Estimating the prevalence of food insecurity of households with children under 15 years, across the globe

Abstract

Target 2.1 of the Sustainable Development Goals (SDG) calls to end hunger in all its forms by 2030. Measuring food security among children under age 15, who represent a quarter of the world’s population, remains a challenge and is infeasible for global monitoring. The SDG framework has agreed to use the Food Insecurity Experience Scale (FIES) to measure moderate and severe food insecurity. Using nationally-representative data from the Gallup World Poll (GWP) survey in 2014–15, we provide the first global and regional estimates of food insecurity among households with children under age 15. In addition, we test the robustness of the FIES against 1) monetary poverty and 2) the Negative Experience Index, a measure of well-being. Finally, we explore trends in per capita income as a determinant of food security (2006–2015) to observe how this relationship fluctuated during the Great Recession.

We find that across 147 countries and four territories, 41% of households with children under age 15 suffer from moderate or severe food insecurity, 19% from severe food insecurity, and 45% reported not having enough money to buy food in the previous 12 months. The relationship between food insecurity, poverty, and well-being varies by region, demonstrating that definitions of food insecurity depend on regional context, and encompass more than monetary poverty alone.

Our findings will ideally encourage and provide motivation for continued global efforts to address food insecurity and monitor progress towards SDGs.

1. INTRODUCTION

To end hunger and achieve food security for all is a prominent target under the Sustainable Development Goals (Target 2.1) (United Nations Economic and Social Council, 2016). From that global commitment follows the challenging task to estimate national prevalence rates of food insecurity: estimates that should be comparable across countries and population groups, and over time.

Holistic measures of food insecurity would ideally assess a broad range of causes, such as food utilization or food quality, and outcomes, such as poor health or shame associated with challenges in accessing food. More specifically, definitions and measures of food insecurity are moving away from focusing on siloes of access, availability and utilization, and instead, accounting for factors such as food sufficiency, nutrient adequacy, cultural acceptability, safety, and certainty and stability of foods (Coates, 2013). Accordingly, the use of indices or scales has become a more common approach to capture these different dimensions, leading to a more comprehensive understanding of food insecurity. Further, because food insecurity risks vary at the individual, household and community levels in different settings, developing a method to comparably measure food insecurity across groups and over time remains a challenge. To track such progress, the global SDG indicators framework includes the Food Insecurity Experience Scale (FIES) to estimate the prevalence of moderate and severe food insecurity (Indicator 2.1.2).

In 2013, the Food and Agricultural Organization of the United Nations (FAO) launched the Voices of the Hungry (VoH) project to monitor food insecurity worldwide (Cafiero, Viviani & Nord, 2018). The FIES was developed to not only accurately measure the severity of food insecurity at the individual or household level, but also provide comparisons of food insecurity across countries and over time. The FIES is based on three existing tools that are used to measure food insecurity in household-based surveys: the United States Household Food Security Survey, the Household Food Insecurity Access Scale, and the Latin American and Caribbean Food Security Scale (Escalal Latinoamericana y Caribena de Seguridad Alimentaria – ELSCA). The FIES is an experience-based metric that reports food-related behaviours on the inability to access food due to resource constraints (Ballard, Kepple, & Cafiero, 2013). In 2014, the FIES was introduced in the Gallup World Poll (GWP), a large-scale population-based annual household survey, covering nearly all countries in the world. For the SDGs, it is of particular interest as it is one of very few indicators that may deliver universal country coverage with annual updates, while reflecting a fundamental quality of human life with strong links to various dimensions of wellbeing. Current global rates of food insecurity are estimated at 27%, with the highest burden in low-income economies, followed by lower-middle income, upper-middle income, and high-income economies (Smith, Rabbitt & Coleman-Jensen, 2017). However, a severe limitation of the FIES estimate is that it is derived from respondents age 15 years and over, implying that individual-level responses from the world’s children below age 15 are left out.

Children below age 15 constitute more than a fourth of the world’s population (World Bank, 2015). Food insecurity has both nutritional and non-nutritional consequences on child well-being. Children who are exposed to food insecurity are more likely to face adverse health outcomes and developmental risk (Cook et al., 2004; Howard, 2011; Rose-Jacobs et al., 2008). Food hardship among children also predicts impaired academic performance, and is positively associated with behaviour problems and experiencing shame of being out of food (Bernal, Frongillo, & Jaffe, 2016; Jaffe, Bernal, & Herrera, 2014; Jyoti, Frongillo, & Jones, 2005; Slack & Yoo, 2005). As experiences of food insecurity can be particularly critical during developmental phases, children cannot be disregarded when monitoring this particular SDG target. However, asking children directly to find out about their experiences of food insecurity, which would have been ideal, is not a viable option in the context of global monitoring, and is infeasible due to a lack of large-scale comparable surveys administered to older children directly. An alternative option would be to estimate the numbers and shares of children below age 15 who live in households with a respondent who is food insecure. Limiting the sample to households with children may account for intra-household inequalities in resource allocation that may not be present in households without children. The objective of this paper is to produce such estimates, determine how the resulting picture differs from the corresponding prevalence rates among adults, and assess the properties and robustness of these estimates by exploring the correlations with a range of alternative wellbeing measures.

We develop the first global estimates of food insecurity among households with children under age 15 using nationally-representative GWP data from 2014 and 2015, the first two years during which the FIES module was administered, and for which data were available at the time of the analysis. We first test the robustness of the FIES measure in comparison to three indicators: 1) the standard GWP measure of food insecurity (“Was there ever a time in the last 12 months when you did not have enough money to buy food?”), 2) household income per capita and 3) the Negative Experience Index (NX Index), a composite measure of respondents’ feelings from any potential negative experiences the previous day. We then present regional estimates of food insecurity among households with children under age 15 and compare some of our country estimates of food insecurity to data from national surveys. Finally, we explore the relationship between household income per-capita and the GWP food insecurity indicator over time (2006–2015), with a particular interest in the period of the Great Recession, which saw highly volatility in global food prices. We expect regions that were hit harder by the food price shocks and the Recession to show higher levels of sensitivity of food insecurity to income.

2. DATA AND METHODS

2.1. Gallup World Poll

Since 2005, the GWP has conducted standardized cross-country surveys in over 160 countries, making it arguably one of the highest frequency and coverage micro-data available. The GWP assesses attitudes and behaviours on topics including well-being, food access, and satisfaction with communities and governments. Approximately 1,000 individuals aged 15 and over in each country are surveyed based on randomly-selected nationally-representative samples. The GWP uses multi-stage sampling, first stratifying countries by population size, then randomly selecting households from each sampling unit, and consequently randomly selecting one individual aged 15 or over in the household for the interview. Sampling weights are applied to make the final sample representative of the total population aged 15 and over. The same core questionnaire is translated into major languages of each country to permit cross country comparisons. Telephone surveys are conducted in countries where telephone coverage exceeds 80% of the population. The core GWP questionnaire includes one question on food security “Was there ever a time in the past 12 months when you did not have enough money to buy food?” For further details about methodology and questionnaire design, please refer to Gallup (2015).

2.2. The Voices of the Hungry (VoH) project and Food Insecurity Experience Scale (FIES)

The FIES has been administered through the GWP since 2014 (Cafiero, Viviani & Nord, 2018). Table 1 details the eight dichotomous questions relating to food insecurity experiences in the preceding 12 months used in the FIES. Individuals are classified based on the total number of affirmative responses ranging from 0–8. However, since the severity and risks of food insecurity experience are expected to vary in different settings and contexts, the raw scores are likely to be incomparable both across time and countries. Due to this incomparability, the raw scores are equated to a global standard to develop the probability of being food insecure at different levels of severity and allow for cross-country comparisons, using Rasch modelling techniques (Rasch, 1960). These probabilities are classified as moderate or severe (FIES-M+), and severe (FIES-S) food insecurity. Figure 1 shows the proportions of respondents by FIES raw scores, and the corresponding mean FIES-M+ and FIES-S proportions for the total sample. The proportions of the adult population at each raw score (summed from 0–8) is shown by the bar graphs, while the FIES-M+ and FIES-S for each raw score total are displayed by the dashed and dotted lines respectively. Thresholds for moderate and severe food insecurity have been set by the VoH project. More information on the development and calculation of thresholds can be found in the VoH technical report (FAO, 2015a).

Table 1:

Questions in the Food Insecurity Experience Scale Survey Module for individuals as fielded in the 2014 Gallup World Poll

Now I would like to ask you some questions about food. During the last 12 months, was there a time when...
Q1	(worried)	... you were worried you would not have enough food to eat because of a lack of money or other resources?
Q2	(healthy foods)	... you were unable to eat healthy and nutritious food because of a lack of money or other resources?
Q3	(few foods)	... you ate only a few kinds of foods because of a lack of money or other resources?
Q4	(skipped meals)	... you had to skip a meal because there was not enough money or other resources to get food?
Q5	(ate less)	... you ate less than you though you should because of a lack of money or other resources?
Q6	(ran out of food)	... your household ran out of food because of a lack or money or other resources?
Q7	(hungry)	... you were hungry but did not eat because there was not enough money or other resources for food?
Q8	(whole day without eating)	... you went without eating for a whole day because of a lack or money or other resources?

Notes: For detailed information on the FIES please refer to the Voices of the Hungry Technical Report (FAO, 2015a)

Figure 1: Proportions of respondents by FIES raw score, mean FIES-M+ and mean FIES-S.

Notes: Proportions and weighted means of FIES-M+ (moderate and severe) and FIES-S (severe) come from the Gallup World Poll and include 289,933 observations (all households) from 147 countries and 4 territories in 2014 and 2015.

2.3. Methods

For the child food insecurity estimates, we use 2014 and 2015 GWP data for 147 countries and four territories from 12 regions: Eastern and Southern Africa, Horn of Africa, West and Central Africa, Central America, North America, South America, East Asia and the Pacific (EAP), South Asia, Southeast Asia (SE Asia), European Union and non-Commonwealth of Independent States (EU/non-CIS), Commonwealth of Independent States (CIS), and Middle East and North Africa (MENA). Household food insecurity estimates are calculated using household weights, for all three food insecurity indicators. We use child weights to estimate the share of children under 15 years living with a respondent who is food insecure, and calculate this among households with at least one child under age 15 years, to account for households that have more than one child. Population estimates of the number of food insecure children in our country sample are calculated using the proportion of children under age 15 in each country from the World Bank Development Indicators and total population numbers provided from FAOSTAT (FAO, 2015b; World Bank, 2015).

The NX Index is in the core GWP questionnaire, and is a composite measure of respondents’ negative experiences from the day before the survey, on experience of five feelings: 1) physical pain, 2) worry, 3) sadness, 4) stress and 5) anger (Gallup, 2015). Log of household income per capita is measured in international dollars. To estimate the relationship between food insecurity and log of household income per capita over time (in International dollars), we used Ordinary Least Squares (OLS) regressions, controlling for age, gender, education (of the main respondent), log of household size, rural residence, and regional fixed effects. We present the results of the coefficient for the log of household income per capita and food insecurity for 2006–2014 globally and from 2007–2014 in South Asia and Sub-Saharan Africa.

