Read on Twitter
I have spent a ton of time on this topic and I pretty much disagree with everything in this thread.
This view is informed by analysis of actual real-world numbers and examination of the underlying economics and characteristics of HSR infrastructure across multiple geographies. https://twitter.com/michaelxpettis/status/1354703238902263808
This view is informed by analysis of actual real-world numbers and examination of the underlying economics and characteristics of HSR infrastructure across multiple geographies. https://twitter.com/michaelxpettis/status/1354703238902263808
@MacroPoloChina growth projections for HSR are reasonable if not outright conservative upon closer examination.
It assumes growth to peak passenger density (in 2035) of about 30% from a 2019 baseline: https://macropolo.org/digital-projects/high-speed-rail/methodology/
It assumes growth to peak passenger density (in 2035) of about 30% from a 2019 baseline: https://macropolo.org/digital-projects/high-speed-rail/methodology/
This is supported by theory (network & J-curve effects) and overwhelming empirical evidence.
I did a thread last year on how the power of network effects will drive increasing efficiency and utilization as the physical density of the network grows. https://twitter.com/GlennLuk/status/1296249678351093760
I did a thread last year on how the power of network effects will drive increasing efficiency and utilization as the physical density of the network grows. https://twitter.com/GlennLuk/status/1296249678351093760
The definitive 2019 World Bank study on HSR in China provides color at the micro-level showing exactly how and why connecting existing point-to-point lines unlocked "bonus" demand. https://twitter.com/GlennLuk/status/1296249692985020418
At the system level, actual ridership data shows how since 2015, network utilization and efficiency have increased even as the length of the network has grown.
Avg. operational track increased 46% from 2015 to 2019. Annual ridership increased 97% over the same period.
Avg. operational track increased 46% from 2015 to 2019. Annual ridership increased 97% over the same period.
This is partly due to something known as the J-curve effect.
It takes time for passengers and urban neighborhoods to adjust to new HSR infrastructure. It takes decades to reach peak utilization.
@MacroPoloChina does a good job showing this:
It takes time for passengers and urban neighborhoods to adjust to new HSR infrastructure. It takes decades to reach peak utilization.
@MacroPoloChina does a good job showing this:
The weighted-average age of a km of HSR track in China is <5 years.
It took Japan's famous original Shinkansen line 40 years to reach peak ridership: https://twitter.com/GlennLuk/status/1299203038268198914
It took Japan's famous original Shinkansen line 40 years to reach peak ridership: https://twitter.com/GlennLuk/status/1299203038268198914
@michaelxpettis also asserts incorrectly that HSR infrastructure cannibalizes existing infrastructure investments.
HSR is mostly complementary with alternative transportation, including conventional rail. https://twitter.com/michaelxpettis/status/1354703241972506624
HSR is mostly complementary with alternative transportation, including conventional rail. https://twitter.com/michaelxpettis/status/1354703241972506624
HSR and airports generally serve different types of trips. While HSR did reduce air traffic along direct point-to-point routes (like Beijing-Shanghai), overall air traffic has continued to grow rapidly.
HSR generally cannot compete with air >1,000 km or indirect connections
HSR generally cannot compete with air >1,000 km or indirect connections
Meanwhile, as HSR was built on purpose-built (typically raised viaducts) track, existing track was freed up for conventional rail and freight.
Prior to HSR build-out, China's rail transportation infrastructure was severely supply constrained.
Prior to HSR build-out, China's rail transportation infrastructure was severely supply constrained.
Meanwhile this "sanity check" form of high-level analysis falls apart under closer examination and belies a real lack of understanding of why HSR track makes economic sense in some places and makes little sense in others https://twitter.com/michaelxpettis/status/1354703247651594240
Here I analyze why HSR makes a lot more sense in China vs. the US.
It's about population density, urban vs. suburban development, land acquisition costs and a bunch of other factors.
It's not just about GDP or land mass. https://lukecapital.substack.com/p/hsr-china-us-contrast
It's about population density, urban vs. suburban development, land acquisition costs and a bunch of other factors.
It's not just about GDP or land mass. https://lukecapital.substack.com/p/hsr-china-us-contrast
Ultimately the most important metrics that matter are utilization and affordability.
And here you can see how China's HSR system stacks up to others https://twitter.com/GlennLuk/status/1297817109733371906
And here you can see how China's HSR system stacks up to others https://twitter.com/GlennLuk/status/1297817109733371906
It's middle-of-the-pack in terms of utilization, but once again it's a very young system compared to more mature systems in Japan and France. https://twitter.com/GlennLuk/status/1297315954930536448
In terms of affordability, which of course adjusts for relative levels of income, HSR tickets are quite affordable and reasonably priced compared to both other countries ... as well as alternative modes of transportation (like air travel).
Significantly, with disposable incomes rising rapidly, travel is getting even more affordable over time.
Today travel by HSR is no longer just for wealthy Tier I city urbanites but pretty much all socio-economic classes.
More on this here: https://lukecapital.substack.com/p/charts-data-1-hsr-affordable
Today travel by HSR is no longer just for wealthy Tier I city urbanites but pretty much all socio-economic classes.
More on this here: https://lukecapital.substack.com/p/charts-data-1-hsr-affordable
This includes migrant laborers. We can infer this from this chart, which shows average trip length for HSR and non-HSR.
Migrant laborers (travel less often, but farther per trip) made up the bulk of non-HSR and up until 2016 you can see how the number was constant.
Migrant laborers (travel less often, but farther per trip) made up the bulk of non-HSR and up until 2016 you can see how the number was constant.
Avg. trip length for non-HSR started to decline in 2016 - strongly suggesting that conventional rail was now being used more for shorter trips — cities that were not yet connected via HSR — instead of long trips taken by migrant laborers.
I remember traveling in HSR 2nd class in 2016 and noticing half the cabin were migrant workers - anecdotal evidence that their disposable incomes had reached a point where time savings and comfort were worth the higher ticket price. https://lukecapital.substack.com/p/can-poor-people-in-china-afford-a
In the long run, conventional rail is likely to be comprised of much shorter trip lengths — connections between smaller towns and cities where it doesn't make sense to connect them via HSR lines.
The threshold for HSR vs. non-HSR connected towns is 200,000.
The threshold for HSR vs. non-HSR connected towns is 200,000.
