Craig Allen: Keynote Speech at Stanford-CSIS Big Data China Annual Conference
Aside from biweekly SCCEI China Briefs, we periodically feature relevant content from other SCCEI programs, like this keynote from Craig Allen, outgoing President of the U.S.-China Business Council
Craig Allen is the President of the U.S.-China Business Council (USCBC). Prior to joining USCBC, Allen had a long and distinguished career in U.S. public service, beginning in 1985 at the U.S. Department of Commerce. Over the years, he served in various diplomatic roles in Beijing and Tokyo, and also served as U.S. Ambassador to Brunei from 2014 to 2018. As President of USCBC, he has met repeatedly with China's top leadership, including Xi Jinping.
Craig Allen delivered the following keynote speech at the Stanford-CSIS Big Data China Annual Conference on “The Turning Point? U.S.-China Relations, Economic Growth, and the Race for Technology Leadership.”
Craig Allen: It truly gives me great pleasure that CSIS, Stanford, and USCBC are so well aligned.
Let me start on a very personal note.
I travel to China frequently and I dread losing contact with my family when I am on the road. I cannot use my WhatsApp. I lose access to my regular news resources. It is hard for me to get around. I even have a hard time paying for things.
I do not have these problems anywhere else in the world.
While my personal issues within China are indeed trivial, they are also emblematic of the unintended consequences of the increasing bifurcation of information systems, and even more so, of big data.
Let me start out with something that everyone in this room knows well. Last week [week of Dec. 2, 2024] was a tough week for U.S.-China technology companies.
On Monday, the U.S. Department of Commerce published 200 pages of new regulation. At the risk of oversimplifying, there were: i) 140 new entity listings; ii) 24 new technology controls; iii) A very significant expansion of the FDPR, over any semiconductor – with no de minimis.
Last week, Congress also brought pressure to bear – with its own vision of how the administration’s national security toolbox should function. While this is not yet finalized, we have a number of possible riders to the NDAA that we are watching – associated with Big Data, including: i) a version of outbound investment that allows Congress to add additional sectors; ii) The BIOSECURE Act; iii) Controls on remote access of AI hardware.
While we do not know yet if the final NDAA bill will include these riders, please rest assured that they will remain on the legislative docket in 2025 for future consideration.
My point is that it is important to recognize that the Congressional/Executive Branch dynamic has repercussions for policy. The Congress is only going to push in one direction. They seem to want to legislate additional controls on Big Data, even if a regulatory approach might be more flexible and appropriate.
On Tuesday last week, the bad news continued. The response from the Chinese side from the semiconductor controls was immediate and much more robust than I had expected. The Chinese response includes:
i. Licensing with presumption of denial to export to America of Gallium, Germanium, Graphite;
ii. Four trade associations told members not to buy U.S. products;
iii. These measures are clearly aimed at the U.S. military and industrial complex;
iv. Also, clearly aimed at sending a signal to E.U. and Japan;
v. Earlier there was an announced introduction of the Chinese FDPR, which could have profound consequences.
In my view, these actions by both governments last week, reciprocal tightening measures in the name of national security, may very well indicate a new era of U.S.-China bilateral technology regulations governing Big Data.
I especially think that we should not underestimate the Chinese response, especially the introduction of a foreign direct product rule and the banning of exports to the U.S. of some critical minerals.
The Chinese measures – especially a prospective robust use of the Foreign Direct Product Rule – could collectively be very disruptive to American commerce and manufacturing.
As this is the big data China annual conference, let me explain where we are from the perspective of leading U.S. technology companies with a presence in China.
Tech is caught in the middle and is increasingly compelled to bifurcate.
We’re at a point in the relationship where companies must either localize, which allows them to capture Chinese domestic markets or decouple, losing opportunities in China but also reducing business and reputational risks associated with China.
What’s driving localization, and what’s driving decoupling?
For both, it’s a mixture of policy and economics.
When we look at localization, we first need to think about China’s pivot from excessive real estate and infrastructure spending to excessive spending on technology and advanced manufacturing.
Under the moniker of “new quality productive forces” and “high quality development,” massive amounts of financial resources and political support have been given by the Chinese government to Chinese technology and manufacturing companies.
Technology, China’s planners hope, will increase the individual wellbeing and incomes, while providing an outlet from the ailing real estate sector.
In a previous speech at Stanford, I have called this a form of techno-utopianism, which none of my Chinese friends have challenged or denied.
A good economic example of the techno-utopian pivot occurred about a year ago when bank credit to manufacturing surpassed credit to real estate.
