The Effect of Financial Development on Unemployment in Emerging Market Countries

Abstract

This study examined the effect of financial development on unemployment in 19 emerging market countries, considering their age groups and gender dichotomy. The data covers the period from 1991–2019. Pooled Ordinary Least Square (OLS), Dynamic OLS, and quantile regression via moments were employed as the estimation methods. Robustness was tested with Fully Modified OLS and Canonical Cointegration Regression (CCR) estimation methods. Our results show that financial development has a conditional mean reducing effect on unemployment and a reducing effect on the distribution of unemployment. However, the reducing effect of financial development on the distribution of unemployment varies across the working-age population and youths. Thus, there is a need to formulate and implement a long-term financial policy to ensure economic growth and guarantee employment for the working-age population and the youths irrespective of gender.

Keywords

Financial development unemployment rate pooled OLS panel quantile regression

Introduction

Unemployment is one of the gravest socio-economic problems facing many emerging market countries (EMCs) in recent times (Horvath & Zhong, 2019; United Nations Department of Economic and Social Affairs [UN-DESA], 2020). The state of unemployment varies across these countries, with some experiencing skyrocketing unemployment rates while others have relatively lower rates. In Brazil, one of the leading economies in Latin America, unemployment rate rose from 6.66% in 2014 to 13.67% in 2020. A similar rise was observed in Chile, where unemployment rose from 6.67% to 11.51% within the same period. Other EMCs such as Colombia, Greece, and Egypt also suffered high rates of unemployment, which stood at 15.44%, 15.47%, and 10.33% in 2020, respectively.¹ However, other EMCs such as China, Czech Republic, Hungry, India, Indonesia, South Korea, Mexico, Palestine, Peru, Philippines, and Romania showed relatively lower unemployment rates.

The unemployment rate is one of the macroeconomic variables being constantly monitored by policymakers as it gauges the soundness of the economy. A high unemployment rate is a manifestation of an economic crisis that often casts doubt on the ability of policymakers to manage the economy effectively. Unemployment is closely related to the welfare of the citizens and the country. A country suffering from a high unemployment rate experiences an increase in not only socioeconomic crises such as a high rate of insecurity, kidnapping, prostitution, and robbery but also civil and political unrest (Oji & Afolabi, 2022). Besides, an unemployed person will likely suffer psychological issues, social exclusion, and health problems. Arising from this, several studies have been conducted to investigate the socio-economic, political, and psychological effects of unemployment (Arango & Flórez, 2020; Folawewo & Adeboje, 2017; Ogunjimi, 2021; Raifu, 2017; Raifu & Abodunde, 2020; Raifu et al., 2020). Among the factors identified is the country’s level of financial development (Ajide, 2020; Bayar, 2016; Raifu, 2019). The question, however, is, how does the level of financial development affect unemployment?

Economists have been interested in how financial development affects the economy (Bagehot, 1873; McKinnon, 1973; Schumpeter, 1911; Shaw, 1973). Levine (2005) argues that financial development improves the service delivery of the financial sector in five ways: the pooling of savings; allocation of capital to productive investments; supervision and monitoring of investments; risk-sharing and diversification; and exchange of goods and services. When these functions perform well, it is believed that financial development spurs growth, investment, and employment. However, the empirical studies to verify the nexus between financial development and the economy have produced mixed results. Most of the existing studies have focused on the effect of financial development on economic growth, investment, poverty, and inequality (Afolabi, 2022; Bijlsma et al., 2018; Isah & Soliu, 2016; Muyambiri & Odhiambo, 2018; Ni & Liu, 2019; Valickova et al., 2015). More recently, several studies have examined the effect of financial development on unemployment (Ajide, 2020; Bayar, 2016; Ibrahiem & Sameh, 2020; Ogbeide et al., 2016; Raifu, 2019). Among these, only Bayar (2016) examined the relationship between financial development and unemployment in EMCs. However, he did not find any significant relationship between financial development and unemployment in the 16 EMCs examined. The nonexistence of the nexus between financial development and unemployment could be attributed to factors such as the variable used to proxy financial development, the category of unemployment considered, and the estimation technique employed. In light of these, this study re-examines the effect of financial development on various categories of unemployment in EMCs.

This study is quite different from other existing studies, especially the study by Bayar (2016) on the EMCs. First, unlike Bayar (2016), who used only domestic credit to the private sector to proxy financial development, we used a comprehensive and broad-based measure of financial development by Svirydzenka (2016). Unlike domestic credit to the private sector, which only captures financial depth, Svirydzenka’s financial development index (also known as IMF financial development index) is a multidimensional measure of financial development that captures financial institutions and financial markets in terms of depth, access, and efficiency. This has shown to be a better measure of financial development in several empirical studies (Islam et al., 2020; Jiang & Ma, 2019; Opoku et al., 2019).

Second, we evaluate the impact of financial development on the aggregate unemployment rate like Bayar (2016). Still, we also extend the analysis to the disaggregated unemployment rate, especially gender unemployment for different age groups. Specifically, the age groups considered were 15–64 years and 14–25 years. Thus, we had six measures of unemployment: total unemployment, female unemployment, and male unemployment for the age bracket 15–64 years (working-age population); and total youth unemployment, female youth unemployment, and male youth unemployment for the age bracket 14–25 years. Examining the impact of financial development on different categories of unemployment has policy implications; it can help policymakers avoid the problem of policy formulation and implementation bias. Thus, being empowered with the knowledge of how financial development affects different categories of unemployed persons would assist them in formulating and implementing all-inclusive financial and labor policies.

Third, most of the estimation techniques used in previous studies, whether time series or panel methods, were based on estimating the conditional mean effects. However, such methods may not provide complete insights into the relationship between financial development and unemployment. In fact, Binder and Coad (2011) assert that these methods may underestimate or overestimate the influence of financial development on unemployment. Given this, it is essential to investigate how financial development affects the distribution of unemployment at different quantiles. This is important because it offers more information on how financial development influences unemployment at different periods. To investigate this, the panel quantile regression method becomes a useful estimation tool. The quantile regression, initially developed by Koenker and Basset (1978), has been improved upon by several researchers. Machado and Silva developed the recent quantile regression in 2019. This method has several advantages over the existing quantile regression methods. Apart from its ability to capture the effect of financial development on the distribution of unemployment over the different quantiles, it can also address the issues of endogeneity and heterogeneity characterizing panel data of countries via the instrumental variable. Besides, it can also accommodate the issue of fixed effects in panel data. Due to its advantages, it has been used to model several economic relationships at the panel level (Anwar et al., 2020; Guo et al., 2022). Hence, we adopted this method to model the effect of financial development on the distribution of unemployment at different quantiles. However, we began the analysis by employing the Pooled Ordinary Least Squares (POLS hereafter) that accounts for individual country and year-specific effects and Dynamic Ordinary Least Squares (DOLS hereafter). As robustness to POLS and DOLS, we also used Fully Modified Ordinary Least Squares (FMOLS hereafter) and Canonical Cointegration Regression (CCR). These methods were used as benchmark estimation methods.

The rest of the study is structured as follows: The next section presents some stylized facts on the relationship between financial development and gender unemployment in EMCs. This is followed by a review of the extant literature on financial development and unemployment. The methods, data sources, and description are presented next, and the subsequent section focuses on the empirical findings and concludes with policy implications.

Stylized Facts About Financial Development and Unemployment

This section focuses on documenting some stylized facts about financial development and unemployment (total and disaggregated unemployment). Figure 1 represents the average level of financial development in 19 EMCs.

Figure 1.

Source: Authors’ analysis using the International Monetary Fund (2020) database.

It clearly shows that the level of financial development varies across EMCs. South Korea has the highest level of financial development, which has been attributed to the series of reforms undertaken since the 1980s that led to a significant reduction in the involvement of the South Korean government in the financial sector (Adnan, 2014). Malaysia follows this with an average level of financial development of 0.58. Thailand occupies the third position, followed by China, Greece, Brazil, Russia, Chile, and Turkey. Peru has the worst level of financial development among the countries. However, this does not connote that the financial sector in Peru is at a low ebb. Its level of financial development is relative to the context of other EMCs. Evidence shows that Peru has recently witnessed credit expansion (Morón et al., 2013).

Figure 2 shows the average total female and male unemployment rate for the adult workforce between 15 and 64 years from 1991–2019. The unemployment rate also differs across the 19 EMCs. Greece has the highest rate of total unemployment, which on average stands at 13.5%. This is not surprising since the Great Financial Crisis of 2008–2009, Greece has witnessed a persistent unemployment rise. This is followed by Colombia, Egypt, Turkey, and Brazil, with an average unemployment rate of 11.6%, 10.3%, 9.1%, and 8.5%, respectively. Pakistan has the least unemployment rate of 1.2%, followed by Thailand at 1.3%. Regarding gender unemployment, female unemployment is higher than male unemployment in almost all countries—a sign of the persistent gender gap in employment. Egypt has the highest female unemployment rate, followed by Greece, Colombia, Brazil, Turkey, and Chile. The female unemployment figure is higher in Egypt due to many factors, including, a contraction in public hiring of workers, long educational pursuit, dwindling economic conditions, and marital demands (Hendy, 2015). On the other hand, male unemployment is higher in Greece than in the rest of the 18 EMCs.

Figure 2.

Source: Authors’ analysis using the World Bank (2020) database.

Figure 3 shows the categories of youth unemployment including total, male, and female unemployment rates. Greece has the highest youth unemployment rate, followed by Egypt, Colombia, Chile, India, Hungary, Brazil, Turkey, and Indonesia. The average youth unemployment rate in Greece stood at 33% while in Egypt at 28%, Colombia at 22%, Chile and India at 19%, Hungary at 18%, and Brazil, Turkey, and Indonesia at 17%. The female youth unemployment rate is high in countries such as Egypt, Greece, Colombia, Chile, and Brazil, with Egypt reporting the highest female youth unemployment rate. The male unemployment rate, on the other hand, is higher in Greece than in the rest of the countries. It is also evident that the youth gender unemployment gap is more prevalent than the observed adult gender unemployment gap.

