copy the linklink copied!1. Ethiopia’s rural-urban transformation process

Ethiopia has achieved sustained economic growth since early 2000s, which led to significant poverty reduction and an overall improvement in well-being in rural and urban areas. The growth was led by a series of structural reforms which started in mid-1990s, and promoted the Agricultural Development-Led Industrialisation (ADLI) strategy. Today Ethiopia’s socio-economic landscape is changing governed by three main trends: an economic, a demographic, and a spatial transformation. This chapter analyses the ways in which these changes are shaping Ethiopia’s rural-urban transformation. It highlights that these three transformations will have a significant impact in rural areas. Moreover, it also highlights that, although Ethiopia’s structural reforms have led to large improvements in rural and urban well-being, the poverty gap between the two territories is now widening. Addressing rural development will require strategies that account for these transformations, and consider the inherent linkages between rural and urban areas.

    

copy the linklink copied!Introduction

Ethiopia’s rural-urban transformation is ongoing. Ethiopia is today a predominantly rural country, with almost 80% of its population residing in rural areas. However, rural areas are transforming following changes at both sub-national and national levels. This chapter analyses the ways in which these changes are shaping Ethiopia’s rural-urban transformation process. In particular, the chapter focuses on the main trends driving this transformation, while taking a close look to the evolution of welfare across rural areas. It argues that fast economic growth and a serious commitment to rural development have increased the well-being of rural populations. However, the welfare gap between urban and rural areas is increasing. Addressing this gap will require accounting for the linkages between rural and urban areas across multiple economic sectors, as well as policy actions that involve different levels of government. Overall, it will require rethinking the current strategies towards rural development.

Ethiopia’s strategic planning has been key to achieving a dynamic economy. Ethiopia has undergone sustained economic growth since the mid-2000s, with a two-digit Gross Domestic Product (GDP), which was consistently higher than the sub-Saharan average. This growth process has been led by the extensive and continuous efforts of the Ethiopian government. The backbone of these policy actions is the Agricultural Development Led Industrialisation (ADLI) strategy. Established in 1994, the ADLI aimed to set the basis for structural transformation by focusing on agricultural growth. The ADLI has guided subsequent national development plans since the early 2000s until today (see Chapter 3 for an extensive analysis on these plans).

However, Ethiopia’s economy and society are changing. These changes are governed by three ongoing transformations that will have major effects on the well-being of rural populations.

The first transformation is demographic. Ethiopia is in the early stages of its demographic transition, i.e. the country’s population will continue to grow between now and 2050, while a large number of people will enter the labour market. The latter will be particularly important for rural areas, as these areas have higher fertility rates.

The second transformation is economic. Although the agricultural sector’s contribution to GDP is decreasing, it will still account for more than two-thirds of employment. In addition, non-farm activities only account for a small share of rural employment. The premature state of the rural non-farm economy questions the sector’s reliability as a potential source of employment opportunities in short or medium term. Overall, structural transformation is taking place at a slow pace.

The third transformation is spatial. Ethiopia will remain a predominantly rural country until 2050. However, it is urbanising rapidly: urban population will almost double by 2030. Although the country is currently characterised by a monocentric urban system, the current urbanisation process is mainly being propelled by intermediary cities. Intermediary cities have a strong potential to contribute to rural development but are confronted with several binding constraints (see Chapter 2).

In addition to these transformations, Ethiopia is confronted with an increasing rural and urban gap in well-being. Indeed, although there have been significant efforts in terms of poverty reduction, the difference in welfare between rural and urban areas is increasing.

Effectively addressing these challenges will depend on the capacity of institutions and policies to adapt to these changes. In practice it will require a paradigm shift in Ethiopia’s approach to rural development. In other words, it will require updating ADLI in order better capture Ethiopia’s new reality.

copy the linklink copied!Ethiopia’s economic transformation

Ethiopia is characterised by a dynamic economy

Ethiopia has achieved sustained economic growth since the mid-1990s. Ethiopia’s gross domestic product (GDP) per capita has experienced sustained growth, with an average annual growth rate of 7.4% between 2004 and 2018 (Figure ‎1.1). Ethiopia’s economic growth outperformed the average of sub-Saharan African countries, which stood at 5.2% during the same period (Figure ‎1.1). Ethiopia’s GDP per capita, however, remains low compared to regional standards. Indeed, in 2018, Ethiopia’s GDP per capita was estimated at almost USD 1 800 (constant 2011 international dollars), representing less than half of that of the regional average (excluding upper-middle-income countries such as South Africa), which reached nearly USD 3 500 (constant 2011 international dollars) during the same year (Figure ‎1.1).

copy the linklink copied!
Figure ‎1.1. Evolution of Ethiopia’s GDP per capita
Figure ‎1.1. Evolution of Ethiopia’s GDP per capita

Note: ETH = Ethiopia; SSA = sub-Saharan Africa. PC = per capita. GDP per capita expressed in constant 2011 international dollars.

Source: World Bank (2019[1]).

A series of structural reforms that started in the mid-1990s are the main drivers of Ethiopia’s successful economic performance. These reforms focused on promoting agricultural development coupled with an unprecedented public investment in pro-poor sectors. In particular, reforms focused on the Agricultural Development-Led Industrialisation (ADLI) strategy. The ADLI is a guiding framework for agricultural policies which promotes the use of land and labour resources. It involves the use of labour-intensive methods, infrastructure investments, health interventions, and capacity building to increase agricultural productivity. Another important reform focused on prioritising government spending on capital projects over consumption, which facilitated an unprecedented investment in both soft and hard infrastructure, despite Ethiopia’s low domestic savings and tax revenues. Economic growth was further supported by increasing international trade, foreign direct investment (FDI) and social improvements (e.g. improvements in educational attainments) (Moller, 2015[2]).

Ethiopia’s structural transformation is still ongoing

Agriculture has consistently been the backbone of the Ethiopian economy, but its contribution to GDP is decreasing. Indeed, since the early 2000s, agriculture’s share of GDP has decreased, whereas the service sector’s contribution has increased. In 1992, the share of GDP coming from agriculture peaked at 64%; since then it has decreased, reaching 31% in 2018. In contrast, the service sector’s contribution to GDP has increased since the early 1980s and has hovered at around 40% since the early 2000s. Industry has historically accounted for the second-lowest share of Ethiopia’s GDP. Nonetheless, since 2011, it has shown a positive growth trend, reaching 27% by 2018 (Figure ‎1.2). The increasing contribution of industry to the economy does not follow from an increase in manufacturing, but rather from a boom in the construction sector (Moller, 2015[2]).

copy the linklink copied!
Figure ‎1.2. Evolution of Ethiopia’s gross value added
Figure ‎1.2. Evolution of Ethiopia’s gross value added

Source: World Bank (2019[1]).

In addition to the decreasing contribution of agriculture to the national economy, there has been a moderate shift in employment out of agricultural activities. Figure ‎1.3 shows that, between 2005 and 2013, the share of employment corresponding to agricultural activities decreased from 80% to 73%. In parallel, employment in the service sector increased from 13% to 20%. However, this shift may be lower once adjusted for differences in the definition of unpaid work in the labour force surveys conducted in 2005 and 2013. This would lead to adjusted employment rates of 77% in agriculture, 8% in industry, and 16% in services in 2013 (Schmidt and Bekele, 2016[3]). In other words, agricultural employment may have only decreased by three percentage points in eight years.

copy the linklink copied!
Figure ‎1.3. Employment across economic sectors, 2005 and 2013
Figure ‎1.3. Employment across economic sectors, 2005 and 2013

Note: Amounts may not add up to 100% due to rounding.

Source: World Bank (2019[1]).

Although Ethiopia’s structural transformation process seems slow, it may be moving faster than in other countries in sub-Saharan Africa. Figure ‎1.4 shows the average annual growth rate of the employment share in agriculture for 16 African countries. Between 2000 and 2016, Ethiopia’s share of employment in agriculture has experienced an annual average growth rate of approximately -2.1%.1 This ranks Ethiopia fifth out of 20 sub-Saharan African countries in terms of the speed of structural transformation. However, when considering only East African countries, Ethiopia ranks first above Rwanda, Kenya and Tanzania.

copy the linklink copied!
Figure ‎1.4. Average annual growth rate of the share of employment in agriculture for period 2000-16
Figure ‎1.4. Average annual growth rate of the share of employment in agriculture for period 2000-16

Source: Authors’ computations based on AUC/OECD (2019[4]) using data from the ASD.

Non-farm activities will be increasingly important for rural development but today they offer limited employment opportunities

Ethiopia’s rural non-farm economy is still at an early stage in its development. In 2011, 23% of rural households had a non-farm enterprise; in contrast to countries like Tanzania (39%), Uganda (42%), Nigeria (53%), and Niger (62%) where non-farm enterprises were more important for rural households (Nagler and Naudé, 2014[5]). Most income generating activities in Ethiopia are directly linked to crop production, while only a small share comes from non-farm activities. Figure ‎1.5 shows the distribution of income sources for rural households in Ethiopia and a selected group of African countries. This figure shows that more than 70% of the income of rural households in Ethiopia comes from crop-production; while non-farm income (non-farm wages and self-employment) only accounts for 6% of rural households. These estimates are based on the Ethiopian Rural Socioeconomic Survey (ERSS), but a study conducted by Bachewe et al. (2016[6]), finds similar shares using the Agricultural Growth Program (AGP) baseline survey.2 This study finds that crop production accounts for almost 72% of rural household income, while non-farm income accounts for almost 11% (non-farm enterprises, 8%; non-farm wages, 3%). These shares significantly differ from other countries in the region. Currently, Ethiopia is the country with the largest share of revenue resulting from crop production; it is followed by Malawi (59%) and Tanzania (53%), while Niger, Nigeria, Uganda stand at the same level (48%). In Nigeria, non-farm revenue accounts for almost 40% of rural household income; in Niger and Uganda, this share is close to 30%, while in Tanzania and Malawi is 20% and 14%, respectively. Ethiopia is the country with the lowest share of non-farm revenue by far.

copy the linklink copied!
Figure ‎1.5. Sources of rural household income across selected African countries
Figure ‎1.5. Sources of rural household income across selected African countries

Note: Non-farm income accounts for non-farm wages and self-employment revenue; others refer to transfers and any sources of income. Data for Ethiopia refers to 2012; Malawi, 2011; Niger, 2010/2011; Nigeria, 2010; Tanzania, 2009; and Uganda, 2009/2010.

Source: Adapted from Davis, Giuseppe and Zezza (2014[7]).

What drives Ethiopian rural households to participate in non-farm activities? Both age and education are factors that influence the decision to participate in non-farm activities in rural areas. Bachewe et al. (2016[6]) find that the age of the household head is positively correlated to crop and livestock income, and negatively correlated to agricultural wage and non-farm enterprises. In other words, older household heads seem to have greater experience for generating income from farm activities while younger heads tend to involve in non-farm activities to secure their livelihoods. Everything else equal, education is positively correlated to diversification, i.e., better skills may facilitate the participation into non-farm enterprises and non-farm activities from which they can get a wage. Moreover, these authors also identify land quality3 as a factor contributing to diversifying income sources, where lower land quality pushes households to engage in non-farm activities. The latter is also consistent with results from a similar study carried out by Schmidt and Bekele (2016[3]). The authors find that push factors such as poor land quality play a key role in households engaging in non-farm activities. Indeed, farmers in less favourable agricultural areas that have limited assets (livestock and land), as well as limited access to financial institutions, are more inclined to engage in non-farm activities. Moreover, in this study, youth households also seem to be prone to engaging in non-farm activities as a coping strategy to address access to, or lack of, land, farm services, and enhancing agricultural technologies.