Qatar and Turkmenistan were excluded from the analysis because the FIES was not fielded there, although the GWP was implemented there in 2014. Afghanistan (2015) and Botswana (2014) were excluded from the analysis because of data missing on the number of adults in the household (GWP indicator: WP12) needed to create the appropriate child weights (Gallup, 2015). Turkey (2014) was excluded for the same reason because of data missing on the number of children in the household (GWP indicator: WP1230) (Gallup, 2015). Not all countries had available data for both years. A complete list of countries, regional groupings and years of data availability is provided in Appendix 1.

3. RESULTS

3.1. Descriptive statistics of sample

The full sample included 289,933 respondents [approximately half (51% or 148,848 respondents) live in households with at least one child under age 15] from 147 countries and 4 territories (Hong Kong, Kosovo, Northern Cyprus, and Puerto Rico). The proportion of households with children under 15, varies by region, ranging from 27% in the European Union/CIS to 80% in the Horn of Africa. Table 2 provides descriptive statistics of our samples of interest, households with at least one child under age 15 compared to all households. Globally, the FIES-M+, FIES-S and the GWP indicator are higher among households with children under 15, averaging 41%, 19% and 45% respectively, compared to 27%, 11% and 32% respectively among all households. On average, respondents in households with children under 15 were younger (36 years vs. 40 years), less likely to be male (46% vs. 49%), and less likely to have some secondary education or higher (57% vs. 73%). The log of household income per capita was slightly lower among households with children under 15 (6.52 vs. 7.54), but higher for log of household size (1.77 versus 1.37), compared to all households. Households with children under 15 were more likely to be rural (69% vs. 62%).

Table 2:

Descriptive statistics for respondents in households with at least one child under 15 years and all households

Variable	Mean (Standard Error)
	All Households	Households with children under 15 years
FIES-M+ (Moderate or severe)	0.27 (0.001)	0.41 (0.001)
FIES-S (Severe)	0.11 (0.000)	0.19 (0.001)
GWP indicator: Not enough money to buy food in the last 12 months	0.32 (0.001)	0.45 (0.001)
Age of respondent	39.50 (0.032)	35.58 (0.036)
Male	0.49 (0.001)	0.46 (0.001)
Education: Secondary or higher	0.73 (0.001)	0.57 (0.002)
Log of per capita income	7.54 (0.004)	6.52 (0.006)
Log of household size	1.37 (0.001)	1.77 (0.001)
Rural household	0.62 (0.001)	0.69 (0.001)
N	289,933	148,848

Note: Data comes from Gallup World Poll (GWP) Surveys and includes 289,933 respondents (148,848 respondents with at least one child under age 15 years in the household) from 147 countries and 4 territories in 2014 and 2015.

The Food Insecurity Experience Scale (FIES) was used to measure moderate or severe (FIES-M+), and severe (FIES-S) food insecurity. The GWP food insecurity indicator is: “Was there ever a time in the last 12 months when you did not have enough money to buy food?”

Estimates are weighted means and the corresponding standard errors.

3.2. Comparing different measures of food insecurity

Figure 2 shows the global and regional prevalence of food insecurity for households with children under age 15, using the FIES-M+, FIES-S, and the GWP indicator, “not enough money to buy food.” The FIES-M+ is the highest in Eastern and Southern Africa (68%), and the lowest in EAP (9%), while the FIES-S is highest in the Horn of Africa (41%) and lowest in EAP (2%). Similar to the FIES-M+, the GWP indicator is the highest in Eastern and Southern Africa (66%) and the lowest in EAP (15%). Across all regions, the magnitude of the estimates of the GWP indicator are more similar to that of the FIES-M+, than FIES-S measure, and show higher food insecurity than the FIES-M+ in all regions except Eastern & Southern Africa, and the Horn of Africa.

Figure 2: Regional estimates of food insecurity prevalence among households with children under 15 years.

Notes:

Data comes from Gallup World Poll (GWP) Surveys and includes 289,933 respondents [148,848 respondents with at least one child under 15 years in the household (HHU15)] from 147 countries and 4 territories in 2014 and 2015. The Food Insecurity Experience Scale (FIES) was used to measure moderate or severe (FIES-M+), and severe (FIES-S) food insecurity. The GWP food insecurity indicator is: “Was there ever a time in the last 12 months when you did not have enough money to buy food?” The GWP selects and surveys one respondent aged 15 years or over per household. Household weights are used for the estimates for all households, and child weights for the HHU15 estimates. The food security indicator is, therefore, the share of children under 15 years living in food insecure households, among households with at least one child under age 15 years.

Sub-Saharan Africa has the highest rates of food insecurity, with the FIES-M+, FIES-S and GWP indicator all above the global average. In the Americas, Central America has the highest prevalence of food insecurity measured by all three indicators (FIES-M+: 4%; FIES-S: 20%; GWP: 55%), followed by South America (FIES-M+: 29%; FIES-S: 10%; GWP: 37%), and then North America (FIES-M+: 23%; FIES-S: 8%; GWP: 27%). In Asia, South Asia has higher rates of FIES-M+ (29% vs. 26%) and FIES-S (12 % vs. 9%) compared to SE Asia, but a lower rate for the GWP indicator (37% vs. 39%). Within Asia and globally, the EAP region has the lowest reported levels of food insecurity, as measured by all three indicators (FIES-M+: 9%; FIES-S: 2%; GWP: 15%). Although lower than the global averages, food insecurity rates are still unacceptably in MENA (FIES-M+: 29%; FIES-S: 10%; GWP: 34%), EU/non-CIS (FIES-M+: 14%; FIES-S: 4%; GWP: 20%) and the CIS region (FIES-M+: 15%; FIES-S: 2%; GWP: 28%). Appendix 5 presents these estimates for all households; the estimates for all households are lower compared to households with children under 15. A list of prevalence estimates by country for all households and households with at least one child under age 15 years is available in Appendix 2.

3.3. Comparing food insecurity with monetary and non-monetary measures of wellbeing

Since the GWP indicator appears to be closer to the FIES-M+ in magnitude, compared to the FIES-S, we correlate the FIES-M+ and the GWP indicator, with income per capita and the NX Index to understand how well the food security indicators capture monetary poverty and a non-monetary measure of well-being. Figure 3 shows results for correlation analysis, where symbols represent the strength of correlation (ranging from 0, no correlation to 1, perfect correlation). Black circles represent the correlation between the NX Index and FIES-M+, and grey circles with the GWP indicator. Black triangles represent the correlation between household income per capita and FIES-M+, and grey triangles with the GWP indicator. Globally and on average, the FIES-M+ and GWP indicator are more strongly correlated with the NX Index (0.30 and 0.23 respectively) compared to household income per capita (0.14 and 0.13 respectively) (Figure 3).

Figure 3: Correlation of food insecurity indicators with income per capita and the Negative Experience Index among households with children under 15.

Note: Weighted means of FIES-Moderate+ and the GWP (“not enough money to buy food”) indicators come from the Gallup World Poll and include 148,848 observations from 147 countries and 4 territories in 2014–15.

Across all regions, the FIES-M+ is more strongly correlated than the GWP indicator to the NX Index. Correlations with the NX Index range from 0.25 in EAP to 0.36 in the Horn of Africa for the FIES-M+, and from 0.18 in the three Sub-Saharan African regions and EAP to 0.30 in SE Asia for the GWP indicator. When looking at correlations with income, GWP does do better (CIS, SE Asia and North America) or the same as FIES-M+ in five regions out of 12. Correlations are strongest between income and FIES-M+ in Central and South America and Eastern and Southern Africa, and globally the overall correlations with income are about the same for both indicators (0.14).

3.4. Prevalence and burden of food insecurity by region

In our sample of 147 countries and four territories, approximately 41% of households with children under age 15 suffered from moderate or severe food insecurity as measured by the FIES-M+, 19% from severe food insecurity, as measured by the FIES-S and 40% reported not having enough money to buy food in the previous 12 months. These estimates translate to roughly 605 million, 260 million, and 688 million children under age 15 who live in households that are moderately, or severely food insecure, or who lived in households that did not have enough money to buy food in 2014–15, respectively. However, the exact number of food insecure children may vary because there are food insecure children who live in food secure households and food secure children who live in food insecure households. Since we do not include countries without data, these estimates are likely to be underestimated at the global level.

Figure 4 presents regional food insecurity prevalence for households with children under age 15 (solid line) and all households (dashed line), and burden for children under age 15 (light shaded bar) and the entire population (sum of light and dark shaded bars) estimates. Although, as seen earlier, the three regions of Sub-Saharan Africa have the highest prevalence of food insecurity as measured by the FIES, South Asia has the highest burden, with 436 million people (162 million children under age 15) living in moderately or severely food insecure households. West and Central Africa (259 million people; 120 million children under age 15), East and Southern Africa (200 million people; 93 million children under age 15), and the Middle East and North Africa (136 million people; 50 million children under age 15) follow close behind.

Figure 4. Prevalence and burden of food insecurity by region.

Note: Data comes from Gallup World Poll (GWP) Surveys and includes 289,933 respondents [148,848 respondents with at least one child under 15 years in the household (HHU15)] from 147 countries and 4 territories in 2014 and 2015. Total population estimates were taken from the Food and Agricultural Organization of the United Nations Statistics for 2014 and 2015, except for Hong Kong, Kosovo and Northern Cyprus, which was taken from Wikipedia. Proportions of population under age 15 years were taken from the World Development Indicators, World Bank database, unless otherwise mentioned in Appendix 4.

3.5. Comparing food insecurity indicators to national data

In this section, we provide sensitivity analyses to understand how the GWP estimates compare to other national data. First, using data from Living Standards Measurement Study (LSMS) in East Africa (Malawi 2013, Uganda 2011–12, Tanzania 2012–13), we compare estimates of food insecurity in the last 12 months, the poverty rate, and the proportion who reported eating less than three meals a day among households with children under age 15 to the FIES-M+ and GWP indicator (Figure 5). The LSMS are conducted by national Governments in collaboration with the World Bank, nationally-representative and routinely used for monitoring key indicators, including income poverty and human capital. These countries were chosen because comparable age-disaggregated data on food insecurity and poverty was available from the Multiple Overlapping Deprivation Analyses (De Neubourg et al., 2013), and were the latest available estimates at the time of this analysis. In all three countries, the under age 15 poverty rates are among the lowest of all indicators at 43%, 26% and 34% in Malawi, Tanzania and Uganda respectively, reflecting a narrowly defined proxy measure of food insecurity. In all countries, the FIES-M+ produced the highest estimates of food insecurity (87%, 60% and 72% in Malawi, Tanzania and Uganda respectively).

