## 3. Private Investment to Rejuvenate Growth

## Source details

**Canonical URL:** [3. Private Investment to Rejuvenate Growth](https://www.imf.org/-/media/files/publications/reo/afr/2018/may/pdf/chap3.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/reo/afr/2018/may/pdf/chap3.pdf.md)
- [Structured JSON version](/-/media/files/publications/reo/afr/2018/may/pdf/chap3.pdf.json)

---

### Private investment trends
- Private investment in sub-Saharan Africa is, on average, 2 percent of GDP lower than in other developing economies.
- Private investment averaged 15 percent of GDP during 2010–16, compared with 22 percent for developing economies in Asia, 18 percent in Europe, 17 percent in Latin America, and 16 percent in the Middle East and North Africa (MENA).
- The gap in private investment relative to other regions has fallen by half since the early 2000s, driven by a decade when private investment grew at an average rate of 14 percent a year.
- Since 2010, private investment slowed, growing on average at 5 percent a year through 2014 and contracting during 2015–16.
- Investment in the region contracted by 4 percent each year on average in 2015–16.
- The decline in private investment was widespread: private investment slowed in two-thirds of countries and fell in half of them.
- By subgroup averages (2010–16):
  - Oil exporters: 14 percent of GDP.
  - Other resource-intensive countries: 17 percent of GDP.
  - Non-resource-intensive countries: 15 percent of GDP.

### Drivers and country experiences
- Commodity prices and resource discoveries:
  - Elevated commodity prices supported large increases in private investment in commodity-exporting countries (example: Nigeria during 2007–14).
  - Discoveries of natural resources supported investment in Equatorial Guinea and Ghana.
  - Some commodity importers benefited from lower commodity prices that created fiscal space for investment (example: Rwanda).
- Resolution of conflicts:
  - The end of long-standing conflicts in Côte d’Ivoire, Ethiopia, Rwanda, and Uganda was followed by marked increases in private investment (post-conflict 10-year averages exceed conflict-period averages for these countries).
- Idiosyncratic and spillover shocks:
  - Country-specific shocks and adverse spillovers from large regional economies (Angola, Nigeria, South Africa—combined GDP weight about 50 percent of the region) contributed to investment declines.
  - Examples: policy and political uncertainty in South Africa; sharp slowdown in credit growth in Kenya; completion of a large mining project in Namibia.

### Empirical determinants (panel estimation, 101 EMDEs, 1980–2015)
- Methodology notes:
  - Dependent variable: private-investment-to-GDP ratio.
  - Estimation method: system generalized method of moments (system GMM).
  - Controls: real GDP growth, public investment as a share of GDP, GDP per capita (PPP), relative price of capital, real interest rate, lagged private-investment-to-GDP ratio, and structural/institutional variables (regulatory quality, insolvency costs, infrastructure, trade openness, financial development, capital account openness).
- Key empirical findings:
  - Real GDP growth raises private investment (accelerator effect) and the impact is nonlinear: private investment increases when real GDP growth is high (above the country historical average), but not when growth is low (below the country historical average).
  - Public investment has an ambiguous effect: it can complement private investment (by providing infrastructure) or crowd out private investment (by competing for scarce financial resources or creating supply bottlenecks).
  - Higher relative price of investment reduces private investment ratios.
  - GDP per capita level and real interest rate were not significant in baseline results.
  - Institutional and structural characteristics strengthen the effect of GDP growth on private investment:
    - Better regulatory quality and lower insolvency costs increase responsiveness of private investment to growth.
    - Better public infrastructure (higher proportion of paved roads; greater access to electricity) increases private sector investment responses to growth.
    - Greater trade openness raises the investment reaction to economic activity.
    - Less open capital accounts are associated with a stronger impact of GDP growth on investment.
    - Higher financial development significantly raises the responsiveness of private investment to GDP growth; very low levels of financial development can be a binding constraint.
  - Persistence: investment ratios display persistence, motivating inclusion of lagged private-investment-to-GDP.

### Magnitudes: impact of a 1 percentage point increase in GDP growth on the private investment ratio (Table 3.1)
- Whole Sample: 0.21
- Low Regulatory Quality (SSA average) – High Regulatory Quality (non-SSA EMDEs average): 0.29 – 0.48
- High Insolvency Cost (SSA average) – Low Insolvency Cost (non-SSA EMDEs average): 0.02 – 0.24
- Higher Proportion of Paved Roads: 0.28
- Higher Access to Electricity: 0.33
- Higher Trade Openness: 0.26
- Lower Capital Account Openness: 0.33
- Higher Financial Development: 0.47

### Policy-relevant implications and recommended approaches
- Promote stronger and sustainable economic activity to trigger the accelerator effect on private investment.
- Improve regulatory quality and insolvency/resolution frameworks to amplify the investment response to growth.
- Increase quality and efficiency of public infrastructure (paved roads, electricity access) to encourage private capital formation.
- Promote trade openness to raise incentives for firms to invest for export markets.
- Foster financial development and deepen domestic financial markets to remove financing constraints; in countries with very low financial development, demand-side recovery alone may not translate into investment.
- Manage public investment carefully: while it can support private investment by improving infrastructure, policymakers should be mindful of potential crowding out when public investment competes with private investment for scarce funds or faces binding supply constraints.
- Diversify sources of financing: promote alternative financing mechanisms for public and private investment (including public–private partnerships) while ensuring associated risks are well managed.
- Attract FDI and consider Special Economic Zones (SEZs) as potential tools to boost private investment, recognizing mixed historical experiences with SEZs.

### Public investment: complementarities and crowding out
- Mechanisms of crowding out:
  - Competing for scarce physical and financial resources via debt issuance, bank credit, higher taxes, or inflation.
  - State enterprises producing output in direct competition with private sector goods and services.
  - Increased macroeconomic instability when public investment is financed through accumulation of unsustainable debt.
- Empirical finding on interaction with financial development:
  - Given observed regional levels of financial development, a 1 percentage point increase in the public investment ratio would lead to a ½ percentage-point contraction of the private investment ratio in the average sub-Saharan African country.
  - The same 1 percentage point increase would lead to a ½ percentage-point increase in private investment in other emerging market and developing economies in the sample.

### Alleviating constraints — Deepening financial systems
- Key constraints and patterns:
  - Bank financing of investment in sub-Saharan Africa is the lowest among regions, while equity financing is the highest.
  - Sub-Saharan Africa has the lowest share of firms that did not apply for a loan because they did not need it and the highest number of firms identifying access to finance as a major constraint.
  - Small and medium-sized firms face greater obstacles to obtaining financing than larger firms.
- Financial structure and indicators:
  - Banking systems dominate; stock exchanges and bond markets remain underdeveloped but expanding rapidly.
  - Banking systems in sub-Saharan Africa have relatively high capital ratios.
  - Regression evidence: y= –1.17**x+ 41.75 (negative association between regulatory capital ratios and credit availability).
  - Regression evidence: y= 1.05***x+ 7.10 (positive relationship between z-score and indicators of credit to the private sector).
  - z-score definition: Z = (ROA +(equity/assets))/(ROA standard deviation).
- Instruments and prerequisites for deeper financial markets:
  - Develop bond markets (registries, central depositories, clearing and settlement systems); ensure a large heterogeneous investor base; maintain a sound banking system and market-determined interest rates.
  - For equity markets, regional integration of stock exchanges can enhance liquidity and efficiency.
  - Improve judicial independence, strengthen investor protection and auditing standards, and reduce constraints in financial market infrastructures.
- Cautions:
  - Financial deepening should proceed cautiously to reduce risks of financial instability; stressed financial systems supply less credit to the private sector.
  - Strengthening institutions and promoting sound judicial, regulatory, and supervisory frameworks is necessary.
  - Fintech could provide leapfrogging opportunities for greater industry efficiency, with positive effects on financial depth and inclusion, but introduces operational, AML/CFT, and cyber-risk trade-offs.

