Germany is losing industrial jobs: BASF employs fewer people than in 1954 and opens an 8.7 billion euro site in China

Thirty Thousand

In Ludwigshafen, in May 2026, full-time positions fell below the thirty-thousand mark. That had not happened since 1954, when West Germany was emerging from reconstruction and the site on the Rhine was becoming what it is today: the largest integrated chemical plant in the world, ten square kilometres, two hundred factories linked by pipes that hand intermediate products from one to the next ³. In the same year, seven thousand kilometres away, the same company inaugurated another site. Zhanjiang, Guangdong province, 26 March 2026: roughly €8.7 billion of investment, Verbund just as in Ludwigshafen, and a company statement flagging two things at once — ahead of schedule and under budget ⁴.

The two pieces of news are not the same fact told twice. They are the two ends of a chain that begins somewhere else, in a pipe that stopped carrying gas, and that along the way ran through chemicals, autos, the German trade balance and finally the state budget. This issue follows that chain link by link, and at each link it says which figure holds it up and which does not. Because one thing has to be said at the outset: the most counter-intuitive part of this story is not the decline. It is that while industrial capacity was being dismantled, the German stock market was hitting record highs. The two curves do not contradict each other. They measure different subjects — and working out which is half the job.

Twenty-Three Years in One Line

The temptation, faced with a crisis, is to make it begin on the day we noticed it. The data collected say something else, and say it precisely. The starting point is 2003. A series of labour-market reforms — the so-called Hartz reforms — changed how unemployment benefits and employment relationships worked. Over the following decade German unemployment fell from 11.3% in 2005 to 5.5% in 2012 ¹². It is the decade in which Germany became, in the European imagination, the machine that works.

Industrial production rose for ten years and peaked in 2018 ³. Then came the pandemic, and from there the curve never climbed back: in 2025 production remained between 10% and 14% below that level, the fourth consecutive year in negative territory, with autos more than twenty percentage points below the peak ³⁴. 2022 is the year in which two things changed at once. Russian gas, which in 2021 covered 55% of German natural-gas imports, disappeared within months ⁸. And the trade balance with China, in surplus for years, recorded a record €84 billion deficit ⁹.

In March 2025 the Bundestag approved, with 512 votes in favour, a constitutional reform that exempts military spending above 1% of GDP from the debt brake, and establishes a €500 billion special fund for infrastructure and climate ¹⁰. Then 2026: Zhanjiang in March, Ludwigshafen below thirty thousand in May. The sequence says something that none of the dates, taken alone, says: the decline in production begins in 2018. Before the gas, before the reversal with China. The energy shock accelerated a curve that was already bending; it did not start it.

Why Germany and Not Elsewhere

The same energy shock hit all of Europe. The damage, however, was not the same everywhere, and the reason lies in the composition of the economy. In 2024 manufacturing accounted for 19.9% of German gross value added. In Italy 16.6%, in Spain 11.9%, in France 10.7%; the European Union average is 15.9% ¹. Put differently: when the blow landed, Germany had almost twice what France had. Gross value added is the right measure because it counts not revenue but wealth actually created. And that German fifth is not spread at random: it is concentrated in the segments that consume the most energy per unit of output.

The bill arrives on time in the Destatis numbers. In 2025 output in energy-intensive industrial segments was 17.8% below 2021 ⁴. Not 17.8% of German industry: 17.8% of precisely the part that burns energy for a living. There is an alternative reading, and it comes from an institution with no stake in the dispute. The International Monetary Fund, in its annual consultation on Germany concluded in February 2026, points to labour costs — more than energy costs — as the main driver of the loss of price competitiveness, together with a marked real appreciation of the exchange rate ¹⁴. It is an objection that recurs, and one worth keeping in hand while following the chain.

The Chain, Link by Link

The first link is gas. In 2021, before the invasion of Ukraine, 55% of Germany's imported gas came from Russia ⁸. Over the course of 2022 that share collapsed. It has to be stated with the right yardstick: this is the share of imports, not of total demand, which also includes domestic production and other sources. The second link is basic chemicals. It is the sector that buys gas twice: as fuel for its plants and as feedstock for its molecules. Between 2022 and 2025, 37 million tonnes of production capacity in Europe were earmarked for closure, roughly 9% of the European total, with 20,000 direct jobs lost and a further 89,000 indirect jobs at risk ². The figure comes from Cefic, the European chemical industry association: it is the primary source for the number, and it is also an interested party in the debate over energy and regulation.

