Oliver Blume floated going over his own board. A 1960 statute and a shareholder register cap his side of the vote at exactly the number he needs to beat.
Volkswagen’s supervisory board meets on Friday, September 4, to rule on a restructuring plan that implies roughly 50,000 more job cuts on top of the 50,000 already agreed. Blume has let it be known he could bypass the board and take the plan straight to shareholders. That exit is close to sealed. Porsche SE, Qatar Holding and every single free-float ordinary share add up to exactly 80.0% of Volkswagen’s votes, and the Volkswagen Act demands more than 80% of the capital represented. Lower Saxony’s 20.0% is not a lobbying position. It is a rounding error away from an absolute veto, and it has been sitting there since 1960.
Everyone is covering this as Germany versus China. Chinese brands took VW’s best overseas market and are now selling into Wolfsburg’s back yard. Tariffs cost the group billions. The German press has settled on “Detroit-ization” as its word of the year.
All true, and all of it explains why Volkswagen needs to change. None of it explains why Volkswagen can’t.
The reason is on the shareholder register, and it turns a restructuring plan into something closer to a purchase negotiation. Blume does not get to decide. He gets to bid.
What Happened
On July 9, Blume put a cost plan to the supervisory board and watched it lose. The vote broke exactly along the lines the structure predicts: ten seats tied to the unions against, plus the two seats belonging to the state of Lower Saxony. Seven votes in favor, all from representatives of the Porsche-Piech family holding company and Qatar’s sovereign wealth fund.
Ten weeks later, Blume spent last week touring VW plants in a roadshow that looked more like a regional election campaign. About 10,000 employees packed a hall in Wolfsburg to hear him defend the plan in front of Daniela Cavallo, who runs the group works council. Placards told him their jobs were not his balance-sheet adjustment. His opening pleasantries got booed.
Olaf Lies, premier of Lower Saxony and one of the board members who voted the plan down, ran his own counter-tour of the factories Blume has hinted may not be needed.
Blume’s message at Emden and Zwickau was blunt. German labor costs run more than double comparable European locations, and even after roughly 20% of cost came out of those plants last year, other plants are still cheaper. He also called closing a plant “always the last and most expensive solution,” which is the sentence of a man who knows what closure costs to buy.
An executive committee meets Thursday, September 3, to hunt for a compromise before the full board sits down on Friday.
The Backstory
The Volkswagen Act of July 21, 1960 was written to privatize a company the state owned, without letting anyone actually take it over. It worked. The European Commission fought it, the European Court of Justice struck down parts of it in 2007, and Germany amended it in 2008 to delete the 20% voting cap and the federal veto.
The one provision Germany kept is the one that matters this week. Under section 4(3), any general meeting resolution that German corporate law would normally pass with a 75% majority instead requires more than four-fifths of the share capital represented at Volkswagen. Written into the articles of association at Article 25(2). Never litigated in this context.
Then there is the share structure, which almost nobody outside Germany reads properly. Volkswagen has 295,089,818 ordinary shares that vote and 206,205,445 preference shares that do not. The preference share is VOW3, the liquid one, the one in the index funds, the one whose ticker sits in every story about this fight. It has no say in any of it.
Ownership of the votes, per the group’s own 2025 annual report: Porsche SE 53.3%, the State of Lower Saxony 20.0%, Qatar Holding 17.0%, free float 9.7%. Lower Saxony also gets to appoint two supervisory board members outright for as long as it holds at least 15% of the ordinary shares. That is a statutory appointment right, not an election result.
The precedent for how this plays out is only two years old. In late 2024 management canceled a jobs guarantee that had been treated as sacred and said VW would have to close a German factory. Four months of negotiation later, the unions had agreed to 50,000 German job cuts across Volkswagen, Audi, Porsche and the CARIAD software unit by 2030, in exchange for a fresh guarantee for everyone left. Every departure voluntary. Every one spread over five years. Blume says about 37,000 of those agreements are already signed.
That is the reference price. Fifty thousand heads, four months, and a promise not to come back for more.
The Plan
Blume came back for more.
The actual target is not a headcount, which is why the 100,000 figure he has publicly pushed back on keeps causing confusion. The target is an overhead ratio. Under the group’s 2030 plan, VW wants around 11 billion euros out of global overhead, taking the automotive overhead cost ratio from roughly 16% today to about 12%. Four points of overhead, four points of margin.
