CFS:2002/Inf.6


COMMITTEE ON WORLD FOOD SECURITY

Twenty-eighth Session

Rome, 6-8 June 2002

The LIFDC Classification - An Exploration

Table of Contents



I. INTRODUCTION

1. The FAO classification of low-income food-deficit countries (LIFDCs) was discussed at the 16th Session of the Committee on Agriculture (COAG) held in Rome during 26-30 May 2001 and at the 27th Session of the Committee on World Food Security (CFS) held in Rome during 28 May-1 June 2001, taking also into account the debates at the COAG. In the context of indicators for the assessment of world food security, the CFS recommended that the Secretariat give consideration to "analysing the suitability and rationale for the classification of low-income food-deficit countries (LIFDCs) as a basis for determining aid allocation".1 This Note is presented for information in response to this recommendation.

2. The specific recommendation of the CFS is considered in Section II. In view of the conclusion reached there and considering the context in which the LIFDC classification was developed originally and used over the years, the rest of the Note provides information and analysis on the LIFDC listing. In doing so, the Note also responds to some of the recent debates in this and other Committees. Section III describes the classification and criteria used. Section IV provides a commentary on the suitability of this classification to the analysis of food insecurity and how it relates to other similar indicators. Section V concludes.

II. THE LIFDC CLASSIFICATION AS A BASIS FOR AID ALLOCATION

3. The CFS recommendation to the Secretariat was to analyse the suitability and rationale for the classification of LIFDCs as a basis for determining aid allocation. To the extent the context is on allocating "aid" in general, it would seem clear that this classification can not be suitable for this purpose. This is principally because the classification, as explained in the next section, is based only on two criteria, i.e. low income and food deficit, while development aid in general seeks to address many other concerns of socio-economic underdevelopment. Not surprisingly, donors use a multiplicity of criteria for aid allocation. Moreover, these criteria typically vary from donor to donor, reflecting various objectives set by donors for providing aid and the weights placed on these objectives. In view of this, it is reasonable to conclude that the LIFDC classification is not suitable in general as a basis for determining aid allocation.

4. While this is so with reference to the aid in general, the LIFDC classification was developed in the context of analyzing food security issues and it was also in this context that this country listing was discussed at FAO Committee meetings, including its use as a basis for the allocation of food aid in relation to food insecurity. Therefore, it makes sense to discuss the classification in this context also. Following an introduction to the history of this classification in Section III, this discussion is undertaken in Section IV.

III. THE LIFDC CRITERIA

5. The LIFDC classification was defined by FAO in the late 1970s in the context of the analysis of food security issues. The key motivation was to create, for various analytical purposes, a sub-category of developing countries that are both low-income and are relatively exposed to external shocks in view of their food-deficit status. It was not necessarily meant to provide a basis for allocating aid. The list has been updated annually since then.

A. THE ORIGINAL CRITERIA

6. As the name indicates, the classification is based on two criteria: low-income, and food-deficit. The low-income criterion is defined in terms of per caput income levels below a ceiling (called "historical cut off") used to determine eligibility for International Development Association (IDA) assistance and 20-year IBRD terms (countries falling under World Bank's operational categories I and II). The IDA window (and the eligibility criterion) was set by the World Bank in the early 1960s in order to provide interest-free loans to countries that have low-incomes and lack creditworthiness for market-based borrowing from both commercial sources and IBRD. The historical cut off level was initially set at US$250 per capita in 1964; this is revised annually to account for inflation. In 2000, it was US$1 445. Annex Table 1 shows these levels since 1985. The IDA-eligible countries also include some small island states, considered to be a special category despite higher per caput incomes.

7. It may be noted the term "low-income" as used here is not the same as the low-income category of the World Bank, for which the income cut-off level was US$ 755 per capita in 1999.

8. The food-deficit criterion is determined by FAO on the basis of the food balance data. Until 1995, food deficit was defined as net trade deficit in cereals, calculated on the basis of the average volume of imports and exports of cereals for the five preceding years.

