach:

The process of forecasting, which has been carried on in London for many years, may be typified by one of its latest developments, namely Col. E. Gold’s Index of Weather Maps. It would be difficult to imagine anything more immediately practical. The observing stations telegraph the elements of present weather. At the head office these particulars are set in their places upon a large-scale map. The index then enables the forecaster to find a number of previous maps which resemble the present one. The forecast is based on the supposition that what the atmosphere did then, it will do again now. There is no troublesome calculation, with its possibilities of theoretical or arithmetical error. The past history of the atmosphere is used, so to speak, as a full-scale working model of its present self.

Richardson, as well as meteorologists Cleveland Abbe and Vilhelm Bjerknes, wanted to infuse the discipline of meteorology with a more scientific approach. Indeed, Abbe wrote in 1901 that the predictions, “merely represent the direct teachings of experience; they are generalizations based upon observations but into which physical theories have as yet entered in only a superficial manner if at all.”

I was unable to find the Index of Weather Maps (if anyone has a digital copy, please get in touch!), but it seems as though Richardson was referring to Ernest Gold’s Aids to Forecasting (1920). The book is so old that I do not have a copy of it either, but a review in Science notes that Gold surveyed the Met Office’s Daily Weather Report for fourteen years of its reports and categorized the weather into similar “indices”. A forecaster could then check the index to find a match for current weather and follow how the weather unfolded in those past examples. The weather prediction would be the closest match to present conditions. However, as the Science reviewer noted, “The weakness of any such classification is, of course, the assumption that like surface isobaric conformations are always followed by similar weather conditions.”

To give you a sense of what these weather reports looked like, below are a few pages for the January 1, 1910 entry. It’s essentially a logbook, where the writer collates all the incoming information from the weather stations, fills out the blank maps to show readings for the wind, pressure (barometer), and temperature, and includes particular notes for the day.

A spread from the Met Office’s Daily Weather Report for January 1, 1910. The layout of the pages is the same for each day, and simply the notes and annotations change. Shown here is one day of many included in the scan. Source

Richardson wasn’t satisfied with searching through an index though. He admired a different publication: The Nautical Almanac (a log of predictions for star positions to help with sea navigation, as well as official dates for holidays). He wrote, “that marvel of accurate forecasting, is not based on the principle that astronomical history repeats itself in the aggregate.” If you take a look at this copy, you will quickly encounter formulas and mathematical expressions. This is in contrast to the weather reports, which were just measurements. While Richardson valued data (this is how he made his predictions, after all), he thought simply comparing to old data was bound to be a dead end. Instead, he wanted to model the weather as a physics problem. And he did, resulting in a framework that helps us forecast the weather much more accurately into the future than we once did.

What’s striking to me about Richardson’s criticism is that it boiled down to this: Pattern matching with data isn’t enough for useful prediction. And while that may seem reasonable, haven’t we just seen a remarkable reversal of that idea in the form of machine learning? Large language models in particular are becoming more and more capable of accomplishing tasks. Can we use machine learning for weather prediction?

There are three reasons to be hopeful about machine learning techniques in weather prediction. First, meteorological stations have rich data sets stretching back decades (and some for over a century). This means extensive training data for models¹. In fact, the European Centre for Medium-Range Weather Forecasts (ECMWF) reports that their forecasts take up terabytes of data, and will significantly increase over time (see Figure 11 of the report)². Second, there’s always a ground truth for weather forecasts. Assuming the observational instruments are working correctly, we can rapidly see if a prediction was right or wrong. This helps when evaluating the performance of a model. Third, running a traditional weather prediction model is expensive. That’s because you must simulate the physics of the atmosphere, which has feedback loops at multiple scales and requires solving many interdependent equations. The exact energy, time, and cost footprints depend on the model, hardware, and kind of simulation (near-term weather versus climate predictions), but climate scientist Tapio Schneider said in a recent article that a climate simulation can consume up to 10 megawatt hours of energy. For context, the article notes that this is about the average energy consumption of a U.S. household in a year. While it’s true that training machine learning models is an upfront cost, there are huge efficiencies in actually outputting a forecast, such as taking one minute on a single TPU to produce a 10-day forecast versus hours on supercomputers with hundreds of machines³.

What caught my attention though was the involvement of the ECMWF in exploring machine learning for their operations. The ECMWF is a research institute but it is also an operational service, providing weather forecasts every day to many countries. It’s one thing to have academics interested in using a new type of tool to solve a problem. It’s quite another for an organization who has a product to consider incorporating a new tool into their workflow.

In 2021, the ECMWF released a ten-year roadmap for how they plan to incorporate machine learning methods into their workflow. They note that, “Machine learning can be used to improve the computational efficiency of weather and climate models, to extract information from data, or to post-process model output, in particular if data-driven machine learning methods can be combined with conventional tools.” Though we have faster computers than ever before, conventional weather models always require more compute, because we want higher resolution, longer forecasts, and more simulations for predictions. Any gain in compute efficiency will translate into increasing these variables.

Many of their efforts and those in the past have occurred in only the past few years, which means there is still a ton to do in this space. The ECMWF’s own initial perspective was that machine learning was an interesting avenue to explore, but “the likelihood of it becoming operational was low, so it was not a wise investment of ECMWF resources.” And yet, they realized that the situation changed in 2022 and 2023, when companies such as NVIDIA, Huawei, and DeepMind released machine learning models that rivaled traditional methods.). These results had enough of an impact on ECMWF that they are now investigating how to incorporate machine learning into their workflow.

They also note that there are many challenges in using machine learning tools for weather forecasting. These include the requirement for custom machine learning solutions for the weather, refactoring code to take advantage of the machine learning libraries in Python or Julia (many conventional weather forecasting code is in Fortran), and barriers in communication between Earth scientists and machine learning researchers. The ECMWF’s plans to address these challenges over the course of their roadmap.

The future of operational weather forecasting probably won’t be just machine learning or just physics-based simulation, but a hybrid of the two. Hopefully, the ECMWF and other organizations can figure out how to augment and complement their workflows such that machine learning becomes an asset, not just a curiosity. I’m excited by the potential reduction in compute time using machine learning, which would allow meteorologists to combine a greater number of simulations to make an ensemble prediction.

Going back to Abbe, Bjerknes, and Richardson, it appears that enough data really can lead to good predictions. What we’ve seen in the past few years with machine learning is that scale matters. To echo what Philip Anderson once said within the context of condensed matter physics, “More is different.”

Endnotes

Note that older data may not be as useful, since it wasn’t as comprehensive as our data collection system today. ↩
They also say that this volume of data is becoming more and more difficult to handle, which poses a problem as they aim to increase the resolution of their forecasts (requiring more data). ↩
A tensor processing unit is a device that Google built specifically for machine learning applications. ↩

Take Care of the Conditions

Mon, 10 Jun 2024 00:00:00 +0000

It was halftime in a close tournament finals. My basketball players gulped water and breathed heavily. It was our fourth game in two days. But we were here, vying for a championship just a week after a disappointing early tournament exit. We were so close, yet there was still work to do.

I spoke before they returned to the court¹. “Guys, we have one last half, and I know we can do this. We can’t control if we win. The other team has just as much of a right to this game. But what we can control is how much energy we bring on defense, how smart we play on offense. We can control how mentally engaged we stay throughout every possession, every minute, and every shift. Are you ready to do it?”

They were. We won the game and the tournament.

As a coach, I was super proud to see my players band together and win. But I’m also aware that we could have lost. Some unlucky bounces, a few clutch shots from the other team, or our opponents maintaining their energy for longer could have made the difference. If any of these had happened, what would I have thought of my team?

I’d be just as proud, because they reached a new level of performance. Becoming champions was sweet, but it wasn’t necessary. Losing would have hurt because I’m competitive at heart², yet what thrilled me is that we did everything to put ourselves in the best position for success. Winning was just a marker of our growth, not the growth itself. In my view, this is what we should be seeking with most youth sports (and probably many other pursuits).

