Minimum Expenditure Baskets - Interim guidance note WFP VAM | Food Security Analysis
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
MINIMUM EXPENDITURE BASKETS Minimum Expenditure Baskets Interim guidance note WFP VAM | Food Security Analysis 1 July 2018
MINIMUM EXPENDITURE BASKETS 2
MINIMUM EXPENDITURE BASKETS MINIMUM EXPENDITURE BASKETS ©July 2018, World Food Programme (WFP), Vulnerability Analysis & Mapping Unit (VAM). All rights are reserved. Reproduction is authorized, except for commercial purposes, provided that WFP is acknowledged as the original source. United Nations World Food Programme Via Cesare Giulio Viola 68/70, Parco de’ Medici 00148, Rome – Italy Arif Husain Chief Economist and Director - Analysis and Trends Service (OSZA) Tel: + 39 06 6513 2014 - e-mail: arif.husain@wfp.org 3
MINIMUM EXPENDITURE BASKETS Contents Aim and scope ..................................................................................................................................... 5 Why a MEB? ........................................................................................................................................ 5 What is a MEB? ................................................................................................................................... 5 Should national poverty lines be used for establishing MEBs? ........................................................... 7 What data are needed to construct a MEB? ....................................................................................... 8 How to construct a MEB...................................................................................................................... 9 How to construct a survival minimum expenditure basket (SMEB) .................................................. 21 How to find a proxy for a MEB when data or time is insufficient ...................................................... 21 Questions or comments? .................................................................................................................. 22 ANNEX 1 .............................................................................................................................................. 23 ANNEX 2 .............................................................................................................................................. 33 4
MINIMUM EXPENDITURE BASKETS Aim and scope This guidance note explains the basic steps for constructing a minimum expenditure basket (MEB). This is a living document that will continue to be updated as more experience is gathered in calculating and using MEBs. MEBs are often constructed in an interagency context where WFP has limited or partial influence on the content. The purpose of this guidance is to provide conceptual clarity and best practices, relying on experience both from the humanitarian and the development field. The aim is to equip WFP staff with what they need when constructing a MEB, whether this takes place in an interagency forum (most likely the Cash Working Group) or is done by WFP on its own. Because the interagency context is the most common one, the guidance cannot be completely prescriptive but needs to allow for some flexibility in its recommendations. Why a MEB? MEBs were originally constructed primarily to identify the percentage of households in a target population who are poor, i.e., cannot meet their essential needs. For WFP, a MEB is useful in a variety of operations but particularly where the organization responds with cash-based transfers (CBT) to meet food needs or a broader set of essential needs through a multisector cash or a multipurpose- cash intervention.1 The MEB can help to achieve the following: ✓ support decisions on transfer value amounts for food and non-food needs, including supporting multi-sector coordination (government, partners and donors); ✓ support population profiling, and in some cases targeting, for multi-sector/multipurpose cash interventions by identifying the characteristics of those who cannot meet their essential needs; ✓ inform decisions on which goods and services to assess in a supply assessment; ✓ monitor immediate and longer-term food security and resilience outcomes by analysing expenditure trends relative to the MEB; and ✓ establish a relevant basket against which to monitor market prices and the cost of living. What is a MEB? A MEB is defined as what a household requires in order to meet their essential needs, on a regular or seasonal basis, and its average cost. Essential/basic needs are defined as essential goods, utilities and services required by households to ensure survival and minimum living standards.2 The monetary threshold established is equivalent to a poverty line. The households whose expenditures fall below this threshold are defined as households who cannot meet their essential needs. The costs of basic needs approach, or establishing a MEB, is fairly new in humanitarian contexts; however, it has long been the most common way to construct national poverty lines.3 Hence, there is often national experience to draw on. 1 Multisector cash refers to a coordinated approach with traditional sector-specific interventions to deliver unrestricted or restricted cash transfers that cover the needs of a household (such as food, water, shelter, livelihood, etc). Multipurpose cash refers to unrestricted cash transfers corresponding to the amount of money a household needs to cover fully or partly a set of basic needs, regardless of the sector they belong to. These include, but are not exclusive to, one-agency interventions. 2 CaLP / DRC/ Mercy Corps / Save the Children / OCHA (2018). Guidance and Toolbox for Basic Needs Assessment. Available at http://www.cashlearning.org/downloads/East%20and%20Central%20Africa%20CVTWG/basic-needs- assessment-guidanceoct17-3.pdf 3 Haughton and Khandker (2009). World Bank Handbook on Inequality and Poverty. Available at http://siteresources.worldbank.org/PGLP/Resources/PovertyManual.pdf 5
MINIMUM EXPENDITURE BASKETS A MEB does not necessary equate to all the essential needs of a household. It is only supposed to capture needs that the household has to cover entirely or partly through the market. For example, in contexts where electricity is considered an essential need but not available for the target population, it cannot be included in the MEB. If shelter is provided in a refugee camp, or public education is provided, these are not captured in the MEB. The choice of a MEB is always somewhat arbitrary. As the World Bank Handbook for Poverty and Inequality outlines, the starting point for establishing a MEB is to estimate the cost of acquiring enough food to meet energy requirements, usually 2,100 calories per person per day, as is the typical threshold also used by WFP. The cost of other essential needs is then added. This can be done using two different approaches or a combination of both: 1) an expenditure-based approach focusing on effective demand; and/or (2) a rights-based approach based on assessed needs. While the expenditure-based approach is used for most national and international poverty lines, the rights- based approach is the principal method followed in the multipurpose cash grants (MPG) guidance developed for humanitarian purposes.4 Essential needs in humanitarian contexts is understood to imply access to full rights as defined by international humanitarian law. According to the MPG guidance, the International