Refused or discouraged? Dutch SMEs and bank loans, 2009–2026

Using the microdata of the EC/ECB Survey on the Access to Finance of Enterprises (SAFE), I compare the bank-loan outcomes of Dutch and euro-area SMEs from 2009 to 2026 and examine which explanations the survey's own variables can and cannot support. All figures are employment-weighted. From 2009 to 2023, 24% of decided Dutch bank-loan applications were rejected, against 9% in the euro-area comparison group. This gap of about 15 percentage points remains after adjustment for firm size, sector, age, self-reported financial developments, ownership, and the size and purpose of the loan. Because Dutch SMEs apply much less often, the ECB's broader financing-obstacle indicator, measured over all firms for which bank loans are relevant, is 13% against 11%, an adjusted difference of 3.3 points: the lower application rate offsets most of the higher rejection rate. The rejection gap persists within groups reporting favourable financial developments, including firms that meet the ECB's "profitable" classification (25% against 6%). It is largest for loans below €250,000, appears for credit lines, and is smaller for trade credit. Under proportional-selection assumptions and stylised omitted-risk scenarios, explaining the gap would require substantial unobserved differences in borrower risk; these exercises do not bound all forms of borrower heterogeneity or applicant selection. The findings do not support an explanation based on the observed covariates, but they cannot distinguish lender behaviour from unobserved borrower risk, selection into application, or cross-country differences in reporting. All reported specifications indicate that the rejection gap narrowed after 2024, but the size of the change varies and its statistical significance depends on the specification; the small recent sample and a coincident survey redesign prevent a firm conclusion. This paper uses data from the EC/ECB Survey on the access to finance of enterprises.