Do FinTech algorithms reduce gender inequality in banks loans? A quantitative study from the USA

Song, Ziheng et al. Article Do FinTech algorithms reduce gender inequality in banks loans? A quantitative study from the USA Journal of Applied Economics Provided in Cooperation with: University of CEMA, Buenos Aires Suggested Citation: Song, Ziheng et al. (2024) : Do FinTech algorithms reduce gender inequality in banks loans? A quantitative study from the USA, Journal of Applied Economics, ISSN 1667-6726, Taylor & Francis, Abingdon, Vol. 27, Iss. 1, pp. 1-18, https://doi.org/10.1080/15140326.2024.2324247 This Version is available at: https://hdl.handle.net/10419/314262 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc/4.0/ Journal of Applied Economics ISSN: (Print) (Online) Journal homepage: www.tandfonline.com/journals/recs20 Do FinTech algorithms reduce gender inequality in banks loans? A quantitative study from the USA Ziheng Song, Shafiq Ur Rehman, Chun PingNg, Yuan Zhou, Patick Washington & Ricardo Verschueren To cite this article: Ziheng Song, Shafiq Ur Rehman, Chun PingNg, Yuan Zhou, Patick Washington & Ricardo Verschueren (2024) Do FinTech algorithms reduce gender inequality in banks loans? A quantitative study from the USA, Journal of Applied Economics, 27:1, 2324247, DOI: 10.1080/15140326.2024.2324247 To link to this article: https://doi.org/10.1080/15140326.2024.2324247 © 2024 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group. Published online: 18 Aug 2024. Submit your article to this journal Article views: 1142 View related articles View Crossmark data Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=recs20 Do FinTech algorithms reduce gender inequality in banks loans? A quantitative study from the USA Ziheng Song a , Shafiq Ur Rehman b,c , Chun PingNg d , Yuan Zhou e , Patick Washington f and Ricardo Verschueren g a Columbia Uplift Inc, Columbia University in the City of New York, Casper, WY, USA; b Lahore School of Business, The University of Lahore, Lahore, Pakistan; c School of Finance and Banking, University of Utara Malaysia, Sintok Kedha, Malaysia; d Tsinghua University, Haidian District, Beijing, China; e Business Administration, Columbia University in the City of New York, New York, USA; f Department of Business Administration, Morehouse College, Atlanta, GA, USA; g Business Administration, University of Gloucestershire, Cheltenham, UK ABSTRACT The potential of FinTech algorithms to decrease gender bias in credit decisions is limited by the impartiality of the data used to train them. If the data is partial or biased, the algorithmic decisionmaking process may also be discriminatory, exacerbating existing inequalities. In this study, the effect of FinTech Firms on reducing gender inequality in bank loans in the USA is examined using a loan application from 60 U.S. banks from 2012 to 2022. We use a twostep system GMM approach to estimate the effect of FinTech algorithms on gender bias in credit decisions, focusing on female loan approval rates. Our results show that by controlling the other factor, banks with credit algorithms significantly increased the loan approval rates and thus reduced gender inequality in bank loans. Specifically, the female loan approval rates increased by 8% after banks adopted FinTech algorithms. We also find that the effect is more substantial for banks with higher baseline gender bias in credit decisions. We also performed the Difference in Difference analysis to analyse the policy shocks and FinTech adoption on bank loans’ gender inequality. Results of the study show that FinTech adoption and policy implications have significantly increased the loans for female borrowers. Our findings suggest that FinTech algorithms can potentially mitigate gender bias in credit decisions and promote gender equality in financial services. ARTICLE HISTORY Received 4 July 2023 Accepted 15 February 2024 KEYWORDS FinTech; algorithms; gender inequality; credit decisions 1. Introduction Gender inequality is a significant issue in the financial services industry, where women often face discrimination in access to credit and financial products. Statistics suggest that gender inequality remains a major issue in the U.S. financial services industry despite some progress in recent years. For instance, according to a report by McKinsey & Company in 2020, women account for only 19% of executive positions in the financial CONTACT Shafiq Ur Rehman [email protected] Lahore Business School, The University of Lahore, University of Utara Malaysia, Sintok Kedha, Malaysia JOURNAL OF APPLIED ECONOMICS 2024, VOL. 27, NO. 1, 2324247 https://doi.org/10.1080/15140326.2024.2324247 © 2024 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group. This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial License (http:// creativecommons.org/licenses/by-nc/4.0/), which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited. The terms on which this article has been published allow the posting of the Accepted Manuscript in a repository by the author(s) or with their consent. services industry in the United States (Madsen, 2020). The same report found that the industry’s overall gender diversity has remained stagnant over the past five years, with only a 1% increase in women’s representation in senior executive positions (Madsen, 2020). Another study by the National Bureau of Economic