Progress Towards Ending Hunger And Malnutrition: A Cross-Country Cluster Analysis

Cluster analysis suggest countries can be classified into six types of food security and nutrition situations:

• Type 1: High hunger and high child undernutrition - Cluster of countries with high prevalence of undernourishment (median: 42 percent) and high rate of child stunting (median: 41 percent).
• Type 2: Moderate hunger but high child undernutrition - Cluster of countries with moderate prevalence of undernourishment (median: 27 percent) but high rate of child stunting (median: 45 percent).
• Type 3: Moderate hunger and moderate child undernutrition - Cluster of countries with moderate prevalence of both undernourishment (median: 20 percent) and child stunting (median: 35 percent).
• Type 4: Moderate child undernutrition and moderate adult overweight - Cluster of countries with moderate prevalence of both undernourishment and child stunting (medians: 14 percent and 23 percent respectively) and moderate prevalence of adult overweight (median: 43 percent).
• Type 5: Low child undernutrition and moderate adult overweight - Cluster of countries with low prevalence of undernourishment and child stunting (median: 5 percent for both) and moderate prevalence of adult overweight (median: 32 percent).
• Type 6: No hunger, but high adult overweight - Cluster of countries with no hunger (median: prevalence of 2 percent for both undernourishment and child stunting) but having high prevalence of adult overweight (median: 55 percent).

Detailed Observed Pathways From Undernourishment To Overweight, 1990-2015 (Fig 4)

Source: David Laborde, Abdullah Mamun, & Rob Vos, Progress towards Ending Hunger and Malnutrition, FAO and IFPRI, 2020.

We have here the detailed observed pathways by country and over time between 1990 and 2015. In a stylized path we can expect that countries tend to take a path over time that reduces the prevalence of undernourishment but increasing the prevalence of overweight. This is, of course, not the desired pathway, as one would hope to see countries move along the preferred path of reducing both the prevalence of undernourishment and the prevalence of overweight. Countries of type [3] and [4] especially tend to suffer from this double burden. At the country level, we also observe that many countries suffering from a double burden of malnutrition.

Stylized Pathways From Undernourishment To Overweight, 1990–2015 (Interactive Version Of Figures 3 And 5)

(Numbering Of Figures Below Refers To That Used In The Report)

Source: David Laborde, Abdullah Mamun, & Rob Vos, Progress towards Ending Hunger and Malnutrition, FAO and IFPRI, 2020.

Evolution Of Size Of FSN Clusters Over Time, 1990-2015 (Fig 2)

Source: David Laborde, Abdullah Mamun, & Rob Vos, Progress towards Ending Hunger and Malnutrition, FAO and IFPRI, 2020.

Most countries that moved out of Types 1 and 2 since 1990 belonged to Type 3 In 2015, as both hunger and child stunting had fallen to moderate levels. This type expanded from 17 countries in 1990 to 28 countries in 2015.

Adult overweight and obesity had risen to notable levels in 10 countries by 2015, accompanied by moderate levels of child stunting. The double burden of malnutrition is heaviest in this type. The number of countries in this type reached a high of 25 in 2000; since, it has shrunk, as many countries moved from Type 4 to Type 5.

This country grouping, having moderate adult overweight but low or no hunger and child undernutrition, is rather heterogeneous. It includes both middle-income economies that by and large have managed to resolve their undernourishment problem but are seeing a rising overweight problem. Only two economies in this type – and in fact in the entire country sample – managed to contain increases in the prevalence of adult overweight (i.e., Japan and the Republic of Korea).

The last cluster contains most of the advanced economies, as well as a large portion of the countries in Latin America, the Middle East and Northern Africa, and Central Asia. By 2015, this type has grown most in size, increasing from 43 to 79 countries between 1990 and 2015. Thus, most countries in this group have reached low to zero undernourishment but are now facing high prevalence of overweight and obesity as their major malnutrition and health problem.

In 2015, there were 8 countries (most in East and Central Africa) with both a high prevalence of undernourishment and a high rate of child stunting, down from 23 countries in 1990.

In 2015, 7 countries (including India, Yemen and several countries in central and eastern Africa) still faced high hunger but had reduced child undernutrition to moderate levels. This group also has shrunk considerably in size, as it consisted of 20 countries in 1990.

Food Security And Nutrition Clusters (Map)

Source: David Laborde, Abdullah Mamun, & Rob Vos, Progress towards Ending Hunger and Malnutrition, FAO and IFPRI, 2020.
Map source: United Nations [United Nations Office for the Coordination of Humanitarian Affairs] map. 7 May 2020
Map disclaimer: The boundaries and names shown and the designations used on this map do not imply official endorsement or acceptance by the International Food Policy Research Institute. (IFPRI).