Figure 3:

Source: Authors’ analysis using the World Bank (2020) database.

The scatter plot graphs in Figures 4 –9 show the correlation between financial development and various categories of unemployment rates, that is, total unemployment rate (Figure 4), male unemployment rate (Figure 5), female unemployment rate (Figure 6), total youth unemployment (Figure 7), male youth unemployment rate (Figure 8), and female youth unemployment rate (Figure 9).

Figure 4.

Source: Authors’ analysis using data from the International Monetary Fund (2020) and World Bank (2020) database.

Figure 5.

Source: Authors’ analysis using data from the International Monetary Fund (2020) and World Bank (2020) database.

Figure 6.

Source: Authors’ analysis using data from the International Monetary Fund (2020) and World Bank (2020) database.

Figure 7.

Source: Authors’ analysis using data from the International Monetary Fund (2020) and World Bank (2020) database.

Figure 8.

Source: Authors’ analysis using data from the International Monetary Fund (2020) and World Bank (2020) database.

Figure 9.

Source: Authors’ analysis using data from the International Monetary Fund (2020) and World Bank (2020) database.

Evidence from all the graphs shows a slightly downward-sloping relationship between financial development and the various categories of unemployment. This implies that a negative relationship exists between financial development and unemployment. Intuitively, an increase in the level of financial development should lead to a reduction in the various categories of unemployment. However, it can be observed that the negative relation between financial development and unemployment varies across countries, and it is likely to be weak in countries such as Greece, Egypt, Colombia, Hungary, and Pakistan. However, a strong negative correlation between the two macroeconomic variables is observed in countries such as South Korea, China, Indonesia, Czech Republic, and Malaysia.

Literature Review

There is a considerable amount of literature on the nexus between financial development and economic growth, investment, poverty, and inequality (Arestis et al., 2015; Bijlsma et al., 2018; Cepparulo et al., 2017; De Haan & Sturm, 2017; Iheonu et al. 2020; Muyambiri & Odhiambo, 2018; Valickova et al., 2015; Younsi & Bechitini, 2018). A recently growing strand of study examines how financial development affects unemployment or employment (Bayar, 2016; Borsi, 2018; Çiftçioğlu & Bein, 2017; Kim et al., 2019). The study has been dichotomized along with the theoretical and empirical nexus between financial crisis or development and unemployment.

From a theoretical perspective, there appears to be no consensus on how financial development affects unemployment. While some theorists posit that financial development promotes employment, others argue that the financial sector is subject to the uncertainty that is detrimental to employment and may even worsen unemployment. Hence, financial development may not promote employment but may rather worsen the unemployment situation. For the proponent of a positive nexus between financial development and employment, it is argued that financial development promotes employment. Pagano and Mica (2012) opined that financial development would promote employment depending on whether factor inputs such as capital and labor are substitutes or complements. If the capital and labor are complements, investors would invest in labor-enhancing technology that spurs labor productivity and promotes employment. Conversely, if they are substitutes, the availability of financial credit would make the investors substitute capital for labor, investing in capital-intensive equipment. This would result in the retrenchment of workers, thereby leading to unemployment.

The proponents of the negative effect of financial development on employment provide three channels through which financial development can negatively influence employment. These channels include credit constraint, imperfect market, and powerful incumbent interest. When firms face credit constraints, especially during a financial crisis, they may find it challenging to produce optimally compared to when there are no financial constraints. In other words, credit constraint leads to low productivity, resulting in the laying-off of workers (Dromel et al., 2010). According to Wasner and Weil (2004), market imperfection results in the prevalence of asymmetric information in the financial sector. Such existence of asymmetric information, in most cases, leads to a high cost of borrowing, which prevents the creation of new businesses and jobs. Acemoglu et al. (2005) contend that existing investors, known as incumbents, who benefit from the current low level of financial development through racketeering and rent-seeking, would not want policy reform that creates competition leading to financial development. These incumbents would ensure that the financial sector remains underdeveloped. Hence, only limited jobs will be created with low-level financial and credit facilities.

Empirically, several studies conducted across countries found mixed results. This could be attributed to several factors, including the measure of financial development used, the level of financial development, the country studied, and the estimation techniques employed. Regarding country-specific studies, Shabbir et al. (2012) investigated the effect of financial development on unemployment in Pakistan using Autoregressive Distributed Lag (ADRL). Their findings revealed that financial development and unemployment in conjunction with other control variables have a long-run relationship. However, financial development was found to be positively related to unemployment. Similar findings were found in studies conducted by Ogbeide et al., (2016) and Ibrahiem and Sameh (2020) for Nigeria and Egypt, respectively. In specific terms, Ogbeide et al. (2016) examined whether resource endowment and financial development determine unemployment in Nigeria. Using the Error Correction Method (ECM) and Ordinary Least Squares (OLS), they documented that financial development worsens unemployment. In the case of Ibrahiem and Sameh (2020), they investigated how clean energy sources and financial development affect unemployment. Their findings established that financial development had a positive effect on unemployment, indicating that financial development worsens unemployment. Aliero et al. (2013), on the other hand, found that formal credit allocation to the rural area abates unemployment in Nigeria. Raifu (2019) found that the reducing effect of financial development on unemployment depends on the measures of financial development. For instance, financial system deposit expressed as a ratio of GDP reduces unemployment in the short-run and the long-run. Other measures of financial development such as credit to the private sector, financial liquidity, and financial stability only have a reducing effect on unemployment in the short run.

Still, from single-country studies, some authors argue that the relationship between financial development and unemployment cannot be linear. Two possible explanations have been presented. First, it is believed that financial development and unemployment behave nonlinearly over time. Second, the response of economic agents to an economic situation such as financial crisis and normalcy varies. In other words, economic agents respond more quickly to a negative event in the financial market than to a positive event. Hence, firms that provide employment are more likely to respond quickly to negative news in the financial sector than the positive news (Shahbaz et al., 2015). Thus, modelling the nexus between financial development and unemployment could yield biased results with low explanatory or testing power. As a result, some studies have attempted to examine the nonlinear relationship between financial development and unemployment. For instance, Ajide (2020) assessed the asymmetric effect of financial development on unemployment in Nigeria by employing linear and nonlinear ARDL estimation methods. The summary of his findings indicates the existence of asymmetric nexus between financial development and unemployment in the short-run and the long-run. The specific results show that positive shocks to financial development led to a reduction in unemployment in the short run. In the long run, both positive and negative shocks to financial development appear to lead to a reduction in unemployment.

We now turn to the review of studies that examine the relationship between financial development and unemployment in the panel of countries. The study conducted by Gatti et al. (2012) for 18 Organization for Economic Co-operation and Development (OECD) countries using the generalized method of moments (GMM) technique reveals that the effect of financial development on unemployment depends on the labor market and financial market conditions. In specific terms, they showed that an increase in market capitalization and decrease in bank concentration tends to reduce unemployment given the rate of labor market regulations, union density, and wage structure or bargaining power. For OECD and non-OECD countries, Pagano and Pica (2012) developed a model that shows how financial development influences employment and labor reallocation. It shows that financial development spurs employment, particularly in non-OECD countries. The positive effect is, however, conditional to the state of the economy. Hence, when the economy is in crisis, there is less employment in the firms that depend on external sources for financing their operations. The studies by Feldmann (2012; 2013) show that the level of sophistication of financial development determines whether financial development would lead to a reduction in unemployment. Thus, he concluded that the higher the level of financial sophistication, the lower the level of unemployment in 21 industrial countries, especially among highly skilled workers. However, for low-skilled workers, the higher the level of financial sophistication, the higher the rate of unemployment.

Even though the earlier review supports the conditional reducing effect of financial development on unemployment, some studies find that financial development hurts employment under certain conditions. Borsi (2018) examined the response of labor market performance to credit contraction in 20 OECD countries using the local projection estimation method. He found that total, youth, and long-term unemployment reacted positively to a decline in private credit. Besides this, he submitted that sporadic boom in credit is associated with some level of unemployment. Therefore, the sources of unemployment in OECD countries are credit contraction and excessive labor market rigidity. Kim et al. (2019) also evaluated the effect of financial development on unemployment in some advanced and developing countries. Their findings confirm that rigid market regulation and bank concentration raise the level of unemployment. Çiftçioğlu and Bein (2017) examined the relationship between financial development and unemployment between 1991 and 2012 in some selected European countries including Germany, France, the UK, Italy, Spain, Greece, Portugal, Poland, Sweden, and Finland. Evidence from running a series of regressions shows that the reducing effect of financial development on unemployment depends on the measures of financial development adopted. Bayar (2016) studied the impact of financial development on unemployment in 16 emerging countries and concluded that no significant relationship exists between financial development and unemployment. This means that financial development does not reduce unemployment in the group of selected emerging countries. He, however, found that Granger causality exists between financial development and unemployment, with the direction of causality running from financial development to unemployment.

The reviews reveal one important fact: the empirical findings are yet to be synchronized. This study contributes to the existing literature, especially on EMCs, by examining the conditional means and distributional effects of financial development on unemployment in 19 emerging economies based on Morgan Stanley Capital International (MSCI) emerging market classification. We consider the effect of financial development on adult (15–65 years) and youth (14–25 years) unemployment across gender: male and female.

