The World has Changed - Time to Debate an Alternative Country Classification Model

The present country classification model has far-reaching normative consequences but as the world has changed its income thresholds look increasingly arbitrary. A debate is needed over what should replace it. The proposal developed below: Let country classification be defined in relation to the global median of GNI per capita, with the middle-income threshold set at the median and the low-income threshold set at a third of the median. Also, redefine the ODA target of 0,7% for advanced economies so that it becomes progressive, demanding more from those with higher capacity to provide assistance.     

Since 1987 the World Bank has classified all countries (and some “territories”) as High Income (HICs), Upper Middle Income (UMICs), Lower Middle Income (LMICs) and Low Income (LICs), using Gross National Income (GNI) per capita, with income thresholds fixed in real terms.

Country classification matters. It determines (sometimes combined with other criteria) eligibility for Official Development Assistance (ODA); prioritization of ODA (LICs and LDCs); eligibility for concessional lending; trade preferences; terms in various global agreements, including on climate. Country classification determines how countries are treated in this world, so watch that space!

With growth in real terms, the composition of the four country categories has changed (see graph above). The number of HICs have more than doubled, to above 80. The number of LICs have more than halved to just 26 countries (something to celebrate!). Haiti and Tanzania were the latest to become middle-income, probably defying the general perception of what a middle-income country “should” look like. At the time of introducing the present country classification system the distribution of the worlds’ economies by GNI per capita used to have two distinct humps, clearly separating the advanced economies from the developing world. This is no longer the case.    

In 1987 more than 90% of the world’s extreme poor lived in LICs. As of today, only approximately four out of ten do so. In fact, less than 10% of the global population live in LICs. Furthermore, with thresholds kept constant, these trends will continue (Ethiopia, Togo, Rwanda and Uganda seem to be next in line to become LMICs).

The LIC-list increasingly looks like a shrinking collection of extremely war-torn economies. For most of its normative uses, it is steadily losing relevance and needs to be replaced at some point. Meanwhile, the Middle-Income category, from Haiti to China, is becoming far too heterogeneous. Also, the group of high-income is looking increasingly heterogeneous, putting into question a unified 0,7% ODA target. Time is ripe for a debate over a new formula for country classification as well as for redefining global ODA commitments.   

When, in the 1970 UN resolution 2626, it was first agreed that the advanced economies should “exert its best efforts” to provide ODA at 0,7 per cent of their GNI, the rationale was also about “disparity” and “aggravation of world tensions”. So, at the time it was not only about absolute income levels, as captured by the present country classification system, but also about relative positions.

OECD’s poverty definition for households is a point of reference. Being income poor in an OECD member state means falling below half the median household income. EU applies a similar definition, but using 60 per cent of the median income. (EU also has a model for internal redistribution among its member countries determined by relative positions in terms of GNI per capita, with a view to reduce disparities and to keep the union together.)

The usual arguments in favour of a relative poverty concept for households are relevant also when dealing with the global community of nations; arguments related to inclusion, cohesion, collective action and redistribution. All of it is needed in a world which is drifting apart and becoming increasingly conflict-ridden, while it urgently needs to come together to address global challenges, the climate not the least.

So, what could an alternative country classification look like?

First, let the present HIC-threshold be replaced by the global median of GNI/capita (7620 USD 2023, World Bank). Countries above the median would no longer be eligible for ODA.

The median of 7.620 USD may appear generous as a threshold. However, the present HIC-threshold (14.006 USD) is far more generous. Some twenty countries eligible for ODA today would be excluded with the median as threshold, among them China, Brazil, Thailand, South Africa and Turkey. However, many of them will soon transit to HIC-status anyway (among them China).

Second, the LIC-threshold (today 1.146 USD) needs to be redefined. We still need a country category to guide preferential treatments of those furthest behind. All thresholds are somehow arbitrary, but a possible candidate would be a third of the median (2540 USD). The LIC-threshold in 1987, when the country classification system was established, was in fact approximately a third of the median of the GNI/capita at the time.

Third, the ODA target needs to be redefined, given the heterogeneity of today’s high-income group. To be fair, the target should apply to all countries above the median, but operate progressively demanding more from the richest. This can be achieved by letting the ODA target apply to the share of GNP that is above the world per capita median, with a slightly higher target than the present 0,7 %, tentatively 0,8%. (Formula: Share of GNI=0,8*(GNIcapita - World median GNIcapita) / GNIcapita).) A country just above the median would then have almost no ODA commitment to live up to. For most traditional donors a target of “0,8% of GNI above the world median”, would demand approximately the same effort as the today´s target of 0,7% of GNI. For a country such as Brazil “0,8% of GNI above the world median” would be equivalent to 0,1% of GNI. For Norway, which is at the top of the GNI per capita table, it would be equivalent to 0,74 % of GNI. (Of course one can imagine alternative formulas to achieve pregressivity; to be looked for is the formula that has the highest chance of being accepted.)       

The countries of the world would then fall into three new categories: High Income (those above the median, where the new ODA target would apply), Low Income (those below a third of the median) and Middle Income (those between a third and half of the median).

The table below summarizes and compares. In terms of numbers, there would be 22 additional High Income countries and 29 countries added to the Low Income group (then becoming very close to the LDC category of today).   

Present country classification (WB, based on 2023 data)

Threshold USD GNI/cap                       No. Countries

HIC                                    >14.006                                                   86

UMIC                                4.516 - 14.006                                       54

LMIC                                 1.147 - 4.516                                         51

LIC                                    <1.147                                                      26

Proposed Country Classification

              Threshold GNI/cap                      USD equiv. 2022      No. Countries

High Income              Above median                               >7.620                      107

Middle Income          Third of median - median           2.403-7.620             55

Low-Income               Below a third of median               <2.403                       55

What would the advantages be?

·       It would be rooted in three basic principles that appear to be reasonable: i) The upper half of the world’s economies should provide some support to the bottom half, which makes sense if ODA is to reduce global disparities.. ii) Those furthest behind should constitute a category worthy of preferential treatment. iii) Global agreements on providing ODA are to be made dependant on the capacity to do so.   

·       The Low-Income category would remain relevant for defining preferential treatment of the countries furthest behind (those below a third of the median). A very clear majority of the world’s extreme poor would reside in these countries, so a relevant category to target from that perspective. 

·       There would be less need to define new arbitrary income thresholds, or to invent other ad-hoc measures, when the present thresholds increasingly become unworkable for operational purposes.

·       ODA commitments would stand a better chance of meeting global acceptance, if rooted in a principle of fairness where most high-income countries are asked to contribute something but the richest are asked to provide more.

Would this imply that ODA and preferential treatments would be perpetuated forever? Not really. It is true that half of the countries will always be below the median, by definition, but being eligible for ODA does not mean being entitled to it. And as for countries below a third of the median, a world can be imagined where there are no such countries at all, if those furthest behind catch up. 

Another argument against may be that the present classification model makes real progress visible, as when countries pass absolute thresholds. However, real growth may mask growing disparities; disparities that the world needs to address. If, over time, there were to be fewer left-behind countries below a third of the median, then that would also be a progress to celebrate. Such progress is made invisible by today’s classification model. 

Comments welcome!

/Göran Holmqvist