Measurement of FCS and CSI Opening notes.. Collecting data in a period of instability Value of this information Dedication of the team involved Importance of getting this information out to provide assistance to Yemenis deeply affected by a tumultuous year Measurement of FCS and CSI 27-28 May 2015 – Roma Tre University

Contents Food consumption score (FCS) Coping Strategy Index (CSI) Explore the questionnaire module Calculation of the FCS Creation of food consumption groups (FCGs) Presenting outputs Coping Strategy Index (CSI) Two types of CSI Questionnaire Module Calculation

Food expenditure Share Explore the questionnaire module Calculation of the FES Creation of Food expenditure groups (FEGs) Presenting outputs

Definitions Household food consumption The consumption patterns (frequency * diversity) of households over the last seven days Household coping strategies Activities to which household resort to obtain food, income, and/or services when their normal means of livelihood have been disrupted I changed (frequency + diversity)

Share of expenditure on food - An indicator of vulnerability to food insecurity “Households that spend a large proportion of their income on food (greater than 75 per cent, say) are vulnerable to food deprivation because, regardless of their current food consumption status, if they were to experience a reduction in income it would likely be accompanied by a reduction in food consumption or the quality of food eaten.” (IFPRI, 2007)

The Food Consumption Score (FCS)

The Food Consumption Score It is calculated by observing the frequency by which households consume various food items over a seven day recall period Households are asked, “in the last seven days, how many days did you eat ______? “ Each food item is put into a category and the categories are given a weight based on its relative nutritional value

Data collection The data required is collected according to usual food items consumed that are specific to the country’s context Food items are grouped into standard food groups The difference between foods and condiments should be captured during the data collection Give example : fish powder in mali

Food consumption module

Calculation steps Using the appropriate thresholds, recode the variable food consumption score, from a continuous variable to a categorical variable. Sum the weighed food group scores, creating the food consumption score. Multiply the value obtained for each food group by its weight and create new weighted food group scores. Sum all the consumption frequencies of food items of the same group and recode the value of each group above 7 as 7. Using standard 7-day food frequency data, group all the food items into specific food groups.

FCS = ∑xi * ai Calculation of FCS Where, FCS Food consumption score xi Frequencies of food consumption = number of days for which each food group was consumed during the past 7 days (7 days was designated as the maximum value of the sum of the frequencies of the different food items belonging to the same food group) ai Weight of each food group

Food groups and weights   Food Item (customize for local context) Food groups Weight 1 Maize , maize porridge, rice, sorghum, millet pasta, bread and other cereals Cereals and Tubers 2 Cassava, potatoes and sweet potatoes 3 Beans. Peas, groundnuts and cashew nuts Pulses 4 Vegetables and leaves Vegetables 5 Fruits Fruit 6 Beef, goat, poultry, pork, eggs and fish Meat and fish 7 Milk yogurt and other diary Milk 8 Sugar and sugar products Sugar 0.5 9 Oils, fats and butter Oil 10 Condiments

Description of weights Food groups Weight Justification Staples 2 Energy dense, protein content lower and poorer quality than legumes, micro-nutrients Pulses 3 Energy dense, high amounts of protein but of lower quality than meats, micro-nutrients, low fat Vegetables 1 Low energy, low protein, no fat, micro-nutrients Fruit Meat and fish 4 Highest quality protein, easily absorbable micro-nutrients, energy dense, fat. Even when consumed in small quantities, improvements to the quality of diet are large. Milk Highest quality protein, micro-nutrients, vitamin A, energy. However, milk could be consumed only in very small amounts and should then be treated as condiment and therefore re-classification in such cases is needed. Sugar 0.5 Empty calories. Usually consumed in small quantities. Oil Energy dense but usually no other micro-nutrients. Usually consumed in small quantities

FCS thresholds The FCS is a continuous variable (minimum of 0 and a maximum of 112) but describing food security with mean values and ranges is not easily interpreted To describe the food security situation, food consumption groups are created based on standard thresholds The thresholds can be adapted based on knowledge of the consumption behaviour in the country / region of interest

