## The Future of Work in Sub-Saharan Africa

## Source details

**Canonical URL:** [The Future of Work in Sub-Saharan Africa](https://www.imf.org/-/media/files/publications/dp/2018/48333-dp1818-the-future-of-work-in-ssa.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/dp/2018/48333-dp1818-the-future-of-work-in-ssa.pdf.md)
- [Structured JSON version](/-/media/files/publications/dp/2018/48333-dp1818-the-future-of-work-in-ssa.pdf.json)

---

### Introduction — context, scope, and approach
- Examines challenges and opportunities of the Fourth Industrial Revolution for sub-Saharan Africa as the region seeks to:
  - create jobs for a booming population,
  - tackle adverse effects of climate change, and
  - face an external environment that may become less supportive.
- Approach combines history, economic modeling, empirical evidence, and scenario analysis.
- Paper structure:
  - Chapters 1–3: labor market developments, Fourth Industrial Revolution impacts, and other big trends (demographics and urbanization, climate change, global economic integration).
  - Chapter 4: three realistic scenarios for the region over the next two decades.
  - Chapter 5: policy ideas to generate the needed jobs.

### Demographics and labor force projections
- Population expected to rise from "1 billion today to 1.7 billion by 2040."
- The working-age population in the region will, on average, experience "a net increase of 20 million per year over the next two decades."
- Sub-Saharan Africa population projections also presented as "1.7 billion by 2040—a 70 percent increase—and to close to 4 billion by the end of the century."
- In comparison, the population of the rest of the world is expected to increase by "about 16 percent between now and 2040."
- Heterogeneity in fertility rates: Mauritius "less than 2"; Niger "over 7."
- Uncertainty in population projections: alternative projections are "as much as 300 million by 2040 and 1.6 billion by 2100" lower than United Nations projections.

### Recent employment trends and heterogeneity
- Since 2000, sub-Saharan Africa has been able to add "an average of 9 million jobs per year."
- Employment-to-population ratio increased from "63 percent in 2000 to 64 percent in 2017."
- Sectoral job creation:
  - Most new jobs were created in sectors with low productivity levels, such as subsistence agriculture and low value-added services.
  - Sub-Saharan Africa has the lowest share of employment in manufacturing in the world.
  - Reallocation toward services did not predominantly favor "modern" services; during the 2000s, the share of modern services in sub-Saharan Africa’s exports of services declined.
- Employment type:
  - "Six of the nine million jobs created annually on average since 2000 have gone to own-account and contributing family workers."
  - "Only 2.6 million have gone to employees being paid for their work."
- Cross-country variation:
  - Employment-to-population rate declined by "13 percentage points in Burkina Faso" while population grew "3.2 percent since 2000" and employment expanded at "2 percent."
  - In Burundi, Côte d’Ivoire, Kenya, Lesotho, Mozambique, and Sierra Leone, population outgrew job creation.
  - In Angola, Botswana, Congo, Ghana, Mali, Senegal, and Zimbabwe, job creation largely outpaced population.
  - South Africa and Eswatini: employment-to-population ratios remained broadly stable at "below 40 percent", less than half of the employment rates in Madagascar, Rwanda, and Tanzania.

### The Fourth Industrial Revolution — definition, risks, and opportunities
- Characterized by breakthroughs in artificial intelligence, machine learning, robotics, and 3-D printing.
- Machine learning and artificial intelligence expand the range of automatable tasks beyond routines; Polanyi’s paradox has limited automation historically but is being broken.
- Fundamental uncertainty: technology may complement workers (boosting convergence) or substitute for workers (causing divergence).
- Empirical estimates of jobs at risk vary widely:
  - Advanced economies: estimates range from 7 percent to 60 percent of jobs at risk.
    - McKinsey Global Institute (2017): "40 percent to 55 percent" of jobs in advanced economies may be at risk.
    - World Bank (2016): "almost 60 percent" of employment at risk in OECD economies.
    - Arntz, Gregory, and Zierahn (2016): "7 percent to 15 percent" of OECD jobs at risk.
  - Sub-Saharan Africa estimates for individual countries range from "40 percent to 60 percent" in some studies (McKinsey Global Institute 2017; World Bank 2016).

### Digital infrastructure and affordability
- Internet penetration in the region remains the lowest in the world—less than half the global average.
- Only Kenya, Nigeria, and Seychelles have penetration levels close to 50 percent, slightly above the global average.
- Cost of fixed broadband is the highest in sub-Saharan Africa compared to other regions.
- Mobile broadband costs in US dollar terms are comparable to other regions, but when scaled by gross national income, sub-Saharan Africa has the least affordable mobile broadband.

### Theoretical model — robots, labor, and convergence/divergence
- Two-region model: advanced economy (ROW) and low-income region (SSA); production uses traditional capital, labor, and "robots" (broadly automation, machine learning, AI).
- Calibration: per capita GDP in advanced economies is 15 times that in sub-Saharan Africa.
- Robots and labor can be complements or substitutes:
  - If complements:
    - Increase in robot productivity can raise per capita GDP more in SSA than in ROW (convergence).
    - Labor share increases if robots complement labor; wage increases can exceed increases in robot stock.
  - If substitutes:
    - Increase in robot productivity raises per capita GDP more in ROW than in SSA (divergence).
    - Labor share falls in both regions; increase in real wages is stronger in ROW, but robot stock rises more, reducing labor share and raising inequality.
- Robot-to-labor ratio scaling: Z/L ∝ w^σ, with examples:
  - Robot-to-labor ratio in ROW is 15 times higher than SSA when robots and labor are complements.
  - Robot-to-labor ratio in ROW is almost 200 times higher than SSA when robots substitute for labor.
- Policy lever: economy-wide productivity growth (infrastructure, skills, institutions) can offset automation headwinds.
  - Historic productivity growth between 2000 and 2014: "0.9 percent" on average would not be enough to close the gap.
  - Potential scenario: trade integration and efficiency gains could reduce the productivity gap by half in two decades, from a factor of 1 to 15 to a factor of 1 to 7.

