Effects of Health Factors on GDP Growth: Empirical Evidence from Saudi Arabia

Abstract

The primary objectives of this study are to examine the presence of long-run equilibrium relationships and short-run dynamic relationships between health factors and GDP growth in Saudi Arabia over a specific time period. By utilizing an annual time series dataset from 1990 to 2019 obtained from the World Bank, the study focuses on four key health factors and employs the Johansen cointegration test and vector error correction model to estimate the relationship between these factors and GDP growth. The results indicate the existence of a long-run equilibrium relationship between the health factors and GDP growth; however, in the short term, the variables are found to be in a state of disequilibrium. Specifically, the study reveals that infant mortality, road traffic mortality, and healthcare expenditure exhibit a strong negative relationship with GDP growth, while the maternal mortality ratio displays a weak positive relationship. The findings of this research hold significant implications for policymakers who are striving to achieve sustainable GDP growth as outlined in Saudi Vision 2030. These findings suggest that policymakers can simultaneously promote higher GDP growth and reduce infant mortality, road traffic mortality, and healthcare expenditure. Although the maternal mortality ratio exhibits a relatively weak positive relationship with GDP, it is still crucial for policymakers to address this issue to enhance sustainable GDP growth, aligning with the objectives set forth in Saudi Vision 2030. Overall, this study bridges a research gap and provides valuable insights that can inform health and economic policies in Saudi Arabia.

1. Introduction

The primary objective of macroeconomics is to increase and maintain the gross domestic product (GDP) of a country. Therefore, economists are continually exploring microeconomic and macroeconomic factors that affect and sustain economic growth over time. Despite numerous empirical and theoretical studies conducted to investigate the various links between GDP growth and health factors, the final determinants contributing to increasing or hindering GDP growth have not yet been identified. Previous research findings make it challenging to determine the weights of the variables in economic relations. For instance, Acemoglu and Johnson [1] found strong negative effects of health on GDP growth, while Lorentzen et al. [2] found strong positive effects despite using similar specifications. Furthermore, while mortality rates (such as infant mortality rates [IMR], road traffic mortality [RTM], and maternal mortality ratio [MMR]) have been decreasing and life expectancy increasing globally in recent decades, it remains unclear theoretically and empirically to what extent health progress has causally impacted positive GDP growth. Lucas [3] and Bloom and Prettner [4] suggested that better health can improve GDP per capita, lower mortality, and lead to longer lives, leading to investment in physical and human capital and sustainable GDP growth. However, Acemoglu and Johnson [1] proposed that mortality reduction drives population growth and can reduce GDP per capita. Thus, it is crucial to assess the impact of changes in mortality rates (IMR, MMR, and RTM) and healthcare expenditures (HCE) on GDP growth to help policymakers formulate policies that foster or prevent these effects. Furthermore, Rocco et al. [5] suggested that the direction and levels of the effects of health on GDP growth vary across countries with their income levels, and the effect varies nonlinearly depending on the stage of demographic change at the beginning of the study period. Therefore, further studies are necessary to produce more specific and robust empirical evidence in a country’s context.

Saudi Arabia is one of the largest economies in the Middle East and North Africa region, with a GDP of $833.54 billion in 2021, up from $164.54 billion in 1980 [6]. This country is very well known for its oil and gas reserves [7]. The economy of Saudi Arabia is heavily reliant on oil, which accounts for around 95% of total export earnings and contributes approximately 40% to the country’s GDP. As a result, fluctuations in the price of oil have a significant impact on economic growth, and this country has faced irregular economic growth for the last few decades, like the overall world economy. In recent years, Saudi Arabia has implemented several initiatives under the banner of “Saudi Vision 2030” to achieve sustainable and robust GDP growth [8]. The main goals of Vision 2030 are: increasing the contribution of non-oil products and services (of GDP) from 16% to 50%; enhancing the country’s global ranking in the Logistics Performance Index from 49th to 25th position; and increasing the private sector’s contribution from 40% to 65% of total GDP [9]. The country is increasingly diversifying its economy by boosting the contribution of non-oil products and services. As a result, the private sector’s contribution to the total GDP has increased from 40% in 2014 to 51.7% in 2021 [6]. Saudi Arabia is also making notable progress in social and environmental development, including gender equality, improved education, higher living standards, enhanced well-being, and environmental laws.

The Government of Saudi Arabia has been working hard to achieve higher economic growth, resulting in a 7.3% increase in real GDP in 2021 compared with the previous year [9]. However, sustainable GDP growth remains a challenge due to a lack of understanding of the determinants of GDP growth. To achieve the objectives of the ‘Saudi Vision 2030’ effectively, it is important to investigate both the positive (prospects) and negative (challenges) determinants of GDP growth. Sustainable GDP growth is an important global indicator, and in recent years, it has become apparent that health-related factors play a vital role in achieving higher and more sustainable GDP growth in Saudi Arabia. However, the existing literature has provided conflicting evidence regarding the relationships between health factors and GDP growth. The lack of a comprehensive and consistent theory on the relationship between health factors and GDP growth has made it difficult for policymakers, the government, and other relevant authorities to address this issue effectively. This study aims to examine the direct and indirect links between health factors (infant mortality rate, maternal mortality ratio, road traffic mortality rate, and healthcare expenditure) and sustainable GDP growth in Saudi Arabia, which has been identified as one of the country’s most vulnerable to changes in health factors [10]. The specific objectives of the study are (i) to determine whether GDP growth and health factors have a long-run equilibrium relationship; (ii) to test for any causal relationship between the variables; and (iii) to provide recommendations to policymakers on formulating effective policies that consider health factors to achieve sustainable GDP growth in Saudi Arabia. It is hoped that the findings of this study will contribute to a better understanding of the relationship between health factors and GDP growth in Saudi Arabia and provide valuable guidance for policymakers.

This study aims to address the broader research problem of understanding the relationship between health factors and GDP growth in Saudi Arabia and how this relationship can be utilized to improve sustainable GDP growth. To achieve this, three research questions have been formulated:

• How are health indicators (such as decreasing IMR, MMR, and RTM and increasing HCE) related to GDP growth in Saudi Arabia in the short and long run, and what is the nature of this impact (positive or negative)?
• Is there a causal relationship between Saudi Arabia’s GDP growth and health factors over time?
• What are the effective policies that should be implemented to address health-related issues and achieve sustainable GDP growth in Saudi Arabia?

