Five, 10-minute abstract-driven presentations.  

Speakers and Presentations:

• Chair: Daniel Sarpong, University of Ghana

• Marco Springmann, University of Oxford, UK
  The costs of healthy and sustainable diets in low- and middle-income countries
  Slides

• William Masters, Tufts University, USA
  The impact of agricultural production, trade and the food system on global nutrient inequalities and implications for nutrient intervention
  Slides/ Recording

• Rebecca Kiwanuka-Lubinda, University of Zambia
  Impact of input subsidies on household food availability in rural Zambia: A gendered perspective
  Slides/ Recording

• Lukasz Aleksandrowicz, Wellcome Trust, UK 
  The cost of adult diets in relation to their healthiness and environmental sustainability: An analysis of the 2011-2012 Indian National Sample Survey Slides/ Recording

• Kalyani Raghunathan, International Food Policy Research Institute (IFPRI)
  How affordable are nutritious diets in India?
  Slides

• Q&A
  Recording

Abstracts:

The costs of healthy and sustainable diets in low- and middle-income countries

Marco Springman, University of Oxford

Introduction:  The importance of dietary changes towards healthier and more sustainable diets is increasingly recognised for reducing environmental impacts and diet-related disease mortality. However, less is known about the economic dimensions of such changes. Here we estimate the costs of healthy and sustainable diets globally and for all major world regions.

Methods:  We paired an international database of food commodity prices with health and sustainable diet scenarios. Price data were adopted from the Organisation for Economic Co-operation and Development (OECD) and projected forward using the International Model for Policy Analysis of Agricultural Commodities and Trade (IMPACT) model. The diet scenarios were adopted from the EAT-Lancet Commission on Healthy Diets from Sustainable Food Systems, and included balanced flexitarian, pescatarian, vegetarian, and vegan dietary patterns. The analysis covered 150 countries and regions.

Findings:  Dietary changes towards balanced flexitarian, pescatarian, vegetarian, and vegan dietary patterns reduced food expenditure in high-income and upper-middle income countries in 2050, had mixed effects in lower middle-income countries, and increased food expenditure in low-income countries. The reductions in high-income countries ranged from 15-30%, and those in upper middle-income countries from 8-19%, with greatest reductions for vegetarian diets, followed by flexitarian, vegan, and pescatarian diets. In lower middle-income countries, food expenditure was reduced for flexitarian, vegetarian, and vegan diets by 1-9%, but it increased by 1% for pescatarian diets. In low-income countries, food expenditure increased by 8-21%, with lowest increases for vegetarian diets, followed by flexitarian, vegan, and pescatarian diets. Dietary changes towards the least stringent but healthy diet, the balanced flexitarian diet, resulted in total annual savings in food expenditure in 2050 of USD 350 billion in high-income countries, USD 170 billion in upper middle-income countries, USD 470 billion in lower middle-income countries, and in increases of USD 150 billion in low-income countries.

Conclusions:  Dietary changes towards healthy and sustainable diets have the potential to reduce food expenditure in high-income and upper middle-income countries, as well as in many lower middle-income countries for changes towards balanced flexitarian, vegetarian and vegan dietary patterns, but not for pescatarian diets. In low-income countries, food expenditure is likely to increase due to high current and projected prices of many healthy and nutritious foods. By 2050, annual investments of about USD 150 billion would be needed to finance healthy and sustainable diets in low-income countries without increasing food expenditure for consumers

The impact of agricultural production, trade and the food system on global nutrient inequalities and implications for nutrient intervention 

William Masters, Tufts University

Introduction:  From 1990 to 2016, the total burden of nutritional deficiencies in DALYs has fallen by over 12%. The remaining burden of 61 million DALYs falls disproportionately on children under 5 in sub-Saharan Africa and South Asia. Many studies address the role of local policies and programs in the fight against nutritional deficiencies. This work addresses the role of agricultural production and international trade. There is evidence of convergence in the number and types of foods consumed across countries, suggesting that global food trade has contributed to increased equality in the per capita availability of nutrients over time.

Methods:  Data are from the Global Expanded Nutrient Supply (GENuS) Model in which national nutrient supplies for 23 nutrients were estimated by calculating and matching the per capita edible food availability of 225 foods to regional food composition tables. The GENuS model is used to determine how changes in production and trade from 1961-2011 have affected the global distribution of nutrients relative to nutrient requirements and to understand the primary foods through which nutrients have been sourced. GENuS data are disaggregated to quantify nutrients at all stages of the FAO food balance relationship. Quantitative data visualizations are used to illustrate changes in the distribution of nutrients over time, notably variations on the Lorenz curve and Pen’s Parade that indicate specific country contributions, relative population sizes and movements within the distributions over time. Contributions from production and trade to the distributions are quantified using the Gini coefficient and Slope Index of Inequality. Annual national nutrient requirements are determined so that changes in inequality are considered relative to national per capita nutrient needs. Nutrients are mapped to their primary food sources to relate changes in inequality to production and trade of commodities and to consider the implications for further large-scale national nutrient intervention.

