How Japan saved Tokyo's rail network from collapse (part 1, 1945-1982)
seungylee14.substack.com
seungylee14.substack.com
One extremely important detail for me when I hear about JNR (Now JR East/West etc) and privatization: during privatization, all the Shinkansen debt in particular was held onto by the public. Extremely expensive infrastructure (now a money printer, at least the Tokyo <-> Osaka segment), just handed to private investors.
Of course JR works well as a whole now and it's not like society is worse off, but it's always left a bit of a poor taste in my mouth.
(For people wanting some more interesting transportation content: here[0] is an excellent overview of IC cards and integration of all the systems in the 2000s)
Wouldn't it be the case that society is now better off because of it? The "poor taste in your mouth" is a mental block from extensive propaganda in the last few years especially against private ownership.
It should be pointed out that privatization of JNR involved a lot of stuff! Regional breakups, the debt load, there's so much stuff going on there that assuming that privatization happened in an orthodox way is much more succumbing to propoganda than my statement.
The coupling of so many things to a single event titled "privatization" means it is extremely hard to untangle everything.
In any case I think "infrastructure-based train company suddenly doesn't have to pay its 65 billion dollar debts, and is able to function better after that" is not surprising, and I do not believe you need a bunch of bankers to figure out a business plan for that. There is an alternate universe where the debt is still taken on by the public, but JNR can continue to be a public good.
I don't disagree. And I'd throw in the dissolution of the JNR labor unions that were preventing mass layoffs (some half of JNR's employees were laid off in the process of privatization) also heavily helped things.
In Canada CN Rail was made debt fee flogged for 2B on the stock market. After 30 years it's now worth much more, so it's been great for shareholders. But a lot of branch lines were cut, freight prices increased as much as possible, and there has been zero movement on passenger rail, both urban and Intercity (well, except Toronto bought a bunch of rails in their city, a tiny fraction of the CN Network for a significant fraction of the privatization gains..)
Privatizing monopolistic infrastructure doesn't usually work well.
There's the option of keeping the natural monopoly part (the rail network itself) in (semi-)public hands and then allowing private train operators to pay for time slots to run their trains on.
The counter-argument is that this creates a very complex interface between public and private, with lots of opportunity for grifting via misaligned incentives. And the experience in the UK doesn't seem that fantastic, with the most recent development being the state taking over much more responsibilities like designing timetables and ticketing, with the private train operators essentially reduced to providing the trains themselves and the train personnel.
Of course the EU is trying to recreate the UK model EU-wide, we'll see how that goes..
There's also the fact that trying to establish any serious amount of competition within a specific route pretty soon runs into capacity limitations, especially since open access operators seem to focus mostly on faster long-distance trains, meaning that on mixed-traffic railways they'll conflict with both freight and regional/local trains. To maximise the capacity of a route, you need to coordinate the schedules of the various trains, which is of course an anathema to the dogma of free, unrestrained competition.
Another thing is that since its impossible to offer direct connections from anywhere to anywhere, especially at anything like a reasonable frequency, the railway system must rely on connections. Attractive connections with not too long waits are only possible between a very limited number of trains. Something like the integrated clockface schedule in Switzerland works only because there's one set of long distance trains that need to be taken care of and integrated with the schedules of regional trains and even other non-railway public transport.
> Of course the EU is trying to recreate the UK model EU-wide, we'll see how that goes..
At least the UK kept a common ticketing system with through-ticketing still available from anywhere to anywhere, plus – albeit with some teething trouble – somewhen in the last decade or so they finally settled on accepting that split tickets purchased independently by a passenger in order to save some money should get the same protection in terms of passenger rights (most importantly: if you miss a connection through no fault of your own, the railway needs to accommodate you, i.e. the ticket remains valid on later trains, you get a taxi/hotel/… if it was the last train of the day, alternatively you have the right to abort the journey and claim a refund, etc. etc.) as through-tickets bought in one transaction.
In the recent EU passenger rights reform, the parliament did try mandating something similar re more through-ticketing, but a number of member states were apparently heavily opposed to this (unfortunately including Germany it seems), so sadly it was dropped…
Only long distance trains are mandated to be self-financing. Competing with local trains means competing with subsidized trains, which doesn't make any economic sense -- if you have a private operating company and want to run local trains, then compete in the bidding processes to get the subsidy to the run them. And they do -- in Germany, about 1/3 of local trains are privately operated. (Btw, for freight its 40%, for intercity trains its below 1%)
Hmm yes, you're right of course. I guess my main beef is how parts of the political word are always saying "Open access competition is great and everything will be super with rainbows and unicorns" and mostly ignoring the fact that most corridors where open-access might be sort of viable are also already mostly full and any serious increase in the number of open access operators would wreak havoc on the schedules of the currently existing services, especially those of regional trains.
