Household food purchasing behaviour

Transcription

1 Household food purchasing behaviour Incomes, Prices and Nutrition Rachel Griffith, Martin O Connell and Kate Smith IFS March 2012 Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

2 Introduction Motivation There is a well established relationship between health outcomes and socioeconomic status Many of these health outcomes are related to diet Tackling diet related health problems is a priority of government One strategy is income transfers to low income households Whether income transfers lead to improved diet depends on relationship between income and food purchasing behaviour Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

3 Motivation Introduction The recent recession led to large declines in household income Employment rates declined Real wages stagnated Asset prices fell Recession was contemporaneous with large price level shock and changes in relative food prices Depreciation of sterling led to increase in price of imported goods World commodity prices rose RPI food basket increased by 10% between Oct and Oct and price increase persisted Price of prepared foods rose by much less than other foods Price of vegetables rose by less (partly because of less imports than other foods - e.g. fruit) Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

4 Introduction What we do in this paper Use income variation from recent recession to identify impact of changes in income on food purchasing behaviour Use QUAIDS demand system to control for impact of contemporaneous changes in relative prices Estimate model using panel data on purchases of very disaggregate food products Allowing for household specific prices And heterogeneity in household preferences Use model to describe the relationship between income (or total food expenditure) and diet And decompose changes in diet over recession into income effect and relative price effect Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

5 Expenditure shares Model Assume preferences for food are weakly separable Household h (from demographic group D) in period t chooses between J food types Share of period t food budget allocated to good j is given by: w htj = α htj + γjk D ln p htk + β D j k ln ( xht Γ D (p ht ) ) + λd j Π D (p ht ) [ ln ( xht Γ D (p ht ) where: α htj = α D 1j + αd 2j d h + α D 3j τ t w htj is household h s period t food expenditure share on j p htj is household h s period t price index for j x ht is household h s period t total food expenditure d h is household h s vector of demographics τ t are year and month dummies )] 2 + ɛ htj Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

6 Model Household specific prices Food type j comprises N j more disaggregate sub food types We follow Lewbel (1989) by assuming: Preferences over food types are weakly homothetically separable Within food type utility function is Cobb-Douglas Implies household h in period t faces price index for food type j given by: p htj = 1 k j N j ( ) ptij whij w i=1 hij where: k j is a scaling factor w hij is household h s (mean) food type j expenditure share on i p tij the period t price of product i belonging to food type j Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

7 Model Engel curves We trace out the relationship between expenditure shares and total food expenditure by holding the constant and prices at their mean levels and varying total expenditure: ŵ htj ( d, τ, p,x ht /Γ D (p ht )) = ˆα D 1j + ˆα D 2j d + ˆα D 3j τ + k ˆγ D jk ln p k + ˆβ D j ( ) xht ln + ˆΓ D (p ht ) ˆλ D [ ( )] 2 j xht ln ˆΠ( p) ˆΓ(p ht ) Denote the associated quantity ˆq htj ( d, τ, p, x ht /Γ D (p ht )) And the quantity of nutrient n in household h s period t bundle of food type j, n htj The Engel curve for nutrient n is then: n ht = j ˆq htj ( d, τ, p, x ht /Γ D (p ht ))n htj Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

8 Model Decomposing income and price shocks We decompose the recession s impact on diet into the effect of falling real expenditure and of changes in relative prices Let T = {0, 1} denote pre 2008 and post 2008, then the change in households h s expenditure share on food type j is given by: where: δ Price hj = ( k + Π D (ph 1) ( δ hj = w 1 hj w 0 hj = (δ Price hj ˆγ jk D ln p1 hk k [ ˆλ D j ˆγ D jk ln p0 hk + δhj Income + δhj Res ) ) ( )] x 0 2 ln h ˆλ D [ j Γ D (ph 0) Π D (ph 0) ) )) ln ( x 0 h Γ D (p 0 h ) )] 2 ( ( x δhj Income = ˆβ D 1 j ln h x Γ D (ph 1) ˆβ D 0 j ln h Γ D (ph 0) + ˆλ D [ ( )] 2 j x 1 Π D (ph 0) ln h ˆλ D [ ( )] j x 0 2 Γ D (ph 1) Π D (ph 0) ln h Γ D (ph 0) Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

9 Data Data Panel of households; includes all food purchased and brought into the home Data records individual transactions; includes prices, quantities and nutritional characteristics Our sample includes 3,050 UK households over the period ; data are longitudinal and we observe all households during every month of Consider how a household allocates its total monthly food expenditure Currently use a household s mean monthly transaction price Price graph Real expenditure distribution Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