Methodology and Data Sources

Methodology

To investigate the effect of financial development on the different categories of unemployment, we employed five main estimation methods: POLS, DOLS, FMOLS, CCR, and Machado and Silva’s (2019) quantile regression via moments (instrumental variable-IV). The POLS, DOLS, FMOLS, and CCR are used as benchmark methods, while Machado and Silva’s (2019) method is used for the main analysis. While the POLS, DOLS, FMOLS, and CCR methods estimate the conditional mean effect of financial development on unemployment, the quantile regression estimates the distributional effects of financial development on unemployment at different quantiles.

Before we present estimation frameworks, we must state that some preliminary tests were performed. These tests included a correlation test, unit root test, cross-sectional dependence test (CDT), and panel Granger causality test. In an empirical study, correlation analysis is conducted to show the strength of the relationship between financial development and unemployment, and to determine whether multicollinearity is present among the regressors. We employed the Im-Pesaran-Shin (Im et al., 2003) unit root test to examine the stationarity property of our variable. The method assumes the non-stationarity of the variables. The null hypothesis is tested against the alternative hypothesis of stationarity of the variables. If the variables are not stationary at the level, the variables actually contain unit roots and only become stationary after the first difference. For the CDT, we used three methods: Frees (1995) cross-sectional dependence (CD), Friedman’s (1937) CDT, and Pesaran’s (2004, pp. 1–39) CDTs to ensure robustness. Ascertaining the existence of CD in the panel dataset is essential because ignoring it could lead to estimation bias (Iheonu et al., 2019). For the causality test, we used Dumitrescu and Hurlin (2012) Granger non-causality designed for heterogeneous panel data models. This panel causality method has some advantages. First, it considers the existence of CD across the panel model. Second, it considers two dimensions of heterogeneity: heterogeneity due to causal relationships and heterogeneity of the regression used for testing Granger causality.

The specification of our model begins with POLS that accounts for individual country- and year-specific effects. Following Raifu et al. (2021) and Raifu (2022), the POLS framework is specified as follows:

u n p l_{i t} = α_{0} + α_{1} f d_{i t} + α_{2} r g d p g r_{i t} + α_{3} \inf_{i t} + α_{4} s s e_{i t} + α_{5} f d i_{i t} + φ d_{i} + ϕ d_{t} + ε_{i t}

(1)

From Equation 1, $u n p l$ is the unemployment rate (for youths and adults of different genders), $f d$ captures financial development, $r g d p g r$ is the real GDP growth rate, $\inf$ is the inflation rate, $s s e$ is the secondary school enrolment, $f d i$ captures foreign direct investment (FDI), and $d_{t}$ is the dummy variable capturing individual country-specific effect as well as a dummy variable that captures year-specific effects. Apart from financial development, we controlled for other explanatory variables in Equation 1, including real GDP growth rate, inflation rate, secondary school enrolment, and FDI. Their inclusion in the model is based on their theoretical relationship with unemployment. Theoretically, Okun’s law stipulates an inverse relationship between unemployment and economic growth. In other words, unemployment is expected to decline when the economy is growing. Therefore, a negative relationship should exist between unemployment and economic growth. In the case of inflation, the Phillip Curve hypothesis also shows that inflation and unemployment are inversely related. Thus, inflation and unemployment are expected, as a priori, to have a negative relationship. Based on growth theory (endogenous growth theory in particular), the more people are enrolled in school, the more human capital is developed, spurring economic growth, labor productivity, and employment, and thereby reducing unemployment. However, poor economic conditions and low education standards in the face of an increase in school enrolment may worsen unemployment (Adejumo et al., 2021). Thus, the relationship between school enrolment and unemployment could range from positive and negative to no relationship. FDI could positively or negatively affect unemployment depending on several factors, such as the political and institutional structure of the host country and the way the FDI is established (Alalawneh & Nessa, 2020). Regarding financial development, we expect a negative effect on unemployment. Improvement in financial development should help firms access finances, enabling them to invest more. With more investment that raises the production level, firms could employ additional workers, thereby reducing the unemployment rate (Ogunjimi, 2019).

Next, we present the DOLS estimation framework as formalized by Kao et al. (1999). In the presence of endogeneity between financial development and unemployment, the POLS estimator may be biased or inefficient.² Thus, the DOLS estimator employs the lead and lag of independent variables to correct for endogeneity bias. Following Neal (2014), Muye and Muye (2017), and Raifu and Adeboje (2022), the panel DOLS is specified as follows:

\begin{array}{l} u n p l_{i t} = α_{0} + α_{1} f d_{i t} + α_{2} r g d p g r_{i t} + α_{3} \inf_{i t} + α_{4} s s e_{i t} + α_{5} f d i_{i t} + \sum_{j = - 1}^{p} μ_{i} Δ f d_{i t} + \sum_{j = - 1}^{p} ϕ Δ r g d p g r_{i t} + \\ \sum_{j = - 1}^{p} φ_{i} Δ \inf_{i t} + \sum_{j = - 1}^{p} γ_{i} Δ s s e_{i t} + \sum_{j = - 1}^{p} λ_{i} Δ f d i_{i t} + ε_{i t} \end{array}

(2)

In Equation 2, all variables remain as previously defined, while $p$ is the optimal lead and lag lengths, and $Δ$ is the change in independent variables.

We will now specify Machado and Silva’s (2019) quantile regression. Assuming that unemployment is denoted as $Y_{i t}$ , whose distribution is conditional on the set of a vector of variables such as financial development, real GDP growth rate, inflation rate, secondary school enrolment and FDI, denoted together as $X_{i t}$ with location-scale variant, the quantile regression for this relationship can be specified as follows:

Y_{i t} = α_{i} + X_{i t}^{'} β + (δ_{i} + Z_{i t}^{'} γ) U_{i t}

(3)

‘here ${(α, β, δ, γ)}^{'} \in R^{2 (k + 1)}$ are the unknown parameters required to be estimated as the conditional probability. The individual fixed effect denoted as $(α_{i}, δ_{i}), i = 1, ..., n$ , $Z$ , k-vector components of $X_{i t}$ , are transformed by differentiating the element $l$ given as $Z_{i} = Z_{i} (X), l = 1, ..., k$ . $X_{i t}$ follows the OLS assumptions of normality that is independently and identically distributed for a fixed $i$ and is also independent across time, $t$ . $U_{i t}$ has three unique characteristics: First, it follows a standard normal distribution assumption across individuals, $i$ , and time, $t$ . Second, it is orthogonal to $X_{i t}$ , that is, it is uncorrelated with $X_{i t}$ . Third, it satisfies moment conditions after being normalized. Given this, Equation 1 can be expressed in quantile regression form as follows:

Q_{Y} (τ / X_{i t}) = (α_{i} + δ_{i} q (τ) + X_{i t}^{'} β + Z_{i t}^{'} γ q (τ)

(4)

As previously stated, $X_{i t}$ are the vector of explanatory variables. $Q_{Y} (τ / X_{i t})$ shows the quantile distribution of our dependent variable, unemployment rate, and it depends on the location regressors $X_{i t}$ . $α_{i} (τ) \equiv α_{i} + δ_{i} q (τ)$ is referred to as the scalar coefficient indicating that the quantile, $τ$ , is the fixed effect of an individual country, $i$ . The individual fixed effects are time-invariant parameters whose homogenous impacts are allowed to differ across the quantiles of conditional distributions of the endogenous variable, $Y$ . $q (τ)$ denotes the $τ - t h$ sample quantile estimated by solving the following optimization problem given as follows:

\min_{q} \sum_{i} \sum_{t} ρ_{i} (R_{i t} - (δ_{i} + Z_{i t}^{'} γ) q)

(5)

where $ρ_{τ} (A) = (τ - 1) A I \{A \leq 0\} + T A I \{A > 0\}$ and denotes the check function. We estimate the quantile regression through the moment conditions of the cross-sectional panel models characterized by endogeneity among the regressors. This approach is referred to as instrumental variable quantile regression (IVQR). In our estimation, we use lags of regressors as instruments. The quantile regression is conducted for different quantiles such as 0.10th, 0.25th, 0.50th, 0.75th and 0.90th quantiles.

Data Sources

We employed the data of 19 EMCs³ selected, based on MSCI classification. The included countries are guided by the availability of relevant data. The data period ranges from 1991–2019. The main dependent variable is the unemployment rate for two different age groups and genders. We use total, male, and female unemployment rates for the age bracket 15–64 years and call it the adult unemployment rate. We also consider the total, male, and female unemployment rate for the age bracket 14–25 years and call it the youth unemployment rate. According to International Labour Organisation (ILO), the unemployment rate is defined as the proportion of the labor force that has no job and is available and actively looking for work. It is an indicator that reflects labor market conditions and also captures the unutilized labor supply in a country.⁴ Financial development is captured by a broad-based IMF financial development index. It is an index that encompasses all aspects of the financial sector such as financial institutions and financial markets, especially in terms of their depth, access, and efficiency. Financial institutions broadly consist of three indicators: financial institution depth (FID), financial institution access (FIA), and financial institution efficiency (FIE). Similarly, financial markets are also classified into three groups, which include financial market depth (FMD), financial market access (FMA), and financial market efficiency (FME). Financial development is, therefore, defined in this sense as a combination of depth (size and liquidity of markets), access (the ability of individuals or companies to have access to financial services), and efficiency (the ability of financial sectors to provide financial services at the lowest cost possible and with sustainable revenue as well as the level of activity of capital markets). The index is computed using a three-step approach, which is the normalization of the variables, aggregation of normalized variables into different sub-indices denoting a particular measurement of financial dimension, and aggregation of sub-indices to the final financial development index (Svirydzenka, 2016). We also make use of real GDP growth rate, inflation rate, secondary school enrolment, and FDI as control variables. Their inclusion, as previously mentioned, is based on their theoretical relationship with unemployment. GDP growth rate is referred to as a change in economic output on a yearly or quarterly basis. It captures how fast the economy is growing. The inflation rate, on the other hand, is defined as the percentage increase or decrease in the general price level during a given period, especially on a monthly or yearly basis. It captures how fast the prices rise or fall during the period. The sources and units of measurement of the variables are presented in Table 1.