### Public-Private Partnerships (PPPs)
- Theoretical benefits:
  - Improve infrastructure quality, bring private sector expertise, and alleviate some financial constraints by expanding financing options.
  - Broadly defined as long-term contracts where the private sector carries significant risks and receives future income streams.
- Regional intensity and composition:
  - Sub-Saharan Africa has the highest average ratio of PPP projects to GDP in the world since 2000: average ratio of 1.4 percent of GDP, compared with 1 percent of GDP in other regions.
  - Distribution within sub-Saharan Africa (average 2000–16):
    - Non-resource-intensive countries: PPPs represented 2¼ percent of GDP on average.
    - Non-oil resource-intensive countries: 1¾ percent of GDP on average.
    - Oil-exporting countries: 1¼ percent of GDP on average.
  - PPPs are mainly concentrated in energy and transportation; in the last five years projects in the energy sector represented the largest share of total PPPs.
- Risks and fiscal management:
  - Since 2006, the value of disputed PPP projects in sub-Saharan Africa averaged ¾ percent of GDP—the highest ratio among emerging market and developing economies.
  - Fiscal risks include bypassing budget constraints, need for public support (capital grants), government-provided debt or revenue guarantees (contingent liabilities), and long-term rigid payment commitments.
- Tools and capacity building:
  - PPP Fiscal Risk Assessment Model (P-FRAM) developed by IMF and World Bank; pilots in Côte d’Ivoire, Mauritius, and Niger.
  - Public Investment Management Assessments (PIMA) conducted in multiple sub-Saharan African countries to identify weaknesses in public investment practices.
- Policy implication: PPPs can be useful financing instruments but require robust institutional frameworks, legal and regulatory arrangements, public investment management capacity, and fiscal risk mitigation measures.

### Foreign Direct Investment (FDI) and SEZs
- FDI patterns:
  - Over the past decade, sub-Saharan Africa has been the main recipient of FDI in percent of GDP among emerging market and developing regions.
  - Ratio of FDI to GDP over the past decade has averaged slightly above 5 percent in sub-Saharan Africa.
  - Cross-country variation: countries with ratios since 2000 well above the regional average of about 4 percent include Cabo Verde, Mauritius, Mozambique, Seychelles, São Tomé and Príncipe, and The Gambia; two-thirds of countries have FDI ratios below the regional average.
- Determinants of FDI:
  - Large domestic markets and natural resources; provision of infrastructure; education level of labor force; openness to trade; macroeconomic and political stability; quality of institutions.
- Special Economic Zones (SEZs):
  - SEZs are a second-best solution compared with economy-wide reforms but can play a catalytic role.
  - Sub-Saharan Africa’s SEZ experience over the past two decades has been mixed; most have been unsuccessful or fallen short of expectations.
  - Reasons for weak performance include reliance primarily on corporate tax holidays with little nontax incentives or regulatory support.
  - Recent positive experiences: Rwanda and Ethiopia, focusing on clusters, competition, quality improvements, and comparative advantages.
  - Target industries include apparel, textile, and agroprocessing; few economies have SEZs in more capital-intensive industries (automotive and aluminum).
  - Potential improvements: integrate SEZs into national/regional strategies; link SEZ investment to domestic firms; improve infrastructure and energy; develop training aligned with SEZ labor needs; promote joint ventures and compliance with global standards.

### International initiatives supporting private investment
- Belt and Road Initiative (BRI):
  - Expected to raise up to $1 trillion in financing from China over 10 years, mainly for infrastructure development.
  - Specific plans for sub-Saharan Africa include transport and energy infrastructure and more SEZs; Kenya has been a focus.
  - At the 2015 Forum on China-Africa Cooperation (FOCAC), China more than doubled its pledges ($60 billion) in both project finance and technical assistance to support Africa’s development.
- G20 Compact with Africa (CwA):
  - Launched in early 2017 to coordinate efforts and facilitate private investment projects with participation from African Development Bank, IMF, World Bank, and countries.
  - Monitoring across three pillars: macroeconomic framework; business framework; financing framework.
  - Eight sub-Saharan African countries joined the CwA: Benin, Côte d’Ivoire, Ethiopia, Ghana, Guinea, Rwanda, Senegal, and Togo.
  - Country-specific priorities include renewable energy and energy efficiency (Ghana); promoting private activity and electricity capacity (Côte d’Ivoire); investor-friendly tax regime and access to finance (Rwanda); regional development poles and export-oriented industry (Senegal); export-oriented industrialization and industrial parks (Ethiopia); policy matrices and prospectuses (Togo, Benin, Guinea).
  - Progress on actual reforms is mixed across participating countries.

### Public investment efficiency, institutions, and PIMA findings
- Public investment efficiency gap:
  - Public investment efficiency in sub-Saharan Africa could be improved by about 35 percent relative to peers.
  - Infrastructure quality perception scores are below regional peers for electricity supply, roads, and railroads.
- Average efficiency scores (Hybrid Indicator):
  - Commonwealth of Independent States: 0.788
  - Emerging and Developing Asia: 0.659
  - Emerging and Developing Europe: 0.727
  - Latin America and the Caribbean: 0.709
  - Middle East, North Africa, Afghanistan, and Pakistan: 0.676
  - Sub-Saharan Africa: 0.642
- Subgroup Hybrid Indicator (selected):
  - CEMAC: 0.511
  - EAC: 0.735
  - WAEMU: 0.619
  - Oil exporters: 0.269
  - Non-resource-intensive countries: 0.698
  - Other resource-intensive countries: 0.656
- Determinants:
  - Cross-country regressions (2000–15) indicate the quality of institutions is the most important factor explaining investment efficiency.
  - A 10 percent increase in the Control of Corruption Index or the Regulatory Quality Index could lead to a reduction in the efficiency gap of about 12 percent.
- PIMA results:
  - Initial PIMA results (21 pilot countries) show SSA regulatory frameworks are similar to other regions but weaker in effective use for central-local coordination, management of PPPs, project appraisal and selection, project management, and monitoring of assets.
  - The investment–growth relationship is stronger in “high-efficiency” countries than in “low-efficiency” countries.

### Developing domestic debt markets (Box 3.3)
- Progress and examples:
  - Côte d’Ivoire, Namibia, and Uganda have more than doubled issuance of local currency government bonds, with stock of local currency bonds in these countries now equivalent to 8.5 percent of GDP on average.
  - Average bond-maturity of issuance rose from 1.5 years to 6.4 years; some countries issue local currency bonds at maturities of or over 15 years (Ghana, Kenya, Namibia, Nigeria, Tanzania).
- Preconditions for sustainable bond-market development:
  - Stable political environment; coordination of debt management and monetary policy; clear legal/regulatory framework; medium-term debt management strategy and a publicly available annual borrowing plan; market interest rates; sound financial system; adequate market infrastructure; diversified investor base; sufficient resources for development.
- Benefits and risks:
  - Benefits: complements external funding and bank funding; supports monetary policy implementation; strengthens financial markets; reduces foreign-exchange risks; mobilizes private savings; facilitates longer-term financing for infrastructure.
  - Risks: foreign investor participation can make markets sensitive to global interest rates and prone to booms and busts; examples of nonresident holdings: about 40 percent of domestic government bonds in South Africa; about 50 percent in Ghana; average of 25 percent for emerging market economies.

### Fintech and private investment (Box 3.4)
- Potential contributions:
  - Use mobile platforms to reduce frictions between savers and investors (examples: M-Pesa’s M-Kesho and M-Shwari in Kenya; Zoona and EasyEquities in South Africa).
  - Improve payments, settlement, and clearing systems; facilitate growth of derivatives, bond, and money markets.
  - Technologies explored: distributed ledger technologies; central counterparties; riskless settlement systems.
- Trade-offs and risks:
  - Efficiency gains can increase operational complexity and operational risk.
  - Fintech may complicate AML/CFT compliance and elevate cyber-risk and other vulnerabilities.
- Regional indicators:
  - SSA leads in mobile money accounts relative to other regions (figures presented in the chapter).

