Policy Research Working Paper 11093
Global Inequality and Economic Growth
The Three Decades before Covid-19 and Three Decades After
Diana C. Garcia Rojas
Nishant Yonzan
Christoph Lakner
Development Economics
Development Data Group
March 2025
Policy Research Working Paper 11093
Abstract
Global income inequality captures income differences increase in global income inequality in at least three decades.
among all individuals around the world. Global inequality The future of global inequality largely depends on how
around the world increased from 1820 to 1990 as incomes incomes grow in various parts of the world. If the trends
in richer countries grew faster than incomes in relatively of the last three decades continue, inequality may increase
poorer countries. However, these trends were reversed as growth in those countries that drove the reduction in
over the three decades starting in 1990. Inequality among inequality now contributes to increasing inequality, since
all citizens of the world decreased as populous and rela- these countries are in the upper part of the global distribu-
tively poorer countries, in particular China, reduced the tion. However, if poorer countries today grow faster than
income gap with richer parts of the world. Growth in aver- their richer peers, global inequality could continue to fall.
age incomes played a critical role in this reduction, with Climate adaptation and mitigation challenges will play an
differences within countries contributing relatively little. increasing role in shaping country-level growth trends and
The Covid-19 pandemic abruptly halted the reduction in thus the changes in global income inequality.
global income inequality and was responsible for the largest
This paper is a product of the Development Data Group, Development Economics. It is part of a larger effort by the
World Bank to provide open access to its research and make a contribution to development policy discussions around the
world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may
be contacted at nyonzan@worldbank.org.
The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development
issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the
names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those
of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and
its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent.
Produced by the Research Support Team
Global Inequality and Economic Growth:
The Three Decades before Covid-19 and Three Decades After
Diana C. Garcia Rojas ⓡ Nishant Yonzan ⓡ Christoph Lakner 1
Keywords: Global inequality, economic growth, shared socioeconomic pathways.
JEL codes: D30, D31, O40
1
The author ordering was constructed through the American Economic Association’s randomization tool
(confirmation code: isYCrHkbRbsc). All authors are with the World Bank. Garcia Rojas can be contacted at
dgarciarojas@worldbank.org, Yonzan can be contacted at nyonzan@worldbank.org, and Lakner can be contacted at
clakner@worlbank.org. We are grateful for comments from Daniel Gerzon Mahler, Branko Milanovic, and
participants at the 2nd III/LIS Comparative Economic Inequality Conference. Prepared for Oxford Handbook of
Income Distribution and Economic Growth, Oxford University Press, edited by Gordon Anderson and Sanghamitra
Bandyopadhyay. We gratefully acknowledge financial support from the UK government through the Data and
Evidence for Tackling Extreme Poverty (DEEP) Research Programme.
1 Introduction
Global income inequality is the inequality in income among all citizens of the world. The levels
and trends of global inequality differ depending on the precise concept used (for example, see
Anand & Segal, 2008; Deaton, 2021). Milanovic (2005) has defined three ways to think about
global income inequality: Concepts 1, 2, and 3 inequalities. Concept 1 inequality considers only
the differences in mean incomes across countries, or just the inequality between countries. Concept
2 adjusts the former for population differences across countries. In other words, it captures
differences in mean incomes across countries, with each country weighted by its population. Both
concept 1 and concept 2 do not account for individuals’ personal income, and thus ignore the
inequality among individuals within a country. Concept 3 considers the interpersonal incomes of
everyone around the world, and thus incorporates both inequality between countries and inequality
within a country. For the discussion below, unless otherwise mentioned, any reference to global
inequality implies this interpersonal version.
The Gini index of the global income inequality was around 50 in 1820 and continuously increased
over the next 170 years to peak at around 70 right before the fall of the Berlin Wall (Milanovic,
2024). This increase in global inequality is largely explained by relatively faster income growth in
the West compared to other parts of the world. Between 1820 and 1990, per capita GDP in Western
Europe grew over 1,000 percent compared to roughly 300 percent in Asia and 125 percent in Sub-
Saharan Africa (Bolt & van Zanden, 2020). The increase in global inequality before 1990 was
entirely due to the differences in income between countries (i.e., Concept 2 inequality) as there is
little data on individual incomes. There are only a few household surveys in the 1980s and they
are largely non-existent (except for a few surveys in Western countries) before that. So, researchers
have relied on historical per capita GDP and population estimates to calculate global inequality for
the period before 1990. This evidence inevitably relies on strong assumptions. Nevertheless, the
historical estimates of global inequality reported above, based on work by Bourguignon &
Morrisson (2002), have been corroborated by others (Lakner & Milanovic, 2016; Van Zanden et
al., 2014).
The focus of this paper is the period after 1990, for which there is more household survey data and
we can thus capture global interpersonal income inequality. Lakner & Milanovic (2016) relied on
surveys conducted around reference years spaced 5 years apart from 1988 to 2008 to construct
their global interpersonal income distribution. In all their reference years besides the first year,
more than 90 percent of the global population was covered by these household surveys. Using an
updated methodology, Mahler ⓡ al. (2022) construct annual distributions of global incomes
starting in 1990 using survey data available from the World Bank’s Poverty and Inequality
Platform (PIP). Their global distribution of income is generated using surveys for 170 countries
covering 97 percent of the global population. Our results below rely on the same annual global
distribution defined in Mahler ⓡ al. (2022) and updated with the latest version of data available
in PIP.
2
The underlying survey microdata in PIP uses income aggregates to capture household welfare for
most industrialized countries as well as countries in Latin America and Europe. For most other
parts of the world, the surveys use a consumption aggregate to capture welfare. Inequality
measured by income tends to be higher than inequality measured by consumption (see Jenkins
2015; Haddad et al. 2024). While the levels of inequality are most likely different, there is evidence
that the trends may be similar (Alvaredo et al., 2023; Kanbur et al., 2022; World Bank, 2024b). In
what follows, we combine the income and consumption distributions to create a global
interpersonal welfare distribution. For ease of reference, we call this the global income
distribution. This is the same distribution that the World Bank uses to track global poverty, shared
prosperity, and national inequality.
This paper first outlines the trends in global income inequality from 1990 to 2019, then discusses
the changes in global inequality due to the Covid-19 pandemic in 2020, and finally, provides
potential pathways for global inequality in the coming decades. We use the income distribution
from PIP to report estimates of global income inequality from 1990 to 2024. For the years after
2024, we draw on growth and population forecasts estimated by Crespo Cuaresma (2017) and
inequality forecasts reported in Rao et al. (2019) to project the global income distribution to 2050.
The two sources provide various scenarios for GDP, population growth, and country-level
inequality changes that follow the Shared Socioeconomic Pathways (SSPs) framework. The SSPs
are scenarios that outline the impacts of socio-economic activities on climate mitigation,
adaptation, and vulnerability. In addition to the SSPs, we provide a baseline forecast of global
income inequality by projecting each country’s income distribution using their historical per capita
GDP growth data, which we refer to as the “business-as-usual” scenario. In other words, the
business-as-usual scenario gives a picture of future global income inequality based on the historical
growth performance.