When an operator goes kaputt there's temporary pain. But now the models are moving even so far that the vehicles aren't even owned by the operators, but the local regional authorities -- so then the operators have little room for making profit, innovating or doing any sort of interesting shananigans. But it does force a lean operation, keeping costs in control.
Indeed, which is precisely what I tried to explain in my parent comment.
> Europe opened up rail services for competition in the 90s, although it took a long time for it to be implemented.
For better or worse, in many countries it's still implemented on paper only. The incumbent state railways have fought long and in many cases successful rearguard actions to delay competition for as long as possible.
> There are different models being used throughout the EU, the UK's franchise-model isn't really used very much. One common model is that a regional public transit agency contracts out operations of different parts of regional rail networks to private (or possibly public) bidders, and it´s pretty transparent to the users -- since the regional transit authority will ensure common branding and ticketing.
Yeah, this is how many municipal or regional authorities tender out bus lines. I think it's also the model the UK railways are moving towards with the introduction of the (cringe) Great British Railways and the end of the franchising system (hastened by the covid crisis):
https://www.theguardian.com/business/2021/may/19/uk-rail-ove...
> It´s not a very "dangerous" model, since the public maintains control over most aspects of the rail network that matter, it doesn't really privatize gains and socialize losses, it just sort of forces a race to the bottom on the profit margins of rail operators.
OTOH, since essentially everything except the running of the actual trains is controlled by the state, there is precious little space for that purported private sector inventiveness and dynamism to make any impact. So what's the benefit instead of just the state directly owning everything and hiring people to run in?
Union busting?
Answering for my own country, the state railways which runs all passenger rail [1] (except for the capital city regional trains which are run via a tendering process similar to how bus services are run), and the state railways seem supremely uninterested in improving regional rail around other cities than the capital, despite interest from other regional transit authorities. With some "real" competition on the rails, other regional authorities could set up tendering for regional rail as well.
[1] Which comes back to the comment I made above about "on paper only". Yes, in principle other operators can operate on the network, but the state company just got a renewed decade long contract from the government to provide essentially all passenger rail service in the country - who's gonna compete with a heavily subsidized operation like that?
Just read recently that Moscow is now in the process of building a couple of dozens metro stations outside of its ~60-miles long ring-road, to add to the existing ~230 stations. I wish someone would also write a more in-depth write-up of that and how do the Russians manage to do it (including how do they manage to keep the whole network running at what I assume to be reasonable costs), out in the West the knowledge of how to maintain such a system at low-ish costs and how to extend it seems to have been lost.
https://www2.deloitte.com/content/dam/insights/us/articles/4...
The 78% share of public transport dwarves figures from any western city. Car utilization is also fairly low.
I think it has something to do with scale. The city is enormous, so cycling and walking (especially in winter) is just not viable and the statistics reflect that. High ridership means more revenue, so hopefully a high percentage of costs covered by fares.
I don't know about Moscow, but I grew up in Warsaw where the subway rolling stock is still in significant part the Metrovagonmash Series 81: https://en.wikipedia.org/wiki/Metrovagonmash_81-717/81-714 - which serves in Moscow's lines as well. The train is fairly noisy and the design is dated, but it didn't change significantly throughout the years, so I guess a lot of money was saved on R&D.
Moscow metropolitan area represents 1/5th of Russia's GDP and 1/7th of countries population. A new station in Moscow potentially benefits a higher proportion of Russians then, say, a station in NYC for Americans. Moscow is also one of the most congested cities in the world (n.1 in 2020 https://www.tomtom.com/en_gb/traffic-index/ranking/). Expanding public transport is probably the only way the city can sustain itself.
Also, metro is extremely popular among the population in general. Building new stations and lines is a crowd pleaser and voters winner. AFAIK lack of widespread public support is the reason many public transportation projects fail in the US.
Before that, Moscow metro had half dozen new stations per decade, now it's half dozen/yr.
Saint Petersburg is still on the former schedule. But the takeway here, there's China who build a few new systems per year, go ask them and maybe employ some of their tech.
Yes, public assumed JNR debt and at the end it is now being paid by tax, estimated to be $240B, a staggering amount. But that debt is from unprofitable lines (they never could be profitable) sold by politicians to residents, not from money-printing Shinkansen. In a sense, it is right and proper that it is paid by tax.
