At a glance
13.7% more R&D spending and 11.4% more R&D staff at funded companies Own R&D funds up by around 20%, every euro of funding triggers at least one euro of R&D 79% of projects certified by the BSFZ, only 2.8% rejected by the tax office Only 49% of companies with a certificate actually collect the money from the tax office The report's recommendation: continue the Forschungszulage permanently On 10 September 2026, the federal government submitted the statutory evaluation report on the Forschungszulagengesetz to the Bundestag (BT-Drs. 21/7999).Report p. III hib press release, 15 Sep 2026 It was prepared by KMU Forschung Austria, WIFO and ifo for the period 2021 to 2025, and for the first time the effect of the Forschungszulage was measured causally, i.e. against control groups of companies that did not apply.Report p. III Report p. 6 If you, as a CFO or founder, have so far thought this sounds too good to be true: the report delivers, across 198 pages, the numbers you can use to check that. Here are the seven answers that matter for your decision.
1. Yes, the Forschungszulage works, and measurably so
What the report says: Companies with an approved application increased their R&D spending by 13.7% and their R&D staff by 11.4% more on average than comparable companies without an application.Report p. 6 hib press release, 15 Sep 2026 Funded companies raised their own R&D funds by around 20%, while funds from public support grew by 16.8%.Report p. 77 This results in a gross leverage effect of around 1.0 to 1.2, meaning every euro of funding triggers at least one euro of internal R&D spending.Report p. III Windfall effects exist, mainly at large companies that already operate close to their optimal R&D level, but according to the report additionality prevails overall, especially at SMEs.Report p. 6
What this means for you: The allowance is not a tax gift that trickles away inside a corporate group, but an instrument that demonstrably finances additional development at companies of your size. For the internal discussion, that is an argument against the reflex of merely pocketing the allowance: companies that plan with it invest more of their own money on average.
2. Half goes into salaries, the other half into capacity
What the report says: Around half of the allowance is absorbed by wage increases, the other half flows into real expansions of R&D capacity.Report p. 6 In the period after funding, the number of R&D employees at funded companies is on average around one fifth higher than in the control group.Report p. 135 An initial effect, i.e. companies starting R&D for the first time because of the allowance, could not be demonstrated by the evaluation.Report p. 6 The instrument stabilises and deepens existing innovation work; it does not initiate it.Report p. 6
What this means for you: The Forschungszulage finances the developers you need anyway and lets you grow faster with the same resources. It is not a reason to invent an R&D department that would not exist without funding. If your team is already working on technically uncertain problems, the leverage is there. If not, the allowance will not create it.
3. Small companies benefit the most in relative terms
What the report says: Small companies and those with lower R&D intensity show relatively stronger increases than large companies.Report p. 6 In absolute terms it looks different, because large companies receive around 55% of the funding volume.Report p. 7 For 2020 to 2024, the federal statistics show an average assessed Forschungszulage of EUR 60,620 per application for micro-enterprises, compared with EUR 126,260 for large companies.Report p. 60 Among applicants, however, the small ones dominate: around 63% are micro or small enterprises, 86% of cases are SMEs and 18% of applicants are start-ups up to five years after founding.Report p. 5
What this means for you: With a team of 8 or 40 people, you are not the exotic case in the system but the norm. Small companies respond more strongly to the incentive even though their amounts are smaller, and the funding rate of 35% for SMEs, raised in 2024, strengthens that incentive further.Report p. 8 An application in the mid five-figure range is not the exception but the typical case.
4. The risk is calculable if the reasoning is right
What the report says: 79% of the projects examined were certified positively or partially positively by the BSFZ, with little fluctuation over the years.Report p. 40 In the second step, the tax offices reject only around 2.8% of assessment applications, and currently five cases are pending in court nationwide.Report p. 63 The average time until the BSFZ issues its decision is 62 days across all years; in 2025 it was 54 days on average up to the end of August.Report p. 39 The catch comes before that: in 59% (2022) to 67% (2024) of projects, the BSFZ requests additional information, 1.3 requests per project on average.Report p. 40 In 57% of these requests on average, the issue is the risk and novelty of the project.Report p. 41
What this means for you: The Forschungszulage is procedurally safe, but not a foregone conclusion. Two out of three applications need to be improved, and almost always it is the same question: why is this technically uncertain and new, rather than merely laborious? Whoever describes market goals instead of technical target values produces requests for additional information and, in the worst case, a partial certificate. Bear in mind: at the BSFZ you can revise and resubmit; only the tax office's decision is final. For comparison: at Zeitmaker, the success rate across all supported applications is 95 %, because interviews with you