Differences in income growth across the world have been key drivers of the changes in global
income inequality. While in the historical period (1820-1990), global inequality increased as
incomes in richer nations grew faster than incomes in relatively poorer nations, the opposite was
true in the three decades starting in 1990. Global inequality decreased in this period as populous
and relatively poorer countries, especially in East Asia, reduced their gap in average incomes with
richer parts of the world. Covid-19 abruptly halted the reduction in global income inequality and
was responsible for the largest increase in global income inequality in at least three decades.
Looking forward, we find that the right mix of growth both across countries and within countries
could reverse the impact of the pandemic on global income inequality.
While the growth in the past three decades reduced global inequality, it also came at a cost to the
climate. We show that more sustainable growth allows for the reduction in global inequality and
at the same time has a relatively small negative impact on the planet. However, this path requires
a rethink of the drivers of growth and places the emphasis on technological change that reduces
the negative impacts on the climate. On a more pessimistic note, we find that global income
inequality could increase if the secular trends that we have seen in the last three decades continue.
Unsustainable and relatively high growth in the same regions of the world that drove the reduction
in global inequality in the period before the Covid-19 pandemic, could, in a turn of events, lead to
3
an increase in inequality globally in the future—dubbed the global inequality boomerang (Kanbur
et al., 2022). Nevertheless, most of our scenarios suggest that the decline in global income
inequality observed since 1990 will moderate over the next three decades or the pace of decline
slows down. 2
The rest of the paper proceeds as follows. Section 2 outlines the data, Section 3 reports the global
income inequality trends in the period before and during the Covid-19 pandemic and looking
forward to 2050. Section 4 discusses the drivers of the trends in global income inequality and
concludes.
2 Data
We use data on income and consumption available from the World Bank’s Poverty and Inequality
Platform (PIP). PIP includes over 2,000 household surveys from 170 countries covering more than
97 percent of the world’s population. PIP directly uses survey micro data wherever available and
supplements these with grouped data. 3 To generate annual distributions for all countries, PIP
interpolates and extrapolates surveys to create a continuous and balanced dataset for all countries
between 1990-2024. 4 The extrapolations and interpolations assume distribution-neutral growth,
which means that income or consumption of each household grows in line with the growth rate in
national accounts in that country, adjusted for the difference in growth rates between national
accounts and surveys. 5 For example, survey data are available for 78 countries (covering 67 percent
of the global population) in 2019. For the remaining 91 countries, PIP interpolates/extrapolates the
closest surveys to 2019 assuming all households grow with the real per capita growth rate of
Household Final Consumption Expenditure or GDP. 6 For 3 percent of the world’s population
residing in countries with no data in PIP, PIP assigns the region’s population-weighted average
distribution to that country.
2
However, it is also important to consider that the GDP growth scenarios under the SSPs are very optimistic relative
to historical performance, especially for low-income countries (on related concerns on the SSPs, see Welch, 2024).
3
There are 165 countries that either have micro data or 400-point equally weighted binned data obtained from micro
data. For an additional 5 economies (China, United Arab Emirates, Algeria, Qatar, and Trinidad and Tobago)
constituting about 18 percent of the world’s population, only quantile shares and the overall mean income or
consumption are available. For these countries, PIP fits a parametric Lorenz curve to recover a full distribution.
4
As of September 2024, the last year with enough data coverage to report a global poverty estimate in PIP is 2022.
Only 4 countries had survey data available for 2023. For 2023 and 2024, PIP also provides nowcasts using a similar
extrapolation methodology from the latest household survey. We rely on these nowcasted distributions for global
income distributions in 2023 and 2024. While data from the 1980s is available in PIP, only a limited number of
countries have surveys in that decade. Hence, this study focuses on the period beginning in 1990.
5
Growth in national accounts has been shown, among a thousand different development indicators, to be the most
reliable predictor of income growth at the household level (Mahler et al., 2021). PIP uses a growth passthrough of 0.7
for countries that have consumption-based surveys and a growth passthrough of 1 for countries that have income
surveys. For further details, see the PIP methodological handbook (https://datanalytics.worldbank.org/PIP-
Methodology/).
6
The República Bolivariana de Venezuela does not have national accounts data for the years after 2014. Hence, PIP
has data for 170 countries for 1990-2014 and for 169 countries after 2014.
4
Given survey availability, PIP combines surveys in which consumption is the welfare measure
with surveys in which income is the welfare measure. This presents challenges as inequality
measured by income has been found to be higher than that measured by consumption (Haddad et
al., 2024; World Bank, 2024b). For this analysis, we follow PIP and combine income and
consumption series and, for ease of communication, refer to both as income henceforth. In
addition, the surveys are designed to capture income at the household level. Hence, we further
assume that the total household income is divided equally among all members of the household,
which means any inequality within a household is unaccounted for. In sum, we measure daily per
capita household disposable income, adjusted for international price differences using the 2017
purchasing power parities (PPPs).
Household surveys often do not fully capture the richest individuals in the distribution due to
underreporting and nonresponse bias (Atkinson & Piketty, 2007; Jenkins, 2017). Surveys are also
known to inadequately capture entrepreneurial/self-employment and capital income, which is often
concentrated at the top (Burkhauser et al., 2015; Flachaire et al., 2023; Piketty et al., 2019; Yonzan
et al., 2022). While most of the literature has focused on income data, there is evidence that
consumption expenditure surveys also underestimate the top because of incomplete coverage of
spending on durables (Aguiar & Bils, 2015). As a result, inequality within countries measured
using survey data is typically lower than inequality measured using data that better captures the
very top income earners, such as survey data appended with administrative tax records (Piketty
and Saez 2006; Saez and Zucman 2016). However, with the data currently available, it is not
possible to do these kinds of adjustments comprehensively for all countries in the world (Anand
& Segal, 2008; Deaton, 2005; World Bank, 2017). 7 Therefore, our analysis is based on survey data
unadjusted for missing top incomes.
For ease of reproducibility, the calculations reported here are performed on the “binned” version
of the global dataset as defined in Mahler ⓡ al. (2022) (MYL henceforth) and updated using the
September 2024 vintage of PIP. The binned global income distribution consists of income vectors
grouped into 1,000 equally weighted observations for each country covering the period 1990-
2024. 8
The analysis for 2020 uses two methodologies. Our preferred estimate is based on the
interpolated/extrapolated data from PIP, which does not account for within-country distributional
7
For most countries, the World Inequality Database (WID, www.wid.world) presents inequality estimates that have
been adjusted for missing top incomes, and the gap between household surveys and national accounts. We do not use
these estimates for several reasons. First, many strong assumptions are needed to go from a household survey
distribution to an annual income estimate that is consistent with national accounts. These assumptions are strongest in
low- and middle-income countries, where the typical survey is infrequent and uses consumption expenditure. Second,
in many of the low- and middle-income countries, information for top incomes, such as from administrative tax
records, is rarely available or might be of limited relevance since the tax base is small and evasion common. Third,
and most importantly, in the WID database, it is not possible to disentangle the adjustment for the missing top from
the adjustment to national accounts. Surveys provide the most appropriate measure of household welfare, since
national accounts aggregates include many other components that are only vaguely related to household welfare
(Anand and Segal, 2008; Deaton, 2005; World Bank, 2017). The extrapolations and interpolations in PIP use growth
rates from national accounts but do not adjust for the differences in levels.
8
Link: https://datacatalog.worldbank.org/search/dataset/0064304/1000_binned_global_distribution.