10 Demographic groups Data Number of Number of Percentage of household-months households households All households 140,338 3, % Family type No kids 58,028 1, % Pensioners 47,228 1, % With kids 35, % Social class A and B 12, % C1 and C2 77,861 1, % D and E 49,622 1, % BMI of main shopper Normal (less than 25) 37, % Overweight (25 to 30) 31, % Obese (greater than 30) 15, % Not recorded 55,383 1, %. Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

11 Data Food types see Table 3.2 Share of Food type Calories per 100g Expenditure Calories Fruits % 5.6% Vegetable % 6.3% Grains % 19.9% Dairy % 13.3% Meats % 12.4% Oils % 7.8% Sweeteners % 6.4% Drinks % 1.9% Prepared % 30.4% Nutritional Information Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

12 Data Food types see Table 3.2 Share of Food type Calories per 100g Expenditure Calories Fruits % 5.6% Vegetable % 6.3% Grains % 19.9% Dairy % 13.3% Meats % 12.4% Oils % 7.8% Sweeteners % 6.4% Drinks % 1.9% Prepared % 30.4% Nutritional Information Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

13 Data Food types see Table 3.2 Share of Food type Calories per 100g Expenditure Calories Fruits % 5.6% Vegetable % 6.3% Grains % 19.9% Dairy % 13.3% Meats % 12.4% Oils % 7.8% Sweeteners % 6.4% Drinks % 1.9% Prepared % 30.4% Nutritional Information Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

14 Data Food types see Table 3.2 Share of Food type Calories per 100g Expenditure Calories Fruits % 5.6% Vegetable % 6.3% Grains % 19.9% Dairy % 13.3% Meats % 12.4% Oils % 7.8% Sweeteners % 6.4% Drinks % 1.9% Prepared % 30.4% Nutritional Information Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

15 All households Results Regression results (1) (2) (3) (4) VARIABLES w_fruit w_veg w_grains w_dairy Logged real expenditure *** *** *** *** ( ) ( ) ( ) ( ) Square of logged real expenditure *** *** *** *** ( ) ( ) ( ) ( ) Constant *** *** 0.195*** 0.295*** ( ) ( ) ( ) ( ) Demographics Yes Yes Yes Yes Time dummies Yes Yes Yes Yes Relative prices Yes Yes Yes Yes (5) (6) (7) (8) VARIABLES w_meat w_oils w_sweeteners w_drinks Logged real expenditure 0.104*** *** *** *** ( ) ( ) ( ) ( ) Square of logged real expenditure *** *** ( ) ( ) (9.07e-05) ( ) Constant *** *** *** *** (0.0103) ( ) ( ) ( ) Demographics Yes Yes Yes Yes Time dummies Yes Yes Yes Yes Relative prices Yes Yes Yes Yes Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

16 All households Results Regression results (1) (2) (3) (4) VARIABLES w_fruit w_veg w_grains w_dairy Logged real expenditure *** *** *** *** ( ) ( ) ( ) ( ) Square of logged real expenditure *** *** *** *** ( ) ( ) ( ) ( ) Constant *** *** 0.195*** 0.295*** ( ) ( ) ( ) ( ) Demographics Yes Yes Yes Yes Time dummies Yes Yes Yes Yes Relative prices Yes Yes Yes Yes (5) (6) (7) (8) VARIABLES w_meat w_oils w_sweeteners w_drinks Logged real expenditure 0.104*** *** *** *** ( ) ( ) ( ) ( ) Square of logged real expenditure *** *** ( ) ( ) (9.07e-05) ( ) Constant *** *** *** *** (0.0103) ( ) ( ) ( ) Demographics Yes Yes Yes Yes Time dummies Yes Yes Yes Yes Relative prices Yes Yes Yes Yes Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

17 All households Results Regression results (1) (2) (3) (4) VARIABLES w_fruit w_veg w_grains w_dairy Logged real expenditure *** *** *** *** ( ) ( ) ( ) ( ) Square of logged real expenditure *** *** *** *** ( ) ( ) ( ) ( ) Constant *** *** 0.195*** 0.295*** ( ) ( ) ( ) ( ) Demographics Yes Yes Yes Yes Time dummies Yes Yes Yes Yes Relative prices Yes Yes Yes Yes (5) (6) (7) (8) VARIABLES w_meat w_oils w_sweeteners w_drinks Logged real expenditure 0.104*** *** *** *** ( ) ( ) ( ) ( ) Square of logged real expenditure *** *** ( ) ( ) (9.07e-05) ( ) Constant *** *** *** *** (0.0103) ( ) ( ) ( ) Demographics Yes Yes Yes Yes Time dummies Yes Yes Yes Yes Relative prices Yes Yes Yes Yes Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

