Systems approaches in policy analysis have had a chequered history. Expecta
tions that 'hard' (quantitative) systems analysis would lead to better answ
ers to policy problems have largely been disappointed. Yet systems theory h
as itself moved on, to embrace the concept of autopoiesis and a variety of
soft systems methodologies. Collectively, these theories offer a way of ana
lysing policy quite distinct from the institutionalist approaches which ten
d to dominate the theory and practice of policy-making. Rather than selecti
ng instruments to fit a particular kind of policy problem (the conventional
approach to policy design) systems analysis suggests that the nature of th
e problem cannot be understood separately from its solution. For policy pro
blems characterised by complexity (such as those concerned with environment
al management and regulation, and urban re-development) using systems conce
pts offers a way of rationalising aspects of existing practice and of sugge
sting directions for improvement.