Hacker News Reader

Comment by rippercushions | Hacker News Reader

The article makes a major mistake: it uses Tokyo Metropolis (Tokyo-to, basically "Tokyo State", area 2,194.07 km2) for comparisons, not central Tokyo (23-ku, aka the former Tokyo City, area 619 km2). So it's basically the same as saying houses in New York are much larger than those in Paris, because you're comparing all of New York State including housing in Buffalo etc with the 20 arrondissements of central Paris.

> Even more striking is that more people in Tokyo live in detached houses compared to apartments (30%) than in New York (16.3%) and Paris (12.3%).

I'm pretty sure this "striking" fact is an artifact of including a whole lotta suburbs, exurbs and farms in the Tokyo stats. Very few people live in detached houses within the 23-ku.

TL;DR: Housing in greater Tokyo is indeed affordable, but as anyone who's seen a "one-room mansion" (read: tiny studio) can attest, it's rarely spacious.

View on HN

Agreed, this article is basing it's numbers off of this (linked) study: https://s3-eu-west-1.amazonaws.com/airdrive-images/wp-conten...

If you view page 6 and 7, you'll see why these numbers are largely irrelevant since the boundaries for all the "cities" except NYC (which is the 5 boroughs) include the large swathes of suburbs/exurbs around the city.

A study looking at 23-wards vs Inner London vs City of Paris would be a lot more informative.

BTW, the writer of the article seems to simply be using these numbers to push his view that real estate prices will continue to go up in Tokyo. Housing Japan, a Japanese real estate agency that caters to foreign investors shows pricing for apartments in the 23 wards have been on a pretty crazy upwards trajectory [1], but 1) this is denominated in JPY pricing, and the exchange rate has tanked by 35% vs USD this past year, and of course, I'd be much more interested in seeing that post revisited in 2023...

[1] https://housingjapan.com/blog/tokyo-residential-real-estate-...

JPY is poised for an astronomical fall. The government cannot both keep servicing debt and buoying the currency . Those houses will appreciate in value JPY denominated and crater in terms of USD.

So....living in Japan with a USD income we should actually wait longer to pull the trigger on buying real estate? I was already sizing up properties across the country, putting together a plan for my family to buy properties for rental income because they are already cheap...

New York "metro" is always tricky. All of New Jersey, southern Connecticut, Staten Island, the Bronx, outer Brooklyn, outer Queens: Who the hell cares. Yes, I will be down-voted for this comment! My point: Defining inner city NYC that is wealthy requires some care and consideration. All the interesting "real estate action" happens in a very small radius from midtown Manhattan.

> because you're comparing all of New York State including housing in Buffalo etc with the 20 arrondissements of central Paris.

But isn't the entire Tokyo Metropolis under an hour by mass transit? That's a far cry from Buffalo. Like, Buffalo cannot commute to NYC by anything other than private aircraft.

If you replace "Buffalo" with "an hour out of NYC" his point still stands.

Except that housing an hour out of NYC seems to be more expensive than the housing the article talks about.

I don't know - an hour out of Penn Station takes you as far as places like Wyandanch on Long Island with the LIRR, or Stamford, CT with the MNR.

I specifically looked at Stamford when I wrote that. Tokyo is in the 350-600sq ft range by the article. You can currently get a fairly spartan 600 sq ft in Stamford for the price listed (one vacancy). That's more than the average the article cites, but in the same range.

After all, you'd have to compare the outskirts of Stamford with the outsirkts of Tokyo.

Well, yeah, if you take the commuter rail line along the densely populated coastline.

If you take the Harlem Line up to the middle of nowhere, you can get to Katonah in an hour on the express trains during peak hours.

And in the other direction, an hour out of Penn Station can get you about halfway across northern NJ even with extensive train traffic.

The theoretical travel time might be an hour, but we can't realistically compare the travel time variances between MetroNorth and Japanese trains.

