The relative price of healthy & unhealthy foods in 177 countries

Transcription

1 The relative price of healthy & unhealthy foods in 177 countries Derek Headey Poverty, Health and Nutrition Division International Food Policy Research Institute (IFPRI)

2 Introduction Non-metabolic risk factors in the 2015 Global Burden of Disease Study: Percentage of total DALYs DIETARY RISKS 20% Air pollution 15% Other environment 1% Sexual abuse 2% Low physical activity 2% Alcohol, drug 9% Occupation 4% Unsafe sex 9% Tobacco 14% WASH 10% MALNUTRITION (mostly non-diet) 14%

3 Introduction Dietary problems vary across regions and income groups: q Low Income Countries (LICs): Protein-energy undernutrition, multiple micronutrient deficiencies; lack of high quality protein (animal sourced foods) q High Income Countries (HICs): Excess intake of carbs, sugar, trans-fats, processed red meat, sodium; low intake of fruit & veg and wholegrains q Middle Income Countries (MICs): Most now have double burdens (as do many LICs)

4 Introduction Are overnutrition and undernutrition problems connected by affordability of healthy and unhealthy foods? q Most work on food prices and nutrition has focused on obesity in specific HICs (e.g. Drewnowski et al.) q 2 studies do comparative analysis (ODI; PURE study) q In LICs & MICs only a few contributions: Ø Demand analysis (own-price elasticities) Ø Biofortification literature (Bouis): MNs expensive! Ø Recent ag-nutrition literature on India and Bangladesh (Headey et al.; Sen et al) & forthcoming work from IANDA

5 Introduction q Use prices for standardized foods in 177 countries, for 2011 q Calorie content is an important factor in making food decisions q In most countries cereals are the main source of calories q Hence we measure calorie-price ratios: Cost of 1 egg calorie to 1 cereal calorie (e.g. rice) q Captures cost of diversifying out of staples at a given income Hypotheses? H0: Law of one price across countries (little price variation) H1_A: Food prices vary a lot across countries More true for perishable fresh fruits & veg and ASFs H1_B: Non-tradable food prices influenced by productivity H1_C: Variation in food prices explains child feeding patterns

6 Data & methods q 2011 International Comparison Program data (ICP), typically used for estimates of national accounts, income & poverty q Cost-of-living, or purchasing power, is at the heart of ICP q Requires price collection for standard definition products ØIncomplete data on 200 food products in 177 countries q National average prices are reported (no vendor-level data) q Supplement with national data sources where needed q Combine with USDA food conversion data for calorie content q Assign foods to 17 groups, measure cheapest food in each q Strengths: High degrees of standardization q Weaknesses: Fruit, veg, pulses and fish are under-populated

7 Data & methods Short name Description Long grain rice - Parboiled Number of units: 1: Unit of measurement: Kilogram: Min: 0.5: Max: 1.2: Brand: Well known: Type: Long grain, white rice (milled rice): Packaging: Pre-packed; paper or plastic bag: Quality: High grade: Preparation: Parboiled : Share of broken rice: Very low (not more than 5%): Other features: Not enriched, not aromatic (fragrant), not sticky: Exclude: Premium rice e.g. Basmati rice, Jasmine rice Short-grained rice Number of units: 1: Unit of measurement: Kilogram: Min: 0.5: Max: 1.2: Brand: Brandless: Type: Short grain White: Packaging: Loose or plastic foiled : Preparation: Uncooked, non-parboiled: Milling: Extra-well-milled: Share of broken rice: Below 50%: Other features: Not enriched, not aromatic (fragrant), not sticky: Exclude: Premium rice e.g. Basmati rice, Jasmine rice White bread Number of units: 500: Unit of Gram: Min: 250: Max: 500: Brand: Not relevant: Type: White bread loaf with Number of units: hard crust, 1: Unit not sliced: of Packaging: measurement: Loose: Ingredients: Min. Kilogram: 90% wheat (white flour); Min: High (first) 0.5: grade flour, on a water base Beef, Fillet Number of units: 1: Unit of measurement: Kilogram: Type: Fillet (a strip of boneless meat): Status: Chilled (refrigerated): Max: 1.2: Brand: Composition: Well Without known: bones: Cut: Type: Middle upper Long part of loin grain, (tenderloin): white Fat content: rice No fat (trimmed) (milled Milk, un-skimmed Pasteurized Number of units: 1: Unit of measurement: Liter: Min: 0.8: Max: 1.5: Brand: Well known: Type: Cow; fresh, un-skimmed milk rice): Packaging: (whole Pre-packed; milk): Packaging: Pre-packed; paper carton: Processing: or plastic Pasteurised; bag: HTST (High Quality: Temperature/Short High Time) treatment : Fat content: Natural (3-4%): Other features: Not fortified, not flavored: Exclude: UHT (Ultra-high temperature) milk: grade: Preparation: Comments: Parboiled Specify brand, weight : observed Share of broken rice: Very low Large size chicken eggs Number of units: 6: Unit of measurement: Piece: Min: 4: Max: 12: Brand: Not relevant: Type: From industrially bred, caged (not more than hen: 5%): Packaging: Other Pre-packed: features: Weight per piece: Not g: enriched, Comments: Specify not weight; aromatic size of unit observed Palm oil Number of units: 1: Unit of measurement: Liter : Brand: Well known: Type: 100% pure, refined : Packaging: Pre-packed; (fragrant), not sticky: plastic bottle: Other Exclude: features: Industrial Premium production: Comments: rice Specify e.g. brand Basmati observed rice, Banana, Standard Number of units: 1: Unit of measurement: Kilogram: Type: Fresh traditional yellow bananas: Packaging: Loose : Length: 14 Jasmine rice - 18 Cm: Quality: Good quality; fresh, Intact (unbroken, unmarked) skin: Exclude: Fair trade fruits, organic fruits, Short finger local banana type: Comments: Specify whether "in-season" or "out-of-season" Sweet Potatoes Number of units: 1: Unit of measurement: Kilogram: Type: Fresh sweet potatoes (Color varies: yellow/orange/purple or light brown): Packaging: Loose : Quality: Good quality; fresh, Intact (unbroken, unmarked) skin: Exclude: Organic vegetables, yam