5
changes for countries that do not have survey data in 2020. We also use the updated version of the
series reported by MYL which accounts for distributional changes in 2020 and thus provides
further insights into what happened during Covid-19. MYL triangulate various sources of data
beyond the survey data available in PIP to get a better picture of the change in global income
inequality between 2019 and 2020.
For the years after 2024, the income distribution in each country is projected based on (i) a
distribution-neutral “business-as-usual” scenario and (ii) the Shared Socioeconomic Pathways
(SPPs). For the former, we use per capita GDP growth rates from the World Bank’s World
Development Indicators supplemented by forecasts from the World Bank and the IMF, which
extend to 2029. 9 Beyond this, each country’s income distributions are projected to 2050 using the
average annual per capita growth rates experienced by the country in the decade before the
pandemic (i.e., 2010-2019). 10 This means that inequality within countries is held fixed at the level
of the last household survey in that country. For this scenario, changes in global inequality
projected after 2024 will only account for the differences in income growth between countries (or
Concept 2 inequality).
For the SSPs-based scenario, we use the country forecasts of per capita GDP and population
reported in Crespo Cuaresma (2017) and the forecasts of inequality reported in Rao et al. (2019).
Both studies provide forecasts that are consistent with the SSPs framework. The SSPs framework
provides important insight into how the world could evolve as countries face climate mitigation
and adaptation challenges. They provide pathways that would lead to different emission and global
warming scenarios. These scenarios are influenced by economic development, income
distributions, other welfare dimensions such as education and health, environmental and ecological
factors, technological change, governance and institutions, non-climate policies, and other factors
such as lifestyles (O’Neill et al., 2017). Some of these categories, such as total factor productivity
(TFP) and education attainment, are also known drivers of within-country income inequality (see
Atkinson & Bourguignon, 2000; Roser & Cuaresma, 2016). Using these factors that are included
in the SSPs, and others that are not directly incorporated but affect income inequality (e.g., trade
and the labor share of income), Rao et al. (2019) predict the evolution of each country’s Gini index.
We rely on the annual changes in a country’s Gini index reported in Rao et al. (2019), and the
annual changes of per capita GDP and population reported in Crespo Cuaresma (2017) to estimate
the changes in each country’s income distribution after 2024. We follow the methodology
described in Lakner et al. (2022) to estimate the changes in each country’s welfare distribution
given the interaction between income growth and inequality changes. 11
The five SSPs capture different scenarios as follows: SSP1 considers low levels of both within-
and between-country inequality mostly driven by high and inclusive growth (for instance, driven
by increased TFP, high levels of educational attainment, and inclusive social policies) that has a
9
Specifically, we use the June 2024 vintage of the World Bank’s Global Economic Prospects, the March 2024 vintage
of the World Bank’s Macro and Poverty Outlook, or the April 2024 vintage of the International Monetary Fund’s
World Economic Outlook.
10
For robustness, estimates of global income inequality based on growth rates using various other historical periods
are reported in Table A2.
11
We assume a linear functional form for the growth incidence curve.
6
minimum impact on the environment. SSP5, on the other hand, considers the same low levels of
within- and between-country inequality but is driven by growth heavily reliant on fossil fuels and
thus provides only limited climate mitigation. SSP3 is the opposite with high levels of between
country inequality and medium-to-high within-country inequality driven by a lack of economic
growth. SSP4 adds a worst-case scenario for within-country inequality to SSP3. Finally, SSP2 lies
between SSP1 and SSP3 with moderate levels of within and between country inequality driven by
moderate growth. We highlight two of the five SSPs in the main text below (SSP1 and SSP4) and
report the results from the rest in the Appendix. SSP1 and SSP4 provide the range of possible
estimates of global inequality in 2050.
3 Results
3.1 The three decades prior to the Covid-19 pandemic
Global income inequality has declined in the past three decades (see Lakner and Milanovic, 2016;
Deaton 2021). Milanovic (2024) argues that this decline was primarily driven by the relatively
faster income growth in populous East Asian countries, particularly China, compared to richer parts
of the world. Figure 1a shows the decline in global inequality since 1990, measured by the Gini
index. 12 The global Gini index fell from 70 points in 1990 to 62 points by 2019, an annualized
decline of -0.42 percent. While the world experienced a fall in income inequality in each of the
three ten-year intervals (1990-00, 2000-10, and 2010-19), the steepest decline took place in the
first decade of the 21st century. Figure 1b reports the global anonymous growth incidence curves
(GICs) for each of the three decades. The graph shows the annualized income growth from the start
year to end year for each of the 100 global percentiles, ranked by income. 13 In all three decades,
we see relatively higher income growth for the bottom 60 percent of the world compared to the top
40 percent. In other words, we see relative income catchup of the poor parts of the global
distribution vis-à-vis the richer parts of the same. This explains the decline in global income
inequality reported in Figure 1a. The annualized global income growth in the middle decade, when
inequality fell the most, was also the highest (3.5 percent) and the annualized global income growth
across percentiles was the lowest in the first decade (1.2 percent), when inequality fell the least.
12
The Gini index is the most widely used measure of inequality that has a range between 0 (most equal) and 100 (most
unequal). While the Gini index is sensitive to the middle part of the distribution, the Mean Log Deviation (Figure 2),
which has a minimum at zero (most equal) and no upper bound, is more sensitive to the bottom tail of the distribution.
13
The GICs shown are anonymous, since the same people are not in each percentile at different points in time.
Following Lakner and Milanovic (2016), we plot the growth in the average income of each percentile group, not the
growth in the value of a particular percentile.
7
Figure 1: Evolution of the global Gini index and Growth Incidence Curves, 1990-2019
Figure 1a: Global Gini index Figure 1b: Global growth incidence curves
Source: Authors’ calculation using Poverty and Inequality Platform (September 2024).
Note: Figure 1a shows the trends in the global Gini index. The Gini index ranges from 0 (most equal) to 100 (most
unequal). Figure 1b shows the annualized growth of each global income percentile for the 1990—2000, 2000—2010,
and 2010—2019 periods. Estimates for the global Gini index is reported in Table A1.
The global picture masks important changes that have happened across countries. To that end,
Figure 2 decomposes total global inequality into the inequality within countries (black bars) and
the inequality between countries (gray bars) using the Mean Log Deviation (MLD). 14 The overall
trends using the MLD (total height of each bar) are almost identical to the trends in the Gini index
reported in Figure 1a. As argued elsewhere (for example, see Lakner & Milanovic, 2016; Mahler
ⓡ al., 2022), the (narrowing of) income differences between countries has driven almost all the
decline in total global inequality over the three decades before the pandemic. 15 Income growth of
populous countries in Asia, in particular China, contributed most to the reduction of inequality
between countries. At the same time, the average (population-weighted) within-country inequality
has not changed. 16 While between-country inequality has declined by 1.5 percent annually, the
within portion declined only 0.01 percent annually. This also means that while the within-country
inequality contributed one-quarter to overall global income inequality in 1990, it contributed one-
third in 2019. Put differently, even with the income catchup that happened in the 30 years before
the Covid-19 pandemic, the differences between countries still account for two-thirds of the
income differences globally. Using the terminology of Milanovic (2015), most of global inequality
is thus explained by a citizenship rent (or penalty) and since most people have little choice over
where they live, this gives rise to a significant global inequality of opportunity.