Humanitarian and Human Rights Law protects the right of crisis-affected persons to food,5 drinking water, soap, clothing, shelter and life-saving medical care. Humanitarian Sphere Standards builds on this definition, adding basic sanitation, contagious disease prevention and education.6 A MEB is not equivalent to a transfer value – although a MEB is a critical factor when determining transfer values. This distinction is important because a MEB remains the same regardless of assistance and funding constraints.7 Most households have their own resources to meet some of their essential needs, so the transfer value will usually be less than the MEB value, covering the gap between own resources, other assistance received and the MEB. The module on Operationalization discusses the gap analysis in more detail. A MEB aims to capture minimum essential needs for average households. It does not account for additional requirements of distinct groups such as pregnant and lactating women, infants, young children, adolescents, the elderly, people living with disabilities, and people with chronic diseases. To estimate impacts of an intervention, the MEB threshold should not be changed over time. The threshold should only be adjusted for price changes. A MEB aims to mainly capture recurrent, regular needs of households. It does not strive to capture ad hoc costs. This can be challenging, particularly in emergency situations when needs are dynamic. 4 UNHCR, CaLP, DRC, OCHA, Oxfam, Save the Children, WFP. Operational Guidance and Toolkit for Multipurpose Cash Grants. See http://www.cashlearning.org/mpg-toolkit/ 5 Defined as energy needs, not considering full nutrient needs (protein, vitamins, minerals etc). 6 Ibid. 7 Note that a distinction is made between needs that are met by public provision (for example free primary education), and temporary assistance (for example an emergency safety net provided by the government). Why the needs for the former should not be included in the MEB, the needs for the latter should (and then be accounted for in the gap analysis). 6
MINIMUM EXPENDITURE BASKETS While this guidance advocates for keeping the MEB fixed to the extent possible, in such situations it might be justified to first create an interim MEB (see section How to find a proxy for a MEB when data or time is insufficient) and when the situation has stabilised, a final one. A similar challenge also concerns needs that are inherently irregular, large and unpredictable, such as health needs. We return to this below where specific needs are discussed. Should national poverty lines8 be used for establishing MEBs? Can we not simply rely on national MEBs? Most countries have their own poverty lines to identify the poor so why not use this poverty line? Whenever possible, the first choice should be to align with government practices. However, this is not always feasible, for three main reasons: ✓ Practices vary widely when it comes to constructing poverty lines. Although MEBs are the most common approach, sometimes poverty lines are set as a share of mean/median income or expenditures or as a fixed percentage of the income or expenditure distribution (although mostly not applicable in low income countries). Sometimes countries exclude non-food items from their MEB. Countries can also have different poverty lines for different purposes. In Zambia, the national poverty line is constructed using the MEB approach based on a simple food basket that meets minimum food needs for a family of six. Imagine this food basket is deemed to cost US$100 per month. This is defined as the food poverty line. To construct the full poverty line, the minimum non-food needs of households are estimated based on the average share of expenditure that households just above the food poverty line dedicate to needs other than food. Let’s say that this corresponds to US$35 per month. The total poverty line is then the sum of the food and non-food lines, which with our hypothetical figures would be US$100 + US$35 = US$135. By contrast, Turkey uses the standard EU approach to measuring poverty, i.e. 50 or 60 percent of median income. However, eligibility for social assistance is based on the household income gap from the national minimum wage. ✓ The target population can be different from the country population and they may have different essential needs because they live in refugee camps or do not have access to the same services as the resident population (for example, public education). ✓ The data that WFP typically can collect through Comprehensive Food Security and Vulnerability Analysis (CFSVAs), Emergency Food Security Assessments (EFSAs), baseline assessments and Post Distribution Monitoring (PDMs) are often much less detailed than the Household Budget Surveys or Living Standard Measurement Surveys used to calculate national MEBs. It is widely observed that the more detailed the questions about expenditures, the higher the reported expenditures.9 8 National poverty lines should not be confused with the international (World Bank-defined) poverty line of US$1.90/day. While the former is defined nationally, the latter is a uniform poverty line that can be applied anywhere in the world. 9 Haughton and Khandker (2009). World Bank Handbook on Inequality and Poverty. Available at http://siteresources.worldbank.org/PGLP/Resources/PovertyManual.pdf 7
MINIMUM EXPENDITURE BASKETS If the poverty line is constructed using detailed data but the assessment of household needs relative to the poverty line is based on less detailed data, errors in the assessment are likely to occur. Also, WFP surveys do not allow for calculation of depreciation of assets which often are accounted for when calculating national poverty lines. Even if the national poverty line cannot be used in the vast majority of cases, especially in humanitarian contexts, elements of the methodology can be replicated. Thus, it is important to know how the national poverty lines are constructed. How about using the weights applied in consumer price indices (CPI)? In most countries, household budget survey data is used to construct the underlying baskets through which consumer prices are measured. A weight that corresponds to average household expenditure patterns is applied to each component in the CPI.10 This basket is not ideal for MEB calculations because it corresponds to average consumption patterns. MEB calculations are based on the consumption levels and patterns of those households who are just able to meet their essential needs. What data are needed to construct a MEB? Ideally, all of the following should be available: ✓ a representative household survey of the target population such as an EFSA, a CFSVA or a pre- assistance baseline survey with an expenditure module that includes food and non-food expenditures (including non-purchased items from own production);11 a good-quality PDM survey can also be used but with care, as expenditure patterns may change once households receive assistance; ✓ focus group discussions with key informants and the target population; ✓ price data series covering the intervention area for food items included in the food basket and price data on key non-food items and services; and/or ✓ price indices from national statistical offices (national data and data collected by WFP and partners can be complementary); and ✓ data based on national Household Budget Surveys or Living Standard Measurement Surveys. 10The CPI weights are mostly available through national statistical offices. 11In WFP surveys, the CARI module is recommended for the MEB. Please look at the module on Essential Needs Assessments in the Essential Needs Guidance package for more information on data requirements. 8
