## _wp1502

## Source details

**Canonical URL:** [_wp1502](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp1502.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp1502.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp1502.pdf.json)

---

### A. Introduction — Context and purpose
- Focus: identify policy reforms to increase productivity growth (TFP) in Sub-Saharan African Small Middle-Income Countries (SMICs).
- SMICs included: Botswana, Cabo Verde, Lesotho, Mauritius, Namibia, Seychelles, and Swaziland.
- Key stylized observations:
  - "Prudent macroeconomic management and improved institutional settings in many of the SMICs in SSA delivered impressive economic performance in the last few decades."
  - GDP per capita of SMICs of SSA on average increased by more than 5 times over the last 30 years compared to only 1.7 times increase for the SSA average excluding South Africa and Nigeria.
  - Per capita real GDP growth for most of these countries has fallen short of the range of 3-4 percent average needed to escape the so called “Middle-Income-Trap”.
- Principal aim: explore policy options that could boost productivity growth in SSA MICs based on cross-country analysis and propose reforms to lift TFP.

### B. Literature review — Determinants of TFP
- Conceptual role: TFP as "unexplained residuals from a production function" and a key driver of long-term growth via resource allocation, innovation, and factor-productivity improvements.
- Main categories of determinants and variables used:
  1. Macroeconomic conditions — Variables: Inflation, government debt, public employment. Empirical note: negative relationship between inflation and productivity growth; effect of government size ambiguous.
  2. Openness and technology creation/transfer — Variables: Trade, FDI, R&D, infrastructure. Empirical note: Trade openness and FDI facilitate technology transfer and learning.
  3. Quality of labor inputs and efficient allocation — Variables: Years of schooling, skill mismatch. Empirical note: human capital and gaps between skill supply and demand affect TFP growth.
  4. Female labor force participation — Variable: Female labor force participation rate. Empirical note: higher female participation may increase TFP if complementary to technology; effect can be concave over time.
  5. Sectoral composition and structural change — Variables: Sector shares of output, economic diversification. Empirical note: structural shift toward more productive sectors generally raises aggregate productivity.
  6. Monetary and financial development — Variables: Credit, market capitalization. Empirical note: financial development enhances productivity through capital allocation and innovation.
  7. Institutional and regulatory factors — Variables: Labor and business regulation indices, doing business indicators, income inequality. Empirical note: stronger institutions and lower political instability raise productivity.

### C. Stylized facts from SMICs analysis
- Historical drivers:
  - Periods of strong growth used capital deepening (infrastructure investment) and FDI to bolster productivity.
  - Recent moderation in growth reflects reduced contribution of TFP and incomplete structural reforms.
- Specific findings:
  - Years with significant FDI inflows helped most SMICs accumulate productive capital and bolster TFP growth; correlation between FDI and TFP growth is positive in most SMICs.
  - Infrastructure expansion (country-wide coverage of electricity, water, telecommunications) constituted a bedrock of productivity growth.
  - Regulatory burdens and weak regulatory systems constrained private sector development; structural constraints include regulatory burden inhibiting private sector development, restrictive labor market regulations, and high financial access costs.
  - Contribution of TFP to growth:
    - generally positively correlated with FDI,
    - generally positively correlated with infrastructure indicators (e.g., phone lines),
    - negatively correlated with regulatory constraints,
    - positively correlated with terms of trade improvement.
- Sectoral and export structure:
  - Except Swaziland, production structures in most SMICs are concentrated in the tertiary sector.
  - Export diversification is low (Herfindahl index reported; lower = greater diversification).
  - Mauritius (2012) percent distribution of GDP by industry group: Agriculture, forestry and fishing, 3.5%; Manufacturing, 16.9%; Construction, 6.3%; Wholesale & retail trade, 12.0%; Accommodation and food service, 6.9%; Information and communication, 4.6%; Transportation and storage, 5.8%; Real estate, 5.6%; Finance and insurance, 10.2%; Education, 4.4%; Human health and social work, 3.8%; Others, 20.0%.

### D. Main analytical contributions and methodological notes
- Two main contributions:
  1. Introduces an index of skill-mismatch as an indicator explaining TFP growth (examines both level and quality of education and the gap between skill supply and demand).
  2. Examines the impact of macro-stability friendly forms of financial inclusion on productivity growth.
- Estimation challenges: endogeneity, cross-country heterogeneity among TFP determinants, and data availability limitations.
- Econometric approaches used to address challenges: Dynamic panel estimation (System GMM), cointegration for heterogeneous panels, panel probit analysis, Group Mean FMOLS. Methodological details documented in Appendix I.