14
One advantage of the MLD over the Gini index is that the former measure can additively decomposed into inequality
contributed by between-group differences and those contributed by within-group differences.
15
This finding is stable using various inequality measures. For instance, the recently proposed mean ratio deviation,
which is even more bottom-sensitive than the MLD, depicts similar between- and within-country inequality trends
(Kraay et al., 2023).
16
It is important to clarify that the within-country inequality reported above is the population-weighted average within-
country inequality across countries. There have been well-documented cases of rises (or falls) in national inequality
in many countries (for example, see Piketty & Saez, 2003; Atkinson et al., 2011).
8
Figure 2: Decomposition of global income inequality into within and between-country
contributions
Source: Authors’ calculation using Poverty and Inequality Platform (September 2024).
Note: Figure decomposes the total global mean log deviation (MLD) into the MLD between countries (gray bar) and
MLD within countries (black bar) countries. The total MLD is the sum of within and between components in each
year. Estimates for the global MLD and decomposition into within and between-country components is reported in
Table A1.
3.2 The impact of the Covid-19 pandemic
The Covid-19 pandemic had the largest negative effect on global income inequality since at least
1990. Figure 3 shows the year-over-year change in within and between components of total global
MLD. In almost all years besides the Asian financial crisis (1997-98) and the Covid-19 pandemic
(2020), between-country inequality was the primary driver of the reduction in global income
inequality. The average within-country inequality increased in most years in the first two decades
and decreased in many years of the last decade before the pandemic. The within-country MLD
went up from 0.244 in 1990 to 0.270 in 2010 and back down to 0.243 in 2019. In both crises,
between-country inequality accounted for the larger share of the increase in overall global income
inequality. In the Covid-19 pandemic, the within-country inequality played an equalizing role,
decreasing by 0.9 points (3.6 percent) from 2019 to 2020. However, the between-country inequality
increased by 2.4 points (4.9 percent) in the same time frame. The single-year increase in between-
country inequality during the Covid-19 pandemic is more than double the increase that happened
during the entire Asian financial crisis. During the latter crisis, the between-country inequality
increased in two consecutive years, by 0.3 points in 1997 and by 1.1 points in 1998, for a total
increase of 2.1 percent.
9
Figure 3: Impact of the Asian financial crisis and the Covid-19 pandemic on global inequality
Source: Authors’ calculation using Poverty and Inequality Platform (September 2024) and Mahler ⓡ al. (2022).
Note: The figure reports the year-over-year change in within-country and between-country mean log deviation (MLD)
(scaled up by 100). Total change in MLD in each year is the sum of the two. The shaded regions are the Asian financial
crisis (1997-1998) and the Covid-19 pandemic (2020). The figure also reports an alternative estimate for the Covid-
19 pandemic from Mahler ⓡ al. (2022) updated with the September 2024 vintage of data from the Poverty and
Inequality Platform (plotted as 2020*).
These findings are corroborated by MYL, whose estimates are replicated in Figure 3 (see the 2020*
bars). Since Covid-19 delayed survey fieldwork in many countries around the world, MYL
combine all available data that could offer some insights into what happened to within-country
inequality during this period. The authors confirm that the between-country share of inequality was
the primary driver of the increase in inequality in 2020 globally. The between-country inequality
increased by 2.8 points, while the within-country inequality decreased by 1.0 points, meaning that
the overall global MLD increased by 1.8 points. In other words, had all households within a country
experienced the same income loss as the average household in that country, inequality globally
would have increased by 2.8 points instead of 1.8 points (or 56 percent higher than the actual
estimated increase). Inequality within countries played a role in moderating the increase in global
inequality in 2020.
MYL report that the increase in overall global inequality was driven by the larger negative growth
shocks experienced by middle- and low-income countries compared to higher income countries.
Furthermore, they show that countries in regions like Latin America, the Middle East, and Sub-
Saharan Africa had both a large negative growth shock and, at the same time, a decrease in within-
10
country inequality from 2019 to 2020. Large social assistance programs in rich and some upper-
middle-income countries mitigated the welfare shocks to households at the bottom of the income
distribution and thus helped reduce inequality in those countries. However, in general this was not
the case in middle-income countries (World Bank, 2022).
To see this, Figure 4 reports the 2019-2020 change in global within- and between-country
inequality, aggregated by income group. 17 On average, we find no change in within-country
inequality for low- and lower-middle-income countries; i.e., the decline in within-country
inequality seen for the world as a whole is driven by upper-middle-income and high-income
countries. The decline in average incomes for high-income countries during the pandemic reduced
between-country inequality since these countries have average incomes above the global mean.
For the other groups of countries, between-country inequality increased, most strongly for lower-
middle-income countries.
Figure 4: Change in global inequality between 2019 and 2020, by income groups
Source: Authors’ calculation using Poverty and Inequality Platform (September 2024).
Note: The figure reports the change in within-country and between-country mean log deviation (MLD) between 2019
and 2020, aggregated by income group. The total change in MLD in each year is the sum of the two. The changes
reported in this figure add up to total global changes reported in Figure 3. We use the baseline estimate from the
Poverty and Inequality Platform (PIP). Income groups are classified using World Bank’s fiscal year 2021
classification, which uses data from 2019.
17
The decomposition is the same as reported in Figure 3, but instead of aggregating within- and between-country
inequality for all countries, we have aggregated by income group. In other words, the sum of all the changes reported
in Figure 4 add up to what is reported in Figure 3. The change in between-country inequality reported in Figure 4 does
not capture the inequality between (say) high-income countries, which would capture the dispersion of average
incomes from the mean of high-income countries. Conversely, it captures the contribution of high-income countries
to global between-country inequality; i.e., the distance of average incomes of high-income countries from the global
mean.
11
3.3 Potential pathways for global income inequality in the next three decades
Projecting forward to 2050 is obviously subject to substantial uncertainty, so we present a range
of scenarios. Figure 5 presents the business-as-usual and two SSPs scenarios for global income
inequality projected to 2050. For the three scenarios, the figure reports the trends in the Gini index.
Inequality estimates from all the SSPs and various choices of growth rates for the business-as-
usual scenarios and using the MLD are available in Table A2. The business-as-usual scenario is
based on distribution-neutral projections of the income distribution in each country using the
average annual historical growth rates experienced by that country for the 2010-2019 period. In
practice it implies growing each person’s income in line with the average growth rate of the
country, and thus, keeping the within-country inequality fixed to the level in the last household
survey.
However, there is no reason to believe that the future trends in a country’s growth will follow the
historical path of the country. For example, there is no reason to believe that countries that have
grown the fastest in the last decade will maintain the same growth rates over the next three decades.
In addition, despite not having an effect in the past, there is no reason to believe that changes in
within-country inequality cannot have a significant impact on global inequality. To that end, SSP1
and SSP4 incorporate inequality changes both across all countries—by varying average growth
rates—and within countries—by changes in their Gini index. SSP1, the best-case for sustainable
and high growth, is referred to as the “optimistic” scenario, and SSP4, with high between- and
within-country inequality combined with low growth, is referred to as the “pessimistic” scenario.
Between 2020 to 2024, the global Gini index decreased from 62.4 to 61.4, or a 0.4 percent annual
decline. This is roughly the same rate of annual decline that was experienced between 1990 and
2019, but lower than the rate of decline in the decade before the pandemic, 0.6 percent.