Standard thresholds Poor food consumption 0 – 21 0-28 Food consumption group Standard threshold Adjusted thresholds with oil and sugar eaten on a daily basis Poor food consumption 0 – 21 0-28 Borderline food consumption 21.5 - 35 28.5 - 42 Acceptable food consumption >35.5 >42.5 Say that even the worse households have to consume sugar and oil

Why 21? A score of 21 is a bare minimum. Scoring below 21 means that a household does NOT eat at least a staple and vegetables on a daily basis and therefore is considered to have a very poor diet. The value 21 is derived from: daily frequency * weight of vegetables + daily frequency * weight of staples (7 * 2 = 14) + (7 * 1 = 7) = 21

Why 35? Households with a FCS between 21 and 35 are considered to have borderline food consumption The value 35 comes from an expected daily consumption of staple and vegetables complemented by a frequent (4 day / week) consumption of oil and pulses. (staple * weight + vegetables * weight + oil * weight + pulses * weight) (7 * 2) + (7 * 1) + (4 * 0.5) + (4 * 3) = 35

How to adapt the thresholds Even though these thresholds are standardized there is room for adjustment based on evidence Consider the basic/minimum food consumption in the country Daily consumption of oil and / or sugar can raise the FCS without adding much nutritional value Based on the data collected and knowledge of the country, try to define the thresholds for poor and borderline consumption The thresholds should only be changed based on evidence and should remain the same if you want to compare the FCGs of different surveys

Example of adapting the thresholds Yemen (2009) Daily consumption of oil and sugar over-inflated the FCS of households with poor food consumption A study by IFPRI sought to relate FCS and actual caloric intake. It was found that an FCS of 28 was highly correlated with a caloric intake of 2,100 kcal / day. The thresholds were adjusted to 28 / 42 accordingly

Presenting Outputs - Compare means Region FCS North 45 Central 38 South 27 We use compare mean when we analyse a continuous variable with a categorical Example: Mean FCS by region FCG by urban/rural Age household head by FCG Urban Rural Poor FC 10% 18% Borderline FC 17% 22% Acceptable FC 73% 60% Age household head Poor FC 36 Borderline FC 45 Acceptable FC 42

Food consumption area graph Borderline food consumption Poor food consumption This graph aids in the interpretation and description of both dietary habits and in determining cut-offs for food consumption groups (FCGs).

Diet composition for each FCG camerron

Deterioration in diet quality over time

Wealth quintiles by FCG cambodia

The Coping Strategy Index (CSI)

Introduction The CSI is a simple indicator that reveals how households manage or cope with shortfalls in food consumption. It is based on a list of possible behaviours (coping strategies) CSI combines The frequency of each strategy (how many times each strategy was adopted?) Their severity (how serious is each strategy?) Two types of CSIs exists: Context-specific (livelihood) CSI Reduced CSI

Why a reduced CSI? The context-specific CSI cannot be used for comparative analysis. From field experience it has been noted that several coping strategies recur across different contexts Therefore....

Reduced CSI A reduced CSI was developed It is based on: A universal list of 5 coping strategies Universal severity weights

Standard module 10. COPING STRATEGIES (core module)   In the past 7 days, were there times when you did not have enough food or money to buy food, did you have to? 0 = Not applied 1 = 1 day 2 = 2 days 3 = 3 days 4 = 4 days 5 = 5 days 6 = 6 days 7 = Everyday 10.01 Rely on less preferred and less expensive food |____| 10.02 Borrow food or rely on help from relative(s) or friend(s) 10.03 Limit portion size at meals 10.04 Restrict consumption by adults in order for small children to eat 10.05 Reduce number of meals eaten in a day

Reduced CSI – Calculation

Reduced CSI: dataset Each strategy is a variable Values are the frequencies Note: If filter question is used, cells for households without food security problems need to be recoded to “0”

Computed variable in the dataset CSI is a continuous variable Mean CSI is the the main descriptive statistic

How can the CSI be used? As a proxy indicator for food security (household food access) To build profiles (identify who have to “struggle” the most) To monitor trends over space and time To compare between beneficiaries and non-beneficiaries

CSI – across regions In which region are housheolds most stressed? In which region are households least stressed?

