AIM: This paper critically appraises relationship arrangements among three broadly conceived sectors: the government sector, the health sector (including researchers) and addictive consumption industries (particularly tobacco, alcohol and gambling). METHOD: Three models for involvement are examined. In the 'tripartite partnership model' health sector agencies engage as co-equals with the government and industry sectors in order to implement public health initiatives such as host responsibility and public education. In the 'non- association model' the health sector engages with government agencies but not with the industry sector. In the 'managed association model' the health sector engages for specific purposes with the industry sector but contact is monitored and managed by government agencies. FINDINGS: Government and industry sectors commonly favour tripartite partnership arrangements. Health sector agencies that opt to engage in these partnership arrangements can encounter conflicts of interest and find their voice subsumed by dominant influences. Furthermore, their partnership compliance generates divisions within the health sector, with partnership dissenters often silenced and excluded from policy processes and funding. The non-association model is the least hazardous to the health sector because it protects against compromise and dominance. The managed association model is an option only when the government sector as a whole is committed strongly and clearly to the public health objectives. CONCLUSION: In contexts where key parts of the government sector are conflicted over their public health responsibilities, health sector engagement in partnership arrangements entails too many risks.
AIM: This paper critically appraises relationship arrangements among three broadly conceived sectors: the government sector, the health sector (including researchers) and addictive consumption industries (particularly tobacco, alcohol and gambling). METHOD: Three models for involvement are examined. In the 'tripartite partnership model' health sector agencies engage as co-equals with the government and industry sectors in order to implement public health initiatives such as host responsibility and public education. In the 'non- association model' the health sector engages with government agencies but not with the industry sector. In the 'managed association model' the health sector engages for specific purposes with the industry sector but contact is monitored and managed by government agencies. FINDINGS: Government and industry sectors commonly favour tripartite partnership arrangements. Health sector agencies that opt to engage in these partnership arrangements can encounter conflicts of interest and find their voice subsumed by dominant influences. Furthermore, their partnership compliance generates divisions within the health sector, with partnership dissenters often silenced and excluded from policy processes and funding. The non-association model is the least hazardous to the health sector because it protects against compromise and dominance. The managed association model is an option only when the government sector as a whole is committed strongly and clearly to the public health objectives. CONCLUSION: In contexts where key parts of the government sector are conflicted over their public health responsibilities, health sector engagement in partnership arrangements entails too many risks.