### E. Findings and policy recommendations (summary)
- Summary findings:
  - Structural reforms are needed to foster TFP growth and accelerate convergence to higher income levels.
  - Slowdown in TFP contribution to growth in many SMICs reduces prospects of graduating to high-income status.
  - Government borrowing to close infrastructure gaps has limits: after a certain threshold, government debt’s marginal impact on productivity growth becomes negative.
- Policy recommendations (explicit list from the paper):
  - Reform the financial sector to improve access and reduce costs (including macro-stability friendly financial inclusion).
  - Reduce regulatory barriers on firms and lower regulatory burden inhibiting private sector development.
  - Improve the quality of public spending, most notably on secondary and tertiary education, to reduce skill mismatch.
  - Alleviate infrastructure bottlenecks, while recognizing limits to productive impact of debt-financed infrastructure beyond a certain debt threshold.
  - Deepen capital markets to enhance efficient capital allocation.
  - Invest in research and development and new technologies to raise innovation and long-term TFP.

### Box 1 — Structural Transformation in Selected SMICs in SSA: Mauritius and Namibia
- Overview:
  - Reallocation from low to high productivity activities underpins structural transformation; drivers of growth may cause both growth and structural change to move together.
  - Period 1995 to 2010: high growth for many SSA countries with varying speeds of structural transformation.
  - Some reduction in agriculture's share of GDP, but employment did not move from agriculture into industry or services (contrast with Asian economies).
  - Dabla-Norris et al (2013) find product and labor market reforms, openness to trade, and access to finance explain variation in sectoral shares.
- Mauritius case:
  - Success factors: diverse and competitive political system supportive of reforms; sequencing of reforms including investing in education/training; flexibility in acquiring skills; attracting FDI (DTT with India); coherence of micro and macro policies.
  - Challenges: public sector administration and efficiency; structural labor market problems including youth unemployment; reliance on European markets for trade and tourism; environmental degradation and weather-related risks.
  - Authorities’ plans: accelerate transformation into a cyber-state; leverage ocean-based sectors.
- Namibia case:
  - Achievements: political and macroeconomic stability; diversification focus on business-oriented infrastructure and support contributing to rising manufacturing activities.
  - Constraints: exports reliant on mining; limited success of EPZs and SME programs; current strategy developing commodity-based value chains.
  - Key challenges: lack of skills in labor market; difficulties obtaining working permits; high regulation and large public sector; socioeconomic dual economy; financial notes: high concentration of bank loans to households, in particular mortgages; concerns about quality of public spending and large public sector.
- Cross-country structural impediments (selected):
  - Repeated lack of economic diversification across SMICs.
  - Labor market issues: high reservation wages and skill mismatch (Botswana); lack of skill (Namibia); high structural unemployment and HIV prevalence (Lesotho); youth unemployment (Mauritius); hiring qualified foreign workers difficult (Seychelles).
  - Financial sector concerns: high concentration of bank exposures and limited financial deepening.
  - Public sector issues: ineffective tax systems, poor quality of public spending, inefficient public investment, presence of large parastatals.
  - Environmental risks: droughts, soil degradation, environmental degradation, weather-related disasters.

### Empirical analysis: methodology, data, and coverage
- Two less explored areas examined: quality of education (beyond level) and relationship between stability-friendly financial inclusion and TFP growth.
- Panel techniques used:
  - Dynamic panel estimation (System GMM) for short-run cross-country evidence and endogeneity control.
  - Cointegration for heterogeneous panels to identify long-run relationships allowing country heterogeneity and adjusting for endogeneity and cross-sectional dependence.
  - Binary response model (panel probit) to measure extensive margin effects; outcome binary for TFP contribution to growth (1 if TFP contributes positively, 0 otherwise).
- Data sources: Penn World Tables (PWT), IMF WEO, World Bank WDI, Economic Freedom of the World (EFW), Barro-Lee.
- Dataset covers 33 upper-middle-income countries for the period 1980–2010. TFP used is human capital augmented TFP from PWT for 30 countries; TFP for Cabo Verde, Seychelles, and Swaziland derived using a World Bank growth accounting tool.

