On the 'Slow' Chinese Internet and the Prospects for China: One More Round

Just off the long-haul Shanghai-Newark-DC National route and stumbling into our house. For another time, perhaps even tomorrow: why Newark has become my new favorite airport for trips into and out of China, apart from the Cory Booker factor and the welcome fact that it's not Dulles.

In this original item on the significance of China's "slow" internet, based on this NYT essay drawn from this book, I argued that the speed of internet access in China, relative to the super-quick networks in Japan, Singapore, and South Korea, was a useful proxy for China's openness and modernization overall. It has some modest direct effect on China's ability to nurture first-rate world research centers, and it is more important as a marker of the ongoing tensions between the security-state and the entrepreneurial forces in China's leadership.

Then in this reply, a number of Chinese and Western tech officials said that the subtler and more intriguing aspect was the difference between (very high) communication speeds inside the Chinese "Great Firewall" and (often very slow) speeds across the Firewall, to sites in the outside world. As one tech official in China wrote:

The big question is not whether or not China can build a world-class society while fighting the internet, the question is whether or not it can do so while building a giant intranet that is China-specific. China is big enough that I think this is something of an open question.

I am going to double-down and say: if, in the long run, internet users in China suffer penalties reaching sites outside the country, then no matter how big and important the Chinese web-o-sphere becomes, it will not be "world-leading" or "world-class," because much of the world will be walled off. To be clear, I hope for China's continuing more complete integration with the rest of the world. That would make some Chinese firms more profoundly "competitive" to Western incumbents -- Apple, Google, GE -- than they are now, but it would also suggest that the "Chinese system" as a whole, and most Chinese people, would more easily interact with the rest of the world.

Now, several further installments on the longer-term significance of this proxy for Chinese openness.
1) From a Chinese-American reader:

I was from China.  I came to US to pursue graduate study and have since settled here.  I am working and raising family in Northeast.

It seems to me indeed Internet speed in terms of accessing servers outside China may not be much of an issue for the first wave of outsourcing (the manufacture outsourcing) which has been going on for the last 15-20 years.  After all what matters at the end of the day is the speed to transport 'goods' across pacific (US centric view (:-)).  

However, perhaps we are now witnessing the second wave of outsourcing (office work, research and development) where the product is no longer goods, rather it's data and information.  Computer systems and applications are integral part of this type of work.   Internet speed in terms of accessing servers outside China will likely matter.

2) From the CTO of a Western tech firm that is expanding operations in China (emphasis added):

Our company runs a service with [several tens of millions of ] subscribers out of California and recently launched a China-only version of our service out of data centers there. We are adding around [tens of thousands] of new Chinese users per day....

The firewall issues going in and out of China get a lot of attention, but the state of networking within China is a bit hairy as well.

There are a small handful of network carriers in China, all with state-backed status (e.g. China Unicom, China Telecom, etc.).

In spite of the state backing, these carriers frequently refuse to carry network traffic from each other. I've even heard unverifiable stories from Chinese Internet professionals of these carriers cutting each others cables when they encroach on each others' turf. [JF note: this resembles some of the stories I've told about trying to create a truly nationwide air-travel network.]

This means that any Internet service in China needs to connect individually to each of these network carriers. You can't just connect to 2 or 3 high-quality carriers and expect them to route traffic properly to the others like you do in the West. Instead, each service needs to deal with direct connectivity to each provider.

The top Internet Data Centers (IDCs) provide some aggregation and intermediation services that allow you to pay them to connect your service to all of the networks, but this costs about 10x as much as comparable Internet connectivity in the US.

In addition, network performance within the country varies widely by geographic locale. East-west networking tends to be much better than north-south. Our users in Guangzhou say they get better performance to [our site in the US] than they do to our servers in Beijing.

All of this translates into some significant inefficiencies for Chinese Internet companies that their western counterparts don't have to bear, independent of The Firewall. I.e. paying 10x as much for unreliable bandwidth as well as higher network engineering labor strikes me as the sort of problem that may eventually result in some market pressure from businesses on the government to provide some level of deregulation of internal network telecommunications.

I.e. it's as if factories in China had to pay 10x as much for unreliable state-run electricity compared to competitors in Mexico or Thailand.

3) From a Western academic with extensive experience inside China and with the Chinese language:

Having [recently] suffered through a year of living and teaching in China,  it is all too painfully obvious to me that the internet controls are harmful to the Chinese people.  My own students often complain, and they do everything they can to circumvent the controls and restrictions, but with only partial success.

I would estimate that well over half of the capability of the internet is unavailable to the Chinese people.  Surely that has a negative effect on the dissemination of knowledge and information in the PRC.

4) And from another person familiar with the Chinese tech world:

I agree, throttling (or mismanaged setups of DNS that unnecessarily limits speeds)  is certainly an issue for power users who depend on hi-quality, consistent internet speeds. And combined with the censorship in China, it's just another reason for outside firms to not even bother trying to crack the Chinese market or to invest it in. It's certainly a boon to local firms since.. there is no net neutrality in China and firms with guanxi or money might be able to purchase better internet speeds.

The big picture, once again, is that so much is happening so fast in China, and we're all trying to make sense of where it might lead. More -- and on a variety of backed-up topics -- after a bout of Eastern Daylight Time zone sleep.