## wpiea2021045-print-pdf

## Source details

**Canonical URL:** [wpiea2021045-print-pdf](https://www.imf.org/-/media/files/publications/wp/2021/english/wpiea2021045-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2021/english/wpiea2021045-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2021/english/wpiea2021045-print-pdf.pdf.json)

---

### I. INTRODUCTION
- Urgency: Advancing structural reforms is urgent to support economic recovery and reverse the socio-economic decline associated with Covid-19.
- Historical growth context:
  - SSA 3.8 percent annual growth since the 1970s translated into per capita income gains of 1.2 percent; at those levels per capita income doubles every 58 years.
  - Per capita growth averaged 1.9 percent since 2010.
- Pandemic impact:
  - The pandemic led to an economic contraction in the region, averaging 3 percent, with significant uncertainty surrounding the recovery.
- Paper objectives:
  - Detail channels through which the pandemic is exacerbating structural rigidities and how reforms could support recovery and inclusion.
  - Investigate the main structural rigidities in SSA countries and quantify potential gains from reforms.
  - Discuss implementation considerations based on country experiences with similar reforms.
- Policy context:
  - Economic impact expected to outlast health impact.
  - Demand management policies remain important but policy space is shrinking due to fiscal and debt effects; monetary easing room is constrained.
  - Supply-side constraints can blunt the efficacy of demand management tools.
  - Reforms take time to implement and bear fruit; design and sequencing matter given deteriorated social conditions.

### II. COVID-19, STRUCTURAL RIGIDITIES, AND REFORMS
- Definition and channels:
  - Structural reforms: policies to improve flexibility and efficiency in combining factor inputs and the institutional/regulatory setup (governance).
  - Rigidities arise from government/public entities, private sector behaviors, or weak institutions.
- Why rigidities matter:
  - Inhibit efficient adjustment, contribute to resource misallocation, and can be mutually reinforcing.
  - Product market rigidities: limited competition, inefficient SOEs, over-regulation, price controls — raise markups, undermine entry and innovation.
  - Labor market rigidities: dualization, high unionization, bargaining mechanisms, skills mismatches — rigid wages and hiring/firing rules harm employment creation.
  - Financial market constraints: limited access, high cost, weak regulation — credit constraints for households and SMEs.
  - Public sector constraints: elevated wages, budget rigidity, large footprint — crowding out private sector and reducing spending efficiency.
  - Complementary policy deficits (health, housing, education, infrastructure) hinder human capital accumulation and entrepreneurship.
- Pandemic amplification channels:
  - Households: job and income losses increase food insecurity and poverty; human capital accumulation declines absent effective safety nets; increased reliance on informal lenders and higher indebtedness.
  - Firms: corporate balance sheets fragilized; closures, regulatory forbearance, and bailouts can create “zombie firms” and reduce restructuring incentives.
  - Public sector: challenge of phasing out support while addressing service delivery and governance weaknesses; financing pressures and risk of continued recourse to central bank financing.
- Reform benefits:
  - Enable restructuring, facilitate firm entry and exit, reallocate factors to productive entities, reduce fiscal and financial risks.
  - Distributional trade-offs exist; complementary policies (education, health, transport, housing) help inclusion but effects on inequality are ambiguous.

### III. STYLIZED FACTS AND DATA
- Empirical scope and data:
  - Panel data for 133 low- and middle-income countries (LICs/MICs) for 1996–2017.
  - Structural indicators cover product markets, labor markets, governance, innovation, and complementary factors (education, health, life expectancy, infrastructure).
  - Distance to frontier measured as difference between country score and top performer among upper-MICs (frontier = 95th percentile among upper-MICs).
  - Controls: income level, region, resource intensity (oil, non-oil), fragility; regressions include year and country fixed effects.
- Regional facts (SSA):
  - Growth averaged 3.8 percent since the 1970s—GDP doubling roughly every 19 years.
  - Per capita growth averaged 1.2 percent—doubling roughly every 58 years.
  - Per capita growth in low-income countries averaged 0.5 percent since the 1970s—doubling every 140 years.
  - Median SSA real GDP per capita stood at around 5 percent of US levels in 2019.
  - SSA share in global exports ~1.5 percent (pre-2010 increase largely driven by mineral fuel).
- Post-COVID constraints:
  - Debt deterioration and increasing number of countries moving to debt distress.
  - Rising debt and external financing reliance limit fiscal space and increase vulnerability.
  - Monetary policy scope limited after easing and record low interest rates; some countries constrained by monetary unions or high dollarization.
  - Weak monetary transmission can be worsened by structural rigidities.
- Performance indicators:
  - Global Competitiveness Index (GCI) average for SSA: 3.5.
  - Product and labor market indices average score: 4.