18 Price elasticities Results Elasticities Price elasticities for the model estimated across all households: Fruit Veg Grains Dairy Meats Oils Sweeteners Drinks Prepared Fruit Veg Grains Dairy Meats Oils Sweeteners Drinks Prepared Notes: expenditure weighted mean elasticities. Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

19 Income elasticities Results Elasticities Income elasticities for the model estimated across all households: Food type Income elasticity Fruit 1.01 Veg 0.99 Grains 0.82 Dairy 0.87 Meats 1.13 Oils 0.79 Sweeteners 0.82 Drinks 1.18 Prepared 1.01 Notes: expenditure weighted mean elasticities. Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

20 Income elasticities Results Elasticities Income elasticities for the model estimated across all households: Food type Income elasticity Fruit 1.01 Veg 0.99 Grains 0.82 Dairy 0.87 Meats 1.13 Oils 0.79 Sweeteners 0.82 Drinks 1.18 Prepared 1.01 Notes: expenditure weighted mean elasticities. Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

21 Income elasticities Results Elasticities Income elasticities for the model estimated across all households: Food type Income elasticity Fruit 1.01 Veg 0.99 Grains 0.82 Dairy 0.87 Meats 1.13 Oils 0.79 Sweeteners 0.82 Drinks 1.18 Prepared 1.01 Notes: expenditure weighted mean elasticities. Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

22 Income elasticities Results Elasticities Income elasticities for the model estimated across all households: Food type Income elasticity Fruit 1.01 Veg 0.99 Grains 0.82 Dairy 0.87 Meats 1.13 Oils 0.79 Sweeteners 0.82 Drinks 1.18 Prepared 1.01 Notes: expenditure weighted mean elasticities. Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

23 Food types: Luxuries Results Estimated Engel curves Real expenditure distribution Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

24 Food types: Necessities Results Estimated Engel curves Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

25 Calories, salt and sugar Results Estimated Engel curves Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

26 Macronutrients Results Estimated Engel curves Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

27 By demographic group Results Estimated Engel curves For different food types: Graphs The Engel curves for households with kids are flatter than the other two family types; while those for pensioner are steeper. At all levels of income social classes A and B have higher expenditure shares on fruit and veg, and spend relatively less on prepared food Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

28 By demographic group Results Estimated Engel curves For different food types: Graphs The Engel curves for households with kids are flatter than the other two family types; while those for pensioner are steeper. At all levels of income social classes A and B have higher expenditure shares on fruit and veg, and spend relatively less on prepared food For the other nutrients: Households with kids purchase a much lower share of the total calories as protein Expenditure on salt falls with social class, as does the share of calories from fat Graphs Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

29 Results Decomposing the effect of the recession Changes over the recession: real expenditure Household type Average % change in real expenditure on food All -2.06% Family type Households without children -2.22% Pensioners -3.12% Households with children -0.35% Social class A and B -1.11% C1 and C2-1.86% D and E -2.60% BMI of main shopper Normal (less than 25) -1.63% Overweight (25-30) -3.45% Obese (30+) -2.80% Price graph Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

30 Results Decomposing the effect of the recession Changes over the recession: real expenditure Household type Average % change in real expenditure on food All -2.06% Family type Households without children -2.22% Pensioners -3.12% Households with children -0.35% Social class A and B -1.11% C1 and C2-1.86% D and E -2.60% BMI of main shopper Normal (less than 25) -1.63% Overweight (25-30) -3.45% Obese (30+) -2.80% Price graph Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

31 Results Decomposing the effect of the recession Changes over the recession: real expenditure Household type Average % change in real expenditure on food All -2.06% Family type Households without children -2.22% Pensioners -3.12% Households with children -0.35% Social class A and B -1.11% C1 and C2-1.86% D and E -2.60% BMI of main shopper Normal (less than 25) -1.63% Overweight (25-30) -3.45% Obese (30+) -2.80% Price graph Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

32 Results Decomposing the effect of the recession Changes over the recession: relative prices Food Average price in Average price in Average % type in in change Fruit % Veg % Grains % Dairy % Meats % Oils % Sweeteners % Drinks % Prepared % Notes: Prices are per kg. Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

33 Results Decomposing the effect of the recession Changes over the recession: relative prices Food Average price in Average price in Average % type in in change Fruit % Veg % Grains % Dairy % Meats % Oils % Sweeteners % Drinks % Prepared % Notes: Prices are per kg. Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