Looking ahead, business-as-usual would mean that the global Gini index stays largely unchanged
after 2024, reaching 60.6 points in 2050. This implies an annual decline of only 0.05 percent, or a
rate that is one-tenth of that experienced prior to the pandemic or in the four years since the
pandemic. Under this scenario, the global Gini is projected to continue declining in the first decade
after 2024 (at a rate of 0.14 percent), however, this is followed by an uptick in global income
inequality wiping out almost all the gains of the earlier period. The uptick is in some part explained
by the populous countries with high historical growth rates, like China, moving from a relatively
poorer position globally where they helped reduce global inequality to a relative richer position
where they will add to global inequality. This upward trend in global inequality due to these
changes has been referred to as the Global inequality boomerang by Kanbur et al. (2022).
Milanović (2018) has termed these long-term cyclical changes inequality waves. 18
Whereas global inequality stagnates in the business-as-usual scenario, in the optimistic scenario,
the global Gini index falls to 55.9 points, or a 0.4 percent annual reduction starting in 2024. The
annual reduction predicted for the optimistic scenario is in line with the inequality reduction seen
in the three decades between 1990 and 2019. The expected decrease in inequality for this scenario
18
In all the historical growth scenarios we considered (see Table A2), the boomerang is much more muted than that
reported elsewhere.
12
comes due to the higher growth rates, especially for poorer countries, combined with lowering of
within-country inequality. This scenario not only is beneficial for income inequality globally but
is also the preferred outcome for lowering climate impacts. While rosy to some extent, a similar
reduction in global inequality is also possible with high growth for poorer countries that is not
environmentally sustainable (see SSP5 scenario in Table A1). Needless to say, the former is
preferred to the latter. The outcome in the optimistic scenario would mean that the level of global
inequality will be back close to the level seen in the 1800s, a full Kuznets-type U-turn
(Bourguignon & Morrisson, 2002; Milanovic, 2024).
The pessimistic scenario predicts a slight uptick in global inequality between 2024 and 2050.
Under this scenario of worsening within-country inequality and low growth, the global Gini index
is expected to reach 62.5, or a 0.07 percent annual increase starting in 2024. Even with this
scenario, global income inequality in 2050 is expected to be lower than the levels of inequality
around the global financial crisis. All other scenarios are within the range encompassed by the
optimistic and pessimistic scenarios (see Table A1).
Figure 5: Potential pathways for global inequality 2025-2050
Source: Authors’ calculations using Poverty and Inequality Platform (September 2024), Crespo Cuaresma, (2017),
Mahler ⓡ al. (2022), Rao et al (2019).
Note: Business-as-usual scenario is based on distribution-neutral projection of country income distributions using
average annual historical growth rates faced by each country in the period 2010-2019; Optimistic (SSP1) and
pessimistic (SSP4) scenarios involve changes in both within-country inequality and across-country average growth
rates. See also Figure 1.
The decline in global inequality in the period before the pandemic was driven by cross-country
income growth differences, with poorer parts of the world growing much faster than the richer
parts. In Figure 6, we look for such trends in the next 30 years. Table 1 summarizes the forecasted
annual average growth rates by various income groups using the 2024 fiscal year classifications.
13
Figure 6. Annualized growth by country’s income level
A. 1990 – 2019 (Historical) B. 2024-2050 (business-as-usual)
C. 2024-2050 (optimistic) D. 2024-2050 (Pessimistic)
Source: Authors’ calculations based on Poverty and Inequality Platform (September 2024), Crespo Cuaresma, (2017),
Rao et al. (2019).
Note: This figure reports the annualized (historic and forecasted) growth rates (percent) for each country ranked by
the initial level of average income from household surveys. Panel A reports historical data for the period 1990-2019,
Panel B reports forecasted data for the period 2024-2050 under the business-as-usual scenario, and Panels C and D
report the scenarios using the Shared Socioeconomic Pathways 1 and 4. Business-as-usual scenario uses the average
annual historical cross-country growth rates for the period 2010-2019. The figure also reports linearly fitted lines
weighted by countries’ population (dashed) in the initial year and not (solid). For readability, only countries with
growth rates between -4 percent and 8 percent are shown in the charts, while the fitted lines use all countries.
For comparison, panel A of Figure 6 shows these differences using annualized growth rates from
1990 to 2019 and ranking countries by their initial level of income (1990 in the historic case).
While the poorest in 1990 fared much better than the richer nations in general, the former did even
better than the latter when growth is adjusted for population size (dashed line). Hence, both the
cross-country average income differences (Concept 1 inequality) and, more so, the population
adjusted cross-country inequality (Concept 2 inequality) decreased over the 30 years before the
pandemic. China’s spectacular growth performance is obviously a big part of this story.
Looking ahead, growth in the business-as-usual scenario is much more muted for the poorest
countries (Figure 6, Panel B). Note that the baseline income level (the horizontal axis) in the
forecasted figures (Panels B-D) is the income level in 2024, where most of the populous poorer
countries in 1990 have now graduated to middle-income status. In the business-as-usual scenario,
as reported in Panel B of Figure 6, we see that the poorest countries in the world have the least
14
growth, while the middle segment of the world has the highest growth. The fitted line is slightly
upward sloping when unweighted and slightly downward sloping when countries are weighted by
their population. In both cases, it is a departure from the historical period, when there was a
negative slope indicating faster growth for (initially) poor countries. The similar growth rates for
the lower and upper middle-income countries (covering three-quarters of the world’s population)
signal that the global distribution will grow somewhat uniformly. The higher growth rates for the
rich countries and UMICs eventually lead to the uptick in global inequality reported in Figure 5.
Table 1: Forecasted average annual growth rates
Average growth rates (%) Average Gini changes
Share of
Business-as-
Income group latest global pop, Optimistic Pessimistic Optimistic Pessimistic
usual
%
Low income 8.9 -0.1 4.8 2.5 -0.4 0.0
Lower middle income 40.1 2.2 4.0 2.6 -0.4 0.1
Upper middle income 35.4 2.1 3.0 2.5 -0.4 0.1
High income 15.7 1.4 1.8 1.7 -0.3 0.3
Source: Authors’ calculations based on Poverty and Inequality Platform (September 2024), (Castaneda Aguilar, 2024;
Crespo Cuaresma, 2017; Rao et al., 2019; World Bank, 2024a).
Note: This table reports the share of global population, the average annual cross-country forecasted growth rates, and
the average Gini changes for the period 2024-2050 under various Shared Socioeconomic Pathway (SSP) scenarios.
Business-as-usual scenario reports the average annual historical cross-country growth rates for the period 2010-2019.
Countries are classified using the latest year of the World Bank’s income classification (fiscal year 2024). The
population shares are reported for 2022, which is the year of the data underlying the fiscal year 2024 income
classification. Growth rates and Gini changes are not weighted by population.