The structure of the paper is as follows: Section 2 offers a comprehensive literature review. Section 3 briefly outlines the data sources, model specifications, methodological tools utilized in this study, detailing the analytical procedures and highlighting the strengths of the vector error correction (VEC) model and the Granger causality test. The results of the econometric tests are presented in Section 4. Section 5 provides a discussion of the study findings and their potential explanations. The practical implications of the study are explored in Section 6. The study concludes in Section 8, summarizing the key findings. Finally, Section 8 concludes the paper by addressing its limitations and suggesting directions for future research.

2. Literature Review and Hypotheses Development

2.1. Theoretical Framework

The relationship between health and GDP (gross domestic product) is complex and multifaceted. While it is difficult to establish causality, several theoretical relationships and empirical studies provide insights into the interactions between health and GDP growth. Here are a few key theoretical perspectives: A major theoretical framework for analyzing the sources of GDP growth is rooted in neoclassical growth theory, which was pioneered by Solow in 1956 [11]. Additionally, the endogenous growth theory, developed by Romer in 1986 [12] and Lucas in 1988 [3], also contributes to our understanding of GDP growth sources. Both neoclassical and endogenous growth theories can be employed to model the effects of human capital on GDP growth. Many studies investigating the impact of health and other forms of human capital on GDP growth have applied the Augmented Solow Model [13,14,15,16]. Consequently, health capital is regarded as a separate input in the production function, similar to reproducible physical capital, serving as a labor-augmenting factor in the production process.

Health is expected to affect the long-run GDP growth of a country because health has been found to be among the most important components for the development and wealth of a country in different previous studies. Economic theories and previous researchers such as Lucas [3], Barro [17], and Folland, Goodman, and Stano [18] suggested that health capital is a component of a potential source of GDP growth. Better health has been found to be a crucial part of overall well-being. Based on economic grounds, good health raises levels of human capital, and this has a positive effect on individual productivity and economic growth rates [19]. Better health increases labor force productivity by reducing incapacity, weakness, and the number of days lost to sick leave. Moreover, healthier workers are physically and mentally more energetic and thus more effective in the labor market. The effect of having less productive labor is stronger in developing countries because a higher proportion of the workforce is engaged in manual labor than in industrial countries [20]. Recently, an interesting possibility has emerged that the health-income correlation is partly explained by a causal link running from health to income. In this line, a theoretical relationship between health and GDP claims that health is a form of capital; investment in health can increase both human and physical accumulation, leading to overall economic growth [4].

From the above theoretical explanation, it is clear that health factors play a crucial role in a country’s GDP growth, both directly and indirectly. For example, several studies have recently reported how the COVID-19 pandemic severely impacted economies and GDP growth in several countries [21,22,23]. These factors can have a positive or negative impact on human capital and productivity, leading to either positive or negative GDP growth. Therefore, the impact of health factors on GDP growth is considered significant for an economy. Given the important role of factors in the GDP growth of a country, what and how health factors change the progress of GDP growth has long been an area of research. The link between health factors and GDP growth has been widely investigated across multiple studies, and researchers’ continuous efforts are expanding to explore the drivers of GDP growth. Despite the growing number of empirical and theoretical studies conducted to investigate the various links between health factors and GDP growth, there is no consensus yet. Recent studies confirm the consequences of health factors on GDP growth [24], while other studies find inconsistent outcomes. Studies conducted in the 19th century reported that health was a variable that significantly explained GDP growth from a statistical perspective [17,25,26]. Empirical studies analyzing the causal relationship between health factors and GDP growth are associated with statistical inconsistencies and biases.

To achieve the objectives of this study, a graphical framework is built, as depicted in Figure 1.

In the following section, we review recent and relevant studies on the relationship between health factors and GDP growth.

2.2. Relationship between Maternal Mortality Ratio (MMR) and GDP Growth

The theory that maternal health care development reduces maternal mortality levels provides a basis for claiming a correlation between maternal mortality and GDP growth. Reducing maternal mortality can have a positive impact on maternal productivity and contribute to GDP growth. Previous studies have found a statistically significant negative relationship between these two variables [5,27,28,29,30,31,32,33]. For example, Trondillo’s study [27], which used panel data from 193 UN countries spanning 53 years, reported a significant negative relationship between maternal mortality and GDP per capita. He found that the effect of maternal mortality cannot be ignored, as it has a considerable impact of 2.32% on GDP per capita. Another study by Frank [31] in Sub-Saharan Africa, using the Pooled Mean Group estimator of Autoregressive Distributed Lag (ARDL) and the Kao test for cointegration, concluded that there is a significant negative relationship between maternal mortality and GDP per capita. However, Amiri and Gerdtham [34], using the same methodology, reported mixed results, with maternal mortality having a significant negative effect on GDP growth in 40% of countries while having no significant effect on GDP growth in 33% of countries. They also argued that 11% of countries have a bidirectional relationship between maternal mortality and GDP growth based on causal analysis. Another study [35] strongly supports the bidirectional relationship between maternal mortality and GDP. Based on this literature, we hypothesize that:

H1. 

There is a significant negative effect of maternal mortality rate (MMR) on GDP growth in Saudi Arabia.

2.3. Relationship between Infant Mortality Rate (IMR) and GDP Growth

The relationship between infant mortality and GDP growth has garnered significant interest in the literature, as labor growth is one of the main factors in GDP growth. Extensive research has examined the impact of infant mortality on GDP growth. The effects of infant mortality are diverse across nations, with low- and high-income nations presenting a greater impact than middle-income nations [5]. Additionally, the impact differs non-linearly depending on the phase of demographic change at the start of the study period [5]. The empirical evidence on the subject is mixed, as demonstrated by the conflicting results of different studies [1]. However, most of the previous research has concluded that infant mortality negatively affects GDP growth [5,14,27,34,36,37,38,39,40,41,42]. For instance, Nishiyama [37] stated that lower GDP is often observed in developing countries with high infant mortality rates compared with developed countries. Similarly, Khan et al. [40] reported a significant negative association between infant mortality and GDP growth. Trondillo [27] also claimed a negative relationship between infant mortality and GDP per capita, with a significant impact of 4.81% on GDP per capita. Additionally, Rocco et al. [5] found that a 10% reduction in infant mortality increases GDP per capita by at least 9.6% in the long run over a quarter-century. Their findings are strongly supported by the previous study conducted by Nevo and Rosen [39], which stated that a 10% reduction in infant mortality could lead to a long-run GDP growth of at least 10.6%. Strittmatter and Sunde [43] also reported that a significant reduction in infant mortality had a significant positive effect on GDP growth in 12 European countries, using data from 1820 to 2010. Similarly, Bloom et al. [14] reported that a one-year increase in population life expectancy leads to a 4% increase in productivity. However, there are some contradictory findings, such as those of Acemoglu and Johnson [1], who studied low- and middle-income countries using panel data for 40 years between 1940 and 1980 and reported contradictory results. They found that improvements in life expectancy had a significant negative effect on GDP per capita. Eboh et al. [44] conducted a study to examine the links between health expenditure, child mortality, and GDP growth in Nigeria using time-series data covering the 1980–2020 sample periods. They found that infant mortality had a positive and significant impact on GDP growth, which is a contradictory finding compared with earlier studies. Based on the existing evidence, we can hypothesize that:

H2. 