Take care of the conditions, and the conditions will take care of the outcomes.

The gist, because I don’t quite remember the exact words I used. ↩
I’m the type of person that has trouble not competing when playing sports socially. Giving anything but my best effort is foreign to me. ↩

Identity's Radius

Tue, 30 Apr 2024 00:00:00 +0000

There are experiences, commitments, environments, people, and roles which will grow your identity. There are also those that will shrink it. You won’t know which you will get in advance, but here are a few points to consider:

First, the larger your identity’s radius, the more resilient you will be in the face of setbacks. For example, if I’m struggling in my work, I can turn to my sports coaching as a way to relieve stress. But if I’m only ever working, then a setback looms large in my life. A larger radius protects against this. That being said, you can have too large of a radius, where you stretch every minute of your life to a pursuit, which gives you little room for margin. The result isn’t resilience, but brittleness.

Second, growing your identity’s radius through volunteering has done wonders in my life. It embeds me within my local community and gives me a sense of satisfaction independent from my professional work. As a coach for teenagers, I can assure you none of my players care if I’m struggling in my work. I need to be there for them, and that’s it!

Third, your identity’s radius is dynamic. You can always change it, and it will sometimes change on its own depending on the seasons of your life. A stressor can warp a small radius and make you only focus on it. A new opportunity can expand the space of possibilities for your identity. The challenge is keeping your radius large enough to not collapse under stressors but small enough such that you still have margin. In my experience, neither parts of the inequality are easy to achieve.

It’s natural that my identity’s radius will need shrink if I need to focus in a particular period of my life. But I try to remind myself that this is intentional and it has an end. Without an end date, I’m setting myself up for a miserable time.

When I get the balance right though, it feels like living on easy mode and makes me grateful for the opportunities I get in life.

The Glue

Fri, 29 Mar 2024 00:00:00 +0000

I was almost seventeen, not an arts and crafts kind of guy, and wasn’t looking to repair anything. So you can imagine my surprise when I unwrapped a bottle of glue as a gift.

This also wasn’t from some kind but clueless family member; it was from my two soccer coaches at the end of the season. In fact, they had gotten a gift for each of my teammates, and we were opening them one by one.

Though my expression likely betrayed my puzzlement, I do have manners. What I probably did was begin making a fool of myself by inspecting the glue like it was a rare and valuable treasure. Thankfully, my coaches took pity on me and presented me their actual gift:

“We got you glue because you hold the team together.”

A decade later, I still remember the feeling of my coaches really seeing me. I wasn’t great at soccer, I was just an athlete who could run a ton. But my coaches hadn’t missed me communicating on the field, nor the effort I gave each time I was on the field. They hadn’t missed how I cared about leading my team. They saw.

My coaches could have just congratulated us on the season and that would have been fine. It would have been expected. Instead, they went further. They may have bought us physical gifts, but what they really gave us was the gift of noticing.

You can’t buy what my coaches gave me. It requires time, care, and investment in your players to even gather the raw materials to create this gift. As a coach now, they are the archetype I choose to follow.

The half-life of the memories of what a coach does on the playing field may be short, but the half-life of such a gift can endure for years. It has for me.

Thank you Mohammad Khan, Dylan Kurt, and Josh Knox for feedback on this essay.

The Constellation of Progress

Thu, 29 Feb 2024 00:00:00 +0000

What does progress mean to you?

In November 2023, I met with my cohort for The Roots of Progress Fellowship in sunny San Francisco to tackle this question. We filled a wall with sticky notes of ideas, examples, and concepts. Certain categories of progress showed up: material, cultural, scientific, and health. We debated and discussed all morning, but ultimately we didn’t come up with a definition of progress that we all agreed on. Honestly, I’m not sure that’s even possible. Instead, we created a constellation of concepts.

That’s useful, but it also means we each carry a subset of this constellation in our heads. When we talk about progress, we’re drawing from this fragment, which can lead to misunderstandings. For example, if one person sees progress as safeguarding the environment and another sees progress as needing to develop and build out infrastructure as quickly as possible to support more people, there may be tension between them.

In my experience in science, one of the most difficult parts of talking with other scientists is establishing a common language. While it’s true that a researcher in a field closely related to mine will likely have overlapping language, that’s not a given (and we will still need plenty of discussion to clear out subtleties).

In a similar way, if we want to make the argument that progress as we envision it is something worth having, it’s worth being clear about what we mean. Otherwise, we will be speaking a foreign language to many people.

I began this essay with the intention of exploring scopes of progress, discussing how they were too limiting, and proposing my own. I left thinking that while I still like my scope (it’s mine, so that’s not surprising!), there’s a lot more nuance to progress that resists tidy scopes, and that’s not necessarily bad.

Scope 1: Progress as raising the material standard of living

I’m super sympathetic to this view of progress. My experience in mathematics and physics makes me appreciate crisp definitions that we can use to craft experiments and measures. If we restrict the scope of progress to raising the material standard of living (as Michael Magoon suggests), then it’s straightforward to derive a bunch of metrics that enable us to compare periods of history and different societies to determine where progress occurred and where it didn’t. Such a definition gives us a firm basis for material progress studies to be part of the social sciences. Don’t underestimate the power of a clear definition.

Within this scope, we can look at technology, the economy, housing, electricity, and infrastructure as levers to create progress. As someone who writes about infrastructure, I care about this view of progress! The downside is that this scope ignores other aspects of progress, such as scientific, moral, and political progress. By pushing these elements out of scope, I worry that we are not capturing what many people would count as progress.

Let me share a thought experiment to illustrate my second issue. Suppose I could survey everyone on their desired material standard of living, and then I could wave a magic wand to grant the world the standard from the highest response. By that definition, everyone should be satisfied with their standard of living. Increasing it further would mean progress (by this scope’s definition), but who would care? We would then have progress while simultaneously not benefiting more from that progress. That seems off to me. I think our idea of progress should be rich enough to not end when we’ve saturated any improvement in material standard of living that people care about.

Scope 2: Progress as more options

In discussing this topic with my cohort at The Roots of Progress, they noted that perhaps we could view progress as the increase in options for humans. I like this framing because it subsumes Scope 1. Increasing the material standard of living is one way to increase the number of options, so people can choose paths that lead to happiness. As Ryan Puzycki suggested to me, this view implies that progress is a means to an end, not an end in itself (this jives with my own view). But the framing of increased options also lets us incorporate aspects of progress that pertain to education, vocation, and culture. When you have more options, you don’t have to take a job that you hate just to pay the bills. You don’t have to serve the needs of others for survival. You can do what fulfills you (which could include serving the needs of others!).

Jason Crawford pointed out to me that this framing is pretty similar to that of “capabilities” from the work of Amartya Sen and Martha Nussbaum, among others. As he explained to me, the idea is that we can’t really measure well-being, nor can we provide it (since it involves individual choice). But we can offer more and more capabilities, such as wealth, knowledge, and human rights. Then, it’s up to everyone to make the best use of those capabilities as they can.

I find this framing to be useful, yet I still see some gaps.

The first issue I have with progress as more options is that more options don’t imply better outcomes. It seems plausible that we could tweak our defaults or the presentation of our options to deliver better outcomes. For example, physical activity is super important for long term health, particularly in kids. Yet under 40% of kids (5-17 years old) in Canada get the recommended hour of moderate to vigorous physical activity per day. While some of this is likely a material standard of living issue (think about the cost of sports), I wonder if there are other levers we can pull which would change our defaults (such as more physical education in school)?

The second issue is that options aren’t all equal. I’m unsure if ruthlessly increasing the number of options is a good proxy for progress. Instead, we probably want some more complicated measure of “quality” for options.