The third link is everything else. Basic chemicals do not sell to consumers: they sell inputs to everyone else. Engineering plastics, solvents, additives, paints, fertilisers, battery components. When the cost of energy enters the price of those inputs, it comes out in the costs of whoever buys them: autos, machinery, electronics, construction. That is why chemicals are not one sector among others in this story, but a junction. And here the BASF link forks, and it is the fork that tells the most precise thing in the whole issue. In Ludwigshafen full-time positions fall below thirty thousand for the first time since 1954 ³. In Zhanjiang an €8.7 billion Verbund site comes on stream ⁴. The capacity has not evaporated: it has moved.

The same movement shows up in Bundesbank data, and it does not concern China alone. German companies in chemicals and basic metals have expanded operations in the United States, plausibly to access lower energy costs, while inward direct investment into Germany has been in sharp decline since 2022 ¹⁶. Curiously, German direct investment in China has itself fallen for the second consecutive year, from €115 billion to €110 billion ¹⁷: the overall flow towards Beijing is thinning while individual bets, such as Zhanjiang, are getting bigger.

The Measured Effects

So much for the chain. Now the numbers that close it, and they are customs numbers: they record goods crossing a border, not forecasts. The reversal with China has a date and a size. The date is 2022, the year of the record €84 billion deficit ⁹ — four years ago, not a couple. The size reads better in the progression: in the first eight months of 2025 the German deficit with China rose by 142.8%, to $17.4 billion against $7.2 billion in the same period a year earlier ¹¹. The first half of 2026 adds the most useful figure of all. German exports to China fell by more than 12%, to just under €37 billion. Imports from China grew by 8.9%, to €91.8 billion. The balance moves from a €40 billion to a €55 billion deficit between the two first halves ⁶.

Those two percentages look almost symmetrical — twelve against nine — and they are nothing of the sort. Imports start from a base two and a half times larger than exports: one percentage point of imports weighs, in absolute terms, two and a half times a point of exports. That is the arithmetic that explains why the gap widens so much faster than the two numbers suggest. Within the aggregate, one segment breaks away. German car exports to China fell by 33% in 2025, to €13.6 billion, against almost €30 billion three years earlier. China has slipped from the second to the sixth largest export market for the German auto industry ⁷. Not a sector hit like the others: three times worse than the average.

On one point our verification stops, and it is better said here than at the end. A figure is circulating according to which German exports to China of machinery, electronics, energy infrastructure and industrial components fell by roughly 10% in 2025. The direction is borne out by the aggregate; the number attached to that specific basket is not. In the material collected there is no sectoral statistic separating electronics from machinery for that year. The disaggregated data are not public in that form — and that is information, not a gap to be papered over. The context has to be added, because it changes the meaning of the phrase "Germany is losing China." In 2025 total trade between the two countries came to €251.8 billion and China was once again Germany's largest trading partner, overtaking the United States ⁵. The volume has not collapsed: the direction has reversed.

What Didn't Propagate

Here the chain breaks, and the break is the most interesting fact in this issue. In 2025 German GDP grew by 0.1% according to the ifo forecasts of December; the International Monetary Fund puts it between +0.2% and +0.3% ¹³. The fourth quarter shows +0.3% quarter on quarter, with unemployment steady at 6.3% and employment down by 25,000 ¹⁵. These are stagnation numbers, not collapse numbers. Over the same period the DAX hit record highs. The contradiction is only apparent, and it dissolves with one figure: roughly 80% of the revenues of the index's companies come from outside Germany, with the US market at 24% ¹⁸. The DAX measures the accounts of multinationals whose production is already spread around the world, not the capacity installed on German soil.

The sharpest proof of this reading sits one floor below. The MDAX, the mid-cap index, more exposed to the domestic market, remains more than 9% below its own record high ¹⁹. Same economy, same year, two indices telling opposite stories because they are looking at different subjects. The last piece is the public budget. On 18 March 2025 the Bundestag approved the debt-brake reform with 512 votes in favour; the Bundesrat gave its green light on the 21st ¹⁰²⁰. Outside the constraint: military spending above 1% of GDP, and a €500 billion fund for infrastructure and climate. One point of chronology is worth making: at the time of the vote Friedrich Merz was not yet chancellor — he had led the negotiation that produced the two-thirds majority, and took office after the coalition with the SPD was formed.

What the Institutions Say

Three institutions have put a diagnosis in writing, and they do not agree. The Bundesbank attributes roughly three quarters of the loss of export market share recorded in recent years to reduced competitiveness among German firms ²¹. This is not a loss of world demand: it is a loss of position within that demand. The OECD, in its 2025 survey of Germany, observes that since the mid-2010s the slowdown in labour productivity is increasingly explained by the decline in total factor productivity, and forecasts energy prices persistently above pre-crisis levels ²².

