HUMAN WELFARE

Poverty
Carlo Azzarri

WHaT are THeSe maPS TellInG uS? 
Almost half of the population of Africa south of the Sahara 
(SSA) lives in extreme poverty, on less than $1�25 per cap-
ita per day�1 Map 1 shows the distribution of the poor and 
highlights areas where over 80 percent of the population 
is extremely poor (for example, parts of Liberia, Nigeria, 
Tanzania, and Zambia)� Map 2 shows the density of extremely 
poor across the continent, highlighting regions that are home 
to more than 100 extremely poor people per square kilome-
ter� Moderate poverty is defined as living on a daily per cap-
ita expenditure between $1�25 and $2�00� Map 3 shows the 
distribution of poor using the $2�00 per day threshold, thus 
including both the moderately and extremely poor� This map 
shows a more even distribution of poor across Africa and 
consistently higher shares of the total population� Map 4 rein-
forces that the most densely populated poor areas are con-
centrated along the coast of western Africa, in much of 
Nigeria, in Malawi, in Ethiopia, and in the countries bordering 
or near Lake Victoria� Figure 1 shows that extreme poverty 
is also highly correlated with certain agroecological zones 
(p� 34)� For example, poverty levels are highest in the warm 
semiarid and subhumid tropical areas immediately south 
of the Sahara and in the tropical warm humid forests of the 
Democratic Republic of the Congo� And, overall, poverty lev-
els are lower in the subtropical zones of southern Africa (for 
example, Namibia and South Africa)�

WHY IS THIS ImPOrTanT?
Poverty prevalence (Maps 1 and 3) is crucial information 
for policymakers and international donors who are setting 
priorities for intervention and investment� Poverty density 
complements prevalence by showing the number of poor 
people per square kilometer (Maps 2 and 4)� These maps 
together answer two important questions: Where is poverty 
a serious problem? Where might investments have the great-
est impact on the highest number of people? Combining 
insights on both prevalence and density allows policymakers 
to more effectively target interventions to reach the great-
est number of the poorest people� Once target populations 
are identified, information on the dominant types of exist-
ing livelihoods and agriculture-related opportunities can be 
helpful in formulating appropriate interventions�

WHaT aBOuT THe unDerlYInG DaTa?
Subnational poverty rates were extracted from 24 nationally 
representative household surveys conducted in various years�  
For countries without survey data, national average poverty 
prevalence extracted from PovcalNet (World Bank 2012) for 
the closest year to 2005 was uniformly applied to the entire 
country� As such, subnational poverty rate distributions reflect 
the relative ranking in the actual survey year, although all val-
ues are expressed in terms of 2005 average purchasing power 
parity exchange rates� Poverty ratios are therefore compara-
ble across countries� Not all current data points refer to 2005, 
with a maximum variance of plus or minus two years for a 
limited number of countries (HarvestChoice 2012)�

WHere Can I learn mOre?

Poverty analysis at the World Bank:  
www�worldbank�org/en/topic/poverty

“Poverty Comparisons over Time and Across Countries in 
Africa�” Sahn and Stifel 2000� 

“Where Will the World’s Poor Live?: An Update on Global 
Poverty and the New Bottom Billion�” Sumner 2012�

FIGURE 1 Poverty headcount ratio by agroecological zone

Warm arid
Cool subhumid

Cool semiarid
Warm semiarid

Warm subhumid
Cool arid

Warm humid
Cool humid

Warm semiarid
Warm subhumid

Warm humid
Cool subhumid

Cool humid
Cool semiarid

Warm arid
Cool arid

Tr
op

ic
Su

bt
ro

pi
c

0 20

Poverty headcount ratio (%)
40 60

Data source: HarvestChoice 2012 and HarvestChoice 2011.
Note: Poverty headcount ratio=the percentage of a population living in house-
holds where consumption or income per person is below the poverty line.

1 The $1.25 and $2.00 poverty lines are the level of total household per capita consumption expenditure (a synthetic indicator of household welfare) expressed 
in terms of 2005 average purchasing power parity exchange rates.

76

http://www.worldbank.org/en/topic/poverty


Data source (all maps): HarvestChoice 2012.

Note: All values are expressed in terms of average 2005 
purchasing power parity rates.

HUMAN WELFARE ATLAS
OF AFRICAN AGRICULTURE RESEARCH & DEVELOPMENT

1
10
20
30
40
50
60
70
80
90
100
No data

Outside focus area

Prevalence (percent)

1
10
20
30
40
50
60
70
80
90
100
No data

Outside focus area

Prevalence (percent)

0
5
10
25
50
100
500 ≤

No data

Number of poor/km²

Outside focus area

0
5
10
25
50
100
500 ≤

No data

Number of poor/km²

Outside focus area

INTERNATIONAL 
FOOD POLICY 
RESEARCH 
INSTITUTE

IFPRI

MAP 1 Share of population living at ≤ $1.25/day  
(extremely poor) 

MAP 2 Poverty density at ≤ $1.25/day (extremely poor)

MAP 3 Share of population living at ≤ $2.00/day (includes 
moderately and extremely poor)

MAP 4 Poverty density at ≤ $2.00/day 
(includes moderately and extremely poor)

77