SteelOnTheNet Insights Podcast

Introduction (00:00–01:52)

Welcome to SteelOnTheNet dot com Podcasts. I'm Dr. Andzhey Kotas, and today is the fifth of August, twenty twenty-six.

Today we're doing something a little different. Instead of one deep dive, we're busting some big steel myths — three of the most persistent fallacies I hear repeated by politicians, by development economists, and sometimes by people inside the industry itself. Specifically: that steel in general is a strategic material, that developing economies need their own steel industry, and that G7 countries need to preserve virgin, iron-ore-based steelmaking.

These aren't fringe views. They shape billion-dollar decisions — trade tariffs, national steel strategies, and investment programmes in developing economies. And in each case, the underlying assumption doesn't survive close scrutiny.

And these aren't costless mistakes in a vacuum, either. The global steel industry is already sitting on a serious capacity glut — the OECD's latest Steel Outlook puts global excess capacity at six hundred forty million tonnes in twenty twenty-five, projected to reach seven hundred forty-five million tonnes by twenty twenty-eight, more than three hundred million tonnes above total OECD steel production. Every one of today's fallacies, acted on, adds to that glut or keeps uncompetitive capacity alive longer than it should. That's the backdrop the whole discussion sits against.

So let's get into it. Three fallacies, each backed by its own detailed analysis on the site. Let's take them one at a time.

Fallacy One: Steel Is a Strategic Material (01:52–08:05)

The first fallacy is the broadest, and probably the most politically powerful: the claim that steel is a strategic material, and that therefore a country must maintain its own steel production capacity.

Now, on the surface this sounds reasonable. Steel goes into tanks, into warships, into nuclear reactors. Surely a country can't be dependent on foreign suppliers for something so fundamental to national security?

Here's the problem. When you actually break down global steel demand by application, the genuinely strategic slice is tiny. In our analysis, perhaps one percent — or less — of global steel consumption goes into applications that meet a rigorous definition of strategic: severe supply concentration among fewer than ten qualified producers, genuine defence or critical-infrastructure essentiality, long capacity ramp-up times, and no viable substitute.

Products that actually meet all four tests are things like naval armour plate, nuclear-grade reactor forgings, and premium seamless tube for gun barrels and aircraft landing gear. Naval armour plate globally totals perhaps one hundred fifty to two hundred thousand tonnes a year. Reactor pressure vessel forgings are measured in the tens of thousands of tonnes. Against global steel production of around one point seven billion tonnes, this is a rounding error.

The other ninety-nine percent — hot rolled coil, rebar, structural sections, galvanised sheet — trades in deep, liquid, competitive global markets. Over a thousand facilities worldwide produce hot rolled coil alone. There is no meaningful supply concentration, no adversary control, and readily available substitutes in most applications. Aluminium increasingly substitutes for steel in automotive and packaging; composites replace it in aerospace; concrete substitutes for structural steel in most construction. None of that substitutability exists for naval armour or nuclear forgings — which is exactly the point. The products that genuinely fail the substitution test are a tiny, identifiable list, not the sector as a whole.

The clearest illustration is British Steel's Scunthorpe works, which has drawn well over five hundred million pounds in government support and loan guarantees, on top of the UK government taking direct control of steelmaking there under special measures in April twenty twenty-five, before fully nationalising the business in July twenty twenty-six, to keep the blast furnaces running. Most of what Scunthorpe produces — hot rolled coil, plate, standard structural sections — is commodity carbon steel available from dozens of European competitors, and none of it needs a rigorous strategic designation to justify import substitution.

Rail is the one genuine complication worth flagging rather than glossing over: Scunthorpe supplies the large majority of Britain's network rail, and rail steel genuinely needs low residual content — particularly low copper — for weldability and fatigue performance, traditionally a strength of virgin steelmaking rather than standard scrap-based production. But that's an argument for protecting access to clean, low-residual semi-finished steel for rail production — not for preserving the whole site's blast furnaces and commodity coil output to get there. We'll come back to why that distinction holds up under fallacy three.