The business-as-usual pattern of growth rates by income group somewhat mirrors the pessimistic
scenario where the middle part of the distribution grows at a similar rate. However, in the
pessimistic scenario, LICs grow much faster, outperforming HICs and implying a catchup on
average of the poorer parts of the global income distribution (see also Panel D in Figure 6). Note
that the SSP-based scenarios also include within-country distributional changes and an increase in
within-country inequality, especially in the richer countries under the pessimistic scenario (Table
1), which could work in an un-equalizing way. The optimistic scenario, on the other hand, clearly
has high pro-poor growth as well as a decline in within-country inequality, which together point
to the considerable decline in global inequality (Figure 5). The trend of the optimistic scenario in
Panel C mirrors the trend from the past three decades (Panel A) and corroborates the finding above
of a similar rate of reduction of global income inequality. While the Gini index decreased by 11%
from 1990 to 2019, under the optimistic scenario, it is expected to improve by 9% from 2024 to
2050.
It is important to consider that the optimistic scenario requires low-income countries to grow at a
rate that is 3 to 5 times higher than the rates experienced by countries in that position in the past
decade. While many middle-income countries that were poor in the past have managed such
income growth rates, this does mean that the growth rates expected for poorer nations under these
scenarios are very optimistic.
15
4 Discussion and Conclusion
Economic growth across countries played a large role in the reduction of global income inequality
from 1990 to 2019. Poorer populous countries, and especially China, were the drivers of this
reduction but there was broad catchup of the poorer parts of the global distribution to the richer
parts. Relatively poorer countries underwent economic progress, which means that they had in
2019 moved closer to the high-income countries compared to where they started in 1990. As Patel
et al (2021) point out, this does not just apply to China, India or selected Asian countries, but
developing countries on average. In some cases, economic growth combined with prudent fiscal
policies has helped reduce income inequality within a country and also helped the country catch
up to the rest of the world. For example, Brazil has more than doubled the daily per capita
household incomes from close to $10 in 1990 to over $21 by 2019 compared to a 56% increase in
the global per capita mean income. In addition, this growth has been progressive. Between 2004
and 2014, the bottom 40 percent of Brazilians grew their income by 6.8 percent compared to 4.5
percent for the average Brazilian (World Bank Group, 2016). This meant that inequality in Brazil
came down from a high of over 60 Gini points in 1990 to 53.5 Gini points in 2019.
Similar to Brazil, Sierra Leone increased average incomes 56 percent between 2003 and 2018,
compared to 34% for the global average. In this period, not only did it move closer to the global
average income, but also the share of income captured by the bottom 40 percent of the population
grew from 17 to 20 percent (a 15 percent increase). In other words, Sierra Leone both decreased
inequality within the country (the Gini index decreased from 40 to 36) and at the same time reduced
the income gap globally.
The Covid-19 pandemic presented different challenges globally. While it was primarily a health
crisis, it also impacted every aspect of life, including economic life. The pandemic halted survey
fieldwork in many countries around the world, limiting the available evidence forever. However,
all the available evidence suggests that global income inequality increased over this period. This
increase was driven by the widening gap in average incomes between countries rather than the gap
in incomes within countries. In fact, average within-country inequality fell in 2020. For instance,
despite the large shock for the overall economy, household incomes in poorer parts of Brazil grew
in 2020. This is in no small part due to the extraordinarily large social protection measures
implemented in the country in 2020. This meant that there was economic support for the most in-
need population during the pandemic and it also meant that within-country inequality decreased in
2020. Unfortunately, the opposite was true in 2021 with the pullback of the emergency protection
measures that were in place during the pandemic. The country saw an increase in within-country
inequality from 48.9 Gini points in 2020 to 52.9 points in 2021, closer to the 2019 level.
High-income countries and some upper-middle-income countries were able to withstand the
economic shock of the pandemic due to social protection measure that were in place during the
pandemic. This meant that the welfare shocks, especially for the poorer households, were
mitigated, at least for a period of time. Poorer countries largely do not have the fiscal space to
carry out these types of social protection measures. While the health impact of the pandemic was
more limited in poorer countries, this was in large part the result of a demographic advantage with
16
a younger population and the Covid-19 virus having a larger impact among the elderly. However,
social protection measures need to be strengthened in these countries if they are to avoid significant
welfare shocks due to any future crises (Abay et al., 2023).
Looking forward, we broadly find three potential trends for global inequality. First, if countries
continue to grow in the manner they did over the last three decades before the pandemic, global
income inequality will likely stagnate initially but trend slightly upwards by 2050. By how much
and when the upturn will start depend on the relative growth and population differences between
large populations in China, India, Sub-Saharan Africa, and the group of rich mostly Western
countries. Second, global inequality could start increasing more significantly if there was
stagnation of country-level growth in the poorer parts of the world and if within-country inequality
increases. Finally, there is also a more positive scenario which would decrease global inequality.
However, this decrease will require very high income growth in the poorer parts of the world
compared to richer parts. The countries that have experienced per capita growth rates under 1
percent most recently would be expected to grow at a rate that is 5 times higher. While a near 5
percent growth rate in poor countries seems unreasonable at first, it is important to consider that
middle-income countries that were poor in the past have managed to sustain even higher levels of
income growth. These lower levels of global income inequality can be achieved either using
unsustainable climate policies or incorporating technological change such that the climate impacts
are reduced at the same time. The latter outcome is beneficial to all.
17
5 References
Abay, K. A., Yonzan, N., Kurdi, S., & Tafere, K. (2023). Revisiting Poverty Trends and the Role
of Social Protection Systems in Africa during the COVID-19 Pandemic. Journal of African
Economies, 32(Supplement_2), ii44–ii68. https://doi.org/10.1093/jae/ejac041
Aguiar, M., & Bils, M. (2015). Has Consumption Inequality Mirrored Income Inequality?
American Economic Review, 105(9), 2725–2756. https://doi.org/10.1257/aer.20120599
Alvaredo, F., Bourguignon, F., Ferreira, F. H. G., & Lustig, N. (2023). Seventy-five Years of
Measuring Income Inequality in Latin America. Inter-American Development Bank.
https://doi.org/10.18235/0005211
Anand, S., & Segal, P. (2008). What Do We Know about Global Income Inequality? Journal of
Economic Literature, 46(1), 57–94. https://doi.org/10.1257/jel.46.1.57
Atkinson, A. B., & Bourguignon, F. (2000). Introduction: Income distribution and economics. In
Handbook of Income Distribution (Vol. 1, pp. 1–58). Elsevier. https://doi.org/10.1016/S1574-
0056(00)80003-2
Atkinson, A. B., & Piketty, T. (Eds.). (2007). Top incomes over the twentieth century: A contrast
between continental European and English-speaking countries. Oxford University Press.
Atkinson, A. B., Piketty, T., & Saez, E. (2011). Top Incomes in the Long Run of History. Journal
of Economic Literature, 49(1), 3–71. https://doi.org/10.1257/jel.49.1.3
Bolt, J., & van Zanden, J. L. (2020). Maddison style estimates of the evolution of the world
economy. A new 2020 update [Dataset].
Bourguignon, F., & Morrisson, C. (2002). Inequality Among World Citizens: 1820–1992.
American Economic Review, 92(4), 727–744. https://doi.org/10.1257/00028280260344443
Burkhauser, R. V., Hahn, M. H., & Wilkins, R. (2015). Measuring top incomes using tax record
data: A cautionary tale from Australia. The Journal of Economic Inequality, 13(2), 181–205.
https://doi.org/10.1007/s10888-014-9281-z
Castaneda Aguilar, R. A. (2024). pip: Stata Module to Access World Bank’s Global Poverty and
Inequality Data (Version 0.10.9.9001) [STATA]. https://worldbank.github.io/pip/
Crespo Cuaresma, J. (2017). Income projections for climate change research: A framework based
on human capital dynamics. Global Environmental Change, 42, 226–236.
https://doi.org/10.1016/j.gloenvcha.2015.02.012
Deaton, A. (2005). Measuring Poverty in a Growing World (Or Measuring Growth in a Poor
World). The Review of Economics and Statistics, 87(1), 1–19.