I don't believe that the point of train lines is to be profitable (just like it's not the point of parks to be profitable, or roads etc), so "wow look at the tax burden" is ... fine for me. Public good, etc. The part I find objectionable is that there were a bunch of assets owned by the public that were transferred to private ownership!
Maybe stuff needs to be broken up, maybe you do need to pay people a lot of money to come work on these hard problems, and privatization is one way of tackling this. But permanent ownership transfer of goods instead of, I don't know, just getting over it and paying public servants good money to do a good job, does leave a bad taste in my mouth.
EDIT: I do think you're right about the contents of the debt though, at least that PDF seems to imply the debt is not from Shinkansen construction.
I think you are moving the goalpost. Yes, privatization itself can be distasteful. But Shinkansen was privatized, not handed out. JR paid for the privilege, and it was fair price (at least it was market price). With the benefit of hindsight, it can be argued it should have been more, but JR took real risk. Money printing wasn't the foregone conclusion.
I agree JNR could have been normalized without privatization, and probably should have. On the other hand, I don't think it was about paying public servants. Normalization of JR was mostly about making obvious but unpopular decisions, and that's in fact one thing the private does better than the democratically accountable public.
Hard to argue that the overall results are bad.
(I do think that "paying a fair price" is not a great argument when you are the central government and can print your currency, but Japan is not the US or Russia and it was the 1960s...)
Yep. A legendary one in my old stomping grounds of Gifu is that there is a shinkansen station in Gifu-Hashima, and not in places that would be far more convenient to have a station (e.g. Ogaki or Gifu City), because a politician local to there was able to decisively dictate the routing and, prior to the route being publicly known, acquired some real estate which was presently undistinguished but would have been extremely valuable if it were right next to a shinkansen station.
(Interested individuals can Google the names if they want them.)
A happier one which I use a lot in explaining why I love Japan: JR refused to close a particular station over one passenger. They ran one train there to pick her up and another train there to send her back every day for +/- a decade, until she graduated, because someone at JR thought that JR must convey a child to their school.
(There are a few more anecdotes about that theme I am aware of, some reported to the press and some of which I know of through other ways, and in this I am in utter agreement with my rail buff friends that the railways see themselves as servants of the public trust in addition to capitalists.)
https://spikejapan.wordpress.com/2012/05/20/minispike-the-en...
It's there to provide an essential service. Which you have to pay for.
Not only is it not a business - and not there to make money for shareholders - running it as if it were a business is ruinously expensive and distracting and hampers the service it's supposed to provide.
The rail efficiency sold to the public is based on projections that (generally) also made them profitable.
Rail efficiency decreases drastically as usage declines, as so much of the embedded energy and emissions is in the fixed cost of building out ROW and track, and ongoing MOW.
You need to operate at a high percentage of capacity in passengers per train, and high percentage of capacity in trains per track.
Efficient rail systems are relatively a highly constrained problem, due to said fixed cost of track, but also due to lack of flexibility to adapt to changes, both in routes (need to move/add track) and capacity (track at full capacity doesn't meet demand, but two tracks exceeds demand at cost of being inefficient). (Operating one track that doesn't fully meet demand increases prices and becomes an expensive but profitable point of stability.)
Which is why eg high-speed rail proposals in the US typically rely primarily on improving existing slow routes, and are limited to midrange distances. Too short and too long both push it to being more costly at reduced efficiency.
"...it was targeted for the axe at the privatization of Japanese National Railways in 1987, just 15 years after completion, but escaped because the road that runs more or less parallel with it and is the only route into Iwaizumi from the south is not wide enough in many places for two cars to pass and is treacherous in winter."
Also, per that article, the number of passengers is quite higher than that, and "per kilometer". Not that is huge, but...
"...the least trafficked line in the whole nation, with average daily ridership per kilometer of 46 people in fiscal 2009"
Anyway, I'm with @rob74 on this, profit shouldn't be the only concern in public transport.
In said case, they might have calculated the cost of upgrading the road through this mountains versus maintaining the train. I guess..
when you think about funding structures, there's a myriad of approaches based on risk and potential outcome.
