Takeaway
The predictions of the death of large cities are overblown and I expect most to recover within five years.
NYC is dead, long live NYC
An article recently proclaimed that NYC is dead forever. Because of the internet, the loss of business, culture, dining, commercial real estate, and colleges in the city has turned from a temporary situation to a permanent outcome. As a result, the article concludes that people are leaving NYC and never coming back.
Similar articles have been written about other large cities, such as SF. A couple of people I've talked to have also reflected the same sentiment of wanting to leave the city they're in. I'm not an expert in urban mobility, and this will be more speculative. Let's take a look at some points that matter regarding the death of large cities.
I'm more optimistic that large cities will survive this. As a way of quantifying that statement - I have 80% confidence that NYC will have a higher population in 5 years than it does today [1]. For context, up until recently NYC has been growing at 0.30% yearly.
Picking a city is like picking a house. It's not a purely financial decision for most. If it was, rents would be much more similar across the country, and it'd be easy to choose the optimal city to live in by comparing the median rent for an apartment of your choice.
People have their own unique priorities when evaluating cities. The person's priorities can change, and so can the city's ability to provide that priority. Once cities no longer give you what you need, you move (if you can).
Usually, the pace at which the person's priority changes has been faster than the pace at which a city's ability changes. You move to a city for your first job after graduating college, and then stay there for a while until you need more space for a family. The city stayed the same, you changed.
Recently though, cities have lost nearly everything that attracted people to them. You may have stayed the same and want the same things, but the city changed. If you recall our pace layer discussion, in this case the "governance", "infrastructure", and "commerce" layers all moved more quickly than expected.
This volatility has led to a change in mindset from people. What used to be a reluctant consideration of remote work locations has turned into an enthusiastic evaluation of how much further your rent money goes in Austin, Texas. For many people, what used to be unthinkable is now the logical choice.
It's important to distinguish these two types of situations though. In the former, your priorities changed, and you'd likely not move back to a city again after moving out. In the latter, your priorities stayed the same, but all the cool spots are closed for now. When they reopen, you'll be back.
It's like going to a restaurant. If you no longer like that style of food, you're not returning. If your favourite dish is sold out all the time, you'd visit less often but still be interested if it came back on the menu.
City factors
Let's look at some factors that make cities attractive, and see what has happened. I'll be drawing from this 1970 paper on urban mobility but mostly adding my own commentary [2]. The main theme is that people's base desires don't change that much.
A few of the top factors:
Work
Education
Social
Services
Vibe/culture
Work
Many people move for the sake of a job; I did too. If that job can now be done anywhere, there's less reason to move. Why would companies want to move back to offices?
Cynically, having an office allows a company to extract more time from their employees. And while some people are more productive at work now that they're working from home, I feel that the productivity of an average worker has gone down since leaving an office. Getting people back eventually is one way to grow faster.
There's also a sense that those random conversations in between meetings are actually important as well. This has led to more companies finding ways to have their teams interact outside of work, such as a socially-distanced team outing. If people were satisfied with just zoom happy hours, we wouldn't see such activities suggested. People want to see others in person.
Remote work also doesn't have to be an all or nothing proposition. If companies ended up letting people do some days in the office and some days out of it, then large cities still matter. Remote work might even directly result in more people moving to cities, should offices do a rotational schedule. e.g. one month group A in the office, one month group B in the office. That implies the city can take twice as many workers as it used to.
The type of work also matters. More physical work can't be done remotely, so those jobs stay in the city where you get economies of scale. We'll get back to this under services, but shouldn't this imply more stickiness to cities?
And one important factor, sometimes missed, is how most of these companies allowing fully remote options have clearly stated they intend to adjust salaries according to the cost of living of the city you relocate to. This is without considering the decreased perks you have working remotely - snacks, meals, other services. Some people will be ok with this, and others will balk.
I do think we'll see many more remote-first or remote-only companies. Zapier, for example, has been remote-only for a while now. They still hold in person company events though. For now, I still think remote work is temporary for most. And the styles of remote work that will stick are still beneficial for cities.
Education
Many people move for school as well, most commonly college. As they've gone remote, there's no reason to be moving out of your parent's place now.
Whether higher education survives covid is a separate discussion; my initial thought is that most will be fine but slowly bleed to death years from now. Assuming they do survive for the next 5 years, it's more likely for them to continue attempting to re-open so that they can resume all the activities that are important for the budget, - college sports being a big one.
If colleges remain remote forever (unlikely), the places that are dependent on colleges will slowly disappear. These places are much more likely to be suburban or rural areas though, by definition. In this case, it seems like any change in education is either bad for both small and large cities, or bad for small cities and much less bad for large ones
Social
Many people also move because of their expectations of the social environment of a city. They want to be around people of the same age, meeting new friends, and have plenty of interaction with others.
There will always be people who age out of this demographic. Once you have a family and realise your social priorities have changed, having a 10 min commute to the bar is less important than having a 10 min commute to your kid's preschool.
That aging out process has been pushed ahead, so a large group of people that thought they were going to make this decision in a few years, are making it now instead. We should see a growing number of people continue to leave large cities. And that's where most people are focusing now, the gross outflow from a city. Probably helps that most of the people writing are around this age group.
What about the group of people younger than them? They're still far away from having their priorities changed. If it works out, we could have a new group of younger people move in to the city at lower rents, causing a resurgence of activity rather than a decline.
The fact that people are holding beach parties amidst covid makes me skeptical of any prediction that assumes people don't want to hang out with other people in person.
Services
I mentioned earlier that physical services are still going to be concentrated. If you're a food distributor, would you rather sell into a population of 1k vs a population of 100k?
Cities get to the size they are by bundling many services together in a convenient package, where the price is mostly the price of rent in that area. People want the convenience of getting anything they want as quickly and easily as possible. Having a grocery store close to the clothing store makes sense in a city, just as how having them together at a strip mall in a rural state makes sense.
As long as we remain attached to physical experiences, needing to get items delivered or people transported, location still matters.
Vibe/culture
It's hard to classify cities based on their vibe, or culture. If you've visited other cities before though, you know that some cities just "feel" differently than others do.
Culture is build via many things, including history, people, stories. It's built from the many small interactions you have with others in the city. In a work from home situation, it doesn't seem obvious to me that all of that will change for a city. Of course, that changes if all the people that make up the culture move away. As mentioned in the other sections, this both doesn't seem to be the case.
Cities have lasted a long time
As I summarised in the Josh Wolfe article, there'll be continued tension from city folks seeing suburban folks enjoy a more normal life, seemingly without consequences. Many will even make the choice to move, finally deciding the tradeoff is worth it.
And that's the crux of the issue - living in a city is always going to be a tradeoff, made consciously or subconsciously by everyone. There's no right or wrong answer, just the one more appropriate for that particular person's situation.
The trend of urbanisation is not a decade long one, nor even a century long one. It's something that has continued happening for 5,000 years. And while the largest city in the world has changed over time, what's also interesting is how so many cities have remained large throughout the years.
Footnotes
I'd compare Aug 2025 vs Aug 2020 if possible, but more likely we'd have to compare Dec 2019 vs Dec 2024 or Dec 2020 vs Dec 2025. That growth rate for NY is lower than I thought at first
I skimmed the paper and didn't fully read it. It's cited by many, and provides a framework for why people move. There's a discussion of what leads people to move, and then how they find and evaluate a new home.