Research in 2019 found that women in the U.S. financial services industry are paid, on average, 50% less than their male counterparts and are also more likely to experience barriers to career progression (Keller et al., 2020). Furthermore, a survey conducted by the Financial Industry Regulatory Authority (FINRA) in 2021 found that women are significantly less likely to hold a professional designation or license in the financial services industry, which can impact their earning potential and career advancement (FINRA Investor Education Foundation, 2021). It is well established in the literature that the financial services industry has historically been male-dominated (Fine et al., 2020; Issa et al., 2022; Mogaji & Nguyen, 2022; Song et al., 2020), which has led to gender bias in credit decisions and has prevented many women from accessing credit on an equal basis as their male counterparts (Blanco-Oliver et al., 2021; De Andrés et al., 2021; Galli et al., 2020; Shoma, 2019). This gender bias can manifest in many ways, including unequal treatment in credit assessment, less favourable loan terms, and a lack of access to credit products that meet women’s unique financial needs (Andriamahery & Qamruzzaman, 2022; Ghosh & Vinod, 2017). FinTech, with its use of algorithms and big data, has the potential to reduce gender bias in credit decisions and promote financial inclusion (Awotunde et al., 2021; Makina, 2019; Ozili, 2021; Purda & Ying, 2022). By relying on algorithms and machine learning, FinTech companies can avoid human biases in traditional credit decision-making processes (Johnson et al., 2019; Lui & Lamb, 2018; Packin & Lev-Aretz, 2018; Yu & Song, 2021). In particular, algorithms can be designed to use a broader range of data sources, including non-traditional ones such as social media and online shopping patterns, which can capture a more comprehensive picture of an individual’s creditworthiness (Hurley & Adebayo, 2016; Makina, 2019; Peng & Zhu, 2021; Purda & Ying, 2022; Sadok et al., 2022). FinTech’s ability to leverage data in this way can also help to address the gender pay gap by enabling lenders to look beyond traditional credit scoring metrics such as income and employment history, which may disadvantage women who have taken time off from work for caregiving or who work part-time. Instead, algorithms can be designed to incorporate a broader range of factors, such as education, job skills, and other indicators of future earning potential (Hurley & Adebayo, 2016; Makina, 2019; Peng & Zhu, 2021; Purda & Ying, 2022; Sadok et al., 2022). While FinTech is not a panacea for gender inequality in financial services, it can potentially play an important role in promoting more significant gender equity in credit decisions and expanding financial inclusion for women (Ahmed, 2021; Kamga et al., 2022). However, it is essential to remain vigilant about the potential for algorithmic bias and ensure that algorithms are designed to be as unbiased as possible (Ludwig & Mullainathan, 2021; Ozili, 2021; Purda & Ying, 2022; Yu & Song, 2021). For instance, previous literature finds that, while innovative, FinTech lending algorithms carry risks of inaccuracy, bias, and discrimination (Sargeant, 2023). Inaccuracies may arise from flawed data inputs or algorithmic errors, leading to unfair loan decisions (Moldovan, 2023). Biases can be embedded within algorithms if they are trained on historical data that reflects 2Z. SONG ET AL. past inequalities, potentially perpetuating discrimination against certain groups. Furthermore, individuals might exploit these systems’ vulnerabilities for personal gain, undermining the fairness and integrity of financial services{ (Moldovan, 2023) #83; Washington, 2022 #79; Rehman, 2023 #78}. To date, previous literature mainly focuses on FinTech and its uses in the financial inclusion of women (Arner et al., 2020; Gabor & Brooks, 2017; Salampasis & Mention, 2018), FinTech and online lending (i.e., peer-to-peer lending (Disemadi et al., 2020; Saiedi et al., 2022). However, limited empirical studies highlight the impact of FinTech-based credit algorithms and their impact on gender inequality. Therefore, this will be the first study highlighting the importance of FinTech credit algorithms to reduce gender inequality in credit decisions. This study aims to fill this research gap by examining the impact of FinTech algorithms on reducing gender inequality in credit decisions. Using 60 U.S. banks’ internal data from 2012 to 2022, we measured the impact of FinTech-based credit algorithms and their impact on both Male and female bank loans. We find the following results. We performed regression separately for both Male and Female, and we found that banks adopting FinTech have higher loan approval rates for male borrowers. For Female, the regression results indicate that the model explains 26.3% of the variance in loan approval rates for female borrowers (R-squared = 0.263). FinTech adoption is found to have a statistically significant positive relationship with loan approval rates for female borrowers (β1 = 0.0206, p < 0.001). We also estimate our empirical results using a 2SS-GMM to reduce the potential endogeneity issue further. The results show that FinTech adoption positively and significantly affects loan approval rates. Suggesting that banks adopting FinTech solutions are more likely to