Therefore, I don’t think it’s enough to give people more options. We need to be mindful of how people make choices given options, or else I think we’re leaving an element of progress on the table.

Scope 3: Progress as creating and applying knowledge

There’s this quote from David Deutsch that I like: “If something is permitted by the laws of physics, then the only thing that can prevent it from being technologically possible is not knowing how.” It gives me hope that we can continue to solve our problems in the future. His quote suggests to me a way to phrase progress:

Progress is the development and application of ideas that lead to sustained increases in flourishing for large groups of people.

Here’s what I like about this phrasing:

It acknowledges both theory (development) and practice (application) which can fuel progress.
It centers ideas, which are flexible enough to include design choices, policy, science, and many other domains. Take the field of metascience. If we are able to find ways to tweak our scientific institutions so that they work better (even with the same financial resources), would you call that progress? I certainly would, which is what this scope allows. See Michael Nielsen’s and Kanjun Qiu’s essay for examples of such tweaks.
It emphasizes “sustained increases” to avoid focusing on noisy short-term trends.
It establishes flourishing as the byproduct and standard of value for progress.

Still, there are weaknesses:

It refers to “large” groups, which raises the question of where to set the threshold.
It doesn’t define a measure of progress. Worse, there’s the same drawback about it being difficult to measure flourishing or well-being as in Scope 2.
You may not agree that flourishing should be the standard of value.
Linking ideas to outcomes isn’t always straightforward, which makes it difficult to decide if an idea did fuel progress.

I think this scope is useful for thinking about some “units” of progress, yet I also can see that it’s not the full story.

As I was writing this essay and discussing it with others, it became clear to me that the word “progress” is overloaded with meaning and nuance. Please forgive me for the physics analogy, but it seems to me that when we say “progress”, we’re only speaking at low resolution. Imagine that the constellation of concepts I mentioned at the top of this essay was an actual constellation of stars. If you look at the stars from too far away, it becomes impossible to distinguish one point of light from another. It’s only when you get closer to the sources that you can resolve them¹:

When concepts overlap, the result is a blur (dashed lines). It's only when we increase the resolution and spread out the concepts can we distinguish them.

Similarly, if we don’t dig deeper into what we mean by “progress”, our constellation of concepts remains a single blurry point of light. That’s not useful for discussions, and it also can also remove nuance from the term. Personally, I feel alienated when I read pieces that are exclusively focused on material growth, as if this is the whole scope of progress. Even if that’s the lever that helps us achieve human flourishing, I think it’s worth being more explicit about it! I know it will help people (like me) that don’t necessarily connect all the dots.

Ryan also raised an excellent point: your standard of value matters a lot. What is it and who is setting it? In Scope 3, human flourishing is my standard of value. That may sound right to you, but maybe it doesn’t. Maybe you think we shouldn’t be only focused on humans. Or maybe “human flourishing” sounds good at first, but then as we dig deeper we discover that my concept of human flourishing and yours split apart like in the animation above. This is why it’s worth drilling down on specifics, or else even those who agree with you may be doing so on the wrong assumptions.

While I like the idea of having a unified definition of progress, that’s probably the theoretical physicist in me. What may be more useful is a definition of material progress, of scientific progress, of technological progress, of economic progress, and so on. The specificity limits scope creep, and lets us deploy smaller, modular definitions that build up to the constellation we call “progress.”

That’s why I now want to ask two questions whenever I hear the word “progress”:

Progress on what, exactly?
Progress by what standard of value?

Thank you Heike Larson, Laura London, and Alex Telford for initial discussions about this idea. Thank you Ryan Puzycki, Jason Crawford, and Laura for feedback on this essay. And thank you Michael Magoon for writing a bunch about the scope of human material progress, which inspired this piece.

Endnotes

In optics, you could also increase the diameter of the lens. ↩

A Lighthouse Cutting Through the Fog of Overwhelm

Tue, 09 Jan 2024 00:00:00 +0000

We might have lost the basketball game, but as a coach, it was one of the most inspiring performances I’ve seen from my team.

We had no business winning. Though we were good, this team was undefeated despite playing some of the other top teams in our league. As I watched our opponents move like athletic machines during warmup, I wondered if I was going to witness a bloodbath. But my team was undaunted. I could tell you about how we played physically, how we shot better than we had all season, how we kept the game close, and how we locked down on defense. But it would be easier to just say, “We played with belief.” If we hadn’t, we wouldn’t have just played average; the other team would have demolished us. We may have lost, but because my players believed in their skills and experience, they made it a much closer game than many would have predicted.

Belief isn’t just for sports. In my view, it’s a vital ingredient for creating change and making progress on some of the biggest problems we face. But this belief cannot grow out of nothing; it needs to stem from substantive data, reasoning, and knowledge. Otherwise, it’s easy to just sit back and let someone else solve our problems. The wrong kind of optimism is a liability, but the right kind propels us forward.

Getting the knowledge requires people doing the difficult work of digging through the scientific record, gathering data, and synthesizing it. It requires people like Hannah Ritchie, the Deputy Editor and Lead Researcher at Our World in Data, a non-profit that’s frankly a jewel of the internet for data exposition about the world’s problems.

As someone who has degrees in geoscience, carbon management, and global food systems, you might think that of course Hannah would be the kind of person to be optimistic. And she is, but that wasn’t always true. In her new book Not the End of the World, Hannah writes about her early universities years:

“In 2010 I started my degree in Environmental Geoscience at the university of Edinburgh. I showed up as a fresh-faced 16-year-old, ready to learn how we were going to fix some of the world’s biggest challenges. Four years later, I left with no solutions. Instead, I felt the deadweight of endless unsolvable problems. Each day at Edinburgh was a constant reminder of how humanity was ravaging the planet.” (Introduction)

She almost quit environmental science because doom dominated her worldview. After all, what’s the point in working in an area when you believe your efforts are futile? But then Hannah watched Hans Rosling present on metrics of human well-being, which changed everything for her. Rosling was showing how the world was actually getting better in many meaningful ways (such as the fraction of people living in extreme poverty and the estimated fraction of newborns that die before the age of five each year).

From there, Hannah began digging into the data more. She began zooming out, looking at historical trends instead of only reading the news. She took a job at Our World in Data, and her work over nearly a decade has led to this book.

Meeting the needs of the present and the future

Before we dive into any of the many metrics of progress that Hannah presents in her book, we need to agree on what we’re even striving for. Hannah’s book is all about sustainability, so she begins by sharing the definition of sustainability from a 1987 United Nations report: “meeting the needs of the present without compromising the ability of future generations to meet their own needs”. In other words: we need to be good citizens and ancestors. I like this definition because it brings in more complexity and nuance than simply focusing all of our efforts on climate.¹

Hannah writes that we’ve always been missing one of the parts of sustainability. For most of human history, we didn’t have the luxury of thinking about the future. Survival was the priority. As Hannah says, “The majority of us think there’s a natural order to death: it’s the old not the young that die. But this sequence is a very recent development.” Take a look at child mortality over history:

For much of the last few thousand years, about half the children died before their fifth birthday. Source

We just weren’t living long, so if we could eke out extra time, we did. The high mortality rate suppressed the world population, limiting our impact on the planet.² For a long time, we burned wood to give us energy, then many people moved to fossil fuels, which unlocked higher standards of living than ever in human history. We achieved the first pillar of sustainability for some.

But two issues emerged. First, a large fraction of the world still didn’t have access to this higher standard of living. And second, the massive increase in population and use of fossil fuels together have led to jeopardizing the needs of future generations. If we want to become sustainable, we need to address both issues.

In Hannah’s view, we actually have a chance at becoming the first sustainable generation. This isn’t just an optimistic perspective, but one based on looking at the long-term trends in human history. When you zoom out, you see that while the world is still terrible in many ways, it’s much better than in the past. And this context is crucial for grounding our worldview and bolstering our confidence that we can tackle the challenges facing us today.