B. REVISIONS IN 1996

9. Since 1996, a broadened definition of food was used to compute the deficit including most of the basic foodstuffs. These were cereals, roots and tubers, pulses, oilseeds and oils other than tree crop oils, meat and dairy products. Second, net imports for a country were aggregated by the calorie content of the individual commodities. Third, for normalization across countries, these aggregate measures were expressed on a per caput basis. Finally, instead of using the average for the past five years, three-year averages were used to determine deficits, as this reflected better the current food situation.

10. A number of considerations were taken into account in this revision. First, while cereals were the major foods imported by developing countries, especially at the margin, imports of other foods were trending upwards strongly, which required a broader definition of foods. Some anomalies were also noted. For example, there were several cases of countries that were food deficit in terms of cereals but surplus under a broader commodity coverage.2 Second, lack of food balance data, especially for a large number of low-income countries, constrained the use of the broader definition of food in the past, which is not the case currently. Third, the switch to three-year averages for the computation of food deficits was motivated by the need to better reflect current food situations.

11. Also from 1996, a third (subjective) criterion was introduced, called the "self-exclusion principle". That is, in addition to the above two criteria, countries that formally objected to being included in the LIFDC category would not be included in the list, even though they fell into that category on the basis of the two statistical criteria.

C. REFINEMENTS INTRODUCED IN 2001

12. In the course of the preparation of the 2001 revision of the list, the Secretariat held a number of internal technical meetings in which it was noted that there was a rationale to avoid countries on the margin changing their LIFDC status too frequently.3 Importantly, there was a need to ensure that graduation from the list reflected sustained improvement in the situation rather than short-term, exogenous shocks. There were cases where a country graduated one year, e.g. due to a bumper crop, but qualified in the next as production returned back to normal.

13. Accordingly, a criterion called "persistence of position" was introduced in 2001. This consideration would postpone the "exit" of a LIFDC from the list, despite it not meeting either the income criterion or the food-deficit criterion, until the improvements were verified for three consecutive years. In other words, a country would graduate from the list in the fourth year after the statistics confirmed its position as no longer qualifying as a LIFDC. During these three years, the country in question would be considered to be in a transitional phase and would be so indicated in the LIFDC list. In the 2001 listing, six countries were in this phase (Annex Table 2). For example, Guatemala entered the transitional phase - although per caput income was higher than the IDA threshold, this was only for the second consecutive year. Bolivia also entered this phase as its net food-exporter status was for the second consecutive year only. Similarly, both India and Nepal were retained as being in the transitional phase as they were net food-exporters for the second consecutive year.

14. The LIFDC list has been recognized implicitly by various FAO Committees when such bodies were informed of the outcome of certain analysis using the listing. For example, food production in the LIFDCs is one of several indicators traditionally used for monitoring global food security by this Committee. The World Food Programme Committee on Food Aid Policies and Programmes adopted the LIFDC category in its Guidelines and Criteria for Food Aid (WFP/CFA: 7/21, June 1979), stating that "in the allocation of food aid resources, donor countries should give priority to low-income food-deficit countries". Also, FAO's Special Programme for Food Security (SPFS) uses this classification for selecting countries for participation in the SPFS with funding from the Organization's own resources, although extra-budgetary resources may be used for implementing the Programme in both LIFDCs non-LIFDCs.

IV. THE LIFDC CLASSIFICATION AND FOOD INSECURITY

15. An often asked question is to what extent does the LIFDC classification provide a good indication of food insecurity ?

16. Food insecurity is a multifaceted problem. Its causes range from inability to produce food for lack of resources, inability to import the shortfall for lack of means and inequity in the distribution of food for various domestic reasons. It is for this reason that no single indicator can adequately capture all dimensions of the problem. It is also the reason why, for example, the CFS monitors global food security on the basis of a number of indicators, food production in the LIFDCs being only one of them.

17. The World Food Summit defined food insecurity at the level of the individual and in terms of sustained access through time. Food insecure people are found in most countries - it is only a matter of degree. The LIFDC was not meant to, and does not, capture this aspect as squarely as the FAO indicator of the prevalence of the population undernourished (Box 1), now updated annually in the FAO publication State of Food Insecurity in the World (SOFI).


