## Promise (Un)kept? Fintech and Financial Inclusion — Section 1

## Source details

**Canonical URL:** [Promise (Un)kept? Fintech and Financial Inclusion — Section 1](https://www.imf.org/-/media/files/publications/wp/2024/english/wpiea2024131-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2024/english/wpiea2024131-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2024/english/wpiea2024131-print-pdf.pdf.json)

---

### Abstract and main findings
- Fintech activity has expanded rapidly (start-up investments into fintech increased from US$1 billion in 2008 to over US$200 billion before the COVID-19 pandemic), while global account ownership rose from 51 percent of the world’s adult population in 2011 to 76 percent in 2021.
- Empirical analysis (panel of 84 countries, 2012–2020; fintech measures exclude cryptocurrencies) finds:
  - The magnitude and statistical significance of fintech on financial inclusion varies by instrument type.
  - Digital lending has a statistically significant negative effect on financial inclusion (coefficient: -0.068***; standard error: [0.015]).
  - Digital capital raising is statistically insignificant (coefficient: -0.823; standard error: [0.349]).
  - The overall impact of fintech (total fintech) is negative and statistically insignificant for the full sample (coefficient: -0.062; standard error: [0.014]).
  - The overall effect of fintech is statistically insignificant (and negative) in advanced economies but positive and statistically highly significant in developing countries (details presented later in the paper).

### Interpretation and context
- Fintech remains small relative to traditional financial intermediation:
  - fintech ~0.1 percent of GDP vs. traditional financial institutions ~59 percent of GDP.
- Possible reasons why fintech has not uniformly expanded financial inclusion:
  - Deep cultural, institutional, and socioeconomic undercurrents.
  - Gaps in access to digital technology infrastructure.
  - Biases in data and algorithms excluding elderly, low-income, women, and minorities.
  - Voluntary financial exclusion due to personal circumstances and preferences.

### Policy implications emphasized
- Policymakers should:
  - Develop an adequate regulatory framework to balance fostering innovation and ensuring equitable treatment.
  - Strengthen financial education.
  - Build strong regulatory institutions with enhanced technological capabilities.
  - Pursue extensive cross-border coordination.
  - Calibrate prudential regulations to ensure a level playing field and effective monitoring and supervision of traditional and emerging financial institutions.

### Data overview
- Sample: unbalanced panel covering 84 countries, 2012–2020.
- Dependent variable: percentage of adults (aged 15 and above) with an account at a formal financial institution (World Bank Global Findex Database).
- Key explanatory variable: volume of fintech transactions (excluding cryptocurrencies) as a share of GDP from the Cambridge Centre for Alternative Finance (CCAF).
  - CCAF divides fintech into: (i) digital lending, and (ii) digital capital raising.
  - Digital lending includes balance sheet lending, peer-to-peer and marketplace lending, debt-based lending, and invoice trading.
  - Digital capital raising includes investment-based crowdfunding and non-investment-based crowdfunding (donation-based, reward-based).
  - CCAF dataset excludes mobile money and internet banking (operated by traditional financial institutions).
- Control variables: real GDP per capita, inflation, trade openness (exports+imports as share of GDP), domestic credit to the private sector (as share of GDP), educational attainments (share of labor force with basic education), government stability, bureaucratic quality (World Bank; International Country Risk Guide).