From a political perspective, the pivot occurred at the last Party Congress [2022], when the Central Science and Technology Commission was created to give support to New Quality Productive Forces. They are systematically expanding R&D in almost all sciences and aggressively streamlining the commercial viability pipeline.
The Chinese are not being shy. They are aiming to leapfrog U.S.; European and Japanese manufacturing and we should not be complacent about their ability to do so.
For MNCs, China’s pivot to tech is both a challenge and an opportunity. If an MNC wishes to ride the wave, the opportunity is often addressed through localization.
To capture opportunities presented by the tech pivot, MNCs are increasingly compelled to localize manufacturing and R&D. Doing so affords them the flexibility needed to chase domestic markets and respond quickly to Chinese customers.
China’s advantages in talent, scale, infrastructure, in addition to the market, make a robust and active presence in the country important to many MNCs’ local, and global competitiveness.
MNCs’ local peers – especially private-sector Chinese companies – are also becoming wickedly competitive. Chinese domestic markets are becoming increasingly competitive, which compounded by slower economic growth and excess supply, often result in severe pricing pressure.
In response, MNCs often rely on quality and seek specialized niches, where competition is not as fierce, but this also requires localized R&D.
Chinese policy also drives localization. A key example that matters for our members are barriers to entry for government and SOE procurement.
Depending on the sector and product, SOEs may be as much as 100% of the marketplace.
I should note the pernicious influence of documents 79 and 551. These are both secret or “neibu” documents. 551 is from the Ministry of Finance and the Ministry of Industry and Information Technology. It is available on the Internet. If you google it, you will see that it bans import of categories of medical equipment.
Document 79 is purportedly a secret document from the State-Owned Assets and Supervision Administration and Supervision Commission (SASAC) that requires State Owned Companies to remove all foreign ICT from companies over a period of time. I do not have a copy of this document – but we see its pernicious results as SOE’s cancel long-standing contracts.
Interestingly, both of these documents seem to be directed at all foreign equipment and all foreign imports. They are not specifically anti-American. And, thus we have common cause with the EU and Japan to protest this unfair treatment.
I should note parenthetically that USCBC has publicly called for the cancellation of 79 and 551.
But secret documents and excessive subsidies to local champions are not our only problems. We have a database of some 130 market access barriers, shifting the playing field in the direction of Chinese players.
An important caveat is that localization does not always equate to access. Many MNCs still face discrimination even after near total-localization. Simply, the playing field is not level. It is downward sloping for SOE and private-sector Chinese companies. It is upward sloping for joint ventures, or products made in China by MNCs. Foreign imports face the steepest upward sloping playing field.
In tech, another important policy is China’s AI registration rules. To run embedded AI applications on ICT products, MNCs partner with local AI firms because U.S. algorithms are often either inferior or fail the registration process.
There are many reasons to be in China. For many USCBC members, China is still the fastest growing end market for technology, and it continues to outpace other regions.
But there are also other headwinds, that sometimes cannot be overcome.
Let’s turn to decoupling, and the subject of export controls.
There can be absolutely no doubt that U.S. export controls have resulted in the widespread abrogation of bilateral commercial partnerships in the tech sector.
Beyond the obvious impacts on sales to sanctioned entities and controlled products, U.S. export controls have had the unintended consequence of accelerating, optimizing, and providing political legitimacy for China’s costly technology self-sufficiency campaigns.
I once had the opportunity to talk to a former MIIT Minister. I said that the Made in China 2025 campaign violated every chapter of the WTO. He did not argue or deny my assertion. But, in reply, he said that that may be so, but it was because of the U.S. export controls. He felt that they had no choice.
In addition to coinciding with the pivot from real estate to tech, export controls have created the perception that American companies are unreliable, which has precipitated active efforts by Chinese industry to “delete A,” or design out products American firms.
The most egregious example of “delete A” is in government and SOE procurement for ICT products.
Through the aforementioned Doc 79, and most recently a series of un-accreditable procurement standards, MNCs have virtually lost all ICT procurement opportunities to either the government or the SOE sector. Now, they are fighting to prevent these initiatives from seeping into the private sector.
Before the imposition of wide-scale export controls, our members report that Chinese firms were dismissive and at best, perfunctory, when the government called for import substitution.
Now, export controls, or the potential threat of supply cutoff from a “subject-to-the Department of Commerce’s Export Administration Regulations supplier” somewhere upstream, have left Chinese companies with little choice but to form new partnerships with European, Japanese or Korean companies or with each other.
This has certainly benefitted Chinese competitors. Look to semiconductor equipment manufacturer Naura for a prime beneficiary of these new partnerships. Their revenue quadrupled in the last two years!