Table 1.

Variables, Units of Measurement and Sources.

Variables	Units of Measurement	Sources
Financial development index	–	IMF (https://data.imf.org/?sk=F8032E80-B36C-43B1-AC26-493C5B1CD33B)
Total unemployment (15–64)	Percentage of total labor force	WDI (2020) version – ILO estimate
Female unemployment (15–64)	Percentage of female labor force	WDI (2020) version – ILO estimate
Male unemployment (15–64)	Percentage of male labor force	WDI (2020) version – ILO estimate
Total youth unemployment rate (14–25)	Percentage of total labor force	WDI (2020) version – ILO estimate
Female youth unemployment rate (14–25)	Percentage of female labor force	WDI (2020) version – ILO estimate
Male youth unemployment rate (14–25)	Percentage of male labor force	WDI (2020) version – ILO estimate
Real GDP growth rate	Percentage	WDI (2020) version
Inflation rate	Percentage	WDI (2020) version
Secondary school enrolment	Gross percentage	WDI (2020) version
Foreign direct investment	Percentage of GDP	WDI (2020) version

Note: IMF = International Monetary Fund. WDI = World Development Indicators.

Empirical Results

This section presents the results of the effect of financial development on unemployment in EMCs. In the preliminary findings, we present the results of summary statistics, pairwise correlation, Im-Pesaran-Shin (Im et al., 2003) unit root test, CD, and Dumitrescu and Hurlin’s (2012) panel Granger non-causality tests. Five main results are presented: POLS, DOLS, FMOLS, CCR, and Machado and Silva’s (2019) panel quantile regression via moments.

Preliminary Results

Descriptive Statistic Results and Pairwise Correlation Results

Table 2 presents the descriptive statistics of the variables used in the analysis. The average total unemployment rate is higher among females than males, although the gender gap in unemployment is relatively low. The youth (14–25 years) are more jobless than adults, and unlike the narrow unemployment gap between female and male adults, the unemployment gap between female and male youths is relatively high. Based on the overall financial development index computed by IMF, the average level of financial development in the selected emerging market stood at 0.42 with a minimum value of 0 and a maximum value of 0.85. The selected countries grow at an average of 4.09%, a relatively robust economic growth. The inflation rate is relatively high in the countries, indicating that most are facing inflationary challenges, apart from the problem of unemployment amid robust economic growth. Average secondary school enrolment stood at 80.40%, ranging from 22.51% to a maximum of 120.65%. Furthermore, the FDI statistics suggest that some countries in emerging markets experienced an outflow of FDI instead of an inflow.

Table 2.

Descriptive Statistics and Correlation Results.

Descriptive Statistics
Variables	Obs	Mean	SD	Min	Max	p1	p99	Skew.	Kurt.
Total unemployment	532	6.242	3.936	0.210	27.47	0.580	21.490	1.568	7.897
Female unemployment	532	7.623	5.958	0.200	31.38	0.300	26.120	1.469	4.988
Male unemployment	532	5.573	3.304	0.210	24.5	0.590	17.850	1.710	9.000
Youth unemployment	532	14.778	8.681	0.830	58.000	0.980	44.300	1.126	5.627
Female youth unemployment	532	17.124	11.984	0.500	63.99	0.590	53.050	1.273	4.492
Male youth unemployment	532	13.455	7.389	0.910	53.55	1.040	39.280	1.188	6.751
Financial development	532	0.424	0.144	0.000	0.853	0.155	0.833	0.625	3.484
Real GDP growth rate	531	4.092	3.829	−14.531	14.231	−8.669	12.308	−1.141	6.333
Inflation rate	529	22.284	137.976	−1.736	2075.89	-.846	409.53	12.379	169.062
Secondary school enrolment	532	80.402	20.347	22.511	120.651	25.27	114.185	−0.897	3.271
Foreign direct investment	532	2.823	4.853	−40.081	54.169	−2.642	11.743	4.330	67.503

Pairwise Correlations Results
Variables	lunemplt	lunemplf	lunemplm	lunemplyt	Lunemplyf	lunemplym	lfd	rgdpgr	inf	lsse	fdi
lunemplt	1
lunemplf	0.973*	1
lunemplm	0.981*	0.917*	1
lunemplyt	0.946*	0.942*	0.922*	1
lunemplyf	0.920*	0.954*	0.869*	0.982*	1
lunemplym	0.934*	0.901*	0.939*	0.985*	0.938*	1
lfd	−0.049	−0.070	−0.022	0.046	0.029	0.069	1
rgdpgr	−0.195*	−0.177*	−0.189*	−0.165*	−0.160*	−0.152*	0.100*	1
inf	0.045	0.044	0.045	0.007	0.015	0.002	−0.079	−0.089*	1
lsse	0.507*	0.511*	0.506*	0.567*	0.584*	0.554*	0.285*	−0.201*	0.052	1
fdi	0.072	0.065	0.079	0.062	0.057	0.068	0.068	0.022	−0.057	0.113*	1

Notes: *shows significance at the 0.05 level.

We also performed a pairwise correlation test. The results in Table 2 indicate that financial development is negatively associated with unemployment, albeit the correlation is not statistically significant. On the other hand, the GDP growth rate is negatively and significantly correlated with unemployment. However, the positive correlation between inflation rate and financial development is not statistically significant. Also, secondary school enrolment and FDI positively correlate with unemployment, albeit only secondary school enrolment is statistically significant. Among the regressors, there is no multicollinearity as the correlation coefficients are relatively low.

lunemplt, lunemplf, lunemplm, lunemplyt, lunemplyf, lunemplym, lfd, regdpgr, inf, lsee, and fdi denote total unemployment, female unemployment, male unemployment, total youth unemployment, female youth unemployment, male youth unemployment, real GDP growth rate, inflation rate, secondary school enrolment and foreign direct investment, respectively.

L preceding some variables means that those variables are transformed in logarithm form.

Unit Root Test and Cross-sectional Dependence Test Results

We performed a panel unit root test to ascertain the stationarity properties of our variables using the Im-Pesaran-Shin unit root test. As seen from the results in Table 3, most of the variables are only stationary after the first difference. These variables include the total unemployment rate, female unemployment, male unemployment, youth unemployment, female youth unemployment, male youth unemployment, and secondary school enrolment. Financial development, GDP growth rate, inflation rate, and FDI are all stationary at levels, implying that these variables do not contain a unit root.

We also performed a CDT. Within the panel data structure, CD usually arises because of the unobserved common factors. The unobserved common factors can come in the forms of both internal and external shocks, such as financial crises, oil prices, and health shocks. The CD is common in empirical studies due to the level of economic integration and globalization of economic activities (De Hoyos & Sarafidis, 2006). The existence of the unobserved common factors has the potential to affect the efficiency of the estimated panel parameters or coefficients. Phillips and Sul (2003) submitted that the efficiency of the estimated panel parameters declines due to the presence of unobserved common factors. In this study, we employed three CDTs: Frees CDT, Freidman CDT, and Pesaran CDT. After running the fixed effects regression for all the models, we carried out the three CDTs, as reported in Table 3. Beginning from Pesaran CDT, it is evident that the null hypothesis of CD is rejected at a 1% significance level in all the models. This implies the existence of CD among the variables in the models of financial development and unemployment in EMCs. This could be attributed to the level of economic integration among the countries. To ascertain the robustness of Pesaran CDT, we also reported the results of Frees CDT and Friedman CDT, which affirm the results we obtained from the Pesaran CDT, suggesting that the null hypothesis of no CD among the variables in the models is rejected.

Table 3.

Unit Root and Cross-sectional Dependence Test Results.

Im-Pesaran-Shin Unit-root Test Results
Variable		Level		First Difference
Total unemployment		−1.23640.8618		−3.1287***0.0000
Female unemployment		−1.51240.3290		−3.3491***0.0000
Male unemployment		−1.20280.8988		−3.1491***0.0000
Youth unemployment		−1.38920.5951		−3.1760***0.0000
Female youth unemployment		−1.6659*0.0977		−3.4011***0.0000
Male youth unemployment		−1.36080.6547		−3.1890***0.0000
Financial development		−1.9559***0.0018		−3.7523***0.0000
Real GDP growth rate		−2.9848***0.0000		−4.0738***0.0000
Inflation rate		−2.8844***0.0000		−3.9915***0.0000
Secondary school enrolment		−1.27450.8925		−4.6103***0.0000
Foreign direct investment		−2.6403***0.0000		−5.8188***0.0000
Cross-Sectional Dependence Test Results
Test	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6
Frees	2.247***0.1870	2.405***0.1870	2.285***0.1870	1.701***0.1870	1.813***0.1870	1.846***0.1870
Friedman	57.893***0.0000	54.698***0.0000	59.957***0.0000	49.409***0.0001	44.022***0.0006	54.305***0.0000
Pesaran	4.694***0.0000	4.927***0.0000	4.466***0.0000	3.137***0.0017	3.193***0.0014	3.022***0.0025

Notes: *, **, and *** denote 10%, 5%, and 1% level of significance, respectively.

The values in parentheses are p-values; the values in parenthesis for Frees test are critical values.