Changes In The Prevalence Of Undernourishment And Adult Overweight Between 1995 And 2015 By FSN Cluster (Fig 6) (Changes In Percentage Points)

Source: David Laborde, Abdullah Mamun, & Rob Vos, Progress towards Ending Hunger and Malnutrition, FAO and IFPRI, 2020.

The above figure shows the trajectory of the prevalence of undernourishment and of adult overweight over the past 20 years.

Most countries belonging to Types 1 and 2 in 1995 managed to reduce the prevalence of undernourishment, though to varying degrees, without much increase in the prevalence of overweight; only North Korea, Zambia, and Tajikistan witnessed a rise in the prevalence of undernourishment.

Countries with moderate adult overweight and child undernutrition (Type 4) in 1995, generally managed to reduce undernourishment, subsequently. However, they also witnessed steep rises in the prevalence of overweight and, as a result, moved either into Type 5 or Type 6. Likewise, most countries that were in Type 5 in 1995 saw significant increases in the rate of adult overweight.

Very few countries belonging to Type 6 in 1995, having already high rates of adult overweight, managed to prevent further increases. In fact, most faced increases in the prevalence of overweight by 10 percentage points or more.

Structural Conditions Clusters (Maps)

Source: David Laborde, Abdullah Mamun, & Rob Vos, Progress towards Ending Hunger and Malnutrition, FAO and IFPRI, 2020.
Map source: United Nations [United Nations Office for the Coordination of Humanitarian Affairs] map. 7 May 2020
Map disclaimer: The boundaries and names shown and the designations used on this map do not imply official endorsement or acceptance by the International Food Policy Research Institute. (IFPRI).

The above figure shows the trajectory of the prevalence of undernourishment and of adult overweight over the past 20 years.

Most countries belonging to Types 1 and 2 in 1995 managed to reduce the prevalence of undernourishment, though to varying degrees, without much increase in the prevalence of overweight; only North Korea, Zambia, and Tajikistan witnessed a rise in the prevalence of undernourishment.

Countries with moderate adult overweight and child undernutrition (Type 4) in 1995, generally managed to reduce undernourishment, subsequently. However, they also witnessed steep rises in the prevalence of overweight and, as a result, moved either into Type 5 or Type 6. Likewise, most countries that were in Type 5 in 1995 saw significant increases in the rate of adult overweight.

Very few countries belonging to Type 6 in 1995, having already high rates of adult overweight, managed to prevent further increases. In fact, most faced increases in the prevalence of overweight by 10 percentage points or more.

Structural Clusters

Five country types, representing a variation on the phases of the agricultural transformation typology

Type A is mainly composed of advanced economies, which have passed the demographic transition to very low birth rates, have a low share of agricultural GDP, and have very high levels of agricultural land and labor productivity. Their institutions are strong, as reflected in low levels of the indicators for corruption and political instability.

Type B consists of slightly less-advanced economies, with relatively low dependence on agriculture and with improved food availability achieved through agricultural intensification and productivity growth. The quality of institutions is moderate.

Type C consists of countries with relative land abundance and high dependence on natural resources, including in metals and minerals. Non-agricultural commodity dependence defines much of the wealth and structural transformation in these economies. Corruption tends to be high, but fiscal space created by natural resource wealth appears to support relative political stability.

Type D, in contrast, consists of countries facing land scarcity, but having generally better agro-ecological conditions and trading options than Type E.

Type E is largely composed of low-income landlocked countries with still some potential for agricultural development at the extensive margin but constrained by relatively poor agro-ecological conditions and high transaction cost to trade.

Evolution Of The Number Of Countries Per Cluster Of Structural Conditions (Fig 7)

Source: David Laborde, Abdullah Mamun, & Rob Vos, Progress towards Ending Hunger and Malnutrition, FAO and IFPRI, 2020.

While defined as structural variables, the related conditions are not static and do change over time. Accordingly, countries have shifted from one cluster to the next depending on the processes of structural transformation. Changes in the composition of the clusters over time are shown

Several structural features are foremost time invariant, such as landlocked status or biophysical agricultural potential.11 Most other variables, however, are time variant, being linked to demographic transitions, economic diversification, improvement or deterioration of institutions, and conflicts.

This clearly shows that the number of rural economies (clusters D and E) has fallen significantly (by one third) between 1995 and 2015. This has happened at different speeds varying by country. Unsurprisingly, the decline has been rapid among the rural economies with a good agricultural potential for intensification and potential for trade (type D), which has facilitated economic diversification, including through export-oriented strategies.