Which livelihood groups are most stressed?

CSI: trend analysis and outcome monitoring Beneficary vs. non-beneficiary

Weaknesses Questions may be cultural sensitive ( could lead to underreporting, further validation of reduced CSI required to assess of valid across cultures) There are no universal thresholds available. Country-specific time intensive if used for the first time

Reference Coping Strategy Index: Field Methods Manual” II edition (2008)

Livelihood based coping strategies module The livelihoods-based coping strategies module is used to better understand longer-term coping capacity of households. Household livelihood and economic status is determined by income, expenditures and assets. By understanding the behaviours households have taken to adapt to recent crises, such as selling productive assets, we can get a rough sense of how difficult their current situation is, and how likely they would be able to meet challenges in the future.

Calculation of groups The module has two columns, one for the list of strategies and one for the possible answers. Each strategy has a severity weight, (1-4), each answer “b” and “c” in the module is then recoded as a 1,2, 3 or 4, (considered as a “yes”).

Severity values for each livelihood strategy Master List ID*** Coping strategy Category 1 Sold household assets/goods (radio, furniture, refrigerator, television, jewelry etc.) 2 = Stress 11 Reduced non-food expenses on health (including drugs) and education 3 = Crisis 9 Sold productive assets or means of transport (sewing machine, wheelbarrow, bicycle, car, etc..) 2 Spent savings 7 Borrowed money / food from a formal lender / bank 15 Sold ​house or land 4 = Emergency 10 Withdrew children from school 18 Sold last female animals 16 Begged 3 Sold more animals (non-productive) than usual

Example ‘Livelihood coping strategies module’ CHECK RANKING

Prevalence of different strategies

Interpretation A higher score indicates more frequent and/or more severe coping strategies. As a proxy indicator for food security (household food access) To build profiles (identify who have to “struggle” the most) To monitor trends over space and time To compare between beneficiaries and non-beneficiaries

Reporting (examples) Livelihood coping If the prevalence of households within the population group using each group of strategies is reported: From the group of stress strategies used, describe which are the most common strategies used. Describe the same for crisis and emergency strategies. If one strategy within a group is much more frequent than others, it is worth describing that.

Expenditure Analysis

Food expenditures Expenditure information is useful as a proxy for wider purchasing power, which is another important component of food access. Households that devote most of the expenditures to buy food are more likely to have difficulties in acquiring adequate amount of food. Generally speaking, the higher shares of total expenditures going towards food, the greater the likelihood a household has poor food access. For households with low levels of income who cannot produce enough food for themselves, buying food becomes, de facto, the main priority.

Classification of households - based on share of expenditure on food Percentage of expenditures on food: < 50% Low 50% - 65% Medium 65% - 75% High >75% Very high (very vulnerable to food insecurity) Thresholds developed by International Food Policy Research Institute (IFPRI)

Available expenditure data – questionnaire module Household food expenditures 30 days recall Household non-food expenditures 30 days, 3 months or 6 months recall Cash estimates of non-purchased food Sources of non-purchased food

