66 increase drug expenditure. However, before the elections of 2002 and 2005 local maxima of growth rates can be observed, which shows that the pharmaceutical industry as well as physicians have no interest in supporting the Red-Green government with low expenditure and even try to counteract their re-election chances. Consequently, we must also consider corporatistic behaviour of the suppliers. One could argue that a different approach is feasible to explain this development: political reform acts usually follow a speci? c schedule. The ? rst year after an election is needed for consolidating the new government coalition, in the second year the political discussion and negotiation process takes place and, ? nally, in the third year the reform act comes into force. This could also explain the cyclical development with strong amplitudes in the third year in power. But several arguments oppose this supposition: ? rstly, between 1983 and 1998 a Christian-liberal coalition constituted the government. It is dif? cult to believe that after every election a consolidation period was necessary. Secondly, the high number of reform acts between 1998 and 2004 shows that the reform-schedule is absolutely ? exible. And thirdly, this approach is not able to explain the reversion of the expenditure cycle after 1998. 3.7. Conclusion The inability of government to contain disproportionate cost increases for pharmaceuticals is mostly attributed to the complex political negotiation and decision-making processes within the German SHI. As a major reason for the unassertive reform process, it is usually alleged that politicians are not able to accomplish strong control instruments against the will of the powerful interest groups of the health care sector. A closer look, however, reveals a different picture. We see political cycles in expenditure developments with strong amplitudes, which re? ect the ability of government to in? uence costs despite the restrictions of decision-making. But if it is possible to affect expenditure development, it is obviously necessary to consider the interests of the controlling protagonists: the health care suppliers, politicians, and especially the powerful pharmaceutical industry. Regarding politicians’ interests, scienti? c literature usually distinguishes between partisan and opportunistic incitements. Partisan theory says that politicians try to serve the interests of their clientele. According to opportunistic theory, politicians temporarily modify politics to improve their election chances. Our econometric tests have shown that health policy in Germany is strongly affected by partisan politics. We also found indications for partisan as well as opportunistic politics in combination with the ambitions of the pharmaceutical industry itself. These econometric results have been con? rmed by empirical observations.