### Export vulnerability to automation — indices and findings
- Constructed two export vulnerability indices mapping occupation automatability to industries to export goods; used two occupation-vulnerability mappings:
  - Frey and Osborne (2017)–based index.
  - Brynjolfsson, Mitchell, and Rock (2018)–based index.
- Divergent results:
  - Frey and Osborne–based index suggests sub-Saharan Africa’s exports are relatively more vulnerable to automation than exports of more developed countries.
  - Brynjolfsson, Mitchell, and Rock–based index suggests sub-Saharan Africa’s exports are relatively less vulnerable compared with advanced economies.
- Reason for divergence:
  - Frey and Osborne emphasize replacement of routine jobs; rate crop production and low-end manufacturing as more susceptible—these are important in SSA noncommodity exports.
  - Brynjolfsson, Mitchell, and Rock emphasize machine learning replacing nonroutine cognitive tasks; rate sophisticated electronics and machinery as more susceptible—prominent in advanced economies’ exports.
- Areas of agreement:
  - Both indices consider apparel and textile manufacturing susceptible to automation.
  - Both indices consider mining less vulnerable in terms of export vulnerability (but mining jobs can still be subject to automation, e.g., drone mining in South Africa).
- Implication: the effect on SSA depends on which tasks and industries technological progress affects and on relative wages and technology prices; export diversification can reduce exposure.

### Other major trends affecting future of work
- Three interacting big trends with great uncertainty:
  - Demographic expansion and urbanization.
  - Severity of climate change impact.
  - Global economic integration.
- Jobs and education:
  - Sub-Saharan Africa must create "16 to 20 million new jobs on average every year."
  - By 2030 the region "will contribute half of the annual increase in the global working age population."
  - Primary completion improved; secondary enrollment rates remain "well below other regions."
- Urbanization:
  - Urban population grew by "about 4.4 percent annually over the last 50 years" and was "close to 400 million at the end of 2015."
  - Average urbanization rate now "exceeds 40 percent."
  - Cities are often congested but not economically dense; many are "consumption cities" with employment concentrated in informal, nontradable services.
- Climate change:
  - Temperatures in Africa have increased "by about 0.5°C over the last 50 to 100 years"; global average increase is "about 1°C."
  - IPCC projections range from "0.2°C per decade" to "0.5°C per decade."
  - IMF (2017) estimates a "1°C increase in temperature" causes a fall in per capita GDP growth of "1.2 percentage points in the same year" for the median low-income country and reduces per capita output by "1.5 percentage points seven years after the shock."
  - Urbanization rate increases by "0.45 percent with a reduction in rainfall of 1 percent."
- Global integration:
  - Nonofficial net private capital flows increased from "$4 billion in the 1980s and 1990s to close to $60 billion in 2017."
  - Intraregional trade rose from "6 percent of total exports in 1980 to 20 percent of GDP in 2016," but overall integration remains about half that of other emerging and developing countries.
  - Risks: protectionist pressures and geopolitical shifts could slow or reverse integration.

### Scenario analysis — methodology and three archetypes
- Method: inductive scenario analysis under Knightian uncertainty to generate plausible narrative scenarios and compare policy implications.
- Three scenarios: "Africa Arisen", "Africa for Africa", and "Africa Adrift".
- Driver summary across scenarios:
  - Technology:
    - Africa Arisen: increases productivity and empowers workers.
    - Africa for Africa: allows for leapfrogging with limited negative impact on jobs thanks to lower wages.
    - Africa Adrift: substitutes for workers, wiping out jobs across the region.
  - Global economic integration:
    - Africa Arisen: deepens.
    - Africa for Africa: reversed globally but offset by strengthened regional integration.
    - Africa Adrift: volatility and trade wars.
  - Climate change:
    - Africa Arisen: largely mitigated by technological advances.
    - Africa for Africa: negative impact on growth.
    - Africa Adrift: devastating, leading to mass migration and obsolete infrastructure.
- Scenario vignettes highlight divergent outcomes for economic structure, world of work, skills required, and governance responses.

### Scenario highlights — selected outcomes
- Africa Arisen:
  - Technology complements labor, productivity rises, services-led employment grows, deepening global integration, expanded digital connectivity, and gradual expansion of social safety nets.
- Africa for Africa:
  - Leapfrogging via cheap technologies and regional integration (e.g., Continental Free Trade Agreement), limited foreign investment, strong regional entrepreneurship, high informality, and weak social safety nets.
- Africa Adrift:
  - Technology substitutes for labor, reshoring of manufacturing, collapse of global cooperation, severe climate impacts (cities run out of water), rising informality and poverty traps, frequent crises, and mass unrest.

### Policy recommendations — five key dimensions and specific actions
- Overarching themes:
  - Get connected.
  - Develop flexible education systems.
  - Urbanize smartly.
  - Boost trade integration on the continent.
  - Expand social safety nets.
- Get Connected:
  - Total annual investment needs for building and maintaining adequate infrastructure in the next 20 years for sub-Saharan Africa are estimated at "$130 to 170 billion."
  - Transportation, electrification, and water and sanitation account for "96 percent" of the total infrastructure investment need.
  - Remaining digital communication investment needs are estimated at "$4 to 7 billion per year."
  - Digital infrastructure will require public–private partnerships given limited fiscal space.
- Develop Flexible Education Systems:
  - Emphasize digital literacy, adaptability, lifelong learning; secondary education becomes more important.
  - Use technology (online centrally designed courses) to keep curricula up to date.
- Urbanize Smartly:
  - Formalize land markets, clarify property rights, improve service delivery, coordinated infrastructure investments, and improved urban transport to mitigate rapid urbanization risks.
  - Use internet of things, big data, and connectivity to improve governance and service delivery.
- Boost Trade Integration on the Continent:
  - Implement the Continental Free Trade Agreement regionwide to realize scale for local firms and attract foreign direct investment.
  - Move swiftly on trade facilitation and regional infrastructure to form a sizable domestic market.
- Expand Social Safety Nets:
  - Prepare for a "gig-style" world of work with frequent job transitions and income volatility.
  - Technology (biometric identification and financial technology) can help targeting and administration; revenue mobilization is key to create fiscal space.

### Modeling details, quantitative insights, and caveats
- Model extension of Berg, Buffie, and Zanna (2018): two regions, discrete time, households own labor (L), capital (K), and robots (Z); robots represent a broad range of technologies.
- Production uses CES between labor and robots and Cobb-Douglas to combine tasks with capital.
- Accumulation rules: K_{t+1} = I^K_t + (1 − δ_K) K_t ; Z_{t+1} = I^Z_t + (1 − δ_Z) Z_t.
- Elasticity of substitution examples: σ = 2 (substitutes), σ = 1 (Cobb-Douglas), σ = 0.8 (complements).
- Quantitative scenarios:
  - No overall productivity growth + doubling of robot productivity: per capita GDP falls slightly in SSA while rising by about "0.5 percent per year" in ROW, causing divergence.
  - Historical productivity growth ("0.9 percent") in both regions: gap in per capita GDP growth rates is smaller but SSA still does not converge.
  - If SSA productivity growth can be raised to "3 percent" through trade liberalization and reduced misallocation, SSA can catch up and undo divergence from robot productivity increases.
- Policy areas that raise overall productivity: improved infrastructure, education, access to finance, and the business environment.
- Literature-based parameter estimates:
  - Ostry, Berg, and Kothari (2018): current account liberalization can increase total factor productivity growth by "0.25" contemporaneously moving from median to 75th percentile, persistence "0.99."
  - IMF (2018a): closing efficiency gaps by reducing misallocation could raise average total factor productivity growth by "1.5 to 2 percentage points" over the medium term.
- Modeling caveats: framework abstracts from within-country inequality, unemployment dynamics, human capital accumulation, government roles, capital market frictions, endogenous innovation, and political economy factors.

