Rich country, poor country

Africa’s mean GDP per capita had a slowly rising long-run trend, consisting of about 20 years of virtual stagnation with an inflexion point upward in the mid-1990s (Figure 8.7).11 Income per person in the average African economy declined towards the middle of the 1990s and then recovered. Weighting by GDP (Figure 8.7b) gives a U-shaped pattern of GDP per capita, reaching a minimum in the mid-1990s.

Africa’s largest economies, measured in terms of GDP, experienced some of its greatest income declines between 1975 and 1994. By 2005, income weighted GDP per capita had not yet recovered to the levels observed in the mid-1970s.

For most individual African economies, GDP per capita registered only modest increases between 1975 and 2005, and many countries - such as the Democratic Republic of Congo, Cote d’Ivoire and Zimbabwe - had declining per capita incomes

Gross domestic product per capita, 1975-2005

Figure 8.7 Gross domestic product per capita, 1975-2005: (a) unweighted data; (b) weighted data

Source: Authors’ computations.

Density of gross domestic product per capita across countries

Figure 8.8 Density of gross domestic product per capita across countries: (a) 1975 compared with 2005; (b) 1975 compared with 1985; (c) 1985 compared with 1995; (d) 1995 compared with 2005

Source: Authors’ computations.

over the period. The SD of income per capita was generally low, and the CV of many countries is close to zero. Most of the variation (69 per cent) in Africa’s average income per person occurs between countries (Table 8.2) rather than within them.

Figure 8.8 shows comparative kernel density plots of GDP per capita for at 10-year intervals between 1975 and 2005. The kernel plot is helpful in identifying shifting patterns in the distribution of country-level incomes per capita, such as the formation of identifiable groups or ‘clubs’ (Quah 1993a, 1993b). The slight movement towards the right of the GDP per capita plot of 2005 compared with that of 1975 (Figure 8.8a) reflects the region’s slow economic growth. Most of that movement took place between 1995 and 2005, when we observe a noticeable slide to the right, reflecting increasing incomes throughout Africa (Figure 8.8d). Since the mid-1990s, the variance of income per capita also appears to have declined.

Since 1975 there has been a pronounced bimodality in the distribution of income among countries in Africa. The twin peaks observed in each of the panels of Figure 8.8 define two groups of countries, rich and poor, which are relatively stable over the 30-year period. The most significant shift towards polarization of the country-level distribution occurred between 1985 and 1995 (Figure 8.8c), a period when many countries were devastated by conflicts. Polarization was reduced somewhat between 1995 and 2005.

The second peak virtually disappears in each 10-year period when we remove Botswana, Cape Verde, Gabon, Mauritius, Namibia, the Seychelles and South

Africa from the data. Thus these seven countries form a stable, rich country club.12 With the exception of South Africa, the members of the club are small economies, and they have relatively little in common with one another. Botswana, Cape Verde and Mauritius had high growth - even by global standards - and low- growth volatility. The Seychelles had more moderate growth with high volatility. Namibia and South Africa barely grew but were highly erratic growers, and Gabon declined with high volatility. Cape Verde, Mauritius and the Seychelles are islands; Botswana is landlocked and resource rich.

Table 8.3 shows, for 43 countries, the ratio of their GDP per capita to that of South Africa, the richest African economy in 1975.

These data show little upward income mobility among African countries. Despite South Africa’s long period of slow growth, only nine countries - Botswana, Cape Verde, Equatorial Guinea, Lesotho, Mauritius, the Seychelles, Sudan, Swaziland and Uganda - had an increase in their GDP per capita of at least 5 percentage points relative to South Africa. Of these, only Sudan and Uganda are not members of the rich country club identified above. Botswana, Mauritius and the Seychelles

Table 8.3 GDP per capita relative to South Africa

Country

1975 (or

earliest

year)

2005 (or most recent year)

Country

1975 (or

earliest

year)

2005 (or most recent year)

Angola

0.19

0.21

Lesotho

0.12

0.30

Benin

0.09

0.10

Madagascar

0.13

0.08

Botswana

0.19

1.12

Malawi

0.06

0.06

Burkina Faso

0.08

0.11

Mali

0.08

0.09

Burundi

0.08

0.06

Mauritania

0.20

0.20

Cameroon

0.18

0.21

Mauritius

0.40

1.14

Cape Verde

0.23

0.52

Mozambique

0.07

0.11

Central African Republic

0.17

0.11

Namibia

0.65

0.68

Chad

0.10

0.13

Niger

0.10

0.07

Comoros

0.19

0.18

Nigeria

0.10

0.10

Congo, Dem. Rep.

