A note on the comparability of business survey indices

The focus of this paper is on issues arising from the comparison of business survey indices (BSI's). Any BSI may be contaminated by false reports or contagion effects so that a crude comparison of BSI's across different groups is not appropriate. A method is therefore proposed for transforming different business survey indices into comparable indices using the generalized birth and death process. The method is carefully designed in order to filter out contagion effects by taking respondents' interaction into consideration. The method also improves the comparability of the BSI's by adjusting the information precision of different survey groups.