There is a significant inverse association between IMR and GDP growth in Saudi Arabia.

2.4. Relationship between Road Traffic Mortality (RTM) and Economic Growth

The improvement of traffic safety rules and regulations, infrastructure development, and awareness programs have significantly reduced the world’s road accident mortality. From 2002 to 2019, the average number of deaths decreased from 19 to 17 per 100,000 people [45]. However, the death rate from road accidents in Saudi Arabia increased from 17.2 to 35.9 per 100,000 people during the same period [45]. This has led to an investigation of the impact of road traffic mortality (RTM) on Saudi Arabia’s GDP growth. According to the World Health Organization (WHO) [46], RTM is a great economic burden worldwide, estimated to cause damages worth 518 billion US dollars and cost between 1% and 3% of the country’s GDP [47]. A few studies have examined the relationship between RTM and GDP growth [5,48,49,50,51,52,53,54]. Dalal et al. [48] found a significant inverse relationship between RTM and GDP growth. Al Turki [49] claimed that RTM is a national problem in Saudi Arabia, hindering GDP progress as it mainly affects the young and economically productive age groups. Bhavan [52] revealed that RTM significantly hinders the long-run improvement of the GDP growth rate in Sri Lanka. Gorea [51] reported that RTM has immediate and long-term consequences for fatalities, their families, and a country’s GDP. Recent research conducted by Zeng et al. [53] in China claimed a significant negative correlation between GDP and RTM. However, Hu et al. [54] found a strong positive correlation between RTM and GDP growth in their study in China using Pearson correlation coefficients. Based on the existing literature, we can hypothesize that:

H3. 

There is a significant inverse association between RTM and GDP growth in Saudi Arabia.

2.5. Relationship between Health Care Expenditure (HCE) and GDP Growth

The relationship between healthcare expenditures and GDP growth has gained significant attention due to the crucial role of human capital development in economic growth. There are at least two ways to describe the link between healthcare expenditures and economic growth. Firstly, healthcare expenditures can be seen as an investment in human capital, which can lead to GDP growth. Secondly, increasing healthcare expenditures may lead to routine health interventions that can improve labor productivity and boost GDP. Some studies have shown that healthcare development can lead to GDP growth and vice versa [55]. However, the relationship between healthcare expenditure (HCE) and GDP growth varies across states in the same country, across a country’s level of development, and across the nature of the country’s factors of production (i.e., labor-intensive or capital-intensive), as well as across data availability, estimation processes, and model specifications in each study [24,56]. Furthermore, there is an ongoing debate about what types of HCE and what ideal level of expenditure are optimal for GDP growth [57]. Several studies have been conducted to explore the true relationship between HCE and GDP growth [16,24,56,58,59,60,61,62,63,64,65]. For instance, Raghupathi and Raghupathi [56] found a positive and strong correlation between HCE, GDP growth, and labor productivity by analyzing Bureau of Labor Statistics data for the period 2003–2014 in the United States of America. Other studies have reported similar results, where higher HCE is linked to higher labor productivity, leading to higher GDP [58,66,67]. Lucian et al. [62] investigated the relationship between HCE and GDP growth in current European Union member countries during the period 1995–2007 and found a positive relationship between investing in population health and GDP growth. Moreover, Onisanwa [63] conducted a study in Nigeria and found a short- and long-run positive and significant bidirectional relationship between HCE and GDP growth per capita. Sethi et al. [68] investigated the short-term and long-term impacts of HCE on the GDP growth of South Asian countries by using data from 1996–2018. They used the ordinary least squares (OLS) method and random effects models, the Johansen–Fisher cointegration test, and the Granger causality test to capture the relationship and confirm a two-way causality between health expenditure and economic growth in the short run only. Meta-regression analysis has also confirmed the key role of HCE in explaining GDP growth across countries [24]. However, Weil [69] argued that increased life expectancy through increasing HCE may lead to population growth, which could reduce GDP per capita. Pauly and Saxena [70] also supported the inverse relationship, stating that higher health expenditure in developed countries may hamper GDP growth due to the overexploitation of productive resources by oversized health sectors. Based on the existing literature, we can hypothesize that:

H4. 

There is a significant positive association between HCE and GDP growth in Saudi Arabia.

3. Materials and Methods

The objective of this study is to investigate the role of health factors in the GDP growth of Saudi Arabia. To achieve this objective, both descriptive and analytical methods were employed to provide a clear and evidence-based understanding of the relationship between health factors and GDP growth.

3.1. Description and Sources of Data

This study used secondary yearly data obtained from the World Bank’s (WB) open data source. Data for all relevant variables, including GDP growth (GDPG), healthcare expenditure (HCE), infant mortality ratio (IMR), maternal mortality ratio (MMR), and road traffic mortality (RTM), were collected for the period 1990–2019 to investigate both short-run and long-run associations among these variables. Relevant secondary data and necessary information were also obtained from WB. GDP growth (GDPG) was used as the dependent variable, while HCE, IMR, MMR, and RTM were treated as independent variables. The variables, their definitions, and their sources [6,71,72,73] are presented in

Table 1. Variables, their descriptions, and data sources.

Variables	Descriptions	Data Sources
Gross domestic product growth (GDPG)	GDP per capita is the gross domestic product divided by the midyear population. GDP is the sum of the gross value added by all resident producers in the economy plus any product taxes and minus any subsidies not included in the value of the products. It is calculated without making deductions for the depreciation of fabricated assets or for the depletion and degradation of natural resources. Data are in current U.S. dollars.	[6]
Healthcare expenditure (HCE)	Current expenditures on health per capita in current US dollars. Estimates of current health expenditures include healthcare goods and services consumed each year.	[6]
Infant mortality rate (IMR)	The mortality rate is the number of infants dying before reaching one year of age per 1000 live births in a given year.	[71]
Maternal mortality ratio (MMR)	The mortality ratio is the number of women who die from pregnancy-related causes while pregnant or within 42 days of pregnancy termination per 100,000 live births. The data are estimated with a regression model using information on the proportion of maternal deaths among non-AIDS deaths in women ages 15–49, fertility, birth attendants, and GDP measured using purchasing power parities (PPPs).	[72]
Road traffic mortality rate (RTM)	Mortality caused by road traffic injuries is estimated by road traffic fatal injury deaths per 100,000 populations.	[73]

.