### Key summary statistics (sample and variable dispersion)
- Financial inclusion (share of adults with an account)
  - Observations: 423
  - Mean: 56.3
  - Std. dev.: 0.3
  - Minimum: 2.5
  - Maximum: 100.0
- Fintech measures
  - Digital lending
    - Observations: 498
    - Mean: 0.1
    - Std. dev.: 0.2
    - Minimum: 0.0
    - Maximum: 1.6
  - Digital capital raising
    - Observations: 840
    - Mean: 0.0
    - Std. dev.: 0.0
    - Minimum: 0.0
    - Maximum: 0.5
  - Total fintech (all instruments)
    - Observations: 862
    - Mean: 0.1
    - Std. dev.: 0.1
    - Minimum: 0.0
    - Maximum: 1.6
- Other sample statistics
  - Real GDP per capita: Observations: 1,197; Mean: 14,860; Std. dev.: 19,598; Min: 270; Max: 108,352
  - Inflation: Observations: 1,156; Mean: 3.7; Std. dev.: 5.4; Min: -4.3; Max: 62.2
  - Trade openness: Observations: 1,198; Mean: 91.3; Std. dev.: 59.7; Min: 18.4; Max: 442.6
  - Domestic credit to the private sector: Observations: 1,096; Mean: 59.3; Std. dev.: 45.3; Min: 4.8; Max: 252.6
  - Educational attainments: Observations: 768; Mean: 46.9; Std. dev.: 16.7; Min: 12.6; Max: 96.6
  - Government stability: Observations: 925; Mean: 7.1; Std. dev.: 1.1; Min: 4.0; Max: 11.0
  - Bureaucratic quality: Observations: 925; Mean: 2.4; Std. dev.: 1.1; Min: 0.0; Max: 4.0

### Empirical strategy
- Baseline panel specification:
  - y_it = β1 + β2 fintech_it + β3 X_it + η_i + μ_t + ε_it
    - y_it: financial inclusion (share of adults with an account).
    - fintech_it: digital lending (share of GDP), digital capital raising (share of GDP), or total fintech (share of GDP).
    - X_it: controls listed above.
    - η_i: country fixed effects; μ_t: time effects.
- Robust inference:
  - Driscoll-Kraay (1998) standard errors used to account for heteroskedasticity, autocorrelation, and cross-sectional dependence in an unbalanced panel with short time dimension.
- Endogeneity:
  - Potential reverse causality addressed via 2SLS-IV approach, instrumenting contemporaneous fintech with its own lags.

### Selected regression results (all countries, baseline with controls)
- Column [1]: Digital lending
  - Digital lending coefficient: -0.068***
  - Standard error: [0.015]
- Column [2]: Digital capital raising
  - Digital capital raising coefficient: -0.823
  - Standard error: [0.349]
- Column [3]: Total fintech
  - Total fintech coefficient: -0.062
  - Standard error: [0.014]
- Other coefficients (columns [1] [2] [3])
  - Real GDP per capita: 0.640*** [0.091]; 0.622*** [0.077]; 0.601*** [0.078]
  - Inflation: -0.005 [0.002]; -0.003 [0.002]; -0.003 [0.001]
  - Trade openness: 0.000 [0.000]; 0.000 [0.000]; 0.000 [0.000]
  - Financial development (domestic credit): 0.001 [0.000]; 0.001 [0.000]; 0.001 [0.000]
  - Educational attainments: 0.002*** [0.000]; 0.002*** [0.000]; 0.002*** [0.000]
  - Government stability: 0.002 [0.001]; 0.001 [0.003]; 0.002 [0.002]
  - Bureaucratic quality: 0.125*** [0.013]; 0.111*** [0.011]; 0.120*** [0.012]
- Sample and fit
  - Number of observations: 127; 167; 176 (columns [1] [2] [3])
  - Number of countries: 69; 83; 84 (columns [1] [2] [3])
  - Country fixed effects: Yes
  - Time fixed effects: Yes
  - R-squared: 0.53; 0.49; 0.50 (columns [1] [2] [3])

### Robustness and subgroup findings
- Results robust to alternative estimation methods including 2SLS-IV with lags as instruments.
- Subsample estimates:
  - Advanced economies: fintech impact on financial inclusion is statistically insignificant (and negative).
  - Developing countries: fintech impact on financial inclusion becomes positive with a statistically highly significant overall effect.
- Additional checks:
  - Estimating the model for the pre-pandemic period yields similar results.
  - Estimating fintech impact on the gender gap in financial inclusion indicates financial inclusion helps close the gender gap, but this effect is not statistically significant across all countries.
  - Alternative measures of financial inclusion (share of adults with savings at a formal institution; share of adults with debit or credit cards) yield comparable results.

