Secondly prescribing (and access to resources-whether drugs or gym) is tied up with evidence - the idea that the doctor has the expertise to weigh up the technical evidence and guide your shared decision-making process. Whilst some people might argue that we can 'test' the effectiveness of going to the gym in the same way as a drug (randomised controlled trial) I believe this is the wrong way of looking at 'social interventions' and that a social prescription will never achieve an equal weight of quantitative evidence base as a pharmaceutical prescription. This immediately puts the social on a sloping playing field of evidence (for the doctor practicing 'evidence-based medicine').
An alternative approach, as Michael Marmot might advocate, is as follows: we know people who live in disadvantaged communities have poorer access to greenspace, opportunities to exercise, and hence higher chance of poorer health. The preventative approach (proportionate universalism) would argue for investment in communities - parks, gardens, gyms etc (as chosen and designed by the local community). Social norms, clubs, park runs etc, then encourage people to make use of these resources.
So, I think social prescribing is good - as far as it goes - but a real preventative approach would invest in communities and community assets, rather than the continual focus on the individual.