French GDP fell 0.1% in the third quarter of 2026. It was the third straight quarter without growth. INSEE put the volume of output at €663.7 billion, in chained previous-year prices and on a seasonally and working-day adjusted basis, down from €664.571 billion in the second quarter and from €665.697 billion in the last quarter of 2025. Public debt, on the separate Maastricht count published on 29 September, stood at €3,595.5 billion at the end of June, equal to 119.0% of GDP.
The first estimate is Informations Rapides no. 263, released at the end of October. The quarterly accounts file that sits behind it, CNT-A2026T3P, is the source of the rates and the euro levels in this piece. Households did not stop spending in the quarter. Consumption rose. The stall is an investment and export story, with inventories swinging hard enough to hide part of the trade gap.
Three quarters, and a 0.2% overhang
The path in the accounts is short enough to read in one line. GDP rose 0.2% in the fourth quarter of 2025, fell 0.2% in the first quarter of 2026, was flat in the second, and fell 0.1% in the third. INSEE said the second-quarter growth rate was not revised after rounding. For the full year 2025, volume GDP rose 0.9%, after 1.4% in 2024. The growth overhang for 2026, the acquis, is 0.2%.
That overhang has a plain meaning. If output in the fourth quarter stayed exactly where it was in the third, annual growth in 2026 would print at 0.2%. A flat fourth quarter does not repair the year. It locks in a year that is already most of the way to a very small number. The first estimate of the fourth quarter is due on 29 January 2027 at 7:30 am. The detailed third-quarter accounts, which can still move these early rates, are due on 27 November 2026 at 8:45 am.
The level matters because the percentage changes are small and the stock of output is large. From the fourth quarter of 2025 to the third quarter of 2026, the chained volume of GDP fell by about €2.0 billion, from €665.697 billion to €663.7 billion. France has spent 2026 giving back a slice of the level it reached at the end of last year. The chart below plots the quarter-on-quarter rates from the first quarter of 2024 through the third quarter of 2026, and the three contributions that add up to the 2026 rates.
Households spent. Firms did not invest.
Household consumption rose 0.4% in the third quarter, after 0.2% in the second and a 0.3% drop in the first. Government consumption rose 0.3%, after 0.4%. Final domestic demand excluding inventories added 0.1 point to GDP, the same contribution as in the previous quarter. INSEE had at one stage put the second-quarter contribution at 0.2 point and then revised it to 0.1. The spending side of the economy was adding a little, and it was adding the same little as in the spring.
Inside household spending, the split is sharp. Consumption of transport equipment rose 3.1%, after 2.5%, with electric vehicles doing the work INSEE names. Energy consumption rose 1.2%, after a 2.0% fall. Food consumption was stable. Spending on capital goods by households fell 2.0%, after a 2.6% rise. People bought more cars and used more energy. They did not buy more equipment.
Gross fixed capital formation fell 0.6%, after a 0.5% fall. That is the third quarter in a row of declining investment: minus 0.8% in the first quarter, minus 0.5% in the second, minus 0.6% in the third. Investment in manufactured goods fell 1.9%, after a flat second quarter. Transport-equipment investment fell 7.0%, after a 2.6% rise. INSEE ties that reversal to a cruising-sailboat delivery that had lifted the previous quarter, and to weaker corporate vehicle fleets. A single delivery can make a quarterly investment rate look healthy and then vanish. The fleets are the part that speaks to company decisions.
The production accounts tell the same story from the supply side. Manufacturing production fell 1.4% after a stable second quarter. Transport equipment production fell 3.5%. Car production fell 1.5%, after a 3.4% fall, a fourth consecutive quarter of decline. Other transport equipment, the category that includes aeronautics, fell 4.5% after a 2.1% rise. Agri-food production fell 1.5%, after 0.6%, and INSEE points to poor summer harvests. Construction output fell 0.4%, after 0.7%. Market services rose 0.3%, after 0.2%. Services were the only large block still expanding.
Trade subtracted 0.6 point. Stocks put 0.4 back.
Exports fell 1.7% after a 2.8% rise. Imports rose 0.1% after 0.9%. Foreign trade subtracted 0.6 point from GDP, after adding 0.6 point in the second quarter. The swing is one of the cleanest facts in the release. A quarter that had been helped by sales abroad gave the whole of that help back, and a bit more.
Exports of transport equipment fell 9.1% after a 20.8% jump, mainly aeronautics. That is the mirror of the production number. France shipped a lot of aircraft and related equipment in the spring and did not repeat it in the summer. A reader who looks only at the second-quarter export rate will think the external sector is a support. A reader who lines the two quarters up will see a delivery cycle.
Inventories moved the other way. The contribution of stockbuilding was plus 0.4 point, after minus 0.7 point. In the first quarter the inventory contribution had been plus 1.0 point, large enough to offset a 0.9 point trade drag and a 0.3 point drag from domestic demand excluding stocks. The arithmetic of 2026 so far is a rotation between stocks and trade, with underlying domestic demand excluding stocks negative in the winter and only barely positive since.