### IV. EMPIRICAL APPROACH — Methodology and Main Results
- Estimation strategy:
  - Extends Prati, Onorato, and Papageorgiou (2013) with broader dataset and reform areas.
  - Estimated growth equation: 풍풏 GDP_i,t − 풍풏 GDP_i,t−1 = a0 + a1 풍풏GDP_i,t−1 + a2 Reform_i,t−1 + μ_i + δ_t + ε_it
  - Reform indices rescaled 1-10; regressions include year and country fixed effects and clustered standard errors by country.
- Reform indices composition:
  - Governance: average of WGI indices (voice and accountability; political stability; government effectiveness; control of corruption; rule of law).
  - Product market: average of WGI regulatory quality and two GCI indicators (goods market efficiency pillar and agricultural policy).
  - Innovation: GCI innovation pillar.
  - Labor market: average of two GCI areas (flexibility of wage determination; cooperation of labor-employer relations).
  - Complementary factors: life expectancy (WDI), primary education enrollment, and quality of infrastructure (GCI).
- Baseline regression selected coefficients (robust standard errors in parentheses):
  - L.GDPpc (log):
    - Column (1): -0.0778*** (0.0139)
    - Column (2): -0.0725*** (0.0119)
    - Column (3): -0.1696*** (0.0539)
    - Column (4): -0.1616*** (0.0487)
    - Column (5): -0.0190*** (0.0047)
    - Column (6): -0.1947*** (0.0590)
  - L.Worldwide Governance Indicator index:
    - Column (1): 0.0188*** (0.0049)
    - Column (2): 0.0270** (0.0104)
  - L.Product Market index:
    - Column (1): 0.0086*** (0.0030)
    - Column (2): 0.0036 (0.0069)
  - L.Innovation index:
    - Column (1): 0.0087* (0.0052)
    - Column (2): 0.0047 (0.0044)
  - L.Labor Market index:
    - Column (1): 0.0028 (0.0042)
    - Column (2): 0.0004 (0.0046)
  - L.Complementary factors index:
    - Column (1): 0.0069*** (0.0023)
    - Column (2): 0.0027 (0.0051)
- Regression sample metrics (by column):
  - Observations: 2,463; 2,475; 1,043; 1,043; 6,036; 1,021.
  - Adjusted R2: 0.0856; 0.0813; 0.141; 0.138; 0.0475; 0.147.
  - No. countries: 131; 133; 104; 104; 133; 102.
- Long-run gains (method):
  - Long-run associated gains computed by multiplying reform coefficient by inverse of (negative of) coefficient on lagged per capita GDP.
  - Improving governance reforms by one point (index 1–10) is associated with an estimated increase in per capita output of about 24 percent in the long run.
  - Complementary factors yield per capita output increasing by about 40 percent in the baseline specification.
- Distance-to-frontier scenario and annualized gains:
  - Frontier defined as 95th percentile among upper-MICs (latest year 2017); gap closed over five years for all reforms except governance (10 years).
  - Cross-country average annual per capita growth gains (SSA, baseline to 95th percentile):
    - Governance, product market, and complementary factor reforms: 1.1-1.3 percentage points per year.
    - Innovation: 0.9 percentage point per year.
    - Labor markets: 0.3 percentage point per year.
  - Less ambitious scenario (reach 75th percentile): overall gains between 0.1-0.9 percentage points per year across reform areas.
  - Heterogeneity:
    - Oil exporters: governance reforms can yield up to 1.8 percentage points per year.
    - LICs tend to exhibit larger gains than MICs.
- Robustness and alternative estimators:
  - Region-time fixed effects and three-year lag intervals produce similar results; three-year lags show larger magnitudes for some reforms.
  - System GMM (robust two-step, collapsed instruments): broadly similar results; passes AR(2) and Hansen tests in reported specifications.
  - Subsample results:
    - Governance particularly important for fragile states.
    - Regulatory quality most important within product market reforms.
    - Quality of infrastructure and life expectancy important within complementary factors.
    - Innovation more important for MICs; governance and product market reforms larger for LICs.

### V. IMPLEMENTATION CONSIDERATIONS
- Political economy and sequencing:
  - Political ownership reduces contestation and reversal risk.
  - Prioritize reforms that deliver visible, immediate gains to build buy-in (e.g., network industries, reductions in banking and telecom prices).
  - Calibrate sequencing to countries’ implementation capacity; address binding constraints first.
  - Combine supply- and demand-side measures to manage distributional impacts (e.g., time-bound payments and training to offset labor reform opposition).
- Complementarities and preconditions:
  - Success requires good macroeconomic policies, governance, reliable infrastructure, and strong human capital.
- Time lags:
  - Product market reform gains tend to peak after 5 years; labor market reforms and innovation can take longer.
- Political sensitivity:
  - Reforms are politically sensitive; less politically sensitive, job-creating reforms offer high growth gains and align with post-pandemic constraints.

### VI. CONCLUSIONS — Quantified Implications for SSA
- Main conclusions:
  - Supply-side reforms can play a major role in aiding recovery and boosting growth by alleviating pandemic-exacerbated structural rigidities for firms, households, and the public sector.
  - Reforms improve medium-term growth, resilience, and diversification, and can support inclusion (distributional effects vary).
  - Priority areas with the largest estimated dividends: governance, product markets, complementary factors (education, health, life expectancy, infrastructure), and innovation.
- Quantified impacts and scenarios:
  - Reforms could boost per capita growth by 0.3–1.3 percentage points over a 5–10-year period; up to 1.8 percentage points for oil exporters.
  - Delivering on these reforms could increase average per capita growth from 1.9 percent in the 2010s to between 2.2 and 3.2 percent.
  - Contextual doubling-time implications:
    - Current per capita growth implies doubling time of around 37 years.
    - Strong reform scenario could reduce doubling time to about 22 years.
  - Regional context note: the region’s 4.4 percent annual average growth since the 2010s has translated into per capita growth of 1.9 percent per year.

### Appendix — Key empirical and data details
- Sample and period:
  - Baseline regressions use annual data for 1996–2017; raw sample of 133 countries.
  - Data sources: WEO June 2020, WDI, WGI, GCI.
- Reform indices rescaled 1–10 for consistency; product market and labor market indicators such as ETCR and EPL are not available for SSA.
- Key robustness / estimator notes (selected table highlights):
  - Table metrics and coefficients summarized in the main regression section above (Observations and No. countries vary by specification).
  - System GMM reported with No. of instruments and AR/Hansen p-values indicating acceptable diagnostics in reported columns.
- Model for dividends (notation):
  - Reform scenario dynamics use ln GDP (y) and Reform (R) with n = 5 (or 10 for governance) to annualize distance-to-frontier closure and compute cumulative per capita level effects via exponentiation of weighted sums of α_2 * (d / n) with (1+α_1) decay factors.