### Empirical results — key quantitative estimates and magnitudes
- General consistency: trade openness and macroeconomic stability have a significant and positive impact on productivity growth.
- Skill-mismatch:
  - Statistically significant and negative impact on productivity (TFP) growth in both short and long-term.
  - Short-run System GMM: For every 10-point increase in the skill-mismatch index, TFP growth tends to decline by about 0.03 percentage points.
  - Cointegration (long-run): An increase in the index by 10 points reduces long-run TFP growth by about 0.2 percentage points.
- Government debt and TFP (concave relationship):
  - Long-run threshold of public debt to GDP ratio when marginal impact on TFP becomes zero is 32-55 percent.
  - Threshold increases to 55 percent when controlling for market capitalization.
  - Short-run threshold estimated in dynamic panel analysis is 80 percent.
  - The 90-percent confidence interval of the debt-to-GDP threshold is (40.3, 86.4).
- Other empirical notes:
  - Ambiguous or counterintuitive results: negative relation between TFP and female participation in some specifications; negative coefficients for credit to GDP ratio and tertiary education in some specifications.
  - Better credit market regulation seems positively correlated with TFP growth but not always statistically significant. Effects of goods market efficiency on TFP growth are ambiguous and not robust across specifications.
- Dynamic panel (System GMM) short-run effects:
  - Skill-mismatch and government debt have statistically significant short-term effects on TFP growth in the model that includes Botswana, Cabo Verde, Mauritius, and Namibia.
  - Based on these results, level of government debt at which positive returns decline to zero is 80 percent (short-run interpretation).

### Consistent with other studies — Determinants and magnitudes
- Macroeconomic stability and trade openness conducive to productivity growth; high inflation reduces TFP growth.
- A large agricultural sector reduces TFP growth; reducing the relative size of the agricultural sector would improve long-run TFP growth.
- FDI: An increase in FDI to GDP ratio by one percentage point increases TFP growth by about 0.1 to 0.6 percentage points in the long run. However, when the share of foreign trade in GDP is included with FDI, the impact of FDI becomes negative.
- Female labor force participation: negative relationship in initial specifications; becomes positive when controlling for agricultural sector share in GDP (suggesting women in sample countries are more involved in low-productivity agriculture).
- Credit to GDP ratio: negative impact in some specifications, potentially reflecting credit expansion followed by financial crises and negative long-run effects of financial crises on TFP.
- Tertiary education coefficient: turned out negative when included together with years of schooling, possibly reflecting quality and efficiency issues.
- Inflation: large coefficient estimates may reflect impacts of inflation on other macroeconomic variables or a non-linear inflation impact not captured by the model.
- FDI coefficient lacks robustness across specifications and changes sign, likely reflecting sample changes due to data availability.

### Panel Probit analysis — probability results and magnitudes
- Higher skill mismatch index decreases the chance of TFP increasing growth in the long run.
- Higher inflation decreases probability that TFP contributes to growth: a one percentage point increase in inflation reduces the probability of TFP contributing to growth by about 2 percent in the long run.
- Probability relation with government debt is concave; coefficients not always statistically significant.
- Less strict labor market regulation increases likelihood of TFP contributing to growth: lowering labor regulation index by one point associated with a 34 percent increase in probability of positive TFP contribution to growth.
- Credit market variables: less strict credit regulation and higher percentage of SMEs with a credit line increase likelihood of TFP increasing growth.
- Selected panel probit regression coefficients (reported excerpts):
  - INF: -0.018**, -0.022*, -0.027***, -0.0315, -0.030*** (standard errors: 0.0086, 0.012, 0.00750, 0.150, 0.011)
  - DBT: -0.0052, -0.0025, 0.063**, 0.082*, 0.015 (standard errors: 0.009, 0.0081, 0.029, 0.047, 0.033)
  - DBT^2/1000: 0.00075, -0.00023, -0.032, -0.048, 0.0038 (standard errors: 0.0046, 0.0042, 0.024, 0.034, 0.029)
  - SKL (skill mismatch): -0.051**, -0.031* (standard errors: 0.021, 0.017)
  - CRG (credit regulation): 0.57***, 0.17 (standard errors: 0.16, 0.24)
  - SMC (SME credit access): 0.033**, (standard error: 0.016)
- Observations and sample coverage for panel probit models: Observations = 112, 107, 49, 22, 53; Number of Countries = 33, 32, 20, 22, 20 (as reported in Table 5).
- Predictive probabilities for selected SSA SMICs (Table 6):
  - Botswana: 69.1, 57.5, 11.8, 13.8, 35.3
  - Cabo Verde: 63.3, 62.0
  - Lesotho: 63.9, 63.8
  - Mauritius: 64.5, 68.0, 71.3, 86.6, 65.9
  - Namibia: 69.3, 63.4, 11.3, 12.8, 34.6
  - Seychelles: 48.2, 29.0
  - Swaziland: 69.4, 92.0

### Interpretation and policy-relevant implications from empirical analysis
- Skill-mismatch: underscores importance of quality of education and relevance of skills to labor market needs; addressing skill-mismatch is crucial for boosting TFP.
- Government debt: positive returns when financing productive public goods at low levels, but high debt levels can crowd out private investment and harm productivity.
- Trade openness, macroeconomic stability, access to finance, and product and labor market reforms matter for productivity and sectoral reallocation.
- Country-specific structural impediments (e.g., high regulation, large public sectors, skill shortages, difficulties hiring foreign workers) require targeted reforms to enable structural transformation.