approve loans, which aligns with improved efficiency of loan approval processes and reduced costs, leading to higher approval rates. The DID analysis results show that policy and FinTech adoption significantly impact bank loans, particularly for the female loan approval rate. Overall, the study’s results suggest that FinTech algorithms significantly reduce gender bias in bank credit decisions. This study contributes to understanding the potential of FinTech algorithms to reduce gender bias in credit decisions in the USA. Specifically, the study provides evidence that banks’ adoption of FinTech algorithms can significantly increase female loan approval rates and reduce gender inequality in bank loans. The study uses a rigorous two-step system GMM approach and ‘Difference in Difference analysis to estimate the effect of FinTech algorithms on gender bias in credit decisions, controlling for other factors that may affect loan approval rates. The findings of this study have important implications for policymakers and financial service providers. The study suggests that adopting FinTech algorithms can be a powerful tool to promote gender equality in financial services. Policymakers can use these findings to promote the adoption of FinTech algorithms in banks and other financial institutions to reduce gender bias in credit decisions. Financial service providers can also use these findings to guide their adoption of FinTech algorithms and to develop strategies to mitigate gender bias in credit decisions. The remaining part of this study is organised as follows. The next section details the literature review, data, methodology and techniques to evaluate gender bias in credit decisions, study results and conclusion. JOURNAL OF APPLIED ECONOMICS 3 2. Literature review The rise of Financial Technology (FinTech) has brought significant changes to the banking industry (Ahmed, 2021; Makina, 2019; Varma et al., 2022). One of the most significant changes is the adoption of algorithms and machine learning in lending decisions (Ozili, 2021). FinTech algorithms are often marketed as an objective way to make lending decisions, which can reduce human bias and discrimination (Lui & Lamb, 2018). However, the impact of FinTech algorithms on gender inequality in bank loans is still a matter of debate (Bartlett et al., 2018; Berg et al., 2022; Chen et al., 2023). A growing body of literature has examined the relationship between FinTech algorithms and gender inequality in the financial industry (Lui & Lamb, 2018). On the one hand, some studies suggest that adopting FinTech algorithms in banks can reduce gender bias in lending decisions (Bartlett et al., 2018; Berg et al., 2022). For example, a study by Hall et al. (2021) found that using machine learning algorithms can reduce gender bias in lending decisions by up to 40%. Similarly, a study by Kozodoi et al. (2022) found that using credit scoring models can reduce gender bias in lending decisions on online platforms. On the other hand, other studies suggest that FinTech algorithms can perpetuate or even amplify gender inequality in lending decisions. For example, a study by Barocas and Selbst (2016) found that algorithmic decision-making can perpetuate discrimination because it relies on historical data, which can reflect and reinforce past discrimination. Another study by Datta et al. (2014) found that algorithmic lending decisions can amplify gender inequality because women are more likely to have gaps in their credit histories and, therefore, may be disadvantaged by algorithms that rely on credit history. Furthermore, studies have also shown that gender inequality in bank loans is still a pervasive issue in the United States. For example, a study by Hall et al. (2021) found that women are less likely to receive loans and more likely to be denied credit than men. Similarly, a study by Cohen-Cole and Gambacorta et al. (2019) found that women pay higher interest rates on loans than men, even after controlling for credit scores and other factors. In light of this literature, it is clear that the impact of FinTech algorithms on gender inequality in bank loans is complex and not yet fully understood. While some studies suggest that FinTech algorithms can reduce gender bias in lending decisions, others suggest that they can perpetuate or amplify gender inequality. It is also clear that gender inequality in bank loans is still a pervasive issue in the United States. Therefore, further research is needed to understand better the relationship between FinTech algorithms and gender inequality in lending decisions. In summary, the literature on FinTech suggests that algorithms and big data are used to streamline credit decisions and reduce human bias in decision-making. However, the gender bias in credit decisions has been well-documented in the literature. 