*Source: IMF staff estimates and text from wpiea2021045-print-pdf.*

### 1.3 percentage points, relative to the 1.9 percent average since 2010. Such growth would

### 1.3 percentage points, relative to the 1.9 percent average since 2010. Such growth would

### I. INTRODUCTION
- Advancing structural reforms is urgent to support economic recovery and reverse the socio-economic decline associated with Covid-19.
- Since the 1970s, Sub‑Saharan Africa’s (SSA) 3.8 percent annual growth translated into per capita income gains of 1.2 percent; at those levels per capita income doubles every 58 years.
- Per capita growth averaged 1.9 percent since 2010.
- The pandemic led to an economic contraction in the region, averaging 3 percent, with significant uncertainty surrounding the recovery.
- The paper’s three objectives:
  - Detail channels through which the pandemic is exacerbating structural rigidities and how reforms could support recovery and inclusion.
  - Investigate the main structural rigidities in SSA countries and quantify potential gains from reforms.
  - Discuss implementation considerations based on country experiences with similar reforms.
- Key contextual points:
  - The economic impact of the pandemic is expected to last longer than the health impact.
  - Demand management policies remain important but policy space is shrinking due to fiscal and debt effects; monetary easing room is constrained.
  - Supply-side constraints can blunt the efficacy of demand management tools.
  - Reforms take time to implement and bear fruit; design and sequencing matter given deteriorated social conditions (higher unemployment, increased poverty and inequality).

### II. COVID-19, STRUCTURAL RIGIDITIES, AND REFORMS
- Covid-19 has exacerbated pre-existing structural weaknesses while accelerating some policy adaptations (digitalization, green finance).
- The pandemic’s economic impact duration and depth depend on:
  - Pre-crisis social and economic conditions.
  - Severity and impact of the unfolding crisis.
  - Countries’ technical and financial capacities to roll out recovery policies.
  - Access to vaccines.
- Possible recovery shapes depend on flexibility and adaptation of markets and agents: L, U, V, W outcomes described; prospects for V-shaped recovery are diminishing.
- Alleviating structural rigidities can improve economic efficiency and inclusion over time, though gains accrue gradually and distributional effects can vary across groups.
- Reforms can improve medium-term growth, resource allocation, diversification, and resilience to shocks.

### III. STYLIZED FACTS AND DATA
- Empirical scope:
  - Panel data for 133 low- and middle-income countries (LICs/MICs) for 1996–2017.
  - Structural indicators relate to products markets, labor markets, governance, innovation, and complementary factors (education, health, life expectancy, infrastructure).
  - Distance to frontier measured as the difference between the country’s score and the top performer among upper-MICs; frontier adjusted for less ambitious reform scenarios.
  - Controls: income level, region, resource intensity (oil, non-oil), fragility; regressions include year fixed effects and country fixed effects.

### IV. EMPIRICAL APPROACH — Methodology and Main Results
- Methodological approach extends Prati, Onorato, and Papageorgiou (2013) using a wider dataset and more reform areas.
- Baseline regression results (associations with GDP per capita growth):
  - Governance, product market, and complementary factor reforms yield the largest per capita growth gains: between 1.1–1.3 percentage points (p.p.) over a 5–year period (10 years for governance).
  - Innovation: 0.9 p.p.
  - Labor markets: 0.3 p.p.
  - LICs stand to gain the most from reforms.
  - Growth could be higher by up to 1.8 p.p. for SSA oil exporters, driven by governance gains.
- Main finding on aggregate potential:
  - Reforms could boost per capita growth by 0.3–1.3 p.p., with up to 1.8 p.p. possible in oil exporters.
  - Delivering on these reforms will increase average per capita growth from 1.9 percent in the 2010s to between 2.2 and 3.2 percent.
  - Per capita income doubling time would fall from the current 37 years to roughly every 22 years under the reform scenario.
- Robustness:
  - Results are robust across income groups and specifications.
  - Findings on governance align with prior estimates (Hammadi et al., 2019) of a potential governance dividend of up to 2 percent for SSA.
- Key assumption:
  - Distance to the frontier largely determines growth gains; countries with weaker structural indicators stand to gain more.

### V. IMPLEMENTATION CONSIDERATIONS
- Structural reforms are wide-ranging and politically sensitive; design and sequencing are critical given worsened social conditions.
- Less politically sensitive reforms that aid job creation offer the highest growth gains and align with post-pandemic political economy constraints.
- Implementation choices include pace (big bang vs. gradual) and attention to distributional and social impacts to maintain political support.
- Reforms can be implemented in the post-pandemic environment, which is characterized by weaker social and distributional outcomes.

### VI. CONCLUSIONS
- Supply-side reforms can play a major role in aiding recovery and boosting growth by alleviating pandemic-exacerbated structural rigidities for firms, households, and the public sector.
- Reforms improve medium-term growth and resilience, facilitate economic diversification, and can support inclusion, though distributional effects differ across groups.
- Priority areas with the largest estimated dividends: governance, product markets, complementary factors (education, health, life expectancy, infrastructure), and innovation.
- The projected per capita growth gains imply substantial reductions in income doubling time and larger long-term welfare gains for SSA, especially for low-income countries and oil exporters.

*IMF Working Paper: JEL Classification Numbers: O4, O11, O50, O55, P16; Keywords: COVID-19, Growth, Development, Structural Reforms, Political Economy, Sub-Saharan Africa; Authors’ E-Mail Address: gganum@imf.org, jthakoor@imf.org*