*Source: Introduction and chapters excerpted from "The Future of Work in Sub‑Saharan Africa" (IMF discussion paper).*

### Introduction �����������������������������������������������������������������������������������������������������������

### Introduction

### Context and scope
- Examines challenges and opportunities of the Fourth Industrial Revolution for sub-Saharan Africa as the region seeks to:
  - create jobs for a booming population,
  - tackle adverse effects of climate change, and
  - face an external environment that may become less supportive.
- Approach combines history, economic modeling, empirical evidence, and scenario analysis.
- Paper structure:
  - Chapters 1–3: labor market developments, Fourth Industrial Revolution impacts, and other big trends (demographics and urbanization, climate change, global economic integration).
  - Chapter 4: three realistic scenarios for the region over the next two decades.
  - Chapter 5: policy ideas to generate the needed jobs.

### Demographics and labor force projections
- Population expected to rise from "1 billion today to 1.7 billion by 2040."
- The working-age population in the region will, on average, experience "a net increase of 20 million per year over the next two decades."
- Migration toward cities is accelerating, increasing the need for urban jobs.

### Recent employment trends
- Since 2000, sub-Saharan Africa has been able to add "an average of 9 million jobs per year."
- Employment-to-population ratio increased from "63 percent in 2000 to 64 percent in 2017."
- Job creation kept pace with population growth over the past two decades, despite demographic pressures that doubled compared to the 1990s.
- Sectoral job creation:
  - Most new jobs were created in sectors with low productivity levels, such as subsistence agriculture and low value-added services.
  - Sub-Saharan Africa has the lowest share of employment in manufacturing in the world.
  - Reallocation toward services did not predominantly favor "modern" services; during the 2000s, the share of modern services in sub-Saharan Africa’s exports of services declined.
- Employment type:
  - "Six of the nine million jobs created annually on average since 2000 have gone to own-account and contributing family workers."
  - "Only 2.6 million have gone to employees being paid for their work."
  - Self-employment remains predominant and is associated with lower productivity and informality.

### Heterogeneity across countries and key statistics
- Large cross-country variation in employment outcomes:
  - Employment-to-population rate declined by "13 percentage points in Burkina Faso" as population grew on average by "3.2 percent since 2000", while employment expanded at "2 percent."
  - In countries such as Burundi, Côte d’Ivoire, Kenya, Lesotho, Mozambique, and Sierra Leone, population outgrew job creation.
  - In Angola, Botswana, Congo, Ghana, Mali, Senegal, and Zimbabwe, job creation largely outpaced population.
  - Joblessness remains severe in South Africa and Eswatini, whose employment-to-population ratios remained broadly stable at the lowest level for the region, "below 40 percent", less than half of the employment rates in Madagascar, Rwanda, and Tanzania.

### Analytical approach
- Chapter 2 uses both a simple economic model and partial equilibrium analysis of exports’ vulnerability to automation to explore technological change.
- Chapter 4 integrates technological, climate, and integration trends into three scenarios for the next two decades.

### Policy recommendations (five key dimensions)
- Governments should take decisive policy actions across five key dimensions today to prepare workers and labor market entrants for the jobs of tomorrow:
  - Get connected — "Governments should invest in infrastructure, both traditional and digital, to provide the basis for the region to leverage the opportunities from the Fourth Industrial Revolution."
  - Develop flexible education systems — "A flexible education system will be critical to keep pace with changing technologies to ensure that workers have the skills to complement machines, and not to be substituted by them."
  - Urbanize smartly — "Governments should take advantage of new technologies to institute effective urban planning, formalize land markets, clarify property rights, improve service delivery, and strengthen governance, accountability, and social infrastructure."
  - Boost trade integration on the continent — "Move forward with trade integration in sub-Saharan Africa, for example through the Continental Free Trade Agreement, but also through investing in regional infrastructure."
  - Expand social safety nets — "As labor markets become more volatile, developing social safety nets can smooth transitions between jobs or 'gigs' for individuals and buffer income volatility."

*Source: Introduction, "The Future of Work in Sub-Saharan Africa" (IMF discussion paper).*

### 1. Employment-to-Population Rate

### 1. Employment-to-Population Rate

### The Present of Work in Sub-Saharan Africa
- The chapter uses history, modeling, and empirical evidence to assess the impact of technology on jobs in sub-Saharan Africa.
- Fundamental uncertainty: Will the Fourth Industrial Revolution take away jobs or create many new products, services, and jobs?
- Two broad pathways:
  - If technology complements workers, sub-Saharan Africa’s convergence with the rest of the world could receive a boost.
  - If technology substitutes for workers, sub-Saharan Africa could diverge from the rest of the world.
- The “classical” manufacturing-export-led development model will need adaptation to constraints and opportunities from the Fourth Industrial Revolution.

### The Fourth Industrial Revolution (definition and characteristics)
- Characterized by breakthroughs in artificial intelligence, machine learning, robotics, and 3-D printing (Schwab 2016).
- Faster, more accessible, more mobile internet; increased data storage and processing capacity.
- Machine learning and artificial intelligence expand the range of automatable tasks beyond routines (Autor 2014; Gibbs 2017).
- Polanyi’s paradox has limited automation historically; machine learning and artificial intelligence are beginning to break this limit.

### Historical Perspective: lessons from previous industrial revolutions
- Previous industrial revolutions produced massive productivity gains and long-term improvements in living conditions.
- Transitions included: improved food provision, rise of international trade, growth of cities, social and cultural transformations, and large-scale reallocation of labor.
- Educational systems were redesigned historically to meet factory skill needs; example: in the United States high school enrollment rose from 9 percent to 73 percent of the youth between 1910 and 1940 (Goldin and Katz 2008).
- Transition periods produced losers: skilled artisans displaced, poor factory working conditions, increased inequality (Deaton 2013; Piketty 2013).
- Industrial revolutions contributed to global divergence; countries that adopted innovations early experienced fast growth (examples cited: Japan, Korea, China, India).