3.2. Model Specification

This study aims to investigate the relationship between health factors and GDP growth in Saudi Arabia by utilizing a modified model of the Cobb–Douglas function [74] with time series data. The model is presented as follows:

$$GDP{G}_t=\alpha +{\beta }_{1}IM{R}_t+{\beta }_{2}MM{R}_t+{\beta }_{3}HC{E}_t+{\beta }_4{RTM}_t+{\in }_t.$$

(1)

where $GDP{G}_t$ represents the GDP growth in year t, α is a constant term, and ${\beta }_1,{\beta }_2,{\beta }_3,and{\beta }_4$ are the slope coefficients of the independent variables, namely $IM{R}_t,$ $MM{R}_t,$ $HC{E}_t,$ and ${RTM}_t,$ respectively. $IM{R}_t$ represents the infant mortality rate in year t, $MM{R}_t$ represents the maternal mortality ratio in year t, $HC{E}_t$ represents healthcare expenditure in year t, ${RTM}_t$ represents the road traffic mortality rate in year t, and ${\in }_t$ represents the error term.

3.3. Analytical Tools

This study is a quantitative empirical analysis that utilizes both descriptive and empirical methods to address the research objectives. Descriptive statistical analysis is used to present background characteristics such as mean and standard deviation, while skewness, kurtosis, and the Jarque–Bera (JB) test of normality are used to assess the normality of the variables. Graphical analysis is conducted to show the trend of the study variables, while percentage change is used to present the degree of change over time. The positive value confirms a percentage increase, whereas the negative value confirms a percentage decrease. For the empirical analysis, the study uses the Augmented Dickey–Fuller (ADF) and Philips–Perron (PP) unit root tests, the Johansen cointegration test, the vector error correction (VEC) model, and the Granger causality test to analyze the annual data series. Eviews (Version 9) is used for data analysis and all tests. These methods allow for evidence-based associations to be made between the studied risks and the risk factors.

3.4. Unit Root Test

At the beginning of our investigation, to detect the order of integration [$I\left(d\right)$, where $d$ is the order of integration] of the variables in the model under study, we tested the stability of the annual data series to avoid errors in estimation [75]. To investigate the stationarity and define the order of integration of each yearly data series, the commonly accepted ADF [76] and PP [77] tests for unit root tests were used. The ADF unit root test estimates the following equation [78]:

$$∆y_t={\alpha }_0+\gamma y_{t-1}+\sum _{i=2}^p{\beta }_i{∆y}_{t-i}+{\epsilon }_i,$$

(2)

where $y_t$ is a random walk with drift around a stochastic trend. If the coefficient $\gamma =0$, then Equation (2) becomes the first difference ($∆$) and has a unit root. The results of the ADF test for unit root are also verified by the PP test for unit root, and the regress of the PP test is as follows:

$$y_t={\alpha }_0+{\gamma }_t+\delta y_{t-1}+\sum _{i=1}^p{\beta }_i∆y_{t-1}+{\in }_t.$$

(3)

In Equation (3), ${\gamma }_t$ may be $0$, $\mu$ or $\mu +{\beta }_t$, and ${\in }_t$ is $I\left(0\right)$ may not be homoscedastic. The ADF and PP tests can be used when the error term $\left({\in }_t\right)$ is not white noise. The hypothesis of the ADF test can be expressed as:

The null hypothesis, $H_0:$ variable has a unit root;

The alternative hypothesis, $H_a:$ variable has no unit root.

If the null hypothesis is rejected at the 0.01, 0.05, or 0.10 level of significance, then it may be said that the variable has no unit root. Therefore, the variable is stationary at a 0.01, 0.05, or 0.10 level of significance. However, if the null hypothesis is not rejected, then the variable has a unit root. In this case, the variable is said to be non-stationary. The non-stationary variable may be converted to stationary by taking the differences (forward difference operator, $∆$).

3.5. Cointegration Test

The cointegration test is an econometric technique that is used to determine a possible correlation between time series processes in the long term. In this study, after examining the stationarity tests, the cointegration test was performed by the Johansen cointegration test [79] to identify the long-run equilibrium relationship among the time series data. The key to the cointegration test lies in selecting the proper form of cointegration test and lag order. This study used three methods, namely Akaike [80], Schwarz [81], and Hannan-Quinn [82], to determine the lag length. These are considered to be the classical procedures to determine the lag length [83]. Two more criteria, namely likelihood ratio (LR) and final prediction error (FPE) [84], were also used for lag order selection.

Cointegration can be employed as the right technique for reframing the time series into a stationary form because cointegration forms a synthetic stationary series from a linear combination of two or more non-stationary series [85]. Johansen’s technique gives two LR tests for the number of cointegrating vectors $\left(r\right),$ which are originated by the trace and maximum eigenvalue tests, as follows:

$${\lambda }_{trace\left(r\right)}=-T\sum _{i=r+1}^n\ln \left(1-{\lambda }_i\right),$$

(4)

$${\lambda }_{max\left(r,r+1\right)}=-T\ln \left(1-{\lambda }_{i+1}\right),$$

(5)

where T is the sample size (for this study, $T=30$), n is the number of endogenous variables, and ${\lambda }_i$ is ith the largest eigenvalue. The above procedures depend on the relationship between the rank of the matrix and its eigenvalues [78]. The hypothesis of this test can be formulated as follows:

The null hypothesis, $H_0:$ no cointegration in the equation;

The alternative hypothesis, $H_a:$ cointegration in the equation.

The acceptance of $H_0$ indicates the short-run relationship, and for this case, the vector autoregressive (VAR) model is investigated. Again, the acceptance of $H_a$ indicates the existence of a cointegration relationship (having long-run relationships among the variables), and for this case, VEC modeling is conducted. After VEC modeling, the Granger causality test is performed between the variables, which is a measure of the capability to predict the future values of a time series by using the past values of another time series.

3.6. Lag Selection Criteria

One of the important responsibilities of the VAR model is to select the lag order because Johansen cointegration tests are sensitive to the lags used. This criterion method is used throughout unrestricted VAR. Therefore, the optimal lag length order is defined by the Akaike information criterion (AIC).

