#83 The US Is About to Seriously Consider Restricting RISC-V
Renewed US Interest In Restricting China’s Access to RISC-V; A Consideration of the FTC’s Ban on Non-Compete Agreements
As the US Department of Commerce announces its intentions to review concerns surrounding China’s access to RISC-V, Satya Sahu highlights why it’s probably not a very good idea for export controls to aim for goals of complete denial of technology.
Arindam Goswami dissects the US Federal Trade Commission’s recent ban on Non-Compete Agreements, arguing that a nuanced, context-specific approach may be better compared to a blanket ban.
Also,
We are hiring! If you are passionate about working on emerging areas of contention at the intersection of technology and international relations, check out the Staff Research Analyst position with Takshashila’s High-Tech Geopolitics programme here.
Course Advertisement: Intake for the 38th cohort of the Graduate Certificate in Public Policy Course (GCPP) has closed, but readers can still apply before the course kicks off on May 4.
The GCPP will equip you with policy fundamentals and connect you to over five thousand people interested in improving India’s governance. Check all details here.
Cyberpolitik: Renewed US Interest In Restricting China’s Access to RISC-V
— Satya Sahu
In a letter sent to lawmakers earlier this week, the US Department of Commerce said that it would “review potential risks” and assess whether the Department could effectively address any potential concerns surrounding China’s work in open-source RISC-V technology.
RISC-V, an instruction set architecture (ISA) that defines a processor's capabilities and execution methods, is a completely open standard managed by the Swiss non-profit organization RISC-V International. In contrast to proprietary ISAs such as x86 and ARM, RISC-V's open-source nature is revolutionising open-source chip development, much akin to the impact of open standards like Ethernet, HTTP, and USB on connectivity and web development.
The open-source nature of RISC-V means that organisations from all around the world, including from India, Europe and China, could contribute to it and implement the standard for multiple use cases without having to pay the costs associated with proprietary ISAs. This has made it very popular with chipmakers/designers designing custom chips for workloads like AI, which do not require the extended instruction sets of x86 or ARM. More importantly, RISC-V is very attractive to countries seeking to build out their domestic semiconductor design ecosystems, like India and China.
This is where the US Department of Commerce’s letter comes in. In October and November last year, US lawmakers categorically asked that US export controls targeting China’s access to advanced computing chips be expanded to include controls on RISC-V technology as well. It would seem that the Department of Commerce has finally heeded this request.
In a blog post from November 2023, my colleague, Rijesh Panicker and I pointed out that attempting to restrict an open-source technology, designed to be collaborative and borderless would be difficult, and also antithetical to the spirit of innovation; something that could backfire easily. However, there are some imperfect means of implementing restrictions:
In its current form, RISC-V has contributors from over 70 countries collaborating on its development and is managed by a non-profit, RISC-V International. It moved its headquarters to Switzerland to signal its neutrality and pre-empt potential geopolitical disruptions amidst the US-China tech war. However, US export controls regulate not only the export of goods and technologies from its mainland but also their subsequent re-export, including know-how, data, and blueprints, if they involve US-origin technology or equipment.
Imposing export controls on open-source technology like RISC-V is a complex endeavour, and it is unclear how US policymakers would achieve that. RISC-V International relies on its members (including US-based Google and Nvidia) to contribute to the ISA and develop software. This provides a potential pathway for the US to control RISC-V's proliferation by imposing controls on these companies.
Effective export or technology controls are possible at different layers from the core architectural standards. For example, the extensions to the RISC-V ISA may depend on underlying patent pools and could be subjected to restriction depending on the use case. Similarly, imposing export controls on the actual manufacturing value chain for chips, such as fabrication technology, may be an effective way to control the dissemination of the technology.
Such a move by the US, driven by national security and technological supremacy concerns, will target yet another facet of China's burgeoning semiconductor industry, which has increasingly turned to RISC-V in response to earlier American sanctions.
We further outline a potential strategy for India to navigate a world with a fractured RISC-V ecosystem if the US were to implement export controls. The full blog post is available here.
But as I had also mentioned previously on Technopolitik:
Unlike physical goods, policing the transfer of knowledge and information, especially in the digital age, is fraught with complexities. Implementing export controls on intangible technological standards poses significant legal and enforcement challenges when diversion risks exist, making the controls largely symbolic. Attempting to stop the blanket proliferation of the open-source ISA is an ineffective measure at best.