MINIMUM EXPENDITURE BASKETS What if data using a Household Economy Approach is available? The Household Economy Approach (HEA) developed by Save The Children is a commonly adopted approach for analyzing food security and livelihoods. It is based on understanding how households normally access income, food and other items/services for survival, established through a baseline analysis. As part of the baseline, the HEA defines livelihood zones where households share similar strategies for obtaining food and income. It also distinguishes households within these livelihood zones in at least three (commonly four and sometimes more) wealth groups. The HEA baseline quantifies the sources of food and income, as well as the expenditure patterns for each wealth group and livelihood zone. The information collected on expenditures can be used as a data source for calculating an MEB. However, due to the relative nature of the wealth cut-offs used, there is no set standard on which group should be the reference for the MEB. If HEA data is utilized, it is important to understand how it was collected – the HEA is simply an analytical framework, not a set method of data collection. Thus, while HEAs are often conducted through qualitative methods (e.g. focus group discussions), they may also be based on quantitative modules in household surveys. The latter yields more rigorous information, however qualitative data can nonetheless be used - but should be cross-checked/triangulated with other sources. How to construct a MEB A MEB is constructed by first estimating the cost of acquiring adequate food, then adding the cost of other essential non-food expenditures. The two principal methods used are the expenditure-based approach and the rights-based approach relying on assessed needs. This guidance argues that while both approaches are viable, depending on the context and the purpose of the MEB, the preferred method should combine elements of both. Box 1: Checklist for constructing a MEB ➢ Identify a food basket ➢ Add a non-food component ➢ Adjust for regional or urban/rural price differences if needed ➢ Consider accounting for differences in household size and composition ➢ Consider seasonal variations, if significant ➢ Validate with stakeholders Expenditure-based MEB Establishing a MEB based on household expenditure data requires calculating the minimum cost of living for households right above the threshold that corresponds to the minimum amount needed to meet essential needs. Including households below this threshold would lead to a basket that does not satisfy essential needs, while including relatively richer households would lead to the inclusion of non- essential needs and therefore inflate the MEB. But what is just enough? In non-humanitarian contexts when the poverty line is based on extensive household surveys, the calorie intake of each household can be calculated. In this case, the ‘poor’ are typically identified as those consuming 2,000-2,200 kcal per person per day, i.e. around 2,100 kcal, or below. In humanitarian contexts, the level of detail required to calculate individual calorie intake is not available 9
MINIMUM EXPENDITURE BASKETS in the surveys used (CFSVAs, EFSAs and pre-intervention baselines) so we need other ways to identify the cohort of households who are just able to meet their needs. This is perhaps the most challenging part of constructing a MEB. How this is done depends on the characteristics of the underlying population and the available data. Below we outline some typical cases: 1. No assistance is present, most households are poor but there are households in the population who are able to meet their essential needs: use the expenditures of households with an acceptable Food Consumption Score12 combined with the criterion that they do not use negative coping strategies (or have a high coping strategy index). Triangulation with other variables such as dwelling quality, asset index or expenditure quintile groups is useful. 2. Households already receive food assistance: to the extent possible, avoid using indicators that are highly influenced by assistance (or such indicators only). For example, in the presence of food assistance, the Food Consumption Score of some poor households might be acceptable. From a holistic essential needs perspective, these households might still be unable to meet their essential needs by themselves and should not be used as reference households when establishing the MEB. In this case dwelling quality, asset index or other similar indicators can also be useful. 3. The majority of the households are very poor and vulnerable so the cohort of households who can meet their basic needs is not big enough to construct a MEB: use a rights-based approach. If survey data is available, make a reality check using these data (and focus group discussion) to understand household consumption patterns. 4. There is a large spread in the target population in terms of well-being: exclude expenditure quintile groups 1 and 5 combined with additional criteria such as acceptable Food Consumption Score, no negative coping and dwelling quality. The reference food basket is then established corresponding to the main consumption patterns of the population “just able to meet their essential needs”.13 Once the basket is established, it has to be priced using local food price data to estimate the total value of this food basket. ANNEX 1 includes an example from Kinshasa on how to construct approximate food baskets from data typically collected by WFP. If the target population is part of and similar to the population living in the country, the reference basket (in terms of actual food items) used for the national poverty line can be used, if available and on the condition that it corresponds to data that WFP collects or has access to through the government and/or partners. The reference basket used for the national poverty line in Zambia includes specified quantities of maize meal, cooking oil, salt, beans, groundnuts, onions, tomatoes, other vegetables, kapenta, bream, milk and salt. WFP collects prices on maize and maize meal, cassava meal, millet, beans and groundnuts. To be able to use the national reference basket, prices for some of the food items have to be obtained from other sources or additional data has to be collected. 12 The FCS is a composite score based on dietary diversity, food frequency, and the relative nutritional importance of different food groups. For more information, see http://documents.wfp.org/stellent/groups/public/documents/manual_guide_proced/wfp271449.pdf?_ga=2.12586487.11 6493132.1499950460-1472687100.1410256493 13 If desired NutVal can be used to set the quantities as per recommended macro nutrient proportions. 10