Non-farm enterprises are more prominent across small towns in Ethiopia. Figure ‎1.6 shows the share of households reporting having at least one non-farm enterprise, as well as the four most frequent types of non-farm activities. Small towns account for a larger share of households with at least one non-farm enterprise. Almost 54% of small town households have non-farm enterprises, compared to 37% in large towns and 18% in rural areas. The most important types on non-farm activities for these households are non-farm business (26%), which account for all off-far activities, including services from home, and shops; processed agricultural products (16%); and trading business (10%), which include trading activities in streets and markets.

copy the linklink copied!
Figure ‎1.6. Households reporting one or more non-farm enterprise by type
Figure ‎1.6. Households reporting one or more non-farm enterprise by type

Note: Non-farm business refers to off-farm activities, including services ran from home, like shops. Trading business include streets and markets. Small towns are those populated centres with less than 10 000 population according to the 2007 Census. Large towns are those populated centres with more than 10 000 inhabitants.

Source: Integrated Surveys on Agriculture Ethiopia Socioeconomic Survey (ESS), 2015-16.

Accessibility is one of the main constraints for developing non-farm enterprises in Ethiopia. Accessibility issues are not only limited to space, however. The most important constraint for Ethiopian households is access to finance. Almost 35% of households across the country report this as their most relevant constraint; followed by access to markets (30%), and transportation (14%) (CSA, 2017[8]). It is important to note that these constraints may be more binding for poorer households or marginalised groups, which in many cases rely on non-farm activities as a coping strategy (Beegle and Christiaensen, 2019[9]).

Since the late 1990s, there has been a transition from informal to formal employment in Ethiopia. According to official sources, the share of population employed in the informal sector4 declined from 50.6% in 1999 to 25.8% in 2013. The female population employed in the informal sector also decreased from 64.8% to 36.5% during the same period (CSA, 2013[10]). Nevertheless, Ethiopian women’s ability to access the labour market remains limited. In 2016, the overall employment rate for females aged 15-49 stood at 33.3%, compared with 88.2% for males in the same age group (ICD, 2016[11]).

Channelling public expenditure to develop physical infrastructure has been a key component of Ethiopia’s recent economic development strategies. Starting with the Growth and Transformation Plan I (GTPI), launched in 2009/10, the Ethiopian government has focused on creating a conducive environment for structural transformation. In particular, government actions have focused on developing the necessary physical infrastructure (such as roads, irrigation and hydropower) for promoting the diversification of economic activities in the country. Between 2007/08 and 2015/16, total public expenditure increased from ETB 71 billion (Ethiopian birr) to ETB 149 billion (constant at 2010 prices), representing an average annual growth rate of almost 10% (Figure ‎1.7). During this period, capital expenditure increased from ETB 36.4 billion to ETB 76.3 billion (constant at 2010 prices), representing, on average, 55% of the total expenditure and around 10% of Ethiopia’s GDP (NPC, 2017[12]). Ethiopia’s public expenditure surpassed the average public expenditure across Africa (which accounted for 7% of total GDP) between 2009 and 2016. Additionally, during the same period, private sector investment on capital projects was higher in Ethiopia, amounting to 18.3%, compared with the average of 15.5% across Africa (AUC/OECD, 2018[13]).

copy the linklink copied!
Figure ‎1.7. Public expenditure by type, 2007/08-2015/16
Figure ‎1.7. Public expenditure by type, 2007/08-2015/16

Note: Data expressed in real ETB billions at 2010 prices.

Source: NPC (2017[12]).

Economic growth paired with social policies has brought significant poverty reduction

Economic growth has translated into significant poverty reduction and overall human development since the mid-1990s. Ethiopia’s poverty head count, i.e. the share of the population living below the national poverty line, fell from 44% in 2000 to less than 30% in 2011, and to 24% by 2017 (Figure ‎1.8). The poverty gap index has followed a similar trend. The poverty gap index provides information about the extent to which the population falls below the poverty line; it measures the average difference (or gap) between the income (expenditure) of individuals who fall below the poverty line and the value of the poverty line. It is expressed as a percentage of the poverty line, meaning that a higher poverty gap index value indicates that the poverty is more severe. In 2000, the mean poverty gap index was close to 12% of the poverty line, but this had almost halved by 2016, reaching close to 7% (Figure ‎1.8).

copy the linklink copied!
Figure ‎1.8. Evolution of poverty in Ethiopia
Figure ‎1.8. Evolution of poverty in Ethiopia

Source: NPC (2017[12]) using estimates from Household Income, Consumption, and Expenditure (HICE) Surveys.

Human development has also increased since the mid-1990s. During the period 2000-10, Ethiopia’s Human Development Index (HDI) showed considerable improvement. The country’s HDI shifted from 0.35 in 2000 to 0.46 in 2013, with an average annual increase of 2.12% (UNDP, 2014[14]). Since the early 2000s, Ethiopia has developed a series of poverty reduction strategies and programmes, along with five-year development plans, that aimed at increasing public expenditure on basic services (e.g. education and health). One of these plans is the Productive Safety Net Programme (PSNP).

Launched in 2005, the PSNP was aimed at targeting food-insecure households through cash and food transfers. The main objective of the programme was to provide more predictable and sustained interventions to households affected by droughts. The PSNP has contributed to agricultural growth through increases in input use and facilitated distribution. Estimates suggest that the impact of transfers has led to reduction in poverty by 7% (Moller, 2015[2]).

The commitment of authorities to expanding social infrastructure has further contributed to positive health and education outcomes. For instance, between 2005 and 2013, the primary school net enrolment rate rose from 68.5% to 85.7%, while the share of immunised children rose from 44.5% to 87.6%, and births attended by trained healthcare workers increased from 12.4% to 23.1% during the same period (UNDP, 2014[14]). It should be noted that Ethiopia financed a large proportion of its investment in infrastructure through public debt, and benefited from debt relief in 2004, thanks to the Heavily Indebted Poor Countries (HIPC) initiative (IMF, 2015[15]).

Ethiopia has also invested extensively in pro-poor sectors. Between 2009 and 2016, spending on education, health, agriculture, roads and water, and sanitation services accounted for 70% of total government spending (Figure ‎1.9). In parallel, investment in public infrastructure such as rural roads granted better access to markets for residents, topping up the investment in agriculture. In fact, between 2010 and 2015, total road length increased by 30%, resulting in better and expanded all-season access to functioning roads (IMF, 2015[15]).

copy the linklink copied!
Figure ‎1.9. Share of public expenditure by anti-poverty action, 2008/09-2015/16
Figure ‎1.9. Share of public expenditure by anti-poverty action, 2008/09-2015/16

Source: NPC (2017[12]).

Poverty reduction has, however, been followed by a small increase in inequality. Indeed, income inequality – measured using the Gini coefficient – has increased since the mid-1990s. In 1995/96, estimates placed Ethiopia’s national Gini coefficient at 0.29, compared with 0.33 in 2015/16 (NPC, 2017[12]). This small increase has mainly been driven by growing differences in terms of income and access to certain services between urban and rural households. Nevertheless, Ethiopia remains the country with the lowest level of inequality in the region. Among East African countries, Ethiopia ranks at the bottom in terms of both the Gini coefficient and the difference in income shares between the richest 10% and the poorest 10% (Figure ‎1.10). Ethiopia’s Gini coefficient (0.33) is followed by Sudan’s and Mauritius’, both estimated at 0.35. At the other end of the ranking, Kenya and Rwanda show the highest values, at 0.48 and 0.50, respectively. In terms of the difference in income shares, Ethiopia stands at 8.6. In other words, the richest 10% of the population is 8.6 times richer than the poorest 10%. This contrasts with countries such as Kenya or South Sudan, where this statistic reaches 22.8 and 25.5, respectively.

Although Ethiopia’s Gini coefficient is low compared with other developing countries, the country’s fiscal system needs to put additional efforts to reduce income inequality. In 2010, taxes, transfers and subsidies contributed to reducing Ethiopia’s Gini coefficient from 32 points (market income) to 30 points (final income). This decrease of two percentage points is among the lowest in a sample of 30 developing countries across the world, and is the lowest among the African countries considered in the sample (Figure ‎1.11).

copy the linklink copied!
Figure ‎1.10. Gini coefficients and difference in income share in East African countries
Figure ‎1.10. Gini coefficients and difference in income share in East African countries

Note: The difference in income share refers to the ratio between the income shares held by the richest 10% to the income held by the poorest 10% in the national income distribution.

Source: AUC/OECD (2018[13]).

copy the linklink copied!
Figure ‎1.11. Gini coefficients for market income and final income
Figure ‎1.11. Gini coefficients for market income and final income

Note: Data refer to different years.

Source: AUC/OECD (2018[13]) using data from CEQ Institute (2019[16]), Commitment to Equity Institute Data Centre on Fiscal Redistribution (database).

Unemployment in Ethiopia has gone down since the mid-1990s. The unemployment rate fell from 8% in 1999 to 5% in 2005, and to 4.5% in 2013 (CSA, 2013[10]). This has also been the case for female and youth (population aged 15-29) unemployment rates. For the former group, the unemployment rate has decreased from 12.5% in 1999 to 7.8% in 2005, and 6.5% in 2013. For the latter, the unemployment rate decreased from 11.1% in 1999 to 6.8% in 2013 (CSA, 2013[10]). Nevertheless, as the next chapter will discuss, unemployment is becoming a predominantly urban phenomenon, which also has a strong regional component.

copy the linklink copied!
Box ‎1.1. Trends in Ethiopia’s governance

Although overall governance has improved in Ethiopia, the country has undergone significant changes in recent years. Between 2005 and 2015, Ethiopia made progress in two out of six indicators: Rule of Law and Control of Corruption out of Rule of Law, Voice and Accountability, Regulatory Quality, Political Stability, Government Effectiveness, and Control of Corruption (World Bank, 2015[17]). Political stability, voice and accountability, and regulatory quality have fluctuated with an overall decreasing trend during the same time period. In addition, the country demonstrated the second-greatest improvement among countries in East Africa in the Ibrahim Index of African Governance between 2006 and 2015 (only outperformed by Rwanda), ranking seventh out of 13 countries in 2015 (Mo Ibrahim Foundation, 2016[18]).

copy the linklink copied!
Figure ‎1.12. Evolution of governance in Ethiopia
Figure ‎1.12. Evolution of governance in Ethiopia

Source: Mo Ibrahim Foundation (2016[18]).

However, Ethiopia has undergone major changes in its governance since 2015. Following three years of political unrest in the country, the former Prime Minister Hailemariam Desalegn resigned, and the current Prime Minister, Abiy Ahmed, was elected in April 2018 as the head of the ruling party, the Ethiopian People’s Revolutionary Democratic Front (EPRDF). The newly elected government implemented substantial political and economic reforms, including the reconciliation with neighbouring Eritrea.

Agriculture will continue to play a key role for development

To say that agriculture plays a pivotal role in Ethiopia’s development path is an understatement. The agricultural sector has evolved since the late 1990s following the adoption of the ADLI. Although it started from a low base, the sector has progressively adopted different types of improved agricultural technology, which has contributed to the increase in the acreage of some of the most important cereals in Ethiopia (Seyoum Taffesse, 2019[19]). Growth in the production of grains, and particularly cereals is key for rural livelihoods.