Contrast that narrower, genuine case with Sheffield Forgemasters, which the UK government acquired for just two point five six million pounds, plus planned investment in forging capacity. That facility holds genuine nuclear-grade forging capability — one of fewer than six comparable facilities in the world, meeting all four strategic tests outright. Rail sits somewhere in between; commodity coil and plate don't come close.

There's also a metallurgical pattern worth noting, because it explains why the one-percent boundary isn't arbitrary. Genuinely strategic products are almost always alloy steels — nickel-chromium-molybdenum armour plate, manganese-molybdenum-nickel nuclear forgings — requiring specialised metallurgical capability that only a handful of producers hold. Commodity products claiming strategic status, by contrast, are plain carbon steels made by processes that are essentially ubiquitous worldwide. That distinction — alloy versus plain carbon — is a far better guide to genuine strategic value than the word "steel" on its own.

The lesson: treating all steel as equally strategic doesn't protect national security. It just makes protectionism politically easier to sell. Smart policy would ask a much narrower question — do we possess the specific alloy metallurgy capabilities, in armour plate, nuclear forgings, and premium seamless tube, that actually matter? — rather than defending an entire commodity sector under a strategic banner it doesn't deserve.

Fallacy Two: Developing Economies Need Their Own Steel Industry (08:05–14:18)

Let's turn to fallacy number two. It's closely related, but aimed at a different audience: the idea that developing economies need domestic steel production to industrialise — that without a steel industry, a country will somehow be left behind, unable to build infrastructure, create jobs, or protect its balance of payments.

This is deeply embedded in development economics, going back to postwar reconstruction and the Soviet industrialisation model. But those historical precedents came from a world with restricted trade, limited capital flows, and genuine geopolitical necessity for self-sufficiency. That world doesn't exist anymore.

Consider the capital involved. Building one hundred million tonnes of integrated steel capacity costs roughly one hundred ten billion pounds. That same capital could instead fund comprehensive fibre-optic networks for two hundred million people, around seventy-five gigawatts of solar generation, thirty thousand kilometres of high-speed rail, or fifteen years of higher-education funding for ten million students annually.

And the return comparisons are not close. Steel investment in developing economies typically generates real returns of four to six percent. Renewable energy generates twelve to fifteen percent. Digital infrastructure, fifteen to twenty percent. Education, twenty to twenty-five percent.

The jobs argument is weaker than it appears too. A typical five-million-tonne integrated steel facility employs three to four thousand workers directly — somewhere around one point five million pounds of capital investment per direct job created, among the highest capital-to-labour ratios in manufacturing. Renewable energy creates fifteen to twenty jobs per million dollars invested. Digital infrastructure, twenty to twenty-five. Labour-intensive manufacturing, thirty to fifty. The same capital deployed elsewhere could create between two and seven times more employment.

And on strategic necessity — we've already covered that one. Ninety-nine percent of steel demand is commodity grade, available on competitive international markets. The genuinely strategic one percent can be secured far more cheaply through targeted stockpiling and dual-source supply contracts than by building an entire steel industry. Singapore and Israel both maintain strategic stockpiles without operating an integrated steel sector at all.

Singapore is really the proof point here. It went from a developing economy to per-capita G-D-P above forty-eight thousand pounds without ever building steel production capacity, choosing instead to specialise in services, electronics, and petrochemicals. Finland shut down most of its integrated capacity and focused on specialty stainless steel, where it holds a genuine technological edge. Denmark never built significant capacity at all.

It's also worth naming why this fallacy persists despite the numbers, because the reasons aren't entirely economic. Steel carries symbolic weight — a blast furnace inauguration photographs better than a software park or a logistics hub, even when the software park delivers far superior returns. That prestige factor helps explain projects like Nigeria's Ajaokuta Steel Company, which consumed nearly six billion pounds and never reached commercial production.