Figure 5. Comparing age-specific measures of food insecurity and poverty in Sub-Saharan Africa.

Note: Weighted means of FIES-M+ and the GWP indicator (“not enough money to buy food”) come from the Gallup World Poll (2014/2015) and include 148,848 observations from households with children under 15, from 147 countries and 4 territories in 2014 and 2015. Under-15 data on food insecurity in the past 12 months, poverty rate and percent who had less than three meals per day come from the Living Standard Measurement Survey from each country [Malawi 2013 (N=8,817), Uganda 2011–12 (N=8,647) Tanzania 2012–13 (N=10,395)].

Secondly, we use data from the European Union Statistics on income and Living Conditions (EU-SILC) survey to provide comparisons in Europe. The EU-SILC is a comparable cross-national survey that provides timely and longitudinal multidimensional microdata on income, poverty, social exclusion and living conditions. Among children under age 15 in Poland, Portugal and the United Kingdom (UK), we look at the proportion who ate three meals a day, and the under-18 poverty rate (Figure 6), compared to the FIES-M+ and GWP food security indicator (both are under age 15 estimates). Despite the differences in age-groups for the indicators, the GWP indicator is the highest in all three countries: 24% in Poland, 25% in Portugal and 20% in the UK. The FIES-M+ is almost similar to the poverty rate in Poland (FIES-M+: 14%; poverty rate: 15%) and Portugal (FIES-M+: 18%; poverty rate: 17%), but more than twice the poverty rate in the UK (FIES-M+: 19%; poverty rate: 9%). In all three countries, the proportion who ate three meals a day produced the lowest estimates of food insecurity, at 1%, 2% and 4% in Poland, Portugal and the UK respectively.

Figure 6. Comparing age-specific measures of food insecurity and poverty in Europe.

Note: Weighted means of FIES-M+ and the GWP indicator (“not enough money to buy food”) come from the Gallup World Poll (2014/2015) and include 148,848 observations from households with children under 15, from 147 countries and 4 territories in 2014 and 2015. Data on percent of children in households where at least one child age 1–15 years did not eat three meals per day (2009) come from the European Union Statistics on income and Living Conditions (EU-SILC) from each country. Data on under-18 poverty rate are 2014 estimates from the Organisation for Economic Co-operation and Development (OECD) Income Distribution Database (IDD).

3.6. Income as a determinant of food security over time (2006–2014)

Figure 7a shows the trends in the global prevalence of food security measured by the GWP indicator, and plots the OLS regression coefficient of the predictor log of household income per capita, for all households and households with children under age 15 (dashed line and solid line respectively). For this part of the analysis, we use food security, measured “had enough money to buy food in the past 12 months” instead of food insecurity to better display the trends on the relationship between food security and income. In addition, since individuals aged 15–24 years were sampled for response to the survey, we also include a comparison to responses from this subsample (dotted line). Regressions control for age, gender, education (all of the main respondent), log of household size, rural residence, and regional fixed effects, however these are not reported in Figure 7. Weighted means of all covariates by year, for each of the groups (all households, households with at least one child under 15 years, and respondents aged 15–24 years) are available in Appendix 3. Across all countries in our sample, food security is more sensitive to income in almost all years among households with children under age 15, as shown by the magnitude of the regression coefficient, and less sensitive among the youth sample. Although prevalence of food security decreases after 2007, during the onset of the Great Recession, food security sensitivity to income peaks slightly in 2008, and then remains fairly constant, until 2011, when we see much greater sensitivity to income (spike in the magnitude of the regression coefficient). We also replicate the analysis focusing on two regions of interest, Sub-Saharan Africa (Figure 7b) and Asia (Figure 7c). In Sub-Saharan Africa, households with children are only slightly more sensitive to income, than all households or adolescents. Similar to global estimates, food security decreases in 2008, before improving in 2009, and decreasing again in 2010.

Figure 7a. Trends in global food security and log of household income per capita 2006–2015.

Note: Weighted means GWP food security indicator come from the Gallup World Poll surveys from 2006–2015 and include a total of 1,231,960 observations from all households, 642,638 observations from households with children under 15, and 253,838 children and youth aged 15–24 years.

The GWP food security indicator is “Did you have enough money to buy food in the last 12 months?” Regressions controlled for age, gender, education, log of household size, rural residence and regional fixed effects.

Figure 7b. Trends in food security and log of household income per capita in Sub-Saharan Africa (2007–2015).

Note: Weighted means GWP food security indicator come from the Gallup World Poll surveys from 2007–2015 and include a total of 246,315 observations from all households, 178,452 observations from households with children under 15, and 74,584 children and youth aged 15–24 years.

The GWP food security indicator is “Did you have enough money to buy food in the last 12 months?” Regressions controlled for age, gender, education (all of the main respondent), log of household size, rural residence and regional fixed effects.

Figure 7c. Trends in food security and log of household income per capita in Asia (2007–2015).

Note: Weighted means GWP food security indicator come from the Gallup World Poll surveys from 2007–2015 and include a total of 260,988 observations from all households, 139,527 observations from households with children under 15, and 49,482 children and youth aged 15–24 years.

The GWP food security indicator is “Did you have enough money to buy food in the last 12 months?” Regressions controlled for age, gender, education (all of the main respondent), log of household size, rural residence and regional fixed effects.

Simultaneously in 2010, food security becomes much more sensitive to income. In Asia, although food security remains fairly stable beginning in 2009, there is high volatility in food security sensitivity to income across all three groups, with peak sensitivity in 2008 and 2011.

4. DISCUSSION AND CONCLUSION

Our estimates present a baseline of the prevalence of food insecurity among households with at least one child under age 15 years compared to food insecurity in all households. Using data from 147 countries and four territories from the 2014 and 2015 GWP, we find that food insecurity, as measured by multiple indicators (FIES-M+, FIES-S, and the GWP indicator “did not have enough money to buy food in the last 12 months”), among households with children under 15 is higher across regions irrespective of the indicator used. This pattern of results is consistent with those for child poverty, where households with children across the globe are disproportionately more likely to be classified as poor or extremely poor, relative to households without children (Silwal et al., 2020; UNICEF & the World Bank Group, 2016), a result that holds even after adjusting for household equivalence scales (Newhouse, Suarez-Becerra, & Evans 2017). In our data, we note that households with children have structural features that make them more at risk of becoming food insecure, such as lower per capita household income, heads with lower education, and more members in the household.

Are reports of household food insecurity from adults an accurate reflection of children’s food insecurity? Findings on differences between adult/caregiver and child reports of food insecurity are inconclusive. Fram et al. (2013) found that parents underreport their child’s experience of food insecurity, compared to the child self-reports in the United States. In Zimbabwe, children were less likely to report food insecurity compared to adult self-reports, suggesting that children are protected from food insecurity in some instances (Kuku, Gundersen, & Garasky, 2011). However, the same study found that the protection did not extend to orphans, children in households with more than eight household members, or children in wealthier households. Factors such as age, orphan status, and gender of caregivers also influence the experiences of children within the same household. The findings from both studies reiterate the need for multiple measures of food insecurity within the household, largely because children and adults may experience food insecurity differently. Nevertheless, while asking children about their own experiences of food insecurity may seem ideal, the cognitive demands may be too high for very young children.

While income is a key component of food insecurity, food insecurity encompasses dimensions that go beyond just income. In our data, this dynamic is borne out by the stronger correlations between the FIES and the GWP indicator with the NX Index across all regions relative to household income, as well as the stronger correlation (relative to the GWP indicator) between the more nuanced FIES scale that captures different dimensions of food insecurity and the multidimensional NX Index. Yet, among households with children, income is a stronger correlate of food insecurity than among all households, and sensitivity of food insecurity to income was heightened immediately after the Great Recession. This excess sensitivity was particularly noticeable in low- and middle-income regions that were already suffering from food and energy price shocks during 2007–2008 (Verick & Islam, 2010). Comparisons with national-level data in Africa and Europe show that definitions of food insecurity that are limited to monetary poverty or quantity of food alone may underestimate the true magnitude of this complex issue. Therefore, efforts to decrease food insecurity will require unpacking contextual factors and tailoring appropriate program and policy responses, many of which will go beyond simply increasing income. This context-specific understanding is especially needed for children, because while increases in income may directly translate to changes in food quantity or quality, other factors such as food allocation decisions or feeding practices in the household, could lead to children being protected from or exposed to food insecurity (Haddad et al., 1996; Hadley, Lindstrom, Tessema, & Belachew, 2008).

Our sample of 147 countries and four territories¹ likely underestimates the true global prevalence and burden of food insecurity, especially if food insecurity rates are higher in the countries that are excluded. However, since the GWP covers the highest population countries, in terms of population coverage, the bias in our estimates is likely small. Estimates provided at the regional-levels should be interpreted with caution because intra-regional differences in food insecurity undoubtedly exist. For example, marginalized groups within countries may face higher levels of food insecurity and may fall out of the survey sample or be underrepresented. Our findings reflect the need to further study how food insecurity reports differ between children’s self-reports and that of guardian and caregivers, whether or not children are insulated from household food insecurity, and how these dynamics differ by age and gender, and across contexts. The current analysis using the FIES provides a good starting point to monitor and understand food insecurity among households with children and expedite progress towards achieving the SDG Goal of ending hunger by 2030.

APPENDIX

A1.

Regional groupings of countries/territories included in analysis (where data on households with at least one child under age 15 was available)

East & Southern Africa	Côte d’Ivoire	South America	Southeast Asia	Kosovo	Tajikistan
Angola*	Gabon	Argentina	Cambodia	Latvia	Ukraine
Botswanaǂ	Ghana	Bolivia	Indonesia	Lithuania	Uzbekistan
Burundi*	Guinea	Brazil	Malaysia	Luxembourg
Kenya	Liberia	Chile	Myanmar	Macedonia	Middle East & North Africa
Madagascar	Mali	Colombia	Philippines	Malta	Algeria*
Malawi	Mauritania	Ecuador	Singapore	Montenegro	Bahrainǂ
Mauritius*	Niger	Paraguay	Thailand	Netherlands	Egypt
Mozambiqueǂ	Nigeria	Peru	Vietnam	Northern Cyprus	Iran
Namibia*	Senegal	Uruguay		Norway	Iraq
Rwanda	Sierra Leone	Venezuela	European Union/Non Commonwealth of Independent States	Poland	Israel
South Africa	Togo			Portugal	Jordan
Tanzania		East Asia & the Pacific	Albania	Romania	Kuwaitǂ
Uganda	Central America	Australia	Austria	Serbia	Lebanon
Zambia	Belize*	China	Belgium	Slovakia	Libyaǂ
Zimbabwe	Costa Rica	Hong Kong*	Bosnia and Herzegovina	Slovenia	Moroccoǂ
	Dominican Republic	Japan	Bulgaria	Spain	Palestine
Horn of Africa	El Salvador	Mongolia	Croatia	Sweden	Saudi Arabia
Ethiopia	Guatemala	New Zealand	Cyprus	Switzerland	Syriaǂ
Somalia	Haiti	South Korea	Czech Republic	United Kingdom	Tunisia
South Sudan	Honduras	Taiwan	Denmark		Turkeyǂ
Sudan*	Jamaica*		Estonia	Commonwealth of Independent States	United Arab Emiratesǂ
	Nicaragua	South Asia	Finland	Armenia	Yemen
West & Central Africa	Panama	Afghanistan*	France	Azerbaijan
Benin		Bangladesh	Germany	Belarus
Burkina Faso	North America	Bhutan	Greece	Georgia
Cameroon	Canada	India	Hungary	Kazakhstan
Chad	Mexico	Nepal	Icelandǂ	Kyrgyzstan
Congo Brazzaville	Puerto Rico*	Pakistan	Ireland	Moldova
Congo Kinshasa	United States	Sri Lanka	Italy	Russia