2.1. Hypothesis development The topic of gender inequality in banking has been a persistent issue for decades, with evidence showing that women have historically faced difficulties in obtaining bank loans compared to men (Ahmed, 2021; Chen et al., 2023; De Andrés et al., 2021; Liu et al., 2022; 4Z. SONG ET AL. Shoma, 2019). A study by the National Women’s Business Council found that women in the United States were less likely to be approved for small business loans than men, even when applying for the same loan amount and under similar credit conditions. This disparity in loan approval rates has been attributed to various factors, including unconscious biases among loan officers and discriminatory lending practices (Blanco-Oliver et al., 2021; De Andrés et al., 2021). In recent years, FinTech companies have emerged as disruptors in the banking industry, offering innovative solutions to traditional banking processes (Sharif Abu Karsh, 2020). FinTech algorithms have been touted as potential tools for reducing gender bias in lending decisions (Adusei & Adeleye, 2022; Perry & Martin, 2022). These algorithms rely on data-driven models to make lending decisions, removing the potential for human biases to influence loan approvals (Adusei & Adeleye, 2022; Perry & Martin, 2022). However, limited research is available on these algorithms’ effectiveness in reducing gender bias in loan approvals. Given the growing importance of FinTech in the banking industry and the persistent issue of gender inequality in bank loans, it is crucial to examine the potential impact of FinTech algorithms on gender bias in lending decisions. This study aims to investigate whether implementing FinTech algorithms in the USA can reduce gender inequality in bank loans, using a quantitative approach to measure the effect of FinTech adoption on loan approval rates for men and women. Therefore, this study’s hypothesis is: H1 Using FinTech algorithms will decrease gender bias in credit decisions and promote gender equality in bank loans in the USA. Several theoretical perspectives support this hypothesis. The first is the social identity theory, which suggests that individuals’ identity, including their gender, can influence their experiences and interactions with others (Blanco-Oliver et al., 2021; Stets & Burke, 2000). The second is the statistical discrimination theory (Fang & Moro, 2011), which proposes that lenders may use demographic characteristics such as gender as a proxy for other unobserved factors that could influence lending decisions (Ahmed, 2021). The third is the technological determinism theory, which argues that technology, including FinTech algorithms, can significantly impact social and economic outcomes, including lending decisions (Oliver, 2011). Furthermore, this hypothesis is consistent with previous empirical research that has found evidence of gender bias in lending decisions, even after controlling for other factors. This research has also highlighted the potential for FinTech algorithms to reduce bias in lending decisions but has also shown that the algorithms can perpetuate or exacerbate existing inequalities if they are not designed and implemented appropriately. Overall, this hypothesis highlights the importance of examining the complex factors that contribute to lending decisions and the potential role of technology in shaping these outcomes. 3. Data and research methodology To estimate the effect of FinTech algorithms on gender bias in credit decisions, we use a logit regression analysis. We compare the loan approval rates for men and women JOURNAL OF APPLIED ECONOMICS 5 before and after adopting FinTech algorithms in 60 U.S. banks for 2012–2022. We obtained the loan application data from the banks’ internal records. We also include bank specific control variables such as bank size. We followed (Cheng & Qu, 2020; Zlatokrilov, 2021) and constructed and measured a bank FinTech index using web crawler technology and word frequency analysis. The FinTech index in our study was constructed using web crawler technology and word frequency analysis, following the methodologies outlined by Cheng and Qu (2020) and Zlatokrilov (2021). This approach allowed us to quantify the extent of FinTech adoption among the 60 U.S. banks included in our analysis throughout 2012–2022. The index measures the presence and intensity of FinTech-related activities and innovations within these institutions based on publicly available data from their websites and financial reports. To construct the FinTech index, we first identified a comprehensive list of keywords associated with FinTech innovations, such as “blockchain”, “peer-to-peer lending”, “machine learning”, “digital lending”, “Artificial intelligence”, and “big-data analytics”. The web crawler then searched the banks’ digital footprints for these keywords, while the word frequency analysis quantified the occurrence of each term. The resulting scores were normalised and aggregated to form a single FinTech index value for each bank. In our model, we proxied all FinTech based algorithms as FinTech adoption. To measure the Loan Approval rate for Male To measure the Loan Approval rate for Females where: Loan Approval Rate is the dependent variable, representing the probability of loan approval for male or female borrowers for a particular bank and year. We used web clever and text-mining techniques to measure the FinTech in banks.FinTech Adoption is a binary variable that takes a value of 1 if the bank has adopted FinTech after 2012 and 0 otherwise. Bank Ownership is a binary variable that takes a value of 1 if the bank is privately owned and 0 if it is government-owned. Bank Size is a continuous variable that represents the size of the bank in terms of its total assets. Bank Profitability is a continuous variable that represents the profitability of the bank. Loan Portfolio Mix is a continuous variable representing the