This article has two objectives: first, it introduces the Swiss Household Panel (SHP) survey as a typical example of a panel study that uses centralized CATI for both contacting sample members and conducting interviews. Second, its randomized interviewer-respondent assignment over waves is described. This is useful to analyze interviewer effects on survey answers in an efficient way because there is no clustering of respondents in interviewers across waves. Similarly, the SHP also uses a randomized assignment of interviewers to each call/contact. This is useful to analyze interviewer effects on obtaining cooperation because there is no clustering of contacts of sample members in interviewers. Although this design offers strong analysis potentials for survey methodologists, both issues are underused. The paper presents examples how to make use of this experimental design to model interviewer and fieldwork effects.