Buying British Is Not Protectionism: It’s an Industrial Policy Decision
The UK government has recently formalized a policy that, according to The Guardian, will prioritize British suppliers for contracts in sectors deemed vital for national security: shipbuilding, steel, artificial intelligence, and energy infrastructure. The measure was welcomed by UK Steel as recognition that steel is a "National Strategic Asset." Meanwhile, The Independent labeled it as a direct subsidy for technological mediocrity. Both viewpoints capture part of the issue, but neither gets to the heart of the mechanics that truly matter.
The question isn’t whether buying from local suppliers is good or bad for national pride. The real question is what this policy does to the state’s cost structure, the competitiveness of the benefiting sectors, and the distribution of value among all players in those supply chains. That’s what I aim to dissect.
The State as a Demand Anchor and Its Impact on Pricing
When a government guarantees contracts to domestic suppliers relatively independent of market prices, it acts as a captive buyer. In basic negotiation theory, a buyer who cannot walk away loses bargaining power. The British state, by declaring that it will prioritize local suppliers in national security sectors, is signaling that its willingness to pay becomes less elastic in those specific markets.
This has a direct consequence: the favored domestic suppliers experience less structural pressure to reduce costs or improve quality. UK Steel may celebrate this political recognition, but if the contract comes regardless of whether its steel costs more than that from South Korea or Turkey, the incentive to invest in productive efficiency weakens. This is not a moral accusation; it’s the mechanics of protected markets.
What makes a preferential public procurement policy sustainable is not the decree itself, but what happens during the protection period. If the prioritized sectors use the demand guarantee to scale capacity, invest in automation, and reduce their cost structure to become genuinely competitive within five or ten years, the state will have bought productive time. Conversely, if these sectors use it to maintain margins without reinvestment, the state will have financed temporary stability at the expense of permanent competitiveness.
The distinction between these scenarios does not depend on the goodwill of industrial executives. It hinges on whether the policy includes binding performance metrics, timelines for price reductions, and exit clauses to reintroduce external competition once sectors hit efficiency thresholds. Without these mechanisms, the policy is a blank check.
Steel and AI in the Same Decree: Two Incompatible Industrial Logics
The most revealing detail of this policy is that it groups together steel, shipbuilding, artificial intelligence, and energy under the same category. Superficially, they all share the umbrella of "national security." However, beneath that label, they operate under radically different productive logics, and applying the same protection model to them is a decision fraught with asymmetric risks.
Steel and shipbuilding are sectors with high entry barriers, capital-intensive physical infrastructure, and decades-long investment cycles. Protecting their local demand makes clear strategic sense: dismantling a steel plant or shipyard is not reversible in the short term, and relying on external suppliers to build frigates or maintain port infrastructure represents a genuine operational vulnerability in an unstable geopolitical context.
Artificial intelligence, in contrast, operates very differently. It is a sector where rapid iteration speed and exposure to global competition are the only forces that maintain product quality. A language model, a cybersecurity tool, or a critical infrastructure analysis system that only competes for British government contracts has a very different set of incentives from a company that must survive in open markets. The Independent pointed out that this amounts to subsidizing technological mediocrity. The criticism is empirically grounded: technology providers living off protected public procurement tend to optimize for the bidding process, not for product quality.
Putting steel and AI under the same protection umbrella combines two problems that require different solutions. One requires demand stability to maintain long-term physical assets, while the other requires competitive pressure to prevent stagnation. Treating them equally sacrifices technological efficiency in the name of a unitary political framework.
The Supply Chain No One Mentioned
A player glaringly absent from the public debate generated by this policy is second- and third-tier suppliers. Those selling components to shipyards, those manufacturing inputs consumed by the steel industry, and those developing the infrastructure layers on which government AI systems run.
A first-tier prioritization policy does not guarantee anything about how the primary beneficiary distributes that value downstream in its own chain. A shipyard that receives guaranteed contracts has incentives to improve its own margins. It can achieve this by investing in efficiency, but it can also do so by squeezing its own component suppliers, who lack the same political protection and are exposed to import competition.
This is not a hypothetical scenario. It reflects the historical pattern of many protected industrialization programs: the first link in the chain strengthens, intermediate links compress, and the employment and value generated at lower levels do not grow proportionally to the contracts coming from above. If the political goal is to systematically bolster British industrial capacity, the policy needs mechanisms that trace the flow of value throughout the chain, not just to the supplier signing the contract with the government.
Well-Designed Protectionism Has an Expiration Date
The difference between an industrial policy that works and one that fossilizes entire sectors does not hinge on whether it protects or not. It lies in whether the protection has exit conditions.
South Korea built its shipbuilding industry with subsidies and state contracts during the 1970s and 80s, but these were incorporated within a model where firms had to meet rising export targets to retain access to benefits. Protection was temporarily designed. Germany supports its automotive industry with combinations of regulation, public investment in infrastructure, and technical standards that raise competitive bars for everyone. The state buys quality, not just origin.
The British policy, as described in its official sources, does not specify the exit mechanisms or performance standards that will condition the continuity of preferential access. This makes it, for now, a statement of intent with incomplete architecture. It could evolve into a sophisticated industrial policy model or remain a short-term political stabilization instrument that placates unions and sectoral lobbies without transforming underlying competitiveness.
A state that pays above-market prices without demanding performance improvement is not investing in national security. It is transferring value from taxpayers to shareholders of protected companies. Conversely, a state that pays market prices or slightly above while requiring metrics for efficiency, investment in R&D, and cost reductions over time is purchasing durable industrial capacity. The difference between the two is not ideological; it’s contractual design.
The sectors granted this protection gain an advantage today that they did not have to compete for. The only way for that advantage not to become a long-term trap is for them to use it to build a value proposition that can survive without it. Those who achieve this will have transformed a government contract into a genuine industrial base. Those who fail to do so will have found in the state the customer that allowed them to avoid precisely the pressure that would have made them better.