3.7. Vector Error Correction (VEC) Model

The cointegration among variables in the model exclusively illustrates a long-run equilibrium association. However, there may be short-run disequilibrium among those variables. The VEC model is applied when a set of variables is found to have one or more cointegrating equations for both short-run changes in the variables and deviations from the long-run equilibrium. The VEC model also requires the mention of several cointegrating equations among the endogenous variables under study. When the data series are found to be stationary at the first difference, the VEC model can be used. The VEC models are parameterizations of $VAR\left(p\right)$ models in levels, and they can be represented symbolically as:

$$∆y_t=\mathsf{\alpha }{\mathsf{\beta }}^{\prime }{\mathrm{y}}_{\mathrm{t}-1}+{\mathsf{\Gamma }}_1\Delta y_{t-1}+{\mathsf{\Gamma }}_2\Delta y_{t-2}+\cdots {\mathsf{\Gamma }}_{p-1}\Delta y_{t+1-p}+{\epsilon }_t$$

(6)

Equation (6) is a VEC model where $∆y_t=y_t-y_{t-1},$ $t=\mathrm{1,2},\dots ..,T;\alpha {\beta }^{\prime }{y}_{t-1}$ is the lagged error correction term (ECT), $y_{t-1}$ is the non-stationary variable, $\mathsf{\Gamma }$ is the matrix of variables, $\beta$ is the cointegrating matrix, ${\beta }^{\prime }$ is the transposed matrix of $\beta$, α is the loading matrix, $p$ is the lag order of the model in its VAR form, and ${\epsilon }_t$ is the normally distributed error term.

In the cointegrating model, if the coefficient is negative in sign and significant, then it indicates a long-run causality running from one variable in the system to the other variables. Short-run causality is indicated when the coefficient of the lagged value of the variable under consideration is not equal to zero.

3.8. Granger Causality Test

Cointegration is a statistical concept that indicates the existence of a long-term relationship between two or more non-stationary time series variables. However, it does not reveal the direction of causality. To explore the direction of causality between variables in a time series dataset, the Granger causality test is employed [86,87]. This method uses probabilistic reasoning to identify patterns of correlation between variables. However, like cointegration, the Granger causality test is only applicable to stationary data series. The Granger causality test is used to confirm the causal relationship between two variables. It is considered a hypothesis test that determines the possibility of one variable forecasting the other [88]. For example, if variable X causes variable Y, it means that X contributes to forecasting Y, indicating a connection that runs from dependent variables to independent ones and vice versa. The null hypothesis is rejected based on a low p-value (<0.05). If one data series causes another, it indicates that the variable causing the other can be modeled to produce a more accurate forecast of the caused variable.

4. Results Analysis

This study investigates the impact of the infant mortality rate (IMR), maternal mortality ratio (MMR), road traffic mortality (RTM), and healthcare expenditure (HCE) on the GDP growth of Saudi Arabia. The study uses annual time series data from 1990 to 2019, and the findings of the econometric analysis are presented in the following sections. The results show that there are long-term relationships among the variables, and there is no significant unidirectional causality between health factors and GDP growth; however, we find a significant unidirectional causality between independent variables from infant mortality rate and road traffic mortality to healthcare expenditure.

Table 2. Descriptive statistics of the variables.

Variables	$\mathit{N}$	$\stackrel{-}{\mathit{x}}\pm \mathbf{S}\mathbf{D}$	Median	Minimum	Maximum	Skewness	Kurtosis	Jarque-Bera	Prob.
GDPG	30	3.541 ± 4.72	2.744	−3.76	15.19	0.899	3.497	4.344	0.114
HCE	30	672.25 ± 401.74	435.998	326.00	1484.59	0.929	2.329	4.88	0.0872
IMR	30	16.36 ± 8.71	14.500	5.70	35.60	0.641	2.338	2.602	0.2723
MMR	30	24.15 ± 7.73	22.000	16.80	46.00	1.233	3.704	8.217	0.0164
RTM	30	22.24 ± 5.82	23.400	16.10	35.90	0.503	2.226	2.016	0.365

presents the descriptive statistics of the selected variables used in the analysis, providing information on their behavior over the years. The statistics include the mean, median, standard deviation, minimum, and maximum values of the data series with the measures of skewness, kurtosis, and the JB test with associated p-values.

Table 3. Percentage changes of the study variables from 1990 to 2019.

Variables	1990	2019	Changes (%)
Gross domestic product growth (GDPG)	15.1934	0.300	−98.0255
Healthcare expenditure (HCE)	326.00	1466.80	349.9387
Infant mortality rate (IMR)	35.60	5.70	−83.9888
Maternal mortality ratio (MMR)	46.00	16.80	−63.4783
Road traffic mortality rate (RTM)	16.00	35.90	124.375

presents the same statistics for the initial year (1990) and the terminal year (2019) of the study, along with the percentage change from 1990 to 2019. This change shows a decreasing trend in GDP growth (change in%: −98.0255), IMR (change in%: −83.9888), and MMR (change in%: −63.4783); and an increasing trend in HCE (change in%: 349.9387) and RTM (change in%: 124.375). To supplement the tabular analysis, graphical analysis was also conducted for the study variables at level and their first difference, as shown in Figure 2, Figure 3, Figure 4, Figure 5 and Figure 6. The figures indicate that all series, except for IMR and MMR, had an increasing trend over the analyzed period. To ensure accurate regression analysis, it is necessary to check for the presence of unit roots in the data series.

To prepare for cointegration analysis of both the dependent variable (GDP growth) and the independent variables (health factors), ADF and PP tests were conducted to determine the order of integration for the data series.

Table 4. Results of unit root tests.

Variables	At Level					At First Difference
	ADF	Prob. a	PP	Prob. a	Decision	ADF	Prob. a	PP	Prob. a	Decision
GDPG	−4.402	0.0017	−4.405	0.0017	I(0)	−5.803	0.0001	−7.274	0.0000	I(1)
HCE	1.193	0.9973	1.629	0.9992		−0.627	0.0455	−5.035439	0.0003	I(1)
IMR	−4.821	0.0009	−16.820	0.0001	I(0)	−5.436	0.0001	−5.918913	0.0000	I(1)
MMR	−2.983	0.0493	−9.576	0.0000	I(0)	−3.699	0.0100	−5.596931	0.0001	I(1)
RTM	1.120	0.9967	0.796	0.9922		−3.348	0.0220	−3.348	0.0220	I(1)

Note: ‘GDPG, Gross domestic product growth’, ‘HCE, Healthcare expenditure’, ‘IMR, Infant mortality rate’, ‘MMR, Maternal mortality ratio’, ‘RTM, Road traffic mortality rate’; ‘ADF, Augmented Dickey–Fuller test for unit root’, ‘PP, Phillips and Perron test for unit root’, ‘^a MacKinnon p-values’.

shows that all variables were found to be statistically significant (p < 0.05) at the first-order difference. The tests revealed that the variables HCE and RTM were not stationary at a level in both unit root tests, as their p-values were greater than 0.05. However, once these series were converted to their first difference (∆), all the data series became stationary (p < 0.05). Thus, all variables were stationary at the first difference [I(1)]. These results were useful for selecting the lag length criterion, as the VAR model cannot be run when all variables are not stationary at the level (i.e., I(0)). Lag selection criteria were applied to choose the lag length to be used in the VEC model. The optimal lag length was determined using AIC, and

Table 5. Lag order selection criteria: VAR order selection criteria (unrestricted VAR).