0.23

0.06

Rwanda

0.09

0.11

Congo, Rep.

0.10

0.11

Senegal

0.15

0.16

Cote d’Ivoire

0.25

0.15

Seychelles

0.76

1.45

Equatorial Guinea

0.13

0.73

Sierra Leone

0.10

0.07

Eritrea

0.09

0.10

Sudan

0.12

0.19

Ethiopia

0.09

0.09

Swaziland

0.32

0.43

Gabon

0.97

0.63

Tanzania

0.05

0.07

Gambia, The

0.16

0.17

Togo

0.18

0.14

Ghana

0.20

0.22

Uganda

0.08

0.13

Guinea

0.19

0.21

Zambia

0.14

0.09

Guinea-Bissau

0.11

0.07

Zimbabwe

0.29

0.18

Kenya

0.10

0.11

Source: Authors’ computations.

Note: The ratio is the fraction of GDP per capita to South Africa’s.

were the only economies to overtake South Africa in terms of per capita income. An important ‘neighbourhood effect’ is apparent in the data: Botswana, Lesotho and Swaziland, all to a large degree integrated into the South African economy, are among the faster converging economies.

Twenty-three countries experienced little or no change in their income levels relative to South Africa, and 11 - including Cote d’Ivoire, Gabon, Madagascar and Zimbabwe - had sharp deteriorations. The increase in income divergence was particularly sharp in Cote d’Ivoire, the Democratic Republic of Congo and Gabon. A majority of resource-rich countries (including oil exporters) did not improve their relative positions (Angola, Chad, Democratic Republic of the Congo, Nigeria and Zambia), providing support to the arguments of Collier (2007) and others that the ‘natural resource curse’ is particularly relevant in Africa.

A more formal test of whether the income per capita of poorer African countries is converging towards the region’s richer ones can be conducted using the following unconditional convergence model:

where Ah) is the mean growth rate of country i, and Y7 is the GDP per capita of country i in 1975. For convergence to occur, poor countries have to grow faster (Barro 1991; Barro and Sala-i-Martin 1991), making the predicted sign of в in Equation 8.4 negative.

The regression, shown in Figure 8.9, offers no evidence of unconditional convergence. The estimated coefficient is not significantly different from zero (в = -0.122, t = -0.29), indicating that in Africa the initial level of income alone has no effect on the growth rate.13

Mean growth as a function of initial conditions Source

Figure 8.9 Mean growth as a function of initial conditions Source: Author’s computations.

Lorenz curves

Figure 8.10 Lorenz curves: GDP per capita Source: Author’s computations.

These results confirm our heuristic evidence of substantial inertia in the income distribution and are consistent with other research. For example, McCoskey (2002) finds no evidence of unconditional convergence in Africa using long-run panel data. We also test for convergence in the period 1995-2005 - due to the structural break that took place in the growth series at about that point - by regressing mean growth during 1995-2005 on income in 1995. The results (P = 0.43, t = 0.19) are essentially the same as those reported for Equation 8.4.14 Africa’s poorer economies were not converging towards the income levels of their richer neighbours, even after the growth acceleration.

The lack of income convergence over the past 30 years has led to an increase in inequality of per capita incomes among countries across Africa. Lorenz curves of the GDP per capita at 10-year intervals from 1975 to 2005 are presented in Figure 8.10. These show increasing income inequality in each 10-year period. The sharpest rise in inter-country inequality took place between 1985 and 1995. This is consistent with the growing polarization of income and the emergence of the rich country club shown in the kernel densities.

The ratio of income of the richest 10 per cent of countries to the poorest 10 per cent of countries rose from 10.5 in 1975 to 18.5 in 2005. In 1975-2000 South Africa’s GDP per capita (then the highest in the region) was 17 times higher than that of Malawi. In 2000-2005, the gap between the highest GDP per capita country, the Seychelles, and Malawi had grown to 24 times.

 
Source
< Prev   CONTENTS   Source   Next >