Grain crops dominate Ethiopia’s agricultural production. In 2018, grain crops accounted for 79% of all crops produced, and almost 88% of all crop area in the country (Table ‎1.1). Among grain crops, cereals account for up to 88% of production, followed by pulses (10%) and oil seeds (2%). Teff, maize and sorghum are the cereals accounting for the largest shares of crop area in the country, at 30%, 23%, and 18%, respectively. Pulses are the second-largest type of grain grown in the country, with some of the most important crops in this group including faba beans, white haricot beans, red haricot beans, and chickpeas (CSA, 2018[20]). In terms of production, root crops are the second-most important group after grains. Sweet potatoes, taro (godere) and potatoes are among the most important crops in this group. In terms of crop area, Table ‎1.1 shows that the second-most important crop after grains is coffee, which accounts for more than 5% of the total crop area in the country. Vegetables and fruits only account for a small share of Ethiopia’s crop area (1.7% and 0.8%, respectively); this is also the case for their shares in terms of production (2.2% and 2.1%, respectively).

copy the linklink copied!
Table ‎1.1. Distribution of production and crop area of agricultural commodities, 2018

Crop

Production

Area

Grain crops

79.03%

87.93%

Root crops

11.36%

1.60%

Sugar cane

3.24%

0.19%

Vegetables

2.23%

1.67%

Fruit crops

2.09%

0.83%

Coffee

1.24%

5.28%

Chat

0.69%

2.24%

Hops

0.12%

0.26%

Total

100%

100%

Grain crops

Production

Area

Cereals

88%

81%

Pulses

10%

13%

Oil seeds

2%

6%

Total

100%

100%

Note: Data for private peasant holdings during the Meher season, which is the main crop season. It accounts for any temporary crops harvested between the months of September and February.

Source: CSA (2018[20]) .

Cereal yields have experienced significant growth since 2000. Figure ‎1.13 presents the yields of teff, sorghum, barley, wheat and maize for the years 2000, 2005, 2010 and 2015. Each of these selected cereals show a positive progression during this period. Overall, cereals’ yields increased from 1.21 kg/ha in 2000 to 1.43 kg/ha in 2005, 1.83 kg/ha in 2010, and 2.32 kg/ha in 2015. In other words, during the period 2000-05, cereal yield increased by 18%, while for both the periods from 2005-10 and 2010-15, the yield increased by 27%.5 However, these numbers hide important differences in the progression of different crops across these periods. For instance, during the period 2010-15, wheat and maize yields increased the most, from 1.83 kg/ha and 2.53 kg/ha in 2010 to 2.54 kg/ha and 3.42 kg/ha in 2015, respectively. In other words, these crops’ yield increased by 38% and 35%. In contrast, during the previous period (2005-10), wheat and maize yield increased by 21% and 16%, respectively, while those of sorghum and teff each increased by 40%.

Smallholders are key actors for agricultural growth in Ethiopia

Agricultural growth has been driven by a mixture of factors that have influenced the performance of smallholders. Smallholders account for most of Ethiopia’s agricultural production. In 2018, 16 million smallholders produced almost 95% of all grain crops in the country (CSA, 2018[20]).6 Ethiopian smallholders are characterised by a very small plot size. In 2015, almost 64% of all holders produced crops in less than 1 ha, and almost 40% of holders produced crops in less than 0.5 ha (Figure ‎1.14).

Smallholders have increasingly adopted modern fertilisers and improved seeds. The Government of Ethiopia (GoE) has made significant efforts in promoting the adoption of modern agricultural technology, which has resulted in larger areas using modern fertilisers and improved seeds. Indeed, between 2005 and 2015, the area using modern fertilisers increased from 40% to 58%, and from 5% to 12% in the case of improved seeds (Figure ‎1.14). Extension services have also played a key role in promoting agricultural technology adoption. During the period 2005-15, the area covered by extension services has also increased, from 22% to 33%. According to Bachewe et al. (2018[21]), agricultural growth has been further propelled by significant public investments which, in addition to expanding extension services, have improved education outcomes, limited negative shocks (through social protection programmes such as the PSNP) and improved market efficiency by expanding the road network. Moreover, Bachewe et al. highlight that external factors such as good weather conditions, high international prices for agricultural commodities, and rapid urbanisation (higher urban demand) further contributed to promoting the adoption of agricultural technology between 2004 and 2015.

The increasing agricultural output in Ethiopia has been a key driver for poverty reduction. Hill and Tsehaye (2015[22]) show that for every 1% increase in agricultural output, poverty was reduced by 0.9%. According to their estimates, agricultural growth contributed to reducing poverty by an average annual rate of 2.2% after 2005 and 0.1% before 2005. However, this effect is only significant in areas located close to cities of at least 50 000 inhabitants. As previously discussed, public investment has been conducive to poverty reduction. However, additional efforts are needed in order to reach remote populations. Access to urban centres and the growth of non-farm sector activities have complemented the welfare gains resulting from investment in public services and infrastructure. Moreover, good climatic conditions and high food prices further contributed to reducing poverty.

copy the linklink copied!
Figure ‎1.13. Yields of selected cereals between 2000 and 2015
Figure ‎1.13. Yields of selected cereals between 2000 and 2015

Note: “Cereals” accounts for teff, barley, wheat, maize, sorghum, finger millet, oats/“aja” and rice.

Source: CSA’s AgSS surveys for the years 2000, 2005, 2010 and 2015.

copy the linklink copied!
Figure ‎1.14. Distribution of smallholders by land size and use of improved agricultural technology between 2005 and 2015
Figure ‎1.14. Distribution of smallholders by land size and use of improved agricultural technology between 2005 and 2015

Source: CSA’s AgSS surveys for the years 2005 and 2015 (CSA, 2018[20]).

There is, however, scope to increase cereals productivity

Ethiopia’s cereal productivity has advanced at a steady pace compared to other countries in the region. In 2000, Ethiopia’s cereal productivity was 78% of Tanzania and 81% of Kenya (Figure ‎1.15); by 2015, Ethiopia’s cereal productivity had outpaced that of these countries, representing 158% of Tanzanian and 144% of Kenyan cereal productivity. However, compared to countries like Vietnam and Egypt, Ethiopia’s productivity remains low. Indeed, by 2015, Ethiopia’s cereal productivity amounted to 46% of Viet Nam and 36% of Egypt.

copy the linklink copied!
Figure ‎1.15. Ethiopia’s cereal productivity compared to selected countries
Figure ‎1.15. Ethiopia’s cereal productivity compared to selected countries

Note: The percentages represent Ethiopia’s cereal productivity (Kg/Ha) over that of the selected countries. Higher than 100% means that Ethiopia’s productivity is higher than that of the selected country.

Source: World Bank (2019[1]).

copy the linklink copied!Ethiopia’s demographic transformation

Ethiopia is going through the early stages of a demographic transition

Since the mid-1950s, Ethiopia’s total population has not stopped growing. In fact, it has increased from 18 million in 1950 to almost 99 million in 2015. This made Ethiopia the most populated country in East Africa in 2017, followed by Tanzania (53 million), Kenya (47 million) and Uganda (40 million) (UNDESA, 2017[23]). This trend has mainly been driven by the spread between birth and mortality rates. This spread has been observed in every country during the early stages of development. The way this spread evolves over time is commonly referred to as a “demographic transition”. The notion of demographic transition follows from observing the demographic processes experienced by most developed countries today. A demographic transition can be divided into a number of consecutive stages, which are strongly interlinked with social and economic changes. There are, however, two main general phases. In the first phase, as the economic situation improves (improvements in access to food supplies, drinkable water and health services become accessible), mortality rates tend to decline. Later on, during the second phase, better access to education, women’s empowerment, higher wages and productive jobs lead to a decline in birth rates. When the latter takes place, population growth slows down.

Ethiopia’s birth and death rates have consistently declined since the 1980s. During the period 2010-15, mortality stood at 7.5 deaths per 1 000 population, compared with 21.4 deaths per 1 000 population in 1980-85 (Figure ‎1.16). Moreover, during the same period, Ethiopia’s death rate was lower than both the sub-Saharan and East African averages of 10.3 and 8.4 deaths per 1 000 population, respectively (UNDESA, 2017[23]). Ethiopia’s birth rate has also decreased, from 45.4 births per 1 000 population in 1980-85 to 33.6 births per 1 000 population in 2010-15 (Figure ‎1.16). Ethiopia’s birth rate from 2010-15 also stands below the sub-Saharan and East African averages (37.9 and 37 births per 1 000 population, respectively). In order to provide a better perspective on Ethiopia’s demographic transition, Annex Figure ‎1.A.1 (in ‎Annex 1.A) compares the demographic transitions of Ethiopia and Viet Nam. This figure shows that, although Ethiopia’s and Viet Nam’s birth rates are converging, there is still a big gap between their rates. In 2010-15, Ethiopia’s birth rate was almost twice as high as Viet Nam’s. Ethiopia’s population will keep growing as long as the gap between the birth and death rates remains; however, decreasing trends for both rates suggest that the country’s rate of population growth will slow down.

copy the linklink copied!
Figure ‎1.16. Ethiopia’s demographic transition
Figure ‎1.16. Ethiopia’s demographic transition

Source: UNDESA (2018[24]).

Ethiopia’s demographic transition is further reflected in the structure of its population

Figure ‎1.17 shows the distribution of the Ethiopian population by age and sex in 2015. The shape of this distribution shows that Ethiopia is characterised by a very young population. Indeed, almost 42% of the total population is under 14 years of age, while the working-age population (those aged 15-64 years old) accounts for 55%.

The nature of Ethiopia’s population structure puts a greater burden on its working-age population. Thus, the working-age group will continue to support the group which is not yet in the labour market. The latter can be expressed as a dependency ratio, i.e. the population outside the labour market (or dependants) over the working age population. In 2015, Ethiopia’s dependency ratio stood at 82%; this value remains only slightly above neighbouring Kenya (78%), but is significantly higher than countries such as Morocco (51%) and Viet Nam (42%), which are more advanced in terms of their demographic transition Annex Figure ‎1.A.3 in ‎Annex 1.A).

copy the linklink copied!
Figure ‎1.17. Ethiopia’s population pyramid, 2015
Figure ‎1.17. Ethiopia’s population pyramid, 2015

Note: Each horizontal bar represents the percentage of the total population of males and females in each age group.

Source: UNDESA (2018[24]).

Ethiopia’s current population structure also brings economic opportunities. Changes in the population’s age structure affect the balance between individuals who are producing and those who are consuming; this implies that countries with a large proportion of both very young and very old people have fewer producers relative to consumers, and vice versa (Mason et al., 2017[25]). This is the basis for the demographic dividend, the contribution of changes in the population’s structure to economic growth. Figure ‎1.18 shows the demographic dividend for Ethiopia. Ethiopia’s demographic dividend started in 2002, and during its first year it potentially contributed 0.068% to economic growth; it will continue to contribute to economic growth until it peaks 22 years later (by 2024), reaching up to 0.92%. It will then slowly decline, and 65 years after it began it will stop (by 2067). Changes in the population structure from then on will have a negative impact on economic growth.

copy the linklink copied!
Figure ‎1.18. Demographic dividend
Figure ‎1.18. Demographic dividend

Note: “Dividend” refers to the contribution of changes in the population structure to economic growth. “Yrs since onset of the dividend” refers to how many years have passed since the first dividend started. In the case of Ethiopia, the first dividend started in 2002, according to United Nations (UN) estimates. The blue bar in the figure marks the year 2020, i.e. 18 years after the beginning of the first dividend.

Source: Mason et al. (2017[25]).

However, the demographic dividend can only materialise if the increase in the labour supply is matched with productive jobs. Moreover, Ethiopia’s demographic transition is taking place along with a rapid urban-rural transformation process, which will further necessitate improving public service delivery and promoting a conducive environment for employment creation in cities of different sizes.

Internal migration plays a key role in Ethiopia’s rural-urban transformation process

Although the share of the population changing residence in the country is small, the most important flow of population is taking place within rural areas. Between 2008-13, close to 35% of the population movements that took place in Ethiopia were rural to rural; rural to urban flows represented 33%, while urban-to-urban migration accounted for 21% of all population flows (Annex Table ‎1.A.4 in ‎Annex 1.A). Although the scale of internal migration7 has not drastically changed since the late 1990s, its patterns have evolved. Notably, the importance of rural-to-urban migration has increased, while that of rural-to-rural migration has decreased. Indeed, in 2005, migration from rural to urban areas accounted for 24% of all migration flows, i.e. seven percentage points lower compared with 2013; in contrast, the share of migration across rural areas accounted for 46% of total migration flows, representing a decrease of more than ten percentage points compared with 2013 (CSA, 2014[26]).