In each chapter, Hannah develops our worldview by studying the past, present, and possible futures for seven big problems: air pollution, climate change, deforestation, food, biodiversity, ocean plastics, and overfishing. She tells us what we used to do, what we currently do, and what we could do. Hannah’s ability to look at a problem and not despair, but go into solution-mode is why I really enjoyed reading Not the End of the World. (This underlying spirit is also why I think Vox named her one of their Future Perfect 50 in 2023.)

Instead of going through each problem sector that Hannah presents, I want to bring out some key threads that weave through the entire book and have shaped how I now think about these problems.

Decoupling

“Is economic growth bad for sustainability?”

For a long time, I probably would have agreed. It seems reasonable: we really kicked off our economies on the back of fossil fuels in the late 19th century, and we’ve continued ever since. It would be a small jump to then say economic growth and fossil fuels are coupled, meaning growth would always be a bad thing for our planet.

And yet, I’d be wrong. In my view, Hannah’s most important idea in the book is that of decoupling our economic growth (and living standards) from environmental destruction:

“And it’s true that in a world without technological change, we’ll be stuck with fossil fuel power, petrol cars and inefficient homes. But as the chapters that follow will show, new technologies are allowing us to decouple a good and comfortable life from an environmentally destructive one. This is what makes it possible for us to be the first generation. In rich countries carbon emissions, energy use, deforestation, fertiliser use, overfishing, plastic pollution, air pollution, and water pollution are all falling, while these countries continue to get richer.” (Chapter 1)

It’s also not just a matter of sneaky accounting where one country “stashes” their environmentally destructive activities in other countries (though Hannah does write that this happens). Rather, as countries grow richer, they have more possibilities for how they power their activities. This gives us a chance at reducing our impacts. For Hannah, the big question is, “whether we can decouple these impacts fast enough.”

I want to show you two examples of decoupling in action. The first is deforestation. In Chapter 4, Hannah uses France as a microcosm of the history of humans chopping down trees for construction, energy, and farmland. The following chart captures the whole saga:

The French depleted forests until they improved farming practices and switched to coal for energy. Source

In the year 1000, forests covered about half of France. During the next three centuries, the population doubled (8 million to 16 million). The result is that the people of France needed a lot more timber and energy, so they chopped down about half of the forests (leaving ~25% of the area in France as forests).

You then see a dramatic increase in forests during the next century or so. That’s because of the Black Death, which reduced France’s population to 10 million. Forests grew back as there was less of a need for timber, energy, and farmland.

But this restoration was only temporary. Midway through the fifteenth century, France’s growing population begins depleting the forests again. France needed wood for ships, land to grow crops (with very low yields), and energy for heating and powering industries. As Hannah writes, it would be easy as a person living in 18-century France to conclude that your home would soon be devoid of forests.³

From our vantage point in time though, we can see that this conclusion would be very wrong. Look at the rebound starting in the 1800s! The fraction of forested area in France made a U-turn, and most importantly, France’s population was still growing. What changed? Farmers improved their craft, slowly increasing yields and also switching to more productive crops (France went from rye to potatoes). The result: more food for the same amount of land. France’s government also switched their policies. They went from incentivizing you to clear the land to having policies for deforestation and getting communities to use more productive farmland. And the big one: France swapped wood for coal for much of their energy needs.

This pattern of deforestation occurs in other countries as well:

“These changes meant that rich countries managed to decouple population–and economic–growth from deforestation. This trajectory is a consistent pattern that we still see across the world today. It follows a country’s journey from less to more industrialised. Deforestation and development are tightly linked when a country is poor, but this relationship eventually breaks down. Once countries get rich enough, forests stage their comeback.” (Chapter 4)

You might be thinking, “Switching wood for coal isn’t that great of a trade.” And I agree that going down the path of fossil fuels created new problems. But I think it also afforded many countries the initial economic growth that now gives us the chance to help everyone grow while limiting downsides such as air pollution.

Clean air is the second example of decoupling I want to highlight. Air pollution comes from burning items, such as wood or other biomass. The burning flings small particles in the air, which then get into our lungs and create all sorts of health problems. But it’s a tradeoff people have been making for hundreds of thousands of years:

“Whether or not they were aware of the health impacts of what they were breathing in, the sacrifices of giving up wood-burning were severe. They needed fuels for cooking, heating, light and safety. Perhaps early death from respiratory infections, cardiovascular disease or lung cancer was worth it for a better life. As we’ll see later, this is a trade-off that billions of people still face today. In the end, having a source of energy nearly always wins.” (Chapter 2)

Once people discovered that coal yielded about double the energy compared to wood (per kilogram), they switched to it as a source for energy. You can see this in the following chart, where wood falls under traditional biomass:

During the 19th and 20th century, people moved from totally using traditional biomass to coal and other fossil fuels. Source

Hannah writes about how the cities of Edinburgh and London exemplified the type of air pollution that burning coal creates. Smoke in homes from cooking, smog in the streets, terrible stenches, and particularly bad events such as the Great Smog of London which killed somewhere around 10,000 people made it clear that there were serious health consequences to burning coal. Hannah calls it a “silent killer, but one that seemed like an inevitable price to pay for progress.”

But today, London is so much better. Environmental regulation helped with this, banning leaded petrol and spurring companies to innovate and create technologies such as sulfur scrubbers for smokestacks on power plants and vehicles that emitted much less pollution.

The key isn’t to just ban everything and throttle the economy. Hannah writes that we don’t have to make the tradeoff between environmental action and economic growth. We can have both, and it’s the trend she sees in many countries worldwide:

“The path that countries follow is a predictable one. Pollution first rises as a country starts to move out of poverty. At this stage, access to energy is the priority. It burns coal, oil, gas without tight restrictions on how clean it needs to be. There are no demands for top-of-the-range power plants with anti-pollution controls, or new car engines with particle filters. Pollution levels continue to rise as more people get electricity, cars, and can afford to heat or cool their homes. The country enters an industrial boom. People have more money and life is getting better. The pollution isn’t pleasant, but the trade-off seems worth it.

But, eventually, the country reaches a tipping point in its pathway to prosperity. Once life is comfortable, our concerns turn to the environment around us. Our priorities shift, and we no longer want to tolerate dirty air. Governments have to shift too; they are forced to take action and reduce levels of air pollution. The curve of air pollution reaches its peak and starts to decline.” (Chapter 2)

This phenomenon is called the “environmental Kuznets curve”, and though Hannah says it isn’t true for all environmental metrics (or other metrics), it does apply for air pollution. She makes the point that while it took a long time for some of the early countries to go through this curve (like the UK and the US), others are going much quicker because of technological advances:

Various pollutants tell a similar tale: emissions go up initially, but then decrease due to innovation and regulation. Source

Compare this to a country like China (note the vertical scale is much larger because China has a bigger population):

Same chart as the UK, but for China. Notice how China may have hit its peak already, in a shorter timescale than the UK. Source

Even better, these advances could help us get some countries through the curve without ever having to go through the “hump”. That would be a huge win, because currently about 4.2 million people die each year from outdoor air pollution, while 3.8 million die from indoor air pollution. Hannah compares this to the number of people who die from smoking (~8 million) and road accidents (~1.3 million). But getting across the curve is only possible because of the combination of environmental regulations and innovation as a response.

Depopulation and degrowth are not viable options

I appreciated that Hannah devoted some space to explain why depopulation and degrowth are not viable strategies for sustainability. The big idea is that our collective environmental impact is the sum of all our individual contributions, so we could either reduce the number of contributions (depopulation) or reduce the value of each impact through shrinking the economy (degrowth).