You can try to estimate ridership, but usually only in the near future as there are major macro affects at play.
every party has a different stake. transit infrastructure can lower housing costs by connecting low cost exurbs to the urban region, which in the us makes locals less interested (probably not so for japan). it can also grow national tax base by providing a competitive central business district and attract international investment, which is something that should appeal to the national government. so it is not as simple as "if locals want it, locals should pay for it", ideally there should be aligned incentives across the board
like any business, investment in transit requires leadership and vision. blindly funding every station proposed is clearly a bad idea. so too is an overly conservative investment regime that takes 0 risk
Overall I agree that it depends on the particular situation, but if it costs several millions for 50 ish travellers from a specific station (an example from in this thread somewhere) that doesn't sound like money well spent.
As far as overdoing it on stations, you’re not entirely wrong but for this reason the Japanese train system has tiers of express service. I’ve heard sometimes it’s worth taking the local train 1 stop so you can then hop on the express train to the city.
Finally, it’s worth noting there are user stories besides commuting to/from downtown business district. Hub and spoke systems like Boston are good for the commute story but fail miserably at displacing the need for car ownership
Almost like in a circulatory system - cutting of the tiny vessels isn’t a recipe for success even if the main system can function for a while without them.
" Modern Monetary Theory (MMT) is a heterodox macroeconomic framework that says monetarily sovereign countries like the U.S., U.K., Japan, and Canada, which spend, tax, and borrow in a fiat currency that they fully control, are not operationally constrained by revenues when it comes to federal government spending.
Put simply, such governments do not rely on taxes or borrowing for spending since they can print as much as they need and are the monopoly issuers of the currency. Since their budgets aren’t like a regular household’s, their policies should not be shaped by fears of rising national debt.
MMT challenges conventional beliefs about how the government interacts with the economy, the nature of money, the use of taxes, and the significance of budget deficits. These beliefs, critics say, are a hangover from the gold standard era and are no longer accurate, useful, or necessary.
MMT is used in policy debates to argue for such progressive legislation as universal healthcare and other public programs for which governments claim to not have enough money to fund. "
By Deborah D'Souza, https://www.investopedia.com/modern-monetary-theory-mmt-4588...
Currently richer areas already subsidize plenty of things like roads, airplane routes and many other services. It’s figuring out below what number of users it is when it is no longer appropriate
In fact, I think there are many things like that in Japan (mailing things comes to mind), where the cost is so low it’s reasonable for anyone. Or where I can go to an out of the way town serviced by one gasoline powered train a day.
You can say that it’s all loss making, and you would be right. Japans public debt increases every year, seemingly without limit. But this has zero effect on people’s lives, whereas closing the only train line that services their home…
I like the "we want to be good train people" argument, of course. It reinforces some faith in humanity. But it would be funny if there was some weird subsidy + very old agreement that everyone is too embarrassed to talk about.
The story in question:
* https://www.mentalfloss.com/article/73525/train-station-kept...
The veracity has been questioned:
* https://www.straitstimes.com/asia/east-asia/remote-hokkaido-...
Railways are OLD. Cities got built around them that's how old they are. And our esteemed ancestors didn't plan 100 years ahead to a future where the population had trebled. You think you can double the tracks in Amsterdam? You'd have to either start tearing down 10s of thousands of buildings or dig tunnels. Very long tunnels.
It was that they could not satisfy the need without it operating really reliably and punctually -- it was simply necessary to have.
[1] https://www.youtube.com/watch?v=Se1XtS8LAKs -- Japanology, a pretty entertaining series in general, if you're interested in the quirks and daily life aspects of Japanese society. And one thing I like about this series is that they usually end each episode with some touching personal story of how someone relates to the overall theme of the episode (oddly enough, because I generally never like the attempt to create a “people story” on the news or documentaries).
Of course, you can't transport dozens of millions of people everyday if you start being unreliable. The whole system would collapse if a train is 10 minutes late because it would block dozens of trains behind it.
Short-term collapses do happen fairly regularly. For fifteen years, until COVID, I commuted every day between Yokohama and Tokyo. In the morning, I went from Yokohama to Shibuya on the Toyoko Line (a non-JR private line); most evenings, I returned home through Shinagawa on JR. I remember many mornings when the trains to Shibuya were delayed by ten or twenty minutes or longer; like other lines into Tokyo, the trains ran at full capacity and at maximum closeness to each other during morning rush hour, and the slightest problem could cause a backup. (Evening rush hour is more spread out, and I experienced fewer delays then.) Anyone who rides Tokyo trains regularly has similar stories.
But it is true that, overall, Japanese trains, including the trains in the Tokyo area, are reliable and even pleasant to ride. In the ongoing discussion of working at home vs. office [1], many people have named commuting as a negative aspect of working in an office. I actually didn’t mind my commute—the walk to and from the stations was good exercise, the bustle of stations like Shibuya and Shinagawa was stimulating, and the train rides gave me time to think, read, and listen to music. While I am now happy to be working at home, I hope the pandemic quiets down enough for me to start feeling confident about going places by train again, both in the Tokyo area and throughout Japan.