Internal migration patterns, however, further differ across regions. Figure ‎1.19 presents the share of population flows across urban and rural areas for all Ethiopian regions. Addis Ababa is the region with the highest share of the population coming from rural areas (59%), followed by Harari and Dire Dawa, for which rural migrants accounted for almost 38%. However, Addis Ababa is not the region with the highest urban-to-urban migration. During the same period, urban-to-urban movements accounted for almost 52% of the internal migration in Harari and Dire Dawa, while for Addis Ababa this represented 41%.

Most population movements in Ethiopia are taking place within regions. From 2008 to 2013, most population movements took place within zones of the same region. This is the case in all regions, with the exceptions of Addis Ababa, Harari, Dire Dawa and Somali. This pattern is clearly identified in Figure ‎1.20. In Oromia, more than 80% of the population movements from 2008-13 took place within the region; in Amhara, Tigray and SNNPR, intra-regional migration accounted for approximately 70% of all population movements. In contrast, in the case of Addis Ababa, Harari, Dire Dawa and Somali, the largest share of migrants moved to a zone outside the region; it is interesting to note that for these four regions, the most frequent destination was Oromia (their neighbour),8 which accounted for 51%, 33%, 29% and 46%, respectively, of the population outflows originating in these regions. Oromia is, in relative terms, the most important destination in the country.

copy the linklink copied!
Figure ‎1.19. Migration across urban and rural areas by region
Figure ‎1.19. Migration across urban and rural areas by region

Note: Migration refers to population movements taking place from 2008-12.

Source: Labour Force Survey (2013), adapted from Schmidt et al. (2018[27]).

The most important driver of internal migration in Ethiopia is the search for employment. Migration between rural and urban areas has been explained as a result of wage differentials between the two areas (Harris and Todaro, 1970[28]), as well as an overall strategy at the household level that helps to maximise the welfare of household members while reducing the risk of external shocks (Rosenzweig and Stark, 1989[29]). Overall, evidence from different countries shows that most migrants (except those from countries facing conflicts or natural disasters) move from one place to another in order to find job opportunities, access to services, and an overall better quality of life. The case of Ethiopia is no different. Among recent migrants in 2013, finding a job was the main reason for changing residence, followed by moving in order to live with family members or due to marriage, and education (CSA, 2014[26]). Although finding a job, family reasons, and education are consistently the main drivers for migration across Ethiopia, in some regions other factors play a relevant role. For instance, shortage of land is more relevant in regions such as Benishangul-Gumuz, Amhara, and Gambela, while displacement linked to war or drought was more important in Somali and Afar.

Internal migration in Ethiopia is influenced by both personal characteristics and place-related factors. Independent of the reason for migrating, young and educated Ethiopians are more likely to migrate than their older and less-educated peers. In 2013, rural migrants were on average 10 years younger than rural non-migrants; they were also characterised by having twice as many years of education compared with non-migrant rural dwellers, and were three times more likely to have attained secondary-level education (Bundervoet, 2018[30]). Holding small sized-land and landlessness are also important drivers for migration (Dominiko, 2016[31]). In addition to individual factors, places characterised by high levels of poverty and limited accessibility to roads provide a less conducive environment for migrants; in other words, credit constraints and high migration costs make rural-to-urban migration less likely (Bundervoet, 2018[30]).

Overall, population mobility appears to be a welfare-maximising strategy in Ethiopia. Evidence from Tanzania suggests that migration is associated with an overall improvement in welfare (Beegle et al., 2011[32]). In the case of Ethiopia, although empirical evidence remains limited, rural-to-urban migrants seem to be better off when compared with non-migrants; consumption of goods other than food tends to more than double while diet further improves for migrants when compared with people who opted to not migrate (Bundervoet, 2018[30]).

copy the linklink copied!
Figure ‎1.20. Internal migration across Ethiopian regions
Figure ‎1.20. Internal migration across Ethiopian regions

Note: Internal migration refers to those population movements within and across regions from 2008-12. Arrows represent the share of population migration from one region to another or within the same region. The thickness of the arrows represents the magnitude of the flow. Colours differentiate regions. The order in which regions are presented around the circle represents total population size in 2013, starting with Oromia (the region accounting for the largest population in the country), moving clockwise until reaching Harari.

Source: Labour Force Survey (2013), adapted from Bundervoet (2018[30]).

copy the linklink copied!Ethiopia’s spatial transformation

Ethiopia’s rural population will continue to grow in parallel with a rapid urbanisation process

Ethiopia is, and will remain until at least 2050, a predominantly rural country. In 2015, the rural population was estimated to be approximately 80.5 million, or 81% of the total population (UNDESA, 2018[24]; CSA, 2013[33]). More importantly, although there are increasing investments to boost manufacturing, as well as ongoing efforts to improve rural electrification, irrigation and mechanisation (which will contribute to the rural-urban transformation), most of the population is expected to reside in rural areas until about 2050 (Figure ‎1.21). Ethiopia’s rural population is expected to reach 102 million by 2030, which represents an average yearly growth rate of 1.6% between 2015 and 2030; growth will then further decrease to 0.65% between 2030 and 2050.9 By the end of 2050, Ethiopia will remain predominantly rural, since the rural population will account for more than 61% of the total population (UNDESA, 2018[24]).

copy the linklink copied!
Figure ‎1.21. Rural and urban populations in Ethiopia, 1950-2050
Figure ‎1.21. Rural and urban populations in Ethiopia, 1950-2050

Source: CSA (2013[33]) and UNDESA (2018[24]).

Ethiopia is one of the least urbanised countries in the region. In 2015, urban areas hosted 20% of the Ethiopian population; this value is lower than the regional averages of sub-Saharan Africa and East Africa, which during the same year stood at 39% and 27%, respectively (UNDESA, 2018[24]). However, comparing urbanisation rates between countries like this has some caveats. The main issue is the different definitions of urban areas across countries, as an area that one country considers to be urban may not be classified as urban by others (see Box ‎1.2 for the definition of urban areas in Ethiopia). Overcoming this caveat requires identifying urban areas – across countries – based on a common criterion that leads to a comparable definition. Datasets such as Africapolis provide this information.10 According to this dataset, Ethiopia’s urban population accounted for 27% of the total population in 2015; in comparison, the urban populations in neighbouring countries are estimated to be 72% in Djibouti, 65% in Kenya and 38% in Tanzania.

copy the linklink copied!
Box ‎1.2. Being urban in Ethiopia

An urban centre in Ethiopia is defined as a locality with 2 000 or more inhabitants. These localities are referred as kebeles, which are the lowest-ranked administrative units with their own jurisdiction. However, areas classified as urban also include the following territorial administrative units regardless of the number of inhabitants:

  • all administrative capitals (includes all regional, zonal, and woreda capitals, as well as localities with urban dweller’s associations (UDAs))

  • municipal towns

  • all localities which are not included either in a) or b) with a population of 1 000 or more persons, and whose inhabitants are primarily engaged in non-agricultural activities (note that localities with a population of fewer than 1 000 persons should be considered as rural).

Urban centres with a population of 100 000 inhabitant or more, as well as regional capitals (irrespective of their population size), are further classified as major urban centres.

Rural areas are, by definition, those areas not classified as urban.

Source: CSA (2014[26]).

Addis Ababa is the most urbanised region in Ethiopia. Among the 11 regions within Ethiopia, Addis Ababa is the only region with 100% of its population classified as urban. However, despite being the most urbanised region in the country, it only represents 3.7% of Ethiopia’s population and close to 18% of the country’s total urban population (CSA, 2013[33]). Addis Ababa is followed by the regions of Dire Dawa and Harari, where the urban population accounts for 63% and 56% of the population, respectively. In contrast, in regions such as Oromia or Somali, urbanisation rates are close to 15% (Figure ‎1.22).

Oromia, Southern Nations, Nationalities, and Peoples’ Region (SNNPR), and Amhara account not only for the largest number of cities with more than 100 000 population, but also for the largest number of small agglomerations. Indeed, in 2015, 19 of the 24 cities with more than 100 000 people in Ethiopia were located in these regions. In contrast, in the regions of Afar, Benishangul-Gumuz, and Gambela, there are no agglomerations with more than 100 000 inhabitants (Table ‎1.2 and Figure ‎1.23). Oromia, SNNPR and Amhara also account for the largest number of agglomerations with fewer than 50 000 inhabitants. Oromia itself accounts for more than 40% of this type of agglomeration in the country, followed by Amhara (38%). It is important to note that Oromia hosts 28% of the urban population in the country; it is followed by Amhara (19%) and Addis Ababa (18%) (CSA, 2013[33]). Moreover, in 2017, Oromia and Amhara together accounted for 60% of the total population of Ethiopia.

copy the linklink copied!
Figure ‎1.22. Urbanisation rate across Ethiopian regions, 2015
Figure ‎1.22. Urbanisation rate across Ethiopian regions, 2015

Source: CSA (2013[33]).

copy the linklink copied!
Table ‎1.2. Number of agglomerations by city size and region

Less than 50 000

Between 50 000 and 100 000

Between 100 000 and 300 000

Between 300 000 and 500 000

More than 500 000

Average city size

Addis Ababa

0

0

0

0

1

3 711 329

Afar

9

0

0

0

0

19 529

Amhara

138

6

3

1

0

26 668

Benishangul-Gumuz

6

1

0

0

0

19 610

Dire Dawa

1

0

1

0

0

145 721

Gambela

4

1

0

0

0

25 678

Harari

0

0

1

0

0

111 073

Oromia

183

13

6

2

0

32 065

SNNPR

58

10

6

1

0

43 191

Somali

16

1

1

0

0

27 739

Tigray

29

5

1

1

0

36 348

Ethiopia

444

37

19

5

1

39 704

Note: Data for 2015.

Source: Authors’ calculations using data from Africapolis (SWAC/OECD, 2018[34]).

Ethiopia’s urban system is heavily skewed towards Addis Ababa

Addis Ababa is the largest city in Ethiopia and the only agglomeration with more than 1 million people. For 2015, according to official statistics, Addis Ababa’s population was projected to account for almost 3.3 million people (CSA, 2013[33]). However, according to Africapolis, Addis Ababa’s population was closer to 3.7 million (SWAC/OECD, 2018[34]). The second-largest city in the country was Adama, followed by Gondar, Mekele and Hawassa (CSA, 2013[33]). Ethiopia’s city rankings – the order of cities based on their population size – varies according to the definition of what a city is. For instance, according to Ethiopia’s National Urban Spatial Development Plan, the second-largest city in 2015 was Mekele, followed by Adama and Gondar (MoUDH, 2016[35]). And according to Africapolis, the second-largest city in 2015 was Harari, followed by Adama and Mekele (SWAC/OECD, 2018[34]). Although these rankings may differ in terms of the position of cities within the ranking, they are consistent with regard to the size difference between Addis Ababa and the following cities in the ranking: Addis Ababa is around 8-10 times bigger than the second-largest city. Moreover, there is a small group of cities characterised by a total population hovering at around 300 000 inhabitants; these include agglomerations such as Mekele, Adama, Dire Dawa, Gondar and Hawassa.

copy the linklink copied!
Figure ‎1.23. Urban agglomerations in Ethiopia
Figure ‎1.23. Urban agglomerations in Ethiopia

Note: City populations for 2015.

Source: Authors’ calculations using data from Africapolis (SWAC/OECD, 2018[34]).

Ethiopia’s urban system strongly relies on its capital city, Addis Ababa. Figure ‎1.24 presents the rank-size distribution for Ethiopian agglomerations in 2015. This figure highlights the role of Addis Ababa as an integral part of the urban system. As discussed, the capital city is very large compared with the following tier of cities in the ranking (Mekele, Adama, Dire Dawa, Gondar and Hawassa); after this group of cities, there is a large number of smaller agglomerations (with at least 10 000 people) scattered around the country that capture the remaining share of Ethiopia’s urban population.