The 50,000 falls out of that as arithmetic, not ambition. Personnel is about 60% of overhead. Close the gap to competitors and that is what the spreadsheet returns. Blume has called it a theoretical deduction rather than an approved plan, and on the numbers he is right. He also said it out loud on the July 24 earnings call, to analysts, where it landed as a cost program. The same figure reaching the works council landed as a threat.
Everything else in the plan follows the same logic of subtraction. Halve a lineup of roughly 150 models across the brands, which would still leave VW with more nameplates than any peer. Prune a portfolio of some 2,000 investments that somehow includes stakes in four German soccer clubs. Take out more than 500,000 vehicles a year of excess European capacity, with Emden, Hanover, Zwickau and Neckarsulm named as sites without secured utilization into the 2030s.
And the sharp end: about half the cost reduction has to come from Germany, and it is aimed at desks rather than lines. Four in five of the roughly 60,000 people in Wolfsburg work in offices. Management, group functions, development, sales. In a city of about 130,000 people that VW effectively built, owns the soccer club in, and funds through the tax base.
The numbers behind the urgency are not soft. First-half 2026 revenue held at 158.1 billion euros, but operating profit fell 12% to 5.9 billion and the margin came in at 3.8%, or 4.3% stripping out special items. Deliveries dropped 6.3% to 4.13 million. China fell 26% in the half and 37% in the second quarter alone. Full-year 2025 was worse in one respect that should terrify anyone: 322 billion euros of revenue converted into 8.9 billion of operating profit, a 2.8% margin, on close to 9 million cars. The whole company is currently worth about 37 billion euros on the market. It sells more than eight times that in a year.
The 80% Wall
Now run the vote.
If the supervisory board refuses consent, German stock corporation law does give the management board a route: it can demand that shareholders resolve on the consent instead. That is the “audacious workaround” being briefed to reporters, and it is real. It is also, at Volkswagen, close to arithmetically foreclosed.
Start with who can vote yes. Porsche SE brings 53.3%. Add Qatar and you reach 70.3%. Add every last share in the free float, assuming all of it turns up and every holder sides with management, and you land on 80.0%.
The statute requires more than 80%.
Not 80%. More than it. A coalition of every shareholder in the company except Lower Saxony hits the number and does not clear it. That is what section 4(3) was designed to do, and it is why the European Commission spent years arguing it was a restriction on capital movement rather than a quirk of drafting.
Turnout makes it worse, not better. The threshold runs on capital represented at the meeting, not capital outstanding. If free-float holders skip the vote, Lower Saxony’s fixed 20.0% becomes a larger share of the capital in the room, and the yes side falls below 80% rather than reaching it. The only scenario where the arithmetic opens up is one where Lower Saxony chooses not to show up, which is roughly as likely as Wolfsburg voting to close Wolfsburg.
There is a second lock behind the first. Under German law the general meeting decides management questions only when the management board asks it to, and it can only say yes or no to a specific resolution. It is not a forum for approving a strategy. So the workaround only reaches items the supervisory board must formally consent to, one at a time, each carrying its own vote and its own veto.
Which reframes the whole standoff. Blume is not choosing between negotiating and forcing. He is choosing between negotiating and losing publicly. The threat has value precisely because he will never test it, in the same way a bidder’s walk-away price only works before anyone checks whether he can actually walk.
Compare it to how governments are buying influence in American deals right now, where Washington has been taking equity and golden shares one transaction at a time. Lower Saxony did not have to negotiate for any of that. It was legislated in 1960 and it renews itself for free every year.
The Business Model Angle
Control rights are a cost line, and they should be priced like one. Volkswagen’s ~3.8% operating margin is not purely a competitiveness number. Part of it is the standing cost of needing supermajority consent to remove cost. Porsche SE, the entity that holds 53.3% of the votes at a business with 322 billion euros of revenue, is itself worth under 9 billion euros on the market. That gap is what the market charges for control it does not believe can be exercised. If your cap table contains a party who can block but never has to pay for blocking, model that as a permanent margin haircut, not as a governance footnote.
Restructuring at a co-determined firm is procurement, not a decision. VW’s 2024 round set an observable price: 50,000 heads, four months, all voluntary, five-year runway, plus a renewed guarantee for the survivors. That is a purchase contract with a delivery schedule. Anyone modeling European industrial restructuring should stop forecasting “will they cut” and start forecasting the clearing price and the lead time, because the volume is negotiable and the timing almost never is. Speed is the thing you cannot buy.