### Annex highlights — methodology and robustness
- Real Investment Index (Annex 3.1):
  - R_t computed recursively using country-specific private-investment shares and PPP-GDP weights; procedures apply to control extreme values.
- Econometric approach (Annex 3.2):
  - Final estimation sample: 101 emerging and developing economies, 1980–2015.
  - Estimator: system GMM with two-step procedure and Windmeijer’s finite-sample correction.
  - Lagged dependent variable instrumented with one to two lags; other regressors instrumented with two lags and more; fixed effects and some institutional variables treated as exogenous.
  - Im-Pesaran-Shin test: null that all panels have a unit root is rejected at less than 0.1 percent significance level.
- Baseline regression summary (selected coefficients and diagnostics):
  - Lagged private-investment-to-GDP ratio coefficients range from 0.793*** to 0.880*** across specifications (strong persistence).
  - Public-investment-to-GDP ratio coefficients mostly negative (examples: –0.557**, –1.362**, –0.546**, –0.521**, –0.625**, –0.514***, –0.471**).
  - Real GDP growth coefficients often positive and significant in several specifications (examples: 0.209*, 0.239*, 0.181*, 0.190***, 0.209*); decompositions show larger growth effects in richer countries (interaction 0.321***).
  - Relative price of investment coefficients negative and significant in many specs (examples: –1.516**, –1.273**, –1.751***, –1.219*).
  - Interaction effects: Real GDP growth × regulatory quality (0.425**); Real GDP growth × cost of resolving insolvency (–0.030***); Real GDP growth × high-paved-roads (0.281*); Real GDP growth × high-access-to-electricity (0.332**); Real GDP growth × high-trade-openness (0.257*); Real GDP growth × low-capital-account-openness (0.331*); Real GDP growth × high-financial-development (0.465***).
  - Nonlinearities: Real GDP growth, squared enters negatively and significantly in several specifications (examples: –0.011**, –0.002**, –0.003*), suggesting diminishing marginal effects of growth on private investment in some models.
  - Observations and instruments vary by model (observations examples: 1,623; 2,194; 2,432; number of countries examples: 100; 101; 99; Hansen test p-values generally indicate instrument validity across models).

### Conclusions and consolidated policy recommendations
- Main conclusions:
  - Sub-Saharan Africa needs to increase private investment to achieve social and development objectives.
  - Private-investment-to-GDP ratios in sub-Saharan Africa remain the lowest compared with other countries at similar levels of economic development.
  - Public investment can support private investment but can also, in specific circumstances, crowd out private investment.
- Policy recommendations:
  - Create a favorable macroeconomic environment: ensure macroeconomic stability; improve current and prospective economic activity; open to trade; deepen financial systems; build efficient public infrastructure.
  - Strengthen institutional environment: strengthen judicial, regulatory, and insolvency frameworks; resolve long-standing conflicts.
  - Mitigate public investment crowd-out risks: promote alternative sources of financing for public and private investment, including deepening domestic financial markets and PPPs; ensure associated risks are well managed.
  - Promote FDI and improve SEZ policy design as described above.

*This chapter was prepared by a team led by Jesus Gonzalez-Garcia and composed of Romain Bouis, Paolo Cavallino, Nkunde Mwase, Hector Perez-Saiz, Ludger Wocken, and Mustafa Yenice.*

### 3. Private Investment to Rejuvenate Growth

### 3. Private Investment to Rejuvenate Growth

### Private investment trends
- Private investment in sub-Saharan Africa is, on average, 2 percent of GDP lower than in other developing economies.
- Private investment averaged 15 percent of GDP during 2010–16, compared with 22 percent for developing economies in Asia, 18 percent in Europe, 17 percent in Latin America, and 16 percent in the Middle East and North Africa (MENA).
- The gap in private investment relative to other regions has fallen by half since the early 2000s, driven by a decade of rapid growth when private investment grew at an average rate of 14 percent a year.
- Since 2010, private investment slowed, growing on average at 5 percent a year through 2014 and contracting during 2015–16.
- Investment in the region contracted by 4 percent each year on average in 2015–16.
- The decline in private investment was widespread: private investment slowed in two-thirds of countries and fell in half of them.
- Oil exporters have the lowest levels of private investment to GDP, averaging 14 percent over 2010–16, compared with 17 percent in other resource-intensive countries and 15 percent in non-resource-intensive countries.
- Weaker private investment has weighed on GDP growth. In oil-exporting countries, declining private investment’s negative impact was compounded by sharp cuts in public investment. In other countries, weaker private investment was partly offset by more public investment, although high debt levels and rising debt servicing costs are constraining fiscal space.

### Drivers and country experiences
- Commodity prices: Elevated commodity prices supported large increases in private investment in commodity-exporting countries (example: Nigeria during 2007–14); discoveries of natural resources also supported investment (Equatorial Guinea, Ghana). Conversely, some commodity importers benefited from lower commodity prices that created fiscal space for investment (Rwanda).
- Resolution of conflicts: The end of long-standing conflicts in Côte d’Ivoire, Ethiopia, Rwanda, and Uganda was followed by marked increases in private investment (post-conflict 10-year averages exceed conflict-period averages for these countries).
- Idiosyncratic and spillover shocks: Country-specific shocks and adverse spillovers from large regional economies (Angola, Nigeria, South Africa—combined GDP weight about 50 percent of the region) contributed to investment declines. Examples: policy and political uncertainty in South Africa; sharp slowdown in credit growth in Kenya; completion of a large mining project in Namibia.

### Empirical determinants (panel estimation, 101 EMDEs, 1980–2015)
- Methodology notes:
  - Dependent variable: private-investment-to-GDP ratio.
  - Estimation method: system generalized method of moments (system GMM) to address endogeneity.
  - Controls include real GDP growth, public investment as a share of GDP, GDP per capita (PPP), relative price of capital, real interest rate, lagged private-investment-to-GDP ratio, and structural/institutional variables (regulatory quality, insolvency costs, infrastructure, trade openness, financial development, capital account openness).
- Key empirical findings:
  - Real GDP growth raises private investment (accelerator effect). The impact is nonlinear: private investment increases when real GDP growth is high (above the country historical average), but not when growth is low (below the country historical average).
  - Public investment has an ambiguous effect: it can either complement private investment (by providing infrastructure) or crowd out private investment (by competing for scarce financial resources or creating supply bottlenecks).
  - Higher relative price of investment reduces private investment ratios. The level of GDP per capita and the real interest rate were not significant in the baseline results.
  - Institutional and structural characteristics strengthen the effect of GDP growth on private investment:
    - Better regulatory quality and lower insolvency costs increase the responsiveness of private investment to growth.
    - Better public infrastructure (higher proportion of paved roads; greater access to electricity) increases private sector investment responses to growth.
    - Greater trade openness raises the investment reaction to economic activity.
    - Less open capital accounts are associated with a stronger impact of GDP growth on investment (the mechanisms are not fully obvious; prior literature offers competing explanations).
    - Higher financial development significantly raises the responsiveness of private investment to GDP growth; very low levels of financial development can be a binding constraint such that firms do not invest even when demand strengthens.
  - Persistence: investment ratios display persistence, motivating inclusion of lagged private-investment-to-GDP.

### Magnitudes: impact of a 1 percentage point increase in GDP growth on the private investment ratio (Table 3.1)
- Whole Sample: 0.21
- Low Regulatory Quality (SSA average) – High Regulatory Quality (non-SSA EMDEs average): 0.29 – 0.48
- High Insolvency Cost (SSA average) – Low Insolvency Cost (non-SSA EMDEs average): 0.02 – 0.24
- Higher Proportion of Paved Roads: 0.28
- Higher Access to Electricity: 0.33
- Higher Trade Openness: 0.26
- Lower Capital Account Openness: 0.33
- Higher Financial Development: 0.47

### Policy-relevant implications and approaches to alleviate constraints
- Promote stronger and sustainable economic activity to trigger the accelerator effect on private investment.
- Improve regulatory quality and insolvency/resolution frameworks to amplify the investment response to growth.
- Increase the quality and efficiency of public infrastructure (paved roads, electricity access) to encourage private capital formation.
- Promote trade openness to raise incentives for firms to invest for export markets.
- Foster financial development and deepen domestic financial markets to remove financing constraints; in countries with very low financial development, demand-side recovery alone may not translate into investment.
- Manage public investment carefully: while public investment can support private investment by improving infrastructure, policymakers should be mindful of potential crowding out when public investment competes with private investment for scarce funds or faces binding supply constraints.
- Diversify sources of financing: promote alternative financing mechanisms for public and private investment (including public–private partnerships) while ensuring associated risks are well managed.
- Attract FDI and consider Special Economic Zones (SEZs) as potential tools to boost private investment, recognizing mixed historical experiences with SEZs.