### Policy recommendations (from the analysis)
- Policymakers should:
  - Develop regulatory frameworks that balance innovation and equitable treatment of individuals and groups.
  - Enhance financial education to improve uptake and usage of fintech products.
  - Strengthen regulatory institutions and supervisory capacity, including technological capabilities.
  - Coordinate extensively cross-border to manage fintech spillovers and risks.
  - Calibrate prudential regulations to ensure a level playing field and effective monitoring and supervision of traditional and emerging financial institutions.

*Prepared by Serhan Cevik, IMF Working Paper WP/24/131, June 2024.*

### Section 1

### Promise (Un)kept? Fintech and Financial Inclusion — Section 1

### Abstract and main findings
- Fintech activity has expanded rapidly (start-up investments into fintech increased from US$1 billion in 2008 to over US$200 billion before the COVID-19 pandemic), while global account ownership rose from 51 percent of the world’s adult population in 2011 to 76 percent in 2021.
- Using a panel of 84 countries over 2012–2020 and direct measures of fintech (excluding cryptocurrencies), the empirical analysis finds:
  - The magnitude and statistical significance of fintech on financial inclusion varies by instrument type.
  - Digital lending has a statistically significant negative effect on financial inclusion (coefficient: -0.068***; standard error: [0.015]).
  - Digital capital raising is statistically insignificant (coefficient: -0.823; standard error: [0.349]).
  - The overall impact of fintech (total fintech) is negative and statistically insignificant for the full sample (coefficient: -0.062; standard error: [0.014]).
  - The overall effect of fintech is statistically insignificant (and negative) in advanced economies but positive and statistically highly significant in developing countries (details presented later in the paper).

### Interpretation and context
- Fintech remains small relative to traditional financial intermediation (fintech ~0.1 percent of GDP vs. traditional financial institutions ~59 percent of GDP) but is rapidly advancing.
- Possible reasons why fintech has not uniformly expanded financial inclusion:
  - Deep cultural, institutional, and socioeconomic undercurrents.
  - Gaps in access to digital technology infrastructure.
  - Biases in data and algorithms excluding elderly, low-income, women, and minorities.
  - Voluntary financial exclusion due to personal circumstances and preferences.
- Policy implications emphasized:
  - Develop an adequate regulatory framework to balance fostering innovation and ensuring equitable treatment.
  - Strengthen financial education.
  - Build strong regulatory institutions with enhanced technological capabilities.
  - Pursue extensive cross-border coordination.
  - Calibrate prudential regulations to ensure a level playing field and effective monitoring and supervision of traditional and emerging financial institutions.

### Data overview
- Sample: unbalanced panel covering 84 countries, 2012–2020.
- Dependent variable: percentage of adults (aged 15 and above) with an account at a formal financial institution (World Bank Global Findex Database).
- Key explanatory variable: volume of fintech transactions (excluding cryptocurrencies) as a share of GDP from the Cambridge Centre for Alternative Finance (CCAF).
  - CCAF divides fintech into: (i) digital lending, and (ii) digital capital raising.
  - Digital lending includes balance sheet lending, peer-to-peer and marketplace lending, debt-based lending, and invoice trading.
  - Digital capital raising includes investment-based crowdfunding and non-investment-based crowdfunding (donation-based, reward-based).
  - CCAF dataset excludes mobile money and internet banking (operated by traditional financial institutions).
- Control variables: real GDP per capita, inflation, trade openness (exports+imports as share of GDP), domestic credit to the private sector (as share of GDP), educational attainments (share of labor force with basic education), government stability, bureaucratic quality (World Bank; International Country Risk Guide).