### Section VI concludes.

### wpiea2021045-print-pdf - Section VI concludes

### II. COVID-19, STRUCTURAL RIGIDITIES, AND REFORMS
- Definition of structural reforms:
  - Policies to improve flexibility and efficiency in combining factor inputs and the institutional/regulatory setup (governance).
  - Can alleviate rigidities from government/public entities (regulations, public sector footprint, natural monopolies), private sector (rent-seeking, uncompetitive behaviors, unionization), or institutions (lack of independence, capacity, state capture).
- Why structural rigidities matter:
  - Rigidities inhibit efficient adjustment and contribute to resource misallocation; can be dynamic and mutually reinforcing.
  - Product market rigidities: worsen business environment, inflate production costs, undermine firm entry and innovation.
  - Labor market rigidities: maintain unemployment and wages above desirable levels, harming competitiveness and job creation.
  - Credit constraints: deter firm entry and competition.
  - Policy weaknesses undermining property rights or factor accumulation contribute to exclusion, unemployment, and slower digitalization and innovation.
- Table 1: Selected structural rigidities and their macroeconomic effects (high-level highlights preserved from table):
  - Product markets: common constraints include Limited competition, Inefficient SOEs, Over-regulation, Price controls, Weak business environment.
    - Conventional effects: High markups and poor service delivery undermine firm entry, innovation, and competitiveness; Distortive subsidies.
    - Possible COVID-effects: Increased concentration due to SMEs being disproportionately hit; Weaker financial situation of SOEs; Zombie firms.
  - Labor markets: common constraints include Dualization, High unionization, Bargaining mechanisms, Skills mismatches, Regulations on workforce management.
    - Conventional effects: Distortive taxes foster inequality; Rigid wages and inability of firms to adjust inhibit efficiency and job creation in formal sector.
    - Possible COVID-effects: Increased informalization; Inability of firms to review business models due to unionization and regulations; Permanent job losses.
  - Financial markets: common constraints include Access, cost, efficiency and Weak regulations.
    - Conventional effects: Credit constraints for households and firms, particularly the poor and SMEs; Weak banking sector with risks (NPLs, failures).
    - Possible COVID-effects: Increased indebtedness, particularly the poor and SMEs, with reliance on informal debt markets; Corporate debt overhang.
  - Public sector: common constraints include Elevated public sector wages, Budget rigidity and composition, Large footprint and crowding out of private sector.
    - Conventional effects: Wage premium and labor incentives to join public sector; Weak efficiency of government spending relative to needs; Spillovers from macroeconomic policies; Crowding out of private sector through reduced business space or higher cost of borrowing.
    - Possible COVID-effects: Increased pressures on government due to weakened social conditions and failure to roll-back COVID policies; Governance slippages arising from emergency procurement; Spillovers from increased financing needs and indebtedness, and recourse to unconventional policies.
  - Complementary policies (Health, Housing, Education, Infrastructure):
    - Conventional effects: Inhibit human capital accumulation, create entrepreneurship barriers, and limit opportunities.
    - Possible COVID-effects: Exclusion from job markets; Deskilling and reduced investment in education and health due to lost income.
- How COVID-19 could exacerbate structural rigidities (three main channels):
  - Households:
    - Job and income losses increase food insecurity and poverty.
    - Absent effective social safety nets, human capital accumulation declines through lower health and education spending, affecting longer-term income prospects.
    - Financial market segmentation can force households to informal lenders and higher indebtedness; reduced incomes can increase non-performing loans.
    - Vulnerable households subject to recurring shocks risk becoming entrenched in a vicious circle of poverty.
  - Firms:
    - Pandemic fragilized corporate balance sheets; many firms closed, others rely on government support or regulatory forbearance.
    - Access to cheap loans and government bailouts, including for SOEs, can reduce incentives to restructure, raising prospect of “zombie firms” that rely on credit to survive.
    - Implications for efficiency, innovation, and broader corporate risks as impaired firms transact.
  - Public sector:
    - Governments face challenges of rolling back COVID-19 support while addressing service delivery and governance weaknesses exposed by the pandemic.
    - Fiscal challenge: phasing out one-off policies to avoid permanent fiscal drag; rolling back support to SOEs with outdated business models.
    - Weakened corporate and social environments increase pressures for longer-than-desirable government support.
    - Absent reforms, financing pressures will remain, especially where government financing needs are elevated and access to finance constrained; continued recourse to central bank financing is risky; countries in debt distress will be more stretched.
- How structural reforms could help:
  - Enable firms to modernize production through restructuring and encourage dynamism by facilitating firm entry and exit.
  - Churning reallocates factors toward more productive entities/sectors and reduces fiscal and financial risks by lowering need to support unviable entities.
  - Large uncompetitive firms/sectors can exert institutional pressure for support, entrenching path dependency and perpetuating poor resource allocation.
- Link between structural reforms and inclusion:
  - Reforms can create trade-offs between efficiency and equity; distributional effects create winners and losers, who may resist reforms.
  - Examples: Rising trade union bargaining power can raise wages but reduce employment and output; product market rigidities often disproportionately hurt the poor via fewer entrepreneurial opportunities and higher prices.
  - Institutional reforms can broaden dissemination of gains, though benefits may materialize at different times for various groups.
  - Complementary policies—education, health, access to job centers (transportation, housing)—are favorable to those excluded from the formal system; these can reduce poverty, but the effect on inequality is not clear-cut.

### III. STYLIZED FACTS
- Pre-COVID growth in Sub-Saharan Africa (SSA):
  - Growth has averaged 3.8 percent since the 1970s—GDP doubling roughly every 19 years.
  - Per capita growth has averaged 1.2 percent—doubling roughly every 58 years.
  - Per capita growth in low-income countries has averaged 0.5 percent since the 1970s—doubling every 140 years—though around 1.5 percent in the 2000s.
  - Such growth levels are inadequate for the region to realize the demographic dividend.
- Relative levels and volatility:
  - Real GDP per capita in the median SSA country stood at around 5 percent of the US levels in 2019; high of about 10 percent for middle-income countries and under 4 percent for LICs and fragile states.
  - All income groups have seen a decline in the share relative to the US since the early 1970s.
  - SSA experiences higher growth volatility relative to other regions; volatility is higher for LICs and resource intensive economies—about twice the levels of advanced economies.
  - SSA non-resource intensive economies have volatility similar to other EMDEs and middle-income countries, although somewhat higher than emerging Asia.
- Post-COVID context:
  - Scope to boost growth post-COVID is constrained from both demand and supply sides.
  - Many countries have seen debt deteriorate significantly due to the pandemic, with an increasing number moving to debt distress.
  - Rising debt and increased reliance on external financing limit fiscal space to boost demand and increase vulnerability to future shocks.
  - Debt acceleration over the past decade has largely eroded gains achieved after HIPC.
  - Scope for monetary policy is limited after easing responses and record low interest rates; some countries have limited control over monetary policy due to monetary unions or high dollarization.
  - Weaknesses in monetary transmission can be worsened by structural rigidities.
  - SSA’s limited integration in global markets and export composition constrain growth and increase vulnerability due to limited diversification.
    - Share of SSA countries in global exports has hovered around 1.5 percent, with the pre-2010 increase driven largely by mineral fuel.
    - Muted commodity prices projected in the near-term create headwinds.
- Main structural rigidities and performance indicators:
  - Global Competitiveness Index (GCI) average for SSA: 3.5, significantly lower than worldwide top performers.
  - Strong performers generally improved 2013–2018 scores; some weak performers fell further behind.
  - Governance remains a challenge; most countries deteriorated relative to 2013, a few improved from low levels.
  - Product and labor market indices show limited changes relative to 2013, with average score of 4 on both components.
    - Product market constraints often manifest as limited competition, SOE presence in network industries, regulations creating insider-outsider dynamics, tariff and non-tariff barriers.
    - Labor market indicators cover wage flexibility, hiring/firing, and how taxes affect incentives to work.

