On Objectivity in Economic Research: A Response to Willis

For some years (since 1976) a research team at the University of Queensland have progressively developed and applied a technique termed the Generation of Regional Input — Output Tables (GRIT), in the compilation of regional input — output tables. GRIT was designed as a hybrid technique, that is, one containing elements of nonsurvey or mechanical methods, but also requiring the insertion into the regional table of sufficient ‘superior’ or ‘survey’ data to ensure an acceptable level of accuracy. Willis has recently published a paper purporting to test the accuracy of the GRIT technique. This paper is a response to Willis's claims, and it is asserted that Willis did not observe the normally accepted guidelines for critical evaluation, by ‘testing’ GRIT purely as a nonsurvey technique, rather than as a hybrid model requiring substantial amounts of region-specific data.