### Key summary statistics (sample and variable dispersion)
- Financial inclusion (share of adults with an account)
  - Observations: 423
  - Mean: 56.3
  - Std. dev.: 0.3
  - Minimum: 2.5
  - Maximum: 100.0
- Fintech measures
  - Digital lending
    - Observations: 498
    - Mean: 0.1
    - Std. dev.: 0.2
    - Minimum: 0.0
    - Maximum: 1.6
  - Digital capital raising
    - Observations: 840
    - Mean: 0.0
    - Std. dev.: 0.0
    - Minimum: 0.0
    - Maximum: 0.5
  - Total fintech (all instruments)
    - Observations: 862
    - Mean: 0.1
    - Std. dev.: 0.1
    - Minimum: 0.0
    - Maximum: 1.6
- Other sample statistics
  - Real GDP per capita: Observations: 1,197; Mean: 14,860; Std. dev.: 19,598; Min: 270; Max: 108,352
  - Inflation: Observations: 1,156; Mean: 3.7; Std. dev.: 5.4; Min: -4.3; Max: 62.2
  - Trade openness: Observations: 1,198; Mean: 91.3; Std. dev.: 59.7; Min: 18.4; Max: 442.6
  - Domestic credit to the private sector: Observations: 1,096; Mean: 59.3; Std. dev.: 45.3; Min: 4.8; Max: 252.6
  - Educational attainments: Observations: 768; Mean: 46.9; Std. dev.: 16.7; Min: 12.6; Max: 96.6
  - Government stability: Observations: 925; Mean: 7.1; Std. dev.: 1.1; Min: 4.0; Max: 11.0
  - Bureaucratic quality: Observations: 925; Mean: 2.4; Std. dev.: 1.1; Min: 0.0; Max: 4.0

### Empirical strategy
- Baseline panel specification:
  - y_it = β1 + β2 fintech_it + β3 X_it + η_i + μ_t + ε_it
    - y_it: financial inclusion (share of adults with an account).
    - fintech_it: digital lending (share of GDP), digital capital raising (share of GDP), or total fintech (share of GDP).
    - X_it: controls listed above.
    - η_i: country fixed effects; μ_t: time effects.
- Robust inference:
  - Driscoll-Kraay (1998) standard errors used to account for heteroskedasticity, autocorrelation, and cross-sectional dependence in an unbalanced panel with short time dimension.
- Endogeneity:
  - Potential reverse causality addressed via 2SLS-IV approach, instrumenting contemporaneous fintech with its own lags.

### Selected regression results (all countries, baseline with controls)
- Column [1]: Digital lending
  - Digital lending coefficient: -0.068***
  - Standard error: [0.015]
- Column [2]: Digital capital raising
  - Digital capital raising coefficient: -0.823
  - Standard error: [0.349]
- Column [3]: Total fintech
  - Total fintech coefficient: -0.062
  - Standard error: [0.014]
- Other coefficients (columns [1] [2] [3])
  - Real GDP per capita: 0.640*** [0.091]; 0.622*** [0.077]; 0.601*** [0.078]
  - Inflation: -0.005 [0.002]; -0.003 [0.002]; -0.003 [0.001]
  - Trade openness: 0.000 [0.000]; 0.000 [0.000]; 0.000 [0.000]
  - Financial development (domestic credit): 0.001 [0.000]; 0.001 [0.000]; 0.001 [0.000]
  - Educational attainments: 0.002*** [0.000]; 0.002*** [0.000]; 0.002*** [0.000]
  - Government stability: 0.002 [0.001]; 0.001 [0.003]; 0.002 [0.002]
  - Bureaucratic quality: 0.125*** [0.013]; 0.111*** [0.011]; 0.120*** [0.012]
- Sample and fit
  - Number of observations: 127; 167; 176 (columns [1] [2] [3])
  - Number of countries: 69; 83; 84 (columns [1] [2] [3])
  - Country fixed effects: Yes
  - Time fixed effects: Yes
  - R-squared: 0.53; 0.49; 0.50 (columns [1] [2] [3])

### Robustness and subgroup findings
- Results robust to alternative estimation methods including 2SLS-IV with lags as instruments.
- Subsample estimates:
  - Advanced economies: fintech impact on financial inclusion is statistically insignificant (and negative).
  - Developing countries: fintech impact on financial inclusion becomes positive with a statistically highly significant overall effect.
- Additional checks:
  - Estimating the model for the pre-pandemic period yields similar results.
  - Estimating fintech impact on the gender gap in financial inclusion indicates financial inclusion helps close the gender gap, but this effect is not statistically significant across all countries.
  - Alternative measures of financial inclusion (share of adults with savings at a formal institution; share of adults with debit or credit cards) yield comparable results.