Findings:  Preliminary results show reduced inequalities in nutrient availability over time, though exceptions exist. Results are captured visually with less bowing in Lorenz curves for more recent years and corresponding flatter slopes in Pen’s Parade graphs. Gini coefficients for inequality also decline with time. When comparing inequalities of production alone versus availability resulting from international trade, we further expect to see increasing inequality in the production of nutrients (through their foods) over time due to increased specialization in agriculture but decreasing inequality in the resulting availability of nutrients for consumption due to the global international trading system. We expect that the difference in these results, which are due to trade, will be shown to increase over time. We also expect that the effect of trade will be greater specifically since the 1990s in which agriculture was included in the reform agenda of the Uruguay Round of multilateral trade negotiations.  The Uruguay Round included the Agreement on Agriculture which established a set of rules and disciplines specifically geared toward reducing trade-distorting support from governments.

Conclusions:  The global system of agricultural production and trade has likely played a significant role in decreasing inequalities in nutrient availability across the world and over time. Changes to policies related to international food trade should be considered in light of their potential impacts on national nutrient availability. Outcomes of the study will be valuable to multiple stakeholders including government policymakers and donors responsible for crafting legislation on food trade policies, allocating budgetary resources for micronutrient intervention and for funding micronutrient intervention research. The results will provide valuable insights as to the importance of trade for supplying essential nutrients globally.

Impact of input subsidies on household food availability in rural Zambia: A gendered perspective 

Rebecca Kiwanuka-Lubinda, University of Zambia

Introduction:  Large-scale agricultural input subsidy programmes (ISPs) have re-emerged across sub-Saharan Africa (SSA) since the early 2000s, buoyed by the argument that mistakes of past agricultural development strategies have been identified and can be rectified. Various studies have focused on the impact of ISPs on smallholder farmers’ welfare, however, there is a gap in understanding the interplay between ISPs and rural household food security from a gendered perspective. This study estimates the impact of gendered participation in the Fertilizer Input Support Programme (FISP) on months of adequate household food provisioning (MAHFP).

Methods:  We used the nationally-representative Rural Agricultural Livelihoods Survey (RALS) of small- and medium-scale farming households in Zambia collected in 2012 and 2015. A total of 8,839 and 7,934 households were interviewed during the 2012 and 2015 surveys, respectively. The key outcome variable used to measure food availability was MAHFP. The independent variables included socioeconomic and demographic characteristics such as gender of the primary decision maker on agriculture production, participation in FISP, land size, land category, education levels of the household head, off-farm income, tropical livestock units, age of the household head, distance to markets in hours, and agro-ecological zones. We used the correlated random effects model to estimate the gendered impact of participation in FISP on household food availability among smallholder farm households in Zambia

Findings:  Results show that, on average, households are food insecure for at least a quarter of the year. A bigger proportion of households with a female decision maker have lower MAHFP than their male counterparts. However, households participating in FISP and having female primary decision makers can increase MAHFP by at least 23.6 percent.

Conclusions:  We recommend that agricultural policies should include deliberate strategies to improve resource allocation for women to enable them participate equitably in programmes like FISP. To enhance household food security in rural Zambia, it is more beneficial to target households with female primary decision makers. Even better, it is imperative to empower women to participate in agricultural decision-making.

The cost of adult diets in relation to their healthiness and environmental sustainability: An analysis of the 2011-2012 Indian National Sample Survey 

Lukasz Aleksandrowicz, Wellcome Trust

Introduction:  Dietary choice is a major driver of environmental change, and unhealthy diets are a leading risk factor for non-communicable disease. Shifting to healthier dietary patterns can offer environmental co-benefits. However, food cost may be a barrier to dietary change, as healthy diets are often more expensive than average diets in both high-income and low- and middle-income countries (LMICs). Little is known about how environmental considerations impact on the cost of healthy diets, particularly in the context of LMICs. We assessed the cost of adult diets that were healthier and more sustainable, compared to average diets in India.

Methods:  We used household food purchase data from the 2011-2012 Indian National Sample Survey (NSS) to approximate dietary intake. Household-level data were converted to individual intakes, using age- and sex-specific dietary energy requirements, for adults aged 18-59. Foods were matched to Indian-specific greenhouse gas emissions (GHGEs), calculated using Cool Farm Tool, and water (WU) and land use (LU), adapted from the Water Footprint Network and Food and Agriculture Organization, respectively. We calculated mean environmental footprints, and used reported cost in the NSS to derive individual dietary cost per day for a range of diets: “adequate” diet (reference average diet scenario, meeting minimum dietary energy requirements), “guideline-adherent” diet (meeting Indian guidelines on dietary energy, % calories from protein, fat, and intake of fruit and vegetables), “healthier-than-average” diet (at least 3 servings of fruit and vegetables/day; higher than median intake), as well as “lower footprint” versions of these (lower GHG, LU, and WU than median Indian dietary footprints). Mixed effects models, to account for clustering within households, tested for differences in cost between the observed healthy and environmentally sustainable diets described above. All calculations were also conducted across quantiles of household expenditure, used as a proxy for income.

