Data_Sheet_3_Drivers of Animal Source Food Consumption: A Biophysical Approach.PDF (126.48 kB)
Download file

Data_Sheet_3_Drivers of Animal Source Food Consumption: A Biophysical Approach.PDF

Download (126.48 kB)
dataset
posted on 24.01.2022, 04:49 authored by Pieter W. Knap

The contributions to energy and protein consumption of animal-source food (ASF) and its commodities bovine, pig, sheep and goat and poultry meat, fish and seafood, dairy products, and eggs were studied by multiple log-log-inverse regression of 1961–2017 consumption (MJ of energy, grams of protein) on income and year within country. The “year” variable implicitly captures time-dependent non-income factors such as prices, climate, agricultural area, urbanization, globalization, gender equality, religion. Fitting the latter six factors explicitly produced unrealistic results, likely due to insufficient within-country variation over time. All consumption patterns differed between countries, and changed over time; these differences and changes were related to income, but considerably more related to time-dependent non-income factors. Within-country estimates of the income elasticity (β) of total energy and protein consumption ranged from −1 to +1: when income increased by 1%, consumption changed by −1 to +1%. The corresponding estimates of the non-income time elasticity (γ) ranged from −0.05 to +0.05% per year: every year, adjusted for income, consumption changed by −0.05 to +0.05%. The β and γ estimates for the contribution of ASF to energy and protein consumption ranged twice as wide as these; those for the contributions of the individual commodities ranged at least three times as wide. The β and γ estimates for those commodities change considerably over time in many countries; their association to each other is very variable too, both between and within countries. Much of this variation takes place at the lower consumption levels. Considering all this, any attempt to forecast the consumption of animal source food (and particularly of its individual commodities) on a more detailed level than globally and on a longer term than a decade should include an income-independent time factor and be very careful with regard to the elasticity coefficients used.

History

References