India is a genuinely harder case than either of those, and it's worth being precise about why, because the usual prestige critique doesn't really apply here. The underlying demand is real: per-capita steel consumption is still only around one hundred kilograms, against a global average closer to two hundred twenty, and consumption has been growing at roughly eight percent a year.

India also doesn't have the deep scrap pool that makes the E-A-F route viable at scale in mature economies — its capital stock is still young, so it's a net scrap importer, not a scrap-rich economy like the G7 countries we'll come to shortly. Add substantial domestic iron ore reserves, and there's a genuine resource-based case for continued ore-based capacity that simply doesn't exist in, say, the UK or Germany.

So this isn't Ajaokuta. But the capital-return and employment arguments still apply regardless of how sound the underlying rationale is. Around seventy-five billion pounds has gone into capacity over two decades, reaching one hundred forty-two million tonnes by twenty twenty-four — the world's second-largest producer — and the sector still needs meaningful tariff protection to stay viable, still employs only around six hundred thousand workers directly against a workforce of four hundred fifty million, at a capital cost exceeding one hundred twenty thousand pounds per job.

Even the government's own target of three hundred million tonnes by twenty thirty is expected to be undershot, while demand keeps outrunning supply regardless. India shows that even a well-founded rationale — real demand, genuine scrap scarcity, genuine ore endowment — doesn't resolve the capital-allocation question on its own. The return on capital and the jobs created per pound still point the same direction as everywhere else in this fallacy: steel is capital-intensive, not labour-intensive, whatever the resource fundamentals underneath it.

None of this means steel investment is always wrong — Indonesia's nickel-based stainless steel sector shows that genuine comparative advantage can justify targeted investment. But the blanket assumption that industrialisation requires a domestic steel industry doesn't hold up. The real question a government should ask isn't "steel or no steel" — it's "steel, or what else could this capital achieve?"

Fallacy Three: G7 Countries Need Virgin Steelmaking (14:18–20:31)

That brings us to fallacy number three, and it moves us to the advanced economies: the claim that G7 nations need to preserve iron-ore-based, or virgin, steel production, because it produces cleaner, higher-quality steel than scrap-based electric arc furnace — or E-A-F — production, which introduces tramp elements like copper.

There's a kernel of truth buried in here, which is exactly what makes this fallacy durable. But the scale is wildly overstated.

Start with scrap availability. G7 countries collectively generate somewhere between one hundred fifty-one and one hundred sixty-seven million tonnes of scrap annually against consumption of two hundred fifty to two hundred eighty million tonnes — a self-sufficiency rate of sixty to sixty-five percent, achievable through E-A-F production, with the remainder covered by imports from allied trading partners. Italy already runs a predominantly E-A-F sector, at fifty-five to sixty percent self-sufficiency, and its producers are consistently profitable without subsidy.

Now, the quality argument. Yes, some applications genuinely need low-residual steel that's harder to achieve from scrap alone — automotive exposed panels, electrical steel for transformers and motors, deep-drawing tinplate for food cans, and, as we touched on earlier, rail. But combined, these applications total under three percent of G7 steel consumption.

The other ninety-seven percent — construction products, standard automotive components, general engineering — tolerates E-A-F metallurgy without difficulty. And even for that residual-sensitive three percent, direct reduced iron, blended with scrap in an E-A-F, can hit the required specifications without a blast furnace at all.

Rail is actually a clean illustration of this, and not just in theory — France has already done it. SNCF's own strategic domestic rail supplier, the Hayange mill in Moselle, historically rolled blast-furnace blooms shipped over from British Steel's Scunthorpe works. Since twenty twenty-one, under a "GREENSTEEL" programme approved directly by SNCF Réseau, Hayange has instead sourced its blooms from the Ascoval electric arc furnace mill in northern France — and by twenty twenty-three it was reportedly the only rail producer in Europe running entirely on E-A-F steel.

France's own national rail operator tested it, certified it, and switched. That's not a hypothetical DRI-EAF pathway — it's a live case of exactly the transition this argument depends on, already running under one of the most safety-conscious procurement regimes in the industry.

