Subsidies for commercial research and development (R&D) aim to fix market failures, but it's tricky to measure their impact. Recent studies show that firms getting support and those not applying aren't random samples. Even firms not getting subsidies can benefit from spillover effects. This makes it hard to create a fair comparison group. New methods are being developed to tackle this issue. It's also important to consider how firms' R&D investments interact and whether spillover effects are accounted for in the market.