Export controls have also given the Chinese government precise direction on what technologies need to be subsidized to minimize exposure to U.S.-origin chokepoint technologies.
This is evident from the way the National IC Industry Fund, or big fund, has shifted its resource commitments from design to tooling and fabrication over the course of several fundraising rounds.
Immediately after the last round of semiconductor rules, four leading trade associations in China and representatives of many MNC’s most prolific customers, called on domestic firms to band together to form new, de-Americanized partnerships and to use caution when procuring U.S. ICT products.
The Cyber Security Association of China has called for cyber security reviews of prominent American suppliers. While these have not yet been triggered, the prospect of such a review should be a concern to all foreign technology suppliers.
Policy developments in both countries have also given rise to a new, and very serious concern: conflict of law.
When our members were asked about compliance challenges of export controls: 34% said export controls caused conflicts between U.S. and Chinese legal regimes.
We are especially worried about (1) our companies’ ability to conduct due diligence investigations in China and (2) about our ability to ensure consistency of supply to valued customers – some of whom are the biggest companies in the world.
About due diligence, 75% of our members said due diligence for compliance purposes was the main challenge of export controls.
Companies are concerned that they will violate China’s anti-espionage law and data security law, because, to get an export license, they must give information to BIS on their customers’ customers and their technological capabilities.
Such information is potentially a state secret. And, we do not wish to put the employees of our companies in harm’s way.
Chinese companies know this, too. Therefore, many Chinese buyers will simply walk away from their American suppliers – if an alternative is available – rather than attempt to use BIS’ licensing system.
With regard to consistency of supply, MNCs are concerned that because when they comply with U.S. export controls, they may unintentionally violate China’s cybersecurity review measures. According to China’s regulation, one must be considered “reliable” to sell to critical information infrastructure operators.
Compliance with U.S. law may potentially lead to American companies being placed on one of China’s two punitive lists: the unreliable entity list or list maintained by the anti-foreign sanctions law.
In short, it is becoming more and more difficult for MNC – especially American MNC – to maintain normal business ties in China in full compliance with both American and Chinese law. This is especially true in the case of big data and indeed all national security sensitive technology.
But it is not easy to walk away. This brings me to the topic of talent, which is perhaps the greatest challenge of the tech competition.
When the U.S. imposed novel restrictions on U.S. persons activities, many engineers were immediately compelled to choose loyalties. Many chose to stay in China with their families and went to U.S. firms’ competitors.
The near-blank check the Chinese government affords its domestic tech champions allows Chinese companies to spend a proportionally greater amount of revenue on R&D. Some estimates are that the Chinese spend 30% of revenue on R&D, in contrast to an average of about 20% of revenue, as reported by several U.S., European and Japanese semiconductor companies.
The majority of R&D spend goes to talent, and it is not uncommon for MNCs to face stiff competition from firms that were founded by their own former engineers.
How to manage through this volatility? Companies need to invest heavily in compliance resources and consider what you need to localize, especially to maintain relationships with key accounts. A key challenge for foreign MNC in China is to protect your talent – which is often mission critical for the entire corporation.
Chinese STEM talent is among the best in the world, and there are huge number of engineers, scientists, technicians in China. They are not only engaged in product localization. Increasingly, they are engaged in franchise-wide, mission critical R&D activities.
Foreign companies need to maintain access to this talent pool, or we will be at a global competitive disadvantage.
Let me conclude by saying that U.S.-China science and technology cooperation has created a cornucopia of good for both countries and for humanity.
However, the very flowers of this past success are producing the seeds for systemic disruption in the very near future.
The fact of the matter is that both countries very much need each other, to continue the tech boom on a sustainable basis in the future.
The fact of the matter is that those companies that were able to create synergies between Shenzhen and Silicon Valley became world beating giants.
Shenzhen and Silicon Valley both had their advantages and collaboration between the two created a holy grail of innovation, productivity growth and prosperity.
But the reality of the matter is that – due to legitimate national security concerns by both governments – this is increasingly less viable.
I do not think that we are at the end of an era. But I do think that it will be increasingly difficult for American and Chinese technology companies to collaborate in a deep manner, due to governments’ deep suspicions and lack of confidence.
Let me end on a note of confidence. Both systems have their advantages. Do not underestimate the enormous innovative potential of Shenzhen, Hangzhou, and Beijing – especially in terms of organizing talent and resources to meet a specific end.
At the same time, the U.S. system – prizing diversity, rule of law, free movement of people and ideas, efficient allocation of capital and the vast rewards for innovation – are overwhelming advantages in any bilateral contest.
At the same time, we must not take our own advantages for granted, nor dilute them through poor governance. Thank you.