Notes:

^*indicates 2014 only

^ǂindicates 2015 only

Afghanistan HHU15 data from 2014 only as it was missing the indicator on number of adults in household (WP12) for 2015, needed to create child weight

Botswana HHU15 data from 2015 only as it was missing the indicator on number of adults in household (WP12) for 2014, needed to create child weight

Iceland FIES was fielded in 2016

Turkey HHU15 data from 2015 only as it was missing the indicator on number of children in household (WP1230) for 2014, needed to create child weight

Appendix 2:

Food insecurity prevalence by region and country in 2014–15, among all households and households with children under 15 years

Country/Region	SHARE OF FOOD INSECURE HOUSEHOLDS				SHARE OF CHILDREN UNDER 15 YEARS LIVING IN FOOD INSECURE HOUSEHOLDS (Among households with at least one child under age 15)
		FIES-M+ (Moderate or severe)	FIES-S (Severe)	GWP indicator “not enough money to buy food in last 12 months”		FIES-M+ (Moderate or severe)	FIES-S (Severe)	GWP indicator “not enough money to buy food in last 12 months”
	N	Mean % (95% CI)	Mean % (95% CI)	Mean % (95% CI)	N	Mean % (95% CI)	Mean % (95% CI)	Mean % (95% CI)
FULL SAMPLE (147 countries and 4 territories)	289,933	26.69 (26.55 – 26.84)	10.77 (10.68 – 10.86)	32.41 (32.23 – 32.58)	148,848	41.28 (41.05 – 41.5)	19.06 (18.89 – 19.22)	45.04 (44.79 – 45.29)
Central America	15,587	42.56 (41.86 – 43.25)	17.29 (16.81 – 17.76)	49.55 (48.76 – 50.33)	9,529	49.37 (48.48 – 50.27)	20.47 (19.84 – 21.11)	54.84 (53.84 – 55.83)
Belize	482	28.14 (24.5 – 31.79)	9.12 (7.12 – 11.13)	31.89 (27.72 – 36.07)	266	37.91 (32.70 – 43.11)	12.45 (9.3 – 15.6)	35.05 (29.28 – 40.82)
Costa Rica	1,958	20.79 (19.19 – 22.39)	4.70 (4.02 – 5.37)	32.47 (30.40 – 34.55)	877	27.34 (24.71 – 29.97)	6.58 (5.42 – 7.74)	39.96 (36.71 – 43.21)
Dominican Republic	1,988	52.79 (50.81 – 54.77)	20.71 (19.33 – 22.09)	53.62 (51.42 – 55.81)	1,125	58.87 (56.29 – 61.45)	24.10 (22.21 – 26)	59.56 (56.69 – 62.43)
El Salvador	1,970	38.75 (36.84 – 40.67)	10.17 (9.24 – 11.11)	48.91 (46.70 – 51.12)	1,195	44.21 (41.69 – 46.73)	12.57 (11.26 – 13.88)	51.78 (48.94 – 54.62)
Guatemala	1,961	40.55 (38.6 – 42.49)	14.23 (13.03 – 15.42)	47.89 (45.68 – 50.11)	1,461	45.69 (43.41 – 47.97)	16.82 (15.34 – 18.29)	51.10 (48.53 – 53.66)
Haiti	895	80.99 (78.8 – 83.18)	68.42 (65.92 – 70.92)	59.74 (56.52 – 62.96)	590	79.70 (76.99 – 82.40)	65.31 (62.19 – 68.43)	57.75 (53.75 – 61.75)
Honduras	1,959	54.09 (52.16 – 56.01)	20.03 (18.73 – 21.33)	62.05 (59.90 – 64.20)	1,427	57.68 (55.46 – 59.89)	21.39 (19.83 – 22.96)	64.59 (62.11 – 67.08)
Jamaica	485	50.94 (46.88 – 55)	24.57 (21.50 – 27.64)	55.77 (51.33 – 60.20)	266	54.35 (48.79 – 59.91)	28.56 (24.26 – 32.86)	59.58 (53.65 – 65.52)
Nicaragua	1,973	44.11 (42.14 – 46.08)	16.71 (15.46 – 17.96)	57.72 (55.53 – 59.90)	1,405	49.62 (47.26 – 51.97)	19.30 (17.73 – 20.87)	62.84 (60.31 – 65.37)
Panama	1,916	30.32 (28.48 – 32.16)	11.13 (10.01 – 12.25)	42.10 (39.89 – 44.32)	917	38.12 (35.29 – 40.95)	15.48 (13.62 – 17.35)	47.02 (43.78 – 50.25)
Commonwealth of Independent States	22,503	13.44 (13.07 – 13.81)	1.46 (1.35 – 1.58)	26.99 (26.41 – 27.57)	10,829	15.16 (14.6 – 15.73)	2.06 (1.87 – 2.26)	28.47 (27.62 – 29.32)
Armenia	1,946	19.79 (18.31 – 21.27)	0.54 (0.41 – 0.66)	40.98 (38.80 – 43.17)	865	22.78 (20.46 – 25.10)	0.55 (0.37 – 0.74)	43.06 (39.75 – 46.36)
Azerbaijan	1,891	6.15 (5.36 – 6.94)	0.78 (0.47 – 1.09)	16.38 (14.71 – 18.05)	983	6.63 (5.53 – 7.74)	0.67 (0.28 – 1.06)	18.60 (16.16 – 21.04)
Belarus	1,853	8.46 (7.43 – 9.5)	0.51 (0.33 – 0.70)	18.38 (16.61 – 20.14)	509	7.27 (5.54 – 9.00)	0.17 (−.04 – 0.38)	15.34 (12.2 – 18.49)
Georgia	1,947	28.41 (26.64 – 30.18)	1.02 (0.86 – 1.17)	52.67 (50.45 – 54.89)	793	30.05 (27.19 – 32.91)	1.18 (0.92 – 1.44)	54.33 (50.86 – 57.81)
Kazakhstan	1,824	7.51 (6.57 – 8.46)	0.73 (0.45 – 1.02)	21.31 (19.43 – 23.19)	935	7.52 (6.20 – 8.83)	0.59 (0.31 – 0.87)	21.45 (18.82 – 24.09)
Kyrgyzstan	1,880	21.00 (19.41 – 22.59)	4.56 (3.86 – 5.26)	30.61 (28.53 – 32.70)	1,316	23.60 (21.64 – 25.56)	5.41 (4.49 – 6.33)	33.77 (31.21 – 36.33)
Moldova	1,860	10.45 (9.41 – 11.49)	2.15 (1.61 – 2.70)	32.77 (30.63 – 34.91)	650	11.86 (9.97 – 13.75)	2.51 (1.52 – 3.51)	33.56 (29.92 – 37.2)
Russia	3,806	7.68 (7 – 8.37)	0.58 (0.41 – 0.74)	14.19 (13.08 – 15.30)	1,292	8.06 (6.83 – 9.28)	0.76 (0.42 – 1.09)	17.24 (15.18 – 19.31)
Tajikistan	1,682	12.95 (11.64 – 14.26)	2.65 (2.05 – 3.24)	24.83 (22.77 – 26.90)	1,443	13.72 (12.27 – 15.17)	2.79 (2.13 – 3.44)	25.52 (23.26 – 27.77)
Ukraine	1,839	16.70 (15.31 – 18.1)	1.60 (1.21 – 2.00)	32.55 (30.41 – 34.70)	607	15.22 (12.91 – 17.52)	1.62 (0.92 – 2.32)	29.56 (25.92 – 33.2)
Uzbekistan	1,975	13.73 (12.47 – 14.99)	2.05 (1.56 – 2.53)	23.85 (21.97 – 25.73)	1,436	14.28 (12.78 – 15.78)	2.05 (1.49 – 2.62)	24.44 (22.22 – 26.67)
East & Southern Africa	23,628	60.92 (60.37 – 61.47)	31.26 (30.78 – 31.74)	61.30 (60.68 – 61.92)	17,121	67.86 (67.25 – 68.48)	36.86 (36.28 – 37.43)	65.90 (65.19 – 66.61)
Angola	965	63.42 (60.9 – 65.94)	19.56 (17.57 – 21.55)	72.19 (69.36 – 75.03)	800	68.52 (65.89 – 71.15)	23.28 (20.97 – 25.6)	76.15 (73.19 – 79.11)
Botswana	1,959	56.48 (54.55 – 58.41)	32.54 (30.84 – 34.25)	63.31 (61.17 – 65.44)	685	67.08 (64.07 – 70.09)	37.88 (35 – 40.76)	70.36 (66.93 – 73.79)
Burundi	988	79.40 (77.26 – 81.55)	40.14 (37.90 – 42.38)	66.68 (63.74 – 69.62)	739	81.52 (79.09 – 83.95)	44.89 (42.33 – 47.46)	71.60 (68.34 – 74.86)
Kenya	1,979	55.86 (53.89 – 57.83)	29.87 (28.21 – 31.52)	56.21 (54.02 – 58.39)	1,393	61.25 (58.96 – 63.55)	34.64 (32.59 – 36.69)	61.98 (59.43 – 64.53)
Madagascar	2,001	52.90 (51.13 – 54.66)	12.68 (11.54 – 13.82)	62.87 (60.75 – 64.99)	1,590	57.79 (55.83 – 59.76)	15.92 (14.52 – 17.32)	67.60 (65.3 – 69.91)
Malawi	1,981	85.29 (83.91 – 86.68)	53.15 (51.53 – 54.77)	70.04 (68.02 – 72.06)	1,655	87.37 (85.95 – 88.78)	55.67 (53.92 – 57.41)	72.47 (70.31 – 74.62)
Mauritius	993	9.40 (7.7 – 11.09)	3.56 (2.61 – 4.51)	15.51 (13.26 – 17.77)	359	16.40 (12.72 – 20.07)	8.34 (5.94 – 10.74)	23.40 (19 – 27.8)
Mozambique	968	63.10 (60.45 – 65.76)	37.54 (35.09 – 39.99)	61.04 (57.97 – 64.12)	801	65.03 (62.10 – 67.96)	41.05 (38.31 – 43.79)	60.88 (57.49 – 64.27)
Namibia	983	61.51 (58.74 – 64.28)	42.06 (39.45 – 44.66)	62.89 (59.86 – 65.91)	733	68.55 (65.49 – 71.61)	48.82 (45.82 – 51.82)	68.23 (64.85 – 71.6)