Section 4 – Household Expenditures   4.01 - Did you spend money on the following items during the last week for domestic consumption? If ‘no’, enter ‘0’ and proceed to next food-item. If ‘yes’, ask respondent to: estimate total cash expenditure on item for the last week. (register the expenses according to local currency) 4.02 - During the last week did your household consume the following items without purchasing them? If no, enter ‘0’ and proceed to next food-item. If yes, ask: What was the main source of the item? 1=own production 2= gathering/hunting 3=donation/food aid/gift 4=credit 5=received in exchange for labour/items 4.03 - Estimated value of the non-purchased food items consumed during the last week (this question refers to the consumption reported in 4.02) (local currency) 1. Cereals (maize, rice, sorghum, wheat, bread) 2. Tubers (sweet potatoes, cassava) 3. Pulses (beans, peas, groundnuts) 4. Fruits & vegetables 5. Fish/Meat/Eggs/poultry 6. Oil, fat, butter 7. Milk, cheese, yogurt 8. Sugar/Salt 9. Tea/Coffee 4.04 - Did you spend money on the following items during the last month for domestic consumption? If none, write 0 and go to next item 4.05 - Estimated expenditure during the last month (register the expenses according to the currency in which it was done) 4.06 - In the past 6 months how much money have you spent on each of the following items or service? Use the following table, write 0 if no expenditure. 4.07 - Estimated expenditure during the last six months 10. Alcohol/Palma wine & Tobacco 17. Medical expenses, health care 11. Soap & HH items 18. Clothing, shoes 12. Transport 19. Education, school fees, uniform, etc 13. Fuel (wood, paraffin, etc.) 20. Debt repayment 14. Water 21. Celebrations / social events 15. Electricity/Lighting 22. Agricultural inputs 16. Communication (phone cards) 23. Savings 24 . Constructions or house repairs

Additional expenditure indicators 1. Total monthly household expenditures The total amount used for all possible household expenditures, converted to a monthly basis 2. Per capita total monthly expenditure The total household monthly expenditure divided by the number of household members 3. Per capita monthly food expenditure The total household monthly food expenditure divided by the number of household members 4. Share of expenditures on specific food and non-food items (i.e., cereals, meat, education, etc.).

Calculation Steps (1) Total expenditures on food items: Sum all the expenditures and estimated expenditures for the purchased and non-purchased food items (e.g., $wheat + $rice +…+$meat = total food exp) Total expenditures on non-food items (collected on monthly basis): Sum of all the expenditures for the non-food items collected on monthly basis (e.g., $kerosene + $rent +…+ $tobacco = total non-food exp short term) Total expenditures on non-food items (collected on long term period): Divide the long term expenditures by 6 and sum. (e.g., $edu/6 + $health/6 +…+ $ceremonies/6 = total non-food exp long term)

Calculation steps (2) Total monthly household expenditures: Sum of all household expenditures Monthly total expenditure on food items + Monthly total expenditures on non-food item (short term) + Monthly total expenditures on non food items (long term)

Calculations steps (3) Share of expenditures on food: Divide the total food expenditures by the total expenditures High Percentage of expenditures on food: The guideline is to recode the above variable into 4 categories: 1 = <50% (low) 2 = 50%–65% (medium) 3 = 65%–75% (high) 4 = >75% (very high (very vulnerable to food insecurity))

Calculations steps (4) Share of expenditures on specific food items: Divide the expenditure of a specific food item by the total expenditure on all food items. Share of expenditures on specific non-food Items: Divide the expenditure of a specific non food item by the total expenditure on all non food items.

Types of variables & Statistics Share of expenditure on food is a continuous variable Analysis is based on: Means (or medians, if the distribution is not normal) Means comparisons by regions by food consumption groups by livelihood groups by wealth quintiles

Types of variables & Statistics Households classified into very high, high, medium and low share of expenditure on food is a categorical variable Analysis is based on: Cross-tabulations by regions by food consumption groups by livelihood groups by wealth quintiles

Concerns Quality, reliability and accuracy of expenditure data have often been questioned. Expenditures vary seasonally (i.e., bulk purchases of staple foods). Cleaning/checking expenditure data is very important especially when it comes to identifying outliers. (consider the per capita values !) Consumption of non-purchased food. The economic vulnerability of households that are home-producers can be underestimated

Considerations for interpretation Some households have a lower share of food expenditure because they rely on their own production. Better-off households may spend a high proportion on costly food items (e.g. meat), which might increase the percentage they spend on food. Investigate the types of items being prioritized!

Reporting: Charts

Reporting: Tables Livelihood groups: Share of total expenditure spent on food (mean): Artisans 50% Unskilled labour 47% Fishermen 44% Agro-pastoralists 42% Food processors 40% Regular employees 39% Agriculturalists Traders 38%

Expenditure share: example

Questions?

Let’s practice!