_This chapter was prepared by a team led by Jesus Gonzalez-Garcia and composed of Romain Bouis, Paolo Cavallino, Nkunde Mwase, Hector Perez-Saiz, Ludger Wocken, and Mustafa Yenice._

### 3. PRIVATE INVESTMENT TO REJUVENATE GROWTH

### 3. PRIVATE INVESTMENT TO REJUVENATE GROWTH

### Public investment: complementarities and crowding out
- Public investment can be complementary to private investment when it provides infrastructure or goods that raise the productivity of private capital.
- Public investment can crowd out private investment through:
  - Competing for scarce physical and financial resources via debt issuance, bank credit, higher taxes, or inflation.
  - State enterprises producing output in direct competition with private sector goods and services.
  - Increased macroeconomic instability when public investment is financed through accumulation of unsustainable debt.
- Empirical finding: the effect of public investment on private investment depends on the degree of financial development (proxied by the Financial Development Index).
  - Given observed regional levels of financial development, a 1 percentage point increase in the public investment ratio would lead to a ½ percentage-point contraction of the private investment ratio in the average sub-Saharan African country.
  - The same 1 percentage point increase would lead to a ½ percentage-point increase in private investment in other emerging market and developing economies in the sample (which are on average much more financially developed than sub-Saharan African countries).
- Policy implication: because of low financial development, large infrastructure gaps, scarce resources, and constraints on availability of foreign financing in sub-Saharan Africa, public investment can potentially crowd out private investment unless alternative financing and institutional measures are pursued.

### Alleviating constraints to private investment — Deepening financial systems
- Evidence that availability of and access to credit are major constraints in sub-Saharan Africa:
  - Bank financing of investment in sub-Saharan Africa is the lowest among regions, while equity financing is the highest.
  - Sub-Saharan Africa has the lowest share of firms that did not apply for a loan because they did not need it and the highest number of firms identifying access to finance as a major constraint.
  - Small and medium-sized firms face greater obstacles to obtaining financing than larger firms.
- Financial structure and indicators:
  - Banking systems dominate; stock exchanges and bond markets remain underdeveloped but have been expanding rapidly.
  - Banking systems in sub-Saharan Africa are characterized by relatively high capital ratios compared with other regions.
  - Negative association observed between regulatory capital ratios and credit availability to firms (see Figure 3.9 regression): y= –1.17**x+ 41.75.
  - Positive relationship between z-score (safety and soundness of the banking system) and indicators of credit to the private sector (see Figure 3.12 regression): y= 1.05***x+ 7.10.
  - z-score definition: Z = (ROA +(equity/assets))/(ROA standard deviation).
- Instruments and prerequisites for deeper financial markets:
  - Develop bond markets (registries, central depositories, clearing and settlement systems); ensure a large heterogeneous investor base; maintain a sound banking system and market-determined interest rates.
  - For equity markets, regional integration of stock exchanges can enhance liquidity and efficiency and bring economies of scale.
  - Improve judicial independence, strengthen investor protection and auditing standards, and reduce constraints in financial market infrastructures.
- Cautions:
  - Financial deepening should proceed cautiously to reduce risks of financial instability; stressed financial systems supply less credit to the private sector.
  - Strengthening institutions and promoting sound judicial, regulatory, and supervisory frameworks is necessary.
  - Fintech could provide leapfrogging opportunities for greater industry efficiency, with positive effects on financial depth and inclusion.

### Public-Private Partnerships (PPPs)
- Theoretical benefits of PPPs:
  - Improve infrastructure quality, bring private sector expertise, and alleviate some financial constraints by expanding financing options.
  - Broad definition: long-term contracts where the private sector carries a significant portion of risks and receives future income streams; private party typically finances, designs, builds, and operates the asset.
- Global and regional experience:
  - Global experience does not uniformly support that PPPs deliver infrastructure more efficiently than public procurement.
  - PPPs imply complex arrangements and fiscal risks that are difficult to evaluate without proper institutional and legal frameworks.
- PPP intensity and sectoral composition in sub-Saharan Africa:
  - Sub-Saharan Africa has the highest average ratio of PPP projects to GDP in the world since 2000: average ratio of 1.4 percent of GDP, compared with 1 percent of GDP in other regions.
  - Distribution within sub-Saharan Africa (average 2000–16):
    - Non-resource-intensive countries: PPPs represented 2¼ percent of GDP on average.
    - Non-oil resource-intensive countries: 1¾ percent of GDP on average.
    - Oil-exporting countries: 1¼ percent of GDP on average.
  - PPPs are mainly concentrated in the energy and transportation sectors; in the last five years projects in the energy sector represented the largest share of total PPPs.
  - ICT projects are often developed under modalities that are not strictly PPPs (limited risk sharing).
- Successful examples and scale:
  - South Africa: 60 power purchase agreement projects over three years, total commitment of 118 billion rand (about 2½ percent of 2017 GDP).
  - SANRAL concessioned 1,288 km of its 19,700-km road network under long-term PPP-type concessions.
- Risks and fiscal management:
  - Since 2006, the value of disputed PPP projects in sub-Saharan Africa averaged ¾ percent of GDP—the highest ratio among emerging market and developing economies.
  - Higher rates of disputed contracts and lower selection quality are related to weaker institutions in public investment management.
  - Fiscal risks of PPPs include bypassing budget constraints, need for public support (capital grants), government-provided debt or revenue guarantees (contingent liabilities), and long-term rigid payment commitments.
- Tools and capacity building:
  - PPP Fiscal Risk Assessment Model (P-FRAM) developed by IMF and World Bank to evaluate potential fiscal costs and risks, including sensitivity analysis; pilots conducted in Côte d’Ivoire, Mauritius, and Niger.
  - Public Investment Management Assessments (PIMA) by IMF and World Bank identify weaknesses in public investment practices; PIMA evaluations have been conducted in Botswana, Burkina Faso, Cameroon, Côte D’Ivoire, Ghana, Liberia, Madagascar, Mauritius, Mozambique, Togo, and Zambia.
- Policy implication: PPPs can be useful financing instruments but require robust institutional frameworks, legal and regulatory arrangements, public investment management capacity, and fiscal risk mitigation measures.

### Foreign Direct Investment (FDI)
- Role of FDI:
  - FDI complements domestic resources and contributes through resource expansion, knowledge and technology transfer.
- Regional outcomes:
  - Over the past decade, sub-Saharan Africa has been the main recipient of FDI in percent of GDP among emerging market and developing regions.
  - Ratio of FDI to GDP over the past decade has averaged slightly above 5 percent in sub-Saharan Africa, higher than Latin America and the Caribbean; other regions show ratios ranging from 2.5 to 4 percent.
- Cross-country variation:
  - FDI-to-GDP ratios tend to be concentrated in some countries, not only resource-intensive ones.
  - Countries with ratios since 2000 well above the regional average of about 4 percent include Cabo Verde, Mauritius, Mozambique, Seychelles, São Tomé and Príncipe, and The Gambia.

*Source: chap3 - 3. PRIVATE INVESTMENT TO REJUVENATE GROWTH (PDF chapter).*

### 1. Disputed Projects

### 1. Disputed Projects

### Disputed and Cancelled Public-Private Partnerships (PPPs)
- Figure 3.16 presents "Selected Regions: Disputed and Cancelled Public-Private Partnerships to GDP, 2006–16" with percent-of-GDP scale from 0 to 1 and regional labels including SSA, MENA, Asia, LAC, EURCIS.
- Source for PPP dispute data: World Bank, Private Participation in Infrastructure Project database.
- Note: EURCIS = Europe and Commonwealth of Independent States; LAC = Latin America and the Caribbean; MENA = Middle East and North Africa; SSA = sub-Saharan Africa.

---

### Benchmarking of PPP Management

### Key findings and implications from benchmarking
- Disputed and cancelled PPPs are quantified relative to GDP across regions (Figure 3.16). Comparative benchmarking is based on World Bank project database indicators.
- Improving PPP project appraisal, selection, monitoring, and risk management is implied by broader recommendations on public investment efficiency (see Box 3.2).

---

### Private Investment to Rejuvenate Growth

### Foreign Direct Investment (FDI) trends and country variation
- Figure 3.18 shows "Selected Regions: Foreign Direct Investment (Three-year averages)" with Percent of GDP on the vertical axis across periods 1999–01, 2002–04, 2005–07, 2008–10, 2011–13, 2014–16 for SSA, Asia, EURCIS, LAC, MENA.
- Figure 3.19 shows "Sub-Saharan Africa: Foreign Direct Investment by Country, Average 2000–16" with percent-of-GDP values for individual countries and classifications: Oil exporters; Other resource-intensive countries; Non-resource-intensive countries; Average.
- Two-thirds of the countries in the region show FDI ratios below the regional average.