### IV. EMPIRICAL APPROACH
- Methodology overview:
  - The factors holding back growth and appropriate reform mix vary by country and require individual assessment.
  - The approach discusses main reforms that could generate highest growth gains for SSA countries by combining distance to frontier in each reform category with estimated general gains from reforms among LICs and MICs.
- Estimation and baseline results:
  - Estimating the impact of structural reforms on growth is challenging due to identification difficulties in disentangling shocks, growth drivers, and economic policies.
  - The assessment follows Prati, Onorato, and Papageorgiou (2013).
  - Data: an unbalanced panel of 133 MICs and LICs over the period 1996-2017.
  - Structural indicators data vary across indicators and countries.

*Source: wpiea2021045-print-pdf - Section VI concludes.*

### Appendix I for definitions and coverage). We estimate the following growth equation:

### wpiea2021045-print-pdf - Appendix I for definitions and coverage). We estimate the following growth equation:

### Growth equation and empirical strategy
- Estimated equation:
  - 풍풏 GDP_i,t − 풍풏 GDP_i,t−1 = a0 + a1 풍풏GDP_i,t−1 + a2 Reform_i,t−1 + μ_i + δ_t + ε_it
  - Where GDP is real GDP per capita in country i in period t, Reform is each structural reform indicator with a lag.
- Identification and controls:
  - Year fixed effects to capture common trends; country fixed effects to capture time-invariant characteristics.
  - Reform variables lagged to mitigate reverse causality concerns.
  - All indicators rescaled to 1-10 (1 = lowest score, 10 = highest).
  - Regressions include clustered standard errors at the country level.

### Reform indices (definitions and composition)
- Governance reform:
  - Average of WGI indices: (i) voice and accountability; (ii) political stability; (iii) government effectiveness; (iv) control of corruption; (v) rule of law. Equal weights.
- Product market reform:
  - Average of WGI regulatory quality and two GCI indicators: goods market efficiency pillar and agricultural policy. Equal weights.
- Innovation:
  - GCI index reflecting resources for cutting-edge products, firm capacity to innovate, and environment conducive to R&D.
- Labor market reform:
  - Average of two GCI areas: (i) flexibility of wage determination; (ii) cooperation of labor-employer relations. Equal weights.
- Complementary factors:
  - Captures human capital and infrastructure: (i) life expectancy from WDI; (ii) primary education enrollment rate; (iii) quality of infrastructure from GCI. Equal weights.

### Baseline regression results (selected coefficients and metrics)
- Key coefficients (robust standard errors in parentheses; significance indicated):
  - L.GDPpc (log):  
    - Column (1): -0.0778*** (0.0139)  
    - Column (2): -0.0725*** (0.0119)  
    - Column (3): -0.1696*** (0.0539)  
    - Column (4): -0.1616*** (0.0487)  
    - Column (5): -0.0190*** (0.0047)  
    - Column (6): -0.1947*** (0.0590)
  - L.Worldwide Governance Indicator index:  
    - Column (1): 0.0188*** (0.0049)  
    - Column (2): 0.0270** (0.0104)
  - L.Product Market index:  
    - Column (1): 0.0086*** (0.0030)  
    - Column (2): 0.0036 (0.0069)
  - L.Innovation index:  
    - Column (1): 0.0087* (0.0052)  
    - Column (2): 0.0047 (0.0044)
  - L.Labor Market index:  
    - Column (1): 0.0028 (0.0042)  
    - Column (2): 0.0004 (0.0046)
  - L.Complementary factors index:  
    - Column (1): 0.0069*** (0.0023)  
    - Column (2): 0.0027 (0.0051)
- Summary regression metrics:
  - Observations: 2,463; 2,475; 1,043; 1,043; 6,036; 1,021 (across columns 1–6 respectively).
  - Adjusted R2: 0.0856; 0.0813; 0.141; 0.138; 0.0475; 0.147.
  - No. countries: 131; 133; 104; 104; 133; 102.
  - Robust standard errors in parentheses. Significance levels: *** p<0.01, ** p<0.05, * p<0.1.

### Long-run gains implied by the model
- Method:
  - Long-run associated gains computed by multiplying reform coefficient by inverse of (negative of) coefficient on lagged per capita GDP.
- Examples of long-run impacts:
  - Improving governance reforms by one point (index 1–10) is associated with an estimated increase in per capita output of about 24 percent in the long run.
  - Complementary factors yield the largest long-run gains in the baseline specification, with per capita output increasing by about 40 percent.