### 3. Binary or multinomial response model: Probit/Logit estimation (methods and notes)
- Dependent indicator: binary index whether TFP contribution is positive/negative.
- Model components:
  - X_it: vector of time-varying country-specific characteristics.
  - α_i: time-invariant country fixed effects.
  - u_it: time-varying unobserved error term.
- Pros: allows non-linear effects; index mitigates measurement errors in TFP contribution.
- Cons: dependent variable contains less information; threshold choice can be arbitrary; complications with endogenous regressors or lagged dependent variables.

### 4. Construction of the Skill Mismatch Index
- Purpose: capture whether changes in TFP relate to matching supply and demand for skills (following Estevão and Tsounta (2011)).
- Index formula: difference between skill demand and supply for each skill level j:
  - S_ijt: percentage of population with skill level j at time t in country i (supply).
  - M_ijt: percentage of employees with skill level j at time t in country i (demand).
- Skill level supply from World Bank educational attainment:
  - primary education = low skilled,
  - secondary education = semi-skilled,
  - college and tertiary education = high skilled.
- Skill level demand approximated by percentage of employees in three sectors:
  - mining and construction (proxy low-skilled),
  - manufacturing (semi-skilled),
  - government and financial services (high-skilled).
- Methodological caveat: treating mining as low skilled may be weak because most mining employees in SMICs are at least medium to high skilled.

_italic: Source — _wp1502 (extracted content)._

### References _______________________________________________________________________________________ 21

### _wp1502 - References

### A. Introduction — Context and purpose
- Focus: identify policy reforms to increase productivity growth (TFP) in Sub-Saharan African Small Middle-Income Countries (SMICs).
- SMICs included in the analysis: Botswana, Cabo Verde, Lesotho, Mauritius, Namibia, Seychelles, and Swaziland.
- Key stylized observations:
  - "Prudent macroeconomic management and improved institutional settings in many of the SMICs in SSA delivered impressive economic performance in the last few decades."
  - GDP per capita of SMICs of SSA on average increased by more than 5 times over the last 30 years compared to only 1.7 times increase for the SSA average excluding South Africa and Nigeria.
  - Per capita real GDP growth for most of these countries has fallen short of the range of 3-4 percent average needed to escape the so called “Middle-Income-Trap”.
- Principal aim: explore policy options that could boost productivity growth in SSA MICs based on cross-country analysis and propose reforms to lift TFP.

### B. Literature review — Determinants of TFP (summary of conceptual variables and empirical findings)
- Conceptual role of TFP: "unexplained residuals from a production function" and a key driver of long-term growth via resource allocation, innovation, and factor-productivity improvements.
- Main categories of determinants and the variables used in the analysis:
  1. Macroeconomic conditions
     - Variables: Inflation, government debt, public employment
     - Empirical note: cross-country studies find a negative relationship between inflation and productivity growth; effect of government size is ambiguous.
  2. Openness and technology creation/transfer
     - Variables: Trade, FDI, R&D, infrastructure
     - Empirical note: Trade openness and FDI facilitate technology transfer and learning.
  3. Quality of labor inputs and efficient allocation
     - Variables: Years of schooling, skill mismatch
     - Empirical note: human capital and gaps between skill supply and demand affect TFP growth.
  4. Female labor force participation
     - Variable: Female labor force participation rate
     - Empirical note: higher female participation may increase TFP if complementary to technology; effect can be concave over time.
  5. Sectoral composition and structural change
     - Variables: Sector shares of output, economic diversification
     - Empirical note: structural shift toward more productive sectors generally raises aggregate productivity.
  6. Monetary and financial development
     - Variables: Credit, market capitalization
     - Empirical note: financial development enhances productivity through capital allocation and innovation.
  7. Institutional and regulatory factors
     - Variables: Labor and business regulation indices, doing business indicators, income inequality
     - Empirical note: stronger institutions and lower political instability raise productivity by improving the investment climate and resource allocation.