### Policy recommendations (from the analysis)
- Policymakers should:
  - Develop regulatory frameworks that balance innovation and equitable treatment of individuals and groups.
  - Enhance financial education to improve uptake and usage of fintech products.
  - Strengthen regulatory institutions and supervisory capacity, including technological capabilities.
  - Coordinate extensively cross-border to manage fintech spillovers and risks.
  - Calibrate prudential regulations to ensure a level playing field and effective monitoring and supervision of traditional and emerging financial institutions.

*Prepared by Serhan Cevik, IMF Working Paper WP/24/131, June 2024.*

### Section 2

### wpiea2024131-print-pdf - Section 2

### Empirical approach and dependent variable
- Dependent variable: financial inclusion measured by the percentage of adults (aged 15 and above) with an account at a formal financial institution.
- Estimation: Driscoll-Kraay standard errors reported in brackets in baseline tables; 2SLS-IV estimations use the contemporaneous measure of fintech instrumented with its own lags. *, **, and *** denote significance at the 10%, 5%, and 1% levels, respectively.

### Key findings — Full sample, advanced economies, and developing countries
- Overall (full sample):
  - Digital lending: statistically significant negative effect on the share of adults with an account at a formal financial institution (negative coefficient reported in text for full-sample specification; digital capital raising statistically insignificant in full-sample).
  - Total fintech (including all instruments) for the entire sample: negative and statistically insignificant in baseline; in 2SLS-IV estimations the estimated coefficient on all fintech instruments as a share of GDP is -0.211** (Driscoll-Kraay/robust SEs in brackets per tables).
- Advanced economies:
  - Estimated coefficients on fintech variables show a pattern of negative effects on financial inclusion, but these are statistically insignificant at conventional levels in baseline and remain statistically insignificant in IV estimations.
  - Table 3 reported specific estimates (selected):
    - Digital lending: -0.081 [0.040]
    - Digital capital raising: -0.269 [0.132]
    - Total fintech: -0.043 [0.034]
    - Real GDP per capita: 0.409*** [0.101], 0.274*** [0.024], 0.360*** [0.099] across three specs
    - Inflation: -0.002 [0.001], -0.007*** [0.002], -0.005*** [0.001]
    - Government stability: 0.003*** [0.000], 0.000 [0.001], 0.001 [0.001]
    - Bureaucratic quality: 0.072*** [0.018], 0.006 [0.003], 0.031 [0.016]
    - Number of observations: 727, 784
    - Number of countries: 28, 29
    - R2: 0.38, 0.28, 0.30
- Developing countries:
  - Impact of fintech on financial inclusion becomes positive and the overall effect of fintech (including all instruments) is statistically highly significant in baseline estimations.
  - Table 4 reported specific estimates (selected):
    - Digital lending: 0.036 [0.044]
    - Digital capital raising: 0.642 [1.300]
    - Total fintech: 0.552*** [0.116]
    - Real GDP per capita: 0.770*** [0.070], 0.587*** [0.047], 0.585*** [0.046]
    - Inflation: -0.006*** [0.001], -0.004*** [0.001], -0.005*** [0.001]
    - Trade openness: 0.005*** [0.001], 0.003*** [0.001], 0.003*** [0.001]
    - Financial development: 0.009*** [0.000], 0.004** [0.000], 0.003* [0.001]
    - Educational attainments: 0.010*** [0.001], 0.003*** [0.001], 0.002*** [0.000]
    - Government stability: 0.044*** [0.003], 0.010** [0.002], 0.012** [0.002]
    - Bureaucratic quality: 0.028*** [0.006], 0.100*** [0.014], 0.042*** [0.006]
    - Number of observations: 559, 092
    - Number of countries: 41, 54, 55
    - R2: 0.95, 0.63, 0.69
- 2SLS-IV (Table 5) highlights:
  - All countries: Total fintech = -0.211** [0.070]
  - Advanced economies (AE): Total fintech = -0.165 [0.132] (statistically insignificant)
  - Emerging markets (EM): Total fintech = 0.105 [0.211] (statistically insignificant)
  - Number of observations: 1,568,492
  - Number of countries: 84, 29, 55 across columns
  - R2: 0.98, 0.76, 0.98