Box
1. FAO's prevalence rate of undernourishment

FAO has been publishing estimates of the prevalence of undernourishment in developing countries for many decades. Since 1999, these estimates are published not only annually but also for many individual countries (125 developing countries and economies in transition in the 2001 issue).1/ By using national-level data on the availability of food calories per capita, requirements and their distribution across population groups, estimates are made of the percentage of people in each country whose average calorie intake falls below the minimum required for living and light activity. The resulting number of undernourished is the prime indicator for monitoring progress towards the World Food Summit goal of halving the number of hungry people by 2015.

Although national-level data are used to derive these estimates, the methodology yields the prevalence of undernourishment in terms of individuals in a country. In this sense, this indicator comes closest to the WFS definition of food insecurity. Most countries in the world have some food insecure people - some may have only a few thousands while others may have millions. When expressed as a percentage of the population, it is possible to classify countries into sub-categories reflecting the severity of food insecurity, e.g. extremely low food insecurity when the prevalence rate is 2.5 % or less, and very high food insecurity when the rate is over 35 %, and so on.

1/ FAO, The State of Food Insecurity in the World (annual publication since 1999).
 

18. How does the LIFDC classification compare with the prevalence indicator ? Although the statistics show that LIFDCs are on the whole relatively food insecure, the relationship is not perfect. For example, the simple correlation coefficient between the prevalence rates and per capita incomes for the LIFDC sample is only negative 0.54, and not negative unity.4 The other LIFDC indicator, net food deficits, relates even more poorly than the prevalence rate, i.e. higher (or lower) levels of food deficits do not necessarily imply higher (or lower) rates of food insecurity (but they do indicate problems in other dimensions of food insecurity).

19. Obviously, the LIFDC classification was not meant to capture the key measure of food insecurity as the prevalence rate does. The rationale behind the LIFDC classification would seem to be that being both food deficit and low-income at the same time is a problem in that the country lacks resources not only to import food but also to produce sufficient amount of food domestically. It is the combination of these two factors that makes these countries food insecure as well as vulnerable to domestic and external shocks. The appropriate use of the LIFDC list is to capture this specific aspect of the food problem.

20. Table 1 below compares various socio-economic characteristics of countries belonging to the LIFDC and three other classifications commonly used for analytical and operational purposes, e.g. the LDCs, the World Bank's low-income countries and Heavily Indebted Poor Countries (HIPCs). As one would expect, there are considerable overlaps among them, despite the fact that the listings were drawn on the basis of different criteria. In large part, this is due to the presence in the LIFDC listing of the LDCs that are also common in the low-income and HIPC categories.

21. These variables are organised along the following lines:

22. The following conclusions can be drawn from Table 1. First of all, the LIFDCs as a category show a better performance on all the economic, trade related and socio-economic variables than the other categories. They have higher per-caput GDP and agricultural GDP on average as well as higher growth rates. Their ability to import food is better than that of any other category and they have better nutrition and health related outcomes than any of the other categories.

23. It is possible that this better performance is due to the presence of China and India in this category. This can be tested by removing these two countries from the LIFDC list and doing a comparison with the other categories. Comparing the LIFDCs without China and India with the LICs minus India5 (Rows 2 and 3 in Table 1), it is clear that the two categories are almost indistinguishable on average, except that the LIFDC category has a higher per-caput GDP and agricultural GDP.

24. This observation seems to suggest that any differences between the LIFDCs and the LICs are ultimately due to the presence of lower middle income countries in the former group, which have a higher per-caput GDP. This insight is borne out if the low income countries are excluded from the LIFDC category. The results are presented in Row 4 of Table 1. It is apparent that the LIFDCs excluding the low-income group are qualitatively different from all the other categories6 on all dimensions.

25. It should also be clear from the table that the HIPC countries and the Least Developed Countries have similar characteristics, except that the HIPC countries have slightly higher GDP per-caput and agricultural GDP per-caput and a worse external debt to GDP ratio than even the LDCs, as would be expected. In terms of nutritional and health outcomes, however, the performance of these two categories generally lags behind that of the others.

