Has Australia Become Obese for the Same Reasons as the US?
Posted: 18 Jun 2007
Date Written: 2006
There are currently two competing views in the US literature on the rise in obesity. Lakdawalla and Philipson, (2002) focus on changes in the technology of food production and work that have altered the relative price of buying and using calories. Cutler, Glaeser and Shapiro (2003) argue that technological change has lowered the fixed costs of calorie production resulting in increased calorie intake and rising body mass. They emphasise increased food intake, particularly snack food, and argue that the effect is more pronounced for households with a non-working spouse. In contrast to the US analyses which focus on changes over time, we use Australian cross section data from the 1995 Australian National Health Survey and the linked National Nutrition Survey to examine the relative impacts on adult body mass of three factors: exercise, energy consumption and labour force attachment of the female. Estimation is complicated by potential endogeneity problems. In the absence of good instruments we use propensity score matching to control for selection. We compare BMI outcomes between treatment observations and controls for different combinations of the three factors. We find evidence that a working spouse is associated with lower own BMI but no evidence for higher partner BMI. We find that exercise has little effect on female BMI but is associated with lower BMI for males with working partners. Persistent reporting bias in energy intake gives rise to perverse and significant effects for calorie intake. Using measured and self-reported height and weight we investigate methods to correct for reporting bias in calorie intake.
Keywords: Obesity, calorie intake, exercise, labour force participation
JEL Classification: D1, I12
Suggested Citation: Suggested Citation