MINIMUM EXPENDITURE BASKETS Once the food component has been established, a non-food component should be added. There is no wholly satisfactory way to add a non-food component as it is difficult to define what a basic minimum is. Unlike food needs, non-food needs cannot be anchored to a specific threshold, such as 2,100 kcal per person per day. The simplest approach is to add a specific allowance for non-food expenditures by using the average food expenditure share in the total expenditures of the population “just able to meet their essential needs”. The total MEB then consists of a food component and a non-food component. A hypothetical MEB could be calculated as follows (figures are per capita per month): ➢ food component = cost of reference food basket = US$20 ➢ food expenditure share in total expenditures for those “just able to meet their essential needs” = 60 % ➢ overall poverty line (MEB) = US$20/0.60 = US$33.3 ➢ non-food component = US$13.3 However, the survey data can also be used to identify specific non-food needs. The precise non-food components can vary slightly by context but would generally include the components discussed in the section on the rights-based MEB below. Care should be taken when it comes to underreporting of expenditures and treatment of expenditures that are irregular in nature. If a MEB is constructed from survey data, not only household expenditures but also purchase/payment on credit needs need to be taken into account, particularly if indebtedness is very prominent in the target population. The ENA module of the Essential Needs guidance package discusses in more detail how to deal with questions on debts in the survey. Care has also to be taken when estimating MEBs from survey data for predominantly rural populations engaged in subsistence farming, as food expenditures can be underestimated, if non-purchased food is not properly estimated and valued. Shelter is a particularly tricky component to treat for sedentary urban populations. If households own their dwelling and they do not pay rent, they might be classified as poor just because they do not have any major shelter expenditures. The module on Operationalization discusses these two issues in the section on gap analysis. ANNEX 1 demonstrates how choices were made in Kinshasa when it comes to non-food expenditures. It is essential to validate the choices made through partner and beneficiary consultation. Rights-based MEB As noted above, in humanitarian contexts the ‘basic needs’ approach has been understood to imply access to full rights as set out by international humanitarian law and the Humanitarian Sphere Standards. The term ‘rights-based MEB’ is derived from this understanding. Generally, these rights comprise access to food, shelter, utilities, non-food items and services (including health, education, transport and communication). Sometimes residency/legal documentation is also included. This exercise is often done by the Interagency Cash Working Group, and each sector contributes with their sectoral needs. In such a case WFP is responsible for defining the food component. The rights-based approach is essentially to define a detailed list of food and non-food needs and price them. The list is typically produced through focus group discussions with the target population, 11
MINIMUM EXPENDITURE BASKETS partners and key informants. The pricing/costing is mainly done based on actual market prices, but sometimes household expenditure data are also used. Food To construct the food basket for a rights-based MEB, the WFP Cash and Voucher Manual offers good starting points:14 a CBT transfer value should correspond to a nutritionally balanced minimum food basket from the market, adapted to locally preferred diets and ensuring access to macro- and micro nutrients in addition to adequate calorie intake. Thus, in comparison to the expenditure-based approach, the rights-based MEB recommends “a nutritionally balanced basket”, while the expenditure based MEB just relies on actual consumption patterns. Despite this conceptual difference, it is still important to recall that also a rights-based MEB is supposed to cover minimum essential needs only, rather than reflect an ideal situation.15 Shelter This is the cost of accommodation that meets basic shelter needs and rights. What this means in practice will depend on the context, driven for example by weather conditions and what is realistically available to the populations (for example, displaced people in northern Nigeria versus Syrian refugees in Turkey where winters are cold). For the Syrian refugee operation in Turkey, the MEB includes the costs of a shelter that meets certain standards, such as a minimum of 3.5 m2 per person, access to a toilet and running water. Utilities These include the cost for basic utilities such as safe drinking water and, depending on the context, electricity. Cooking gas/fuel is often included in non-food items. Non-food items These reflect basic household needs for cooking, clothing, hygiene and general household items. The recommendation here is to focus on recurrent needs, even if in some contexts also one-off needs are included. The precise list has to be defined in consultation with beneficiaries, partners and the government. In practice, these lists can look very different (see examples in Tables 1 and 2). In some refugee contexts, UNHCR has already defined lists of specific items, which can be taken into account for the MEB. 14 Available at http://documents.wfp.org/stellent/groups/public/documents/newsroom/wfp274576.pdf 15 A tool that can be used to estimate the costs of meeting nutrient requirements of a household composed of different members based on composition and prices of locally available foods is the Cost of the Diet (CotD) that has been developed by SC-UK and used by WFP’s ‘Fill the Nutrient Gap’ team of the Nutrition Division. The CotD uses linear optimization and takes needs for 16 different nutrients into consideration. However, it is important to note that the cost of a diet that meets all nutrient needs of a household’s different members costs typically 50-100 percent more than a diet that meets energy needs and hence might not correspond to actual consumption patterns of poor people. 12
MINIMUM EXPENDITURE BASKETS Table 1 shows an example list of non-food items from the refugee operations for Syrians in Turkey. A similar list for Kakuma refugee camp in Kenya is in Table 2. Table 1: Non-food items included in the refugee operations for Syrians in Turkey Quantities per HH Type Product (of 6 pax) per month Hygiene Toilet paper 24 rolls Hygiene Toothpaste 4 tubes/ 100ml Hygiene Toothbrush 6 toothbrushes Laundry Hygiene 1.5kg detergent Liquid dish Hygiene 750ml detergent 6 packets of 10 pads Hygiene Sanitary napkins per packet Hygiene Individual soap 12 pieces of 125g Disinfectant / Hygiene 500ml cleaning fluid Hygiene Shampoo 650ml Hygiene Diapers 88 per packet Clothes & Other N/A household items Cooking LPG Bottle 12kg bottle gas Note: Non-food items were defined to include only recurrent needs (per month per six-person household) Source: Hobbs (2016). MEB/SMEB calculation for Syrians living in Turkey, September 2016. Report commissioned by the Cash-Based Interventions Technical Working Group in Turkey. Table 2: Non-food items specified by UNHCR for Kakuma refugee camp in Kenya (per capita) cover both one-off and recurrent needs Quantities per Type Products capita/household Hygiene Soap 250 mg (per month) one piece (per 2 Cooking Kitchen set years) one piece (per 2 Other Mosquito net years) 20-litre rigid one piece (per 6 Other plastic Jerry can months) Synthetic Other one piece (per year) sleeping mat Other Woollen blanket one piece (per year) Source: Kimetrica (2016). Refugee Household Vulnerability Study: Kakuma refugee camp Services This includes costs to access basic services such as healthcare, education, transport and communication. 13
