ALTERNATIVE PROCEDURES FOR CONVERTING QUALITATIVE RESPONSE DATA TO QUANTITATIVE EXPECTATIONS - AN APPLICATION TO AUSTRALIAN MANUFACTURING

This paper analyses and extends alternative procedures for converting qualitative expectations responses to quantitative expectations. A num ber of conversion procedures is investigated, including the probabilit y model, the time-varying parameter probability model, and the regress ion approach. The informational content of the survey expectations is compared with simple time series models. It is found that the expectat ions models are superior for many series, both in terms of producing l ower forecast root mean square error (RMSE) values and in detecting tu rning points in the actual data. Survey expectations are also tested f or rational expectations in aggregate using the orthogonality test.