8 Data & methods Cereals (N=13) Roots/tubers (N=3) Legumes (N=2) va-rich fruits/veg (N=7) DGL veg. (N=3) Other Veg (N=4) Other fruit (N=8) Baby cereal (N=1) Milk (N=5) White meat (N=4) Red meat (N=16) Eggs (N=2) Fish (N=5) Oils/fats (N=9) Sugar (n=2) Potato chips (N=1) Coca-Cola (N=1) Starchy staples Rice (5), bread products (5), maize flour, maize, tortilla Brown potatoes, sweet potatoes*, cassava Healthy vegetal foods Lentils (dry), mung beans (dry) Carrot, sweet potatoes*, red pepper, grapefruit, mango, papaya, peach Spinach, Lettuce, Green Cabbage Cauliflower, tomato, onion, eggplant Apple (2 varieties), avocado, grapes, orange, lemon, watermelon, melon Baby cereal: dry mix, typically fortified Animal sourced foods (ASFs) Pasteurized fresh milk (3 types), condensed milk, powdered milk Whole chicken (2 types), chicken breast, chicken leg Beef/veal (7), Lamb/mutton (4), Pork (4), Goat Large brown eggs, medium brown eggs Fresh Carp, Mackerel or Tilapia; canned Sardines or Tuna (HIC only) Unhealthy condiments, junk food Butter (3), sunflower, olive, palm, soybean, peanut, vegetable White sugar, brown sugar Plain salted potato chips Coca Cola, bottled

9 Data & methods We analyze the data in several steps 1: Differences in calorie-prices across foods & regions ØClassify countries into geographical sub-regions, but pool HICs ØFocus on differences in mean calorie-prices across food groups, and variation across countries within food groups ØThen look at differences in calorie-prices across regions 2: Linking calorie-prices to child diets & stunting ØDHS data on children s 24hr food group recall: countries ØUse WHO data on prevalence of stunting ØAll regressions control for GDP per capita

10 Price differences across income groups F1. Population-weighted mean CPRs by income level Panel A: CPRs across low, middle and high income countries LIC MIC HIC

11 Price differences across income groups 700% 600% Panel B: Ratio of mean CPRs in LICs and MICs to mean CPRs in HICs Eggs & poultry 5 times more expensive in LICs! Dairy & red meat twice as expensive! Fish relatively cheap! 500% 400% 300% 200% LIC MIC 100% 0%

12 Price differences across regions High income E. Euro C. Asia South America Central America Small Islands M. East N. Africa East Asia South Asia East Africa Sth. Africa West Africa Central Africa N Plant-based foods Root/tuber Legumes VA-rich FVs DGL veg Other Veg Other fruit Baby cereal Animal-sourced foods Dairy White meat Red meat Eggs Fish/seafood Unhealthy processed foods Oils/fats Sugar Potato chips Coca-Cola

13 Price differences across regions Dairy Eggs Legumes/nuts HIC ECA SAM CAM MNA EAS SAS WAF EAF SAF MAF HIC ECA SAM CAM MNA EAS SAS WAF EAF SAF MAF HIC ECA SAM CAM MNA EAS SAS WAF EAF SAF MAF Poultry meat Red meat Fish HIC SAM SIS EAS WAF SAF HIC SAM MNA SAS EAF MAF HIC ECA SAM CAM MNA EAS SAS WAF EAF SAF MAF

14 Prices and children s dietary diversity Robust regressions of the consumption of different food groups against foodspecific CPRs & GDP per capita (children months old), LICs & MICs Root/tuber va fruit/veg Other fruit DGL Veg Legumes Log food-specific CPR ** Log GDP per capita 0.10*** *** -0.10*** 0.01 R-squared N Dairy Eggs Fish Meat Flesh food Log food-specific CPR -0.12** -0.15*** Log GDP per capita 0.19*** 0.07*** -0.07** 0.11*** 0.06** R-squared N

15 Prices and child stunting Robust regressions of stunting rates against CPRs and GDP per capita (children 0-59 months old), LICs and MICs Root/tuber va fruit/veg Other fruit DGL Veg Legumes Log food-specific CPR Log GDP p.c *** -9.19*** -9.35*** -9.07*** -9.21*** R-squared N Dairy Eggs Fish Meat Flesh food Log food-specific CPR 4.80*** 5.01** *** 8.53*** Log GDP p.c *** -7.71*** -9.47*** -8.24*** -7.99*** R-squared N

16 Conclusions q First paper to measure food prices in a comparative, highly standardized cross-country framework q Main findings Ø Food prices highly variable across regions and income levels Ø Most foods relatively cheap in HICs, fruits & veg exceptions Ø Unhealthy foods expensive in LICs, but cheap in MICs Ø ASFs much more expensive in LICs & MICs; esp. eggs, meat Ø Dairy and egg prices explain child consumption & stunting

17 Conclusions Important policy implications: Ø Foods rich in protein & micronutrients are expensive in LICs Ø Prices strongly associated with local productivity levels, and agroecological constraints (e.g. livestock disease) Ø High prices might constrain behavioral interventions, but income transfers could complement such interventions Important research implications: Ø How do calorie-relative food prices vary within countries? Ø How do calorie-relative food prices vary over time? Ø What production, value chain or trade interventions could most effectively reduce prices in different kinds of settings?

18 Thank you