Because most of them are driving.
For example, you want to go west you'd go to Shinjuku (station on the Yamanote line) and take the Chuo. North? Either Ikebukuro or Ueno, on the north-west/north-east of the Yamanote line. Shibuya (south-west), Shinagawa (south-east) and Tokyo (east) stations are also within this. And those are the biggest stations in Tokyo, all connected through a circle line (in both directions) with trains that are ~2 mins from each other and completely full on peak times.
In certain combinations it might be faster to go through "smaller" lines crossing Tokyo Station, like from Shibuya to Tokyo I'd get the Ginza Line because it goes straight. But this circle + rays of lines that happens thanks of Yamanote IMHO would have collapsed if there was instead only a single central station in the middle managing these movements between the major lines from outside Tokyo.
Although the Yamanote line is a circle, it's very much an off-centre one that swings out through the suburbs on one side and not the other. And it's worth noting that Japan is pretty far behind in terms of integrated ticketing for occasional journeys (although commuter passes are integrated between different companies).
Firstly, Japan actually has very good integrated ticketing, namely you don't buy tickets at all. You wave your phone/card over the stall on entry and wave it again at whatever station you exit from, and the system calculates the fare at exit. (And it's discounted from the price of actually buying tickets.) Japanese train station gates allow more people to move in and out of the station than any other model of station I have seen in any country.
Secondly, the over-focus on integrated ticketing misses that the important thing is integrated payment. It wasn't the case until some number of years ago, but nowadays everything but the most rural of rural lines has a single shared national system for IC card payment that is easy to use. The IC cards can even be used at most grocery stores and restaurants. They even work for foreigners and are a great alternative to cash to avoid having to calculate payments.
You missed having to go through interchange gates or, more commonly, exit and re-enter when you're changing lines. The fact that ticket payment is easy mitigates the damage somewhat, but the fact remains that tickets aren't integrated and this has genuine costs; people often end up travelling a longer way around so that they can stay within a single operator's network, because the fare structure incentivises them to do that.
That’s the trap, it’s extremely easy to change networks.
> the price differences being so marginal it didn't even matter. I don't care that I'm spending 30 yen (roughly $0.30) extra for a more direct route.
That was not my experience at all. The price difference was routinely 100% or more, especially for cross-city trips where the “direct” route would have multiple network swaps, whereas a slightly less direct route (adding maybe 1/10th the trip length or less) would have 0~1 network changes.
In fact, just looking at the fare structures you can’t be correct: they start at 170~180Y, and reach double the initial fare at something like 40km. I would expect JR to be similar. The entry fare is an outsized component of the total, and will be the vast majority of it unless you travel very long distances. Even more so if you network switch as the “distance counter” resets on every network change.
So if you make a single trip with one network change you’ll be paying 350Y minimum, versus under 200 if you remain in-network and the distance traveled is below 10km or so.
But that's as fast as walking through them so it's not an issue. Thus it's "integrated" even though it's not fully integrated in the normal sense. In other words the lack of "integration" disappears.
A good example on-topic is that I've heard that in London you can pay with your credit card straight away on the underground tickets, tap it and enter and then tap it to exit (and you get charged), no need to even buy tickets.
Note: I'm comparing this to Europe, I've heard that Japan is amazing compared to the US so I guess CC and payment systems in EU should be even more amazing for the US readers over here?
Details here: https://pedestrianobservations.com/2017/05/13/the-yamanote-l...
TRA has also been poorly managed and doesn’t have a great safety record compared to other developed nations. I worry that as more and more people want to demonstrate their new wealth by driving, ridership on TRA will fall, and without integration with HSR it can’t even serve as a feeder network and we’ll be stuck with a high-speed rail too far outside the cities to be useful without driving and parking at the station.
I think "car culture" is usually a good explanation for big picture decisions for transportation -- where society invests, how the government develops transportation-related policies, what people will protest locally against -- but a poor explanation for why people make day to day transportation decisions.
Pretty much anytime you see people hardly using public transit (or biking or walking) in an area, you can zoom in and discover that, shocker of shockers, it really sucks in one or more ways. People are mostly pretty rational when it comes to those day-to-day decisions, I think. Yes, it may seem silly that people will drive their huge truck to their office job a mere mile and a half away, but if you zoom in, you find that even their residential neighborhood has huge, wide streets that people speed around in, making biking feel uncomfortable and dangerous. And when everyone else has huge trucks going fast, being in a huge truck does make you feel quite a bit safer.
