A thread on the scale of US versus Chinese infrastructure spending, following the big new US bill.

My assumption was that even with the new $1trn commitment, the US would still significantly lag. That's partly true. But not entirely, which is where it gets interesting. (1/x)

Start with the headline numbers. (With the caveat that apples-to-apples comparisons are *very hard* because of data gaps, especially from China, plus slightly different definitions of infrastructure by each country's statistics agency.)


So the headline:

*US public spending on infrastructure is about $550bn/year (2.5% of GDP).
*China's is about $2.8trn/year (nearly 20% of GDP).

Not entirely surprising when China has built 30,000kms of high-speed rail in past decade and the US zero... but, still, gulp.


Now, here's where things get a little more surprising. First, incremental increase in Chinese infrastructure spending has slowed sharply since 2017: annual growth has averaged just about 2% y/y. The US could now be headed for 5%+ y/y on infra (depending on implementation). (4/x)

Second, when narrowing down to specific categories, spending gap between the two isn't quite as extreme as headline implies. In transportation, for example, China's annual investment is about two-times bigger than in the US. And that gap will now narrow. (5/x)


(For what it's worth, I think China's broader headline infrastructure data is inflated by cost of land acquisition plus some GDP massaging by local officials. Since 2018 China's data definition of infra has changed, trying to address some of these problems.) (6/x)

Third, and most surprisingly, stock of capital in China and the US is now remarkably similar (*depending on which gauge you use*).

Best apples-to-apples is nominal capital-to-GDP ratio in local currency. Here's that: (7/x)


(Note that chart in the preceding tweet includes residential investment... data limitations, sorry. I'm relying on Penn World Table, best international growth comparison database.) (8/x)

Accepting that limitation, the big picture:
China does not look horribly over-invested; super-fast investment growth, but it was catching up from a low starting point.
And the US, for its part, does not look horribly under-invested. (9/x)

But there are other gauges. If you use the Penn World Table's real data (which controls for inflation and, most notably, assumes that China's growth is much lower than officially reported), the picture is starkly different. Here's what it looks like: (10/x)


From this vantage, China's trend does look crazy, while the US trend is gradually downhill. Now, I think Penn (using Conference Board work) may have over-corrected China data. So perhaps something between the nominal & real charts in this thread would be more accurate. (11/x)

And if that is so, then the divergence in investment trends that we may now be seeing (close to 0% infrastructure growth in CN and a decent-sized push in the US ) seems, well, reasonable. (12/x)

One sad note for lovers of bullet trains. The US is not, alas, going to come anywhere close to what China has built. Lion's share of transport spending in the US still goes to roads. And rail is much more expensive to build in the US: https://transitcosts.com/what-does-the-data-say/ (13/x)

If you've made it this far and want to read more about the new US infrastructure bill and what it means for the economy, here's my analysis on that. TLDR: no bullet trains, but it is important and could be a real boost to growth. (14/x) https://www.economist.com/united-states/joe-biden-passes-the-less-contentious-half-of-his-legislative-agenda/21806187

And a piece from 2018 about China's infrastructure slowdown, which, I think, still holds up reasonably well. (15/15) https://www.economist.com/china/2018/09/20/is-chinas-infrastructure-boom-past-its-peak