### Anxiety and hope: empirical estimates of jobs at risk
- Wide uncertainty in estimates for advanced economies: existing jobs at risk range from 7 percent to 60 percent.
  - McKinsey Global Institute (2017): 40 percent to 55 percent of jobs in advanced economies may be at risk from automation.
  - World Bank (2016): almost 60 percent of employment is at risk of automation in OECD economies.
  - Studies using individual-level task data (Arntz, Gregory, and Zierahn 2016): only 7 percent to 15 percent of OECD jobs are at risk.
- For sub-Saharan Africa:
  - McKinsey Global Institute (2017) and World Bank (2016) estimate employment in sub-Saharan Africa is a little less susceptible to automation than in advanced economies but still sizable, with estimates for individual countries in the region ranging from 40 percent to 60 percent.
- Demographic difference:
  - Advanced economies face declining working populations and thus may embrace automation to sustain output with fewer workers.
  - Sub-Saharan Africa has rapidly growing working populations and needs to add new jobs year after year for most of this century.
- Examples of technology uptake and local solutions across sub-Saharan Africa include mobile money (East Africa), phone-based recruitment (Biscate in Mozambique), Vula Mobile (South Africa), and other innovations spanning agriculture, healthcare, education, logistics, and finance.

### Digital infrastructure and affordability (observations)
- Internet penetration in the region remains the lowest in the world—less than half the global average (Figure 8).
- Only a few countries (Kenya, Nigeria, and Seychelles) have achieved penetration levels of close to 50 percent, slightly above the global average.
- The cost of a fixed broadband connection is the highest in sub-Saharan Africa compared to other regions (Figure 9).
- Mobile broadband costs in US dollar terms are comparable to other regions, but when scaled by gross national income, sub-Saharan Africa has the least affordable mobile broadband.

### The Theory: modeling the impact of the Fourth Industrial Revolution
- Model features two regions: an advanced economy and a low-income region (sub-Saharan Africa) that trade.
- Production uses traditional capital, labor, and “robots” (robots broadly include automation, machine learning, artificial intelligence).
- Regions differ in overall productivity; model calibrated so per capita GDP in advanced economies is 15 times that in sub-Saharan Africa (matches gap between sub-Saharan Africa and the United States).
- Because wages are lower in sub-Saharan Africa, production there uses relatively more labor and fewer robots.
- Robots and labor can be complements (robots help workers) or substitutes (robots replace workers).
  - Example complement: digital technology in agriculture that helps farmers treat pest infestation.
  - Example substitute: self-checkout machines replacing cashiers.
- Modeling robot productivity doubling:
  - Firms invest in robots and traditional capital and increase production.
  - Both regions see an increase in per capita GDP in the long term.
- Long-term regional outcomes depend on complementarity vs. substitutability:
  - If labor and robots are complements:
    - Increase in per capita GDP is larger in sub-Saharan Africa than in the advanced economy (convergence).
    - Sub-Saharan Africa benefits because lower wages make it more profitable to invest in robots where combined with relatively cheap labor.
  - If labor and robots are substitutes:
    - Increase in per capita GDP is larger in the advanced economy than in sub-Saharan Africa (divergence).
    - Introducing robots is most profitable where wages are high because they save on costly labor.
- Labor share dynamics across steady states:
  - Labor share increases only if robots complement labor.
  - When robots complement labor, wage increases exceed increases in robot stock, raising labor share in both regions, with larger increases for sub-Saharan Africa.
  - When robots substitute for labor, robot stock increases more than wages, leading to a fall in labor share in both regions; increase in real wages is stronger in the advanced economy, but per capita GDP increases even more, so labor share falls more there, raising inequality.
- Policy lever: overall productivity growth (influenced by infrastructure, skills, institutional and regulatory framework) can be large enough to achieve convergence even if automation is a headwind.
  - Historic productivity growth realized between 2000 and 2014: 0.9 percent on average.
  - Historic productivity growth of 0.9 percent would not be enough to close the gap with advanced economies.
  - Potential productivity growth scenario: trade integration and efficiency gains could reduce the productivity gap by half in two decades, from a factor of 1 to 15 to a factor of 1 to 7, bringing sub-Saharan Africa’s income position relative to the United States to slightly ahead of where India currently is (Figure 12).

### The numbers: assessing the impact on exports from sub-Saharan Africa
- Concern: sub-Saharan Africa may be less vulnerable to direct automation of jobs due to economic structure and wage levels, but may be indirectly affected via exports if automation in advanced economies enables reshoring.
- Constructed two export vulnerability indices to gauge vulnerability of sub-Saharan Africa’s exports to automation, based on Frey and Osborne (2017) and Brynjolfsson, Mitchell, and Rock (2018); indices mapped from automatability of occupations to industries to export goods.
- Divergent index results:
  - Frey and Osborne–based index suggests sub-Saharan Africa’s exports (and those of low-income and developing countries) are relatively more vulnerable to automation than exports of more developed countries (Figure 13).
  - Brynjolfsson, Mitchell, and Rock–based index shows the opposite: sub-Saharan Africa’s exports are relatively less vulnerable compared with advanced economies (Figure 14).
- Explanation of divergence:
  - Frey and Osborne emphasize technologies that replace routine jobs; they rate crop production and low-end manufacturing (food, paper, wood manufacturing) as relatively more susceptible to automation—these constitute much of sub-Saharan Africa’s noncommodity exports.
  - Brynjolfsson, Mitchell, and Rock emphasize machine learning replacing nonroutine cognitive tasks; they rate sophisticated electronics and machinery as relatively more susceptible—sectors prominent in advanced economies’ export baskets.
- Areas of agreement across indices:
  - Both indices consider apparel and textile manufacturing susceptible to automation.
  - Both indices consider mining less vulnerable (note: index refers to export vulnerability; this does not imply that jobs in mining are not subject to automation—example: use of drone mining in South Africa).
- Implications:
  - The impact on sub-Saharan Africa depends on whether technology complements or substitutes different types of labor.
  - Technological advances in low-end manufacturing (per Frey and Osborne) could render the traditional export-led growth model infeasible.
  - Speed of these changes depends on pace of technological progress and evolution of wages relative to price of new technologies.
  - Export diversification is emphasized to reduce exposure to automation shocks.