Hannah thinks both approaches are misguided, and I agree with her. When it comes to the population, we’ve already hit peak population growth⁴ (the rate at which our population grows each year). You can see this in the following chart:

The rise and fall of the world population growth rate. Source

The magenta curve shows the growth rate, while the turquoise-filled area shows the world population. If the population for a given year is P and the growth rate is G%, then the population the next year will be P(1+G/100). When the growth rate is positive, that means the population will be growing exponentially. But if you look at the projections from 2023, you see that the population is still growing, though more and more slowly. Once that growth rate goes negative around 2100, the population starts to decrease.

For our environmental impact, this means there won’t be an uncontrollable explosion in population. There are then some who want to go even further and deliberately lower the population (she quotes Paul R. Ehrlich’s roughly 1 billion figure from his book The Population Bomb). Hannah writes that any policy you wanted to implement globally to lower the population, this won’t happen nearly on the timescale we would need to make a dent in our environmental problems. The only way to get our numbers to “1, 2, or 3 billion people would mean killing billions or stopping people from having any children at all”. I don’t want to live in such a world. Do you?

To reign in our collective impact then, we will have to focus on slicing our individual impact. Hannah puts it well:

“The end goal that we’re aiming for is to reduce our impacts per person to zero–or at least very close to zero. If we’re to build a sustainable world for the future then we all have to tread with the lightest of footprints.” (Chapter 1)

If we can do that, population numbers will matter much less.

Hannah then tackles the degrowth mindset, which links economic growth with high environmental impact lifestyles (think: fossil fuels, extracting resources, using land, eating meat, etc.). In addition to decoupling economic growth from sustainability (as we saw in the previous section), she notes that global wealth isn’t even close to enough to put everyone on the $30/day poverty line of rich countries like Denmark. We’d have to make the global economy five times bigger to achieve this, which is what we’d want to do if we’re trying to satisfy the first pillar of sustainability. This fact was very surprising to me, and it made me see economic growth in a different light. As Hannah quips, “A world without any economic growth would remain a very poor one. A world with degrowth would be even worse.”

Overlap

Despite the book being an exploration of seven big problems, what I learned from reading is that they aren’t independent. Instead, they overlap. Making progress in one area often means making progress in several others. This isn’t to downplay the work ahead of us, but it does make our quest for sustainability a little easier.

For example, burning stuff is responsible for our air pollution. Burning less would be good for our lungs. But it would also help our efforts against climate change, since much of that stuff is fossil fuels like coal. Moving to cleaner energy sources would do good on multiple fronts.

This isn’t an isolated instance; it’s the pattern. Hannah has a section about eating meat, which is my favourite example of overlap from the book. Eating meat (beef in particular) is one of the most emissions-heavy foods we can eat. Shifting our diets by eating a bit less beef will reduce the number of cattle we have, which directly lowers methane emissions. But then there are the knock-on effects. Fewer cattle means we don’t have to feed as much of our food to them and can instead feed it to people (41% of the food we produce goes to livestock). We also won’t need as much farmland (currently occupying about half our usable land on Earth), so we can cut down fewer forests, which then limits biodiversity loss.

As usual, Hannah has a good perspective:

“The problems we’re facing are tightly interconnected. The worry is that this gives us impossible trade-offs; we’ll be forced to prioritise one problem at the expense of another. But it isn’t the case; instead, these interdependencies mean we can solve a lot in one go. Move to renewable or nuclear energy to improve air pollution and climate change; eat less beef to improve climate, deforestation, land use, biodiversity and water pollution. Improve crop yields to benefit the climate and humans.” (Conclusion)

Better, not perfect

The book’s subtitle is “How We Can Be the First Generation to Build a Sustainable Planet”, and Hannah delivers. For each problem, Hannah writes “How we got to now”, then follows it up with “Where we are today”, and finishes with paths we can take. This infuses each chapter with a mix of history and hope. Instead of ending each chapter with the feeling that we were screwed, I got the sense that we had a daunting challenge ahead of us that was solvable. As Seth Godin would say, these are problems, not situations.

I spoke to Hannah for an upcoming podcast episode of CarbonSessions, and she made a very good point about these paths: none of them are perfect. No matter what we do, there will always be tradeoffs. But that shouldn’t stop us from taking action and implementing better—yet imperfect—solutions. For example, eating meat is a huge burden on our yearly emissions budget (in addition to all of those other concerns I highlighted). It would be fantastic if all of us who have a rich variety of food would switch to plant-based diets. Yet the reality is that diet is a big part of our identity, and most of us won’t just completely change what we eat. What is much more realistic is finding tasty plant-based meals that could substitute a few of our lunches and dinners each week. And that can make a large difference: I learned that we can actually make huge gains just by shifting to a less beef-intensive diet, without cutting all meat completely. That’s a huge win when we are searching for ideas that whole populations could adopt.

Personally, I am not a vegetarian yet. But I have slowly started shifting my diet, and I’m at the point where roughly 50% of my meals are plant-based. Am I minimizing my emissions from my diet? No, but 50% is not nothing! Speaking with Hannah, it was clear to me that she is much more pragmatic about finding better solutions over perfect solutions.

My one counterargument is that we could get trapped in a local minima where we implement an imperfect solution that now blocks us from moving to an even better solution. The blocks I’m thinking of are more social and political in nature, but I think they could pose issues despite not being about technological or scientific breakthroughs.

Data is useless in isolation

If I tell you that one in ten people don’t get enough calories, what’s your reaction? Hopefully you agree this is a terrible fact of the world. How can we solve this problem? If you only have that statistic, then you might think, “We’re not growing enough food.” You might then focus your efforts on policy favouring more production from farmers. But then you might learn that we actually produce enough food to feed the world twice over (over 5000 calories a day per person, see Figure 1). The problem isn’t supply, but food loss, inequality, and feeding food to animals and machines. With this second piece of information, you’ll probably pursue very different strategies.

What Hannah’s book shows over and over again is that any particular data point is useless on its own. To make sense of any number, you need to understand the constellation of data around it. As Hannah told me in our conversation, you need the right context.

It’s even more relevant when thinking about these global problems. By definition, the numbers often aren’t going to be in the hundreds but in the thousands, millions, and billions. These are hard scales to relate to! The only example I can think of where many of us regularly interact with the lower end of these scales is money. But otherwise, these numbers are on alien scales, making them all seem “big”.

Yet a million and a billion are very different quantities, and they matter when it comes to talking about solutions. Hannah gave me the example of looking at CO2 emissions. If you see a statistic that some process unleashes tens of thousands of tons of CO2 per year, that looks bad. But when you put that number beside our 37.15 billion tons of CO2 emissions in 2022, it suddenly doesn’t seem so big.

If we want to tackle all of these problems, we cannot waste all our energy on solutions that give us small wins. We need to triage, and data helps inform us on where the biggest benefits may be. Data and science will never tell us what to do, but they can tell us where we stand to gain the most.

Throughout the book, Hannah’s message is simple: We aren’t screwed. In fact, we’re in a great position to become the first sustainable generation. Her book is one of hope, but also practicality. It’s not that we are simply going to cross our fingers and wish ourselves across the finish line. There are actions we need to take as people (such as shifting our diets), but also actions we need to take as countries (moving to cleaner energy, prioritizing air quality, protecting biodiversity, and more). Both need to happen, but especially the systematic changes on a national and global scale.

What I love the most about Hannah’s work is that it never shies away from the problems we’re up against, yet it also doesn’t settle for despair. Not the End of the World cuts through a lot of the overwhelm I feel when it comes to the environment. But it did more: the book reminded me that there’s plenty of tasks left to do, and each one will bring us closer to a sustainable world.

Thank you so much to Elle Griffin for feedback on this piece, as well as Hannah Ritchie for our conversation about her book and the context around it.