Causality Test Results

Another critical test is the causality test between financial development and unemployment rates. Dumitrescu and Hurlin’s (2012) Granger non-causality test was employed. The Granger non-causality test assumes no causality between two variables, financial development and unemployment in this case. This is tested against the alternative hypothesis that stipulates that financial development does Granger-cause unemployment for at least one-panel vector autoregression (VAR). We computed the p-value using 500 bootstrap replications. The results are presented in Table 4, which does not find any evidence to support that financial development Granger causes unemployment. However, we find evidence of unidirectional causality running from unemployment to financial development. Precisely, total unemployment, female unemployment, youth unemployment, and female youth unemployment Granger cause financial development. There is no evidence of causality between financial development and male unemployment (total and youth).

Table 4.

Dumitrescu and Hurlin (2012) Granger Non-causality Test Results.

Variable	Causality Test Output
Financial development vs total unemployment	0.2242 (0.9660)
Total unemployment vs financial development	4.4513* (0.0680)
Financial development vs female unemployment	−0.0645 (0.9800)
Female unemployment vs financial development	6.0301** (0.0180)
Financial development vs male unemployment	0.0712 (0.9900)
Male unemployment vs financial development	3.3006 (0.1700)
Financial development vs youth unemployment	−0.9180 (0.7840)
Youth unemployment vs financial development	4.1724* (0.0940)
Financial development vs female youth unemployment	−0.2210 (0.9300)
Female youth unemployment vs financial development	4.9521** (0.0380)
Financial development vs male youth unemployment	−0.6332 (0.8680)
Male youth unemployment vs financial development	3.6548 (0.1420)

Notes: H₀: Financial development does not Granger-cause unemployment and vice versa.

H₁: Financial development does Granger-cause unemployment and vice versa.

*p-values computed using 500 bootstrap replications.

*, **, and *** denote 10%, 5%, and 1% level of significance, respectively.

The values in parentheses are p-values.

Main Results

POLS and DOLS Results

The results of POLS, accounting for country- and year-specific effects, are presented in Table 5 while DOLS results are presented in Table 6 including the results of FMOLS and CCR. The results of FMOLS and CCR are included for robustness. Starting from POLS results, it is evident that financial development has a negative and significant effect on all categories of unemployment. Precisely, a percentage increase in the level of financial development has a negative and significant effect on total unemployment, female unemployment, male unemployment, youth unemployment, female youth unemployment, and male youth unemployment by 0.66%, 0.87%, 0.65%, 0.65%, 0.69%, and 0.70%, respectively. This means that an increase in financial development leads to a reduction in unemployment for both the adult working groups (between 15−64 years) and the youths (between 14−25 years), irrespective of their gender. A cursory observation of the estimates reveals that financial development has a greater impact on female unemployment than male unemployment for adults and youths. Similarly, its effect on total youth unemployment is greater than on total adult unemployment. Our findings are in line with some empirical studies that indicate that financial development reduces unemployment (Çiftçioğlu & Bein, 2017; Feldmann, 2012; 2013; Gatti et al., 2012; Pagano & Pica, 2012). Measuring financial development by the level of financial sophistication, Feldmann, 2012; 2013) concluded that it takes a high level of financial sophistication for financial development to reduce unemployment. In the non-OECD countries considered, Pagano and Pica (2012) submitted that only an increase in market capitalization and a decrease in bank concentration reduces unemployment. However, the study by Bayar (2016) on some selected EMCs showed that financial development has no significant effect on unemployment. Contrary to our findings that support the reducing effect of financial development on unemployment, his findings could be adduced to the use of a narrow measure of financial development (credit to the private sector) and the total unemployment rate of those countries. Credit to the private sector only captures the financial depth without considering the multi-dimensional nature of financial development (Svirydzenka, 2016). The centrality of the message of our findings is that holistic financial development could be pro-employment.

Table 5.

POLS Results: The Effect of Financial Development on Unemployment.

	Total Unemployment	Female Unemployment	Male Unemployment	Youth Unemployment	Female Youth Unemployment	Male Youth Unemployment
Financial development	−0.662***	−0.872***	−0.652***	−0.651***	−0.692***	−0.698***
	(0.102)	(0.118)	(0.104)	(0.089)	(0.097)	(0.094)
Growth rate	−0.019***	−0.012**	−0.025***	−0.018***	−0.013***	−0.022***
	(0.005)	(0.006)	(0.005)	(0.004)	(0.005)	(0.005)
Inflation rate	−0.000	−0.000	−0.000	−0.000	−0.000	−0.000
	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)
Secondary school enrolment	−0.150	−0.141	−0.095	0.236**	0.392***	0.138
	(0.122)	(0.141)	(0.125)	(0.107)	(0.117)	(0.113)
Foreign direct investment	0.004	0.004	0.004	0.004	0.005	0.004
	(0.003)	(0.004)	(0.003)	(0.003)	(0.003)	(0.003)
Constant	1.430***	1.086*	1.297**	0.662	−0.244	1.221**
	(0.540)	(0.625)	(0.551)	(0.471)	(0.517)	(0.499)
Observation	529	529	529	529	529	529
R-squared	0.849	0.865	0.816	0.884	0.898	0.854
Country dummies	YES	YES	YES	YES	YES	YES
Year dummies	YES	YES	YES	YES	YES	YES

Notes: The values in parentheses are standard error value of estimate coefficient.

***p < .01, **p < .05, *p < .1.

Table 6.

DOLS, FMOLS and CCR Results: The Effect of Financial Development on Unemployment.

	Total Unemployment	Female Unemployment	Male Unemployment	Youth Unemployment	Female Youth Unemployment	Male Youth Unemployment
DYNAMIC ORDINARY LEAST SQUARES
Financial development	−0.160***	−0.31***	−0.160***	−0.080***	−0.150***	−0.080***
	(−4.170)	(−5.360)	(−4.500)	(−3.330)	(−2.070)	(−4.430)
Growth rate	−0.040***	−0.03	−0.040***	−0.040***	−0.030***	−0.050***
	(−4.730)	(−1.200)	(−5.030)	(−3.150)	(−1.060)	(−5.970)
Inflation rate	−0.050***(−2.630)	−0.050***(−4.310)	−0.050***(−2.870)	−0.060***(−4.380)	−0.0740***(−4.000)	−0.060***(−6.720)
Secondary school enrolment	−0.53(1.320)	−0.380**(2.03)	−0.670(1.250)	−1.060***(3.510)	−1.260(1.250)	−0.900***(5.210)
Foreign direct investment	−0.010**(−2.430)	0.000***(−4.150)	0.000**(−2.080)	−0.010**(−2.360)	−0.010**(−2.170)	0.000***(−2.880)
FULLY MODIFIED ORDINARY LEAST SQUARES
Financial development	−0.470***(−7.260)	−0.460***(−5.340)	−0.430***(−7.330)	−0.130***(−5.890)	−0.090***(−3.750)	−0.160***(−6.190)
Growth rate	0.000***(−4.910)	0.010(0.110)	−0.010***(−8.090)	−0.010***(−9.070)	0.000**(−2.510)	−0.020***(−11.450)
Inflation rate	−0.030***(−11.050)	−0.020***(−7.800)	−0.030***(−13.050)	−0.030***(−14.470)	−0.020***(−12.210)	−0.030***(−15.850)
Secondary school enrolment	0.490***(3.160)	0.390**(2.650)	0.480***(3.180)	0.290***(10.190)	0.270***(10.050)	0.350***(9.960)
Foreign direct investment	0.010***(4.060)	−0.000(1.430)	0.020***(5.420)	0.010***(2.810)	−0.000(0.000)	0.020***(4.31)
CANONICAL COINTEGRATION REGRESSION
Financial development	−0.420***(−4.770)	−0.390**(−2.310)	−0.440***(−4.970)	−0.210**(2.600)	−0.170***(7.560)	−0.330***(2.770)
Growth rate	−0.010***(−14.550)	−0.000***(−7.290)	−0.020***(−16.150)	−0.010***(−18.080)	−0.010***(−10.180)	−0.020***(−18.750)
Inflation rate	−0.020***(−23.240)	−0.020***(−22.220)	−0.030***(−20.980)	−0.020***(−23.590)	−0.020***(−22.130)	−0.020***(−23.800)
Secondary school enrolment	−0.240***(4.700)	−0.150***(3.180)	−0.180***(5.590)	0.200***(22.250)	0.260***(24.550)	0.360***(23.250)
Foreign direct investment	−0.010***(−0.330)	−0.010***(−0.100)	0.000***(−0.510)	−0.010(−1.340)	−0.010*(−1.760)	−0.000*(−1.710)

Notes: The values in parentheses are t-test values showing the level of significance of the variables in the models.

***, **, and * denote 1%, 5%, and 10% level of significance, respectively.

The results of the control variables are also reported in Table 5. The effects of real GDP and inflation rate follow a priori expectation regarding signs of the estimated coefficients. However, only the estimates of the real GDP growth rate are statistically significant. Thus, a 1% rise in the real GDP growth rate leads to a reduction in total unemployment, female unemployment, male unemployment, youth unemployment, female youth unemployment, and male youth unemployment by 0.02%, 0.01%, 0.03%, 0.02%, 0.01%, and 0.02%, respectively. Secondary school enrolment produces a mixed effect on unemployment. While it has an insignificant negative effect on total female and male unemployment of the working-age population (adults), it has a positive and significant effect on total female and male youth unemployment. This result is not surprising, especially in the youth, as they constitute the highest number of the population enrolled in secondary school. Besides, in most countries, including emerging countries, youths constitute the greater part of the population. Aside from this, youth unemployment is usually high in most emerging countries. Hence, the more this set of people is enrolled in school in anticipation of prospects in the face of growing limited opportunities in the labor market, the more they are likely to end up in the pool of unemployment, thus, making the unemployment situation worse. In the case of FDI, our results at this point do not show any significant effect of FDI on unemployment, even though the effect is positive.