Addis Ababa’s predominant role makes Ethiopia’s urban system fairly monocentric. Urban systems with a tendency to concentrate economic functions and populations in a small number of cities are considered to be monocentric, while those with a more even distribution are said to be polycentric.11 Figure ‎1.25 presents the extent of polycentricity for Ethiopia:12 when this measure is higher than 1, the urban system tends to be polycentric; when it is lower than 1, it suggests a monocentric urban system. Ethiopia has a coefficient of 0.77, suggesting that it has a monocentric national urban system further characterised by a strong primacy.13

copy the linklink copied!
Figure ‎1.24. Rank-size distribution of Ethiopia
Figure ‎1.24. Rank-size distribution of Ethiopia

Note: The population of Hawassa has been replaced by CSA estimates due to potential issues with the estimates from Africapolis.

Source: Authors’ calculations using data from Africapolis (SWAC/OECD, 2018[34]).

Ethiopia is not the only country in East Africa with a capital city dominating the national urban system. Figure ‎1.25 compares polycentricity and primacy for nine East African countries. Both Uganda and Sudan are characterised by a high primacy: their largest cities are around nine times bigger than the second-largest cities in their urban systems. As is the case for Ethiopia, these two countries also show a strong tendency towards monocentricity. In contrast, countries such as South Sudan and Rwanda show more polycentric urban systems and comparatively lower primacy levels.

The monocentric structure of Ethiopia’s urban system, as well as its high primacy, suggests that there is scope for other cities to play more important roles within the urban system. Notably, this could release some pressure from Addis Ababa and allow other cities in the urban system to accommodate higher-value economic activities and inhabitants. Indeed, it seems that cities in Ethiopia tend to be relatively small. In 2015, only 16 out of the 78 agglomerations identified in Ethiopia’s National Urban Spatial Development Plan had more than 100 000 inhabitants, while according to Africapolis, only 24 out of 509 agglomerations in Ethiopia fell into this category.

copy the linklink copied!
Figure ‎1.25. Polycentricity and primacy across East African countries
Figure ‎1.25. Polycentricity and primacy across East African countries

Note: Polycentricity is measured through the coefficient estimate resulting from regressing the city ranking over the city’s total population (both variables expressed in logs) for each country. See in ‎Annex 1.A for detailed results. Primacy is measured as the ratio of the population of the largest city in the country over the population of the second-largest city.

Source: Authors’ calculations using data from Africapolis (SWAC/OECD, 2018[34]).

Urban population will double within the next 15 years

Ethiopia is urbanising rapidly. It took Europe 110 years to increase its urban population from 15% in 1800 to 40% in 1910 (AfDB/OECD/UNDP, 2016[36]), whereas Ethiopia will experience this change in half that time.14 By 2025, the urban population is expected to account for 24-29% of Ethiopia’s total population; this number will reach up to 30-40% by 2035 (Figure ‎1.26). Although these percentages remain small compared with other advanced economies and are close to the sub-Saharan African average as of 2019 (39%), they imply a significant change in terms of the total number of urban residents. Within the next 15 years, the urban population is expected to double, increasing by 80-118%.15 In other words, by 2035, city authorities will have to address the needs of 18-29 million new urban dwellers who will demand access to electricity, water, sanitation, housing, education, etc.

copy the linklink copied!
Figure ‎1.26. Urbanisation forecasts, 2015-35
Figure ‎1.26. Urbanisation forecasts, 2015-35

Source: CSA (2013[33]), Ozlu et al. (2015[37]) and MoUDH (2016[35]).

However, Ethiopia is not the only country in the region experiencing a rapid urbanisation process. Ethiopia’s urban population is expected to grow at an annual rate of 4.5% on average between 2015 and 2035 (Figure ‎1.27), but it will be outpaced by countries such as Tanzania, Uganda and Burundi, which are all expected to grow at rates higher than 5%. However, of particular note in this figure is the scale at which Ethiopia’s urbanisation process is taking place. Ethiopia’s urban areas accounted for 19 million people in 2015, representing the largest urban hub in East Africa. By 2050, Ethiopia is expected to be the country with the second-largest number of urban dwellers in East Africa (just after Tanzania), reaching almost 75 million people (Figure ‎1.27). See Box ‎1.3 to compare Ethiopia’s urbanisation process to another key contributor to urbanisation in its respective region, Viet Nam.

copy the linklink copied!
Figure ‎1.27. Urban growth in East Africa, 2015-30
Figure ‎1.27. Urban growth in East Africa, 2015-30

Note: Dotted lines represent the average of the variables in each axis.

Source: Authors’ calculations using data from UNDESA (2018[24]).

copy the linklink copied!
Box ‎1.3. Urban growth in Ethiopia and Viet Nam

How does Ethiopia’s urbanisation process compare with that of Viet Nam? The economies of Ethiopia and Viet Nam both reached growth rates close to 7% in 2018, making them key actors in their respective regions (World Bank, 2019[1]). However, these economies are at different stages of their economic transformation process. Viet Nam’s economy relies on a growing services sector that accounts for the largest share of economic activity, while agriculture accounted for less than 15% of its economy in 2018. Ethiopia’s service sector is evolving, and has overtaken industrial activities, reaching close to 36% in 2018; however, agriculture remains a key sector for the economy, representing almost 34% of its GDP (World Bank, 2019[1]).

Although both countries remain predominantly rural, differences in their economic activity described above further translate into the rural-urban spectrum, in particular when it comes to the types of agglomerations driving the urbanisation process. In 2018, Viet Nam’s urban population represented approximately 35% of its total population, while Ethiopia’s urban population was close to 20% of its total population (UNDESA, 2018[24]). Nevertheless, from 2015 to 2035, cities in Viet Nam with more than 1 million inhabitants will account for 65% of the country’s urban population growth, followed by cities with populations of between 300 000 and 500 000 inhabitants, which are expected to account for 24% of urban population growth. However, in the case of Ethiopia, cities with fewer than 300 000 inhabitants will account for more than 60% of urban population growth, while Addis Ababa (the only city with more than 1 million inhabitants) will only account for 19%.

copy the linklink copied!
Figure ‎1.28. Urban population growth by agglomeration size in Viet Nam and Ethiopia, 2015-35
Figure ‎1.28. Urban population growth by agglomeration size in Viet Nam and Ethiopia, 2015-35

Source: Authors’ calculations using UNDESA (2018[24]).

Ethiopia’s smaller agglomerations are contributing the most to urbanisation and playing a growing role in the economy

Intermediary cities are driving Ethiopia’s urbanisation process. In 2015, Addis Ababa accounted for 19% of the whole urban population in Ethiopia; this share, however, is expected to decrease between 2020 and 2035, reaching close to 11% by 2035 (Figure ‎1.29). Indeed, although Addis Ababa’s population will continue to grow during this time, it will do so at a lower rate compared with small and medium-sized cities. Cities with fewer than 50 000 inhabitants will continue to account for the largest share of Ethiopia’s urban population between 2020 and 2035, going from 51% in 2015 to 40% in 2035. Nevertheless, intermediary or medium-sized cities with populations ranging from 100 000 to 500 000 people will experience the highest average annual growth rates, which are estimated to be 10.21% between 2015 and 2025 and 8.18% between 2025 and 2035 (Figure ‎1.29). They will be followed by small cities with populations ranging from 50 000 to 100 000, which are expected to grow by 6.30% and 5.47% during these two periods, respectively.

The rapid growth of intermediary cities has significant implications for both well-being of rural population and the structural transformation of Ethiopia. However, as will be discussed in Chapter 2, intermediary cities are characterised by a consistent lack of adequate infrastructure, a limited capacity of government officials, and a recurring financing gap. In order to generate economies of scale and limit the cost associated with congestion and pollution, the supply of infrastructure and the provision of public services must catch up with population growth. Moreover, the extent to which intermediary cities contribute to development further depends on the accessibility to rural areas, as well as other agglomerations within the national urban system.

copy the linklink copied!
Figure ‎1.29. Urbanisation growth rates by agglomeration size
Figure ‎1.29. Urbanisation growth rates by agglomeration size

Source: Schmidt et al. (2018[27]).

Ethiopian authorities have actively invested in roads and better telecommunication, which has translated into better accessibility across the country. In fact, as part of its expenditure in pro-poor sectors, the GoE has extensively invested in road development across the country. For example, between 2010/11 and 2015/16, the average annual expenditure on roads reached ETB 22.7 billion, and accounted for 43% of the total pro-poor expenditure (Endale, 2019[38]). In certain regions, however, the population still faces significant challenges in terms of accessing education and health services that are usually provided in large and medium-sized cities. Figure ‎1.30 shows the amount of time needed to reach a populated centre with at least 100 000 inhabitants. At the country level, on average, a populated centre with at least 100 000 inhabitants can be reached within 7 hours. This value, however, hides significant regional differences. In Oromia, the average travel time is 5 hours; while the regions of Gambela and Somali face the longest travel times in order to reach a populated centre with 100 000 inhabitants: 9 and 10 hours on average, respectively.

copy the linklink copied!
Figure ‎1.30. Accessibility to cities of at least 100 000 inhabitants
Figure ‎1.30. Accessibility to cities of at least 100 000 inhabitants

Note: Data on both city populations and accessibility are from 2015.

Accessibility refers to the time (in hours) needed to reach a population of 100 000 inhabitants or more.

Source: Authors’ calculations using Geographic Information Systems (GIS) data from Africapolis (SWAC/OECD, 2018[34]) and Weiss et al. (2018[39]).

Although average time to reach medium-sized and large cities is high, a large share of the population can access these populated centres within 5 hours. Figure ‎1.31 shows the percentages of the population that can reach cities with 100 000 and 250 000 inhabitants by time threshold. Almost 64% of the population can reach an agglomeration with at least 100 000 people within 3 hours. Moreover, 44% of Ethiopians can access a city with at least 250 000 inhabitants within this time frame. In contrast, almost 12% of the population needs at least 7 hours to reach a city with 100 000 people; while 10% of the population needs 9 hours or more to reach a populated centre with at least 250 000 inhabitants.

copy the linklink copied!
Figure ‎1.31. Accessibility to cities with 100 000 and 250 000 inhabitants
Figure ‎1.31. Accessibility to cities with 100 000 and 250 000 inhabitants

Note: Data for 2015.

Source: Authors’ calculations using Geographic Information Systems (GIS) data from Africapolis (SWAC/OECD, 2018[34]) and Weiss et al. (2018[39]).

Urban expansion is making cities less dense, further affecting rural areas

Ethiopia’s rural-urban transformation is taking place by low-density urban growth. Urban population growth has further translated into a consistent expansion of built-up areas across Ethiopia. Built-up areas can be broadly considered as roofed constructions above ground that are used for sheltering humans, animals or materials; the production of economic goods; or the delivery of services (Pesaresi et al., 2019[40]). As cities grow, built-up areas expand. However, in many developing regions, built-up areas are growing at a faster rate than the urban population, making cities less dense (Angel et al., 2016[41]). This commonly leads to increasing transportation and service delivery costs and deterioration in overall environmental quality. Moreover, rapid city expansion can create tensions and even conflicts between urban and rural dwellers.

Ethiopian cities seem to be following this process. Estimates suggest that, between 2000 and 2010, Addis Ababa’s density decreased by an annual average of 3.1%, while its built-up area expanded annually by an average of 5.9% (Angel et al., 2016[41]). However, this phenomenon is not limited to the capital city. A large number of Ethiopian cities are facing a rapid expansion of built-up areas that have outpaced the cities’ population growth rates. Figure ‎1.32 shows the average yearly growth rate of both built-up areas and the population across Ethiopian cities between 2000 and 2014. The line extending from the origin to the top right corner represents a 1:1 growth trajectory, i.e. when both population and built-up area growth are increasing at the same rate. The figure shows that for a significant share of cities, the built-up area is expanding at a faster rate than the urban population. On average, across all agglomerations, the built-up area is growing twice as fast as the population. However, the built-up area expansion process seems to differ across cities of different sizes. Between 2000 and 2014, cities with more than 200 000 inhabitants (in the year 2000), experienced built-up area growth that was, on average, 14% faster than the population growth, while cities with fewer than 200 000 inhabitants, built-up area expansion overtook population growth by an average of 82%.16

copy the linklink copied!
Figure ‎1.32. Urban expansion and population growth, 2000-14
Figure ‎1.32. Urban expansion and population growth, 2000-14

Note: Both built-up area and population growth rates account for yearly average growth between 2000 and 2015.