Endnotes

As she mentioned just last week, “While most journalists have only asked me climate questions about my forthcoming book, that is only one chapter of it… Climate is not the only problem! […] In fact, climate is not even the biggest problem for human wellbeing right now (although that could change). Air pollution kills millions every year.” ↩
Though not entirely. As Hannah writes in her chapter on conservation, with a population of only up to 5 million people (~1/2000 the number now), humans hunted hundreds of large mammals to extinction. There’s a lot we were able to do with small numbers. ↩
My friend Malcolm Cochran has a nice essay about deforestation and automobiles in New England. ↩
I would note though that just because we’ve hit a peak in population growth, we aren’t destined to never go higher again. ↩

Data Has a Friction Problem

Wed, 08 Nov 2023 00:00:00 +0000

As a runner, I could tell you things about my town that few others would know. I could describe the best homes for glorious views of the sunrise. I could tell you about our street lights based on the bubbles of light I see on dark mornings. I could show you the terrible (and sometimes non-existent) shoulders for cyclists, runners, and walkers from having to navigate them. I could even tell you when my town is spreading gravel from the tactile experience of having rocks dig into the soles of my shoes.

If I only drove around my town, I wouldn’t go past most homes, my headlights would mask the darkness of missing streetlights, my winter tires would grip most icy roads, and the gravel would be such a minor sensation that I would barely notice it.

Running gives me a detailed window into my town, but it’s a narrow one. What I know about traffic, for example, only applies to the early morning on my usual route. If I tried to generalize my experience to my region, province, or country, the narrow window would transform into the tiny glazed windows you find in public restrooms, which block out all details.

As Max Roser, the founder of Our World in Data, writes in his essay “The limits of our personal experience and the value of statistics”, “It’s tempting to believe that we can simply rely on personal experience to develop our understanding of the world. But that’s a mistake.” Instead, he suggests that, “To see what the world is like, we need to rely on other means: carefully-collected global statistics.”

But Roser doesn’t stop there. He points out that even making a huge effort to talk with a bunch of people every single day doesn’t make a huge dent in the problem. You will slightly expand the potential experience vista, but there are still so many perspectives that will be invisible to you.

News also informs our worldview, so in theory we could expand our horizon by learning about events from outside our life. But have you ever seen the news!? Roser puts it well:

“The news reports on the unusual things that happen on a particular day, but the things that happen every day never get mentioned. This gives us a biased and incomplete picture of the world; we are inundated with detailed news on terrorism but hardly ever hear of everyday tragedies like the fact that 16,000 children die every single day. … The problem is not so much what the news media covers but what it does not cover. Those left in darkness are often poor and powerless and geographically far from us. What we see in the news is not nearly enough to understand the world we are living in.”

Statistics correct the distortion of our personal experience to give us a clear view of our world. Roser’s point is that we should therefore value statistics just as much as our personal experience to inform our worldview.

Roser’s essay focuses on the value of statistics, which I agree is important. I value statistics a lot, and I think others should too! But Roser only briefly touches on the current challenges in obtaining and understanding statistics as an end user: “These statistics exist, but they are often neither accessible nor understandable. They are buried in spreadsheets, hidden behind paywalls, and presented in academic jargon.” I think these challenges are critical to valuing statistics and worth honing in on.

When I recently wrote about infrastructure, I recounted the 1998 ice storm that incapacitated a huge swath of my province’s electrical grid. While I was writing that piece, I wanted to know: How many homes used electrical heating in Québec at the time?

A simple question, right?

I first read a report by StatCan Plus and learned that roughly 65% of Québec households use electricity as their primary energy source (through baseboard heaters and radiant heating). I then clicked through to the data, figured out how to customize my query to Québec, and then saw this:

Primary heating source for Québec households. Source

We can see that for 2021, the two numbers do indeed add to 65%. But what’s going on with the “heat pump” row? I know they can be electric; I have one! So why isn’t it part of the total? I had no clue¹.

I learned that this data comes from the Households and the Environment Survey. Within it I found the variable “Heating system energy type of dwelling, category”. Seemed promising, so I read the whole text. But it was a waste of my time; it’s not discussing heating systems at all, just distinguishing different types of dwellings. I looked around for a few more minutes before giving up.

I went down a rabbit hole for this simple question, and found a partial answer and more questions. And thankfully I was able to find data at all. Navigating the StatCan website can be tricky at times, but in my mind it’s better than finding which of the 13-ish agencies in the US have the particular data I’m interested in². And it’s way better than coming across a $169 USD/month paywall from Statista.

How often do you think people will put in all this work to locate, access, understand, and interpret the data?

I persisted to find this data for my essay, yet I still wanted to give up! While I agree with Roser that we need to increase the value of statistics, I think we need to massively reduce the effort in obtaining them. Each point of friction is an invitation to abandon the search.

If we don’t, then is it so surprising that people turn to their personal experience? It’s always there, there’s no paywall, it will confirm their beliefs, and it requires little effort. That sounds like a winning combination to me.

In my view, to increase the value of statistics we need to make data easier to work with for the typical person. That suggests two pillars of action: teach data literacy (how to find, interpret, and understand data using statistics) and make our current datasets much more accessible.

The former is a more idealistic suggestion, but I think it’s worth mentioning. Working with data isn’t easy. Apart from statistical skills, you also need to know where to look, what to look for, and how to question data. I personally think this could start in primary and secondary school. Imagine a few classes where students ask questions and learn to answer them by finding and wrestling with data sources. This would teach students more than basic probability, but also how to dive deeper into the uses and limits of data.

The practical action we can take is to make data easier to access. Our World in Data does an excellent job of this; I have mixed feelings about my governments’ websites. Though I do like the articles on StatsCAN Plus, I find the whole site confusing at times to navigate. For my provincial government, they treat French as the only first-class citizen, increasing the barriers for non-French speakers. Just look at the English page for our population:

There’s an awful lot of French for the English version. Source

I appreciate that governments have a ton of data and designing a repository is challenging and perhaps not high on their priority list. That’s where non-profit organizations like Our World in Data could come in. They’ve shown how it’s possible to take data from all over the world and present it in a way that’s accessible and beautiful. Our World in Data focuses on pressing global problems, but I’d love to have the same care and attention to data at smaller scales and other topics. What would it take to have My Province in Data? Relative to overhauling the entire government website, some funding and the right individuals could make just as much of an impact in increasing data accessibility.

My colleague Maxwell Tabarrok suggests an analogy to supply and demand. Roser’s framing is about increasing the demand for statistics, raising its value relative to news media and personal experience. Though I agree demand is important, I’m arguing for the supply side: I find it difficult to imagine people valuing statistics if accessing them is an exercise in frustration. We need to remove the friction of finding, accessing, and interpreting data. Once we do, I think more people will be able to appreciate how statistics is the spell that expands our horizon of personal experience into a more representative worldview.

Thanks to Max Tabarrok, Rob Tracinski, and Tommy Dixon for feedback on drafts of this essay.

Endnotes

I stopped here for that essay (and found a source from my Québec government for 1998), but let’s keep going for fun. ↩
I plan to explore this in a future essay. ↩

Posthaste

Wed, 01 Nov 2023 00:00:00 +0000

Seamen hurry near the harbour, making final preparations for their vessel. They crunch through a blanket of red, orange, and yellow leaves. Their destination: Europe. A cool wind reminds them to keep working, anything to keep warm against the falling temperatures of autumn. It’s October 15th, 1657, and it’s departure day for the ship docked in the St. Lawrence River near Québec City.

A view of Québec City around 1760, a hundred years later. Note the harbour and the small size of the settlement, even then. Created by Hervey Smith. Source

Today’s about more than just this ship, though. It’s the final outbound trip to Europe before the cold and ice arrive, making travel down the mouth of the river impossible. For the winter, the colonists in Québec City will have to be self-reliant.

Before they raise the anchor though, several colonists have approached the sailors, asking if they could carry their correspondence. This is their final chance to send any letters to their loved ones back in Europe.