### C. Stylized facts from SMICs analysis
- Historical drivers of TFP in SMICs:
  - During periods of strong growth, SMICs used capital deepening (infrastructure investment) and FDI to bolster productivity.
  - Recent moderation in growth reflects reduced contribution of TFP to growth and incomplete structural reforms.
- Specific empirical and descriptive findings:
  - Years with significant FDI inflows helped most SMICs accumulate productive capital and bolster TFP growth; correlation between FDI and TFP growth is positive in most SMICs.
  - Infrastructure expansion (country-wide coverage of electricity, water, telecommunications) constituted a bedrock of productivity growth.
  - Regulatory burdens and weak regulatory systems constrained private sector development and hampered productivity growth; structural constraints include:
    - regulatory burden inhibiting private sector development,
    - restrictive labor market regulations,
    - high financial access costs.
  - Contribution of TFP to growth is:
    - generally positively correlated with FDI,
    - generally positively correlated with infrastructure indicators (e.g., phone lines),
    - negatively correlated with regulatory constraints,
    - positively correlated with terms of trade improvement.
- Sectoral and export structure:
  - With the exception of Swaziland, production structures in most SMICs are concentrated in the tertiary sector.
  - Export diversification is low (Herfindahl index reported; lower = greater diversification).
  - Example percent distribution of GDP by industry group (2012) for Mauritius: Agriculture, forestry and fishing, 3.5%; Manufacturing, 16.9%; Construction, 6.3%; Wholesale & retail trade, 12.0%; Accommodation and food service, 6.9%; Information and communication, 4.6%; Transportation and storage, 5.8%; Real estate, 5.6%; Finance and insurance, 10.2%; Education, 4.4%; Human health and social work, 3.8%; Others, 20.0%.

### D. Main analytical contributions and methodological notes
- Two main contributions of the paper:
  1. Introduces an index of skill-mismatch as an indicator explaining TFP growth (examines both level and quality of education and the gap between skill supply and demand).
  2. Examines the impact of macro-stability friendly forms of financial inclusion on productivity growth.
- Estimation challenges highlighted:
  - Endogeneity, cross-country heterogeneity among TFP determinants, and data availability limitations.
  - Use of several econometric techniques to account for endogeneity and heterogeneity and to provide robustness (methodologies for modeling TFP are documented in Appendix I).

### E. Findings and policy recommendations
- Summary findings:
  - Structural reforms are needed to foster TFP growth and accelerate convergence to higher income levels.
  - There is a slowdown in the contribution of TFP to growth in many SMICs, reducing their potential to graduate to high-income status.
  - Government borrowing to close infrastructure gaps has limits: after a certain threshold, government debt’s marginal impact on productivity growth becomes negative.
- Policy recommendations to boost productivity growth (explicit list from the paper):
  - Reform the financial sector to improve access and reduce costs (including macro-stability friendly financial inclusion).
  - Reduce regulatory barriers on firms and lower regulatory burden inhibiting private sector development.
  - Improve the quality of public spending, most notably on secondary and tertiary education, to reduce skill mismatch.
  - Alleviate infrastructure bottlenecks, while recognizing limits to productive impact of debt-financed infrastructure beyond a certain debt threshold.
  - Deepen capital markets to enhance efficient capital allocation.
  - Invest in research and development and new technologies to raise innovation and long-term TFP.

*Source: _wp1502 - References (extracted content).*

### Box 1. Structural Transformation in Selected SMICs in SSA—Mauritius and Namibia

### Box 1. Structural Transformation in Selected SMICs in SSA—Mauritius and Namibia

### Overview of structural transformation in SSA SMICs
- The reallocation of economic activity from low to high productivity activities lies at the heart of the rapidly growing work on structural transformation. An alternative view is that drivers of growth cause both growth and structural change to move simultaneously.
- The period 1995 to 2010 was characterized by high growth for a significant number of countries in SSA, most of which have experienced some degree of structural transformation, albeit at different speeds.
- Although there has been some reduction in the share of agriculture in the GDP in SSA SMICs, employment has not moved from agriculture into industry or services. This contrasts with Asian economies that have registered strong growth.
- An empirical study by Dabla-Norris et al (2013) suggests that product and labor market reforms, openness to trade, and access to finance are factors that explain the variation in sectoral shares across countries.

### Country cases: Mauritius
- Success factors cited for Mauritius:
  - A diverse and competitive political system supportive to the economic reforms.
  - Better sequencing of reforms, particularly investing in appropriate education and training, which enhanced the absorptive capacity and buttressed the authorities’ resolve to create new sectors.
  - Flexibility in acquiring necessary skills in the labor market, attracting FDI (DTT with India), and the coherence of micro and macro policies.
- Challenges noted:
  - Public sector administration and efficiency of public service.
  - Structural labor market problems, including youth unemployment.
  - Reliance on European markets for trade and tourism.
  - Environmental degradation and weather-related risks.
- Authorities’ plans and opportunities:
  - Plan to accelerate transformation into a cyber-state.
  - Leverage opportunities from the vast potential of ocean-based sectors.