34 Results Decomposing the effect of the recession Changes over the recession: relative prices Food Average price in Average price in Average % type in in change Fruit % Veg % Grains % Dairy % Meats % Oils % Sweeteners % Drinks % Prepared % Notes: Prices are per kg. Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

35 Expenditure shares Results Decomposing the effect of the recession (1) (2) = (3) + (4) + (5) Food Expenditure Percentage point Change in share due to type share in change to Price Income Residual Fruit 9.37% Veg 10.83% Grains 8.63% Dairy 13.04% Meats 18.58% Oils 2.02% Sweeteners 0.86% Drinks 4.84% Prepared 31.83% Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

36 Expenditure shares Results Decomposing the effect of the recession (1) (2) = (3) + (4) + (5) Food Expenditure Percentage point Change in share due to type share in change to Price Income Residual Fruit 9.37% Veg 10.83% Grains 8.63% Dairy 13.04% Meats 18.58% Oils 2.02% Sweeteners 0.86% Drinks 4.84% Prepared 31.83% Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

37 Expenditure shares Results Decomposing the effect of the recession (1) (2) = (3) + (4) + (5) Food Expenditure Percentage point Change in share due to type share in change to Price Income Residual Fruit 9.37% Veg 10.83% Grains 8.63% Dairy 13.04% Meats 18.58% Oils 2.02% Sweeteners 0.86% Drinks 4.84% Prepared 31.83% Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

38 Expenditure shares Results Decomposing the effect of the recession (1) (2) = (3) + (4) + (5) Food Expenditure Percentage point Change in share due to type share in change to Price Income Residual Fruit 9.37% Veg 10.83% Grains 8.63% Dairy 13.04% Meats 18.58% Oils 2.02% Sweeteners 0.86% Drinks 4.84% Prepared 31.83% Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

39 Results Calorie allocation across food types Decomposing the effect of the recession (1) (2) (3) (2) = (4) + (5) + (6) Food Calories Change Fixed Allowing reallocation due to type in to fall Price Income Residual Fruit Veg Grains Dairy Meats Oils Sweeteners Drinks Prepared Total Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

40 Results Calorie allocation across food types Decomposing the effect of the recession (1) (2) (3) (2) = (4) + (5) + (6) Food Calories Change Fixed Allowing reallocation due to type in to fall Price Income Residual Fruit Veg Grains Dairy Meats Oils Sweeteners Drinks Prepared Total Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

41 Results Calorie allocation across food types Decomposing the effect of the recession (1) (2) (3) (2) = (4) + (5) + (6) Food Calories Change Fixed Allowing reallocation due to type in to fall Price Income Residual Fruit Veg Grains Dairy Meats Oils Sweeteners Drinks Prepared Total Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

42 Results Calorie allocation across food types Decomposing the effect of the recession (1) (2) (3) (2) = (4) + (5) + (6) (7) Food Calories Change Fixed Allowing reallocation due to (5) (3) type in to fall Price Income Residual Diff Fruit Veg Grains Dairy Meats Oils Sweeteners Drinks Prepared Total Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

43 Results Decomposing the effect of the recession Purchase of calories by demographic group (1) (2) (3) (2) = (4) + (5) + (6) Food Calories Change Fixed Allowing reallocation due to type in to fall Price Income Residual All Family type No kids Pensioners With kids Social class A and B C1 and C D and E BMI group Normal Overweight Obese Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

44 Results Decomposing the effect of the recession Purchase of calories by demographic group (1) (2) (3) (2) = (4) + (5) + (6) Food Calories Change Fixed Allowing reallocation due to type in to fall Price Income Residual All Family type No kids Pensioners With kids Social class A and B C1 and C D and E BMI group Normal Overweight Obese Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

45 Results Decomposing the effect of the recession Purchase of calories by demographic group (1) (2) (3) (2) = (4) + (5) + (6) (7) Food Calories Change Fixed Allowing reallocation due to (5) (3) type in to fall Price Income Residual Diff All Family type No kids Pensioners With kids Social class A and B C1 and C D and E BMI group Normal Overweight Obese Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

46 Results Decomposing the effect of the recession Purchase of nutrients for all households (1) (2) (3) (4) (3) = (5) + (6) + (7) Nutrient per adult per day % Fixed Allowing reallocation due to type in in change fall Price Income Residual Calories % -2.9% 0.2% -1.9% -1.5% Salt (g) % -2.9% 2.4% -0.0% -5.1% Sugar (g) % -2.9% -2.5% -4.0% 0.9% Fibre (g) % -2.9% 1.7% -2.1% -1.2% Protein (g) % -2.9% -0.1% -2.2% -1.0% Fats (g) % -2.9% 1.4% -1.0% -3.4% Carbs (g) % -2.9% -0.5% -2.6% -0.2% Nutritional Information Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