### Interpretation of empirical results
- Heterogeneous effects:
  - Fintech endeavors in developing countries have helped expand financial inclusion (positive and statistically highly significant overall effect in baseline), but fintech remains small relative to traditional financial institutions.
  - In advanced economies, fintech’s estimated negative effects on financial inclusion are generally statistically insignificant.
- Control variables:
  - Real GDP per capita: positively and significantly correlated with financial inclusion across all countries and subsamples.
  - Inflation: negative association with financial inclusion, especially in developing countries.
  - Trade openness: positive and statistically significant only in developing countries.
  - Financial development (domestic credit to the private sector as a share of GDP): negative coefficient across specifications but not statistically significant in baseline (some positive and significant coefficients in developing-country specifications).
  - Educational attainments: positive contribution to financial inclusion.
  - Institutional and political variables: reduce financial exclusion, with greater magnitude and statistical significance in developing countries.

### Mechanisms, limitations, and persistent exclusion
- Potential channels limiting fintech’s inclusive impact:
  - Gaps in access to digital technology infrastructure.
  - Biases in data and algorithms used in fintech solutions.
  - Cultural, institutional, and socioeconomic undercurrents that persistently exclude groups (elderly, low-income, women and minorities).
  - Voluntary financial exclusion because of personal circumstances and preferences.
- Magnitude and evolution:
  - Fintech is growing rapidly but is still small relative to traditional financial institutions; therefore, its nature and magnitude of effects on financial inclusion are expected to evolve over time.

### Policy implications and recommendations
- Policymakers should:
  - Acknowledge potential risks and threats from fintech while fostering innovation.
  - Develop an adequate regulatory framework that balances fostering innovation and ensuring equitable treatment of individuals and groups.
  - Provide better financial education.
  - Strengthen regulatory institutions with enhanced technological capabilities.
  - Pursue extensive cross-border coordination.
  - Calibrate prudential regulations to ensure a level playing field and effective monitoring and supervision of traditional and emerging financial institutions.

*Source: Author's estimations as presented in the provided Section 2 content.*

### Section 3

### wpiea2024131-print-pdf - Section 3

### References cited in Section 3

- Vucinic, M. (2020). “Potential Influence of Fintech on Financial Stability: Risks and Benefits,” Journal of Central Banking Theory and Practice, Vol. 9, pp. 43–66.  
- Wang, R., J. Liu, and H. Luo (2021). “Fintech Development and Bank Risk Taking in China,” European Journal of Finance, Vol. 27, pp. 397–418.  
- Yang, T., and X. Zhang (2022). “FinTech Adoption and Financial Inclusion: Evidence from Household Consumption in China,” Journal of Banking and Finance, Vol. 145, 106668.  
- Zhang, X., J. Zhang, G. Wan, and Z. Lou (2020). “Fintech, Growth and Inequality: Evidence from China’s Household Survey Data” Singapore Economic Review, Vol. 65, pp. 75–93.  
- Zhu, X., S. Asimakopoulos, and J. Kim (2020). “Financial Development and Innovation-Led Growth: Is Too Much Finance Better?” Journal of International Money and Finance, Vol. 100, 102083.  

*wpiea2024131-print-pdf - Section 3*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2024/english/wpiea2024131-print-pdf.pdf_