MINIMUM EXPENDITURE BASKETS Healthcare costs are difficult to estimate since they are inherently irregular, large and unpredictable. Typically, however, the rule of thumb is that only basic minimum needs are covered, such as two visits per year to the doctor; sometimes expenses for critical events, deliveries, baby kits and medicines are also included. Again, it is worth emphasizing that even if a need is not covered by the MEB, it doesn’t mean that these needs don’t need to be met. Education costs cover school fees, materials, uniforms and transport, depending on what households have to pay themselves and what is publicly available. Transport and communication needs are often specified as the average transport and communication costs from household surveys and then validated with the communities; communication needs can also be specified as the cost of a SIM card with a specific amount of data per household per month. Overcoming some challenges: constructing a hybrid MEB These two different approaches have advantages and disadvantages. The advantage of the expenditure-based approach is that it is fairly straightforward to carry out if survey data on the population are available. The disadvantage is that it can be difficult to put into practice when the target population is generally poor (such as in a refugee situation prior to assistance), so that the number of households who have sufficient expenditures to be just above the poverty line is very small. Also, it does not facilitate sector-specific interventions, which are often desired in humanitarian contexts. Finally, it does not have a rights-based perspective. For example, the food patterns for those “just able to meet their essential needs” are typically just taken as they are with no nutritional requirements reflected in the reference basket.16 The advantage of the rights-based approach is that it can be carried out without survey data at the construction stage (survey data is needed however to monitor the MEB). The disadvantage of the rights-based approach is that even if it is based on assessed needs, the effective demand of households can look quite different from the “needs” that results from this approach. 16 A study conducted in Nepal showed that the food poverty line is well below the so-called ‘nutrient poverty’ line, see Perrine Geniez, Astrid Mathiassen, Saskia de Pee, Nils Grede, and Donald Rose (2014). Integrating food poverty and minimum cost diet methods into a single framework: A case study using a Nepalese household expenditure survey. 14
MINIMUM EXPENDITURE BASKETS In the original MEB constructed for the Syrian refugee operation in Turkey, the share devoted to education expenditure was 17 percent while the average education expenditure share in the pre- assistance expenditure data is under 2 percent (this does not vary much with household vulnerability status). The large expenditure share in the MEB reflects the costs for transportation to schools in rural areas where no buses are available and where the only way for households to send children to school is to hire private transport. To assure children’s right to education, the transport costs are counted for in the MEB. Since the MEB was constructed to assure full access to all rights, the high education expenditure is justified. However, this can be problematic if we want to compare the theoretical costs for basic needs as estimated by humanitarian partners with the actual consumption choices of households. Even if households are assisted, there is nothing to say that they actually start hiring private transport to get their children to school, i.e. the principal ‘need’ identified by humanitarian actors will not necessary translate into effective demand if the MEB is used as a basis for transfer value calculations. While an important access problem has been identified, other complementary interventions will likely be needed to address it. If there is a large difference between the constructed MEB and actual expenditure patterns, it will be difficult to compare the two once an operation is in place, and household expenditures will need to be monitored against the MEB. Another disadvantage is that a rights-based MEB can easily become an instrument for different partners to compete for and secure funding. There is a substantial incentive to include excessively high sectorial needs if sector-specific interventions are envisaged. Finally, if the MEB is very detailed but the expenditure module in the household surveys is very crude, comparison is difficult and therefore the practical use of the MEB can be limited. Many households may fall below the poverty line simply because of the discrepancy in methods between the MEB and the survey data collection rather than for any other reason. This is very similar to the issue previously discussed regarding the use of national poverty lines. The non-food component in the Turkey MEB includes the items outlined in Table 1. The WFP expenditure module asks about expenditure on non-food items in two questions: (1) hygiene items and (2) ‘other’ (including clothing, shoes, tobacco etc.). This difference in detail between the MEB and the expenditure data may make it difficult to evaluate the expenditure data against the MEB. There is no straightforward solution to the problems outlined above but it is good to keep them in mind when deciding which method to use to establish the MEB. An inflated MEB is of limited use for operational and monitoring purposes, thus WFP should always advocate for a realistic, operational MEB. One way to ease some of the concerns and make the MEB operationally useful is to combine the two approaches, keeping the rights-based lens but also making sure that the MEB is consistent with demand behaviour. Of course, this is subject to the availability of expenditure data. As stated above, the MEB should only capture needs that households will realistically cover through the market; it should not be an attempt to monetize all the needs of a population. If education is difficult to access 15
MINIMUM EXPENDITURE BASKETS for the target population (using the example above), adding an education component in the MEB to facilitate education support will not necessarily solve the access problem. If electricity is not available for the target population, the MEB cannot include electricity even if in some contexts it should be considered an essential need. A food basket component that corresponds to an ideal basket may or may not achieve adequate nutrition when it is operationalized because of the consumption choices of households and food allocation within households. Other complementary nutrition interventions may be needed instead, such as providing certain nutritious foods for specific target groups such as young children (6-23 months), in-kind or through a commodity-specific voucher, and social behaviour change communication to raise awareness and stimulate people to make better choices for health and nutrition. Just because a need has not been captured by the MEB, does not mean that this need should not be considered an essential need and supported through complementary interventions. As shown in Table 3, the approach to take when constructing a MEB will depend on the purpose of the MEB, the possibility of collecting data and the status of the population to be assisted. Sometimes an expenditure-based approach is the best solution, but if a multi-stakeholder response is envisaged, a hybrid MEB is likely to be the best solution, as long as data are available. 16