Complete denial as a goal of chip export controls is impractical when it comes to plugging leakages, enforcement, and monitoring. It will be interesting to see what course of action the Department of Commerce decides on. Perhaps it will attempt to mandate hardware-enabled governance mechanisms (HEMs) for all chips designed in the US. Again, as outlined here, this course of action is also fraught with implementation, legal, and geopolitical concerns:
Current export controls targeting China are generally internationally accepted, as they do not interfere with post-acquisition ownership rights; HEMs change this calculus. By introducing the capability to modify or disable a chip's functionality remotely, these mechanisms will engender persistent concerns among international consumers and governments. The apprehension is about more than losing control over purchased critical technological hardware and its advertised capabilities; it is also about the broader implications of US influence on countries' sovereignty.
The effectiveness of HEMs hinges on the possibility of a fragile international consensus. Without broad global buy-in, implementing a HEMs-based strategy will alienate vital US allies and estrange prospective partner countries. Such a move could also cede ground in the very technological domains and markets the US seeks to control in its bid to reduce Chinese sway on chokepoints in the global value chains for semiconductors and other transformative technologies.
Instead of seeking absolute control over every single layer of computing hardware to deny its adversaries access to technology, the US will be better off outpacing them.
India is set to host the Quad Leaders' Summit in 2024. Subscribe to Takshashila's Quad Bulletin, a fortnightly newsletter that tracks the Quad's activities through the Indo-Pacific.
Your weekly dose of All Things China, with an upcoming particular focus on Chinese discourses on defence, foreign policy, tech, and India, awaits you in the Eye on China newsletter!
The Takshashila Geospatial Bulletin is a monthly dispatch of Geospatial insights for India’s strategic affairs. Subscribe now!
Technomachy: A Consideration of the FTC’s Ban on Non-Compete Agreements
— Arindam Goswami
On April 23, 2024, the Federal Trade Commission (FTC) in the US issued a final rule banning non-compete agreements (NCAs) nationwide, arguing that they are an unfair method of competition. After the rule’s effective date, existing NCAs will be unenforceable, except for senior executives, and no new NCAs can be enforced or entered into with anybody. What do we make of this rule? What could be its consequences, both seen and unseen? Let us look at both sides of the argument.
The Good…
It seems rather intuitive that as per free market principles, in the marketplace of labour, the fundamental freedom of workers to switch jobs should be protected. Employers and workers should be able to bargain with each other on wages, working conditions and other aspects, in a free and fair manner, without any coercion and the sword of NCAs hanging over them. By hurting the mobility of workers, NCAs are accused of depressing wages, inhibiting innovation, preventing the growth of startups and businesses, and thereby hurting both businesses and labour. This negatively affects competitive conditions in labour and product and services markets. It leads to market concentration and higher prices for consumers.
Through sheer experience and explicit training, a worker gains skills and, thereby, value while working. However, due to NCAs, she cannot leverage this skill and value for better wages, working conditions, and increased value. Her career choices are restricted. Apparently, low-wage, low-education workers are disproportionately affected.
These agreements have gained renewed attention due to concerns about alleged reductions in labour’s share of national income and worker salaries not rising despite growth in the US economy. President Joe Biden’s July 9, 2021, executive order, Promoting Competition in the American Economy, directed the FTC to consider employing its statutory rule-making authority “to curtail the unfair use of non-compete clauses and other clauses or agreements that may unfairly limit worker mobility.”
There is another aspect of how this affects businesses, too. Due to non-competes, businesses are prevented from hiring the most relevant talent. There is an opportunity cost involved in this—the time and money spent retraining workers when a readymade talent pool is not allowed to be used due to NCAs. The FTC and its chair, Lina M. Khan, take this argument further to say that non-competes stymie the free flow of ideas, hamper innovation, and prevent business growth.
“The FTC estimates that the final rule banning non-competes will lead to new business formation growing by 2.7% per year, resulting in more than 8,500 additional new businesses created each year. The final rule is expected to result in higher earnings for workers, with estimated earnings increasing for the average worker by an additional $524 per year, and it is expected to lower health care costs by up to $194 billion over the next decade. In addition, the final rule is expected to help drive innovation, leading to an estimated average increase of 17,000 to 29,000 more patents each year for the next 10 years under the final rule.”
After all, innovation cannot flower in a controlled environment. That seems pretty intuitive, pretty obvious, doesn’t it?
The Bad..
But hold on! The other side, the ones not in favour of such a blanket ban, say non-competes, are, in fact, necessary for innovation. The argument is that businesses would invest in innovation, risky R&D investments and training only if they reasonably believe that their benefits would first accrue to them, not their competitors. In the cut-throat world of competition, access to intellectual property (IP) and skilled human resources is of foremost importance.