Rwanda	1,988	46.96 (44.99 – 48.93)	18.76 (17.38 – 20.15)	49.35 (47.15 – 51.55)	1,478	55.59 (53.31 – 57.88)	23.68 (21.96 – 25.39)	55.33 (52.79 – 57.87)
South Africa	1,955	41.18 (39.31 – 43.05)	18.58 (17.18 – 19.98)	52.58 (50.36 – 54.80)	1,157	46.52 (44.10 – 48.94)	20.77 (18.87 – 22.66)	55.90 (53.03 – 58.77)
Tanzania	1,955	54.35 (52.38 – 56.33)	27.54 (25.89 – 29.20)	54.56 (52.35 – 56.77)	1,444	60.17 (57.92 – 62.43)	31.69 (29.7 – 33.68)	58.04 (55.5 – 60.59)
Uganda	1,960	67.10 (65.24 – 68.96)	38.48 (36.75 – 40.21)	62.77 (60.62 – 64.91)	1,492	71.67 (69.63 – 73.71)	42.90 (40.9 – 44.9)	65.97 (63.56 – 68.38)
Zambia	1,970	76.22 (74.59 – 77.86)	43.18 (41.47 – 44.89)	66.35 (64.26 – 68.44)	1,567	79.60 (77.87 – 81.33)	46.13 (44.23 – 48.02)	67.80 (65.49 – 70.12)
Zimbabwe	1,976	60.95 (59.03 – 62.88)	30.84 (29.24 – 32.44)	66.15 (64.06 – 68.24)	1,587	65.74 (63.65 – 67.83)	35.76 (33.91 – 37.6)	71.71 (69.49 – 73.92)
East Asia & the Pacific	21,444	6.25 (5.98 – 6.53)	0.98 (0.88 – 1.08)	11.08 (10.66 – 11.5)	7,212	9.28 (8.71 – 9.85)	1.61 (1.39 – 1.84)	14.95 (14.12 – 15.77)
Australia	1,989	9.73 (8.55 – 10.91)	2.33 (1.81 – 2.84)	9.31 (8.03 – 10.59)	421	15.95 (12.78 – 19.12)	4.89 (3.2 – 6.57)	13.02 (9.79 – 16.24)
China	8,645	4.11 (3.78 – 4.44)	0.48 (0.37 – 0.59)	6.63 (6.10 – 7.15)	3,444	5.27 (4.65 – 5.88)	0.88 (0.64 – 1.11)	7.85 (6.95 – 8.75)
Hong Kong	997	8.75 (7.26 – 10.24)	1.10 (0.65 – 1.55)	7.06 (5.47 – 8.65)	271	10.93 (7.70 – 14.16)	1.49 (0.48 – 2.49)	7.37 (4.24 – 10.5)
Japan	1,999	2.01 (1.49 – 2.53)	0.50 (0.26 – 0.74)	5.62 (4.61 – 6.63)	403	1.37 (0.48 – 2.25)	0.26 (−.14 – 0.65)	6.84 (4.37 – 9.32)
Mongolia	1,898	17.75 (16.35 – 19.16)	1.27 (0.96 – 1.57)	36.29 (34.12 – 38.45)	1,158	19.91 (18.02 – 21.79)	1.56 (1.13 – 1.99)	38.08 (35.28 – 40.88)
New Zealand	1,987	6.69 (5.73 – 7.65)	2.28 (1.73 – 2.83)	9.49 (8.20 – 10.78)	484	10.93 (8.58 – 13.29)	3.17 (1.84 – 4.49)	15.06 (11.87 – 18.26)
South Korea	1,936	6.40 (5.51 – 7.29)	0.88 (0.59 – 1.17)	18.47 (16.74 – 20.20)	421	4.81 (3.15 – 6.46)	0.43 (0.07 – 0.78)	15.91 (12.4 – 19.41)
Taiwan	1,993	3.59 (2.94 – 4.25)	0.70 (0.40 – 0.99)	10.14 (8.81 – 11.46)	610	5.00 (3.58 – 6.42)	0.83 (0.31 – 1.35)	12.73 (10.08 – 15.38)
European Union/Non Commonwealth of Independent States	72,743	10.34 (10.14 – 10.53)	2.37 (2.28 – 2.45)	15.63 (15.37 – 15.9)	19,622	14.00 (13.57 – 14.43)	3.57 (3.36 – 3.77)	19.86 (19.3 – 20.42)
Albania	1,958	37.21 (35.33 – 39.09)	9.25 (8.28 – 10.22)	53.01 (50.80 – 55.23)	766	45.58 (42.47 – 48.68)	12.85 (11.07 – 14.64)	62.42 (58.98 – 65.85)
Austria	1,995	5.35 (4.59 – 6.12)	1.73 (1.32 – 2.15)	7.68 (6.51 – 8.85)	454	7.66 (5.75 – 9.58)	2.73 (1.61 – 3.86)	10.16 (7.37 – 12.95)
Belgium	2,034	7.85 (6.79 – 8.92)	2.71 (2.15 – 3.26)	11.01 (9.65 – 12.37)	667	9.72 (7.69 – 11.74)	3.64 (2.51 – 4.77)	16.21 (13.41 – 19.01)
Bosnia Herzegovina	1,955	10.97 (9.71 – 12.23)	1.60 (1.24 – 1.95)	20.65 (18.85 – 22.44)	551	9.96 (7.75 – 12.18)	1.58 (0.89 – 2.28)	23.69 (20.13 – 27.25)
Bulgaria	1,919	14.35 (13.01 – 15.69)	1.39 (1.05 – 1.73)	29.25 (27.21 – 31.29)	416	25.80 (22.01 – 29.59)	3.97 (2.82 – 5.12)	39.89 (35.16 – 44.61)
Croatia	1,952	6.59 (5.66 – 7.52)	0.85 (0.55 – 1.15)	10.97 (9.58 – 12.36)	433	4.87 (3.20 – 6.55)	0.15 (0 – 0.29)	7.33 (4.87 – 9.8)
Cyprus	2,010	15.04 (13.65 – 16.44)	4.59 (3.85 – 5.34)	18.73 (17.02 – 20.43)	664	17.57 (15.01 – 20.13)	5.71 (4.27 – 7.15)	23.92 (20.67 – 27.18)
Czech Republic	1,948	6.75 (5.81 – 7.69)	1.21 (0.83 – 1.59)	14.20 (12.64 – 15.75)	445	8.94 (6.61 – 11.26)	2.14 (1.09 – 3.19)	15.87 (12.46 – 19.27)
Denmark	1,996	4.99 (4.16 – 5.81)	0.68 (0.42 – 0.93)	5.80 (4.78 – 6.83)	551	8.51 (6.48 – 10.54)	1.19 (0.62 – 1.76)	7.64 (5.42 – 9.87)
Estonia	1,952	8.13 (7.12 – 9.14)	0.24 (0.13 – 0.35)	17.16 (15.49 – 18.84)	473	11.15 (8.73 – 13.56)	0.21 (0.03 – 0.4)	21.15 (17.46 – 24.84)
Finland	1,986	9.19 (8.11 – 10.26)	2.53 (1.99 – 3.08)	9.69 (8.38 – 10.99)	274	7.73 (5.41 – 10.05)	0.88 (0.15 – 1.61)	8.61 (5.26 – 11.95)
France	1,979	6.51 (5.58 – 7.44)	1.67 (1.21 – 2.12)	11.13 (9.74 – 12.51)	486	6.02 (4.29 – 7.74)	1.34 (0.48 – 2.2)	12.23 (9.31 – 15.15)
Germany	1,992	3.61 (2.91 – 4.31)	1.02 (0.66 – 1.38)	5.24 (4.26 – 6.22)	418	4.91 (3.16 – 6.66)	1.19 (0.38 – 2)	7.27 (4.77 – 9.76)
Greece	1,999	12.93 (11.64 – 14.22)	1.52 (1.16 – 1.88)	20.13 (18.37 – 21.89)	446	15.74 (12.70 – 18.78)	2.13 (1.28 – 2.98)	24.61 (20.59 – 28.62)
Hungary	1,925	9.22 (8.11 – 10.33)	1.15 (0.82 – 1.48)	16.19 (14.54 – 17.83)	465	11.84 (9.34 – 14.33)	1.20 (0.52 – 1.87)	16.59 (13.2 – 19.98)
Iceland	594	8.12 (6.12 – 10.13)	2.25 (1.32 – 3.18)	11.26 (8.71 – 13.81)	225	8.36 (4.94 – 11.79)	3.46 (1.5 – 5.42)	11.64 (7.41 – 15.86)
Ireland	1,991	10.10 (8.93 – 11.27)	4.18 (3.44 – 4.91)	11.18 (9.79 – 12.56)	601	17.94 (15.21 – 20.67)	7.29 (5.59 – 8.99)	20.84 (17.58 – 24.1)
Italy	1,966	7.51 (6.52 – 8.5)	0.93 (0.63 – 1.22)	13.93 (12.40 – 15.47)	450	8.79 (6.50 – 11.07)	1.13 (0.42 – 1.85)	18.09 (14.52 – 21.66)
Kosovo	1,809	13.45 (12.11 – 14.78)	3.31 (2.64 – 3.98)	18.95 (17.14 – 20.76)	967	16.03 (14.07 – 17.99)	4.26 (3.25 – 5.27)	22.73 (20.09 – 25.38)
Latvia	1,909	9.82 (8.69 – 10.95)	0.74 (0.51 – 0.98)	18.82 (17.07 – 20.58)	437	12.89 (10.12 – 15.66)	1.15 (0.52 – 1.78)	21.45 (17.59 – 25.31)
Lithuania	1,869	18.50 (16.95 – 20.06)	3.12 (2.50 – 3.75)	12.20 (10.71 – 13.68)	456	21.53 (18.12 – 24.93)	4.55 (2.98 – 6.11)	16.82 (13.37 – 20.26)
Luxembourg	1,983	5.27 (4.44 – 6.1)	2.34 (1.81 – 2.88)	5.79 (4.76 – 6.82)	513	7.25 (5.19 – 9.30)	4.51 (2.94 – 6.08)	5.99 (3.93 – 8.06)
Macedonia	1,966	13.23 (11.88 – 14.59)	3.78 (3.11 – 4.45)	20.73 (18.93 – 22.52)	573	20.97 (17.91 – 24.02)	6.80 (5.16 – 8.45)	29.37 (25.63 – 33.11)
Malta	2,005	5.35 (4.5 – 6.21)	1.60 (1.16 – 2.04)	7.56 (6.40 – 8.72)	540	10.50 (8.18 – 12.82)	3.47 (2.29 – 4.66)	13.74 (10.83 – 16.65)
Montenegro	1,945	12.06 (10.78 – 13.34)	1.73 (1.33 – 2.13)	16.86 (15.20 – 18.53)	578	17.93 (15.08 – 20.77)	2.52 (1.68 – 3.36)	19.61 (16.37 – 22.86)
Netherlands	2,001	5.03 (4.2 – 5.86)	1.43 (1.01 – 1.86)	9.94 (8.63 – 11.26)	553	8.78 (6.70 – 10.85)	2.66 (1.53 – 3.78)	13.52 (10.66 – 16.38)
Northern Cyprus	1,961	24.35 (22.71 – 26)	7.74 (6.78 – 8.70)	36.88 (34.75 – 39.02)	650	29.55 (26.57 – 32.52)	9.42 (7.59 – 11.26)	40.25 (36.47 – 44.03)
Norway	1,995	4.61 (3.8 – 5.42)	1.23 (0.85 – 1.61)	6.09 (5.04 – 7.14)	480	5.15 (3.44 – 6.86)	1.66 (0.76 – 2.57)	5.65 (3.57 – 7.72)
Poland	1,909	9.61 (8.45 – 10.78)	2.12 (1.62 – 2.62)	17.26 (15.56 – 18.96)	512	14.17 (11.45 – 16.89)	2.91 (1.78 – 4.03)	24.02 (20.31 – 27.73)