Source: Authors’ calculations based on the Global Human Settlement Layer Urban Centres Database (GHS-UCDB) produced by Eurostat, Florczyk et al. (2019[42]).

copy the linklink copied!The evolution of well-being in rural areas

There have been important improvements in rural well-being, but there is an increasing gap between rural and urban areas

Rural poverty has been significantly reduced since the late 1990s. Between 2000 and 2016, the rural poverty head count – i.e. the share of the rural population considered poor – decreased from 45% to 25% (Figure ‎1.33). Moreover, the intensity of poverty in rural areas also decreased during this period. The poverty gap across rural areas – i.e. the ratio by which the mean income of the poor falls below the poverty line – decreased from 12% in 2000 to 7.4% in 2016 (Figure ‎1.33).

Despite the success of poverty reduction in rural areas, since the mid-2000s, the gap between urban and rural areas has increased. Poverty in urban areas has been lower since the mid-1990s. In 2000, the poverty head count in urban areas was close to 33%, i.e. almost 14 percentage points lower than in rural areas; by 2005, this difference was four percentage points, but by 2016, the difference had increased by almost 11 percentage points (Figure ‎1.33). A similar process has taken place with the poverty gap. In 2005, the difference between urban and rural areas was less than one percentage point, but by 2016, this difference had reached almost four percentage points.

copy the linklink copied!
Figure ‎1.33. Evolution of poverty in rural and urban areas
Figure ‎1.33. Evolution of poverty in rural and urban areas

Note: Difference refers to the difference between urban and rural areas for either poverty head counts or the poverty gap in each year.

Source: Authors’ calculations using data from NPC (2017[12]).

Beyond monetary poverty, rural areas are particularly affected by limited access to basic services and lower levels of human capital. Table ‎1.3 shows a selected number of indicators describing differences in health, household characteristics and education across urban and rural areas in 2016. Overall, rural areas have lower outcomes in all these areas. At the household level, in 2016, access to electricity remained an area for further improvement, since less than 9% of rural households had access to electricity; in contrast, more than 93% of urban households had access to this service. This is also the case for access to water, since only 6% of rural households had access to water in their premises, compared with 77% of urban households. Moreover, more than half of rural households have to spend 30 minutes or more round trip in order to fetch water.

Ethiopia has maintained a steady progress in achieving Goal 2 of the Millennium Development Goals (MDGs). Indeed, primary school enrolment reached 93% in 2014. However, despite the improvements, enrolment in secondary school only reached 20.2% in the same year (UNDP, 2014[43]). Furthermore, education is a welfare dimension in which striking differences between rural and urban areas prevail: almost half of the population (aged 6 years or older) in rural areas has no education, while in urban areas the illiterate population only represents 19% (Table ‎1.3). There is also a big gap between rural and urban areas in terms of attendance to secondary and higher education institutions. In 2016, the percentage of the Ethiopian population with secondary education living in urban areas was five times higher than that of the population living in rural areas. Moreover, during the same year, less than 1% of the rural population had attended higher education, compared with 17% of the population in urban areas.

In spite of these marked differences across urban and rural areas, there has been significant progress in some of these indicators. Table ‎1.3 also shows the growth rate of the selected group of indicators between 2011 and 2016. Access to water has been an area of significant progression for rural households, increasing by 34% during this period. Access to mobile phones shows a similar progression, growing by almost 30% among rural households. Indeed, in 2016, 47% of rural households had access to a mobile phone. Secondary education has also progressed in rural areas, growing by 17% during the period 2011-16.

copy the linklink copied!
Table ‎1.3. List of selected welfare indicators across urban and rural areas

2016

Change, 2011-16 (%)

Welfare indicators

Rural

Urban

Rural

Urban

Children stunted

39.9

25.4

-2.89

-4.21

Households possessing a mobile telephone

47.2

88.0

29.82

6.18

Households possessing a telephone

0.6

15.2

24.57

-4.36

Households possessing a television

2.1

59.4

13.81

7.13

Households with electricity

8.4

93.3

11.84

1.83

Households with water on the premises

5.6

76.8

33.92

8.79

Households using an improved water source

56.5

97.3

6.26

0.59

Households with water 30 minutes or farther away round trip

52.6

12.6

-3.36

-7.79

Population aged 6 years and over who attended higher education

0.9

17.2

0.00

7.65

Population aged 6 years and over who attended primary education

47.8

41.3

0.86

-3.90

Population aged 6 years and over who attended secondary education

4.3

21.9

16.54

7.16

Population aged 6 years and over with no education

46.9

19.4

-1.70

-2.48

Note: All indicators refer to the share (percentage) of households or people. The general fertility rate accounts for the 3 years preceding the survey expressed as the number of live births per 1 000 females of childbearing age (those between the ages of 15 and 44 years).

Source: ICD (2016[11]). The DHS Program STATcompiler.

Although monetary poverty has decreased, multi-dimensional poverty remains high

Compared to neighbouring countries, Ethiopia shows a high level of deprivation across different welfare dimensions at the national level. Overall, differences in welfare can be measured through the Multidimensional Poverty Index (MPI). The MPI captures differences across dimensions of well-being – health, education and living standards – providing a general picture of the extent of deprivation. The MPI ranges from 0 to 1, with 1 representing a high level of deprivation. The MPI is the product of two partial indices: the headcount ratio and the intensity of poverty.17 In 2016, at the national level, Ethiopia’s MPI sat at 0.48, the highest value across those East African countries for which data are available (Figure ‎1.34). The total incidence of multidimensional poverty stood at 84%, i.e. 84% of the population in Ethiopia is considered multidimensionally poor (Alkire, S et al., 2019[44]). This number contrasts with the monetary poverty estimate from Figure ‎1.36, in which only 24% of the population is considered poor.

Ethiopia also shows a large gap in multidimensional poverty between urban and rural areas. In 2016, rural areas’ MPI stood at 0.55, while in urban areas it stood at 0.16 (Figure ‎1.34). Breaking up this value by incidence and intensity shows a more challenging picture in terms of future development. Indeed, the incidence, i.e. the share of multidimensionally poor people in rural areas, reached almost 92% of the total rural population; in contrast, urban areas’ incidence was close to 16% (Alkire, S et al., 2019[44]). In other words, amongst the almost 74 million people living in rural areas in 2016, close to 68 million were multidimensionally poor. Rural areas are also more affected in terms of intensity, i.e. the average proportion of deprivation experienced. During the same year, poor people in rural areas were deprived in almost 60% of the dimensions, while urban people were deprived in 43% of them. It is important to keep in mind that these estimates account for urban and rural areas at the national level. Estimates across Ethiopian regions presented in Box ‎1.4 show that important disparities prevailed across multiple regions.

copy the linklink copied!
Figure ‎1.34. Multidimensional poverty in selected East African countries, 2016
Figure ‎1.34. Multidimensional poverty in selected East African countries, 2016

Source: Elaborated by the authors using data from Alkire et al. (2019[44]).

Deprivation in rural areas is mainly a consequence of low education levels and low living standards. Annex Figure ‎1.A.4 in ‎Annex 1.A presents the percentage of people who are poor and deprived across the dimensions of welfare considered by the MPI for both urban and rural areas (see also Box ‎1.4 for more on MPI across different Ethiopian regions). This figure also shows the contribution of each dimension to overall poverty. The first thing to notice is that, although the percentage of people who are poor and deprived is systematically higher in all dimensions for rural areas, the contribution of each dimension to overall poverty differs across places. For instance, in rural areas, almost 58% of the population is multidimensionally poor and has a malnourished person at home, compared with 24% in urban areas; almost 60% of the rural population lives in households where no member has completed more than 5 years of schooling, compared with 14% in urban areas; more than 85% of the population in rural areas is multidimensionally poor and lacks electricity, drinking water or adequate sanitation facilities, while at most 35% of the urban population lacks access to one of these services (left side of Annex Figure ‎1.A.4). Nutrition and years of schooling are the dimensions that contribute the most to both urban and rural poverty. However, nutrition has a higher weight in urban areas than in rural areas, contributing 24% and 18%, respectively (right side of Annex Figure ‎1.A.4). Conversely, limited access to electricity, drinking water and assets contributes more to overall poverty in rural areas than in urban areas.

copy the linklink copied!
Box ‎1.4. MPI across Ethiopian regions

Needless to say, there are important differences in terms of MPI across regions. Figure ‎1.35 presents the poverty headcount and intensity for all Ethiopian regions. In 2016, the region experiencing the lowest multidimensional poverty was Addis Ababa (MPI = 0.05), followed by Dire Dawa (0.29), Harari (0.30) and Gambela (0.35). Not surprisingly, these regions are also the ones with the highest shares of urbanisation. There are, however, striking differences in the extent and intensity of multidimensional poverty across regions. In Addis Ababa, only 15% of the population is considered multidimensionally poor, whereas in Dire Dawa and Harari, this share is higher than 50%, and in Gambela it reaches up to 70%. In spite of the latter, the intensity of multidimensional poverty in Gambela (49) is lower than in Dire Dawa (57) and Harari (54). In contrast, Somali and Afar are the regions with the highest MPI (0.57). In both regions, close to 90% of the population experiences multidimensional poverty.

copy the linklink copied!
Figure ‎1.35. Headcount and intensity of MPI across Ethiopian regions, 2016
Figure ‎1.35. Headcount and intensity of MPI across Ethiopian regions, 2016

Source: Elaborated by the authors using data from Alkire et al. (2019[44]).

The reduction in multidimensional poverty is advancing faster in urban areas. Between 2011 and 2016, multidimensional poverty was reduced in both urban and rural areas. However, the MPI in urban areas decreased from 0.23 in 2011 to 0.16 in 2016; i.e. the MPI experienced an average annual decrease of 7% during this period. The overall rural MPI went from 0.61 to 0.54, representing an average annual decrease of 2%. For both urban and rural areas, this change was propelled by a decrease in the proportion of people who ceased to be multidimensional poor (the incidence of poverty) in each area, not by the intensity of poverty (Figure ‎1.36). In urban areas, the proportion of people considered multidimensionally poor decreased from 50% to 37% between 2011 and 2016. In rural areas, this value only decreased by five percentage points, going from 97% to 92%. For both areas, the intensity of poverty experienced an average annual decrease of approximately 1.3%.

copy the linklink copied!
Figure ‎1.36. Annualised change in multidimensional poverty, 2011-16
Figure ‎1.36. Annualised change in multidimensional poverty, 2011-16

Source: Elaborated by the authors using data from Alkire et al. (2019[44]).