47 Results Decomposing the effect of the recession Purchase of nutrients for all households (1) (2) (3) (4) (3) = (5) + (6) + (7) Nutrient per adult per day % Fixed Allowing reallocation due to type in in change fall Price Income Residual Calories % -2.9% 0.2% -1.9% -1.5% Salt (g) % -2.9% 2.4% -0.0% -5.1% Sugar (g) % -2.9% -2.5% -4.0% 0.9% Fibre (g) % -2.9% 1.7% -2.1% -1.2% Protein (g) % -2.9% -0.1% -2.2% -1.0% Fats (g) % -2.9% 1.4% -1.0% -3.4% Carbs (g) % -2.9% -0.5% -2.6% -0.2% Nutritional Information Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

48 Results Decomposing the effect of the recession Purchase of nutrients for all households (1) (2) (3) (4) (3) = (5) + (6) + (7) Nutrient per adult per day % Fixed Allowing reallocation due to type in in change fall Price Income Residual Calories % -2.9% 0.2% -1.9% -1.5% Salt (g) % -2.9% 2.4% -0.0% -5.1% Sugar (g) % -2.9% -2.5% -4.0% 0.9% Fibre (g) % -2.9% 1.7% -2.1% -1.2% Protein (g) % -2.9% -0.1% -2.2% -1.0% Fats (g) % -2.9% 1.4% -1.0% -3.4% Carbs (g) % -2.9% -0.5% -2.6% -0.2% Nutritional Information Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

49 Results Decomposing the effect of the recession Purchase of nutrients for all households (1) (2) (3) (4) (3) = (5) + (6) + (7) Nutrient per adult per day % Fixed Allowing reallocation due to type in in change fall Price Income Residual Calories % -2.9% 0.2% -1.9% -1.5% Salt (g) % -2.9% 2.4% -0.0% -5.1% Sugar (g) % -2.9% -2.5% -4.0% 0.9% Fibre (g) % -2.9% 1.7% -2.1% -1.2% Protein (g) % -2.9% -0.1% -2.2% -1.0% Fats (g) % -2.9% 1.4% -1.0% -3.4% Carbs (g) % -2.9% -0.5% -2.6% -0.2% Nutritional Information Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

50 Results Decomposing the effect of the recession Purchase of nutrients for all households (1) (2) (3) (4) (3) = (5) + (6) + (7) Nutrient per adult per day % Fixed Allowing reallocation due to type in in change fall Price Income Residual Calories % -2.9% 0.2% -1.9% -1.5% Salt (g) % -2.9% 2.4% -0.0% -5.1% Sugar (g) % -2.9% -2.5% -4.0% 0.9% Fibre (g) % -2.9% 1.7% -2.1% -1.2% Protein (g) % -2.9% -0.1% -2.2% -1.0% Fats (g) % -2.9% 1.4% -1.0% -3.4% Carbs (g) % -2.9% -0.5% -2.6% -0.2% Nutritional Information Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

51 Conclusions What next? Use model to conduct ex ante policy evaluation What would be the effect of a policy which increases/decreases the food expenditure of some groups by a given amount Currently assume within food type preferences are homothetic But we have estimated Engel curves for different types of meats and they show evidence that preferences with meat are not homothetic We could allow for more than 9 food types, but then we would have many more zeros in data Instrument for total food expenditure? We control for demographics and demographic specific time effects Possibly could include household fixed effects Combining data with EFS to allow for incorporation of food in vs. food out choice in model Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

52 Price of food Appendix Notes: Price of food consumed in the home from the RPI. Back: Data Back: Changes over recession Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

53 Appendix Distribution of logged real expenditure Notes: Distribution of household-months. Truncated at the 1 st and 99 th percentiles. Back: Data Back: Engel curves Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

54 Appendix Engel curves for different food types Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26 Back

55 Appendix Engel curves for calories Back Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26

56 Appendix Engel curves for other nutrients Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26 Back

57 Appendix Nutritional content see Table 3.2 per 100g Calorie share Food type Calories Salt Sugar Fibre Protein Fat Carbs Fruits % 5% 90% Vegetables % 14% 69% Grains % 11% 74% Dairy % 48% 28% Meats % 56% 6% Oils % 99% 1% Sweeteners % 0% 99% Drinks % 6% 86% Prepared % 41% 47% Back: Food types Back: Decomposition Griffith, O Connell and Smith (IFS) Resource allocation within households March / 26