### Determinants of FDI (literature synthesis)
- Factors that help attract FDI include:
  - large domestic markets and natural resources
  - the provision of infrastructure
  - the level of education of the labor force
  - openness to trade
  - macroeconomic and political stability
  - the quality of institutions
- Policymakers could foster stronger FDI inflows into sub-Saharan Africa by improving macroeconomic and political stability; providing better infrastructure services and a more skilled labor force; and improving the institutional environment.
- Note: some countries have other important sources of financial flows (portfolio and loans), including Kenya, Senegal, and South Africa.

### Special Economic Zones (SEZs)
- SEZs are a second-best solution compared with economy-wide reforms but can play a catalytic role in promoting structural transformation.
- Sub-Saharan Africa’s SEZ experience over the past two decades has been mixed; most have been unsuccessful or fallen short of expectations.
- Reasons for weak performance include reliance primarily on corporate tax holidays with little else in nontax incentives or regulatory support; taxes are not the only factor in investment location decisions.
- Recent more positive experiences include Rwanda and Ethiopia, with a focus on developing clusters, fostering competition and quality improvements, and relying on comparative advantages.
- Industries targeted by SEZs in the region include apparel, textile, and agroprocessing (Ethiopia, Ghana, Kenya, Madagascar, Malawi, Mauritius, Seychelles, Zimbabwe). Few economies (Mozambique, Namibia, Nigeria, South Africa, Zambia) have established SEZs in more capital-intensive industries (automotive and aluminum).
- Potential ways to increase SEZ effectiveness:
  - integrate SEZ programs into national and regional development strategies
  - promote investments better linked to domestic firms
  - encourage stronger ownership by foreign investors
  - improve provision of infrastructure and energy
  - promote joint ventures between local corporations and foreign investors
  - develop training and education aligned with SEZ labor requirements
  - improve compliance with global production and environmental standards
  - ensure SEZs catalyze transformation of the broader economy

### International initiatives supporting private investment
- Belt and Road Initiative (BRI)
  - A framework to connect China with south, central, and west Asia, Europe, and Africa through trade, infrastructure, investment, and finance.
  - Expected to raise up to $1 trillion in financing from China over 10 years, mainly for infrastructure development.
  - Specific plans for sub-Saharan Africa include transport and energy infrastructure and more SEZs; Kenya has been a focus (maritime ports and railways). Ethiopia, Mozambique, South Africa, and Tanzania are seeking active involvement.
  - At the 2015 Forum on China-Africa Cooperation (FOCAC), China more than doubled its pledges ($60 billion) in both project finance and technical assistance to support Africa’s development.
- G20 Compact with Africa (CwA)
  - Launched in early 2017 with cooperation from the G20, African Development Bank, IMF, World Bank, and participating countries to coordinate efforts and facilitate private investment projects.
  - A G20-supported monitoring mechanism (with IMF and World Bank support) will assess progress on commitments across three pillars:
    - macroeconomic framework: maintaining macroeconomic stability while providing for adequate investment in infrastructure
    - business framework: making countries more attractive for private investors
    - financing framework: increasing availability of financing with reduced costs and risks
  - Eight sub-Saharan African countries joined the CwA: Benin, Côte d’Ivoire, Ethiopia, Ghana, Guinea, Rwanda, Senegal, and Togo (and three more in the rest of Africa).
  - Country-specific priorities under the CwA:
    - Ghana: renewable energy and energy efficiency; structural reform of the energy sector including debt restructuring and privatization plans; in-depth assessment of private sector opportunities and constraints (IFC 2017).
    - Côte d’Ivoire: promoting private activity and employment; increasing electricity capacity while maintaining financial sustainability; projects to support value addition in cocoa industry.
    - Rwanda: investor-friendly tax regime without eroding tax base; strengthen government responsiveness to private sector; instruments to ease access to finance; improve coordination between national development authorities; quarterly investor roundtable; investor response mechanism.
    - Senegal: develop regional development poles with special economic development zones; promote export-oriented industry and job creation for youth and women.
    - Ethiopia: align CwA with its growth and transformation plan; priorities include export-oriented industrialization, industrial parks, and plug-and-play business environments.
    - Togo: joined after preparing policy matrix and investment prospectus to improve private investment conditions.
    - Benin and Guinea: developing policy matrices and implementation requirements; bilateral G20 partner involvement is under preparation.
  - Progress on actual reforms is mixed; participating countries are at various stages and some joined only recently.

---

### Conclusions and Policy Recommendations

### Main conclusions
- Sub-Saharan Africa needs to increase private investment to achieve social and development objectives.
- Private investment has increased since 2000, but private-investment-to-GDP ratios in sub-Saharan Africa remain the lowest compared with other countries at similar levels of economic development.
- Many countries engaged in large public infrastructure projects to address infrastructure gaps; public investment can support private investment but can also, in specific circumstances, crowd out private investment.

### Policy recommendations
- Create a favorable macroeconomic environment:
  - ensure macroeconomic stability
  - improve current and prospective economic activity
  - open to trade
  - deepen financial systems
  - build efficient public infrastructure
- Strengthen institutional environment:
  - strengthen judicial, regulatory, and insolvency frameworks
  - resolve long-standing conflicts (resolution is typically followed by increases in private investment)
- Mitigate public investment crowd-out risks:
  - promote alternative sources of financing for public and private investment, including deepening domestic financial markets and PPPs
  - ensure associated risks of PPPs and other financing mechanisms are well managed
- Promote FDI and improve SEZ policy design (see SEZ recommendations above)

---

### Box 3.1 — Policy Reform and Private Investment Growth

### Analytical framework and key findings
- Focus: relationship between policy reform (macro stability and institutional reforms) and private investment growth; sample spans 97 emerging market and developing economies over 1996–2015 (excluding populations < 3 million).
- Definitions:
  - Spurts and setbacks in governance follow World Bank (2017).
  - For macroeconomic variables, a spurt (setback) = a two-year decrease (increase) bigger (smaller) than the mean minus (plus) one standard deviation in the public-debt-to-GDP ratio or inflation.
  - Episodes with simultaneous improvements and setbacks across measures are excluded.
- Method: panel regression where dependent variable is real private investment growth; regressors are dummy variables for spurts (t) and setbacks (s) over specified windows; includes time fixed effects and country fixed effects; robust standard errors identified with asterisks.
- Main finding: private investment increases after key improvements in public debt, inflation, and the quality of institutions; setbacks are generally anticipated by investors who curtail investment.

### Event study: selected regression coefficients (Dependent Variable: Private Investment Growth)
- Period t − 1 of reform spurt: 1.15 ; 1.35
- Period t of reform spurt: 1.46 ; 1.23
- Period t + 1 of reform spurt: 2.42 ; 1.29 *
- Period s − 1 of reform setback: −3.99 ; 1.25 ***
- Period s of reform setback: −1.51 ; 1.15
- Period s + 1 of reform setback: 1.89 ; 1.23
- Period s + 2 of reform setback: −0.01 ; 1.10
- Number of observations: 1582
- R-squared: 0.135
- Note: Regression includes country and time fixed effects. Robust standard errors. Significance levels: ***p < 0.01; **p < 0.05; *p < 0.1.

---

### Box 3.2 — Public Investment Efficiency in Sub-Saharan Africa

### Summary and key statistics
- Improving public investment efficiency could contribute to stronger economic growth and achievement of social and development goals.
- Public investment efficiency in sub-Saharan Africa compares unfavorably with other regions and could be improved by about 35 percent.
- Raising efficiency would require improving the quality of institutions and strengthening:
  - planning and selection of PPPs
  - credibility of multiyear budgeting
  - effectiveness of project appraisal and selection
  - monitoring of projects during implementation
  - registration of infrastructure assets
- Infrastructure quality in sub-Saharan Africa (perception measures) is scored below regional peers for electricity supply, roads, and railroads.