Table 6 presents the results of the DOLS estimation method, including the results of FMOLS and CCR for robustness and validation of DOLS results. It can be deduced that all variables are correctly signed; that is, they follow a priori expectations. For instance, an increase in financial development leads to a reduction in all categories of unemployment—youths and adults. This means that the results of DOLS are not significantly different from POLS, thereby reinforcing each other. Also, the results of FMOLS and CCR reinforce the results of DOLS and POLS. Thus, irrespectively of estimation methods, financial development reduces unemployment in EMCs. This could be attributed to several factors. One is the level of integration of the financial sector into the economies of EMCs compared with the developing countries where the financial sector is fragmented from the real economy or is yet to be developed to the extent of influencing the real sectors of the economy. It can also be attributed to the development of new financial instruments in the emerging financial sector that smoothly facilitate investment and by extension raise workers’ productivity and employment of new workforces. Furthermore, efficient allocation of financial resources across different sectors of the economy may be a result of the quality of regulations by different regulatory authorities associated with the financial sector.

The impact of the real GDP growth rate on unemployment is similar to the POLS results because the real GDP growth rate negatively influences unemployment. This suggests that the economic growth recorded in EMCs lead to a significant reduction in unemployment across different genders of diverse ages. Unlike the results obtained from the POLS, the effect of inflation on unemployment follows a priori expectation because it negatively influences unemployment. Moreover, in the DOLS model, secondary school enrolment now is negatively and significantly related to total unemployment (working-age population), total youth unemployment, and female youth unemployment. Quantitatively, an increase in enrolment in secondary school would lead to reductions in total unemployment, total youth unemployment, and female unemployment by 0.38%, 1.06%, and 0.90%, respectively. This finding appears to be contrary to the results from POLS. It implies that different estimation techniques may yield different outcomes since each estimation is based on different assumptions.

Quantile Regression via Moments Results

POLS and DOLS, including FMOLS and CCR, provide estimates that show the impacts of financial development on unemployment without considering how financial development affects the distribution of unemployment at different quantiles. Such results, as stated by Binder and Coad (2011), do not provide complete information about the influence of financial development on unemployment, which makes it impossible to elicit cogent policy information for policymakers. Hence, we used the quantile regression method developed by Machado and Silver (2019) to capture the effect of financial development on the distribution of unemployment at different quantiles. The method allows us to address the endogeneity problem characterizing the relationship between financial development and unemployment and it also enables us to deal with the heterogeneity problem in the panel of countries. The lags of independent variables are used as instruments. The results are presented in Table 7, which shows that the effects of financial development vary on the distribution of unemployment based on their categorization as is obvious for the working population (15–64 years), especially for total and female unemployment. Financial development has a significant negative effect on unemployment in all quantiles. Precisely, an increase in the level of financial development would lead to a reduction in total unemployment by 1.04%, 0.95%, 0.86%, 0.76%, and 0.72% at 0.10th, 0.25th, 0.50th, 0.75th and 0.90th quantiles, respectively. Similarly, an increase in financial development would result in a reduction in female unemployment by 1.238%, 1.234%, 1.227%, 1.222% and 1.219% at 0.10th, 0.25th, 0.50th, 0.75th, and 0.90th quantiles, respectively. However, a cursory look at the estimated coefficients shows that the effect of financial development on total and female unemployment declines as the quantiles increase. In the case of the effect of financial development on male unemployment, financial development enters with negative effects at a lower quantile (0.10th quantile). The negative effect continues until the middle quantile. After 0.50th quantile, financial development still has a negative effect on male unemployment but is statistically insignificant. Generally, our results suggest that the financial sector has the potential to reduce unemployment in EMCs, which has important policy implications. Thus, it is essential to design long-term policies that would create enabling environment for the financial sector to grow and serve as a harbinger of economic growth and employment.

Table 7.

Quantile Regression Results: The Effect of Financial Development on Unemployment.

Quantile Regression Via Moment (IVQR)		Quantile
Dependent Variable	Independent Variables	Location	Scale	0.1	0.25	0.50	0.75	0.9
Total unemployment	Financial development	−0. 845***(0.003)	0.113(0.396)	−1.038 ***(0.004)	−0.948***(0.002)	−0.859 ***(0.001)	−0.763 ***(0.004)	−0.718 ***(0.008)
	GDP growth rate	−0.040(0.199)	−0.009(0.863)	−0.024(0.732)	−0.031(0.453)	−0.038(0.143)	−0.046(0.285)	−0.050(0.380)
	Inflation rate	−0. 0002(0.897)	−0.0002(0.863)	−0.000(0.980)	−0.0001(0.986)	−0.0002(0.919)	−0.0004(0.697)	−0.0004(0.326)
	Secondary school Enrolment	1.553**(0.011)	−0.208(0.273)	1.907***(0.000)	1.743(0.000)	1.578***(0.000)	1.402***(0.003)	1.320***(0.009)
	Foreign direct investment	0.010(0.946)	−0.013(0.580)	0.032(0.843)	0.021(0.881)	0.011(0.929)	0.0004(0.997)	−0.004(0.964)
	Constant			−8.370**(0.000)	−7.107***(0.000)	−5.847(0.001)	−4.497**(0.043)	−3.867(0.112)
Female unemployment	Financial development	−1.227(0.131)	0.007(0.993)	−1.238***(0.000)	−1.234***(0.000)	−1.227***(0.000)	−1.222***(0.000)	−1.219***(0.006)
	GDP growth rate	−0.010(0.980)	0.020(0.941)	−0.041(0.548)	−0.029(0.562)	−0.010(0.825)	0.005(0.941)	0.015(0.855)
	Inflation rate	−0.001(0.931)	−0.0002(0.867)	−0.0002(0.958)	−0.0003(0.907)	−0.0005(0.742)	−0.001***(0.445)	−0.001(0.133)
	Secondary school Enrolment	2.230(0.708)	0.002(1.000)	2.227***(0.000)	2.228***(0.000)	2.230***(0.001)	2.231**(0.021)	2.232*(0.059)
	Foreign direct investment	−0.060(0.951)	−0.027(0.747)	−0.016(0.547)	−0.033(0.320)	−0.060(0.181)	−0.080(0.139)	−0.094(0.122)
	Constant			−9.796 ***(0.000)	−9.415***(0.000)	−8.792***(0.002)	−8.330**(0.047)	−8.010(0.117)
Male unemployment	Financial development	−0.658***(0.000)	0.167(0.315)	−0.964***(0.002)	−0.793**(0.011)	−0.671**(0.038)	−0.527(0.130)	−0.423(0.257)
	GDP growth Rate	−0.039(0.362)	−0011(0.531)	−0.018(0.816)	−0.030(0.406)	−0.038(0.554)	−0.048(0.687)	−0.055(0.733)
	Inflation rate	−0.0002(0.812)	−0.0002(0.786)	−0.000(0.977)	−0.000(0.980)	−0.000(0.920)	−0.000(0.746)	−0.000(0.396)
	Secondary school Enrolment	1.370***(0.001)	−0.301***(0.002)	1.923***(0.000)	1.613**(0.026)	1.394(0.301)	1.135(0.587)	0.947(0.719)
	Foreign direct investment	0.005(0.955)	−0.010(0.601)	0.024(0.908)	0.013(0.939)	0.005(0.971)	−0.004(0.976)	−0.010(0.920)
	Constant			−8.447***(0.000)	−0.644**(0.031)	−4.955(0.399)	−3.245(0.729)	−2.008(0.866)
Youth unemployment	Financial development	−0.754*(0.086)	0.055(0.690)	−0.841***(0.000)	−0.792***(0.000)	−0.745***(0.009)	−0.703(0.109)	−0.676(00.208)
	GDP growth rate	−0.014(0.753)	−0.021(0.177)	0.020(0.672)	0.001(0.973)	−0.017(0.678)	−0.033(0.551)	−0.043(0.520
	Inflation rate	−0.001(0.825)	0.000(0.984)	−0.001(0.908) ***	−0.001(0.871)	−0.001(0.772)	−0.000(0.362)	−0.000(0.604)
	Secondary school Enrolment	2.009**(0.036)	−0.556***(0.000)	2.891(0.000)	2.395***(0.000)	1.926***(0.005)	1.498(0.191)	1.233(0.389)
	Foreign direct investment	−0.008(0.672)	−0.008***(0.000)	0.005(0.773)	−0.002(0.905)	−0.009(0.644)	−0.016(0.462)	−0.020(0.376)
	Constant			−11.815***(0.000)	−9.031***(0.000)	−6.402**(0.045)	−3.997(0.451	−2.514(0.705)
Female youth unemployment	Financial development	−0.925***(0.008)	−0.252**(0.012)	−0.535**(0.015)	−0.710***(0.000)	−0.940***(0.000)	−1.157***(0.000)	−1.305***(0.000)
	GDP growth rate	−0.020(0.534)	0.005(0.648)	−0.028(0.370)	−0.024(0.334)	−0.020(0.339)	−0.016(0.477)	−0.012(0.623)
	Inflation rate	−0.000***(0.004)	−0.000***(0.000)	0.000(0.349)	−0.000(0.745)	−0.000(0.001)	−0.001***(0.000)	−0.001***(0.000)
	Secondary school Enrolment	2.222***(0.003)	−0.045***(0.038)	2.291***(0.000)	0.260***(0.000)	2.219***(0.000)	2.181***(0.000)	2.154***(0.000)
	Foreign direct investment	−0.008(0.887)	−0.005(0.854)	0.001(0.994)	−0.003(0.953)	−0.008(0.823)	−0.012(0.507)	−0.015(0.055)
	Constant			−8.640***(0.000)	−8.274***(0.000)	−7.790***(0.000)	−7.337***(0.000)	−7.025***(0.000)
Male youth unemployment	Financial development	−0.373***(0.012)	−0.002(0.984)	−0.369***(0.046)	−0.371***(0.006)	−0.373(0.250)	−0.375(0.466)	−0.376(0.573)
	GDP growth rate	−0.025(0.348)	−0.018(0.298)	0.005(0.940)	−0.010(0.813)	−0.026(0.463)	−0.039(0.457)	−0.050(0.491)
	Inflation rate	−0.000(0.884)	−0.000(0.933)	−0.000(0.988)	−0.000(0.957)	−0.000(0.876)	−0.000(0.638)	−0.001(0.195)
	Secondary school Enrolment	1.375***(0.000)	−0.442***(0.000)	2.090***(0.000)	1.745***(0.000)	1.360(0.830)	1.036(0.497)	0.779(0.692)
	Foreign direct investment	−0.005(0.741)	−0.008***(0.000)	0.008(0.687)	0.002(0.939)	−0.005(0.830)	−0.011(0.681)	−0.016(0.587)
	Constant			−7.807***(0.000)	−5.838***(0.000)	−3.640(0.400)	−1.791(0.795)	−0.324(0.971)

Notes: *, ** and *** denote 10%, 5% and 1% level of significance, respectively.