Deaton, A. (2021). COVID-19 and Global Income Inequality. LSE Public Policy Review, 1(4), 1.
https://doi.org/10.31389/lseppr.26
18
Flachaire, E., Lustig, N., & Vigorito, A. (2023). Underreporting of Top Incomes and Inequality: A
Comparison of Correction Methods using Simulations and Linked Survey and Tax Data. Review
of Income and Wealth, 69(4), 1033–1059. https://doi.org/10.1111/roiw.12618
Haddad, C. N., Mahler, D. G., Diaz-Bonilla, C., Hill, R., Lakner, C., & Lara Ibarra, G. (2024). The
World Bank’s New Inequality Indicator: The Number of Countries with High Inequality (English).
WPS 10796.
http://documents.worldbank.org/curated/en/099549506102441825/IDU1bd155bac16d78143af18
8331f87564a9d6c8
Jenkins, S. P. (2015). World income inequality databases: An assessment of WIID and SWIID. The
Journal of Economic Inequality, 13(4), 629–671. https://doi.org/10.1007/s10888-015-9305-3
Jenkins, S. P. (2017). Pareto Models, Top Incomes and Recent Trends in UK Income Inequality.
Economica, 84(334), 261–289. https://doi.org/10.1111/ecca.12217
Kanbur, R., Ortiz-Juarez, E., & Sumner, A. (2022). The Global Inequality Boomerang. SSRN
Electronic Journal. https://doi.org/10.2139/ssrn.4114720
Kraay, A., Lakner, C., Özler, B., Decerf, B., Jolliffe, D., Sterck, O., & Yonzan, N. (2023). A New
Distribution Sensitive Index for Measuring Welfare, Poverty, and Inequality. Washington, DC:
World Bank. https://doi.org/10.1596/1813-9450-10470
Lakner, C., Mahler, D. G., Negre, M., & Prydz, E. B. (2022). How much does reducing inequality
matter for global poverty? The Journal of Economic Inequality, 20(3), 559–585.
https://doi.org/10.1007/s10888-021-09510-w
Lakner, C., & Milanovic, B. (2016). Global Income Distribution: From the Fall of the Berlin Wall
to the Great Recession. The World Bank Economic Review, 30(2), 203–232.
https://doi.org/10.1093/wber/lhv039
Mahler, D. G., Castaneda Aguilar, R. A., & Newhouse, D. (2021). Nowcasting Global Poverty.
The World Bank. https://doi.org/10.1596/1813-9450-9860
Mahler, D. G., Yonzan, N., & Lakner, C. (2022). The Impact of COVID-19 on Global Inequality
and Poverty. The World Bank. https://doi.org/10.1596/1813-9450-10198
Milanovic, B. (2005). Worlds Apart. Princeton University Press; JSTOR.
http://www.jstor.org/stable/j.ctt7t4v9
Milanovic, B. (2015). Global Inequality of Opportunity: How Much of Our Income Is Determined
by Where We Live? Review of Economics and Statistics, 97(2), 452–460.
https://doi.org/10.1162/REST_a_00432
Milanović, B. (2018). Global inequality: A new approach for the age of globalization (First
Harvard University Press paperback edition). The Belknap Press of Harvard University Press.
Milanovic, B. (2024). The three eras of global inequality, 1820–2020 with the focus on the past
thirty years. World Development, 177, 106516. https://doi.org/10.1016/j.worlddev.2023.106516
19
O’Neill, B. C., Kriegler, E., Ebi, K. L., Kemp-Benedict, E., Riahi, K., Rothman, D. S., Van Ruijven,
B. J., Van Vuuren, D. P., Birkmann, J., Kok, K., Levy, M., & Solecki, W. (2017). The roads ahead:
Narratives for shared socioeconomic pathways describing world futures in the 21st century. Global
Environmental Change, 42, 169–180. https://doi.org/10.1016/j.gloenvcha.2015.01.004
Patel, D., Sandefur, J., & Subramanian, A. (2021). The new era of unconditional convergence.
Journal of Development Economics, 152, 102687. https://doi.org/10.1016/j.jdeveco.2021.102687
Piketty, T., & Saez, E. (2003). Income Inequality in the United States, 1913-1998. The Quarterly
Journal of Economics, 118(1), 1–41. https://doi.org/10.1162/00335530360535135
Piketty, T., Yang, L., & Zucman, G. (2019). Capital Accumulation, Private Property, and Rising
Inequality in China, 1978–2015. American Economic Review, 109(7), 2469–2496.
https://doi.org/10.1257/aer.20170973
Rao, N. D., Sauer, P., Gidden, M., & Riahi, K. (2019). Income inequality projections for the Shared
Socioeconomic Pathways (SSPs). Futures, 105, 27–39.
https://doi.org/10.1016/j.futures.2018.07.001
Roser, M., & Cuaresma, J. C. (2016). Why is Income Inequality Increasing in the Developed
World? Review of Income and Wealth, 62(1), 1–27. https://doi.org/10.1111/roiw.12153
Van Zanden, J. L., Baten, J., Foldvari, P., & Van Leeuwen, B. (2014). The Changing Shape of
Global Inequality 1820–2000; Exploring a New Dataset. Review of Income and Wealth, 60(2),
279–297. https://doi.org/10.1111/roiw.12014
Welch, I. (2024). The IPCC Shared Socioeconomic Pathways (SSPs): Explained, Critiqued,
Replaced. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.4681042
World Bank. (2017). Monitoring Global Poverty: Report of the Commission on Global Poverty.
Washington, DC: World Bank. https://doi.org/10.1596/978-1-4648-0961-3
World Bank. (2022). Poverty and Shared Prosperity 2022: Correcting Course. The World Bank.
https://doi.org/10.1596/978-1-4648-1893-6
World Bank. (2024a). Poverty and Inequality Platform (Version 20240627_2017_01_02_PROD)
[Dataset]. https://pip.worldbank.org/
World Bank. (2024b). Poverty, Prosperity, and Planet Report 2024: Pathways Out of the
Polycrisis. https://www.worldbank.org/en/publication/poverty-prosperity-and-planet
World Bank Group. (2016). Poverty and Shared Prosperity 2016: Taking on Inequality.