Fewer countries have moved out of Type E, which consists of landlocked countries with poor agricultural potential and weak integration into regional and global value chains. Among countries moving out of Type E, a few have progressed towards Type C, but most have become part of Type B.

Changes In The Prevalence Of Undernourishment And Adult Overweight Between 1995 And 2015 By Structural Conditions (Fig 8)

Source: David Laborde, Abdullah Mamun, & Rob Vos, Progress towards Ending Hunger and Malnutrition, FAO and IFPRI, 2020.

The above figure shows the change in the prevalence of undernourishment and that in adult overweight between 1995 and 2015 by countries and their cluster position in 1995.

As expected, it shows that the more advanced and transforming countries (Types A and B) tended to show increases in the prevalence of adult overweight, while maintaining low to zero rates of hunger and undernutrition. For the other structural clusters (C, D, and E), the figure shows a rather dispersed pattern of rates of reduction in undernutrition and increase in adult overweight, which could suggest that the impact of structural conditions is strongly mediated through policy.

In the policy space six groups can be described as follows:

Type [α] includes many large agricultural exporters (both among developed and developing countries in 2015) who’s policies have been characterized by moderate public support and low tariffs. It is the level of protection of the agricultural sector that differentiates this group from the next.

Type [γ] which has much higher tariffs and is currently (2015) comprised of several large middle-income countries, like India, Bangladesh, and Mexico.

Type [β] also shows – on average – a slightly higher degree of agricultural protection than Type [α], but countries in this type provided much stronger support to the agricultural sector through public spending allocations and investments in rural infrastructure, rather than through tariff protection.

Type [κ] includes countries where agriculture is heavily subsidized, such as in Japan, the Republic of Korea and some European countries.

Type [δ] includes countries with relatively high tariffs on agricultural products but which provide limited farm support and already have broad, though still incomplete coverage of rural infrastructure. Gender disparity was still great in this cluster.

Type [ω] includes most sub-Sahara African countries and least developed countries in Asia, such as Afghanistan. They are more protectionist than average, have invested relatively little in agriculture, have insufficient rural infrastructure, and still have large gender disparity.

Evolution Of The Size Of Policy Clusters Between 1990 And 2015 (Fig 9)

Source: David Laborde, Abdullah Mamun, & Rob Vos, Progress towards Ending Hunger and Malnutrition, FAO and IFPRI, 2020.

Above we see shifts in the size of the clusters, reflecting significant shifts in policy stances by countries over time. Mostly, shifts have been from Type [γ] to Type [α], reflecting significant lowering of tariffs on agricultural products. Seventeen countries made this shift over the past 25 years.

A second set of countries (14) increased farm support (both subsidies and import tariffs) and stepped up public investment in rural infrastructure, thus moving into Type [β]. A smaller number of countries (10) moved from cluster [γ] to [β], as they invested further in rural infrastructure, increased direct support to farmers, but reduced agricultural protection. Seven countries shifted from Type [δ] to Type [γ] after partially liberalizing agricultural trade, while eight countries increased protection and shifted from Type [α] to Type [γ].

Policy Clusters (Maps)

Source: David Laborde, Abdullah Mamun, & Rob Vos, Progress towards Ending Hunger and Malnutrition, FAO and IFPRI, 2020.
Map source: United Nations [United Nations Office for the Coordination of Humanitarian Affairs] map. 7 May 2020
Map disclaimer: The boundaries and names shown and the designations used on this map do not imply official endorsement or acceptance by the International Food Policy Research Institute. (IFPRI).

Policy Stance And FSN Performance Over Time, 1990-2015 (Fig 10)

Source: David Laborde, Abdullah Mamun, & Rob Vos, Progress towards Ending Hunger and Malnutrition, FAO and IFPRI, 2020.

Figure above shows a simple comparison for all countries between the prevalence of undernourishment and the prevalence of overweight, on the one hand, and the FSN policy stance (and changes therein), on the other. It clearly shows that – though also making progress – the countries in policy cluster [ω] and to a lesser extent those in Groups [δ] and [κ] continue to have highest levels of undernourishment. Policy cluster [ω] has low investment in rural infrastructure, limited direct farm support, average protection of agriculture and wide gender disparities.

Good performers in terms of lower rates of undernutrition have followed both more and less protective agricultural trade policies, though all have good to very good infrastructure and greater gender parity. The policy typology is mostly relevant to identify efforts at reducing food energy deficits (hunger) and child undernutrition, as we lack identifiers for policy efforts directed at reducing overweight and obesity.