The values in parentheses are p-values.

For youth unemployment (14–25 years), the results are slightly different but similar, to some extent. For female youth unemployment, financial development has a negative effect on unemployment in all quantiles, that is, financial development reduces female youth unemployment in all quantiles. Thus, an increase in financial development would reduce female youth unemployment by 0.54%, 0.71%, 0.94%, 1.16%, and 1.31% at 0.10th, 0.25th, 0.50th, 0.75th, and 0.90th quantiles, respectively. In the case of total youth unemployment, financial development reduces unemployment up to 0.50th quantiles. Subsequently, the effect of financial development on total youth unemployment remains negative but statistically insignificant. Similarly, the reducing effect of financial development on the distribution of male youth unemployment stopped at 0.25th quantile. Even though the influence of financial development remains negative on the distribution of male youth unemployment, it is not statistically significant. Specifically, an increase in financial development reduces the male youth unemployment rate by 0.369% and 0.371% at 0.10th and 0.25th quantiles, respectively.

Thus, our investigation reveals that financial development does not only have a conditional mean reducing effect on unemployment (adults and youths across different genders) but also has a progressive or incremental conditional distribution effect on unemployment. Hence, financial sector development can help solve unemployment by channeling credits to the productive sectors of the economy that would spur growth and engender employment.

Conclusion and Policy Recommendations

We conducted an in-depth analysis of the effect of financial development on unemployment in 19 EMCs for adult (15–64 years) and youth (14–25 years) unemployment across genders—female and male. The availability of data, especially unemployment and financial development, largely determined the countries and years included in this study. Financial development was proxied by a broad-based financial development index computed by IMF from the series of measures of financial development, including financial institutions and markets. The unemployment categories considered were total adult and youth unemployment rate, female adult and youth unemployment rate, and male adult and youth unemployment rate. We also controlled for regressors such as real GDP growth rate, inflation rate, secondary school enrollment and foreign direct investment. The four variables were included because they have established theoretical linkages with unemployment. A series of preliminary analytical tests were conducted, which included descriptive statistics, correlation, unit root test, CDT, and Granger non-causality test. For the main analysis, we employed three estimation techniques: POLS, DOLS, and Machado and Silva’s (2019) panel quantile regression via moments. However, to verify the veracity of the results of DOLS, we also used FMOLS and CCR estimation methods.

We found some interesting results. First, there is evidence of CD among the panel of the countries, signifying that the emerging countries have strong integration and, hence, they can be affected by unobserved common factors such as oil price shocks, financial shocks, and health shocks. Second, unidirectional causality exists between financial development and unemployment, especially total adult and youth unemployment, and female adult and youth unemployment. The direction of causality runs from unemployment to financial development. Third, the results from POLS show that financial development has a negative and significant effect on all categories of unemployment. This suggests that financial development reduces all categories of unemployment: adult and youth, irrespective of gender. The DOLS results were robust to the POLS, FMOLS, and CCR results. Fourth, based on quantile regression results, it is evident that financial development, generally, has a negative effect on unemployment for the working-age population (adults). However, the negative effect varies across the distribution of the categories of unemployment. For total and female unemployment of the working-age population, financial development has a negative effect on the distribution of unemployment in all quantiles. However, financial development reduces male unemployment for the working-age population up to 0.50th quantile. In the case of youth unemployment, financial development has a reducing effect on total and male youth unemployment up to the 0.50th and 0.25th quantiles, respectively.

The policy implications of these findings are straightforward. First, financial development does not only have a conditional mean effect on unemployment but also affects unemployment over time. Thus, a broad-based financial sector development policy framework must be developed in EMCs, given its potential for sustainable growth and employment creation. In addition, the policy for financial sector development must not only be a short-term phenomenon usually adopted during the period of the economic or financial crisis but also be a consistent long-term policy. Second, given the high youth unemployment compared to adult unemployment in EMCs, policymakers need to implement and facilitate youth-friendly programs that would generate employment opportunities for youths and empower them to be employers of labor. In the same vein, the gender gap in employment in EMCs should be bridged to provide equal opportunities for men and women to contribute and benefit from the labor market. This calls for modifying the extant labor laws that discriminate between male and female labor to ensure gender equity in employment.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflict of interest concerning this research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and /or publication of this article.

Notes

ORCID iD

Isiaka Akande Raifu

References

Acemoglu

, Johnson

, & Robinson

J. A.

(2005). Institutions as a fundamental cause of long-run growth. In Aghion

& Durlauf

(Eds), Handbook of economic growth (pp. 385–472). Elsevier. https://doi.org/10.1016/S1574-0684(05)01006-3

Adejumo

O. O.

, Asongu

S. A.

, & Adejumo

A. V.

(2021). Education enrolment rate vs employment rate: Implications for sustainable human capital development in Nigeria. International Journal of Educational Development , 83(4), 1–13. https://doi.org/10.1016/j.ijedudev.2021.102385

Adnan

(2014). Financial development and growth the case of South Korea. International Proceedings of Economics Development and Research , 76(1), 1–4.

Afolabi

J. A.

(2022). Financial development, trade openness and economic growth in Nigeria. Iranian Economic Review , 26(1), 237–254.

Ajide

F. M.

(2020). Asymmetric influence of financial development on unemployment in Nigeria. Ilorin Journal of Economic Policy , 7(2), 39–52.

Alalawneh

M. M.

, & Nessa

(2020). The impact of foreign direct investment on unemployment: Panel data approach. Emerging Science Journal , 4(4), 228–242.

Aliero

H. M.

, Ibrahim

S. S.

, & Shuaibu

(2013). An empirical investigation into the relationship between financial sector development and unemployment in Nigeria. Asian Economic and Financial Review , 3(10), 1361–1370.

Anwar

, Siddique

, Dogan

, & Sharif

(2021). The moderating role of renewable and non-renewable energy in environment-income nexus for ASEAN countries: Evidence from method of moments quantile regression. Renewable Energy , 164(C), 956–967.

Arango

L. E.

, & Flórez

L. A.

(2020). Determinants of structural unemployment in Colombia: A search approach. Empirical Economics , 58(5), 2431–2464.

10.

Arestis

, Chortareas

, & Magkonis

(2015). The financial development and growth nexus: A meta-analysis. Journal of Economic Surveys , 29(3), 549–565.

11.

Bagehot

(1873). Lombard street: A description of the money market . Scribner.

12.

Bayar

(2016). Financial development and unemployment in emerging market economies. Scientific Annals of Economics and Business , 63(2), 237–245.

13.

Bijlsma

, Kool

, & Non

(2018). The effect of financial development on economic growth: A meta-analysis. Applied Economics , 50(57), 6128–6148.

14.

Binder

, & Coad

(2011). From average Joe’s happiness to miserable Jane and cheerful John: Using quantile regressions to analyze the full subjective well-being distribution. Journal of Economic Behavior & Organization , 79(3), 275–290.

15.

Borsi

M. T.

(2018). Credit contractions and unemployment. International Review of Economics & Finance , 58(C), 573–593. https://doi.org/10.1016/j.iref.2018.06.004

16.

Cepparulo

, Cuestas

J. C.

, & Intartaglia

(2017). Financial development, institutions, and poverty alleviation: An empirical analysis. Applied Economics , 49(36), 3611–3622.

17.

Çiftçioglu

, & Bein

M. A.

(2017). The relationship between financial development and unemployment in selected countries of the European Union. European Review , 25(2), 307.

18.

De Haan

, & Sturm

J. E.

(2017). Finance and income inequality: A review and new evidence. European Journal of Political Economy , 50(C), 171–195. https://doi.org/10.1016/j.ejpoleco.2017.04.007

19.

De Hoyos

R. E.

, & Sarafidis

(2006). Testing for cross-sectional dependence in panel-data models. The Stata Journal , 6(4), 482–496.

20.

Dromel

N. L.

, Kolakez

, & Lehmann

(2010). Credit constraints and the persistence of unemployment. Labour Economics , 17(5), 823–834.

21.

Dumitrescu

E. I.

, & Hurlin

(2012). Testing for Granger non-causality in heterogeneous panels. Economic Modelling , 29(4), 1450–1460.

22.

Feldmann

(2012). Financial system sophistication and unemployment in industrial countries. International Journal of Finance & Economics , 18(4), 319–338.

23.

Feldmann

(2013). Financial system sophistication and unemployment around the world. Applied Economic Letter , 20(16), 1491–1496.

24.

Folawewo

A. O.

, & Adeboje

O. M.

(2017). Macroeconomic determinants of unemployment: Empirical evidence from economic community of West African states. African Development Review , 29(2), 197–210.