### Average efficiency scores by region (Table 3.2.1)
- Commonwealth of Independent States: Physical Infrastructure 0.935 ; Quality Infrastructure 0.716 ; Hybrid Indicator 0.788
- Emerging and Developing Asia: Physical Infrastructure 0.501 ; Quality Infrastructure 0.788 ; Hybrid Indicator 0.659
- Emerging and Developing Europe: Physical Infrastructure 0.753 ; Quality Infrastructure 0.708 ; Hybrid Indicator 0.727
- Latin America and the Caribbean: Physical Infrastructure 0.580 ; Quality Infrastructure 0.769 ; Hybrid Indicator 0.709
- Middle East, North Africa, Afghanistan, and Pakistan: Physical Infrastructure 0.472 ; Quality Infrastructure 0.791 ; Hybrid Indicator 0.676
- Sub-Saharan Africa: Physical Infrastructure 0.460 ; Quality Infrastructure 0.803 ; Hybrid Indicator 0.642

### Average efficiency scores by sub-Saharan groups (Table 3.2.2)
- Sub-Saharan Africa: Physical Infrastructure 0.460 ; Quality Infrastructure 0.803 ; Hybrid Indicator 0.642
- CEMAC: Physical Infrastructure 0.305 ; Quality Infrastructure 0.625 ; Hybrid Indicator 0.511
- EAC: Physical Infrastructure 0.487 ; Quality Infrastructure 0.874 ; Hybrid Indicator 0.735
- WAEMU: Physical Infrastructure 0.369 ; Quality Infrastructure 0.814 ; Hybrid Indicator 0.619
- Oil exporters: Physical Infrastructure 0.196 ; Quality Infrastructure 0.594 ; Hybrid Indicator 0.269
- Non-resource-intensive countries: Physical Infrastructure 0.446 ; Quality Infrastructure 0.858 ; Hybrid Indicator 0.698
- Other resource-intensive countries: Physical Infrastructure 0.602 ; Quality Infrastructure 0.813 ; Hybrid Indicator 0.656

### Determinants of public investment efficiency
- Cross-country regressions (2000–15) indicate the quality of institutions is the most important factor in explaining investment efficiency.
- Explanatory variables include:
  - quality of institutions (World Economic Forum indicators: control of corruption and regulatory quality)
  - official development assistance
  - percentage of urban population
  - dependence on natural resources (dummy for countries rich in nonrenewable resources)

---

*Source: Regional Economic Outlook: Sub‑Saharan Africa (chapter excerpts).*

### 3. PRIVATE INVESTMENT TO REJUVENATE GROWTH

### 3. PRIVATE INVESTMENT TO REJUVENATE GROWTH

### Public investment efficiency, institutions, and natural resources
- Estimates show a positive correlation between public investment efficiency and the quality of institutions and a negative association between dependence on natural resources and public investment efficiency.
- Strengthening institutions could reduce the public investment efficiency gap in sub-Saharan Africa:
  - A 10 percent increase in the Control of Corruption Index or the Regulatory Quality Index could lead to a reduction in the efficiency gap of about 12 percent.

### PIMA (Public Investment Management Assessment) findings for sub-Saharan Africa
- Initial PIMA results (21 pilot countries) indicate sub-Saharan African countries have generally similar regulatory frameworks compared with the average in other regions.
- Comparative strengths (SSA vs. Non-SSA, n = 10 shown in figures):
  - Slightly better frameworks in: national and sectoral planning; multiyear budgeting; project management.
- Comparative weaknesses and gaps in effective use:
  - Weaker regulations in: central-local coordination; management of PPPs; regulation of firms; monitoring of assets.
  - Regulations exist but are not used effectively in: management of PPPs; multiyear budgeting; project appraisal and selection; project management; monitoring of assets.
- Figures referenced:
  - Figure 3.2.2: SSA (n = 10) vs Non-SSA (n = 10) — PIMA regulatory framework scores across 15 areas (Fiscal Rules; National & Sectoral Planning; Central-Local Coordination; Management of PPPs; Company Regulation; Multiyear Budgeting; Budget Comprehensiveness; Budget Unity; Project Appraisal; Project Selection; Protection of Investment; Availability of Funding; Transparency of Execution; Project Management; Monitoring of Assets).
  - Figure 3.2.3: Distinguishes SSA regulatory presence vs SSA regulatory effectiveness across the same 15 areas.

### Investment–growth relationship within SSA
- The relationship between investment and growth is stronger in “high-efficiency” countries than in “low-efficiency” countries (Figure 3.2.4).
- Figure 3.2.4 plots average annual per capita GDP growth (2010–15) against average annual public investment, percent of GDP (2010–15) for SSA countries, distinguishing low-efficiency (blue) and high-efficiency (red) countries.

### Policy implications to improve public investment efficiency
- Potential areas for strengthening public investment management to increase efficiency:
  - Strengthening planning and selection of PPPs.
  - Increasing credibility of multiyear budgeting.
  - Improving effectiveness of project appraisal and selection.
  - Enhancing monitoring of projects during implementation.
  - Improving registration and monitoring of infrastructure assets.

### Developing domestic debt markets in sub-Saharan Africa (Box 3.3)
- Drivers and progress:
  - Governments have increased domestic issuance to finance growing budget deficits because of: limitations of direct banking financing; limited availability of foreign aid/concessional loans; awareness of risks of foreign-currency borrowing.
  - Côte d’Ivoire, Namibia, and Uganda have more than doubled issuance of local currency government bonds, with the stock of local currency bonds in these countries now equivalent to 8.5 percent of GDP on average.
  - Average bond-maturity of issuance rose from 1.5 years to 6.4 years; some countries (Ghana, Kenya, Namibia, Nigeria, Tanzania) issue local currency bonds at maturities of or over 15 years.
- Preconditions for sustainable bond-market development:
  - A stable political environment for credible policymaking.
  - A suitable environment for domestic issuance and effective coordination of debt management and monetary policy.
  - A clear, modern legal and regulatory framework for government securities and market transactions.
  - Adherence to sound debt management policies: medium-term debt management strategy and a publicly available annual borrowing plan.
  - Government commitment to pay market interest rates (avoid creating a captive investor base or intervening to manage yields).
  - A sound financial system; banks typically are initial investors.
  - Adequate market infrastructure for clearing, settlement, and custody.
  - A diversified investor base with varied risk preferences and horizons.
  - Availability of sufficient resources (staff and capacity) for bond market development; acknowledgement of start-up costs (higher yields, greater rollover risk).
- Benefits and risks:
  - Benefits: complements external funding and bank funding; can support monetary policy implementation; strengthens financial markets; reduces foreign-exchange risks; enables private savings mobilization; facilitates longer-term financing for infrastructure.
  - Risks/financial-stability implications: foreign investor participation can diversify investor base and extend maturities but makes markets sensitive to global interest rates and prone to booms and busts.
  - Examples of nonresident holdings: about 40 percent of domestic government bonds in South Africa; about 50 percent of domestic government debt in Ghana; compares with an average of 25 percent for emerging market economies.

### Fintech and private investment (Box 3.4)
- Potential contributions of fintech:
  - Use existing mobile platforms to reduce frictions between savers and investors (e.g., M-Pesa’s M-Kesho and M-Shwari in Kenya; Zoona and EasyEquities in South Africa).
  - Improve infrastructure of financial markets: payments, settlement, and clearing systems—underdeveloped in SSA relative to other regions—thereby reducing systemic, credit, and liquidity risks and facilitating growth of derivatives, bond, and money markets.
  - Technologies explored for efficiency gains: distributed ledger technologies; central counterparties to improve derivatives market functioning; riskless settlement systems to reduce trading frictions.
- Trade-offs and risks:
  - Efficiency gains can increase operational complexity and operational risk.
  - Fintech may complicate compliance (AML/CFT) and elevate cyber-risk and other vulnerabilities.
- Figures:
  - Figure 3.4.1 shows selected regions’ mobile subscriptions per 100 people and mobile money accounts per 1,000 adults; SSA leads in mobile money accounts relative to other regions.

### Annex 3.1 — Real Investment Index: methodology highlights
- Decomposition for each country i and year t:
  - Total annual real investment growth is decomposed into contributions of private and public components using country-specific private-investment shares.
  - The regional total investment growth rate is the purchasing-power-parity GDP-weighted average of country components.
- Real Investment Index R_t computed recursively:
  - R_t = R_{t-1} * (1 + private component growth)^{alpha} * (1 + public component growth)^{1-alpha}, starting from R_{t0} = 1, with alpha equal to the PPP-GDP-weighted average share of private investment across countries.
- Procedures to control for extreme values are applied when computing regional private and public investment growth rates.