Lag	LogL	LR	FPE	AIC	SC	HQ
0	−436.2063	NA	1.07 × 108	32.68195	32.92192	32.75330
1	−252.7657	285.3521	892.4079	20.94561	22.38542	21.37374
2	−227.8634	29.51386	1112.504	20.95284	23.59251	21.73775
3	−169.8234	47.29183 *	182.4814 *	18.50544 *	22.34495 *	19.64712 *

Note: * indicates lag order selected by the criterion, LR: Likelihood Ratio test statistic (each test at 5% level), FPE: Final Prediction Error, AIC: Akaike Information Criterion, SC: Schwarz Information Criterion, HQ: Hannan-Quinn Information Criterion.

indicated that most criteria suggested a lag of three. Therefore, the optimal lag order was three for this yearly time series dataset, which was used to estimate the Johansen cointegration test and the VEC model.

The selection of an appropriate form of cointegration test and lag order is crucial to the cointegration test. In this study, the Johansen cointegration test was used to determine the number of cointegration equations, which examine the long-term correlation between a dependent variable (GDPG) and independent variables (HCE, IMR, MMR, and RTM). The results, presented in

Table 6. Johansen cointegration test for GDP growth as a dependent variable and health factors as independent variables.

Hypothesis	Eigenvalue	Trace Statistics	0.05 Critical Value	Prob. a	Max-Eigen Statistics	0.05 Critical Value	Prob. a
None *	0.999859	358.9761	69.81889	0.0001	230.5754	33.87687	0.0001
At most 1 *	0.905974	128.4006	47.85613	0.0000	61.46879	27.58434	0.0000
At most 2 *	0.723651	66.93186	29.79707	0.0000	33.43839	21.13162	0.0006
At most 3 *	0.514572	33.49347	15.49471	0.0000	18.79086	14.26460	0.0090
At most 4 *	0.431915	14.70261	3.841466	0.0001	14.70261	3.841466	0.0001

Note: ‘* rejection of the hypothesis at 0.05’, ‘^a MacKinnon p-values’.

, confirmed the rejection of the null hypothesis of no cointegration of GDP growth and health factors. Specifically, both the computed trace and maximum eigenvalue statistics, along with their corresponding critical values, indicated that the null hypothesis of no cointegration could be rejected under both tests at the 0.05 significance level. Moreover, both tests indicated five cointegrating equations at the same significance level. The presence of cointegration suggested the existence of stable and long-run equilibrium relationships among the variables under study, which is crucial for validating Granger causality. Without stationary variables and cointegration, Granger causality tests are not valid. The cointegrated time series have an error-correction representation, which reflects the long-run adjustment mechanism. Based on the evidence of the existence of cointegration relationships, VEC modeling was conducted [89].

The VEC model is used to determine the number of cointegration equations among variables, which was established by conducting the Johansen cointegration test. Results from this analysis indicate that the variables have long-run equilibrium relationships; however, in the short term, the variables are in disequilibrium. The VEC model is utilized to express the short-term imbalance and dynamic structure. The objective of using the VEC estimation model is to adjust for short-run and long-run changes in variables and deviations from equilibrium relationships.

Table 7. Results of cointegration equation.

Cointegrating Equation:	Co-Efficients
GDPG(−1)	1.000000
IMR(−1)	3.667891 [10.2649]
MMR(−1)	−0.408164 [−1.29368]
HCE(−1)	0.016317 [12.3420]
RTM(−1)	0.711331 [4.26702]
Constant	−76.66228

Note: ‘GDPG, Gross domestic product growth’, ‘IMR, Infant mortality rate’, ‘MMR, Maternal mortality ratio’, ‘HCE, Healthcare expenditure’, ‘RTM, Road traffic mortality rate’; t-statistics in [ ].

displays both the short-run and long-run relationship effects and cointegration equation results of GDP growth and health factors. The appropriate lag order of the VEC model is two, selected using the FPE criterion [84], since the lag order of VAR is three. The VEC model’s cointegration vector $\beta =(13.6679-0.40820.01630.7113)$ is shown in Table 7, and the adjustment parameters matrix α is presented in

Table 8. Vector error correction model estimation results and test for health variables.