### Dividends from reforms — scenarios and magnitudes
- Distance-to-frontier approach:
  - Frontier defined as the 95th percentile among upper-MICs for each indicator (latest year available 2017).
  - Assume gap closed over five years for all reforms except governance (assumed 10 years).
  - Dividends computed by multiplying baseline coefficient estimates with the annualized distance-to-frontier advance.
- Cross-country average annual per capita growth gains (SSA, baseline to 95th percentile):
  - Governance, product market, and complementary factor reforms: 1.1-1.3 percentage points per year.
  - Innovation: 0.9 percentage point per year.
  - Labor markets: 0.3 percentage point per year.
- Less ambitious scenario (reach 75th percentile):
  - Overall gains between 0.1-0.9 percentage points per year across reform areas.
- Heterogeneity:
  - Oil exporters: magnitudes larger; governance reforms can yield up to 1.8 percentage points per year.
  - LICs tend to exhibit larger gains than MICs under the distance-to-frontier approach.

### Robustness checks and alternative estimators
- Alternative specifications:
  - Region-time varying fixed effects: similar results to baseline.
  - Three-year lag intervals: results robust; coefficients from reforms display larger magnitudes; product market reforms maintain expected sign but lose significance.
- System GMM (robust two-step, treating reforms as endogenous; “collapsed” instruments used):
  - Results broadly similar to baseline, with some reforms showing larger coefficients.
  - Tests: passes absence of second order serial correlation; Hansen test indicates instruments valid.
  - Arellano-Bond difference GMM gives similar results but Sargan test rejects instrument validity.
- Subsample and heterogeneity results:
  - Interactions by income group: all reform areas remain important across income groups; complementary reforms more important for LMICs.
  - Fragile states: governance reforms particularly important.
  - Sub-indicator unpacking: government effectiveness and political stability most important within governance; regulatory quality most important within product market reforms; quality of infrastructure and life expectancy important within complementary factors.
  - MICs vs LICs: innovation more important for MICs; governance and product market reforms have larger positive coefficients for LICs.

### Implementation considerations and policy design
- Key political economy and implementation factors:
  - Political ownership: full government ownership reduces contestation and reversal risk.
  - Sequencing: address binding constraints; calibrate sequencing to implementation capacity.
  - Composition: start with reforms delivering immediate gains for population to build buy-in (e.g., network industries; reductions in banking and telecom prices).
  - Synergies: combine supply- and demand-side measures (e.g., time-bound payments and training to offset labor reform opposition).
  - Complementarities required for success: good macroeconomic policies and governance; reliable infrastructure; strong human capital.
- Lags to realize gains:
  - Recognition and decision lags, legislative lags, and effectiveness lags affect time to reach frontier.
  - Evidence: product market reform gains tend to peak after 5 years; labor market reforms and innovation can take longer.

### International experience (lessons)
- Case summaries (selected lessons):
  - Australia and New Zealand: comprehensive multi-decade reforms improved competition and productivity; benefits realized over extended periods.
  - Mauritius: export-led diversification focusing on binding constraints and FDI; recent headwinds from weakening fundamentals.
  - Mexico: wide-ranging reforms in 2013–14 initially promising, with delayed implementation and some reversals limiting gains.
  - Chile: "first-generation" reforms boosted income; current priorities include inclusion, diversification, product market regulation, skills, and innovation.

### Conclusions — quantitative implications for SSA
- Summary implications:
  - Structural reforms are critical post-Covid-19 to restore and boost growth and resilience.
  - SSA stands to gain most from product market, governance, and innovation reforms; labor market reforms yield smaller gains on average.
- Quantified impacts:
  - Depending on composition, structural reforms can boost SSA’s per capita growth by an additional 0.3 to 1.3 percentage points over a 5-10-year period.
  - Higher gains up to 1.8 percentage points for oil exporters.
  - Context: the region’s 4.4 percent annual average growth since the 2010s has translated into per capita growth of 1.9 percent per year.
  - A strong reform scenario could boost per capita growth to above 3 percent annually, reducing doubling time of per capita income from around 37 years to about 22 years.

*Source: IMF staff estimates from wpiea2021045-print-pdf (Appendix I and associated tables and text).*

### References

### References

### Major cited works on structural reforms, governance, and macroeconomic effects
- Adhikari, B., R. Duval, B. Hu, and P. Loungani. 2018. “Can Reform Waves Turn the Tide? Some Case Studies using the Synthetic Control Method,” Open Economies Review, 29 (4): 879-910.
- Alesina, A., and A. Drazen. 1991. “Why are Stabilizations Delayed?” American Economic Review, 81: 1170–1188.
- Arellano, M., and S. Bond. 1991. “Some Tests of Specification for Panel Data: Monte Carlo Evidence and an Application to Employment Equations,” The Review of Economic Studies, 58(2): 277-297.
- Blanchard, O., and F. Giavazzi. 2003. “Macroeconomic Effects of Regulation and Deregulation in Goods and Labor Markets.” Quarterly Journal of Economics, 118: 879–907.
- Causa, O, M. Hermansen, and N. Ruiz. 2016. “The Distributive Effects of Structural Reforms.” OECD Economics Department Working Paper 1342, OECD, Paris.
- Cerra, V., and S.C. Saxena. 2008. "Growth Dynamics: The Myth of Economic Recovery." American Economic Review, 98 (1): 439-57.
- Da Silva, A., A. Givone, and D. Sondermann. 2017. “When Do Countries Implement Structural Reforms?” ECB Working Paper No 2078, June.
- David, A., T. Komatsuzaki, and S. Pienknagura. 2020, “The Macroeconomic Effects of Structural Reforms in Latin America and the Caribbean,” IMF Working Paper No. 20/195, Washington DC.
- Drazen, A., and V. Grilli. (1993). “The Benefit of Crisis for Economic Reforms,” American Economic Review, 83: 598–607.
- Duval, R., and D. Furceri. 2018. “The Effects of Labor and Product Market Reforms: The Role of Macroeconomic Conditions and Policies.” IMF Economic Review 66 (1): 31–69.
- Duval, R., and P. Loungani. 2019. “Designing Labor Market Institutions in Emerging and Developing Economies: Evidence and Policy Options.” IMF Staff Discussion Note 19/04, Washington, DC.
- Egert, B. 2018. “The Quantification of Structural Reforms: Extending the Framework to Emerging Market Economies”, CESifo Working Paper No. 6921.
- Fabrizio, S., D. Furceri, R. Garcia-Verdu, B.G. Li, S. Lizarazo, M. Mendes Tavares, F. Narita, A. Peralta-Alva, 2017. "Macro-Structural Policies and Income Inequality in Low-Income Developing Countries," IMF Staff Discussion Notes 2017/001, IMF, Washington DC.
- Hausmann, R., D. Rodrik, and A. Velasco, 2005, "Growth Diagnostics," Kennedy School of Government, Harvard University, Cambridge, Massachusetts; and Hausmann, R., B. Klinger, and R. Wagner. 2008. “Doing Growth Diagnostics in Practice: ‘A Mindbook’”. CID Working Paper No. 177, September.
- IMF (multiple): 2015a; 2015b; 2016a; 2016b; 2019; 2019b; 2020 — covering demographic dividend, structural reforms, supply-side boosts, resilience to natural disasters, competition and competitiveness, and IMFC communiqué.
- Jordà, O., S. R. Singh, and A. M. Taylor, 2020, The Long Hangover of Pandemics, Finance and Development, Vol57(2).
- Ostry, J. D, A Prati, and A Spilimbergo. 2009. “Structural Reforms and Economic Performance in Advanced and Developing Countries,” IMF Occasional Paper No. 268. Washington D.C.
- Ostry, J. D., A. Berg, and S Kothari. 2018. “Growth-Equity Trade-Offs in Structural Reforms,” IMF Working Paper 18/05. Washington D.C.
- World Bank. 2020. Global Economic Prospects. The World Bank, Washington D.C.
- Zettelmeyer, J., 2006. "Growth and Reforms in Latin America; A Survey of Facts and Arguments," IMF Working Papers 06/210, International Monetary Fund.
- (Additional referenced works cover country reform experiences—New Zealand, Australia, Mexico, Chile—and methodological contributions on panel estimators, synthetic controls, and political economy of reform.)