Going to Stamford is maybe like 1 hour, with frequent catastrophic delays and problems, such that we can maybe model it as like 1.5hr +/- 20 minutes, whereas an hour on a Japanese train is like 1.01hr +/- 30 seconds.

Wildly different user experiences.

OK sure, but if you want to talk about wildly different user experiences, you also need to consider the likelihood of being physically rammed into the carriage by the platform staff so the doors can close when riding a rush-hour train out of Shinjuku station or whatever.

New Jersey and Long Island are not more expensive than what the article talks about.

Technically NYC commuter rail goes out about 1.5hrs.

Technically Tokyo Metropolis includes some islands several hours by plane south of Tokyo, including Iwo Jima of WW2 fame. Their combined population is only a rounding error compared to the mainland though.

I don't have numbers, but I doubt that 12.3% of people living in Paris proper (the 20 arrondissements) live in detached houses. Yes, there are villas in the city, but they're extremely rare. I don't see them housing more than 1 in 10 people.

So their numbers seem indeed dubious.

Edit: I found official stats (French only): https://www.insee.fr/fr/statistiques/2011101?geo=DEP-75#chif... Table "LOG T2". "maisons" = "house", "appartements" = "apartments"

In 2019, less than 1% of housing stock were "houses".

Aren’t the backstreets of the less central ku filled with houses? I mean Setagaya, Ota, Edogawa, Adachi, Suginami, etc.

I used to pick stations at random to arrive at and explore, and what I always saw was endless houses. Only the areas closer to bigger stations had manshon.

I think you might be conflating Shibuya/Roppongi/Shinjuku with less central parts of the 23 special wards.

even in Shinjuku there are cheap houses if you look. My friend rents an old house with renovated interior in Shimoochiai for what I thought was surprisingly cheap

Those are huge wards by themselves. There are plenty of busy as well as residential places. For example, Setagaya-ku - stations like Sangenjaya, Jiyugaoka or Futakotamagawa are far cry from the quiet residential areas. There's a lot of any kind of housing there - bigger, smaller, houses and mansions alike.

And the apartment as a comparison to houses doesn't help much, as Japanese houses tend to be smaller than western. Source is me, I've been living in Japan for over 10 years now.

Western being what exactly? Houses are rarely spacious in a lot of what qualifies as 'west'. You can't compare The Netherlands and Belgium with Germany, let alone the US.

Sorry I should have specified, I was referring to Aussie homes(my home country) or US. My bad.

But even then I'd probably guess Japanese are smaller. I have not looked at the numbers and will go look now though, I should have before commenting.

Edit: you're right(as I had assumed being familiar), Netherlands average from a quick search is roughly on par.

Okutama, Tokyo is part of "Tokyo Metropolis". https://en.wikipedia.org/wiki/Okutama,_Tokyo . The furthest I've been is Hachioji, which was about 1 hour by train from 23-ku. Hachioji is roughly half as far as Okutama. In New York City terms, getting to Hicksville, NY takes about 1 hour by train. For Bay Area folks, both Dublin / Pleasanton and Bay Point / Pittsburg ... or even Antioch are about an hour away by BART.

Is 1 hour by train a reasonable metric to use given the differences in trains

yes, because we are measuring how satisfied the people will be with their living arrangements not how fast the trains are. As a person I don't care if the train is fast or slow I care how long it takes me to get home - if it takes me 1 hour to get home with a slow train or 1 hour with a fast train it is the same to me - I live 1 hour from work.

Almost. Take reliability of trains into account too. If you are physically closer you have backup options (buses, other trains, bicycle perhaps!). Sydney has such slow transport that if you need to make connections it can be quicker to cycle the same distance. The UK people live not just in suburbs but entirely different cities to commute to London, but using trains that if they go wrong - and they will - it’ll be long and/or expensive commute that day.

> Take reliability of trains into account too.

I think it's fair to say that Japan is the country in the world with the most reliable trains

Switzerland, but Japan is up there indeed!

> Sydney has such slow transport that if you need to make connections it can be quicker to cycle the same distance.