### Other major trends affecting the future of work in sub-Saharan Africa
- Identified three additional big trends, each with great uncertainty:
  - Demographic expansion and urbanization.
  - Severity of climate change impact.
  - Global economic integration.
- Key points:
  - Working-age population is rising sharply, creating an opportunity for a demographic dividend contingent on improved educational outcomes.
  - Urbanization is advancing rapidly but has not yet delivered substantial economic benefits.
  - Climate change threatens growth, undermines food security, and accelerates migration.
  - The region has become more integrated with the rest of the world and with itself but still lags behind other regions.
- Demographic statistics:
  - Total population increased from 470 million in 1990 to 1 billion today.
  - Based on UN projections, sub-Saharan Africa’s population will continue growing (projection text cut off in source).

*Source: IMF staff and referenced studies as presented in the chapter "Technological Change and the Future of Work in Sub-Saharan Africa."*

### 1.7 billion by 2040—a 70 percent increase—and to close to 4 billion by

### Other Big Trends Shaping the Future of Work in Sub-Saharan Africa

### Demographics and population growth
- Sub-Saharan Africa population projections: "1.7 billion by 2040—a 70 percent increase—and to close to 4 billion by the end of the century."
- In comparison, the population of the rest of the world is expected to increase by "about 16 percent between now and 2040."
- The share of working age population began to increase in the region in 1985 and is expected to continue increasing for several more decades, creating an opportunity to capture a demographic dividend.
- Heterogeneity across countries:
  - Mauritius already has a fertility rate of "less than 2."
  - Niger has a fertility rate of "over 7," the highest in the world.
  - Some countries have almost completed demographic transitions (Mauritius, South Africa); others have barely started (Central African Republic, Nigeria).
- Uncertainty in population projections:
  - Wittgenstein Centre and International Institute for Applied Systems Analysis projections are significantly lower than United Nations projections by "as much as 300 million by 2040 and 1.6 billion by 2100."
  - Main differences stem from assumptions about women’s access to education and employment and pace of child health improvements.

### Jobs, education, and workforce needs
- Sub-Saharan Africa must create "16 to 20 million new jobs on average every year."
- By 2030 the region "will contribute half of the annual increase in the global working age population."
- Education indicators:
  - Important gains in primary completion rates.
  - Secondary enrollment rates remain "well below other regions" and are a forward-looking indicator of likely education outcomes for the next generation.

### Urbanization and agglomeration economies
- Urbanization trends:
  - Urban population has grown by "about 4.4 percent annually over the last 50 years" and was "close to 400 million at the end of 2015."
  - The average urbanization rate now "exceeds 40 percent."
- Urbanization characteristics and challenges:
  - Cities are on average congested but not economically dense, reflecting inadequate residential housing near city centers, highly fragmented and unreliable transportation, and limitations in clearing labor and product markets.
  - Urbanization in sub-Saharan Africa is concentrated in "consumption cities," with employment mostly in informal, nontradable services rather than manufacturing-led "production cities."
  - Nondiversified, resource-intensive export structures distributed mainly in cities raise food prices and put upward pressure on formal sector wages relative to income levels.

### Climate change and environmental impacts
- Historical and projected temperature changes:
  - In Africa, temperatures have increased "by about 0.5°C over the last 50 to 100 years."
  - This compares with a global average increase of "about 1°C."
  - Projections from the Intergovernmental Panel on Climate Change range from "0.2°C per decade" to "0.5°C per decade."
- Expected consequences identified by the Intergovernmental Panel on Climate Change:
  - Reduction in the availability of water resources on average; increased need for coordination in water management for shared basins.
  - Increased food insecurity due to more extreme weather, declines in soil moisture, and water stress affecting agriculture and fisheries.
  - Pressures to migrate away from uninhabitable areas and loss of biodiversity.
  - Increased risk of vector- and water-borne diseases for populations with inadequate health care.
  - Sea level increases putting coastal areas and transportation networks at risk.
- Economic impacts:
  - IMF (2017) estimates the median low-income country experiencing a "1°C increase in temperature" would suffer a fall in per capita GDP growth of "1.2 percentage points in the same year."
  - The impact is long-lasting, "reducing per capita output by 1.5 percentage points seven years after the shock" for the median low-income country.
  - Main channels: lower output in agriculture and manufacturing; little impact on services.
- Climate-driven migration and urbanization:
  - Urbanization rate increases by "0.45 percent with a reduction in rainfall of 1 percent."
- Suggested mitigants:
  - Formalization of land markets, clarifying property rights, effective urban planning, coordinated infrastructure investments, and improved urban transport to mitigate rapid urbanization risks.

### Global economic integration
- Integration trends:
  - Sub-Saharan Africa’s trade openness accelerated in the mid-1990s; export destinations expanded to emerging markets, particularly China.
  - Financial integration: nonofficial net private capital flows increased from "$4 billion in the 1980s and 1990s to close to $60 billion in 2017."
  - Intraregional trade rose from "6 percent of total exports in 1980 to 20 percent of GDP in 2016," but remains low.
  - Measured by exports, number of trade partners, and partner weight, sub-Saharan Africa’s integration is "about half that of other emerging and developing countries."
- Risks to future integration:
  - Recent protectionist pressures (higher tariffs between the United States and China, and the United Kingdom leaving the European Union) raise concerns that further global economic integration could slow or reverse, reducing opportunities for less integrated economies.

### Scenario analysis and policy implications
- Key points from scenario analysis:
  - Sub-Saharan Africa has little or no control over how technology, demographics and urbanization, climate change, and global economic integration will play out, but policy matters for shaping responses.
  - With the right infrastructure and skill-sets, technology offers opportunities for leapfrogging development and mitigating adverse effects, including from climate change.
  - A flexible education system is critical to empower people to work with and benefit from new technology rather than be displaced.

*Source: IMF staff and cited data within the chapter.*

### Introduction

### Introduction

### Scenario analysis methodology
- Scenario analysis is used in situations with large and multifaceted uncertainty to illustrate alternative outcomes rather than one specific baseline projection.
- It is particularly useful under Knightian uncertainty, when events are so unpredictable that it is impossible to attach quantifiable risk probabilities to different outcomes.
- The emphasis is on creating plausible, relevant, and challenging scenarios to help decision-making under uncertainty (Ramírez and Wilkinson 2016).
- The specific approach used in this paper is the inductive method: start from available data and ideas about the future and generate insights through a bottom-up process to develop qualitative narrative scenarios that capture key ideas, trends, and interplay of outcomes difficult to capture quantitatively.
- Scenarios are most valuable when considered as a set to compare, contrast, and assess whether they capture the broad range of most likely outcomes and to generate and test different policies in alternative possible worlds.