Washington, DC: World Bank. https://doi.org/10.1596/978-1-4648-0958-3
Yonzan, N., Milanovic, B., Morelli, S., & Gornick, J. (2022). Drawing a Line: Comparing the
Estimation of Top Incomes between Tax Data and Household Survey Data. The Journal of
Economic Inequality, 20(1), 67–95. https://doi.org/10.1007/s10888-021-09515-5
20
6 Appendix
Table A1: Estimates of global inequality
MLD MLD
Year Gini MLD
Within Between
1990 69.780 0.997 0.244 0.753
1991 69.695 0.990 0.243 0.747
1992 69.631 0.982 0.243 0.739
1993 69.822 0.982 0.251 0.731
1994 69.589 0.969 0.253 0.716
1995 69.319 0.959 0.258 0.700
1996 68.902 0.942 0.261 0.681
1997 68.946 0.944 0.260 0.684
1998 69.304 0.958 0.263 0.695
1999 69.455 0.959 0.266 0.693
2000 69.251 0.954 0.266 0.689
2001 69.024 0.945 0.268 0.677
2002 68.614 0.929 0.269 0.661
2003 68.302 0.917 0.268 0.648
2004 67.699 0.896 0.266 0.629
2005 67.301 0.881 0.267 0.614
2006 67.044 0.874 0.267 0.607
2007 66.824 0.868 0.270 0.598
2008 66.292 0.851 0.270 0.581
2009 65.809 0.837 0.269 0.568
2010 65.138 0.815 0.270 0.545
2011 64.415 0.790 0.265 0.525
2012 63.831 0.773 0.265 0.507
2013 63.004 0.749 0.256 0.493
2014 62.672 0.741 0.254 0.487
2015 62.530 0.738 0.251 0.487
2016 62.542 0.740 0.249 0.491
2017 62.153 0.730 0.251 0.479
2018 61.756 0.719 0.247 0.472
2019 61.831 0.723 0.243 0.480
2020 62.351 0.738 0.234 0.504
2021 61.967 0.734 0.231 0.502
2022 61.792 0.728 0.233 0.496
2023 61.675 0.726 0.233 0.493
2024 61.412 0.720 0.233 0.487
Source: Authors’ calculations based on Poverty and Inequality Platform (September 2024).
21
Table A2: Estimates of global Gini index for various scenarios
Business Business Business Business
as usual as usual as usual as usual
Year SSP 1 SSP 2 SSP 3 SSP 4 SSP 5
(2010- (1990- (1990-2019 (1990-2019
2019) 2019) FY95) FY24)
2025 61.34 61.34 61.34 61.34 61.26 61.32 61.37 61.42 61.24
2026 61.27 61.27 61.27 61.27 61.10 61.22 61.33 61.43 61.07
2027 61.17 61.17 61.17 61.17 60.94 61.13 61.29 61.44 60.89
2028 61.08 61.07 61.07 61.07 60.78 61.04 61.25 61.45 60.72
2029 60.98 60.97 60.97 60.97 60.62 60.94 61.22 61.47 60.54
2030 60.88 60.88 60.97 61.09 60.46 60.85 61.18 61.48 60.36
2031 60.78 60.80 60.96 61.22 60.28 60.76 61.16 61.54 60.16
2032 60.70 60.73 60.96 61.34 60.09 60.66 61.15 61.61 59.96
2033 60.62 60.68 60.96 61.47 59.91 60.56 61.12 61.67 59.77
2034 60.55 60.64 60.95 61.59 59.72 60.46 61.10 61.73 59.59
2035 60.49 60.62 60.95 61.72 59.53 60.36 61.07 61.80 59.40
2036 60.43 60.60 60.95 61.85 59.32 60.24 61.06 61.88 59.18
2037 60.39 60.60 60.94 61.97 59.10 60.12 61.04 61.95 58.96
2038 60.35 60.62 60.94 62.10 58.88 60.00 61.03 62.02 58.74
2039 60.32 60.65 60.93 62.23 58.65 59.89 61.02 62.10 58.52
2040 60.30 60.69 60.93 62.35 58.43 59.77 61.01 62.17 58.29
2041 60.28 60.74 60.92 62.48 58.19 59.63 61.00 62.22 58.05
2042 60.28 60.81 60.92 62.61 57.94 59.50 60.99 62.27 57.82
2043 60.28 60.90 60.91 62.73 57.70 59.36 60.99 62.32 57.58
2044 60.29 60.99 60.91 62.86 57.45 59.22 60.98 62.36 57.33
2045 60.31 61.11 60.90 62.99 57.19 59.08 60.97 62.41 57.09
2046 60.34 61.23 60.90 63.12 56.95 58.94 60.97 62.43 56.86
2047 60.38 61.38 60.89 63.25 56.70 58.79 60.97 62.46 56.62
2048 60.43 61.54 60.89 63.38 56.44 58.65 60.96 62.48 56.39
2049 60.48 61.71 60.89 63.52 56.19 58.50 60.96 62.51 56.15
2050 60.55 61.89 60.89 63.65 55.93 58.35 60.96 62.53 55.91
Source: Authors’ calculations based on Poverty and Inequality Platform (September 2024), Rao et al (2019), Crespo Cuaresma
(2017).
Note: This table reports the global Gini index for the period 2020-2050 under various business-as-usual and Shared Socioeconomic
Pathway (SSP) scenarios. Gini for the years 2020-2023 is the same across all scenarios. The business-as-usual (2010-2019) uses
average annual historical country growth rates for the years 2010-2019 (reported in the paper); the business-as-usual scenario
(1990-2019) uses average annual historical country growth rates for the years 1990-2019 ; business-as-usual scenario (FY95) uses
average annual historical growth rates for the years 1990-2019 associated with the income group the country belonged to as defined
in the World Bank’s FY1995 income group classification; and, business-as-usual scenario (FY24) uses average annual historical
growth rates for the years 1990-2019 associated with the income group the country belonged to as defined in the World Bank’s
FY2024 income group classification. SSP1 (optimistic scenario) and SSP4 (pessimistic scenario) are included in the main text.
22
Table A3: Other forecasted average annual growth rates
Business-as-usual growth rates (%) Average growth rates (%)
Share of From 1990 From 1990 to From 1990 to
Income group latest global to 2019 2019 (Income 2019 (Income SSP 2 SSP 3 SSP 5
pop, % (Country) group FY24) group FY95)
Low income 8.9 0.8 1.1 2.5 3.8 2.4 5.6
Lower middle income 40.1 2.2 2.2 2.4 3.2 1.9 4.7
Upper middle income 35.4 2.8 2.7 2.2 2.5 1.5 3.7
High income 15.7 1.7 1.9 1.5 1.5 0.9 2.3
Source: Authors’ calculations based on Crespo Cuaresma (2017) and PIP.
Note: This table reports the share of global population and the average annual cross-country forecasted growth rates for the period
2024-2050 under Shared Socioeconomic Pathway (SSP) scenarios 2, 3 and 5. Business-as-usual scenarios reports the average
annual historical cross-country growth rates for the period 1990-2019. Countries are classified using the World Bank’s income
classification for fiscal year 1995 (FY95) and fiscal year 2024 (FY24). The population shares are reported for 2022, which is
the year of the data underlying the fiscal year 2024 income classification.
Table A4: Other forecasted average Gini changes
Average Gini changes
Share of
Income group latest global pop, SSP 2 SSP 3 SSP 5
%
Low income 8.9 -0.2 0.1 -0.5
Lower middle income 40.1 -0.2 0.1 -0.5
Upper middle income 35.4 -0.1 0.1 -0.5
High income 15.7 0.0 0.3 -0.3
Source: Authors’ calculations based on Crespo Cuaresma (2017) and PIP.
Note: This table reports the share of global population and the average annual cross-country forecasted Gini changes for the period
2024-2050 under Shared Socioeconomic Pathway (SSP) scenarios 2, 3 and 5. Countries are classified using the latest year of the
World Bank’s income classification (FY24).
23