### Country cases: Namibia
- Achievements:
  - Political and macroeconomic stability has helped improve living standards.
  - Active in enhancing diversification with a focus on business-oriented infrastructure and support that contributed to rising manufacturing activities.
- Constraints and policy history:
  - Exports have relied on mining; past initiatives included export processing zones (EPZ) and small-and-medium enterprise development programs with limited success.
  - Current strategy focuses on developing commodity-based value chains to enhance growth and economic diversification.
- Key challenges:
  - Lack of skills in the labor market.
  - Difficulties in obtaining working permits for the private sector.
  - High regulation and a large public sector; socioeconomic challenges of a dual economy hinder opening the labor market to achieve high skills needed to boost economy-wide productivity and potential growth.
- Financial and public sector notes from structural impediments table:
  - High concentration of bank loans to households, in particular mortgages.
  - Quality of public spending (not getting value for money) and large public sector.

### Structural impediments to productivity enhancement (selected points for SMICs)
- Common cross-country impediments (from Table 1):
  - Lack of economic diversification is repeatedly cited for Botswana, Namibia, Lesotho, Swaziland, Cabo Verde, Mauritius, Seychelles.
  - Labor market issues include high reservation wages and skill mismatch (Botswana), lack of skill (Namibia), high and persistent structural unemployment and HIV prevalence among labor force (Lesotho), youth unemployment (Mauritius), shortage of adequate skills and complex procedures for hiring qualified foreign workers (Seychelles).
  - Financial sector concerns include high concentration of bank exposures and limited financial deepening.
  - Public sector issues include effectiveness of tax system, quality of public spending, large and ineffective public sectors, inefficient public investment, and presence of large parastatals.
  - Environmental risks such as droughts, soil degradation, environmental degradation and weather-related disasters affect several SMICs.

### Empirical analysis: methodology and data
- Two relatively less explored areas examined:
  - Quality of education (beyond level of education) because lack of relevant skills contributes to persistently high unemployment in many SSA SMICs.
  - Relationship between stability-friendly forms of financial inclusion and TFP growth.
- Panel data techniques used:
  - Dynamic panel estimation (System GMM) to provide short-run cross-country evidence and control for endogeneity.
  - Cointegration for heterogeneous panels to identify long-run relationships allowing country heterogeneity and adjusting for endogeneity and cross-sectional dependence.
  - Binary response model (panel probit analysis) to measure extensive margin effects; outcome is binary for TFP contribution to growth (1 if TFP contributes positively to growth, and 0 otherwise).
- Data sources:
  - Penn World Tables (PWT), IMF’s World Economic Outlook (WEO), World Bank’s World Development Indicators (WDI), Economic Freedom of the World (EFW) project, Barro-Lee database.
  - Dataset covers 33 upper-middle-income countries for the period 1980–2010. TFP used is human capital augmented TFP from the latest PWT for 30 countries; TFP for Cabo Verde, Seychelles, and Swaziland was derived using a World Bank growth accounting tool.

### Empirical results: key findings and exact quantitative estimates
- General consistency: traditional variables such as trade openness and macroeconomic stability have a significant and positive impact on productivity growth.
- Skill-mismatch:
  - Statistically significant and negative impact on productivity (TFP) growth in both short and long-term horizons.
  - Short-run System GMM result: For every 10-point increase in the skill-mismatch index, TFP growth tends to decline by about 0.03 percentage points.
  - Cointegration (long-run) result: An increase in the index by 10 points reduces the long-run TFP growth by about 0.2 percentage points.
- Government debt and TFP:
  - Relationship is concave: at low levels, debt has a positive impact on TFP growth; at high levels, its impact becomes negative.
  - Long-run threshold of public debt to GDP ratio when marginal impact on TFP becomes zero is 32-55 percent. (The threshold increases to 55 percent when controlling for market capitalization.)
  - Short-run threshold estimated in dynamic panel analysis is 80 percent.
  - Additional reported interval: The 90-percent confidence interval of the debt-to-GDP threshold is (40.3, 86.4).
- Other empirical notes:
  - Some ambiguous or counterintuitive results were found: negative relation between TFP and female participation in the labor force in some specifications; negative coefficients for credit to GDP ratio and tertiary education in some specifications.
  - Better credit market regulation seems positively correlated with TFP growth but is not always statistically significant. Effects of goods market efficiency on TFP growth are ambiguous and not robust across specifications.
- Dynamic panel (System GMM) specific statistically significant short-run effects:
  - Skill-mismatch and government debt have statistically significant short-term effects on TFP growth in the model that includes Botswana, Cabo Verde, Mauritius, and Namibia.
  - Based on these results, the level of government debt at which positive returns decline to zero is 80 percent (short-run interpretation).