Error Correction	Δ(GDPG)	Δ(IMR)	Δ(MMR)	Δ(HCE)	Δ(RTM)
Cointegration equation	−1.540225 (0.39155) [−3.93369]	−0.003702 (0.00718) [−0.51525]	0.240182 (0.08459) [2.83938]	−16.20806 (8.72106) [−1.85850]	0.252136 (0.16557) [1.52281]
Δ(GDPG(−1))	0.237060 (0.24785) [0.95647]	−0.000529 (0.00455) [−0.11623]	−0.148580 (0.05355) [−2.77484]	7.241893 (5.52044) [1.31183]	−0.033821 (0.10481) [−0.32270]
Δ(GDPG(−2))	−0.246046 (0.15614) [−1.57580]	−0.000289 (0.00287) [−0.10096]	−0.092157 (0.03373) [−2.73202]	−0.272160 (3.47775) [−0.07826]	−0.001615 (0.06603) [−0.02447]
Δ(IMR(−1))	−12.56020 (14.1171) [−0.88972]	0.824344 (0.25903) [3.18237]	−4.318597 (3.04984) [−1.41601]	173.3437 (314.434) [0.55129]	6.612241 (5.96964) [1.10764]
Δ(IMR(−2))	−34.80565 (16.9289) [−2.05599]	−0.011618 (0.31063) [−0.03740]	13.43568 (3.65731) [3.67365]	−523.8543 (377.063) [−1.38930]	2.028268 (7.15868) [0.28333]
Δ(MMR(−1))	0.051208 (0.87227) [0.05871]	−0.024452 (0.01601) [−1.52775]	−0.165098 (0.18844) [−0.87611]	−37.19748 (19.4283) [−1.91460]	0.329417 (0.36885) [0.89308]
Δ(MMR(−2))	−0.480208 (0.88821) [−0.54065]	0.003150 (0.01630) [0.19326]	−0.041692 (0.19189) [−0.21727]	−20.31339 (19.7834) [−1.02679]	0.181386 (0.37560) [0.48293]
Δ(HCE(−1))	0.031130 (0.01242) [2.50632]	−3.25 × 10−5 (0.00023) [−0.14247]	−0.003400 (0.00268) [−1.26718]	−0.030118 (0.27665) [−0.10887]	−0.002438 (0.00525) [−0.46415]
Δ(HCE(−2))	0.022837 (0.01382) [1.65296]	6.23 × 10−5 (0.00025) [0.24561]	−0.009932 (0.00298) [−3.32775]	0.071061 (0.30772) [0.23092]	−0.008281 (0.00584) [−1.41747]
Δ(RTM(−1))	2.060809 (0.72841) [2.82919]	0.014707 (0.01337) [1.10033]	−0.250197 (0.15737) [−1.58991]	30.97105 (16.2241) [1.90895]	0.077495 (0.30802) [0.25159]
Δ(RTM(−2))	−0.350043 (0.66747) [−0.52443]	0.005031 (0.01225) [0.41082]	−0.150864 (0.14420) [−1.04622]	19.62198 (14.8668) [1.31985]	−0.561913 (0.28225) [−1.99083]
C	−54.42331 (14.7353) [−3.69340]	−0.149121 (0.27038) [−0.55153]	9.993350 (3.18340) [3.13920]	−434.1424 (328.204) [−1.32278]	10.72452 (6.23107) [1.72114]
R2	0.734468	0.989232	0.723306	0.458886	0.471737
Log likelihood (LL): −211.8849 Akaike information criterion (AIC): 20.50999 Schwarz criterion (SC): 23.62960

Note: ‘GDPG, Gross domestic product growth’, ‘IMR, Infant mortality rate’, ‘MMR, Maternal mortality ratio’, ‘HCE, Healthcare expenditure’, ‘RTM, Road traffic mortality rate’; Standard errors in ( ) and t-statistics in [ ], ‘Δ the first difference’.

. Inspection of t-statistics reveals that the coefficients of IMR, RTM, and HCE are statistically significant. As a result, the effects of infant mortality, health expenditure, and road traffic mortality on GDP growth are opposite. Therefore, the long-run cointegration equation can be written as follows:

$$\begin{gathered} GDP{G}_{t-1}=-3.667891IM{R}_{t-1}+0.408164MM{R}_{t-1}-0.016317HC{E}_{t-1} \\ -0.711331RT{M}_{t-1}+76.66228. \end{gathered}$$

(7)

Based on Equation (7), it is apparent that, holding other factors constant, a 1% increase in IMR will result in a 3.667891% decrease in GDP growth. Conversely, a 1% increase in MMR leads to a 0.408164% increase in GDP growth. Additionally, each percentage point increase in HCE will lead to a 0.016317% decrease in GDP growth, while a 1% increase in RTM will result in a 0.711331% decrease in GDP growth. For the developing countries, Nishiyama [37] identified that a 1% increase in per capita GDP lowers IMR by 0.255%. Bhavan [52] also found that the accident costs were about 1% higher than the fatality index, which was negatively associated with the change in the economic growth rate by 0.79% in Sri Lanka.

The VEC model, which consists of a set of linear equations (Equation (8)), estimates five separate models, one for each endogenous variable, based on the remaining endogenous variables and other exogenous variables. The focus of this study is primarily on the first equation of the VEC model. The matrix $\alpha =\left(-1.5402-0.00370.2402-16.20810.2521\right)$ contains the adjustment coefficients that correspond to a disequilibrium in the above cointegration equation. The results of the VEC model suggest that there is a long-run equilibrium relationship between variables, as indicated by the estimated parameter of the error correction term (ECT), which is negative (−1.540225) and statistically significant (t-statistics = −3.9337). The negative sign of the element α shows that GDP growth is an endogenous variable that aligns with the theoretical framework and indicates that the model is dynamically stable. This study confirms that there is a strong long-run equilibrium causality between GDP growth and health factors. Moreover, the coefficient of ECT is −1.540225, which implies that any deviation from the long-run equilibrium in the previous period is corrected in the current period at a speed of 154.023%. In other words, the adjustment from disequilibrium to equilibrium occurs at an extremely fast rate. Additionally, the predictor variables influence the dependent variable by 73% $R^2=0.7345)$. The VEC model is as follows:

$$\begin{gathered} \left[\begin{array}{c}∆GDP{G}_t\\ ∆IM{R}_t\\ ∆MM{R}_t\\ ∆HC{E}_t\\ ∆RT{M}_t\end{array}\right]=\left[\begin{array}{c}\begin{array}{c}-54.42331\\ -0.149121\end{array}\\ 9.993350\\ -434.1424\\ 10.72452\end{array}\right]+\left[\begin{array}{c}0.237060-0.000529-0.1485807.241893-0.033821\\ -12.560200.824344-4.318597173.34376.612241\\ 0.051208-0.024452-0.165098-37.197480.329417\\ 0.03113-0.0000325-0.00340-0.030118-0.002438\\ 2.0608090.014707-0.25019730.971050.077495\end{array}\right]\left[\begin{array}{c}∆GDP{G}_{t-1}\\ ∆IM{R}_{t-1}\\ ∆MM{R}_{t-1}\\ ∆HC{E}_{t-1}\\ ∆RT{M}_{t-1}\end{array}\right] \\ +\left[\begin{array}{c}-0.246046-0.000289-0.092157-0.272160-0.001615\\ -34.80565-0.01161813.43568-523.8543523.85438\\ -0.4802080.003150-0.041692-20.313390.181386\\ 0.0228370.0000623-0.0099320.071061-0.008281\\ -0.3500430.005031-0.15086419.62198-0.561913\end{array}\right]\left[\begin{array}{c}∆GDP{G}_{t-2}\\ ∆IM{R}_{t-2}\\ ∆MM{R}_{t-2}\\ ∆HC{E}_{t-2}\\ ∆RT{M}_{t-2}\end{array}\right] \end{gathered}$$

(8)

The results in Table 8 indicate that the VEC model has a fitting degree of $R^2>0.5$5 and small values for the AIC and SC criteria, suggesting that the model estimation is reasonable.