Add the third-quarter contributions and the GDP rate falls out. Domestic demand excluding inventories, 0.1. Foreign trade, minus 0.6. Inventories, plus 0.4. The sum is minus 0.1, which is the headline. Stocks kept the quarter from being a much deeper decline. They are also the line most likely to reverse when the detailed accounts arrive on 27 November, because early stock estimates are built from partial information.
Prices: 3.4% on a provisional October print
The same morning as the GDP release, INSEE published a provisional consumer-price flash, Informations Rapides no. 266. On that estimate, the consumer-price index rose 3.4% in the year to October, after 3.1% in September. On the month, prices rose 0.4%, after a 0.3% fall in September. The harmonised index, the one used for European comparisons, rose 3.8% on the year after 3.4%, and 0.4% on the month after a 0.3% fall.
Energy is the acceleration. The provisional breakdown puts energy up 24.1% on the year, after 21.3% in September, a fourth consecutive month of faster energy inflation, pulled by gas and petroleum products. Food rose 1.2%, after 1.4%, with fresh products slowing. Services rose 2.7%, after 2.4%, pulled by accommodation and transport. Manufactured goods were still slightly down on the year, at minus 0.2% after minus 0.3%. Tobacco rose at the same annual pace as in September.
INSEE’s own warning on that page is operational. The flash uses a narrow set of price observations and estimates for some tariffs that are not yet in. The provisional index must not be used for contractual indexation. The final October index is due on 13 November 2026 at 8:45 am.
Jobs were already softer before this GDP print
The labour-market number that is actually out is still the second quarter. Informations Rapides no. 192, published on 7 August, put the ILO unemployment rate at 8.3% of the labour force. That was up 0.2 point on the quarter and 0.7 point on the year. The number of unemployed people rose by 62,000 over the quarter, to 2.7 million, and by 261,000 over the year. From that publication onward, the national rate covers France including Mayotte, a wider field than the old “France excluding Mayotte” series.
The rise was across ages. The rate for people aged 15 to 24 rose 0.4 point, to 21.6%. The rate for people aged 25 to 49 rose 0.2 point, to 7.5%, the highest since the first quarter of 2021. The rate for people aged 50 or over rose 0.3 point, to 5.5%, the highest since the first quarter of 2022. The next labour-force release, covering the third quarter, is scheduled for 10 November 2026 at 7:30 am. Until that morning, 8.3% is the rate, and it describes April to June, not July to September.
The dashboard INSEE refreshed on 30 October also carries two softer activity surveys. The business climate for all sectors was 98 in October, up 1 point on the month, with the next reading due on 20 November. Household confidence was 80 in October. Production capacity utilisation in the fourth-quarter survey was 81.5%. A climate reading of 98 sits below the long-run average of 100 that INSEE builds into the indicator. It is a small improvement on September, in a quarter when measured output fell.
Debt is a second-quarter fact: €3,595.5 billion, 119.0%
The debt release is Informations Rapides no. 239, dated 29 September. At the end of the second quarter, Maastricht debt was €3,595.5 billion, up €59.6 billion on the quarter, after a €75.8 billion rise in the first quarter. As a share of GDP it was 119.0%, after 117.5% in the first quarter and 115.7% at the end of 2025, when the stock was €3,460.1 billion.
Cash moved the net figure by more than the gross figure. General government cash holdings fell by €6.8 billion, after rising €21.3 billion in the first quarter. Net debt therefore rose by €68.2 billion, more than the €59.6 billion rise in gross debt, and stood at €3,366.8 billion, or 111.4% of GDP, after 109.7%. The central government’s contribution to the gross debt rose by €53.0 billion. Social security funds added €8.4 billion, the same quarterly increase as in the first quarter, with Urssaf Caisse Nationale up €4.8 billion and Cades up €4.0 billion. Local government debt fell by €1.7 billion.
INSEE prints a warning that is easy to skip. Quarterly debt comes from a thinner accounting source than the annual accounts, and it can be revised for several quarters. The change in debt in a quarter is not the deficit. The government holds financial assets, some liabilities sit outside the Maastricht definition, and the quarterly deficit is seasonally adjusted while the debt stock is not. A €59.6 billion rise in the debt stock is not a €59.6 billion deficit.
The ratio is also a second-quarter ratio. It uses GDP through June. The third-quarter drop in volume GDP does not, by itself, rewrite 119.0%. INSEE will publish the third-quarter debt figures on 18 December 2026 at 8:45 am. Until then, 119.0% and €3,595.5 billion are the numbers, and they are three months older than the GDP print that made the headlines on 30 October.
What the finance minister said, and what the budget is asking for
Roland Lescure, the economy and finance minister, went on France Info the morning of the GDP release. AFP’s account, carried by Boursorama, quotes him directly. “Ça fait trois trimestres que la croissance n’est pas au rendez-vous.” Growth has not shown up for three quarters. “La conclusion de tout ça, c’est qu’en 2026, l’économie française est à l’arrêt.” His conclusion is that in 2026 the French economy is at a standstill.