MINIMUM EXPENDITURE BASKETS Table 3. Approaches to establishing MEBs, data requirements, pros and cons, and when to use which approach Approach Data requirements Pros and cons When to use + straightforward to carry out - problematic when everybody is - when WFP is the only actor CFSVA, EFSA, household pre- poor - when WFP wants to monitor Expenditure-based MEB assistance baseline data or PDM, - might not reflect essential needs expenditure patterns in relation to survey data from partners from a rights-based perspective the MEB + survey data is not needed - if a multi-stakeholder response - effective demand can be different envisaged Detailed prices/cost for food and from assessed needs > comparison Rights-based MEB - when “everybody is poor” and non-food expenditures with monitoring data is hard survey data are not available - big incentives for partners to inflate sector-specific needs Detailed prices/cost for food and - if a multi-stakeholder response + combining the rights-based lens non-food expenditures, CFSVA, envisaged and survey data are with an approach consistent with Hybrid approach EFSA, household pre-assistance available demand patterns baseline data or PDM, survey data - when “everybody is poor” and - data intense from partners survey data are available 17
Other important considerations ➢ Consider price changes over time To be operationally useful, the MEB must be updated over time to account for price changes faced by the target group. If inflation is high, this might have to be done on a monthly basis, if it is low, once a year might be enough. This should be planned for when the MEB is constructed to assure that the costs of the MEB components can be updated. A simple solution is to adjust the MEB with the national/sub-national CPI or its components. However, in some crisis contexts, CPIs are not updated or relevant for the target population. Urban areas are often over-represented in the national CPI or prices and costs faced by, for example, displaced populations can be very different from the national price level. In poor contexts where food constitutes a large part of the household expenditures, the price developments of food and fuel are central when it comes to capturing price changes. Based on WFP food price data collection for basic food items (and fuel), a price index for key items can be constructed and used to estimate cost changes in the MEB. In contexts where shelter is a major part of household expenditures, development in shelter costs should also be captured. Sometimes the main purpose of the MEB is to monitor price developments in the absence of a CPI. The main purpose of the MEB in Somalia is to construct a monthly price index in the absence of an official CPI. Price monitoring for a basket of basic food and non-food items is performed. The MEB represents a set of food items comprising 2,100 kilocalories/per capita/day and non-food items such as such as water, kerosene, firewood, soap and cereal grinding costs. The MEB contains four sub-baskets; two baskets cover the rural and urban towns in the North West and the other two cover the rural and urban towns in the rest of the country. ➢ Adjust for regional or urban/rural price differences if needed If beneficiaries are concentrated in one area, there is clearly no need to adjust for regional price differences. However, if they are spread out in urban/peri-urban and rural areas throughout the country, adjusting for differences may be vital. This means that different MEBs have to be constructed for different regions or for rural or urban/peri-urban areas. There are a few approaches for this. 1. Price the baskets based on available price data in different regions or urban/rural areas. For the food basket, this is possible using the VAM price database or other similar price series. For non- food items, housing, utilities and services, this can be more challenging and may rely on price data collection by partners or require new data collection. 2. For some countries, price data provided by the national statistical office are useful. In the case of Turkey, regional purchasing power parity indices were used to provide price estimates for components of the MEB for which direct price information was not available. 3. Use approximations from expenditure data. If the household survey has sufficient regional coverage, the expenditure levels in different regions can be explored, using the cohort of households just above the poverty line. Care should be taken in using this method, particularly if the sample size is very small by region.
➢ Account for differences in household composition and economies of scale in household consumption In the context of food assistance, households are most often given assistance according to their size on a per capita basis (even though different household members have different needs). When establishing a MEB, this is problematic. The needs of a household grow with each additional member but, due to economies of scale in consumption, not in a proportional way. Some goods consumed within a household, such as food, are “private“ in character - once a person has consumed it, no one else can consume the same - while there are other goods that are “public”, as they can be consumed commonly among household members. Hence, needs for housing space, electricity, etc., will not be three times as high for a household with three members than for a single person. The aggregation of many needs will compound the problem if they are treated per capita and are particularly problematic in contexts where food needs do not constitute a major part of the household essential needs. Using a per capita based MEB to facilitate targeting or transfer value calculations is not recommended as big households will always be found to be vulnerable by construction as the per capita approach assumes that the household expenditures increase proportionally for each added household member. In many contexts, this is implicitly dealt with by calculating a MEB for only the most common household size. However, if the MEB is to be used for operational purposes, this is not very practical. Figure 2 shows expenditures by household size compared to one-person households based on the Vulnerability Assessment of Syrian Refugees in Lebanon (VaSyr) 2017. Only when the household size reaches five does the average expenditure double compared to a one-person household and it takes eleven members to triple the expenditures of a one-person household. Figure 2: Increase in household expenditure by household size compared to one-person households 3.5 3.0 Multiplication factor Double 2.5 Triple 2.0 1.5 1.0 1 2 3 4 5 6 7 8 9 10 11 Household size Total Food Non-food Outside the humanitarian world, the most common solution to this problem is to use equivalence scales. Mostly, these do not only take into account economies of scale, but also different needs of children and adults. A typical equivalence scale measures the number of adults (sometimes adult 19
