This blog is part two of a response to James Park's article on markets in education at the progress@HSE blog. The examples Park gives are far from mechanistic. Funding differentiated by ability levels and income is more individualised than flat vouchers or funding systems that are not pupil-led. It’s more nuanced. In practice, there are various realistic ways to calibrate the voucher to ensure that there are incentives to compete for all pupils. This is discussed in more detail in upcoming work. Second, Park does not like the idea of online education. Yet, online education holds great promise for a future technological revolution that could increase productivity significantly. It also holds the potential to produce a less mechanistic system, since it gives stronger opportunities to produce tailor-made programmes for individual pupils. It’s dangerous to be too attached to a specific idea of how education should be carried out. We need innovation also in schools and a precondition for such innovation is that it’s allowed to occur. In contrast to Park’s concerns, it’s possible to maintain peer interactions also when using online education. Companies could easily and often quite cheaply set up online learning facilities to accommodate for many pupils at once. It could be very similar to a regular classroom, but in non-traditional buildings and with teaching mostly via computers. Pupils would then have similar access to peers as they would otherwise. The main point is that there are various ways education can be carried out – and that we must allow experimentation to produce significantly higher quality. In sharp contrast to Park’s argument, a market-based approach to education is not mechanistic, but rather nuanced and dynamic. Monopolies are generally far more stultifying than competitive situations. This holds true also in education.