I believe that would come under the time to get home calculation. Aside from that all these other points would seem to be to Japan's benefit.

I'd say yes, an hour to work - while long - is pretty acceptable to most. Many caveats here such as reliability of transport etc but it does come down to time.

I think there’s an adage of urban planning that the median commute is always 30 minutes independent of how much transport infrastructure you build.

An hour commute to work is most definitely not acceptable to most, especially post-pandemic. Median commute in the US is about 26 minutes and many people are now used to working from home some or all of the time. Hour-long commutes are outliers for a relatively small percentage of people in a small number of cities.

Fair enough, my point of view may be skewed by living here for quite awhile

Possibly. But close to an hour commute (or more) is probably the reality most places if you work downtown in a city and don’t live within the city limits. Even within NYC getting to the financial district from Brooklyn or Staten Island is probably hitting close to that.

Average commutes in the US are brought down by urbanites who live close to the office and suburbanites who live a modest drive away from an office park.

The data does not support your statement. For example: https://www.geotab.com/time-to-commute/

Sure, if you cherry pick the worst possible commutes in the worst commuting cities you're going to find pockets of ~1hr commuters, but the broader claim simply isn't true and is perpetuated by people who are themselves trying to justify their overly long commutes. The large majority of people value their time too much to put up with commutes over about 40 minutes.

That data for people commuting into cities seems pretty consistent with what I said. A minority have a less than 30 minute commute. A significant majority aren’t more than an hour. I certainly agree that more than an hour isn’t common though 10-15% isn’t nothing. I did that part-time for about 18 months and it really isn’t sustainable.

Also presumably the number goes up if you exclude people who live within the metro whereas I specifically excluded that group.

1 hour is a fair measure in my opinion, the average commute time is about 30 minutes in the US:

> In 2019, the average one-way commute in the United States increased to a new high of 27.6 minutes [27.6]

Which would mean the reasonable "maximum" is going to be about an hour one-way. Some will commute further but it'll drop off very fast. You can actually see it in action if you graph house prices on a map around an urban area; once you are outside commuting distance from the center prices start going down. People will still commute to the edges of the metro, but the numbers who want to decreases.

And a train can create "pockets" further away by mileage.

[27.6] https://www.census.gov/newsroom/press-releases/2021/one-way-...

Not fair comparison, Okutama is literally in the middle of a mountain range with no much around, but you can get there combining a bunch of express trains in a very straightforward combination, which seems pretty hand-picked. Hicksville to NYC is 56km by car. Okutama from Tokyo Station is 96km.

I actually compared train-to-train distance to Hachioji, not Okutama.

Japan also has a declining population, extremely low immigration, and minimal foreign housing purchases.

These are the opposite factors of what urban US is facing.

Even Tokyo metro area population has been in decline and is forecast to decline, as NYCs continues growing.

Tokyo housing-space-per-resident being higher is probably also an artifact of lower household formation / childrearing than western cities.

That said Tokyo has different zoning which allows for smaller apartment units than NYC/SF/etc and should be applauded for that.

> it uses Tokyo Metropolis (Tokyo-to, basically "Tokyo State", area 2,194.07 km2) for comparisons, not central Tokyo (23-ku, aka the former Tokyo City, area 619 km2). So it's basically the same as saying houses in New York are much larger than those in Paris, because you're comparing all of New York State including housing in Buffalo etc with the 20 arrondissements of central Paris.

As others have said, difference between NYC and NYS is much much larger than that...

But for reference, Ottawa, Canada, is one of the largest cities in Canada by area at 2778 km^2. But the Greater Toronto Area, what many call "Toronto", is ~7000km^2 and only has ~6 million people...

Toronto proper is only 630km^2, with 3 million people. The scale here is reasonable, Tokyo has excellent transportation options that make it realistic to live anywhere in that region.

Your point that we have to be careful about the basis of comparison is well-taken, but your analogy is not sound. New York state is 141,000 km2, it's over 1/3 the size of all of Japan.

Reply on news.ycombinator.com