loan type mix in the bank’s portfolio. The year is a continuous variable that represents the year of the observation. β0 is the intercept or constant term of the model. β1, β2, β3, β4, β5, and β6 are the coefficients of the 6Z. SONG ET AL. independent variables, representing the effect of each variable on the loan approval rate (ε) is the error term or random error, representing the unobserved factors that affect the loan approval rate but are not captured by the independent variables. Table 1 reports the descriptive statistics of the variables used in the regression analyses. Loan Approval Rate for Male and Female: The mean approval rates for male and female borrowers are around 75% and 67%, respectively. The approval rates for both genders are relatively narrow (25th percentile = 71% for males and 63% for females; 75th percentile = 79% for males and 71% for females). The skewness values suggest that both variables are moderately skewed to the right, meaning that most observations are clustered around the lower end of the distribution. The kurtosis values indicate that both variables have a relatively flat distribution compared to a normal distribution. FinTech Adoption The mean adoption rate is around 0.5, which suggests that approximately half of the banks in the dataset have adopted FinTech. The skewness value is close to zero, indicating that the distribution is approximately symmetric. The kurtosis value is negative, suggesting that the distribution is slightly flatter than a normal distribution. Bank Ownership: The mean ownership rate is around 0.5, which suggests that the sample is evenly split between private and government-owned banks. The skewness value is close to zero, indicating that the distribution is approximately symmetric. The kurtosis value is negative, suggesting that the distribution is slightly flatter than a normal distribution. Bank Size: The mean bank size is approximately 1 million, ranging from 500,000 to 1.5 million. The skewness value suggests that the variable is moderately skewed to the right, indicating that most observations are clustered around the lower end of the distribution. The kurtosis value indicates that the distribution is relatively flat compared to a normal distribution. Bank Profitability: The mean profitability rate is around 0.1, ranging from 0.06 to 0.14. The skewness value is close to zero, indicating that the distribution is approximately symmetric. The kurtosis value is negative, suggesting that the distribution is slightly flatter than a normal distribution. Loan Portfolio Mix: The mean portfolio mix is around 0.5, ranging from 0.3 to 0.7. The skewness value is close to zero, indicating that the distribution is approximately symmetric. The kurtosis value is negative, suggesting that the distribution is slightly flatter than a normal distribution. 4. Results and discussion 4.1. Impact of FinTech adoption on bank loans Before conducting our regression analysis, we assessed multicollinearity among our explanatory variables using variance inflation factors (VIFs). The values were 1.33 and 1.36 for both regression models, which indicated that high multicollinearity did not exist. We then used CD and Fisher tests to determine the appropriate estimation technique for our data and found panel data estimations. The result values for the Critical value for the CD test at 5% significance level = 0.382. Our calculated CD statistic value of 0.434 is greater than the critical value. Therefore, we cannot reject the null hypothesis of no crosssectional dependence and conclude that our data has cross-sectional dependence. We also performed the Fisher Test. The calculated Fisher statistic value, 12.54, is greater than JOURNAL OF APPLIED ECONOMICS 7 Loan Portfolio Mix, and Year are statistically significant at the 1% level, while the coefficients for Ownership and Bank Size are statistically significant at the 5% level. The coefficient of determination (R-squared) is 0.404, which means that this model explains 40.4% of the variation in the natural logarithm of the Loan Approval Rate (See Table 7). Using the natural logarithm of the Loan Approval Rate as the dependent variable transforms the variable into a logarithmic scale, which can have several advantages. For example, it can make the relationship between the dependent and independent variables more linear, improving the model’s fit. It can also help to reduce the influence of extreme values in the dependent variable, which can be helpful if the variable is highly skewed or has a long tail. However, using the natural logarithm of the Loan Approval Rate can also have some drawbacks. For example, it can be more challenging to interpret the coefficients of the independent variables since they are now measured in terms of percentage changes in the Loan Approval Rate rather than absolute changes. It can also make it more difficult to compare the results of this model to other studies that use the level of the Loan Approval Rate as the dependent variable. In general, whether to use the natural logarithm of the Loan Approval Rate or the level of the Loan Approval Rate as the dependent variable will depend on the specific research question and the properties of the data. Researchers should carefully consider the pros and cons of each approach and choose the most appropriate for their needs. 