Over the past two decades Australian meat consumption patterns have shifted away from red
meat consumption and toward chicken and pork consumption. Inspection of relevant price and
consumption data appears to explain the shift. However, the magnitude of relative price
changes for each meat, in addition to the large body of literature dedicated to aggregate
Australian meat demand, may indicate that non-price and non-income factors have contributed
to the changes in consumer preferences over this period. This thesis provides the first known
application of the generalised almost ideal demand system (GAIDS) to Australian meat
consumption data to test for the existence of pre-committed (non-price and non-income driven
demand) meat consumption among Australian consumers. A range of GAIDS model
specifications are estimated to empirically test for the impact of non-discretionary demand shift
factors on Australian meat demand, and to investigate alternative and novel methods of
demand-shift index construction. Non-discretionary demand shift factors incorporated into the
GAIDS include seasonal and time trend factors, and consumer interest in farm animal welfare,
improvements in lamb eating quality, and lamb advertising campaigns. Attention is dedicated
specifically to the determinants of Australian consumer lamb demand via the incorporation of
the lamb eating quality assurance and advertising indices into the GAIDS, and an assessment
of the determinants of Australian consumer willingness to pay for quality graded lamb.
Evidence is found for the existence of pre-committed chicken consumption among Australian
consumers when jointly estimated with seasonal and time trend factors. Results from this thesis
support improved demand modelling of Australian meat consumption using the GAIDS;
demonstrate the potential usefulness of alternative demand-shift indices; lend support to the
notion that the lamb quality assurance program may have prevented larger falls in lamb
consumption and expenditure; and highlight the critical requirement of data quality.
Recommendations are provided for a more formal approach to GAIDS model development and
specification when dealing with underlying data issues.