### Annex 3.2 — Determinants of private fixed investment ratios: empirical approach
- Baseline dynamic fixed-effects panel specification explains private fixed investment-to-GDP ratio (I/Y) by:
  - Lagged I/Y, public fixed investment-to-GDP ratio (IG/Y), real GDP per capita in PPP (Ypc), relative price of capital (PI/PY), real interest rate (IR), real GDP growth (g), plus country (η_i) and year (γ_t) fixed effects.
- Final estimation sample:
  - 101 emerging and developing economies over the years 1980 to 2015.
  - Countries listed in the sample include Algeria, Angola, Antigua and Barbuda, Argentina, Armenia, Azerbaijan, The Bahamas, Bahrain, Bangladesh, Barbados, Belarus, Belize, Benin, Bhutan, Bolivia, Bosnia and Herzegovina, Botswana, Brazil, Bulgaria, Burkina Faso, Burundi, Cabo Verde, Cameroon, Central African Republic, Chad, Chile, China, Colombia, Comoros, Democratic Republic of the Congo, Republic of the Congo, Costa Rica, Côte d'Ivoire, Croatia, Djibouti, Dominica, Dominican Republic, Ecuador, Egypt, Equatorial Guinea, Ethiopia, Gabon, The Gambia, Ghana, Grenada, Guatemala, Guinea, Guinea-Bissau, Haiti, Honduras, Hungary, India, Indonesia, Iran, Jordan, Kenya, Kuwait, Kyrgyz Republic, Lebanon, Lesotho, Malaysia, Mauritius, Mexico, Moldova, Mongolia, Republic of Montenegro, Morocco, Mozambique, Myanmar, Namibia, Nepal, Nicaragua, Niger, Oman, Panama, Peru, Poland, Romania, Russia, Rwanda, São Tomé and Príncipe, Senegal, Serbia, Seychelles, Sierra Leone, South Africa, Sri Lanka, St. Lucia, St. Vincent and the Grenadines, Suriname, Swaziland, United Republic of Tanzania, Thailand, Togo, Tunisia, Uganda, Ukraine, Uruguay, Venezuela, Yemen, Zambia.
- Estimation methodology:
  - System GMM estimator (Arellano and Bover 1995; Blundell and Bond 1998) with two-step procedure and Windmeijer’s finite-sample correction.
  - Lagged dependent variable treated as predetermined and instrumented with one to two lags; other regressors treated as endogenous and instrumented with two lags and more; fixed effects and some institutional variables treated as exogenous.
  - Instrument validity tested using the Hansen test; number of instruments kept lower than number of countries to limit weakening of the Hansen test.
  - Im-Pesaran-Shin test: null that all panels have a unit root is rejected at less than 0.1 percent significance level.

*Source: chap3 - 3. PRIVATE INVESTMENT TO REJUVENATE GROWTH*

### Annex 3.2. Determinants of Private Fixed Investment Ratios in Emerging and Developing

### Annex 3.2. Determinants of Private Fixed Investment Ratios in Emerging and Developing Economies

### Econometric methodology and data
- Model specification:
  - Dynamic fixed-effects panel data equation with the private-fixed-investment-to-GDP ratio (I/Y) explained by its lagged value and traditional determinants: public-fixed-investment-to-GDP ratio (IG/Y), real GDP per capita in purchasing power parity (Ypc), the ratio of the deflator of gross fixed investment to the GDP deflator (PI/PY), the real interest rate (IR), and real GDP growth (g).
  - Country fixed effects (ηi) and year fixed effects (γt) included; εi,t is the error term.
  - Final estimation sample: 101 emerging and developing economies over the years 1980 to 2015.
- Estimator and diagnostics:
  - System GMM estimator (Arellano and Bover 1995; Blundell and Bond 1998) used to address Nickell (1981) bias and endogeneity.
  - GMM regressions performed using the two-step procedure with Windmeijer’s finite-sample correction.
  - Lagged dependent variable treated as predetermined and instrumented with one to two lags.
  - Other regressors treated as endogenous and instrumented with two lags and more; fixed effects and some institutional variables treated as exogenous.
  - Validity of instruments tested using the Hansen test, with the number of instruments kept lower than the number of countries as suggested by Roodman (2009).
  - Unit-root testing: The null hypothesis of the Im-Pesaran-Shin test that all panels have a unit root is rejected at less than 0.1 percent significance level.
  - Serial correlation: Absence of serial correlation of residuals tested using the AR(2) test; AR(1) test is rejected in all regressions, indicating first-order serial correlation of the differenced error term.

### Baseline regression findings (summary of main effects)
- Persistence:
  - The private-investment-to-GDP ratio is persistent. Lagged private-investment-to-GDP ratio coefficients (across columns): 0.793***, 0.764***, 0.800***, 0.811***, 0.781***, 0.858***, 0.775*** (t-statistics in parentheses: (10.58), (8.88), (9.69), (11.23), (10.43), (14.30), (10.03)).
- Public investment crowding-out:
  - Public-investment-to-GDP ratio has a significant negative coefficient consistent with crowding-out: –0.557**, –1.362**, –0.546**, –0.521**, –0.625**, –0.514***, –0.471** (t-statistics: (–2.45), (–2.37), (–2.48), (–2.36), (–2.03), (–2.76), (–2.53)).
  - Crowding-out is mitigated when financial development is higher (interaction in column (2)).
- Real GDP growth:
  - Real GDP growth is statistically and economically significant in several specifications: coefficients reported include 0.209*, 0.239*, 0.181*, 0.190***, 0.209* (t-statistics: (1.86), (1.91), (1.83), (2.86), (1.87)).
  - Decomposition: a 1 standard deviation increase in real GDP growth (+6.2 percent) translates into a 1.3 percentage point increase in the investment ratio (as reported).
  - Growth effects are larger in richer countries (those with an average GDP per capita above the sample median $5,072 in 2011 PPP terms); interaction coefficient for high income country: 0.321*** (t-statistic (3.50)); interaction for low income country: –0.263 (t-statistic (–0.87)).
- Relative price of investment:
  - The relative price of investment reduces private investment ratios with coefficients: –1.516**, –0.999, –1.273**, –1.751***, –1.219*, –0.761, –1.354** (t-statistics: (–2.39), (–1.32), (–2.07), (–2.70), (–1.82), (–1.23), (–2.00)).
- Other controls:
  - Real GDP per capita in logs: coefficients reported 2.885, –0.406, 0.691, 1.777, 1.689, 1.821, 2.071 (t-statistics: (0.96), (–0.13), (0.30), (0.55), (0.81), (1.15), (0.48)) — not statistically significant in baseline specifications.
  - Real interest rate: coefficients –0.034, –0.022, –0.037, –0.025, –0.013, 0.045, –0.031 (t-statistics: (–1.01), (–0.72), (–0.86), (–0.89), (–0.44), (0.72), (–1.06)) — not statistically significant.
  - Financial Development Index level: –25.888 (t-statistic (–1.52)) in column (2) not significant by itself.
  - Financial development interacts positively with public investment ratio: coefficient 5.946** (t-statistic (2.02)).
  - Trade openness and capital account openness not significant in baseline columns reported (trade openness coefficient 0.027 (1.13); capital account openness 0.096 (0.29)).
- Robustness and specification metrics:
  - Observations by column: 2,194; 2,185; 2,194; 2,194; 2,194; 2,432; 1,185 (as listed in the table under Observations — note: exact column alignment preserved from source).
  - Number of countries by column: 101; 100; 101; 101; 101; 99; 100.
  - Number of instruments by column: 51; 58; 54; 54; 60; 69; 54.
  - AR(2) test p-values: 0.693; 0.544; 0.603; 0.687; 0.835; 0.994; 0.662.
  - Hansen test p-values: 0.303; 0.305; 0.347; 0.237; 0.097; 0.146; 0.17.

### Nonlinearities, interactions with institutions, and country-group heterogeneity
- Nonlinear GDP growth effect:
  - Decomposing GDP growth into low and high levels indicates a nonlinear effect (column (3)); definitions: for each country, real GDP growth is considered high (low) if it is above (below) the country-specific historical mean measured over the estimation period.
  - Coefficients reported for low and high growth: Low real GDP growth 0.372 (t-statistic (1.25)); High real GDP growth 0.228* (t-statistic (1.90)).
- Interactions with institutional indicators:
  - Using World Bank Doing Business, Worldwide Governance, and International Country Risk Guide indicators (available only from end-1990s/mid-2000s) reduces sample size.
  - Effect of GDP growth on investment is larger when regulatory quality is higher and when the cost of resolving insolvency (as percent of real estate property value of the firm) is lower (columns (1) and (2) in Annex Table 3.2.2).
  - Caution: regulatory quality indicator is perception-based; results interpreted with caution.
- Country-group classification (following Servén 2003):
  - Countries classified as high/low in infrastructure (paved roads and access to electricity), trade openness, financial development, and capital account openness according to whether country-average level is above/below sample median; allows different GDP growth coefficients across groups.
  - Results indicate a positive effect of GDP growth on investment in groups with:
    - High levels of paved roads (column (3)),
    - High access to electricity (column (4)),
    - High trade openness (column (5)),
    - Low capital account openness (column (6)),
    - High level of financial development (column (7)).
  - Regressions include square of real GDP growth to control for potential correlation between better institutions and higher growth; inclusion does not alter significance of interaction coefficients.