One of these people is probably Marie Guyart, a founding member of the Ursuline convent here in Québec. As Jane E. Harrison recounts in the introduction of her PhD thesis (where I learned of this story), this is the final time Marie will be able to communicate with her son Claude before the ice closes her off from her homeland. After today’s letter, she will have to wait until the snow melts, the plants bloom, and the river thaws out before voyages begin anew.

The communication cadence between a colonist in Québec (New France) and a resident of Europe would be agonizing by our modern standards of email. “They’re taking forever to reply” wasn’t hyperbole but a fact of life. Remember: a voyage across the ocean took weeks or months, even during the summer. This delay increased the stakes of each letter, turning letter-writing into a sort of art form. There wasn’t any official postal service at the time to deliver mail throughout the colonies and Europe (though there were couriers and some individuals who would). Instead, people would often ask ships to carry their correspondence, perhaps paying them a small fee for the service. Postal organizations would only really begin in the 18th century.

Here in Québec, one big change occurred in the 18th century with the advent of a postal service connected to the burgeoning British colonies. As Harrison explains (page 2), people could continue writing through the winter by routing their letters south to New York, where ships could then leave for Europe. This development was progress in solidifying postal routes and a service throughout the colonies and enabling contact with Europe year-round. It was a slow process towards acceptance for people though, since they often could find others willing to do the job even more quickly. But over time, the postal services became the standard.

The route to New York. From the Canadian Museum of History (Andrée Héroux). Source

Steamships and railways helped propel the postal service into a faster and standardized system that people could rely on instead of entrusting their mail to random travelers (though there was still plenty of that). For example, when Canada ran its first railway line in 1836, delivery from Montréal to New York decreased by 5 hours¹ (a 7% improvement). In my view, while technology was important, it was also important to establish regular shipping and railway routes that people could plan around. Communication delays decreased, increasing the connectivity of people who lived far away.

Still, correspondence is a far-cry from conversation. That’s fine if you’re placing an order. It’s less great if you want to keep in touch with a loved one who is far away. Electricity changed that. Over the last two centuries, humans invented and implemented the telegraph², the telephone, the radio, and finally the internet. Each one required time and effort to establish the infrastructure, but we gained enormous connectivity with people far away from us. These inventions also expanded our range of possibilities. The telegraph accelerated the post. The telephone and radio accelerated delivery of our voice. And the internet opened up the possibility for real-time video calls.

These days, I don’t think of physical space as a barrier to communication. I have friends who I’ve literally never met in the physical world. I’ve talked, laughed, and built relationships with them online, all because of our ability to harness the speed of light for our communication infrastructure. There are downsides to such a culture³, but connecting with friends isn’t one of them.

If we take a broad view of history, we’ve increased our potential for sending messages across vast distances in a timely manner. Not all of the world is connected to the same degree, but most regions are more connected than they were in the past. What once could be a perilous journey across the ocean to deliver a message from a colonist in New France to Europe is now just a few taps and seconds away. This can create the illusion that connectivity is only an implementation issue since the speed of light is effectively instantaneous relative to distances on our planet.

Looking to a (hopefully not-so-distant) future in which we settle⁴ on other planets, how hard could expanding our communication network to encompass them really be?

The answer: very hard. The problem is that space is big. No, I mean really big. Just traveling to the nearest star at light speed would take over four years (and by the end, you would curse the laws of physics for such a sluggish cosmic speed limit). If we want to become a multi-world species, I wonder how the speed of light will fracture our conversations with loved ones during a trip, our calls back home when we move to new worlds, and our cadence of receiving news from other planets.

Many of us are used to hopping on a call with another person, talking in real-time. But we won’t be able to do that if we’re trying to contact someone on another planet or far-away station. It takes around twenty minutes for a light signal to travel from Earth to Mars, over half an hour for Jupiter, and over an hour for Saturn⁵. And we’re still talking about the solar system here!

Distance is proportional to the time delay, but my enjoyment of a conversation is not. Anything up to a few seconds of delay is annoying but manageable. Ten seconds, though? I don’t think I’m going to want to have many of those exchanges. Electricity once bridged distant locations to turn correspondence into conversation, but becoming a multi-planetary civilization might force us back to a slower mode of conversation.

Imagine a young engineer traveling to Mars to begin her career. For the first few days, conversations with her parents remind her of just having a very laggy internet connection. After a week, they’re interrupting each other all the time due to roughly a five-second delay. After two weeks, she and her parents learn to pause for several awkward moments each time one of them speaks. After a month, the engineer begins reading a book while she waits for her parents’ reactions, nearly half a minute late. By the second month, she tells her parents they should simply start exchanging video messages⁶.

Unless we’re communicating with nearby worlds, I suspect we will revert back to the cadence of email and letters, where we don’t expect a reply any time soon. That means no video calls, no online courses with synchronous sessions, and a lot more recorded video messages (or even just text).

I also wonder if it will make our worlds more isolated. After all, if I can only communicate with you through a delay, will I want to put in the effort to do so? Maybe, but I’d probably rather make friends with those on nearby worlds with a bearable time delay. Perhaps sending long-form messages with lengthy time delays will become an art form like it did for the colonists in New France. Or maybe we’ll all become experts at using asynchronous tools like remote workers currently use.

For official communication from organizations such as governments, I’m unsure what will happen. My guess would be that organizations will focus their scope around worlds they can rapidly communicate with. After all, you don’t want to wait on a reply for a time-sensitive issue because the head office of your judicial system is light years away.

The common theme: Our experience of time will probably limit the practical sphere of connectivity.

We’ve done really well at pushing our communication infrastructure to the limits of the laws of physics. Saturating them allowed us to temporarily convert correspondence into conversation. But when it comes to increasing connectivity in an expanding civilization, space always wins.

Note: I took some liberties in painting the scene surrounding October 15, 1657. However, the date and departure of the vessel do match the records. I learned of this fascinating account from Jane E. Harrison’s PhD thesis, “An Intercourse of Letters: Transatlantic correspondence in early Canada, 1640-1812”.

Thank you to Malcolm Cochran, Rob Tracinski, Heike Larson, and Camilo Moreno-Salamanca for feedback on earlier drafts.

Endnotes

My source claims this was a 7% improvement, but I think this may be because a railway doesn’t imply a train. While there appeared to be what I’d call trains (a locomotive), this source suggests that the initial railway used horse-drawn carriages. I suspect that would explain the small 7% improvement. If you know more, please let me know! ↩
There was a transcontinental US postal service called the Pony Express, who did incredible work going across the US to deliver mail, but had the unfortunate timing to begin only a year before the US implemented a telegraph system, rendering the Pony Express obsolete. ↩
I won’t touch on it in this essay, but I think Henry David Thoreau’s words from Walden in 1854 about the telegraph nicely capture the downsides of increased connectivity: “We are eager to tunnel under the Atlantic and bring the Old World some weeks nearer to the New; but perchance the first news that will leak through into the broad, flapping American ear will be that the Princess Adelaide has the whooping cough.” ↩
The vast expanse of space will also severely limit the extent and speed to which we can even set up civilizations on other worlds. Doing so will probably require many scientific breakthroughs. Here, I’m assuming we are able to do it. ↩
I’m using the distances from this source (the orbits of the planets change the distance, but they’re roughly in the right range). Ten million kilometres is about 0.556 light minutes. Mars is about 380 million kilometres (around 21 light minutes) away. Jupiter is about 596 million kilometres (around 33 light minutes) away. And Saturn is about 1.396 billion kilometres (around 77 minutes) away. ↩
I’m using the distances in the flight simulation here (from “Getting There and Back”) and converting from astronomical units (au) to light minutes. The distances are 0.01 au after a week, 0.02 after two weeks, 0.05 au after 31 days, and 0.14 au after 61 days. ↩

Infrastructure is the Concealed Wiring of Our Civilization

Wed, 25 Oct 2023 00:00:00 +0000

Each winter, I set a goal of not slipping on ice during my runs. Each winter, I’ve failed. Bruised hips and scraped knees tell those tales. Darkness is the primary factor. A typical run begins long before there’s even a hint of morning light. My steps are never fully confident, the snow concealing potential danger on the road.