Where does Ethiopia stand in terms of multidimensional poverty reduction compared with other countries? Ethiopia fares very well when it comes to urban areas, but there is room for improvement among rural populations. Figure ‎1.37 shows the value of Ethiopia’s MPIs in urban and rural areas for 2011 (vertical axes) and their relative change between 2011 and 2016 (horizontal axes); this figure compares these estimates with those of Bangladesh, Cambodia, the Democratic Republic of the Congo, Haiti, India and Peru.18 Ethiopia and Bangladesh started at the same level and achieved similar decreases in multidimensional poverty among urban areas, of 7.2% and 6.4% on average per year, respectively. Among the seven included countries, Ethiopia achieved the third-largest reduction in MPI in urban areas. When it comes to rural areas, Ethiopia started with the highest MPI of the group (0.61), followed by the Democratic Republic of the Congo (0.54). Both countries achieved similar decreases in MPI, of 2.1% and 2.3% on average per year, respectively. Ethiopia outperformed only India in reducing MPI across rural areas.

copy the linklink copied!
Figure ‎1.37. Changes in MPI across selected countries
Figure ‎1.37. Changes in MPI across selected countries

Note: These figures are based on those countries for which the MPI has been estimated at two periods of time. The vertical axis accounts for the value of the MPI during the first survey. The horizontal axis accounts for the annual relative change in the MPIs between the first and second time period. For all countries, the change has been negative. Thus, these figures refer to the decrease in the MPI across urban and rural areas in positive terms. The periods for which the MPI has been estimated differ across countries, as follows: Bangladesh: 2004-14; Cambodia: 2010-14; Democratic Republic of the Congo: 2007-13/14; Ethiopia: 2011-16; Haiti: 2012-16/17; India: 2005/06-15/16; Peru: 2006-12.

Source: Elaborated by the authors using data from Alkire et al. (2019[44]).

copy the linklink copied!Conclusions

Ethiopia is a fast-growing economy with strong potential for development. Addressing rural development will be key to achieving national development goals and, more importantly, for improving the well-being of a large share of the population. The three transformations described in this chapter will bring important changes in both the economy and the society that will further affect rural areas. Rural development will require policy actions that account for these transformations through an integrated and holistic approach.

Ethiopia has implemented a series of strategies that led to reduction in poverty and promoted economic development across rural areas in the country. However, rural areas are still confronted with limited development opportunities and are increasingly lagging behind urban areas. As will be discussed in Chapter 3, the current policy framework for rural development has a strong sectoral focus that do not account for the roles and potential of urban areas. Indeed, although urban development has received increasing attention in the last 5-10 years, rural and urban policies appear as mutually exclusive interventions. Addressing the challenges and reaping the opportunities resulting from the tree mayor transformations experienced by Ethiopia will require going beyond the rural-urban divide and create strategies that create policy complementarities between urban and rural areas. To this end, additional attention should be paid to the roles of intermediary cities for rural development (see Chapter 2).

copy the linklink copied!Annex 1.A. Additional figures and tables
copy the linklink copied!
Annex Figure  ‎1.A.1. Demographic transitions in Ethiopia and Viet Nam
Annex Figure  ‎1.A.1. Demographic transitions in Ethiopia and Viet Nam

Source: UNDESA (2018[24]).

copy the linklink copied!
Annex Figure 1.A.2. Population growth rates in selected East African countries and region of Sub-Saharan Africa
Annex Figure 1.A.2. Population growth rates in selected East African countries and region of Sub-Saharan Africa

Source: UNDESA (2018[24]).

copy the linklink copied!
Annex Figure 1.A.3. Total dependency ratio in selected countries
Annex Figure 1.A.3. Total dependency ratio in selected countries

Note: Total dependency ratio = ((age 0-14 + age 65+) / age 15-64). De facto population as of 1 July of the year indicated.

Source: UNDESA (2018[24]).

copy the linklink copied!
Annex Table 1.A.1. Rank-size regression results for East African countries

Country

β

Standard Error

t

P>t

R-squared

Eritrea

-1.07786

0.163242

-6.6

0

0.9248

Ethiopia

-0.77527

0.11648

-6.66

0

0.7512

Kenya

-1.04327

0.187172

-5.57

0.001

0.9012

Rwanda

-1.27259

0.170013

-7.49

0

0.9158

Sudan

-1.08058

0.323564

-3.34

0.01

0.7734

Somalia

-1.29749

0.079386

-16.34

0

0.9599

South Sudan

-0.68496

0.022274

-30.75

0

0.9823

Tanzania

-1.11059

0.207861

-5.34

0.001

0.8929

Uganda

-1.0349

0.31318

-3.3

0.011

0.7762

Note: Estimates from Ordinary Least Square (OLS) regression with robust standard errors considering the ten largest agglomerations in each country.

Source: Authors’ calculations using data from Africapolis (2018[34]).

copy the linklink copied!
Annex Table 1.A.2. Access to cities with at least 100 000 inhabitants, by region

Regions

1 hour

2 hours

3 hours

4 hours

5 hours

6 hours

7 hours

8 hours

9 hours

≥10 hours

Total

Addis Ababa

100%

0%

0%

0%

0%

0%

0%

0%

0%

0%

100%

Afar

1%

9%

10%

13%

9%

13%

11%

12%

5%

16%

100%

Amhara

14%

15%

16%

19%

13%

8%

6%

4%

2%

4%

100%

Benishangul-Gumuz

0%

1%

9%

27%

19%

15%

10%

7%

6%

6%

100%

Dire Dawa

94%

3%

0%

0%

2%

0%

0%

0%

0%

0%

100%

Gambela

0%

0%

14%

16%

6%

6%

18%

18%

3%

18%

100%

Harari

100%

0%

0%

0%

0%

0%

0%

0%

0%

0%

100%

Oromia

25%

26%

21%

11%

9%

4%

2%

1%

1%

1%

100%

SNNPR

33%

29%

19%

8%

3%

2%

1%

1%

1%

3%

100%

Somali

6%

7%

5%

8%

4%

7%

7%

8%

9%

40%

100%

Tigray

15%

28%

23%

14%

7%

5%

4%

1%

0%

2%

100%

Note: Percentages may not sum to 100% due to rounding.

Source: Authors’ calculations using Geographic Information Systems (GIS) and data from Africapolis (SWAC/OECD, 2018[45]) and Weiss et al. (2018[39]).

copy the linklink copied!
Annex Table 1.A.3. Alternative typologies for classifying settlements in Ethiopia

Typology

Share

Extended typology

Share

Urban centres

28.3%

Urban – close

24.8%

Urban – remote

3.5%

Suburbs

0.1%

Suburbs – close

0.1%

Towns

6.3%

Towns – close

4.3%

Towns – remote

2.0%

Villages

4.0%

Villages – close

2.7%

Villages – remote

1.3%

Rural

61.3%

Rural – close

31.8%

Rural – remote

29.5%

Total

100.0%

100.0%

Note: “Close” and “remote” define those settlements where inhabitants can reach a city with at least 100 000 population in less than or more than 3 hours, respectively.

Source: Authors’ calculations using data from (Pesaresi et al., 2019[40]).

copy the linklink copied!
Annex Table 1.A.4. ‎Types of internal migration in Ethiopia (%)

March 1999

March 2005

June 2013

Rural-rural

37.6

46.0

34.5

Rural-urban

23.5

24.3

32.5

Urban-rural

15.7

12.1

11.6

Urban-urban

23.2

17.7

21.3

Source: Adapted from (CSA, 2014[26]).

copy the linklink copied!
Annex Table 1.A.5.  Components of multidimensional poverty by type of area

Area

MPI

Incidence (%)

Intensity (%)

Vulnerable

Severe

Urban

0.16

36.83

43.47

18.14

10.89

Rural

0.55

91.82

59.61

7.23

70.49

National

0.49

83.50

58.54

8.88

61.48

Note: Vulnerable: the proportion of people who experience deprivation across ‘’20 to 33.3% of the weighted indicators’’; Severe: proportion of people experiencing deprivation in 50% or more of the poverty dimensions

Source: Elaborated by the authors using data from Alkire et al. (2019[44]).

copy the linklink copied!
Annex Figure 1.A.4. Multidimensional poverty by type of deprivation
Annex Figure 1.A.4. Multidimensional poverty by type of deprivation

Source: Elaborated by the authors using data from Alkire et al. (2019[44]).

References

[36] AfDB/OECD/UNDP (2016), African Economic Outlook 2016: Sustainable Cities and Structural Transformation, OECD Publishing, Paris, https://dx.doi.org/10.1787/aeo-2016-en.

[44] Alkire, S et al. (2019), “The Global Multidimensional Poverty Index (MPI) 2019”, OPHI MPI Methodological Notes 47, Oxford Poverty and Human Development Initiative, University of Oxford, https://ophi.org.uk/wp-content/uploads/OPHI_MPI_MN_47_2019_vs2.pdf.

[48] Alkire, S. et al. (2017), “Multidimensional Poverty Reduction Among Countries in Sub-Saharan Africa”, Forum for Social Economics, Vol. 46/2, pp. 178-191, http://dx.doi.org/10.1080/07360932.2017.1310123.

[41] Angel, S. et al. (2016), Atlas of Urban Expansion, Lincoln Institute, https://www.lincolninst.edu/publications/other/atlas-urban-expansion-2016-edition.

[4] AUC/OECD (2019), Africa’s Development Dynamics 2019: Achieving Productive Transformation, African Union Commission, Addis Ababa/OECD Publishing, Paris, https://dx.doi.org/10.1787/c1cd7de0-en.

[13] AUC/OECD (2018), Africa’s Development Dynamics 2018: Growth, Jobs and Inequalities, African Union Commission, Addis Ababa/OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264302501-en.

[21] Bachewe, F. et al. (2018), “Agricultural Transformation in Africa? Assessing the Evidence in Ethiopia”, World Development, Vol. 105, pp. 286-298, http://dx.doi.org/10.1016/j.worlddev.2017.05.041.

[6] Bachewe, F. et al. (2016), “Non-farm income and labor markets in rural Ethiopia”, No. ESSP Working Paper 90, International Food Policy Research Institute (IFPRI) and Ethiopian Development Research Institute (EDRI), Washington, D.C. and Addis Ababa, Ethiopia, http://ebrary.ifpri.org/cdm/ref/collection/p15738coll2/id/130375.

[32] Beegle et al. (2011), “Migration and Economic Mobility in Tanzania: Evidence From a Tracking Survey”, The Review of Economics and Statistics, Vol. 93/3, pp. 1010–33, http://documents.worldbank.org/curated/en/607811468305945623/pdf/946700JRN0Box30LIC00160rest0a000105.pdf.

[9] Beegle, K. and L. Christiaensen (2019), Accelerating Poverty Reduction in Africa, World Bank, Washington, DC, https://openknowledge.worldbank.org/handle/10986/32354.

[47] Brezzi, M. and P. Veneri (2015), “Assessing Polycentric Urban Systems in the OECD: Country, Regional and Metropolitan Perspectives”, European Planning Studies, Vol. 23/6, pp. 1128-1145, http://dx.doi.org/10.1080/09654313.2014.905005.

[30] Bundervoet, T. (2018), Internal Migration in Ethiopia : Evidence from a Quantitative and Qualitative Research Study, World Bank, Washington, DC., http://documents.worldbank.org/curated/en/428111562239161418/Internal-Migration-in-Ethiopia-Evidence-from-a-Quantitative-and-Qualitative-Research-Study.

[16] CEQ (2019), Commitment to Equity Institute Data centre on Fiscal Redistribution (database), Comiitment to Equity (CEQ) Institute, https://commitmentoequity.org/datacenter (accessed on 22 December 2019).

[20] CSA (2018), “Agricultural Sample Survey (2018)”, Central Statistical Agency, Addis Ababa.

[8] CSA (2017), Integrated Surveys on Agriculture Ethiopia Socioeconomic Survey (ESS), Central Statistical Agency and Living Standards Measurement Study (LSMS), World Bank.

[26] CSA (2014), Analytical Report on The 2013 National Labour Force Suervey Statistical Bulletin, Central Statistical Agency, Addis Ababa, http://www.csa.gov.et/survey-report/category/34-nlfs-2013.

[10] CSA (2013), Labour Force Survey 2013, Central Statistical Agency, http://catalog.ihsn.org/index.php/catalog/5870.

[33] CSA (2013), Population Projection of Ethiopia for All Regions at Wereda Level from 2014 – 2017, Central Statistical Agency, Addis Ababa, https://www.scribd.com/document/343869975/Population-Projection-At-Wereda-Level-from-2014-2017-pdf.