 

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

Country Grouping

No. of Countries

Real GDP per capita, mean 1997-99
(US $)

Real Agric GDP per capita, mean 1997-99
(US $)

Median Real GDP per capita Growth, 1992-99
(Percent)

Median Agric. Real GDP per capita Growth, 1992-99
(Percent)

External Debt/GDP mean
1997-99 (Percent)

Food Aid (Cereals)/
Cereal Imports, mean
1997-99 (Percent)

Food Imports/ Export Earnings mean 1997-99 (Percent)

Net Inter-national Reserves/ Food Imports, mean 1997-99

Proportion of Under-nourished, mean 1997-99

Under-5 Mortality Rate (per 1,000), 1999

Children under
weight (% of total under-5)

(1) LIFDC

81

603

125

4.3

1.6

45.2

22.5

6.5

16.7

23.4

79.4

27.8

  

(2) LIFDC excl.
China & India

79

618

135

1.3

0.3

87.8

25.1

12.9

4.2

24.3

108.7

33.7

 

(3) LIC excl. India

62

491

116

1.4

0.5

90.1

26.9

12.5

4.3

24.9

114.3

35.5

  

(4) LIFDC excl. LIC

24

798

143

8.9

3.1

23.4

5.3

4.4

22.8

11.5

38.7

9.6

  

(6) LIC

63

464

111

4.3

1.6

60.8

23.5

7.9

11.5

30.3

104.2

40.2

  

(7) HIPC

41

332

99

1.2

0.6

122.6

36.6

17.0

4.1

34.3

153.3

30.0

  

(8) LDC

48

284

82

2.1

0.6

103.5

46.5

18.8

3.8

36.1

148.2

40.0

Sources: Columns 2, 3, 4, 5, 6, 8, 9, 11,12: World Development Indicators 2001

             Columns 8, 9: FAOSTAT

             Column 10: The State of Food Insecurity in the World, 2001

 

Notes:  Column 9: This variable indicates how many years of imports of food could be covered by the available international reserves.

            Column 12: Latest Observation 1994-99. Underweight is defined as more than 2 standard deviations below median weight for age.

 

V. CONCLUDING REMARKS

26. The purpose of this Note was to respond to a specific request by the 27th Session of the CFS to analyse and consider the suitability for the LIFDC classification as a basis for determining aid allocation. It was concluded in Section II that the classification would not be suitable for allocating aid in general, in view of the fact that development aid seeks to address many other aspects of socio-economic underdevelopment not covered by the LIFDC criteria.

27. Since the LIFDC classification was conceived in the context of food security, a brief commentary was made on the extent to which this classification provides a good indication of the status of food insecurity, and how this category compares with similar other listings. The main conclusion was that there were considerable overlaps among the LIFDC and other listings, e.g. the LDCs, the World Bank's low-income group and HIPCs. Moreover, although the LIFDC countries are, on the whole, relatively more food insecure, the classification is not as close an indicator of food insecurity as is the prevalence rate of undernourishment.

28. The main conclusion is that different country classifications and indicators address different aspects of food insecurity. Rarely does a single criterion suffice, especially when the problem is multifaceted as is food insecurity. Drawing upon several indicators or country classifications in making decisions relating to food security is important.

Undisplayed Graphic

 

Undisplayed Graphic

________________________

1 Paragraph 9 of the Report of the Twenty-Seventh Session of the Committee on World Food Security, 28 May - 1 June 2001.

2 A typical example was that some countries imported cereals for livestock feeding and exported livestock products in turn. There were opposite cases also, i.e. surplus in cereals but food deficit overall.

3 Annex Table 1 shows which countries entered into and exited from the LIFDC list since 1985.

4 Of the 82 LIFDCs in 2002, 47 are least-developed countries (LDCs) with 40% of the population undernourished on average (population-weighted). The prevalence rate for the rest of the LIFDCs was 11% (17% average for all LIFDCs).

5 China is not included in the LICs.

6 However this category, LIFDC minus LIC, still includes China. It is clear that the average reserve ratio will be lower if China is excluded, but all the other variables will be even higher since this category includes a number of transition economies as well as countries from the Near East and North Africa region.