### Scenario set and interacting drivers
- Three scenarios for the future of work in sub-Saharan Africa are considered: “Africa Arisen”, “Africa for Africa”, and “Africa Adrift”.
- Key drivers that evolve and interact differently across scenarios: technology, global economic integration, climate change, and demographics.
- Summary of driver impacts across scenarios:
  - Technology:
    - Africa Arisen: increases productivity and empowers workers.
    - Africa for Africa: allows for leapfrogging with limited negative impact on jobs thanks to lower wages.
    - Africa Adrift: substitutes for workers, wiping out jobs across the region.
  - Global economic integration:
    - Africa Arisen: deepens.
    - Africa for Africa: reversed—with some offset from strengthened regional integration.
    - Africa Adrift: gives way to volatility and trade wars.
  - Climate change:
    - Africa Arisen: largely mitigated by technological advances.
    - Africa for Africa: has a negative impact on growth.
    - Africa Adrift: devastating, leading to mass migration and obsolete infrastructure.

### Scenario: Africa Arisen
- Picture: rural smallholder farmer with a simple mobile phone; peak sorghum harvest a bumper season; transition to self-sufficiency in staple grain via new seeds, text message irrigation tips, and access to credit. Urban relatives send remittances; daughter maintains IT infrastructure for an African service company headquartered elsewhere on the continent.
- Technology as a complement to labor:
  - Investment in new technologies improved productivity of low-skilled workers and created a small but growing manufacturing base integrated with global value chains.
  - In agriculture, technology raised yields and overall output; mechanization and big data reduced rural labor demand and spurred urban migration.
  - Greater digital connectivity within national borders, regionally, and globally spurred employment in services.
- Deepening global integration:
  - Breakthrough in global trade negotiations begins new era of goods and services integration.
  - Development of a stable playing field for services trade accelerated global economic integration, supported by financial flows, particularly foreign direct investment.
- Climate change adaption:
  - Global cooperation and investment produced mitigation and adaption strategies (e.g., seeds resistant to drier weather conditions).
  - Movement from less habitable areas increased urban pressures.
- A service-led economy:
  - Urbanization created thriving businesses, mostly in services; digitally connected youth near innovation hubs developed and adapted new technologies.
  - Integration into global value chains and stable trading environment enabled trade in global markets.
- A vibrant but volatile world of work:
  - Employment dominated by small-scale services businesses serving a global market and multinational companies integrated into global value chains.
  - Dominant occupations include professional services, tech entrepreneurs, and low-skilled services.
  - Workers face frequent job transitions driven by changes in global trends.
- Urbanization challenges and government response:
  - Investment in transportation networks and low-cost housing increased but cities remain bottlenecks due to congestion and limited efficient, low-cost housing.
  - Governments invested in physical and digital infrastructure; growing tax base and technology improved tax revenues and service delivery.
  - Governments slowly expanded social safety nets as the revenue base grew, reducing informality and increasing incentives to contribute to national health, employment, and pension schemes.
- Skills demand:
  - Innovation, creativity, and continual skill updating are rewarded; digital literacy and access to the digital world are basic necessities.

- Timeline highlights (as presented):
  - 2020 2025 2030 2035 2040
  - Events noted: Breakthrough in global trade negotiations; Countries commit to ambitious investments in climate change mitigation and adaption; Sub-Saharan Africa becomes home to the fastest growing online start-ups; Universal social safety nets introduced in some sub-Saharan African countries; Global investment in sub-Saharan Africa’s services sector reaches the highest in the world.

### Scenario: Africa for Africa
- Picture: Kali helps her father run a small business providing services to local farmers; she manages drones that survey fields and collect data; dreams of starting a business to help youth across Africa.
- Leapfrogging via simple and cheap technologies:
  - Information technologies and open source know-how are adopted for production and service needs.
  - Distributed ledger technology improves public sector efficiency and facilitates financial sector development.
  - Active social media increases government accountability; digital communication enables education and knowledge sharing.
- Rise of regional integration as globalization stalls:
  - Technological shifts and job displacement in advanced economies spurred inward-looking policies and trade wars.
  - Reversal of globalization accompanied by slowed foreign investments and official development assistance.
  - African leaders deepened regional integration via implementation of the Continental Free Trade Agreement; public-private partnership built a high-speed railroad connecting Mombasa and Cape Town.
- Climate impacts and adaptation:
  - Climate change exacerbated existing difficult weather; frequency and severity of droughts and floods affected agriculture.
  - Technologies such as smart water management, improved weather forecasting, and digital connectivity mitigated some impacts but global effort remained lacking.
- Economic structure:
  - Agriculture adopted innovative technologies improving productivity from a low base, producing excess rural labor and urban migration.
  - Industry saw small-scale, locally based manufacturing aided by low wages and 3D printing-like technologies; limited access to financing constrained further industrial growth.
  - Service sectors became primary employers; new services proliferated, including information and communication, retirement communities, and entertainment.
  - Domestic mass production of electric scooters charged from solar-powered microgrids emerged to meet urban transport demand.
- Entrepreneurship and finance:
  - Regional integration and connectivity fostered entrepreneurship and grassroots innovation.
  - Financial sector development was slow; businesses relied on self-financing and crowdfunding; domestic savings were insufficient for large-scale expansion.
  - Open-source education provided self-education opportunities but required initial skill sets.
- World of work:
  - Vibrant private sector dominated by small-scale businesses and low-productivity jobs.
  - High informality mitigated by information technologies but low job security and weak social safety nets left workers vulnerable; poverty remained widespread, especially in rural areas.

- Timeline highlights (as presented):
  - 2020 2025 2030 2035 2040
  - Events noted: Multilateral talks to end the trade war collapse; Continental Free Trade Agreement comes into force; Entrepreneur begins production of electric “bodaboda” scooters; High-speed railroad connects Mombasa and Cape Town; Afrovision celebrates its 15th anniversary.

