Published: 16/06/2005, Volume II5, No. 5960 Page 24 25

Dr D Kernick, GP, Exeter

John Appleby and Simon Stevens argued over whether the government knows how much to spend on the NHS (Debate, page 18, 19 May). But both arguments are wrong as they are underpinned by the fundamental assumption that - although increasing bangs diminish with increasing bucks - putting more in equals getting more out.

This naively assumes there is a simple relationship between the causes and effects that govern organisational behaviour.

From this perspective, the NHS is viewed as a machine, set within a discourse of design, regularity and control. Policy-makers can stand outside the system and engineer it towards defined objectives using a system of carrots and sticks to keep the mechanism on track.

But if the system fails to respond to inputs in a simple way, it is probably best not viewed as a machine.

Studies are beginning to suggest that the NHS behaves like a nonlinear, complex system - and that we should not be surprised if things do not turn out as expected. Small inputs can have large and unintended consequences, large inputs can have no impact at all. So increased NHS investment may produce more output, less output or no change at all. If this is not appreciated, dysfunctional consequences will be even greater as frustrated policy-makers seek evergreater levers of control.