5.2. Conclusion Gender inequality is a significant issue in the financial services industry, where women often face discrimination in access to credit and financial products. It is well established in the literature that the financial services industry has historically been male-dominated. FinTech, with its use of algorithms and big data, has the potential to reduce gender bias in credit decisions and promote financial inclusion. FinTech can avoid human biases in traditional credit decision-making processes by relying on algorithms and machine learning. This study hypothesises that implementing FinTech algorithms will reduce gender inequality in bank loans in the USA. Using 60 U.S. banks’ internal data throughout 2012–2022, we measured the impact of FinTech based credit algorithms and their impact on bank loans both for Male and Female. Find the following results. We performed regression separately for both Male and Female. We found that banks that adopt FinTech have higher loan approval rates for male borrowers. For Female, the regression results indicate that the model explains 26.3% of the variance in loan approval rates for female borrowers (R-squared = 0.263). FinTech adoption is found to have a statistically significant positive relationship with loan approval rates for female borrowers (β1 = 0.0206, p < 0.001). We also estimate our Table 7. Shows the alternative specifications of the dependent variable. Variable Coefficient Standard Error t-statistic p-value Intercept −0.263 0.112 −2.35 0.020 FinTech Adoption 0.147 0.029 5.11 <0.001 Ownership 0.076 0.024 3.14 0.002 Bank Size 0.094 0.029 3.23 0.001 Bank Profitability 0.177 0.030 5.88 <0.001 14 Z. SONG ET AL. empirical results using a two-step GMM to reduce the potential endogeneity issue further. The results show that FinTech adoption has a positive and significant effect on loan approval rates, which suggests that banks that have adopted FinTech solutions are more likely to approve loans. This aligns with the idea that FinTech can improve the efficiency of loan approval processes and reduce costs, leading to higher approval rates. The DID analysis results show that policy and FinTech adoption significantly impact bank loans, specifically for female loan approval rates. Overall, the study’s results suggest that FinTech algorithms significantly reduce gender bias in bank credit decisions. This study contributes to understanding the potential of FinTech algorithms to reduce gender bias in credit decisions in the USA. Specifically, the study provides evidence that banks’ adoption of FinTech algorithms can significantly increase female loan approval rates and reduce gender inequality in bank loans. The study uses a rigorous two-step system GMM approach and Difference in Difference analysis to estimate the effect of FinTech algorithms on gender bias in credit decisions, controlling for other factors that may affect loan approval rates. The findings of this study have significant implications for policymakers and financial service providers. Our results suggest that adopting FinTech algorithms can be a powerful tool for promoting gender equality in financial services by reducing gender bias in credit decisions. Policymakers can utilise these findings to promote the widespread adoption of FinTech algorithms in banks and other financial institutions. Financial service providers can also leverage these results to guide their adoption of FinTech algorithms and develop strategies to mitigate gender bias in credit decisions. These actions could ultimately lead to greater financial inclusion and equal access to credit for all individuals, regardless of gender. Disclosure statement No potential conflict of interest was reported by the author(s). Funding The work was supported by the Center for Accelerating Financial Equity (CAFE). Notes on contributors Ziheng Song, associated with Columbia University and Columbia Uplift Inc. in Wyoming, is likely engaged in academia and industry, focusing on practical applications of academic research. Shafiq Ur Rehman works across the Lahore School of Business, The University of Lahore, and the School of Finance and Banking at the University of Utara Malaysia, contributing to business and finance education. Chun PingNg at Tsinghua University is involved in economic research, potentially influencing economic policies. Yuan Zhou at Columbia University engages in diverse academic disciplines. Dr. Patrick Washington at Morehouse College focuses on leadership and ethical business strategies. JOURNAL OF APPLIED ECONOMICS 15 Ricardo Verschueren at the University of Gloucestershire likely combines teaching and research in business, social sciences, or humanities. To verify their professional achievements and publications, one would need to access academic databases or contact them directly through their provided emails. ORCID Shafiq Ur Rehman http://orcid.org/0000-0003-1980-4167 References Adusei, M., & Adeleye, N. (2022). Credit information sharing and non‐performing loans: The moderating role of creditor rights protection. International Journal of Finance & Economics, 27 (4), 4756–4769. https://doi.org/10.1002/ijfe.2398 Ahmed, S. (2021). A gender perspective on the use of artificial intelligence in the African Fintech ecosystem: Case studies from South Africa. Andriamahery, A., & Qamruzzaman, M. (2022). Do access to finance, technical know-how, and financial literacy