### Scenario: Africa Adrift
- Picture: Adamu, a night-shift security worker watching surveillance drones; lives 10 kilometers from work in a mega slum; long hours, low wage, precarious full-time employment; older children risk joining mass protests; younger children attend a local school offering teacher-guided online courses with hopes of migrating for better jobs.
- Technology replaces labor:
  - Rapid technological change led to reshoring of manufacturing to advanced economies and rendered many African infrastructure projects worthless.
  - Deployment of robots and artificial intelligence in Asia, Europe, and North America changed global trade, increasing self-sufficiency in advanced economies.
  - Sub-Saharan Africa was severely affected: stranded assets, high debt, low domestic revenues, no fiscal space, and a painful decline in external financing.
- Global (dis)integration:
  - International cooperation collapsed into zero-sum competition, trade wars, and intellectual property disputes.
  - The global trading system collapsed; trade within the region dwindled.
  - Censorship and state control of media became prevalent; elites strengthened hold on institutions.
- Severe climate impacts:
  - Climate change hit harder and faster than anticipated; a number of major cities literally ran out of water in the late-2020s.
  - Declining humidity rendered cocoa cultivation unfeasible (additional impacts continue in the scenario narrative).
- Socioeconomic outcomes:
  - Development policies thwarted by global developments, leaving economies stagnant and indebted.
  - Informal jobs in subsistence agriculture and low productivity services remained dominant.
  - Frequent economic crises and dependence on primary commodities fueled an “each for their own” mentality and mass civil unrest.

- Timeline highlights (as presented):
  - 2020 2025 2030 2035 2040
  - Events noted: Trade wars and disruptions in global financial markets become the norm; Censorship of social media is on the rise; Big African cities run out of water; The longest road-railway bridge in Africa collapses, disrupting north-south movement of passengers and goods; Mass civil protests sweep Africa prompting leaders to gather at the conference of the African Civil Society Network.

*Source: IMF staff.*

### 2030. Farmers were forced to switch from cash to subsistence crops. As water

### The Future of Work in Sub-Saharan Africa

### Africa Adrift: 2030 scenario — key findings
- Farmers were forced to switch from cash to subsistence crops as water reservoirs depleted.
- Rapid migration strained cities and poor infrastructure; limited fiscal resources prevented adequate climate change mitigation.
- Some relatively new infrastructure projects collapsed because design failed to take into account changing climate conditions.
- Many living in extreme poverty sought to migrate to Europe and beyond, but destination countries did not have enough job opportunities to warrant letting anybody in.
- Informality remains the norm: an affluent minority study abroad or use international online universities and control commodity monopolies; most people provide low-pay services for elites.
- Lack of formal urban jobs fuels crime and terrorism; conflicts emerge over access to commodities and water.
- Private security becomes the largest employer in sub-Saharan Africa.

### Rapid urbanization and poverty traps
- Rural: weather dependency, subsistence farming, and poor access to water, electricity, and roads create a rural poverty trap.
- Urban: lack of affordable housing, transportation, and perennial high food prices create an urban poverty trap.
- Mass migration to cities during declining moisture episodes raises food prices, inhibits formal-sector job creation, and exacerbates urban poverty traps.
- Only those with access to informal networks or remittances can buffer these trends.

### Education and technology
- Education quality declined; government reforms failed to introduce enough flexibility to match labor market needs.
- Incorporating new skills demanded by the labor market took a long time.
- Global online courses and accreditation offer a route out of poverty, but access is limited by affordability.
- Social media becomes a critical organizing tool; mass demonstrations lead governments to become more accountable and improve social service delivery from a low base.

### Scenario framework — summary of contrasts
- Three scenario archetypes summarized: "Africa Arisen", "Africa for Africa", and "Africa Adrift" (the latter described above).
- Technology, global economic integration, and climate change interact differently across scenarios, producing divergent outcomes for:
  - Economy structure (e.g., urban innovation hubs versus commodity-dominated economies).
  - Future of work (project-based professional services versus subsistence activities).
  - Skills for success (STEM, creativity, entrepreneurship versus endowments and connections).
  - Challenges (income security and automation risk versus terrorism, conflict, and infrastructure obsolescence).
- Infrastructure and Rest of the World roles differ across scenarios, from providing financing and migration opportunities to obsolescence and commodity-focused relations.

### Policy recommendations (overarching themes)
- Get connected.
- Develop flexible education systems.
- Urbanize smartly.
- Boost trade integration on the continent.
- Expand social safety nets.

### Detailed policy actions and rationale
- Get Connected
  - Connectivity includes digital infrastructure beyond roads, railways, and ports.
  - Digital infrastructure and a supportive business environment enable new forms of business and higher efficiency, especially in services and agriculture facing climate change.
  - Total annual investment needs for building and maintaining adequate infrastructure in the next 20 years for sub-Saharan Africa are estimated at $130 to 170 billion (African Development Bank Group 2018).
  - Transportation, electrification, and water and sanitation account for 96 percent of the total infrastructure investment need.
  - Remaining digital communication investment needs are estimated at $4 to 7 billion per year.
  - Digital infrastructure is necessary to take advantage of the Fourth Industrial Revolution and will require public–private partnerships given limited fiscal space.

- Develop Flexible Education Systems
  - Key skills: digital literacy, adaptability, lifelong learning; secondary education becomes more important.
  - Education systems must be flexible given uncertainty over specific future skills.
  - Technology (online centrally designed courses) can ensure up-to-date curricula reach every classroom.

- Urbanize Smartly
  - Cities are the hotbeds of the Fourth Industrial Revolution; sub-Saharan African cities currently lack sufficient infrastructure and are often oriented toward local consumer services.
  - Urban planning and development, aided by internet of things, big data, and connectivity, can improve service delivery, governance, and social infrastructure.
  - Traditional policies (formalizing land markets, clarifying property rights) remain necessary and can be supported by technology.

- Boost Trade Integration on the Continent
  - Regional trade integration would drive development, especially if the global environment weakens.
  - The Continental Free Trade Agreement requires regionwide implementation to realize scale for local firms and attract foreign direct investment.
  - Sub-Saharan Africa can form a sizable domestic market if countries move swiftly on trade facilitation and regional infrastructure.

- Expand Social Safety Nets
  - A likely “gig-style” world of work implies frequent job transitions and income volatility; many in SSA already live with such volatility in subsistence and informal sectors.
  - Family and social networks provide some backup; expanding safety nets to offer income security is a major challenge.
  - Technology (biometric identification and financial technology) can help with targeting and administration, but revenue mobilization is key to create fiscal space.