Non-compete agreements can also facilitate the development of specialised labour markets in specific industries or geographic regions. By restricting employee mobility, these agreements may incentivise firms to concentrate their operations in specific locations, leading to agglomeration economies and knowledge spillovers that benefit overall industry competitiveness.
And the Ugly..??
But again, all of this needs to be tempered with evidence. While some studies back the argument that non-competes cause an increase in training offered by companies, it is also found that such training causes a decrease in wages and worse wage outcomes over time, due to firm-specific training that cannot be generalised across firms or sectors. Also, there are other laws and rules to protect trade secrets.
We need to evaluate whether we truly require NCAs for this and whether they can prevent IP theft (if that is the intention). There is also a national security angle to these arguments — threats of IP theft and luring of human resources by rival countries (like China). However, NCAs don’t seem to be the most effective way of preventing this since such activity will probably be done surreptitiously, and there are already other possible governmental actions to deal with this.
Another line of argument against the ban talks about legal issues that might beseech this rule. As per this argument, Section 6(g) is a tool to aid only the investigative and procedural authority of the FTC, and doesn’t provide it any standalone substantive rule-making authority. Is this, therefore, an unconstitutional delegation of legislative authority? It can look at issues on a case-by-case basis, and not issue such a blanket ban. Historically, too, non-competes have been evaluated on a case-by-case basis through state law adjudication. Some states outright ban NCAs and some have restricted them, specifying circumstances and context, ensuring procedural protections. This, it could be argued, is a better way to handle NCAs, because different industries and sectors operate in different scenarios, and a one-size-fits-all approach may not be the best. Decentralised rule-making and adjudication are, in general, sound principles to follow along with free market principles.
Legal Issues
These are important legal issues to examine from a public policy perspective. These will set precedents; hence, defining limits to bureaucratic rule-making and understanding the decentralisation and federal aspects of rule-making and adjudication are essential. As economist Brian Albrecht points out, legal policy must be flexible enough to course-correct after considering new information. As Thomas Sowell writes in his book “Knowledge and Decisions”:
“the most fundamental question is not what decision to make but who is to make it—through what processes and under what incentives and constraints, and with what feedback mechanisms to correct the decision if it proves to be wrong.”
Legal challenges will also examine whether the FTC sufficiently considered pro-competitive justifications of NCAs, like protection of trade secrets, human capital investment, promotion of innovation and R&D, preservation of customer relationships (preventing soliciting former clients and customers), fostering of specialised labour markets, reduced costs and better quality for customers, etc. Some, like FTC staff economist John McAdams, have argued that there is no definite empirical and research consensus on NCA’s positive or adverse impact on salaries, free flow of ideas, competition, firm entry, innovation and R&D. In fact, they argue that it negatively impacts the above. It also raises concerns about error costs and economic welfare, since it doesn’t do a proper cost-benefit analysis of NCAs, which a case-by-case adjudication allows to do. There is also the problem of asymmetric enforcement because the US Department of Justice also has jurisdiction over competition policy, but only the FTC would enforce FTC rules.
What is, therefore, the way forward? Mandating increased transparency about NCAs before signing a contract would give better control to workers. This could be done under Section 19 of the FTC Act, which (unlike Section 6(g)) authorises the FTC to enact substantive rules to combat unfair or deceptive acts or practices. This would be better able to withstand judicial scrutiny. Businesses themselves could take the cue and restrict NCAs only to specific sections of the workforce and certain situations, instead of having an overarching NCA, and that too in a very transparent manner. Laws for the protection of trade secrets are a better way to handle the problems related to innovation. Case-by-case adjudication under state laws is better than a one-size-fits-all approach.
As is always the case with public policy, a more nuanced, middle-way approach is what is needed.
Admissions are now open for Takshashila’s Postgraduate Programme in Public Policy. This comprehensive, multidisciplinary, and entirely online 48-week program is tailored for those aiming to delve deep into both the theoretical and practical aspects of public policy.
What We're Reading (or Listening to)
[Takshashila Blog] China’s Electronics Domestic Substitution Mandate Is a Sign of Weakness, by Pranay Kotasthane and Amit Kumar
[Opinion] China’s Military-Civil Fusion Space Program, by Rakshith Shetty and Ashwin Prasad
[Article] Huawei’s New Phone Runs Latest Version of Made-in-China Chip, by Vlad Savov