offer women empowerment through women’s entrepreneurial development? Frontiers in Psychology, 12, 776844. https://doi.org/10.3389/fpsyg.2021.776844 Arellano, M., & Bover, O. (1995). Another look at the instrumental variable estimation of errorcomponents models. Journal of Econometrics, 68(1), 29–51. Arner, D. W., Buckley, R. P., Zetzsche, D. A., & Veidt, R. (2020). Sustainability, FinTech and financial inclusion. European Business Organization Law Review, 21(1), 7–35. https://doi.org/ 10.1007/s40804-020-00183-y Awotunde, J. B., Adeniyi, E. A., Ogundokun, R. O., & Ayo, F. E. (2021). Application of big data with fintech in financial services. In Fintech with artificial intelligence, big data, and blockchain (pp. 107–132). Springer. Barocas, S., & Selbst, A. D. (2016). Big data’s disparate impact. California Law Review, 671–732. https://doi.org/10.2139/ssrn.2477899 Bartlett, R., Morse, A., Stanton, R., & Wallace, N. (2018). Consumer-lending discrimination in the era of fintech. Unpublished working paper. University of California, Berkeley. Berg, T., Fuster, A., & Puri, M. (2022). Fintech lending. Annual Review of Financial Economics, 14 (1), 187–207. https://doi.org/10.1146/annurev-financial-101521-112042 Blanco-Oliver, A., Reguera-Alvarado, N., & Veronesi, G. (2021). Credit risk in the microfinance industry: The role of gender affinity. Journal of Small Business Management, 59(2), 280–311. https://doi.org/10.1080/00472778.2020.1844487 Blundell, R., & Bond, S. (1998). Initial conditions and moment restrictions in dynamic panel data models. Journal of Econometrics, 87(1), 115–143. Chen, S., Doerr, S., Frost, J., Gambacorta, L., & Shin, H. S. (2023). The fintech gender gap. Journal of Financial Intermediation, 54, 101026. https://doi.org/10.1016/j.jfi.2023.101026 Cheng, M., & Qu, Y. (2020). Does bank FinTech reduce credit risk? Evidence from China. PacificBasin Finance Journal, 63, 101398. https://doi.org/10.1016/j.pacfin.2020.101398 Datta, A., Tschantz, M. C., & Datta, A. (2014). Automated experiments on ad privacy settings: A tale of opacity, choice, and discrimination. arXiv Preprint arXiv: 1408 6491, 2015(1), 92–112. https://doi.org/10.1515/popets-2015-0007 De Andrés, P., Gimeno, R., & de Cabo, R. M. (2021). The gender gap in bank credit access. Journal of Corporate Finance, 71, 101782. https://doi.org/10.1016/j.jcorpfin.2020.101782 Disemadi, H. S., Yusro, M. A., & Balqis, W. G. (2020). The problems of consumer protection in fintech peer to peer lending business activities in Indonesia. Sociological Jurisprudence Journal, 3 (2), 91–97. https://doi.org/10.22225/scj.3.2.1798.91-97 16 Z. SONG ET AL. Fang, H., & Moro, A. (2011). Theories of statistical discrimination and affirmative action: A survey. Handbook of Social Economics, 1, 133–200. https://doi.org/10.1016/B978-0-444-531872.00005-X Fine, C., Sojo, V., & Lawford‐Smith, H. (2020). Why does workplace gender diversity matter? Justice, organisational benefits, and policy. Social Issues and Policy Review, 14(1), 36–72. https:// doi.org/10.1111/sipr.12064 FINRA Investor Education Foundation. (2021). Financial well-being in America: Women 2021. Financial Industry Regulatory Authority. https://www.finra.org/sites/default/files/2021-02/ FINRA-Foundation-Financial-Well-Being-in-America-Women-2021.pdf Gabor, D., & Brooks, S. (2017). The digital revolution in financial inclusion: International development in the fintech era. New Political Economy, 22(4), 423–436. https://doi.org/10.1080/ 13563467.2017.1259298 Galli, E., Mascia, D. V., & Rossi, S. P. (2020). Bank credit constraints for women‐led SMEs: Self‐ restraint or lender bias? European Financial Management, 26(4), 1147–1188. https://doi.org/10. 1111/eufm.12255 Gambacorta, L., Huang, Y., Qiu, H., & Wang, J. (2019). How do machine learning and non-traditional data affect credit scoring? New evidence from a Chinese Fintech firm. Ghosh, S., & Vinod, D. (2017). What constrains financial inclusion for women? Evidence from Indian micro data. World Development, 92, 60–81. https://doi.org/10.1016/j.worlddev.2016.11. 011 Hall, P., Cox, B., Dickerson, S., Ravi Kannan, A., Kulkarni, R., & Schmidt, N. (2021). A United States fair lending perspective on machine learning. Frontiers in Artificial Intelligence, 4, 695301. https://doi.org/10.3389/frai.2021.695301 Hurley, M., & Adebayo, J. (2016). Credit scoring in the era of big data. Yale JL & Tech, 18, 148. Issa, A., Zaid, M. A., Hanaysha, J. R., & Gull, A. A. (2022). An examination of board diversity and corporate social responsibility disclosure: Evidence from banking sector in the Arabian Gulf countries. International Journal of Accounting & Information Management, 30(1), 22–46. https://doi.org/10.1108/IJAIM-07-2021-0137 Johnson, K., Pasquale, F., & Chapman, J. (2019). Artificial intelligence, machine learning, and bias in finance: Toward responsible innovation. Fordham L Rev, 88, 499. Kamga, C. C. K., Tchouassi, G., Tomo, C. P., & Danzoumbé, B. (2022). Financial inclusion and economic growth in the Central African economic and monetary community: Implications on gender balance. In Mainstreaming diversity, equity, and inclusion as future workplace ethics (pp. 200–218). IGI Global. Keller, W., Molina, T., & Olney, W. W. (2020). The gender gap among top business executives. Journal of Economic Behavior & Organization, 211, 270–286. https://doi.org/10.1016/j.jebo. 2023.04.017 Kozodoi, N., Jacob, J., & Lessmann, S. (2022). Fairness in credit scoring: Assessment, implementation and profit implications. European Journal of Operational Research, 297(3), 1083–1094. https://doi.org/10.1016/j.ejor.2021.06.023 Liu, J. J., Daly, K., & Mishra, A. V. (2022). Board gender diversity and bank risks: Evidence from Australia. Economic Analysis and Policy, 76, 1040–1052. https://doi.org/10.1016/j.eap.2022.10. 