Step here, step there.

Keep my feet light.

Slow down before turning.

Watch out for snow-filled potholes.

It’s part muscle memory, part good intuition, part reflexes, and part guessing game. A small relief: On my most frequent road for dark winter running, the 9 lampposts emit bubbles of illumination. They reassure my steps, if only briefly. But near the end of the cul-de-sac lies a stretch where there are no more bubbles and it’s just me with the darkness.

This year I have hope though: My town installed an extra lamppost, providing one last bubble of light. Perhaps this year will be the year I succeed.

Either way, I soon won’t even notice the lamppost as being “new”. In a few years, I may not even remember that there was a time without its light. But it will make me safer.

That’s the nature of infrastructure: when operational, it’s an invisible supporting actor in our lives. It’s unassuming, quietly working in the background to enable everything else in my life. Imagine if I had to manually turn on my lampposts, or send a request to my town to activate these particular lights at a certain time! It’s absurd, which is why infrastructure functions best when it’s invisible.

In fact, it’s when things go very wrong—a natural disaster, a human accident, or a tipping point in a system—that we sadly remember. Infrastructure failures expose the concealed wiring of our civilization.

Such an event happened in Québec (and eastern Ontario) in 1998. Its scale was so massive that my older family members still reference it. Overnight from January 4th to 5th, freezing rain poured down on the province. This isn’t so unusual. But the freezing rain didn’t stop. Instead, it kept coming. And coming. And coming. It went on for six more days, at which point some areas had received up to 100mm of freezing rain. In comparison, Canada’s previous largest ice storm pelted down only 30-40mm of freezing rain.

To give you a glimpse of the scale of the problem, take a look at this photograph during that time (from Hydro-Québec):

What happens when freezing rain turns trees into ice sculptures. Source

These storms don’t just make roads and sidewalks slippery. Freezing rain clings to surfaces, weighing them down. The rain transformed trees and transmission towers into ice sculptures whose structures suddenly had to bear quite a bit more weight. Trees snapped, their branches snagging on power lines. Over 1000 transmission towers fell, many buckling over their own weight, knocking out the electric grid.

When one transmission tower fell, it brought down multiple. Source

Hydro-Québec, my province’s main electricity provider (along with other organizations), began repairing the broken power lines and electrifying the province again. But this took time. Millions of Québecers had no electricity for days and over a hundred thousand had none several weeks later (p. A3 of the January 28, 1998 issue of Globe and Mail). The winter is a bad time for the electricity to go out when around ⅔ of the households use it for primary heating. Several people died from hypothermia or from carbon monoxide poisoning through burning dangerous fuels indoors (p. 17). As you can see in the chart below from Statistics Canada, the extent of the storm was vast and many regions were out of power for more than a week.

A map of the storm over southern Québec and eastern Ontario. Note the many red areas on the map where customers had no power for over a week. Source

Our electric grid had several flaws which turned an extreme event into a disaster: the transmission lines could only withstand about 45mm of ice (many regions received that as a minimum), one falling transmission tower could unleash a domino effect on many more, and overgrown vegetation meant snapping trees could damage the power lines.

Failing infrastructure can be so devastating because we take it for granted. Like thinking you’re at the bottom of the stairs when you still have one step remaining, we’re left flailing when infrastructure breaks because we completely depend on it.

It’s also easy to forget that we built every single piece of infrastructure we have. It wasn’t always there. The problem is that we’ve inherited a lot of it from the past. It usually blends so well in the background that we treat infrastructure like the laws of physics rather than systems we can build, sculpt, and change. A lot of it was there before we were even born! But there was a time before smartphones and the internet, a time before weather forecasts, a time before highways, a time before traffic lights, a time before air conditioners, a time before cars, a time before electricity, a time before democracy and judicial systems, and even a time before modern waste management and toilets.

Each advancement required ingenuity, human coordination, and effort. They didn’t happen by accident, and the underlying problems were thorny (and still are in some regions). When we only think about infrastructure during times of crises or failures, I fear we forget how vital it is to our lives. Infrastructure is boring and invisible when it’s successful, and that’s kind of the point.

Yet I don’t want us to forget about infrastructure. I want us—those who benefit from so much of it—to learn how it shapes our lives. Doing so will remind us of two lessons. First, we’ve made progress on big problems before. Just consider our multi-century effort to electrify our homes, buildings, and streets. We turned a problem of energy access into the electric grid we have today. That should strengthen our confidence that we can make progress on our current challenges (and for more people). Second, infrastructure creates new problems. If we want to be capable of fixing them, we need to notice infrastructure and understand how it works.

In the case of the 1998 ice storm, Hydro-Québec learned from its mistakes and rebuilt a more robust and resilient electrical grid that will do better if a similar storm appears. Success will mean we don’t notice. But we should. The first step to improving and building new infrastructure is to notice its presence and to appreciate all that it enables (both the good and the bad!).

Infrastructure isn’t just about technology, either. As Paul Edwards writes in his essay “Infrastructure and Modernity: Force, Time, and Social Organization in the History of Sociotechnical Systems”:

Power outages or traffic jams cause most of us to think of downed power lines or inadequate roads, rather than to question our society’s construction around and dependency on them.

If technologies are the “atoms” of infrastructure, then social organization is the glue that holds them together. Infrastructure is social and technological. Edwards points this out right in his title: infrastructure is sociotechnical, and you can’t pry them apart. This suggests that understanding infrastructure requires going further than learning about technological progress. It also requires understanding institutions and social organization.

When millions had no power during the 1998 ice storm, Canada launched Operation Recuperation, deploying over 15 000 members from the Canadian Forces (more than any other instance within the country in history). The troops were crucial to the relief effort, yet they weren’t infrastructure in the form of technology. Rather, they were an institution that Canada could call upon when disaster struck. In this sense, I consider them just as much part of infrastructure as the electric grid that Hydro-Québec rebuilt. They both enabled the storm to recede into history and the back of our minds.

Here’s a list of questions I don’t worry about: Where will I get my food? Do I have clean fuel to cook? Can I store food for later in a refrigerator or freezer? Where can I get clean water? Will my shower serve me hot water? Can I read at night with an electric light? How can I talk to my faraway friends who I haven’t seen in years? Infrastructure is a big part of why I don’t worry about these questions.

Right now, it’s a privilege to not have to worry. But I want to turn this into a basic fact of life for everyone. I think part of the solution is building literacy in infrastructure, so that those like me, who’ve often taken infrastructure for granted, can appreciate just how far we’ve come, then identify the issues with our infrastructure and give us the confidence to go even further. Our world still has many problems left to turn into infrastructure.

A big question I do worry about is: How can we decarbonize our civilization while still growing and creating opportunities for everyone to flourish? It’s an immense challenge, one that can seem paralyzing amidst the flood of dismaying news about our climate. But I think by learning about infrastructure and appreciating its history, we can inspire people to make progress on these current problems, just like we did in the past.

Progress transforms problems into invisible infrastructure, and my hope is that there will be a future generation where this question has faded into the background of their lives.

Note: Parts of Paul N. Edwards’s great essay “Force, Time, and Social Organization in History of Sociotechnical Systems” and an article entitled “Invisible infrastructure is the background to our modern lives” by Sarah Bell, Charlotte Johnson, Kat Austen, Gemma Moore, and Tse-Hui Teh inspired this piece.

Thank you so much Elle Griffin, Tina Marsh Dalton, Rob Tracinski, Anthony D’Apolito III, Becky Isjwara, Harrison Moore, Paige Lambermont, and Anna Mackenzie for feedback on this draft.