Another example, my hometown of the South SF bay area has excellent weather for biking, and the more populated areas are largely flat*, but biking there is still uncommon, and if you try to do so yourself you quickly discover why.
It's largely unpleasant and very dangerous-feeling, with fragmented, quarter-assed infrastructure that often seems designed to try and kill you. Who the hell feels comfortable biking on painted bike lanes on Lawrence Expressway next to 50+ mph traffic, I'll never know. And what little infrastructure there is often randomly disappears, like bike lanes that just sort of curve into the sidewalk (and not even a sloped edge, so I'm not sure what they expect you to do there). I personally got hit twice by cars the last year before I moved to Munich, including once when my son was on my bike (police came out, but they didn't even bother ticketing the driver who t-boned by rear wheel, of course).
* Yes, obviously there are hills around the valley, but most of the people and stuff is in the flat parts
That said, there are plenty of times that friends will want to drive, for the sake of driving around, to places where transit would do just fine. To me, it looks like the current generation is far wealthier than the last one, and being able to own and drive a car is a sign of having made it. Trains are awesome to me, having grown up in a place without public transit, but some of my friends act like riding TRA is more like a poverty decision. I may be reasoning from a too small a sample set of just my friends, but the voluntary gridlock weekends to go ‘camping’ drive me nuts.
About bicycles: In Japan hitting something with your car is an offense in itself. So, even if a cyclist or pedestrian is breaking traffic rules, if you hit them you will be liable to pay between 10-100% of their medical costs+damages.
Politics for a polity with less dysfunction (or at least different dysfunction) than ours.
If you live anywhere in the Tokyo metro area owning a car is silly. I had a friend they paid near $800 a month to park because he liked to drive outside of Tokyo once a month. I looked at him and said "you can rent cars in Japan, correct?"
The "Commuting Five Directions Operation" (通勤五方面作戦, tsukin-gohomen-sakusen). That last character "sen" means battle. That word means more like "wartime/military operation" which reflects the urgency.
...maybe you mean "cruise"?
Now, in the case of "crusades", the word has very negative connotation in Arab/Muslims countries. So, when George Bush said the war on terror was a "crusade", the vast majority of political commentators in the Arab world took this literally as religious war between the Christian and Islamic civilizations.
On top of the challenges they faced, I guess JNR had enough veterans with memories from previous careers and it reflected that.
I'll use this moment to remind people that Japanese is not Chinese and bears little relation to it despite borrowing parts of it's writing system and some vocabulary. (China also reverse-borrowed Japanese invented vocabulary.)
Each of those characters can be broken down further into their radicals. And, those radicals tell you a lot about the meaning. When I lived in Japan I could identify anything botanical by the tree radical even if I didn't know the character.
Every single native Japanese speaker will look at that character and see that "word" contains the word for war (it's a beautiful and striking character).
Every single word in every single language has a long history with dozens of influences.
Japanese is not Chinese. But, the influences of that language are all over Japanese. Almost every character has a Japanese and at least one or more Chinese pronunciations.
It is absolutely not correct to say "it is a single word and cannot be broken down any further."
Source: I'm not a native speaker, but I've got a degree in Japanese, and lived there for two years, and have been studying it for thirty years.
This is only partially true and only for some words. In some words characters were chosen almost at random to match a loan word's sound from some other language than chinese (before katakana began to be used extensively). In other words the meaning has heavily drifted from the originally chosen characters and the current meaning bears little resemblence to it's original meaning. In other cases yes you can determine the origin of the word but it's not helpful to determine it's modern uses (for example the over use of the 女 (woman) radical in many words that were thought to be "related to women" with a sexist meaning). Also in some words the kanji were dropped and replaced with other kanji of a similar sound so as to simplify the number of kanji down to the current "2000" core kanji used in all official documents. https://en.wikipedia.org/wiki/Japanese_script_reform
I do still think a Japanese speaker would see "sen" at the end of that word and notice the connection to a military root word.
When you say partially true about the radicals, I'm a bit surprised. Is it really the case that this is only for some words? I believed it to be true for most words; that the radicals almost always influenced a thought, usually a thought that was true at the very least when the character was introduced. The thought might be archaic and sexist (that's a good point) but that it would be understood and noted by the reader. You are saying that's not the case with most radicals, or in the rare case more often?