### Modeling approach and quantitative insights
- Model extension of Berg, Buffie, and Zanna (2018): two regions (SSA and ROW), discrete time, no uncertainty; households own labor (L), capital (K), and robots (Z).
- Robots represent a broad range of Fourth Industrial Revolution technologies; production uses CES between labor and robots to produce tasks, and Cobb-Douglas to combine tasks with capital.
- Household utility uses discount factor β and parameter τ for intertemporal elasticity of substitution (notation preserved).
- Accumulation rules: K_{t+1} = I^K_t + (1 − δ_K) K_t ; Z_{t+1} = I^Z_t + (1 − δ_Z) Z_t.
- Elasticity of substitution examples: σ = 2 (substitutes), σ = 1 (Cobb-Douglas), σ = 0.8 (complements).
- Robot-to-labor ratio scaling: Z/L ∝ w^σ, implying regions with higher wages have higher robot-to-labor ratios.
  - Example magnitudes: robot-to-labor ratio in ROW is 15 times higher than SSA when robots and labor are complements, and almost 200 times higher when robots substitute for labor.
- Transition dynamics and financial flows:
  - When robots and labor are substitutes, higher demand for robots in ROW channels resources out of SSA, leading to a fall of GDP per capita in SSA in the short term, along with declines in wages and labor share.
  - SSA can run a current account surplus in the transition, accumulating foreign assets; in the final steady state, consumption in SSA can be greater than output, financed by interest income on accumulated assets.
  - If robots and labor are complements, resources may flow into SSA and GDP per capita can grow more rapidly in the short term.

- Quantitative scenarios for robot productivity increase:
  - If there is no overall productivity growth in both regions, a doubling of robot productivity leads to per capita GDP falling slightly in SSA but rising by about 0.5 percent per year in ROW, leading to further divergence.
  - If overall productivity grows at the historical rate seen between 2000 and 2014 in both regions, the gap in per capita GDP growth rates is much smaller, but SSA still does not converge.
  - If productivity growth in SSA can be pushed up to a level of 3 percent through trade liberalization and reduced misallocation and efficiency gains, SSA can catch up and undo divergence from the robot productivity increase.
  - The analysis highlights that economy-wide productivity gains could have a greater impact on income than robot productivity increases in SSA.

### Calibration and policy implications from the model
- Calibration matches relative GDP per capita between regions and the robot share in income in ROW by choosing σ, total factor productivity (A_i), and labor share in task production (e).
- Policy areas associated with higher overall productivity: improve infrastructure, education, access to finance, and the business environment.
- Estimates and literature-based parameters cited in the modeling discussion:
  - Ostry, Berg, and Kothari (2018) estimate current account liberalization can increase total factor productivity growth by 0.25 contemporaneously if moving from the median to the 75th percentile in the index, with a persistence of 0.99.
  - IMF (2018a) estimates that closing efficiency gaps by reducing misallocation could raise average total factor productivity growth by 1.5 to 2 percentage points over the medium term.
- Modeling caveats:
  - How to model technological change remains an open question; projection of robot productivity increases is difficult.
  - The framework treats labor either as complement or substitute to robots, but real-world outcomes will vary across worker types.
  - The model abstracts from within-country inequality, unemployment, human capital accumulation dynamics, the role of government, capital market frictions, endogenous innovation, and political economy factors.

*Source: IMF staff calculations and chapter content from The Future of Work in Sub-Saharan Africa.*

### 1. Growth Decomposition: SSA2. Growth Decomposition: ROW

### 1. Growth Decomposition: SSA2. Growth Decomposition: ROW

### Empirical Approach: Automation Vulnerability Index (export vulnerability)
- Overall export vulnerability for a country is calculated as the export-share weighted sum of industries’ automatability indices.
- Industry-level automation vulnerability calculation (using occupation vulnerabilities indexed by j):
  - A  V  
    i

    j
      =    ∑  
    o
     s  h  
    o,i
       AV  
    o

    j
- Definitions and steps:
  - AV_o_j is the automation vulnerability of occupation o (two types used, indexed by j: (1) based on Brynjolfsson, Mitchell, and Rock (2018); and (2) based on Frey and Osborne (2017)).
  - s h_o,i is the share of industry i workers who are in occupation o in the United States (derived from the US Industry-Occupation Matrix).
  - Using a concordance of products and industries, country exports by products are mapped to exports by industry to compute s h_c,i, the share of a country’s exports coming from industry i.
  - Country-level automation vulnerability index is obtained by combining industry export shares with industry automation vulnerability:
    - AV  
      c

      j
        =    ∑  
      i
     s  h  
    c,i
       AV  
    i

    j

### Contributions to Scenario Planning Approach
- Scenarios were developed in a workshop facilitated by Shirin Elahi (Scenarios Architecture) and Alberto Behar (IMF).
- Insights distilled from interviews with a broad range of experts informed the scenarios, including contributions from:
  - Julius Akinyemi (Massachusetts Institute of Technology)
  - Stone Atwine (Eversend)
  - Brahima Coulibaly (Brookings Institute)
  - Stefan Dercon (University of Oxford)
  - Shanta Devarajan (World Bank)
  - Peter Draper (University of Adelaide)
  - Andreas Freytag (University of Jena)
  - Idayat Hassan (Centre for Democracy and Development)
  - Remi Jedwab (George Washington University)
  - Susan Lund (McKinsey Global Institute)
  - Arthur Muliro (Society for International Development)
  - Elizabeth Nyamayaro (UN Women)
  - Thomas Schäfer (Volkswagen)
  - Catarina Tully (School of International Futures)

### Figure reference
- Figure 31. Projected Average Annual Growth Rate (visual provided in source; axis range shown in source: –1.0 to 5.0)

### Key methodological notes
- Two occupation-vulnerability mappings were used: Brynjolfsson, Mitchell, and Rock (2018) and Frey and Osborne (2017).
- The US Industry-Occupation Matrix is used to translate occupation-level automation vulnerability into industry-level vulnerability.
- Country export structures (product-to-industry concordance and s h_c,i) are crucial to compute country automation vulnerability indices.

### References (selected from source)
- Acemoglu, Daron, and Pascual Restrepo. 2018. “Modeling Automation.” AEA Papers and Proceedings 108: 48–53.
- Brynjolfsson, Erik, Tom Mitchell, and Daniel Rock. 2018. “What Can Machines Learn, and What Does It Mean for Occupations and the Economy?” AEA Papers and Proceedings 108: 43–47.
- Frey, Carl Benedikt, and Michael Osbourne. 2017. “The Future of Employment: How Susceptible are Jobs to Computerization?” Journal of Technological Forecasting and Social Change 114 (C): 254–80.
- McKinsey Global Institute. 2017a. “A Future that Works: Automation, Employment, and Productivity.”
- OECD and other working papers and IMF chapters cited in source.

*The FuTure oF Work In Sub-SAhArAn AFrICA — excerpt content unit*

---


_Source: https://www.imf.org/-/media/files/publications/dp/2018/48333-dp1818-the-future-of-work-in-ssa.pdf_