### Interpretation and policy-relevant implications from the analysis
- Skill-mismatch findings underscore the importance of the quality of education and relevance of skills to labor market needs; addressing skill-mismatch is crucial for boosting TFP growth.
- Government debt has potential positive returns when financing productive public goods (infrastructure, public goods) at low levels, but high debt levels can crowd out private investment and harm productivity.
- Trade openness, macroeconomic stability, access to finance, and product and labor market reforms matter for productivity and sectoral reallocation.
- Country-specific structural impediments (e.g., high regulation, large public sectors, skill shortages, difficulties in hiring foreign workers) need targeted reforms to enable structural transformation.

*Source: IMF staff (Box 1, "_wp1502 - Box 1. Structural Transformation in Selected SMICs in SSA—Mauritius and Namibia").*

### 26. Consistent with other studies, we found that macroeconomic stability, a small

### Consistent with other studies, we found that macroeconomic stability, a small agricultural sector, and trade openness are conducive to TFP growth.

### Key empirical findings on determinants of TFP
- Macroeconomic stability and trade openness are conducive to productivity growth; high inflation reduces TFP growth.
- A large agricultural sector reduces TFP growth; reducing the relative size of the agricultural sector would improve long-run TFP growth.
- An increase in FDI to GDP ratio by one percentage point increases TFP growth by about 0.1 to 0.6 percentage points in the long run. However, when the share of foreign trade in GDP is included with FDI, the impact of FDI becomes negative.
- Female labor force participation shows a negative relationship with TFP growth in initial specifications; when controlling for the share of agricultural sector in GDP the impact of female participation turns positive (suggesting women in sample countries are more involved in the low-productivity agricultural sector).
- Credit to GDP ratio has a negative impact in some specifications, potentially reflecting credit expansion followed by financial crises in parts of the sample and highlighting negative long-run effects of financial crises on TFP growth.
- The coefficient on tertiary education turned out negative when included together with years of schooling, possibly reflecting the quality and efficiency of tertiary education in the sample countries.
- Large coefficient estimates on inflation may reflect impacts of inflation on other macroeconomic variables influencing TFP growth or a non-linear inflation impact not captured by the model.
- The FDI coefficient is not robust across specifications and changes sign in other specifications, probably reflecting sample changes due to data availability.

### Panel Probit analysis — probability results and magnitudes
- Higher skill mismatch index decreases the chance of TFP increasing growth in the long run.
- Higher inflation decreases the probability that TFP contributes to growth: a one percentage point increase in inflation reduces the probability of TFP contributing to growth by about 2 percent in the long run.
- The relation between the probability of TFP contributing to growth and government debt is concave, although coefficients are not always statistically significant.
- Less strict labor market regulation increases the likelihood of TFP contributing to growth: lowering the index of labor regulation by one point is associated with a 34 percent increase in the probability of positive TFP contribution to growth.
- Credit market variables: less strict credit regulation and a higher percentage of SMEs with a credit line increase the likelihood of TFP increasing growth.
- Selected regression coefficient magnitudes reported (panel probit excerpts):
  - INF: -0.018**, -0.022*, -0.027***, -0.0315, -0.030*** (standard errors: 0.0086, 0.012, 0.00750, 0.150, 0.011)
  - DBT: -0.0052, -0.0025, 0.063**, 0.082*, 0.015 (standard errors: 0.009, 0.0081, 0.029, 0.047, 0.033)
  - DBT^2/1000: 0.00075, -0.00023, -0.032, -0.048, 0.0038 (standard errors: 0.0046, 0.0042, 0.024, 0.034, 0.029)
  - SKL (skill mismatch): -0.051**, -0.031* (standard errors: 0.021, 0.017)
  - CRG (credit regulation): 0.57***, 0.17 (standard errors: 0.16, 0.24)
  - SMC (SME credit access): 0.033**, (standard error: 0.016)
- Observations and sample coverage for panel probit models: Observations = 112, 107, 49, 22, 53; Number of Countries = 33, 32, 20, 22, 20 (as reported in Table 5).

### Predictive probabilities for selected SSA SMICs (Table 6)
- Botswana: 69.1, 57.5, 11.8, 13.8, 35.3
- Cabo Verde: 63.3, 62.0
- Lesotho: 63.9, 63.8
- Mauritius: 64.5, 68.0, 71.3, 86.6, 65.9
- Namibia: 69.3, 63.4, 11.3, 12.8, 34.6
- Seychelles: 48.2, 29.0
- Swaziland: 69.4, 92.0

### Policy-relevant implications and recommendations
- Maintain macroeconomic stability (including low and stable inflation) to support TFP growth.
- Promote trade openness to enhance productivity growth.
- Reduce the relative size of low-productivity agricultural activities and facilitate structural transformation away from agriculture to improve long-run TFP growth.
- Address skill mismatches in the labor market by improving the quality of education and public spending on education to better align skills with job requirements.
- Reduce regulatory burden on firms (less strict labor and credit market regulation) to increase the likelihood that TFP contributes to growth; easier access to finance for SMEs may be beneficial.
- Monitor government debt levels: the impact of government debt on productivity growth is significant and nonlinear, turning negative when government debt reaches a certain threshold of GDP.
- Exercise caution in interpreting FDI effects due to lack of robustness across specifications and sample changes caused by data availability.