### Thematic focus of the references
- Structural reforms: governance, product market, labor market, innovation, complementary factors.
- Measurement and quantification of reforms for emerging and developing countries.
- Macroeconomic impacts of reforms on GDP per capita growth, inequality, and resilience to shocks.
- Political economy and implementation timing (crises as opportunities, delays in stabilization).
- Methodologies: panel data methods (Arellano–Bond), synthetic control, growth diagnostics, system GMM.

### Appendix I: Data

### Sample period and scope
- Baseline regressions: annual frequency data spanning 22 years (1996–2017).
- WDI indicators: availability starts as early as 1960.
- Data sources: WEO June 2020, World Development Indicators (WDI), Worldwide Governance Indicators (WGI), Global Competitiveness Indicators (GCI).
- Five reform types compiled: governance, product market reforms, innovation, labor market reforms, complementary factor reforms.
- Product market and labor market indicators such as ETCR and EPL are not available for SSA countries.

### Country coverage
- Raw sample: 133 countries (including middle-income and low-income countries).
- Regression samples vary with data availability and specification.

### Table A1: Variables, sources, and sample periods (selected entries; each subindicator rescaled from 1 to 10)
- Growth: GDP per capita growth rate, unit — WEO June 2020 — 1951-2025
- GDP per capita: Measured at PPP, log — WEO June 2020 — 1951-2025
- Governance reform: Average of political stability, government effectiveness, control of corruption, voice and accountability, and rule of law — WGI — 1996-2017
- Product market reform: Average of regulatory quality, goods market efficiency, and agricultural policy (rescaled from 1 to 10) — WGI, GCI — 1996-2017
- Goods Market efficiency — GCI — 2007-2017
- Agricultural Policy — GCI — 2007-2017
- Innovation (Innovation pillar) — GCI — 2007-2017
- Labor market reform: Average of flexibility of wage determination and cooperation of labor-employer relation — GCI — 2007-2017
- Complementary factors: Average of life expectancy, primary education enrollment and quality of infrastructure — WDI, GCI — 1960-2017
- Life expectancy, rescaled to a score — WDI — 1960-2017
- Quality of infrastructure — GCI — 1999-2017
- Note: All indicators rescaled from 1 to 10 for consistency.

### Table A2: Country list by income group (partial representation)
- Low income examples: Afghanistan, Bangladesh, Benin, Burkina Faso, Burundi, Cambodia, Central African Republic, Chad, Comoros, Eritrea, Ethiopia, Gambia, The, Guinea, Guinea-Bissau, Haiti, Kenya, Kyrgyz Republic, Liberia, Madagascar, Malawi, Mali, Mauritania, Mozambique, Myanmar, Nepal, Niger, Rwanda, Sierra Leone, Tajikistan, Tanzania, Togo, Uganda, Zimbabwe, Samoa, Solomon Islands, Sudan, Syrian Arab Republic, São Tomé and Príncipe, Timor-Leste, Tonga, Ukraine, Uzbekistan, Vanuatu, Vietnam, Yemen, Rep., Zambia.
- Lower middle income and Upper middle income columns list many additional countries (see source for full lists).

### Appendix II: Estimations of Dividends from Reforms

### Model setup and notation
- Rename ln GDP to y and Reform to R.
- Reform scenario equation (1):
  - y_it^R − y_i,t−1^R = α_0 + α_1 y_i,t−1^R + α_2 R_i,t−1^R + μ_i + γ_t + ε_it
- Reordered (2): y_it^R = α_0 + (1+α_1) y_i,t−1^R + α_2 R_i,t−1^R + μ_i + γ_t + ε_it
- Baseline (no reforms) (3): y_it^B = α_0 + (1+α_1) y_i,t−1^B + α_2 R_i,t−1^B + μ_i + γ_t + ε_it
- Subtraction leads to (4): y_it^R − y_it^B = (1+α_1)(y_i,t−1^R − y_i,t−1^B) + α_2(R_i,t−1^R − R_i,t−1^B)
- Assumption that reforms occur in t-1 and do not affect income until the next period (5): y_i,t−1^R = y_i,t−1^B
- Define distance-to-frontier gap (6): R_i,t−1^R − R_i,t−1^B = d_i,t−1 / n, where n is the time to close the gap.
- Resulting short expression (7): y_it^R − y_it^B = α_2 (d_i,t−1 / n)

