The data behind the debate over Uganda’s income status
DI's Zach Christensen and Dean Breed look into the contested statistics behind the dispute between UBOS and the World Bank about Uganda's income status.
In July 2022, Ugandan President Yoweri Museveni announced that the country had achieved lower-middle-income status, a claim that the World Bank quickly disputed. We examine the contested data and explain why settling such debates can be far from straightforward.
Quantifiable economic growth, or lack thereof, carries huge importance for countries, despite recognition of its limitations as a measure of economic and social progress. Even the data used to calculate the size of a country’s economy and its population can be controversial and have important consequences, raising questions about the methodology used, who has ownership of a country’s statistics, and how this data should be used to inform decision-making.
The World Bank assigns every country to one of four income categories: low, lower-middle, upper-middle or high. These categorisations are updated every year, based on data that measures a country’s Gross National Income (GNI) per capita in US$ against the thresholds for each category established by the bank. While these categories have no explicit operational implications for the bank, they remain important and influential globally.
Countries can move between categories when there are material changes in national income or population, as well as due to exchange rate fluctuations against the US$, or changes in the thresholds set out by the World Bank.
Uganda: A low-income or lower-middle-income country?
President Museveni’s announcement was based on data from the Uganda Bureau of Statistics (UBOS), showing that Uganda had grown to the level of a lower-middle-income country, meeting the World Bank’s income threshold of US$1,046 per capita for the 2021–2022 financial year (FY). The World Bank quickly disputed this claim based on its own calculations. Table 1 presents the contrasting figures from UBOS and the World Bank; more detail on the differences in the calculations is available in this methodology note.
Table 1: Income per capita data for Uganda from UBOS and the World Bank
|UBOS||World Bank||Difference||Why does this difference exist?|
|Income per capita||US$1046||US$840||US$206|
|Constituent parts: National accounts (total income)||GDP US$45.7 billion (July 2021–June 2022)||GNI US$39.5 billion (2021 calendar year)||UBOS’s figure is US$6.2 billion (16%) larger||Firstly, different measure of income: UBOS used GDP as a measure of the size of its economy, whereas the World Bank used GNI, consistent with its standard approach and the thresholds. Secondly, while GDP is not always larger than GNI, in the case of Uganda, the World Bank also estimated GDP to be higher than GNI, (by around 2%), but its estimate of Uganda’s GDP differed to that of UBOS. For more information on the numbers presented in this blog, see the methodology note. Thirdly, different time period: UBOS based its calculations on FY 2021–2022 (July to June), while the World Bank used the calendar year 2021.|
Constituent parts: Total population (total adults and
|43.7 million||47.1 million||
The World Bank’s estimate is 3.4 million
|Different adjustments to census data: Both used the 2002 and 2014 national census data as the basis of their 2021 estimates. The World Bank adjusted this data upwards due to post-enumeration surveys, based on which the UN believe there is an undercounting of around 2.3 million people, and projected a steeper growth in population since 2014.|
Compared with the World Bank, UBOS used more recent estimates for the size of Uganda’s economy and measured it using GDP rather than GNI. Combined with lower estimates for the total size of the population, UBOS calculates that Uganda has crossed the World Bank’s threshold for a lower-middle-income country.
Why does this matter?
Like most thresholds in international statistics, the definition of a lower-middle-income country can seem arbitrary. Sometimes falling above or below a key threshold can depend on technical methodological choices that are unlikely to reflect meaningful changes for most of the population. The limitations of any income figures to describe meaningful progress for people and societies have been well articulated. A nation’s or government’s performance should be assessed by multidimensional measures of progress, beyond just material productivity, and that also take into account the sustainability of the planet. Moreover, these aggregated national figures tell us nothing about how different people within a country are doing, and who is left behind.
However, this does not mean that these thresholds are inconsequential. There is potentially an incentive for a country to understate GNI per capita to fall just below an income status threshold. The World Bank considers GNI per capita as part of its criteria to determine eligibility for highly concessional funding from the International Development Association (IDA). One study argues that aid-dependent countries are likely to report GNI per capita numbers just below the threshold for eligibility for IDA funding, but scrutiny of the numbers by the international community seeks to remove that bias. The argument that thresholds are set too low, decreasing aid eligibility, has also led to protests and ad campaigns targeting the World Bank.
But there can be substantial political benefits for governments that can claim they have enabled their country to reach a higher income status. According to a study of authoritarian regimes, some authoritarian leaders sought political gain by undermining the independence of national statistical offices when it suited their needs, manipulating GDP numbers by up to 30%.
As the debate in Uganda demonstrates, the stakes around national income categories, as well as other headline national and sub-national indicators of progress, can be high and the data used to calculate them strongly contested. The UN Statistics Division has established a set of fundamental principles for national statistics; these seek to establish a common understanding of good statistical practice for all national data and include the right to comment on the misinterpretation and misuse of statistics by others.
In all countries, greater transparency, consultation with local stakeholders, independent national statistical offices and more resources for national statisticians could go a long way to addressing many of the controversies that can arise.
For further details on the differences between UBOS and the World Bank's calculations, please see our methodology note.
- 1The thresholds are updated every year; the threshold for lower-middle-income status for 2022–2023 (FY) has been set at US$1,086.
Multidimensional poverty: Measures and frameworks to leave no one behind
An overview of multidimensional approaches to poverty measurement to help practitioners and policymakers identify who is left behind in their context.
Why we need a multidimensional approach to measuring poverty that leaves no one behind
DI's Deborah Hardoon explains how multidimensional approaches to poverty measurement can help to promote inclusion and identify inequality
Inequality, measuring it and why it matters for poverty reduction
This briefing explores the relationship between poverty and inequality, and shows that measuring, understanding and reducing inequality should be central to efforts tackle poverty, discrimination and exclusion.