### Methodologies used (summary)
- Three empirical methodologies were applied:
  - Dynamic Panel Estimation: Dynamic-Panel Generalized Method of Moments (Difference and System GMM). Pros: allows lagged dependent variable, controls for endogeneity; Cons: requires assumptions (e.g., instruments validity), stationarity of included variables.
  - Panel Co-integration Analysis: Pedroni’s seven statistics used to test cointegration with group mean and pooled tests; individual cointegration OLS regressions estimated for each country with adjustments to compute standard normal statistics.
  - Group Mean Fully-modified Ordinary Least Squares (FMOLS): estimates country-specific long-run relations and computes group mean FMOLS. Pros: suitable for non-stationary data, allows heterogeneous dynamics and cointegrating slopes, provides long-run estimates; Cons: computationally complex and depends on assumption of exact unit roots for regressors.

*Italic: Source — _wp1502 - 26. Consistent with other studies, we found that macroeconomic stability, a small agricultural sector, and trade openness are conducive to TFP growth.*

### 3. Binary or multinomial response model: Probit/Logit estimation

### 3. Binary or multinomial response model: Probit/Logit estimation

### Model specification and interpretation
- The dependent indicator is an index indicating whether the TFP contribution is positive/negative (i.e. [positive/negative indicator]), or whether TFP falls in a specific range.
- Explanatory components:
  - X_it: a vector of time-varying country-specific characteristics (including structural policy factors and other factors influencing the TFP).
  - α_i: the time-invariant country fixed effects (both observed and unobserved).
  - u_it: the unobserved error term which are time-varying.
- Probit/logit frameworks model a discrete outcome derived from an underlying continuous TFP contribution by applying thresholds to construct an index variable.

### Pros of Probit/Logit estimation
- The probit/logit estimations allow for non-linear effects of the explanatory variables.
- An index variable can mitigate the problem of measurement errors in the TFP contribution.

### Cons and assumptions required
- The dependent variable contains less information (as a continuous dependent variable is converted to a discrete dependent variable).
- The choice of thresholds in the multinomial response model can be very arbitrary.
- The estimations are more complicated when there is an endogenous regressor and/or a lagged dependent variable.

### Sectoral shift analysis (complementary empirical approach)
- Purpose: uses ideas of structural transformation and shifts to dynamic sectors that often lead to a boost in TFPs.
- Pros:
  - The sector shift analysis identifies how each sector contributes to the total productivity growth.
  - This is a decomposition exercise, so it is not subject to econometric assumptions.
- Cons:
  - The analysis requires information on sectoral productivity and employment.
  - It may not fully answer policy questions of interest; policy implementation implications relate to labor force reallocation and labor market flexibility (not directly how public sector size affects the TFP contribution).
- Reference note in source: See World Bank (2008)’s Unleashing Prosperity, pp.176-177.

### 4. Construction of the Skill Mismatch Index
- Purpose: assess whether changes in TFP can be captured by country differences in matching supply and demand for skills, following Estevão and Tsounta (2011).
- Index construction: the skills mismatch index for each country i at time t is calculated by taking the difference between the skill demand and supply for each skill level j:
  - S_ijt: the percentage of the population with skill level j at time t in country i (skill level supply).
  - M_ijt: the percentage of employees with skill level j at time t in country i (skill level demand).
- Skill level supply:
  - Constructed from World Bank educational attainment data using:
    - primary education = low skilled,
    - secondary education = semi-skilled,
    - college and tertiary education = high skilled.
- Skill level demand:
  - Approximated by the percentage of employees in three key sectors:
    - mining and construction (to proxy low-skilled workers),
    - manufacturing (for semi-skilled workers),
    - government and financial services (for high-skilled workers).
- Source caveat (methodological limitation): Although the Estevão and Tsounta (2011) method of estimating skill supply is reasonably robust based on educational attainment, the measures of skill demand and skill intensity have weaknesses, including treating mining sector as low skilled in the skill-intensity spectrum when most of the mining sector employees in SMICs are at least medium to high skilled.

*Source: _wp1502 - 3. Binary or multinomial response model: Probit/Logit estimation*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp1502.pdf_