He called the stall a “conjonction inédite,” an unusual combination, of internal and external shocks. He named the summer heatwave and the tensions around the Strait of Hormuz that are feeding the surge in fuel prices. Then the budget line, again in his words: “Aujourd’hui, si on n’a pas le redressement budgétaire qu’on a mis dans le budget, on ira dans le mur.” If the repair written into the budget does not happen, France will hit the wall. He said he would repeat it to the end: France needs a budget, and it needs a budget deficit of 5% or less. That, he said, is what it will take to get back on a growth path and to persuade investors and companies that “la France ne va pas à la crise,” that France is not heading into a crisis.
AFP’s report of the same interview says the 2027 budget bill, under examination in the National Assembly, aims to bring the public deficit to 5% of GDP after 5.4% in 2026. Adoption, the agency wrote, looks very uncertain, for lack of a majority and with the presidential election approaching. The 5% and the 5.4% are the government’s stated objective for the bill and its description of 2026. They are not an INSEE outturn. INSEE has not published a 2026 deficit ratio in the releases used here.
On 13 October, presenting the budget to the Assembly, Lescure had already put the growth forecast for this year at 0.5%, revised down, and had said the 2026 deficit would reach 5.4% of GDP against a 5% figure originally voted. The text of that presentation is on Vie-publique. The 30 October interview did not walk those targets back. It tied them to the new GDP print and to a warning about what happens if the Assembly does not pass a budget.
The OFCE October forecast, published on 20 October under the direction of Éric Heyer and Xavier Timbeau, is a different document from the official estimate. It put French growth at 0.4% in 2026 and 0.5% in 2027, after 0.9% in 2025, against 1.1% and 1.2% for the euro area. Its quarterly table had the third quarter of 2026 at plus 0.1% and the fourth at plus 0.2%. The first official estimate, ten days later, was minus 0.1% for the third quarter. The OFCE also had the public balance at minus 5.5% of GDP in 2026 and minus 5.4% in 2027. Those are forecast deficits, not accounts.
Borrowing costs, oil, and the rate the ECB has already set
The financing backdrop was already tight before this GDP print. On 20 October, FinanceFeeds reported that the French 10-year yield stood near 4.93%, against about 3.42% for Germany, with the spread over Bunds at 150.8 basis points. The euro was near $1.12. Moody’s review was scheduled for 23 October, with the agency then at Aa3 and a negative outlook. This article does not update that rating. A later rating action that is not on a page opened for this piece is not treated as a fact here.
The oil link in Lescure’s interview is the same shock sitting in the energy-price flash. On Monday 2 November, Brent was trading near $110 a barrel, about 3% below Friday’s $113.43 settlement, with the move tied to talk of a reopening of the Strait of Hormuz. The Monday oil note is a market level, not a French pump price. It is the external price that INSEE’s energy line, up 24.1% on the year in the provisional October index, is already recording inside French inflation.
Interest rates are the other half of the debt arithmetic. On 10 September the European Central Bank raised the deposit rate to 2.50%, the main refinancing rate to 2.65%, and the marginal lending rate to 2.90%, with the new levels taking effect on 16 September. President Christine Lagarde said policy would be set on a “data-dependent and meeting-by-meeting” basis, and that the Governing Council was “not pre-committing to a particular rate path.” The decision and that wording are the last ECB step this desk has from a full report. French debt at 119% of GDP, rolling at a 10-year yield that was already near 4.93% on 20 October, meets that rate environment whether or not the Assembly passes the 5% deficit target.
What can still move, and what cannot
Four dates do the work from here. On 10 November, INSEE publishes third-quarter unemployment. On 13 November, it publishes the final October price index, the one that is allowed to index contracts. On 27 November, it publishes the detailed third-quarter accounts, which can revise the minus 0.1% and, in particular, the inventory contribution. On 18 December, it publishes third-quarter Maastricht debt, the first debt ratio that will sit on GDP through September. The first estimate of fourth-quarter GDP follows on 29 January 2027.
A revision that lifted the third quarter from minus 0.1% to zero would leave the first quarter at minus 0.2% and the second at zero. Three quarters without a plus sign is the description that survives a small revision. A revision to inventories, or to the aeronautics export drop, would change the composition more than the headline. Consumption at plus 0.4% and investment at minus 0.6% say different things about tax receipts and about how durable a 5% deficit target is.
The debt ratio is still a June figure. The first quarter added €75.8 billion to the stock and the second added €59.6 billion. The third-quarter drop in volume GDP enters the ratio only when INSEE publishes it on 18 December. Until then, 119.0% should be quoted as the end-June ratio.
Lescure has said the standstill breaks if the Assembly passes the repair in the 2027 bill, and that the deficit has to be 5% or less. AFP has said the majority for that bill is not there, with the presidential election close. The accounts settle what the vote is about: a year whose first three quarters produced minus 0.2%, zero, and minus 0.1%, with debt at 119.0% of GDP the last time it was measured.
This is not financial advice.