### Multi-period expressions and converting to levels
- Period t+1 (9): y_i,t+1^R − y_i,t+1^B = (1+α_1)(α_2 d_i,t−1 / n) + α_2 d_i,t / n
- Exponentiation to level ratios (11):
  - Y_i,t+1^R / Y_i,t+1^B = exp((1+α_1)(α_2 d_i,t−1 / n) + α_2 d_i,t / n)
- Path of income per capita under reform (12):
  - Y_i,t+1^R = Y_i,t+1^B * (exp((1+α_1)(α_2 d_i,t−1 / n) + α_2 d_i,t / n))
- Estimation performed for each of the 5 years of reforms (10 years for governance reforms).
- General n-period expression (13) for n = 5 or 10 captures persistent but declining effects of past reforms on per capita income.
- Exponentiated n-period level (14) provides Y_i,t+n^R as baseline path times an exponential of the weighted sum of α_2 * (d / n) terms with (1+α_1) decay factors.

### Annex Tables — Key estimation results and coefficients (selected)

### Table 4: Structural Reforms and per capita GDP growth, by income group (selected coefficients)
- L.GDPpc (log): coefficients in columns (1)–(5): -0.0961***, -0.0938***, -0.1660***, -0.1604***, -0.0214*** (standard errors reported in source).
- L. Reform: 0.0272**, 0.0125***, 0.0063, 0.0107, 0.0018.
- Constant terms: 0.7562***, 0.8037***, 1.5319***, 1.4654***, 0.1725***.
- Observations: 3,375; 3,388; 1,539; 1,539; 8,440 (respectively per column).
- No. countries: 179, 182, 150, 150, 182 (respectively).
- Adjusted R2: 0.123, 0.121, 0.159, 0.162, 0.0562.

### Table 5: Structural Reforms and per capita GDP growth in MICs and LICs, with oil and fragile state interactions (selected)
- L.GDPpc (log): -0.0802***, -0.0710***, -0.1718***, -0.1610***, -0.0219*** (cols (1)–(5)).
- L.Reform: 0.0104**, 0.0074**, 0.0136***, 0.0025, 0.0099***.
- Fragile * L. Reform (col 1): 0.0217**; Fragile * L. Reform (col 5): -0.0094***.
- Observations: 2,463; 2,475; 1,043; 1,043; 6,036.
- No. countries: 131; 133; 104; 104; 133.
- Adjusted R2: 0.0890, 0.0841, 0.143, 0.136, 0.0507.

### Table 7: Baseline specification in MICs (selected)
- L.GDPpc (log) coefficients across columns: -0.0709***, -0.0656***, -0.1805***, -0.1710***, -0.0178***, -0.2132***.
- L.Worldwide Governance Indicator index: 0.0143** (col 1), 0.0358** (col 2).
- L.Innovation index: 0.0111* (col 3).
- L.Complementary factors index: 0.0067** (col 5).
- Observations vary by column: 1,843; 1,855; 787; 787; 4,532; 765.
- No. countries: 98, 100, 78, 78, 100, 76.
- Adjusted R2 range up to 0.162.

### Table 8: Baseline specification in LICs (selected)
- L.GDPpc (log): -0.0963***, -0.0934***, -0.0820*, -0.0923*, -0.0271***, -0.1075* (cols (1)–(6)).
- L.Worldwide Governance Indicator index: 0.0285*** (col 1), 0.0183* (col 2).
- L.Product Market index: 0.0149*** (col 3), 0.0006 (col 4).
- L.Complementary factors index: 0.0055* (col 5), -0.0049 (col 6).
- Observations: 620, 620, 256, 256, 1,504, 256.
- No. countries: 33, 33, 26, 26, 33, 26.
- Adjusted R2 up to 0.0878.

### Table 9: Baseline specification with three-year lag intervals (selected)
- Dependent variable: lngdp(t)-lngdp(t-3)
- L3.GDPpc (log): -0.1970***, -0.1889***, -0.4259***, -0.4109***, -0.0649*** (cols (1)–(5)).
- L3.Worldwide Governance Indicator index: 0.0304*** (col 1).
- L3.Innovation index: 0.0176** (col 3).
- L3.Complementary factors index: 0.0164*** (col 5).
- Observations: 2,460; 2,472; 1,040; 1,040; 6,033.
- No. countries: 131; 133; 104; 104; 133.
- Adjusted R2 up to 0.296 (col 3).

### Table 10: System GMM estimator (selected)
- Dependent variable: Real per capita growth
- L.GDPpc (log): -0.0137***, -0.0074**, -0.0176*, -0.0067, -0.0108** (cols (1)–(5)).
- L.Worldwide Governance Indicator index: 0.0222*** (col 1).
- L.Product Market index: 0.0086* (col 2).
- L.Innovation index: 0.0290* (col 3).
- L.Complementary factors index: 0.0067* (col 5).
- Observations: 2,463; 2,475; 1,043; 1,043; 6,036.
- No. countries: 131; 133; 104; 104; 133.
- No. of instruments: 105; 106; 90; 90; 142.
- Serial correlation p-values for AR(1): 0.1410; 0.1400; 0.0320; 0.0304; 0.00237.
- Serial correlation p-values for AR(2): 0.7250; 0.6360; 0.4200; 0.3350; 0.0675.
- Hansen Test p-values: 0.2800; 0.1040; 0.3450; 0.4970; 0.567.

### Annex Figures
- Annex Figure 1: Per capita growth dividends and distance to the frontier (percentile references: 95th percentile, 75th percentile shown in figure).
- Annex Figure 2: Cross-country Distribution of Dividends from Structural Reforms (in terms of per capita GDP growth).

*References and annex content as listed in the source PDF.*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2021/english/wpiea2021045-print-pdf.pdf_