China's Economy is 60% Smaller Than We Thought

This video has been sponsored and approved by Incogni.

LIKE CHATTING ECON WITH ME?
△ I have a private Discord server for Senior and Chief economist Patrons / members.
Otherwise I sometimes hang out in two Discord servers:

SOURCES (including links to paper):

CALCULATIONS:

Timestamps:
0:00 - introduction
1:22 - official figures
3:31 - evidence
5:13 - sponsor
6:31 - true GDP
8:15 - Pro's & Con's

Made Possible by Research Assistance from Carlo Humpert
In Memory of Macrocat (Henry)
Narrated and produced by Dr. Joeri Schasfoort (University of Groningen)

[ScottWaltonDev]

I'm so sorry to hear about Macrocat. I've lost cats to being hit by cars before and it's really horrible. Wishing you all the best Joeri.

[schumanhuman]

There was a similar approach to research carried out by Prosper Australia looking at under reported vacant properties in Melbourne in 2017, but rather tha using lights they looked at water records which revealed many supposedly occupied properties had less water usage than a dripping tap.

[DAAN]

Thanks for providing actual quality content while protecting the researchers values AND explaining them to everyone. The world needs more people like you!

[aaronduff]

I know I’m late but great vid man I’m sorry for your loss it’s never easy losing a family member who’s been with you for a long time. I’m new to the channel but I love the effort you give to these videos and your cat who was always apart of them. Best wishes and prayers for you and your cat ❤

[docopoper]

What I would love to see is some independent studies trying to calculate the same deflation factor using different data sources. Then we could compare their error bars.

[tobiasL]

If this number is really in the same ballpark, that's remarkable, a 50% smaller economy is quite insane!

[best]

Hey man, sad for your loss! Remember the good work that you are doing for us millions of people and that will give you strength to carry on

[basilcharleston]

One of the reasons may indeed be that ghost cities were counted in GDP but once built turn into a mere deadweight. Another (and more speculative) one is that China may have overstated its population, especially the number of working age people.

[frogsandmushrooms]

I love this channel and the information it presents... but i also loved the glimpses of Macrocat. He contributed to the vibe of the channel and did a good job keeping things relaxed and not over-serious. Thank you for mentioning that he did not suffer... but my heart is still in mourning for Henry. RIP Macrocat and rest easy.

[alexf]

Another reason (in addition to the one I posted on air quality) for China's night lights not growing as fast as gdp could be that a lot of Chinas economic activity was building "Ghost Cities" and properties that were not inhabited, but instead held as a second home. These would not generate as much light, as only infrastructure surrounding the uninhabited skyscrapers would be lit.

[tomdagan]

Hi, great video! One question I have though, is how is it not more noticeable and has no evidence other than the fact it has less lights than expected? I could see how it makes sense for even something like 20%, but more than half of a country’s economy being fictional is bound to make people notice sooner and in more forms, isn’t it?

[jamesodell]

One must also consider the switch to LED street lighting. LED light can be directed down and throws less light into the night sky which could affect the results.

[WilliamChan]

Rest in peace, Henry. It's tough to lose your friends, especially before their time, so wishing you the absolute best

[fernrfernandez]

Couple of issues with this study. First, the relationship with night light and GDP has been shown to depend on the level of urbanization. For example, study by Xaquín S.Pérez-Sindín et al. (2021) found that the relationship between satellite measured night light and GDP was strongest for urban areas (coefficient of 0.71 with an R2 of 0.5 for 500K centers vs.0.57 and an R2 of 0.32 for areas with 20-50K). Although having a large population, China is still less urbanized than the US and most Western European countries. Urbanization in China is ~61% vs. 82% in the US, with the difference being much larger during the 1990 - 2010 period.

Additionally, Martinez assumes that the relationship between night light and GDP is linear. There's no reason to assume this as you might expect that increased GDP in highly developed countries might be near a saturation point; at some point this relationship will break down as economic growth is less tied to population growth and increased light generation (just an aside, health expenditure as a fraction of GPD in the US is 19% and has been a major contributor to US GDP growth. How is that tied to light generation?).

Further, the R2 values are generally low (~0.3), indicating that there are a host of other factors far greater than night light that account for changes in GDP. Again, linear regression with values this low can't be used to conclude something as specific as China lying about its GDP numbers. Finally, the use of the "Freedom Index" is a little dubious. It seems unlikely that something as simplistic and ideological as a "Freedom Index" will capture any nuance in something as complex as a modern government. Correlation is not causation. In complex systems (especially anything as complex as economies and government) one will always find an endless number of variables that are weakly correlated. However, explaining how these variables relate mechanistically and causally is the hard part, which Martinez fails to do.

When coupled with the low R2 values and low correlation there really isn't anything specific one can say about how governments lie about GDP measures. If this was a paper in a prestigious science journal it would have been rejected, not because its methods are wrong, but rather because its conclusion cannot be justified on the basis of the results and analyses presented. The only thing we can say with certainty is that there is a relationship between night light and GDP (with plenty of unexplained variance) and lots of variability between countries, regardless of whether we subjectively classify them as free or not.

[stokeynathu]

thanks for sharing this paper and gives a insightful remarks.

[jkuang]

So no China Threat right? Yeah! Let's get back to enjoy our lives!

[masenformen]

Thank you so much for having Henry lighten us up all this time. This wonderful cat brought joy to you and so many others here. I'm know he too is grateful for all the love you gave him.

Rest in Peace, dear Macrocat.
F

[PBoyle]

Great video Joeri. Very sorry to hear about your cat. I hope you are doing well.

[CaliforniaDreams-ebsx]

Aside from light activity, I think there has been other metrics used such as food production/consumption, steel usage, and other commodities that people can try to rely on for GDP growth

[hr]

Very intriguing. Thank you for sharing buddy