males) deemed to constitute a household. For example, a child under 15 is counted as a fraction of an adult (e.g. 0.5). The effective household size is then the sum of these adult-equivalent fractions.17 However, not only the household size needs to be translated into adult equivalents but also the quantified needs, i.e. the MEB (see ANNEX 2 for an example using data from Lebanon). While not very complicated as such, using equivalence scales might prove challenging since the same approach has to be used in both assessments and monitoring and ideally also be translated into operational decision making. The alternative approach, recommended here, is relying on the fact that while WFP often (but not always) gives assistance on per capita basis, many partners give assistance per household, regardless of household size. Thus, a pragmatic solution is to calculate the food component in the MEB on per capita basis and the rest of the needs as a flat rate regardless of household size. This approach was taken for the calculation of the multi-purpose cash assistance in Lebanon (although the MEB itself is still for a five-person household). This will in practice balance the fact that household needs do increase as members are added but not in a proportional way. A variation of this approach is to calculate the food component on per capita basis and a household-size specific non-food component as described in ANNEX 1 for the MEB in Kinshasa. The disadvantage of this pragmatic approach is that the MEB will vary by household size, but given its simplicity, this is still the most preferred approach. ➢ Consider seasonal variations if significant In many countries where WFP works, seasonal price variations are substantial. This calls for considering the construction of different MEBs for the harvest and the lean season. In other contexts, household needs change with the seasons. In Turkey where winters are cold, households have additional needs for heating and warm clothes to survive. If a survey is undertaken when prices are high or winters are cold without adjusting the MEB, it is likely that there is a decrease in the percentage of people below the MEB as households have higher needs or are confronted with higher prices and thus have higher expenditures. In Turkey, it was estimated that household needs during the winter would result in a 48 percent increase in minimum expenditures. ➢ Validate with stakeholders and make reality checks In most context the construction of the MEB take place in an interagency context, implicitly resulting in a validation process. However, not all clusters might be engaged in the Cash Working Group, resulting in limited by-in and understanding of the MEB unless they are specifically consulted. Government stakeholders and development partners are another essential group unless they already have been involved in the process. Endorsement by government counterparts will be needed because of existing government safety nets or policies regarding minimum wages. If, for example, the targeted populations are refugees or IDPs and the transfer value that is established based on the MEB is higher than social assistance provided by the government to its own population, there is a high risk of tensions with the government. Development partners might on their side wonder why a MEB is needed in addition to the national poverty line. Dialoguing with partners such as the World Bank is therefore essential. In addition to government and partner validation, general reality checks are 17 One common equivalence scale is the OECD scale: it assigns the weight 1 to the household head, 0.7 to all additional adults, and 0.5 to all children. A household with five people, say, two adults and three children, consists of 3.2 adult equivalents (1+0.7+0.5+0.5+0.5). This is a common scale used in many developing and developed countries. Another common scale is to give weight 1 to each adult and different weights to children depending on their age. For the official poverty line in Zambia, the following weights are given to children: 0-3 years: 0.36, 4-6 years: 0.62, 7-9 years: 0.76, and 10- 12 years: 0.78. 20
essential. How does the constructed MEB compare with teacher salaries or typical daily labour rates? For the MEB to be of practical use in operational contexts these checks are essential. How to construct a survival minimum expenditure basket (SMEB) In addition to the MEB, a SMEB is often constructed. A SMEB can serve at least two purposes. First, together with the MEB, it can be used to classify households into different vulnerability categories. Households under the SMEB are classified as severely vulnerable, households between the SMEB and the MEB as moderately vulnerable, and households above the MEB as not vulnerable. This information can then be used for prioritization of beneficiaries or for monitoring purposes. Second, in some cases where the MEB has been considered as “too high” to facilitate transfer value calculations, compared to national social assistance, the SMEB has instead been used. The principal difference between a MEB and the SMEB is that while the MEB implies full access to rights as defined by humanitarian law, the SMEB is just the minimum amount required to maintain existence and cover lifesaving needs. In the context of national poverty lines, a distinction between a full and a food poverty line is often made. In the humanitarian context, the distinction between the MEB and the SMEB has been very different in different contexts, to the point that they might not be useful to use as a reference. The experiences so far within WFP mostly come from the Syrian refugee context. ANNEX 1 illustrates how a MEB and a SMEB in Kinshasa were constructed. In Jordan, the SMEB includes food, shelter, water and sanitation. In Lebanon, only health and education are excluded from the SMEB while other needs are covered with smaller amounts in the SMEB compared to the MEB. For example, the SMEB has a less diverse food basket than the MEB, corresponding to the WFP food basket. Debt repayments are also covered in the SMEB. In Turkey, a similar approach to that in Lebanon was taken. The SMEB in Turkey relies more heavily on actual expenditure patterns than the MEB. The recommendations here are the following: 1. Before constructing a SMEB in an interagency context, the population of interest should be consulted. According to them, what do households need at the very minimum to maintain existence and cover lifesaving needs? 2. If there is no interagency SMEB or there is agreement among partners, the MEB/SMEB distinction can follow that of many national poverty lines, i.e. the SMEB corresponds to the minimum food expenditure basket. How to find a proxy for a MEB when data or time is insufficient A full MEB can be challenging to construct in a sudden onset emergency or if data are scarce or unavailable. Here are some thoughts on how this could be resolved. However, from a ‘do no harm’ perspective, it is important to underline that proxies should only be used on interim when no other solutions are available. 1. Use the national MEB. 21
2. A MEB is a good preparedness measure and should be constructed before an emergency. While both prices and availability will be affected by an emergency, it is still likely to provide a useful starting point. 3. The expenditure-based MEB essentially only requires (a) an approximate value of the food basket, and (b) an estimate of the average expenditure share that households use on food. Even if no survey data are available, this information should be available to the Country Office or can rapidly be collected or approximated. 4. Consider using the minimum wage as a proxy. Bear in mind that while the MEB captures household-level needs, the minimum wage is individual-level income so an assessment of how many minimum wages are needed per household depending on the household size is required. It is also advisable to find out how the minimum wage has been constructed. The advantage with this approach is that it is aligned with government approaches. Questions or comments? We are keen to learn more from you. Please contact the VAM or CBT Programme teams in HQ for any questions or comments. 22