The International Monetary Fund is the one that breaks away most. In its February 2026 consultation it identifies labour costs, more than energy, as the primary driver of the loss of price competitiveness, and flags a real appreciation of the exchange rate; German unit labour costs have deteriorated relative to France and Spain ¹⁴. A working paper from June 2026 quantifies it: roughly 60% of the growth shortfall is explained by lower potential, roughly 40% by cyclical factors ²³. That 40% is the number that keeps the question open. Eroded potential does not come back when the cycle turns; a cyclical component does.

Among the theses circulating about the German case, some do not rest on verifiable measurements and must be recorded as the judgements of those who express them. One analyst interviewed argues that Europe risks becoming "a large open-air museum" if it does not act now. Another states that in the less competitive industrial segments real wage growth will be nil or very weak in the coming years. A third describes a vicious circle in which difficulty in reforming, falling growth, lost tax revenue and lost public confidence feed one another. Friedrich Merz, for his part, has observed that China builds infrastructure in a few months where the Union needs years. These are assessments, not measurements: none of the data collected confirms or refutes them.

Opinion reported as such, not verified by the editorial team.

From the Notebook

1954 marks a beginning, not just an end. The year Ludwigshafen passed thirty thousand full-time positions is the same year the site was completing its postwar reconstruction. That the threshold is crossed again, downwards, seventy-two years later is not a symbol chosen by anyone: it is the company's own accounting, picked up by the local press in Mannheim, confronting a number it had not recorded in two generations ³. And at the end of 2025 the count was still around 32,190: the threshold fell in 2026, not before.

Heads or positions. The thirty-thousand figure does not count people, it counts full-time positions. These are two different measures — part-timers, mixed contracts and apprenticeships are in the first and not in the second — and the difference is not cosmetic when you are building a seventy-year time series. The source specifies Vollzeitstellen: full-time positions. Anyone who translates it as "employees" is changing the unit of measurement without saying so.

Zhanjiang finished early. In its statement of 26 March 2026 BASF notes that the site was completed ahead of schedule and under budget ⁴. It is a detail that goes almost unnoticed in the European conversation about costs, and it concerns a variable that gets little discussion: not what it costs to produce, but what it costs to build the plant in which production will then take place.

The disinvestment paradox. While BASF opens in Zhanjiang, German direct investment in China falls for the second consecutive year, from €115 billion to €110 billion ¹⁷. The overall flow thins while individual stakes swell: fewer bets, each one larger. That is not the same thing as "German industry is relocating to China," and it is harder to sum up in a headline.

The peak that moves. Almost every source uses 2018 as the high point of industrial production. One does not: a 2026 analysis places it in 2017 and observes that the slowdown was already under way before the pandemic, adding that German output today is at the same level as in 2005 ²⁴. A year's difference seems small, but it shifts the boundary between "shock-driven crisis" and "decline already in motion."

The dispute that has lasted twenty years. On the 2003 reforms the literature has reached no agreement. One reading judges them "clearly positive" in terms of economic efficiency, with the unemployment numbers as support ¹². A study published in 2019 reaches the opposite conclusion on causation: the measures are said to have shortened the typical duration of unemployment without reducing its overall level, with wages falling by around 4%, and wage moderation is said to have begun in 1995 — eight years before the reforms ²⁵. What is missing is an up-to-date meta-analysis reconciling the two strands: on recent Europe's most debated reform, the verdict is still open.

The chart we could not draw. Of the 37 million tonnes of chemical capacity closed in Europe, it would be useful to know how much was actually relocated elsewhere and how much simply lost. The Cefic report gives the total, not the breakdown by destination. The chart is absent because the data do not exist in public, and that is itself a fact: nobody keeps an account of where European industrial capacity ends up.

What Remains Open

The spending package is there: €500 billion for infrastructure and climate, plus defence outside the constitutional constraint. The stock market is at record highs. GDP is growing by a whisker. The question none of the data collected can yet close is where that money will end up: in the productive capacity that has been dismantled, or in the accounts of those who have already moved it elsewhere. The ifo forecasts point to +0.8% in 2026 and +1.1% in 2027 ¹³ — a recovery, not a return.

There are two numbers that will answer before any statement does. The next Destatis release on the balance with China will say whether the fifteen billion of deficit added in twelve months is a step or a slope. And the next set of Ludwigshafen accounts will say whether below thirty thousand is the floor or merely a level passed through. In 1954 that site rose above thirty thousand positions while Germany was rebuilding itself. In 2026 it crosses back down through them while, seven thousand kilometres away, another one opens ahead of schedule. The two dates do not describe a rise and a fall: they describe a centre of gravity that has moved.