25.

Frees

E. W.

(1995). Assessing cross-sectional correlation in panel data. Journal of Econometrics , 69(2), 393–414.

26.

Friedman

(1937). The use of ranks to avoid the assumption of normality implicit in the analysis of variance. Journal of the American Statistical Association , 32(200), 675–701.

27.

Gatti

, Rault

, & Vaubourg

A. G.

(2012). Unemployment and finance: How do financial and labour market factors interact? Oxford Economic Papers , 64(3), 464–489.

28.

Guo

, You

, & Lee

C. C.

(2022). CO2 emissions, income inequality, and country risk: Some international evidence. Environmental Science and Pollution Research International , 29(9), 12756–12776.

29.

Hendy

(2015). Women’s participation in the Egyptian labour market: 1998–2012. In

Assaad

& Krafft

(Eds), The Egyptian labour market in an era of revolution . Oxford Scholarship Online. https://doi.org/10.1093/acprof:oso/9780198737254.001.0001

30.

Horvath

, & Zhong

(2019). Unemployment dynamics in emerging countries: Monetary policy and external shocks. Economic Modelling , 76(C), 31–49. https://doi.org/10.1016/j.econmod.2018.07.017

31.

Ibrahiem

D. M.

, & Sameh

(2020). How do clean energy sources and financial development affect unemployment? Empirical evidence from Egypt. Environmental Science and Pollution Research , 27(18), 22770–22779.

32.

Iheonu

C. O.

, Agbutun

S. A.

, Omenihu

C. M.

, Ihedimma

G. I.

, & Osuagwu

V. N.

(2019). The impact of governance quality on mortality rates in sub-Saharan Africa. African Population Studies , 33(1), 1–15.

33.

Iheonu

C. O.

, Asongu

S. A.

, Odo

K. O.

, & Ojiem

P. K.

(2020). Financial sector development and investment in selected countries of the Economic Community of West African States: Empirical evidence using heterogeneous panel data method. Financial Innovation , 6(1), 1–15.

34.

K. S.

, Pesaran

M. H.

, & Shin

(2003). Testing for unit roots in heterogeneous panels. Journal of Econometrics , 115(1), 53–74.

35.

International Monetary Fund. (2020). Financial development index database . https://data.imf.org/?sk=F8032E80-B36C-43B1-AC26-493C5B1CD33B

36.

Isah

, & Soliu

(2016). Financial development and income inequality: A review of literature [Paper presentation]. International Conference on Social Sciences and Law, Nile University of Nigeria, Abuja, 11–12 May 2016.

37.

Islam

M. A.

, Khan

M. A.

, Popp

, Sroka

, & Oláh

(2020). Financial development and foreign direct investment: The moderating role of quality institutions. Sustainability , 12(9), 3556.

38.

Jiang

, & Ma

(2019). The impact of financial development on carbon emissions: A global perspective. Sustainability , 11(19), 5241.

39.

Kao

, Chiang

, & Chen

(1999). International R&D spillovers: An application of estimation and inference in panel cointegration. Oxford Bulletin of Economics and Statistics , 61(1), 691–709.

40.

Kim

D. H.

, Chen

T. C.

, & Lin

S. C.

(2019). Finance and unemployment: New panel evidence. Journal of Economic Policy Reform , 22(4), 307–324.

41.

Koenker

, & Bassett

Jr (1978). Regression quantiles. Econometrica , 46(1), 33–50.

42.

Levine

(2005). Finance and growth: Theory and evidence. In

Aghion

& Durlauf

(Eds), Handbook of economic growth (Vol. 1, pp. 865–934). Elsevier. https://doi.org/10.1016/S1574-0684(05)01012-9

43.

Machado

J. A.

, & Silva

J. S.

(2019). Quantiles via moments. Journal of Econometrics , 213(1), 145–173.

44.

McKinnon

R. I.

(1973). Money and capital in economic development . Brookings Institution.

45.

Morón

, Salgado

, & Seminario

(2013). Regional financial development and firm growth in Peru (IDB Working Paper Series No. IDB-WP-398). Inter-American Development Bank.

46.

Muyambiri

, & Odhiambo

N. M.

(2018). The impact of financial development on investment: A review of international literature. Organizations and Markets in Emerging Economies , 9(2), 167–192.

47.

Muye

I. M.

, & Muye

I. Y.

(2017). Testing for causality among globalization, institution and financial development: Further evidence from three economic blocs. Borsa Istanbul Review , 17(2), 117–132.

48.

Neal

(2014). Panel cointegration analysis with xtpedroni. The Stata Journal , 14(3), 684–692.

49.

, & Liu

(2019). Financial liberalization and income inequality: A meta-analysis based on cross-country studies. China Economic Review , 56(C), 1–17. https://doi/org/10.1016/j.chieco.2019.101306

50.

Ogbeide

F. I.

, Kanwanye

, & Kadiri

(2016). Revisiting the determinants of unemployment in Nigeria: Do resource dependence and financial development matter? African Development Review , 28(4), 430–443.

51.

Ogunjimi

J. A.

(2019). Impact of public debt on investment: Evidence from Nigeria. DBN Journal of Economics and Sustainable Growth , 2(2), 1–28.

52.

Ogunjimi

J. A.

(2021). The role of small and medium scale enterprises in tackling unemployment in Nigeria. DBN Journal of Economics and Sustainable Growth , 4(1), 74–92.

53.

Oji

C. E.

, & Afolabi

J. A.

(2022). Economic and peace effects of terrorism in the 21st century. International Journal of Public Law and Policy , 8(2), 144–157. https://doi.org/10.1504/IJPLAP.2022.10044160

54.

Opoku

E. E. O.

, Ibrahim

, & Sare

Y. A.

(2019). The causal relationship between financial development and economic growth in Africa. International Review of Applied Economics , 33(6), 789–812.

55.

Pagano

, & Pica

(2012). Finance and employment. Economic Policy , 27(69), 5–55.

56.

Pesaran

M. H.

(2004). General diagnostic tests for cross-section dependence in panels (Cambridge Working Papers in Economics No. 0435). University of Cambridge. https://www.econ.cam.ac.uk/research-files/repec/cam/pdf/cwpe0435.pdf

57.

Phillips

, & Sul

(2003). Dynamic panel estimation and homogeneity testing under cross- section dependence. The Econometrics Journal , 6(1), 217–259.

58.

Raifu

I. A.

(2017). On the determinants of unemployment in Nigeria: What are the roles of trade openness and current account balance? Review of Innovation and Competitiveness: A Journal of Economic and Social Research , 3(4), 5–30.

59.

Raifu

I. A.

(2019). The effect of financial development on unemployment in Nigeria: Do measures of financial development matter? DBN Journal of Economics and Sustainable Growth , 2(2), 1–35.

60.

Raifu

I. A.

(2022). COVID-19 pandemic, government policy responses and stock market performance in some selected African countries. DBN Journal of Economics and Sustainable Growth , 4(3), 24–65.

61.

Raifu

I. A.

, & Abodunde

T. T.

(2020). Does unemployment still follow hysteresis hypothesis in Nigeria? Evidence from linear and nonlinear unit root test methods. Journal of International Cooperation and Development , 3(1), 79–79.

62.

Raifu

I. A.

, & Adeboje

O. M.

(2022). Labour force participation and unemployment rate: Does discouraged worker effect hypothesis or unemployment invariance hypothesis hold in Africa? African Journal of Economic and Management Studies . 13(2), 284–305. https://doi.org/10.1108/AJEMS-07-2021-0317

63.

Raifu

I. A.

, Aminu

, & Folawewo

A. O.

(2020). Investigating the relationship between changes in oil prices and unemployment rate in Nigeria: Linear and nonlinear autoregressive distributed lag approaches. Future Business Journal , 6(1), 1–18.

64.

Raifu

I. A.

, Kumeka

T. T.

, & Aminu

(2021). Reaction of stock market returns to COVID-19 pandemic and lockdown policy: Evidence from Nigerian firms’ stock returns. Future Business Journal , 7(1), 1–16.

65.

Schumpeter

J. A.

(1911). The theory of economic development . Harvard University Press.

66.

Shabbir

, Anwar

, Hussain

, & Imran

(2012). Contribution of financial sector development in reducing unemployment in Pakistan. International Journal of Economics and Finance , 4(1), 260–268.

67.

Shahbaz

, Van Hoang

T. H.

, Mahalik

M. K.

, & Roubaud

(2017). Energy consumption, financial development and economic growth in India: New evidence from a nonlinear and asymmetric analysis. Energy Economics , 63(C), 199–212. https://doi.org/10.1016/j.eneco.2017.01.023

68.

Shaw

E. S.

(1973). Financial deepening in economic development . Oxford University Press.

69.

Svirydzenka

(2016). Introducing a new broad-based index of financial development (IMF Working Paper No. 2016/005). International Monetary Fund. https://doi.org/10.5089/9781513583709.001

70.

United Nations Department of Economic and Social Affairs. (2020). World economic situation and prospects: June 2020 (Briefing No. 138). https://www.un.org/development/desa/dpad/publication/world-economic-situation-and-prospects-june-2020-briefing-no-138/

71.

Valickova

, Havranek

, & Horvath

(2015). Financial development and economic growth: A meta-analysis. Journal of Economic Surveys , 29(3), 506–526.

72.

Wasmer

, & Weil

(2004). The macroeconomics of labour and credit market imperfections. American Economic Review , 94(4), 944–963.

73.

World Bank. (2020). World development Indicators. 2020 version . https://databank.worldbank.org/source/world-development-indicators

74.

Younsi

, & Bechtini

(2018). Economic growth, financial development and income inequality in BRICS countries: Evidence from panel granger causality tests . University Library of Munich.