### Additional notes on controls considered
- Other control variables tested but not significant and not altering main results include:
  - Inflation, real effective exchange rate index, terms of trade, oil prices interacted with an oil-exporter dummy, foreign direct investment, estimates of stocks of public and private capital, public consumption as share of GDP, public external debt, and current account as share of GDP.

*Source: Annex 3.2. Determinants of Private Fixed Investment Ratios in Emerging and Developing Economies (chap3 - Annex 3.2).*

### Annex Table 3.2.1. Determinants of Private Investment Ratios in Emerging Market and Developing Economies:

### Annex Table 3.2.1. Determinants of Private Investment Ratios in Emerging Market and Developing Economies:

### Baseline regressions (Arellano and Bond system-GMM)
- Dependent Variable: Private-Investment-to-GDP Ratio.
- Models reported: (1) through (7). Year fixed effects: Yes (all models).
- Lagged dependent variable (Private investment-to-GDP ratio, one-year lagged):
  - (1) 0.878*** (14.11)
  - (2) 0.877*** (7.78)
  - (3) 0.824*** (9.92)
  - (4) 0.879*** (14.53)
  - (5) 0.867*** (11.52)
  - (6) 0.880*** (12.97)
  - (7) 0.873*** (11.98)
- Public-investment-to-GDP ratio:
  - (1) -0.508** (-1.84)
  - (2) -0.414** (-2.28)
  - (3) -0.528*** (-2.70)
  - (4) -0.546** (-2.33)
  - (5) -0.488** (-2.17)
  - (6) -0.340* (-1.76)
  - (7) -0.471** (-2.14)
- Real GDP per capita in logs:
  - (1) 1.465 (1.63)
  - (2) -0.480 (-0.18)
  - (3) 1.324 (0.52)
  - (4) 3.017 (0.85)
  - (5) 1.458 (0.98)
  - (6) 0.767 (0.29)
  - (7) 3.665 (0.92)
- Relative price of investment in logs:
  - (1) 0.193 (0.12)
  - (2) -1.127 (-1.41)
  - (3) -1.322 (-1.41)
  - (4) -1.203 (-1.57)
  - (5) -1.001 (-1.43)
  - (6) -0.530 (-0.78)
  - (7) -1.219 (-1.37)
- Real interest rate:
  - (1) 0.211 (2.14)
  - (2) 0.036 (0.19)
  - (3) -0.059* (-1.66)
  - (4) -0.001 (-0.01)
  - (5) -0.008 (-0.25)
  - (6) 0.028 (0.59)
  - (7) 0.008 (0.24)
- Real GDP growth:
  - (6) 0.542* (1.67)
  - (7) 0.692** (2.00)
- Regulatory quality (single coefficient reported):
  - -3.566 (-1.37)
- Real GDP growth × regulatory quality:
  - 0.425** (2.39)
- Cost of resolving insolvency (% of estate):
  - 0.026 (0.27)
- Real GDP growth × cost of resolving insolvency:
  - -0.030*** (-3.35)
- High-paved-roads country:
  - 4.840 (1.09)
- Real GDP growth × low-paved-roads country:
  - 0.215 (1.29)
- Real GDP growth × high-paved-roads country:
  - 0.281* (1.93)
- High-access-to-electricity country:
  - -2.936 (-0.64)
- Real GDP growth × low-access-to-electricity country:
  - 0.107 (0.46)
- Real GDP growth × high-access-to-electricity country:
  - 0.332** (1.99)
- Trade openness:
  - 0.017 (1.01)
- Real GDP growth × low-trade-openness country:
  - 0.262 (1.09)
- Real GDP growth × high-trade-openness country:
  - 0.257* (1.78)
- Capital account openness:
  - 0.111 (0.21)
- Real GDP growth × low-capital-account-openness country:
  - 0.331* (1.77)
- Real GDP growth × high-capital-account-openness country:
  - 0.217 (1.47)
- Financial Development Index:
  - -10.365 (-0.98)
- Real GDP growth × low-financial-development country:
  - -0.119 (-0.74)
- Real GDP growth × high-financial-development country:
  - 0.465*** (3.28)
- Real GDP growth, squared:
  - (1) -0.011** (-2.18)
  - (2) -0.001 (-0.70)
  - (3) -0.002** (-2.09)
  - (4) -0.003* (-1.95)
  - (5) 0.000 (0.32)

### Sample sizes, instruments, and specification tests
- Observations by model:
  - (1) 1,623
  - (2) 778
  - (3) 2,113
  - (4) 2,194
  - (5) 2,194
  - (6) 1,432
  - (7) 2,185
- Number of countries by model:
  - (1) 100
  - (2) 89
  - (3) 98
  - (4) 101
  - (5) 101
  - (6) 99
  - (7) 100
- Number of instruments by model:
  - (1) 45
  - (2) 32
  - (3) 59
  - (4) 60
  - (5) 58
  - (6) 60
  - (7) 60
- AR(2) test p-value by model:
  - (1) 0.979
  - (2) 0.863
  - (3) 0.407
  - (4) 0.743
  - (5) 0.944
  - (6) 0.939
  - (7) 0.591
- Hansen test p-value by model:
  - (1) 0.425
  - (2) 0.402
  - (3) 0.364
  - (4) 0.210
  - (5) 0.696
  - (6) 0.216
  - (7) 0.392

### Key empirical findings (from coefficients and interactions)
- Strong persistence in private investment: lagged private-investment-to-GDP ratio coefficients are large and statistically significant in all models (coefficients range from 0.824*** to 0.880***).
- Public investment shows a negative association with private investment in most specifications (coefficients range from -0.546** to -0.340*).
- Real GDP growth has a positive and statistically significant association with private investment in models (6) and (7) (0.542* and 0.692**).
- Interaction effects indicate that the growth impact on private investment is conditional on institutional and structural characteristics:
  - Positive and significant interaction: Real GDP growth × regulatory quality (0.425**).
  - Negative and significant interaction: Real GDP growth × cost of resolving insolvency (-0.030***).
  - Positive and significant interactions with infrastructure and openness measures:
    - Real GDP growth × high-paved-roads country: 0.281*.
    - Real GDP growth × high-access-to-electricity country: 0.332**.
    - Real GDP growth × high-trade-openness country: 0.257*.
    - Real GDP growth × low-capital-account-openness country: 0.331*.
    - Real GDP growth × high-financial-development country: 0.465***.
- Nonlinear growth effect: Real GDP growth, squared enters negatively and significantly in several specifications (e.g., -0.011**, -0.002**, -0.003*), suggesting diminishing marginal effects of growth on private investment at higher growth rates in some models.

### Variable definitions and data sources (selected)
- Private fixed gross capital formation (percent of GDP): IMF, World Economic Outlook database; United Nations National Accounts.
- Public gross fixed capital formation (percent of GDP): IMF, World Economic Outlook database; United Nations National Accounts.
- Real GDP growth: IMF, World Economic Outlook database; United Nations National Accounts.
- Real GDP per capita, in purchasing power parity: IMF, World Economic Outlook database; United Nations National Accounts.
- Relative price of investment (capital formation price index to GDP deflator): Penn World Tables 9.0.
- Real interest rate: World Bank, World Development Indicators.
- Regulatory quality: World Bank, Doing Business Indicator database.
- Cost of resolving insolvency (percentage of business real estate): World Bank, Worldwide Governance Indicators database.
- Roads paved, percent of total roads; Access to electricity, percent of population; Trade openness: World Bank, World Development Indicators; IMF WEO as indicated above.
- De jure financial openness (Chinn–Ito Index): Chinn and Ito (2006), updated July 2017.
- Financial Development Index: Svirydzenka (2016).

*Source: IMF staff calculations. Estimates use the Arellano and Bond system—generalized method of moments estimator. Robust z-statistics in parentheses. *p < .10; **p < .05; ***p < .01.*

---


_Source: https://www.imf.org/-/media/files/publications/reo/afr/2018/may/pdf/chap3.pdf_