[7] Davis, B., S. Giuseppe and A. Zezza (2014), Income Diversification Patterns in Rural Sub-Saharan Africa Reassessing the Evidence, http://econ.worldbank.org.

[31] Dominiko, T. (2016), “A Qualitative Study on Internal Migrants in Ethiopia: Causes, Experiences, Perceptions, Challenges and Future Life Orientations”, Journal of Social Economics, Research Academy of Social Sciences,, Vol. 3/2, pp. 54-72, https://ideas.repec.org/a/rss/jnljse/v3i2p2.html.

[38] Endale, E. (2019), “Financing social protection in Ethiopia: Long term perspective”, OECD, Paris, https://doi.org/10.1787/24140929.

[42] Florczyk et al. (2019), GHS Urban Centre Database 2015, multitemporal and multidimensional attributes, R2019A, European Comission, Joint Research Centre (JRC), https://data.jrc.ec.europa.eu/dataset/53473144-b88c-44bc-b4a3-4583ed1f547e (accessed on 22 December 2019).

[28] Harris, J. and M. Todaro (1970), “Migration, Unemployment and Development: A Two-Sector Analysis”, The American Economic Review, Vol. 60/1, pp. 126-142, https://www.jstor.org/stable/pdf/1807860.pdf?refreqid=excelsior%3A7a3ee5d2d8a1b95021d13b575538094b.

[22] Hill, R. and E. Tsehaye (2015), Ethiopia Poverty Assessment, World Bank Group, Washington, DC, http://documents.worldbank.org/curated/en/131011468247457565/Ethiopia-Poverty-assessment.

[11] ICD (2016), Demographic and Health Survey, funded by USAID, https://www.statcompiler.com/en/ (accessed on 22 December 2019).

[15] IMF (2015), IMF Country Report: The Federal Republic of Ethiopia, Selected Issues, International Monetary Fund (IMF), Washington DC, https://www.imf.org/en/Publications/CR/Issues/2016/12/31/Federal-Democratic-Republic-of-Ethiopia-Selected-Issues-43427.

[25] Mason, A. et al. (2017), “Support Ratios and Demographic Dividends: Estimates for the World”, Technical Paper No. 2017/1, United Nations Population Division, New York, https://www.un.org/en/development/desa/population/publications/pdf/technical/TP2017-1.pdf.

[46] Meijers, E. and S. Sandberg (2008), “Reducing regional disparities by means of polycentric development: Panacea or Placebo”, Scienze Regionali, Vol. 7/2, pp. 71–96, https://www.researchgate.net/publication/241758007_Reducing_Regional_Disparities_by_Means_of_Polycentric_Development_Panacea_or_Placebo.

[18] Mo Ibrahim Foundation (2016), A Decade of African Governance 2006-2015: Ethiopia Insights, Mo Ibrahim Foundation, London, http://s.mo.ibrahim.foundation/u/2017/03/08200153/Ethiopia-Insights-2016-IIAG.pdf.

[2] Moller, L. (2015), “Ethiopia’s Great Run: The Growth Acceleration and How to Pace It”, Washington, DC, World Bank Group, p. 164, http://documents.worldbank.org/curated/en/693561467988949839/Ethiopia-s-great-run-the-growth-acceleration-and-how-to-pace-it.

[35] MoUDH (2016), National Urban Development Spatial Plan, Ministry of Urban Development & Housing, Addis Abeba, Ethiopia.

[5] Nagler, P. and W. Naudé (2014), Non-Farm Entrepreneurship in Rural Africa: Patterns and Determinants.

[12] NPC (2017), Ethiopia’s Progress Towards Eradicating Poverty: An Interim Report on 2015/16 Poverty Analysis Study, National Planning Commission, Addis Ababa, Ethiopia.

[37] Ozlu, M. et al. (2015), Ethiopia - Urbanization review : urban institutions for a middle-income Ethiopia, World Bank Group, Washington, D.C., http://documents.worldbank.org/curated/en/543201468000586809/Ethiopia-Urbanization-review-urban-institutions-for-a-middle-income-Ethiopia.

[40] Pesaresi, M. et al. (2019), GHS settlement grid, updated and refined REGIO model 2014 in application to GHS-BUILT R2018A and GHS-POP R2019A, multitemporal (1975-1990-2000-2015), R2019A., European Commission, Joint Research Centre (JRC), http://dx.doi.org/10.2905/42E8BE89-54FF-464E-BE7B-BF9E64DA5218.

[29] Rosenzweig, M. and O. Stark (1989), “Consumption Smoothing, Migration, and Marriage: Evidence from Rural India”, Journal of Political Economy, Vol. 97/4, pp. 905-926.

[3] Schmidt, E. and F. Bekele (2016), “Rural youth and employment in Ethiopia”, International Food Policy Reserach (IFPRI), Washington DC December, http://ebrary.ifpri.org/cdm/singleitem/collection/p15738coll2/id/131006.

[27] Schmidt, E. et al. (2018), Ethiopia’s Spatial and Structural Transformation Public Policy and Drivers of Change, International Food Policy Reserach (IFPRI), Washington DC, http://ebrary.ifpri.org/utils/getfile/collection/p15738coll2/id/132728/filename/132940.pdf.

[19] Seyoum Taffesse, A. (2019), “The Transformation of Smallholder Crop Production in Ethiopia, 1994–2016”, in The Oxford Handbook of the Ethiopian Economy, Oxford University Press, http://dx.doi.org/10.1093/oxfordhb/9780198814986.013.24.

[34] SWAC/OECD (2018), Africapolis, http://www.africapolis.org/ (accessed on 22 December 2019).

[45] SWAC/OECD (2018), Africapolis, http://www.africapolis.org/ (accessed 22-12-2019).

[24] UNDESA (2018), World Urbanization Prospects: The 2018 Revision, Online Edition., United Nations, Department of Economic and Social Affairs, Population Division, https://www.un.org/development/desa/publications/.

[23] UNDESA (2017), World Population Prospects: The 2017 Revision, DVD Edition., United Nations, Department of Economic and Social Affairs, Population Division.

[43] UNDP (2014), “Accelerating Inclusive Growth for Sustainable Human Development in Ethiopia”, National Human Develoment Report (2014) Ethiopia, pp. 1-97, http://hdr.undp.org/sites/default/files/nhdr2015-ethiopia-en.pdf.

[14] UNDP (2014), National Human Development Report - Ethiopia, United Nations Development Programme, Addis Ababa, Ethiopia, http://hdr.undp.org/sites/default/files/nhdr2015-ethiopia-en.pdf.

[39] Weiss, D. et al. (2018), “A global map of travel time to cities to assess inequalities in accessibility in 2015”, Nature, Vol. 553/7688, pp. 333-336, http://dx.doi.org/10.1038/nature25181.

[1] World Bank (2019), World Development Indicators, https://datacatalog.worldbank.org/dataset/world-development-indicators (accessed on 22 December 2019).

[17] World Bank (2015), World Bank Worldwide Governance Indicators (database), https://www.quandl.com/data/WWGI-World-Bank-Worldwide-Governance-Indicators (accessed on 22 December 2019).

Notes

← 1. Due to methodological differences that allow for cross-country comparison, the estimates from the Africa Sector Database (ASD) may not necessarily match the ones from Ethiopia’s Labour Force Survey. However, contrasting these estimates and comparing across countries allows us to have a better understanding of Ethiopia’s ongoing structural transformation.

← 2. This study focused on the regions of Tigray, Amhara, Oromiya, and SNNP.

← 3. Based on soil fertility and slope measures.

← 4. According to the Central Statistical Agency CSA (2014[26]), the informal sector is considered as a group of production units, such as household enterprises or unincorporated enterprises owned by households. People engaged in subsistence farming and those who work in private households are exempted. Government employees, government development organisation employees, non-governmental organisation (NGO) employees, and members of cooperatives were considered part of the formal sector. Employers, private organisation employees, the self-employed, and apprentices were asked about whether the business/enterprise they were engaged in fulfilled the following criteria: a) the enterprise has an accounting book; b) the enterprise has a licence; or c) the product/service of the enterprise is marketable. Employed persons who satisfy at least one of the above conditions (‘a’ or ‘b’) were considered as working in the formal sector. For those who did not fulfil either ‘a’ or ‘b’ but did fulfil ‘c’, the activity was considered as informal. Those who did not know the main activity/business/enterprise, with respect to the criteria above, were considered as “not identified”.

← 5. The percentage increase was calculated by the authors as follows: [(Yield c,t+5 / Yield c, t)-1]*100.; where C stands for crop and t for the years 2000, 2005 and 2010.

← 6. The remaining share was mainly produced by commercial farms.

← 7. In order to define whether the person interviewed was a migrant or not, the Labour Force Survey asked a different question depending on the location where the survey took place. For urban areas, the question asked referred to the number of years the person had been continuously residing in the town or city, while in rural areas, it referred to the number of years the person had been continuously residing in the rural part of their woreda of enumeration (CSA, 2013[10]) .

← 8. Whether those cities are just a departure point for migration or a transition point requires further study.

← 9. Calculations made by the authors using data from UNDESA (2018[24]).

← 10. The Africapolis dataset estimates urban growth across various agglomerations size in Africa. Africapolis is built by detecting continuously built-up areas using satellite images, and then defining a continuously built-up area as an area with less than 200 m between buildings and constructions. The boundaries of the identified area are then overlaid with a map of the smallest officially defined local administrative units in the country. The populations of all local administrative units that are covered by at least 50% of the identified continuously built-up area are added up and counted as the population of the urban agglomeration. The minimum population threshold used by Africapolis to consider an agglomeration as urban is 10 000 inhabitants. The agglomeration takes the name of the local administrative unit that is highest in the administrative hierarchy and/or population. See https://www.africapolis.org/data for detailed information about the methodology.

← 11. However, it is not adequate to think about polycentricity and monocentricity in a discrete way; instead, it is better to consider it as a gradient from low to high population concentration. The extent to which a country or a region is considered polycentric or monocentric is commonly captured by analysing the relationship between cities’ populations and their corresponding rankings within an urban system. This is referred to as the rank-size distribution.

← 12. Polycentricity is measured as the coefficient (β) resulting from regressing city rank over city population under the following specification:

log r a n k i = α + β   l o g ( p o p u l a t i o n i )

When the coefficient β is higher than 1, the system tends to be polycentric; when it is lower than 1, it tends to be monocentric. The slope is estimated using the top ten largest cities in order to avoid bias linked to the large number of small towns (Meijers and Sandberg, 2008[46]). This exercise builds on Africapolis data that do not rely on administrative borders, which allows better separation of the nodes of the urban system. In particular, this avoids the mistake of capturing places that are part of a single integrated area, such as a municipality (Brezzi and Veneri, 2015[47]).

← 13. Urban primacy refers to the size of the largest city with respect to other agglomerations in the country. In this case, primacy is measured as the ratio of the population in the largest city over the population of the second-largest city in the country.

← 14. Ethiopia’s share of urban population is expected to increase from 14% to 40% between 2005 and 2050.

← 15. Authors’ calculations using data estimates from the CSA (2013[33]), Ozlu et al. (2015[37]) and MoUDH (2016[35]).

← 16. Between 2000 and 2014, the mean value of the average yearly population growth and built-up area growth across all cities was 2.54% and 4.64%, respectively. The built-up area growth rate is thus 82% higher than the population growth rate.

← 17. The headcount ratio is the share of poor people in the population; the intensity shows how much deprivation poor people experience on average. See (Alkire et al., 2017[48]) for additional information on the MPI.

← 18. These countries were selected due to the availability of MPI estimates across time. Please note that periods between MPI estimates differ across countries.

Metadata, Legal and Rights

This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area. Extracts from publications may be subject to additional disclaimers, which are set out in the complete version of the publication, available at the link provided.

https://doi.org/10.1787/a325a658-en

© OECD and PSI 2020

The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at http://www.oecd.org/termsandconditions.