The economics tell the same story. Blast furnace operations in Europe and the UK face structural cost disadvantages of one hundred to two hundred pounds or more per tonne against Asian integrated mills, driven by carbon pricing, energy costs, and labour costs.

Meanwhile E-A-F producers post real profits. Nucor, in the United States, generated one point six billion pounds of net income in twenty twenty-three. Steel Dynamics generated one point four billion.

Over the past decade, those two companies posted profits in nine or ten years out of ten. Their integrated competitors — US Steel and Cleveland-Cliffs — managed profitability in only six or seven of those ten years, despite tariff protection.

Japan is the genuine exception, and it's worth naming why, because it proves the rule rather than breaking it. Japanese integrated steelmaking is profitable through export manufacturing integration, superior blast furnace productivity, and Pacific Basin market access — location-specific and firm-specific advantages that simply don't transfer to Scunthorpe, or to Dunkirk, or to Gary, Indiana.

It's worth being specific about defence consumption too, because this is where the fallacy does the most political work. Across G7 economies, defence steel consumption sits well under one percent of total consumption — the United States, with the world's largest defence budget, consumes something like eight hundred thousand to one million tonnes a year against total consumption of ninety-five to one hundred million tonnes. And of that defence total, sixty to seventy percent is standard structural sections, commodity plate, and fasteners with no special requirements at all — perfectly well served by E-A-F production. Only a genuinely small slice, under five percent of defence steel, needs the specialty alloy capability we discussed under fallacy one.

And the genuinely strategic alloy capabilities — nuclear forgings, naval armour, aerospace alloys — these don't come from blast furnaces at all. They come from specialised, independent operations: Sheffield Forgemasters, Framatome Le Creusot, Japan Steel Works. Protecting those costs on the order of fifty to one hundred fifty million pounds a year across a G7 economy — a fraction of the five hundred million to eight hundred fifty million pounds or more it costs to keep a single uncompetitive blast furnace running. Conflating the two — commodity blast furnace output and genuine strategic alloy capability — is what keeps this fallacy alive.

Conclusion (20:31–22:23)

So, three fallacies, and in each case the underlying claim usually survives on assumption rather than evidence. Steel is not, in the main, a strategic material — ninety-nine percent of it trades like any other commodity, and the genuinely strategic sliver deserves far more focused protection than it typically gets. Developing economies don't need their own steel industry to industrialise — the capital almost always earns a better return elsewhere. And G7 countries don't need to preserve virgin steelmaking to maintain quality or security — scrap-based production, backed by imports and targeted DRI-EAF capacity, covers the overwhelming majority of demand.

There's a sharper lesson worth drawing out from Hayange specifically. Rail steel genuinely needs low residuals — that requirement was never in question. What was in question was whether a blast furnace was the only route to meeting it, and nobody had actually tested that until SNCF Réseau did. The alternative turned out to work, and turned out to be cheaper.

That's the discipline this podcast keeps coming back to across all three fallacies: even where the underlying requirement is genuine, it's worth checking whether the assumed solution is really the only one — and even where the underlying rationale is sound, it's worth checking whether the numbers still stack up once you actually do the sums.

None of this happens in a vacuum, either. The industry is already carrying the capacity overhang we opened with — six hundred forty million tonnes of excess capacity today, heading toward seven hundred forty-five million by twenty twenty-eight on the OECD's own numbers. Every fallacy we've covered today, acted on, makes that worse: subsidising commodity capacity that never needed protecting, building steel industries where the capital would do more good elsewhere, propping up blast furnaces that scrap-based production could replace. Getting these arguments right isn't just good capital allocation for the country making the call — it's what the global industry needs to work its way through its own overcapacity problem.

If you're a government, an investor, or a steel producer working through one of these arguments in a live decision, that's exactly the kind of analysis we do. Get in touch through SteelOnTheNet dot com.

This has been a SteelOnTheNet Podcast. I'm Dr Andrzej M Kotas. For full show notes and related resources, visit steelonthenet.com/insights/podcasts. Thank you for listening.