Unravelling gossamer with boxing gloves: problems in explaining the decline in smoking.

For about three decades countries such as Australia, Great Britain, and the United States have been turning up the heat on tobacco advertising. Encouraging, sometimes dramatic falls in consumption have followed. On any given day in 1993 smokers in such countries are exposed to a welter of news, information, persuasion, and policies designed to turn them off smoking. For a long time explanations and evaluations of the effects of these policies and interventions have been tied to oversimplified causal models when the reality is rather more messy and complicated. Four factors largely explain the reluctance of researchers to move beyond these models: the reductionist tradition of science; the explanatory privileging of recent events and factors; pragmatic concern for policy "tractable" factors; and the relation of funding to the evaluative process. Broader research approaches to understanding changes in complex behaviours such as smoking are required--for example, qualitative methods.