Portugal	2,016	14.79 (13.41 – 16.17)	4.16 (3.48 – 4.84)	19.04 (17.32 – 20.75)	793	18.15 (15.73 – 20.57)	4.87 (3.75 – 6)	24.96 (21.95 – 27.98)
Romania	1,915	18.95 (17.37 – 20.54)	5.62 (4.78 – 6.45)	37.58 (35.41 – 39.75)	470	26.32 (22.71 – 29.94)	9.09 (6.95 – 11.24)	48.81 (44.27 – 53.35)
Serbia	1,923	9.67 (8.52 – 10.82)	1.46 (1.09 – 1.82)	20.63 (18.82 – 22.44)	466	11.35 (8.89 – 13.82)	1.95 (1.03 – 2.86)	25.34 (21.38 – 29.31)
Slovakia	1,923	5.80 (4.92 – 6.68)	0.85 (0.54 – 1.16)	9.52 (8.21 – 10.84)	454	10.82 (8.38 – 13.27)	1.74 (0.78 – 2.71)	16.10 (12.71 – 19.49)
Slovenia	1,996	12.17 (10.93 – 13.41)	1.34 (1.02 – 1.66)	13.73 (12.22 – 15.24)	607	12.73 (10.42 – 15.03)	1.49 (0.87 – 2.11)	14.98 (12.13 – 17.83)
Spain	1,990	6.09 (5.21 – 6.98)	1.39 (1.01 – 1.76)	11.71 (10.29 – 13.12)	585	12.01 (9.65 – 14.37)	3.36 (2.32 – 4.41)	13.17 (10.42 – 15.92)
Sweden	1,985	4.52 (3.73 – 5.31)	0.77 (0.49 – 1.04)	4.97 (4.01 – 5.93)	514	4.05 (2.63 – 5.46)	0.56 (0.08 – 1.04)	3.42 (1.85 – 5)
Switzerland	1,500	3.92 (3.06 – 4.78)	0.56 (0.29 – 0.82)	4.58 (3.52 – 5.64)	335	5.00 (3.17 – 6.82)	0.19 (−.12 – 0.49)	7.94 (5.03 – 10.85)
United Kingdom	1,992	9.53 (8.39 – 10.68)	4.72 (3.89 – 5.55)	10.89 (9.52 – 12.26)	354	19.46 (15.79 – 23.13)	10.40 (7.53 – 13.28)	20.02 (15.83 – 24.21)
Horn of Africa	6,573	60.62 (59.55 – 61.69)	36.19 (35.2 – 37.17)	53.00 (51.79 – 54.2)	5,260	65.49 (64.33 – 66.66)	41.05 (39.94 – 42.17)	57.44 (56.11 – 58.78)
Ethiopia	1,971	49.13 (47.23 – 51.04)	11.35 (10.29 – 12.40)	45.40 (43.20 – 47.60)	1,482	53.50 (51.29 – 55.72)	13.76 (12.45 – 15.08)	49.36 (46.81 – 51.91)
Somalia	1,786	48.52 (46.39 – 50.64)	28.16 (26.41 – 29.91)	54.37 (52.06 – 56.68)	1,512	53.43 (51.11 – 55.76)	33.01 (31.02 – 34.99)	57.97 (55.48 – 60.46)
South Sudan	1,897	91.58 (90.58 – 92.57)	76.86 (75.55 – 78.18)	62.73 (60.56 – 64.91)	1,661	92.02 (90.99 – 93.06)	77.29 (75.92 – 78.66)	64.23 (61.92 – 66.53)
Sudan	919	44.54 (41.57 – 47.51)	20.71 (18.60 – 22.82)	46.48 (43.25 – 49.71)	605	53.47 (49.81 – 57.13)	25.38 (22.59 – 28.16)	55.65 (51.68 – 59.62)
Middle East & North Africa	32,838	24.09 (23.68 – 24.5)	7.96 (7.73 – 8.2)	29.00 (28.51 – 29.49)	19,474	29.43 (28.86 – 30)	10.39 (10.05 – 10.72)	33.70 (33.04 – 34.37)
Algeria	999	6.42 (5.32 – 7.53)	1.29 (0.76 – 1.82)	6.49 (4.96 – 8.02)	579	6.77 (5.27 – 8.27)	1.43 (0.65 – 2.2)	5.98 (4.04 – 7.91)
Bahrain	1,975	16.80 (15.35 – 18.25)	5.52 (4.73 – 6.31)	21.31 (19.50 – 23.12)	1,284	18.20 (16.31 – 20.08)	6.37 (5.34 – 7.4)	22.14 (19.87 – 24.42)
Egypt	1,987	23.63 (21.95 – 25.32)	9.66 (8.59 – 10.72)	15.09 (13.52 – 16.67)	1,165	27.31 (25.01 – 29.61)	10.61 (9.18 – 12.04)	15.68 (13.59 – 17.77)
Iran	1,983	48.71 (46.82 – 50.61)	8.54 (7.70 – 9.38)	53.65 (51.45 – 55.85)	1,006	56.05 (53.38 – 58.72)	11.57 (10.22 – 12.92)	57.43 (54.37 – 60.49)
Iraq	1,971	42.77 (40.89 – 44.65)	18.70 (17.39 – 20.00)	47.19 (44.98 – 49.39)	1,597	46.77 (44.67 – 48.88)	21.26 (19.75 – 22.77)	51.56 (49.11 – 54.01)
Israel	1,925	5.47 (4.65 – 6.3)	0.95 (0.62 – 1.28)	8.62 (7.36 – 9.87)	863	7.39 (5.91 – 8.87)	1.89 (1.15 – 2.63)	11.94 (9.77 – 14.11)
Jordan	1,969	28.11 (26.24 – 29.97)	13.00 (11.77 – 14.23)	35.24 (33.13 – 37.35)	1,277	34.81 (32.37 – 37.25)	15.87 (14.24 – 17.51)	41.92 (39.21 – 44.63)
Kuwait	1,982	12.25 (11.02 – 13.48)	4.37 (3.64 – 5.09)	19.13 (17.40 – 20.87)	1,025	11.03 (9.42 – 12.64)	3.39 (2.54 – 4.24)	20.17 (17.71 – 22.63)
Lebanon	1,959	6.83 (5.89 – 7.76)	1.68 (1.22 – 2.13)	9.86 (8.54 – 11.18)	756	9.63 (7.82 – 11.45)	2.80 (1.88 – 3.73)	13.70 (11.25 – 16.16)
Libya	998	29.89 (27.41 – 32.37)	10.59 (9.11 – 12.07)	31.17 (28.29 – 34.05)	695	34.71 (31.63 – 37.78)	11.92 (10.05 – 13.8)	33.85 (30.33 – 37.38)
Morocco	2,026	25.98 (24.32 – 27.63)	5.08 (4.37 – 5.79)	31.91 (29.88 – 33.94)	1,279	28.18 (26.01 – 30.34)	5.41 (4.51 – 6.32)	34.28 (31.68 – 36.88)
Palestine	1,987	28.52 (26.72 – 30.33)	9.13 (8.13 – 10.14)	37.77 (35.63 – 39.90)	1,357	33.48 (31.17 – 35.78)	12.15 (10.76 – 13.53)	41.28 (38.66 – 43.91)
Saudi Arabia	1,993	23.22 (21.6 – 24.85)	7.93 (6.99 – 8.86)	28.85 (26.85 – 30.84)	1,313	26.69 (24.62 – 28.75)	8.68 (7.49 – 9.88)	31.36 (28.84 – 33.87)
Syria	303	46.67 (42.12 – 51.21)	22.57 (18.94 – 26.21)	37.22 (31.75 – 42.70)	245	49.29 (44.22 – 54.36)	24.74 (20.55 – 28.92)	36.88 (30.8 – 42.97)
Tunisia	1,980	18.86 (17.27 – 20.44)	12.28 (10.97 – 13.59)	23.16 (21.30 – 25.02)	1,089	18.95 (16.78 – 21.11)	12.95 (11.13 – 14.77)	24.63 (22.06 – 27.19)
Turkey	1,969	31.81 (30.06 – 33.57)	5.73 (5.02 – 6.45)	47.64 (45.43 – 49.85)	408	33.39 (29.50 – 37.29)	6.82 (5.15 – 8.49)	45.81 (40.95 – 50.66)
United Arab Emirates	2,865	12.18 (11.15 – 13.21)	7.02 (6.25 – 7.79)	19.70 (18.24 – 21.16)	1,817	11.05 (9.79 – 12.31)	6.77 (5.79 – 7.74)	20.75 (18.88 – 22.62)
Yemen	1,967	43.59 (41.69 – 45.49)	9.98 (9.00 – 10.96)	48.57 (46.36 – 50.78)	1,719	48.19 (46.15 – 50.22)	11.03 (9.93 – 12.13)	52.74 (50.38 – 55.11)
North America	6,335	17.39 (16.57 – 18.22)	5.34 (4.9 – 5.78)	21.48 (20.46 – 22.49)	2,062	23.24 (21.62 – 24.86)	8.07 (7.12 – 9.01)	27.48 (25.55 – 29.4)
Canada	2,001	8.31 (7.23 – 9.39)	2.00 (1.53 – 2.47)	9.88 (8.57 – 11.19)	513	11.92 (9.34 – 14.51)	3.27 (2.12 – 4.42)	12.98 (10.07 – 15.9)
Mexico	1,865	28.18 (26.45 – 29.91)	8.96 (7.92 – 9.99)	36.63 (34.45 – 38.82)	949	34.93 (32.36 – 37.49)	12.79 (11.06 – 14.51)	42.46 (39.31 – 45.61)
Puerto Rico	484	18.20 (15.02 – 21.37)	7.49 (5.60 – 9.39)	32.69 (28.50 – 36.89)	92	19.41 (12.13 – 26.70)	7.22 (2.77 – 11.68)	36.37 (26.36 – 46.39)
United States	1,985	16.11 (14.64 – 17.58)	4.74 (4.01 – 5.47)	16.05 (14.44 – 17.67)	508	19.62 (16.42 – 22.81)	6.68 (4.96 – 8.41)	20.04 (16.55 – 23.53)
South America	19,660	22.23 (21.72 – 22.75)	6.90 (6.61 – 7.18)	29.37 (28.74 – 30.01)	9,690	28.95 (28.15 – 29.76)	9.64 (9.18 – 10.11)	37.07 (36.11 – 38.03)
Argentina	1,994	13.97 (12.59 – 15.35)	4.12 (3.43 – 4.82)	24.28 (22.39 – 26.16)	887	22.17 (19.67 – 24.68)	6.68 (5.38 – 7.97)	34.51 (31.38 – 37.65)
Bolivia	1,980	27.78 (26.02 – 29.55)	12.12 (10.90 – 13.34)	34.36 (32.26 – 36.45)	1,283	33.48 (31.18 – 35.78)	14.36 (12.75 – 15.98)	38.92 (36.25 – 41.6)
Brazil	1,977	16.11 (14.71 – 17.52)	2.80 (2.27 – 3.33)	19.23 (17.49 – 20.97)	798	25.04 (22.42 – 27.66)	5.18 (4.06 – 6.3)	28.06 (24.94 – 31.18)
Chile	1,994	13.44 (12.12 – 14.76)	3.48 (2.86 – 4.10)	20.19 (18.42 – 21.95)	878	17.00 (14.79 – 19.22)	5.20 (4.04 – 6.35)	23.49 (20.68 – 26.3)
