Reducing health harms of foods high in fat, sugar or salt: economic modelling – final report

Footnotes

1. The life stage categories classify households by the age of the head of the household and the presence of dependent children in the household. The 'pre-family' are from 16 years old to 34 years old without children (childless couples over the age of 35 years are automatically included in the empty nester); 'young family' same age but with children; 'middle family' are 35 years old to 44 years old with children; 'older family' are those older than 44 years old and with children and 45+ without children is the remaining group (i.e., other dependents, empty nesters and retired). The classification was provided with the dataset.

2. The Scottish Index of Multiple Deprivation identifies the level of multiple deprivation in small areas across all of Scotland in a consistent way. These areas can then be grouped into quintiles (fifths). Quintile 1 refers to the most deprived area, and quintile 5 refers to the fifth least deprived area.

3. These ranges are defined by Kantar Worldpanel.

4. The life stage categories classify households by the age of the head of the household and the presence of dependent children in the household. The 'pre-family' are from 16 years old to 34 years old without children (childless couples over the age of 35 years are automatically included in the empty nester); 'young family' same age but with children; 'middle family' are 35 years old to 44 years old with children; 'older family' are those older than 44 years old and with children and 45+ without children is the remaining group (i.e., other dependents, empty nesters and retired). The classification was provided with the dataset.

5. Data from treatments 1 were used to test the robustness of the results from treatments 2, 3 and 4. They can also be used to assess (1) the effect of price discounts (comparing the results from treatment 1 and 2), and (2) the joint effect of discounting the prices and advertising the discounts (comparing the results from treatment 1 and 4), and (3) the joint effect of discounting the prices and restricting advertising price discounts in the case of products with higher level of sugar, fat, or salt (comparing treatments 1 and 3). For ease of presentation and to avoid confusion, only the results on the effect of restricting advertising price discounts on consumers' choices are presented and discussed (i.e., results from treatments 2, 3, and 4).

6. Remember that 500 respondents participated in each treatment. Each respondent was provided with three choice sets of three alternatives each. Therefores, the total number of observed alternatives in the data is 4,500 (i.e., 500 x 3 x 3). All the observed alternatives choices were classified as healthy, unhealthy or mixed. For example, in treatment 2 (Table 4), 2127 chcolates were classified as heathy, 2,248 chcolates as mixed, 125 chcolates as unhealthy. Not all 4,500 observed alternatives are different. Many of them are repeated across respondents and choice sets.

7. The models were estimated as cross sections; a preliminary version was estimated using fixed effects (i.e., as a panel data model); however, due to the similarities in the results, this paper uses the cross section to increase the number of observations as in the panel dataset those households with only one observation are eliminated from estimation.