The shares that trade are not the shares that decide. Every institution holding VOW3 owns the economics of this fight and none of the votes. That split shows up wherever founders, families, states or dual-class structures separate cash flow from control, and it is why activism is cheap to launch and expensive to win in those names. Before underwriting a turnaround thesis, check whether the security you are buying can actually vote for the turnaround.
The Risk
The obvious counter is that this structure is not a bug someone forgot to fix. It is the reason Volkswagen still exists.
Co-determination is what allowed VW to take 50,000 people out of Germany without a single forced layoff, without a strike that stopped production, and without the years of litigation an American company would have booked in the same move. Cavallo’s line at the town hall was not sentimentality. “We, too, have a right to the returns we make, like a dividend for the workforce” is a claim on residual value, made by a counterparty who funded the original plant. Wolfsburg’s 1930s factory was built partly with money seized from unions after Hitler banned them, and the works council has made that the moral basis of its claim for eighty years. It is a stronger argument than most CEOs would like it to be.
There is also a data point that cuts hard against the structural thesis. Under exactly the same governance, the same board and the same union, the Porsche brand group lifted first-half operating profit 45% to 1.2 billion euros and pushed its margin from 5.2% to 8.0%, with volume down 11%. Governance did not stop that. So the honest version of the argument is narrower than “co-determination is the problem.” It is that co-determination sets a speed limit, and the speed limit is fine when your product is a 911 and murder when Chinese exporters are relocating a domestic glut into your home market while BYD’s cost base is structurally below yours.
And the steelman for the union has one more leg. Blume’s own plan concedes that closing a plant is the most expensive option. If the fixed cost is closure charges and the variable win is a few points of overhead, the works council’s demand for a product strategy before a headcount number is not obstruction. It is asking whether the plants would be full under a better plan, which is the same question that separates margin by absorption from margin by subtraction everywhere else in the industry.
The counter-argument to all of it is timing. Speed limits are survivable when the road is straight. VW is absorbing a China collapse, a tariff regime that keeps re-pricing itself and a competitor set that ships product on eighteen-month cycles. The structure that made the last forty years stable is being asked to clear a decision in weeks.
Quick Questions
Can Blume actually be overruled by shareholders? Only on specific items the supervisory board must consent to, one vote at a time, and only if more than 80% of the capital represented agrees. With Lower Saxony voting no, the maximum possible yes vote is 80.0%.
Why does Lower Saxony have this much power on 20%? Because the 1960 Volkswagen Act raised the qualified-majority threshold from the standard 75% to more than 80%, which converts a 20% stake into a blocking minority. The state also appoints two board seats directly while it holds 15% or more of the ordinary shares.
Is 100,000 job cuts the real plan? No. About 50,000 are already agreed through 2030 and roughly 37,000 of those agreements are signed. The second 50,000 is a derived figure from closing an overhead gap of about four percentage points, not an approved program.
Does the free float have any leverage? Almost none. The 9.7% of ordinary shares in free float can only matter as the marginal votes in a coalition that still cannot clear the threshold. The much larger pool of preference shareholders has no vote at all.
What actually breaks the deadlock? A price. Either the works council accepts a bigger cut in exchange for a longer guarantee, or Lower Saxony trades votes for plant commitments. Both are transactions, which is why Blume spent last week campaigning rather than filing.
The Business Model Analyst Take
The most useful thing about the Volkswagen fight is that it prices something most companies never have to price: the cost of needing permission.
Every business has stakeholders who can slow it down. Very few have one whose consent is written into national law, renewed annually at zero cost, and immune to being bought, diluted or outvoted. Volkswagen’s is, and the market has done the math even if the coverage hasn’t. A company selling 322 billion euros of cars a year trades at roughly 37 billion, and the holding company that “controls” it is worth less than nine.
Blume is not fighting BYD this Friday. He is running a bid process against a counterparty who cannot lose, and hoping the words “extraordinary general meeting” move the reserve price a little. It is a good bluff. It is not a plan.
The lesson generalizes past Wolfsburg. When you build a business whose survival depends on being able to shrink quickly, check first whether anyone else holds a key to the door. Growth hides that question for decades. Contraction asks it in a single afternoon.