Colombia	1,975	25.52 (23.76 – 27.27)	9.05 (8.05 – 10.06)	31.82 (29.77 – 33.88)	1,000	33.18 (30.50 – 35.87)	12.45 (10.87 – 14.03)	36.96 (33.96 – 39.96)
Ecuador	1,973	25.40 (23.73 – 27.07)	9.73 (8.65 – 10.81)	40.46 (38.29 – 42.63)	1,154	30.31 (28.03 – 32.60)	11.42 (9.92 – 12.93)	48.27 (45.38 – 51.16)
Paraguay	1,952	19.37 (17.93 – 20.81)	1.14 (0.89 – 1.40)	17.29 (15.61 – 18.97)	926	21.90 (19.70 – 24.10)	1.69 (1.25 – 2.14)	24.43 (21.66 – 27.21)
Peru	1,930	33.47 (31.58 – 35.36)	11.03 (9.95 – 12.11)	41.50 (39.29 – 43.70)	1,062	38.81 (36.18 – 41.44)	12.84 (11.34 – 14.34)	46.93 (43.92 – 49.93)
Uruguay	1,978	15.94 (14.48 – 17.4)	5.09 (4.30 – 5.89)	21.70 (19.89 – 23.52)	722	27.64 (24.69 – 30.59)	9.26 (7.53 – 10.98)	34.91 (31.42 – 38.39)
Venezuela	1,907	32.04 (30.15 – 33.93)	10.64 (9.57 – 11.72)	43.73 (41.50 – 45.96)	980	35.34 (32.60 – 38.08)	13.78 (12.09 – 15.48)	47.01 (43.88 – 50.14)
South Asia	17,244	25.82 (25.24 – 26.4)	10.64 (10.25 – 11.02)	34.63 (33.92 – 35.34)	11,190	30.13 (29.38 – 30.88)	12.86 (12.34 – 13.37)	38.30 (37.4 – 39.2)
Afghanistan	1,759	45.79 (43.86 – 47.73)	19.07 (17.56 – 20.59)	44.73 (42.40 – 47.06)	871	47.78 (45.03 – 50.53)	20.60 (18.39 – 22.82)	40.80 (37.53 – 44.07)
Bangladesh	1,903	29.44 (27.57 – 31.3)	12.17 (10.97 – 13.38)	32.95 (30.84 – 35.07)	1,445	34.51 (32.29 – 36.74)	15.08 (13.57 – 16.6)	37.67 (35.17 – 40.17)
Bhutan	1,987	2.58 (2.05 – 3.12)	0.28 (0.12 – 0.43)	7.68 (6.51 – 8.86)	1,288	3.32 (2.53 – 4.11)	0.69 (0.37 – 1.01)	8.56 (7.03 – 10.09)
India	5,532	22.46 (21.47 – 23.45)	12.40 (11.65 – 13.14)	27.45 (26.27 – 28.62)	3,365	27.46 (26.09 – 28.84)	16.10 (15.03 – 17.17)	30.98 (29.41 – 32.54)
Nepal	2,027	25.39 (23.76 – 27.02)	7.73 (6.81 – 8.66)	49.19 (47.01 – 51.37)	1,456	32.27 (30.22 – 34.32)	10.10 (8.87 – 11.33)	58.62 (56.09 – 61.15)
Pakistan	1,978	42.17 (40.4 – 43.95)	15.75 (14.40 – 17.09)	43.49 (41.30 – 45.68)	1,640	46.86 (44.87 – 48.85)	18.74 (17.16 – 20.32)	46.69 (44.28 – 49.11)
Sri Lanka	2,058	21.41 (19.82 – 22.99)	5.12 (4.40 – 5.84)	50.04 (47.88 – 52.20)	1,125	24.35 (22.13 – 26.57)	5.93 (4.88 – 6.98)	50.91 (47.99 – 53.84)
Southeast Asia	15,711	20.77 (20.21 – 21.34)	6.10 (5.8 – 6.4)	34.68 (33.94 – 35.42)	8,965	29.17 (28.34 – 30)	9.21 (8.73 – 9.69)	45.09 (44.05 – 46.12)
Cambodia	1,979	48.69 (46.81 – 50.57)	18.74 (17.35 – 20.12)	67.59 (65.53 – 69.66)	1,484	51.19 (49.01 – 53.37)	20.23 (18.59 – 21.88)	71.70 (69.4 – 73.99)
Indonesia	1,959	17.02 (15.56 – 18.48)	5.19 (4.39 – 6.00)	26.94 (24.97 – 28.91)	1,352	19.52 (17.67 – 21.36)	6.00 (4.95 – 7.05)	30.63 (28.17 – 33.09)
Malaysia	1,948	16.49 (15.01 – 17.97)	8.05 (7.04 – 9.06)	18.85 (17.11 – 20.59)	936	20.75 (18.35 – 23.14)	12.20 (10.39 – 14)	22.07 (19.41 – 24.73)
Myanmar	2,035	13.14 (11.93 – 14.35)	1.55 (1.16 – 1.94)	46.02 (43.85 – 48.19)	1,278	16.96 (15.27 – 18.65)	2.15 (1.57 – 2.73)	53.26 (50.52 – 56)
Philippines	1,993	43.97 (42.09 – 45.85)	10.84 (9.85 – 11.82)	64.16 (62.05 – 66.27)	1,376	51.52 (49.26 – 53.78)	13.98 (12.66 – 15.31)	71.06 (68.66 – 73.46)
Singapore	1,940	3.13 (2.41 – 3.85)	1.09 (0.69 – 1.49)	5.76 (4.72 – 6.80)	635	3.84 (2.45 – 5.24)	1.51 (0.7 – 2.32)	5.08 (3.37 – 6.79)
Thailand	1,903	6.81 (5.78 – 7.84)	2.26 (1.73 – 2.79)	18.53 (16.79 – 20.28)	677	9.91 (7.88 – 11.95)	3.23 (2.15 – 4.32)	17.74 (14.85 – 20.62)
Vietnam	1,954	15.86 (14.47 – 17.24)	0.85 (0.58 – 1.12)	27.30 (25.32 – 29.28)	1,227	17.63 (15.82 – 19.44)	1.01 (0.64 – 1.38)	30.23 (27.66 – 32.8)
West & Central Africa	35,667	54.42 (53.97 – 54.87)	26.39 (26.02 – 26.76)	57.13 (56.62 – 57.64)	27,894	59.10 (58.6 – 59.6)	29.05 (28.62 – 29.48)	60.07 (59.5 – 60.65)
Benin	1,949	55.61 (53.7 – 57.53)	23.63 (22.18 – 25.08)	63.93 (61.79 – 66.06)	1,457	55.49 (53.29 – 57.69)	23.50 (21.83 – 25.18)	65.66 (63.21 – 68.1)
Burkina Faso	1,936	42.01 (40.14 – 43.89)	16.73 (15.39 – 18.06)	46.17 (43.95 – 48.39)	1,571	44.55 (42.45 – 46.65)	18.56 (17 – 20.11)	47.19 (44.72 – 49.66)
Cameroon	1,946	53.92 (51.99 – 55.85)	27.69 (26.12 – 29.26)	63.52 (61.38 – 65.66)	1,464	56.00 (53.80 – 58.21)	28.89 (27.07 – 30.7)	66.09 (63.66 – 68.52)
Chad	1,968	64.33 (62.53 – 66.12)	25.67 (24.16 – 27.17)	66.33 (64.24 – 68.42)	1,702	67.55 (65.67 – 69.43)	27.83 (26.18 – 29.48)	67.21 (64.98 – 69.44)
Congo Brazzaville	1,933	61.86 (60.03 – 63.69)	37.30 (35.62 – 38.98)	63.57 (61.43 – 65.72)	1,289	64.24 (62.03 – 66.45)	39.35 (37.29 – 41.41)	64.72 (62.1 – 67.33)
Congo Kinshasa	1,854	72.26 (70.57 – 73.96)	36.67 (34.92 – 38.42)	71.25 (69.19 – 73.31)	1,585	75.31 (73.60 – 77.03)	37.37 (35.48 – 39.25)	74.68 (72.54 – 76.82)
Cote d’Ivoire	1,941	52.06 (50.16 – 53.96)	19.40 (18.04 – 20.75)	60.42 (58.24 – 62.59)	1,432	56.00 (53.84 – 58.17)	21.43 (19.8 – 23.05)	64.04 (61.55 – 66.53)
Gabon	1,951	58.55 (56.63 – 60.48)	36.75 (34.98 – 38.53)	60.24 (58.07 – 62.41)	1,286	64.20 (61.91 – 66.50)	41.18 (38.97 – 43.39)	62.09 (59.43 – 64.74)
Ghana	1,949	51.20 (49.26 – 53.14)	20.19 (18.82 – 21.56)	60.27 (58.10 – 62.45)	1,293	54.29 (51.92 – 56.66)	20.78 (19.13 – 22.43)	62.92 (60.29 – 65.56)
Guinea	1,945	69.71 (68 – 71.43)	32.08 (30.47 – 33.70)	67.59 (65.51 – 69.68)	1,787	72.79 (71.08 – 74.50)	34.37 (32.67 – 36.07)	69.65 (67.51 – 71.78)
Liberia	1,905	86.36 (84.95 – 87.77)	55.21 (53.60 – 56.83)	66.88 (64.77 – 69.00)	1,595	88.54 (87.12 – 89.95)	56.39 (54.66 – 58.11)	67.25 (64.95 – 69.56)
Mali	1,935	24.86 (23.25 – 26.47)	4.30 (3.61 – 4.98)	29.07 (27.05 – 31.10)	1,787	26.58 (24.87 – 28.28)	4.25 (3.55 – 4.95)	29.28 (27.17 – 31.39)
Mauritania	1,883	26.34 (24.67 – 28.01)	8.82 (7.79 – 9.85)	36.97 (34.79 – 39.15)	1,604	28.95 (27.08 – 30.82)	10.27 (9.07 – 11.47)	38.58 (36.19 – 40.96)
Niger	1,867	58.83 (57.02 – 60.64)	21.32 (19.85 – 22.78)	65.78 (63.62 – 67.93)	1,725	61.93 (60.10 – 63.76)	22.59 (21.04 – 24.14)	69.08 (66.9 – 71.27)
Nigeria	1,901	53.18 (51.25 – 55.11)	24.82 (23.28 – 26.35)	64.61 (62.46 – 66.76)	1,137	56.27 (53.86 – 58.68)	25.36 (23.36 – 27.36)	66.11 (63.35 – 68.86)
Senegal	1,961	30.38 (28.72 – 32.04)	8.34 (7.42 – 9.27)	40.83 (38.66 – 43.01)	1,798	34.68 (32.88 – 36.48)	10.14 (9.09 – 11.2)	43.47 (41.18 – 45.76)
Sierra Leone	1,917	77.98 (76.42 – 79.54)	57.21 (55.44 – 58.98)	56.63 (54.41 – 58.85)	1,694	80.84 (79.26 – 82.43)	61.72 (59.88 – 63.55)	54.41 (52.03 – 56.78)
Togo	1,933	64.43 (62.57 – 66.29)	31.45 (29.84 – 33.06)	66.40 (64.30 – 68.51)	1,329	68.02 (65.83 – 70.21)	34.29 (32.3 – 36.28)	69.36 (66.88 – 71.84)