Findings:  Prevalence of healthy diets was low (2% and 11% for guideline-adherent and healthier-than-average diets, respectively), and even lower for the environmentally-sustainable versions of these (<1% and 4%, respectively). Adherence to individual healthy guideline components increased at higher expenditure quartiles. Both guideline-adherent and healthier-than-average diets had higher GHGEs, WU, and LU than average diets. The mean cost, in Indian Rupees (Rs) per day, of guideline-adherent and healthier-than-average diets was substantially higher than the average adequate diet (Rs 40 and 50, versus Rs 33, respectively). Healthier-than-average and guideline-adherent diets with lower footprints had a mean cost of Rs 35 and 43, respectively; less expensive than those with solely health considerations, though still more expensive than the adequate diet. Mixed effects models showed a small, but statistically significant, difference in price between the adequate diet and the healthier-than-average and guideline-adherent diets with lower environmental footprints (0.2 and 1.8 rupees/day, respectively). The difference in cost between adequate and both health-oriented diets with lower footprints, progressively decreased across lowest to highest expenditure quartiles; in the highest expenditure quartile, a healthier-than-average diet with lower footprints was slightly cheaper than an adequate diet. Further work is assessing the reasons for this.

Conclusions:  This is the first descriptive study, to our knowledge, to assess the cost of diets with varying health and environmental characteristics in India, and is an initial step in better understanding uptake barriers of improved diets. We have highlighted the higher cost, for many individuals, of both healthy and environmentally sustainable diets, though also the opportunity for the highest-income individuals to adopt improved diets in a cost-neutral way. Future work should assess these trends in more regional granularity, better understand the reasons for these dynamics, and examine solutions across food systems to improve affordability of healthy and environmentally-sustainable diets.

How affordable are nutritious diets in India? 

Kalyani Raghunathan, International Food Policy Research Institute (IFPRI)

Introduction:  India accounts for one-third of global stunting and has extremely high rates of maternal and child anemia and underweight. There are many hypotheses for the high rates of malnutrition in India, but surprisingly few focus on the affordability of diets. In this paper, we use nationally-representative data on food prices and male/female wages to construct measures of the affordability of an India-specific recommended diet, as well as basic calorie costs. We analyse the affordability of diets across space, time, and gender, and the relationships between affordability of diets, state GDP, and dietary intake.

Methods:  Our main research questions are: (1) How affordable are nutritious diets, and how does this vary across time, space, and gender? (2) Does dietary affordability improve with economic growth? (3) Is dietary affordability associated with improved dietary diversity for young children, and improved intake of micronutrient-rich foods for women and men? We use district-level data from the National Sample Survey (NSS) Rural Price and Wage Collection Survey (RPC) from 2005 to 2011. This dataset has prices for more than 140 edible commodities and wages for 18 occupations for both men and women. This dataset is combined with state-level GDP data from the Reserve Bank of India and dietary information from the 2015-16 National Family Health Survey. We use India’s food-based dietary guidelines to calculate our primary outcome, the Cost of Recommended Diet (CoRD). We measure the Cost of Caloric Adequacy (CoCA) as a secondary benchmark outcome. Affordability is measured by estimating CoRD and CoCA relative to gender-specific wages for unskilled labourers. We use descriptive statistics and regression models to describe the food components of CoRD and CoCA, trends across space, time, and gender, and relationships between affordability of diets, district GDP, and dietary intake indicators.

Findings:  Common foods in each food group in the CoRD and CoCA calculations largely accord with expectations. Rice, wheat, and maize are the most common cereals, and, as would be expected, pulses are the cheapest forms of protein. CoRD is substantially more expensive than CoCA, reflecting the higher costs of diverse diets including micronutrient-rich foods. The CoRD measure appears to be fairly stable over time but exhibits considerable seasonal variation, with troughs around January-February, and peaks in the summer months. This lines up with seasonal variation in the availability of foods, particularly dark green leafy vegetables. Overall hours of work have remained stable for male laborers, but have declined for female unskilled labourers, while male unskilled labour wages trended upwards after mid-2009. Diets that meet nutritional requirements are expensive relative to daily wages for unskilled labourers: CoRD constitutes approximately 50% of the male and about 70-90% of the female daily wage, with some improvement in affordability for male labourers. However, there is considerable variation in the affordability of CoRD across states, with CoRD becoming more affordable in Tamil Nadu, Rajasthan, and Andhra Pradesh but less affordable in Madhya Pradesh, Gujarat, and Chhattisgarh.

Conclusions:  More attention needs to be paid to how economic access to nutritious food influences diets and nutrition, especially for malnourished rural populations. For India, we show that nutritious diets are expensive relative to the wages of the rural poor, much more so than minimum calorie requirements. Moreover, accounting for tastes and preferences would likely raise the actual food costs incurred by households. We discuss the implications of our findings for government food policies, particularly food subsidies that are currently limited to cereals and pulses (in some states only), as well employment policies that influence the wages of the poor.