010 Ludwig, J., & Mullainathan, S. (2021). Fragile algorithms and fallible decision-makers: Lessons from the justice system. Journal of Economic Perspectives, 35(4), 71–96. https://doi.org/10.1257/ jep.35.4.71 Lui, A., & Lamb, G. W. (2018). Artificial intelligence and augmented intelligence collaboration: Regaining trust and confidence in the financial sector. Information & Communications Technology Law, 27(3), 267–283. https://doi.org/10.1080/13600834.2018.1488659 Madsen, S. R. (2020). Takeaways from“Diversity wins: How inclusion matters” by McKinsey & Company. Makina, D. (2019). 14 - The potential of FinTech in enabling financial inclusion. In D. Makina (Ed.), Extending financial inclusion in Africa (pp. 299–318). Academic Press. JOURNAL OF APPLIED ECONOMICS 17 Mogaji, E., & Nguyen, N. P. (2022). Managers’ understanding of artificial intelligence in relation to marketing financial services: Insights from a cross-country study. International Journal of Bank Marketing, 40(6), 1272–1298. https://doi.org/10.1108/IJBM-09-2021-0440 Moldovan, D. (2023). Algorithmic decision making methods for fair credit scoring. IEEE Access. Oliver, M. (2011). Technological determinism in educational technology research: Some alternative ways of thinking about the relationship between learning and technology. Journal of Computer Assisted Learning, 27(5), 373–384. https://doi.org/10.1111/j.1365-2729.2011.00406.x Ozili, P. K. (2021). Big data and artificial intelligence for financial inclusion: Benefits and issues. Artificial Intelligence Fintech, and Financial Inclusion. https://doi.org/10.2139/ssrn.3766097 Packin, N. G., & Lev-Aretz, Y. (2018). Learning algorithms and discrimination. In Research handbook on the law of artificial intelligence (pp. 88–113). Edward Elgar Publishing. Peng, W., & Zhu, F. (2021). Trust building and credit reporting with big data in the digital age. The Palgrave Handbook of Technological Finance, 809–836. Perry, V. G., & Martin, K. (2022). Algorithms for all: Has digitalisation in the mortgage market expanded access to homeownership? Available at SSRN 4126409. Purda, L., & Ying, C. (2022). Consumer credit assessments in the age of big data. In Big data in finance: Opportunities and challenges of financial digitalization (pp. 95–113). Springer International Publishing. Rehman, S. U., Al-Shaikh, M., Washington, P. B., Lee, E., Song, Z., Abu-AlSondos, I. A., & Allahham, M. (2023). FinTech adoption in SMEs and bank credit supplies: a study on manufacturing SMEs. Economies, 11(8), 213. Sadok, H., Sakka, F., & El Maknouzi, M. E. H. (2022). Artificial intelligence and bank credit analysis: A review. Cogent Economics & Finance, 10(1), 2023262. https://doi.org/10.1080/ 23322039.2021.2023262 Saiedi, E., Mohammadi, A., Broström, A., & Shafi, K. (2022). Distrust in banks and fintech participation: The case of peer-to-peer lending. Entrepreneurship Theory and Practice, 46(5), 1170–1197. https://doi.org/10.1177/1042258720958020 Salampasis, D., & Mention, A.-L. (2018). Chapter 18 - FinTech: Harnessing innovation for financial inclusion. In D. Lee Kuo Chuen, & R. Deng (Eds.), Handbook of blockchain, digital finance, and inclusion, volume 2 (pp. 451–461). Academic Press. Sargeant, H. (2023). Algorithmic decision-making in financial services: Economic and normative outcomes in consumer credit. AI and Ethics, 3(4), 1295–1311. https://doi.org/10.1007/s43681022-00236-7 Sharif Abu Karsh, Y. A. (2020). The new era of financial technology in banking industry. Journal of Southwest Jiaotong University, 55(4). https://doi.org/10.35741/issn.0258-2724.55.4.54 Shoma, C. D. (2019). Gender is a human rights issue: The case of women’s entrepreneurship development in the small and medium enterprise sector of Bangladesh. Journal of International Women’s Studies, 20(7), 13–34. Song, Z., Mellon, G., & Shen, Z. (2020). Relationship between racial bias exposure, financial literacy, and entrepreneurial intention: An empirical investigation. Journal of Artificial Intelligence and Machine Learning in Management, 4(1), 42–55. Stets, J. E., & Burke, P. J. (2000). Identity theory and social identity theory. Social Psychology Quarterly, 63(3), 224–237. https://doi.org/10.2307/2695870 Varma, P., Nijjer, S., Sood, K., Grima, S., & Rupeika-Apoga, R. (2022). Thematic analysis of financial technology (Fintech) influence on the banking industry. Risks, 10(10), 186. https://doi. org/10.3390/risks10100186 Washington, P. B., Rehman, S. U., & Lee, E. (2022). Nexus between regulatory sandbox and performance of digital banks—A study on UK digital banks. Journal of Risk and Financial Management, 15(12), 610. Yu, T. R., & Song, X. (2021). Big data and artificial intelligence in the banking industry. In Handbook of financial econometrics, mathematics, statistics, and machine learning (pp. 4025–4041). Zlatokrilov, L. S. (2021). Do words create reality? The development of fintech-banking as seen in financial reports. 18 Z. SONG ET AL.