ANNEX 1 The Minimum Expenditure Basket in Kinshasa The MEB in Kinshasa was constructed by WFP as a part of an urban assessment focussing on household essential needs. A hybrid approach (mix of an expenditure-based and a rights-based approach) was applied to construct the MEB. Annex Table 1 shows how the MEB components were determined and how the expenditure components included in the MEB/SMEB were analysed. These are further discussed below. Annex Table 1: MEB and SMEB components and how they were derived MEB Subcomponents Method Calculation Household Inclusion in component size MEB and SMEB adjustment Food • Maize meal Expenditure- Mean Per capita MEB: Inclusion • Cassava based approach expenditure for adjustment of all food • Beans for each the cohort of items that are • Vegetable oil subcomponent, households with being • Fish scaling to 2100 Food consumed and • Sugar kcal per person Consumption calorie-scaling • Vegetables per day Score < 70 and on all calorie- • Fruits reduced coping relevant items • Dairy strategy index < SMEB: 15 • Condiments Inclusion of cereals, tubers, beans, oil, scaled to 2100 kcal. Rent 2-room housing Rights-based, Typical monthly 1 room for No using qualitative rent of a 2-room 1-2 person information apartment HH, 3 rooms for 8+ HH Education • School fees Rights-based Median school None No • School meals (exp.-based fees and material and material determination cost per month of amounts) per child Health • GP’s visit Rights-based, Median cost of Per capita No • Over-the- using qualitative one doctor’s visit counter information per person per medication year Monthly amount for OTC medication Other non- • Water Expenditure- Median Adjustment SMEB includes food Items • Cooking fuel based expenditure of using an only water, • Lighting (subcomponent median expenditure- cooking fuel, • Hygiene s included in household size based lighting, products line with what with Food household- hygiene • Electricity typically is Consumption size products defined to be Score < 70 and 23
• Transport part of essential reduced coping adjustment • Clothes needs) strategy index < scale • Communicati 15 on For the expenditure-based components of the MEB, the aim was to analyse households that can generally be considered above the poverty line, while at the same time not being too wealthy. For this MEB we chose a theoretically justifiable, yet somewhat arbitrary upper and lower thresholds of household exclusion. On one hand, we are excluding households with a Food Consumption Score over 70 (a mid-range value between the threshold for adequate diet and the maximum value), thereby filtering out the households that are consuming an unusually rich and expensive diet with daily intake of several protein-rich food items. On the other hand, we are excluding households with a food-based coping strategy index above 15, so as to ensure that households with highly negative coping behaviour are not included as reference households for the MEB threshold. For the adjustments to smaller and larger households, we are advocating for a per-capita adjustment for the food basket and a specific household size adjustment scale for the expenditure-based non- food items. For the rights-based parts, we will argue for specific adjustments. Food Basket We distinguish between a standard food basket that reflects actual consumption patterns as observed in the selected cohort of households (MEB), and a survival basket ensuring macronutrient sufficiency (SMEB). Annex Table 2: Calculation of the food basket for Kinshasa MONTHLY URBAN FOOD BASKET average kg food- food- Food item caloric value Survival basket elements Kg consumed/month Kcal consumed/ day consumed / month Kg/month rescaled basket Kcal/day rescaled SMEB MEB capita spending Rounded Rounded Kg price Survival basket basket /100g Per Food Item Maize 6898 900 7.7 360 920 982 8.2 7400 982 8.7 8000 Cassava 1808 500 3.6 342 412 440 3.9 1900 440 4.1 2100 Beans 2089 1300 1.6 335 179 192 1.7 2200 192 1.8 2400 Veg Oil 2341 2300 1.0 890 302 322 1.1 2500 322 1.2 2700 Fish 4306 2500 1.7 76 44 47 1.8 4600 Sugar 1817 2200 0.8 400 110 118 0.9 1900 Vegetables 2826 2800 Fruits 833 800 Dairy 799 800 Condiment 2509 2500 Meals Outside House 2144 Total 28371 Total kcal 1967 2100 27400 1936 2100 15200 24
The MEB and the SMEB food basket calculation are described in Annex Table 2. The process of deriving the final amounts for each food item is as follows: Annex Figure 1: MEB and the SMEB food baskets for Kinshasa (Congolese Franc) 1. For each of the food expenditure categories, 30,000 27,400 we decide whether they have a significant caloric relevance in the overall food intake of a 2,500 25,000 person. In the case of Kinshasa, we decided that cereals, tubers, pulses, oil, animal protein and 2,800 sugar contribute significantly to people’s calorie 20,000 1,900 intake. 15,200 4,600 2. For each of the calorie-relevant food groups, 15,000 we select the most commonly consumed food 2,700 2,500 item of this group in the local context. For 2,400 10,000 2,200 Kinshasa, maize meal is used as the cereal, 2,100 1,900 cassava flour for tubers, red beans for pulses, vegetable oil for the oil component, and white 5,000 8,000 7,400 sugar. As part of our market visits, retail prices for these food items were collected. - 3. Knowing the average per capita expenditure Maize Manioc from the household survey and the market price on each component, we can approximate the Beans Veg Oil quantities consumed per person per month. Fish Sugar Next, we can calculate the daily caloric value Vegetables Fruits associated with these quantities and derive the Dairy Condiment total kcal value of this average basket. In the case of the urban zone of Kinshasa, this amounts to 1967 kcal, which is good indication that the expenditure data is likely to be fairly accurate. 4. Nevertheless, we perform an adjustment of this basket by scaling the caloric values proportionally to ensure the sum of all food items reaches the standard daily average caloric requirement of 2100 kcal.18 On the basis of this caloric value, we then convert all values back to monthly data to derive a monetary value for each food item. 5. In addition to the calorie-relevant food items, we add the mean expenditure on other food categories that households regularly consume. As these food items have little overall caloric share on people’s diets (given the quantities consumed or the type of the foods), we do not include them when we rescale the calories. However, we can assume that most households would still consume these food items as part of their usual diets. In addition, even if these food items have little significance for overall calorie intake, they are very important as part of 18 In the guidance, it is argued that WFP data cannot be used for establishing whether households consume enough calories or not, yet here the data is used to establish an average approximate food basket of 2100 Kcal. The difference here is that the data is only used to construct an average basket which is different from judging the sufficiency on individual household level. 25
You can also read