Europe’s economy grew last quarter and its companies failed anyway. EU bankruptcies rose 5.7 percent between the first and second quarters of 2026 while new business registrations fell 0.5 percent, according to figures Eurostat published on 17 August. The euro area looked worse on both counts, with insolvencies up 6.9 percent and registrations down 0.1 percent.

Three days earlier the same statistical office reported quarterly GDP growth of 0.4 percent in the euro area and 0.5 percent across the EU, with employment 0.5 percent higher than a year ago. Two releases, one week, opposite moods. Both are correct, and reconciling them requires understanding what the business demography series actually counts.

Registrations and bankruptcy declarations are administrative events. A company enters the count when a court or registry records the filing, which can happen many months after the business stopped trading. The series therefore lags the conditions that caused the failure, and a quarter of rising insolvencies often reflects pressure that peaked in late 2025 rather than trouble that started in June.

The sectoral breakdown carries more information than the headline. Education and social work recorded the sharpest increase in bankruptcies at 21.1 percent, followed by transport at 11.4 percent and financial services at 6.8 percent. Those are three different stories. Care and education providers depend heavily on public contracts and municipal budgets that tightened through 2025. Transport operators face freight rates that never recovered and a parcel market being reshaped by new customs charges.

Registrations tell a similarly split story. Information and communication was the outlier, up 8.8 percent, and construction gained 1.0 percent. Industry lost the most ground at 3.6 percent, with accommodation and food services down 3.4 percent and education and social services down 3.2 percent. New software and consultancy firms need a laptop and a registration fee. New industrial firms need capital that is still expensive.

That asymmetry matters for anyone reading the numbers as evidence of entrepreneurial vitality. A registration count treats a single-person consultancy and a factory as one unit each. When the mix shifts toward low-capital services, the total can hold up while the productive base thins.

National figures diverge sharply beneath the EU average. Greek insolvencies jumped by roughly 46 percent, a swing that owes as much to court processing and reform of national insolvency procedure as to underlying business conditions. Comparability across member states remains the weakest part of this dataset. Some countries push distressed firms through formal insolvency quickly, others let them dissolve informally, and neither approach is visible in the aggregate.

Policymakers in Brussels will read the release against two live files. The first is the Commission’s competitiveness agenda, which promises to cut administrative costs for smaller companies and depends on registration data to claim results. The second is the long-stalled effort to harmonise corporate insolvency law, where the difficulty has always been that member states treat business failure as a national moral question rather than a technical one.

Neither file moves on a single quarter of data. What the release does establish is that aggregate growth and firm-level distress are running together, which usually means the gains are concentrated. Employment held up because larger firms kept hiring, not because the population of businesses expanded.

The next release will show whether the second quarter marked a turn or a bump. Until then, the honest reading is narrow. More European companies entered formal failure procedures than in the previous quarter, fewer new ones registered, and the economy still grew. All three can be true at once, and the discomfort that causes is the point of publishing them together.