Notes:

Data comes from Gallup World Poll (GWP) Surveys and includes 289,933 respondents [148,848 respondents with at least one child under 15 years in the household (HHU15)] from 147 countries and 4 territories in 2014 and 2015.

The Food Insecurity Experience Scale (FIES) was used to measure moderate or severe (FIES-M+), and severe (FIES-S) food insecurity. The GWP food insecurity indicator is: “Was there ever a time in the last 12 months when you did not have enough money to buy food?”

The GWP selects and surveys one respondent aged 15 years or over per household. Food insecurity estimates are, therefore, based on individual responses.

Estimates are weighted means and the corresponding 95% Confidence Intervals. We used household weights for the estimates for all households, and child weights for the estimates for households with at least one child under the age of 15 years. The child food security indicator is, therefore, the share of children under 15 years living in food insecure households, among households with at least one child under age 15 years

Afghanistan data for households with at least one child under age 15 years are for 2014 only as it was missing the indicator for number of adults in household (WP12) for 2015, needed to create the child weight

Botswana data for households with at least one child under age 15 years are for 2015 only as it was missing the indicator for number of adults in household (WP12) for 2014, needed to create the child weight

Turkey data for households with at least one child under age 15 years are for 2015 only as it was missing the indicator for number of children in household (WP1230) for 2014, needed to create the child weight

Appendix 3:

Weighted means for key variables for all countries by year

Variable	ALL HOUSEHOLDS
	2006	2007	2008	2009	2010	2011	2012	2013	2014	2015
	Mean (SE)	Mean (SE)	Mean (SE)	Mean (SE)	Mean (SE)	Mean (SE)	Mean (SE)	Mean (SE)	Mean (SE)	Mean (SE)
Age of respondent (years)	42.03 (0.18)	40.28 (0.08)	40.07 (0.06)	39.30 (0.05)	40.37 (0.05)	40.08 (0.04)	41.55 (0.04)	40.71 (0.05)	41.71 (0.05)	41.54 (0.05)
Male	0.44 (0.01)	0.45 (0.00)	0.46 (0.00)	0.47 (0.00)	0.46 (0.00)	0.46 (0.00)	0.46 (0.00)	0.47 (0.00)	0.46 (0.00)	0.47 (0.00)
Education	0.76 (0.01)	0.88 (0.00)	0.84 (0.00)	0.78 (0.00)	0.80 (0.00)	0.82 (0.00)	0.84 (0.00)	0.82 (0.00)	0.82 (0.00)	0.85 (0.00)
Log of per capita income	7.44 (0.01)	7.81 (0.01)	7.74 (0.01)	7.52 (0.00)	7.65 (0.00)	7.67 (0.00)	7.85 (0.00)	7.57 (0.00)	7.61 (0.01)	7.76 (0.01)
Log of household size	0.89 (0.01)	1.12 (0.00)	1.23 (0.00)	1.37 (0.00)	1.34 (0.00)	1.33 (0.00)	1.23 (0.00)	1.29 (0.00)	1.22 (0.00)	1.23 (0.00)
Rural household	0.57 (0.01)	0.50 (0.00)	0.55 (0.00)	0.55 (0.00)	0.55 (0.00)	0.54 (0.00)	0.59 (0.00)	0.61 (0.00)	0.61 (0.00)	0.59 (0.00)
N	8,312	45,817	95,921	126,917	142,108	175,497	222,063	133,251	139,733	142,341
	HOUSEHOLDS WITH AT LEAST ONE CHILD UNDER AGE 15 YEARS
	2006	2007	2008	2009	2010	2011	2012	2013	2014	2015
	Mean (SE)	Mean (SE)	Mean (SE)	Mean (SE)	Mean (SE)	Mean (SE)	Mean (SE)	Mean (SE)	Mean (SE)	Mean (SE)
Age of respondent (years)	38.28 (0.19)	36.25 (0.09)	35.60 (0.06)	34.92 (0.05)	35.67 (0.05)	35.42 (0.04)	36.05 (0.04)	36.08 (0.05)	36.20 (0.05)	36.25 (0.05)
Male	0.43 (0.01)	0.42 (0.00)	0.45 (0.00)	0.46 (0.00)	0.45 (0.00)	0.46 (0.00)	0.45 (0.00)	0.46 (0.00)	0.45 (0.00)	0.45 (0.00)
Education	0.69 (0.01)	0.87 (0.00)	0.79 (0.00)	0.71 (0.00)	0.72 (0.00)	0.74 (0.00)	0.74 (0.00)	0.74 (0.00)	0.72 (0.00)	0.75 (0.00)
Log of per capita income	7.21	7.62	7.21	7.03	7.07	7.08	7.18	6.98	6.88	7.01
	(0.02)	(0.01)	(0.01)	(0.01)	(0.01)	(0.00)	(0.00)	(0.01)	(0.01)	(0.01)
Log of household size	0.93	1.32	1.59	1.74	1.70	1.71	1.63	1.67	1.62	1.64
	(0.01)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)
Rural household	0.59	0.48	0.57	0.57	0.58	0.58	0.62	0.65	0.67	0.64
	(0.01)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)
N	5,230	20,158	46,623	69,119	78,351	94,490	110,932	72,024	70,649	75,062
	AGES 15–24 YEARS
	2006	2007	2008	2009	2010	2011	2012	2013	2014	2015
	Mean	Mean	Mean	Mean	Mean	Mean	Mean	Mean	Mean	Mean
	(SE)	(SE)	(SE)	(SE)	(SE)	(SE)	(SE)	(SE)	(SE)	(SE)
Age of respondent (years)	19.89	19.57	19.60	19.71	19.69	19.69	19.65	19.69	19.68	19.76
	(0.08)	(0.03)	(0.02)	(0.02)	(0.02)	(0.01)	(0.01)	(0.02)	(0.02)	(0.02)
Male	0.45	0.47	0.48	0.49	0.48	0.48	0.48	0.48	0.47	0.47
	(0.01)	(0.01)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)
Education	0.80	0.86	0.78	0.77	0.78	0.78	0.79	0.77	0.77	0.77
	(0.02)	(0.01)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)
Log of per capita income	7.43	7.57	7.26	7.23	7.25	7.27	7.35	7.11	6.98	7.16
	(0.03)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)
Log of household size	1.20	1.30	1.44	1.59	1.58	1.57	1.49	1.52	1.46	1.47
	(0.02)	(0.01)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)
Rural household	0.56	0.48	0.54	0.53	0.55	0.56	0.60	0.63	0.63	0.62
	(0.01)	(0.01)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)
N	1,323	9,331	20,871	28,884	30,288	38,209	43,328	27,189	27,181	27,234

Notes:

Data comes from Gallup World Poll Surveys from 2006 to 2015. Estimates are from all countries in a given year, and therefore may not be comparable over time Estimates are weighted averages and the corresponding standard errors

Appendix 4: Notes on sources and calculations

Population estimates

All population estimates came from FAOSTAT [Accessed November 17, 2016], except for:

• Hong Kong for 2014 were provided by FAO and came from Wikipedia; 2015 estimates came from the Hong Kong census and statistics department (http://www.censtatd.gov.hk/hkstat/sub/so20.jsp)

• Kosovo for 2014 were provided by FAO and came from Wikipedia; 2015 estimates were calculated using an annual population growth rate of 0.009 (from the World Development Indicators, 2013)

• Northern Cyprus for 2014 were provided by FAO and came from Wikipedia (based on the 2011 census); we used the same estimate in 2015 because of missing data on the annual population growth rate

• All estimates of proportion of children under age 15 in a country came from the World Development Indicators, World Bank Database [Accessed May 25, 2017], except for:

• Estimates for Northern Cyprus and Taiwan came from UN Stats, as provided by FAO; we used 2014 estimates for both 2014 and 2015

• Only the 2015 estimate were available for Kosovo from the WDI database, so we used the 2015 estimate for 2014 as well

Appendix 5: Regional estimates of food insecurity prevalence among all households.

Notes:

Data comes from Gallup World Poll (GWP) Surveys and includes 289,933 respondents from 147 countries and 4 territories in 2014 and 2015. The Food Insecurity Experience Scale (FIES) was used to measure moderate or severe (FIES-M+), and severe (FIES-S) food insecurity. The GWP food insecurity indicator is: “Was there ever a time in the last 12 months when you did not have enough money to buy food?” The GWP selects and surveys one respondent aged 15 years or over per household. Household weights are used for the estimates for all households, and child weights for the HHU15 estimates.