The study utilized the Granger causality test to determine the causal relationship between the variables under investigation. The results indicate that all the data series were stationary in the first difference, establishing the variables as I(1). As a result, the Granger causality test is valid for the cointegrated variables, and it can be used to predict possible causal relationships between them.

Table 9. Granger causality test between the variables.

Null Hypothesis	Observations	F-Statistic	Prob.	Decision	Direction of Causality
IMR does not Granger Cause GDPG	27	0.96736	0.4275	Accept	None
GDPG does not Granger Cause IMR		0.35266	0.7877	Accept	None
MMR does not Granger Cause GDPY	27	0.54904	0.6546	Accept	None
GDPG does not Granger Cause MMR		0.60092	0.6219	Accept	None
HCE does not Granger Cause GDPPG	27	0.41099	0.7469	Accept	None
GDPG does not Granger Cause HCE		1.44223	0.2603	Accept	None
RTM does not Granger Cause GDPG	27	0.78972	0.5138	Accept	None
GDPG does not Granger Cause RTM		1.04253	0.3953	Accept	None
MMR does not Granger Cause IMR	27	1.06395	0.3866	Accept	None
IMR does not Granger Cause MMR		0.56913	0.6418	Accept	None
HCE does not Granger Cause IMR	27	1.44118	0.2605	Accept	None
IMR does not Granger Cause HCE		7.51008	0.0015 **	Reject	IMR to HCE
RTM does not Granger Cause IMR	27	1.57923	0.2257	Accept	None
IMR does not Granger Cause RTM		1.07094	0.3838	Accept	None
HCE does not Granger Cause MMR	27	0.40361	0.7520	Accept	None
MMR does not Granger Cause HCE		1.35554	0.2849	Accept	None
RTM does not Granger Cause MMR	27	0.32052	0.8104	Accept	None
MMR does not Granger Cause RTM		0.39244	0.7598	Accept	None
RTM does not Granger Cause HCE	27	2.70998	0.0724 *	Reject	RTM to HCE
HCE does not Granger Cause RTM		0.95854	0.4315	Accept	None

Note: ‘GDPG, Gross domestic product growth’, ‘IMR, Infant mortality rate’, ‘MMR, Maternal mortality ratio’, ‘HCE, Healthcare expenditure’, ‘RTM, Road traffic mortality rate’, ‘* significant at 10%’, ‘** significant at 5%’.

presents the results of the Granger tests using three lags. The analysis revealed a unidirectional causality between infant mortality and healthcare expenditure, suggesting that any changes in infant mortality would lead to changes in healthcare expenditure. Furthermore, there is unidirectional causality between road traffic mortality and healthcare expenditure, indicating that any changes in road traffic mortality would lead to changes in healthcare expenditure.

5. Discussion

The impact of mortality and healthcare spending on individual income is well established. However, at the national level, while it is evident that poorer countries generally have a lower health status, the connection between changes in income and health outcomes is unclear. The death of an infant is a tragic event for both the family and the community [90]. In 2022, Saudi Arabia’s infant mortality rate (IMR) was 5.497 per 1000 live births, showing a 3.54% reduction from the previous year, and the trend is gradually declining [91]. Despite this decline, the IMR remains a significant issue, and there are no straightforward solutions. One question that arises is whether better child health leads to higher GDP growth outcomes in Saudi Arabia, whether such a relationship exists for infant mortality in Saudi Arabia, and, if so, what the causal effect may be. The study’s findings indicated a significant negative relationship between the infant mortality rate (IMR) and GDP growth in Saudi Arabia. Therefore, the second hypothesis (H2), proposing an inverse association between IMR and GDP growth in Saudi Arabia, was supported and accepted. This finding is consistent with previous studies [5,14,27,34,36,37,38,39,40,41,42,91,92], which have also reported a detrimental impact of IMR on economic growth. For example, Klobodu et al. [92] discovered that child health causes GDP growth in six sub-Saharan African countries (Burkina Faso, Togo, Ghana, Ivory Coast, Botswana, and South Africa). In addition, Lawal et al. [93] identified significant positive effects of infant and maternal mortality on GDP growth. Furthermore, Ogunjimi and Adebayo’s study [94] found no causal relationship between real GDP and IMR. Instead, they identified a unidirectional causal relationship between GDP and health expenditure.

Maternal mortality continues to be the primary cause of death and disability among women of reproductive age, with potentially significant but inadequately documented economic implications [5]. In Saudi Arabia, the MMR was 16.80 per 100,000 live births in 2019, down from 17.00 in 2013 [91]. Interestingly, the relationship between MMR and GDP growth during economic booms in Saudi Arabia was found to be unstable. Although there is no significant relationship between MMR and GDP growth at a 5% level, it is significant at 10% levels, and the direction of the relationship is interestingly positive. The first hypothesis (H₁) of this study posited that there would be a negative relationship between the maternal mortality rate (MMR) and GDP growth in Saudi Arabia. However, the results of the study led to the rejection of this hypothesis, contradicting the findings of earlier research [5,27,28,29,30,31,32,33], which reported a negative association between MMR and GDP growth. There are several reasons why there may be a positive relationship between the maternal mortality ratio (MMR) and GDP growth in Saudi Arabia. One contributing factor to the increase in GDP despite the rise in maternal mortality may be the fact that the rate of increase in maternal mortality is lower than the rate of increase in female labor force participation. During the earlier part of the study period, women’s participation in the labor force was minimal, so the impact of maternal mortality on the country’s GDP growth was insignificant. However, the Saudi government launched Vision 2030, which includes various programs and policies aimed at promoting women’s participation in the workforce. As a result, women’s labor force participation has increased to 30%, despite the increase in maternal mortality. This may have contributed to the overall increase in GDP. Another reason could be the recent increase in oil prices, which has had a significant impact on eliminating the negative effect of the MMR on GDP growth. Additionally, the new projects that Saudi Arabia is investing in are also boosting GDP growth. In conclusion, although the MMR has a low level of significant positive relationship with GDP, it is important for policymakers to still consider lowering the MMR as the Saudi Arabian government has a strategy to transform the economy from a factor-driven nation that heavily depends on oil exports to an efficiency-driven nation that produces standard products and services. Lowering the MMR is essential to sustain GDP growth in an efficiency-driven economy, promote economic development, and achieve sustainable development goals. This can be achieved by investing in healthcare systems, increasing access to maternal healthcare services, and implementing policies that support maternal and child health.

Accidents, while unexpected, have become a significant social and economic burden worldwide. The study findings indicated a strong negative relationship between road traffic mortality (RTM) and GDP growth in Saudi Arabia, thereby providing support for the third hypothesis (H₃) that a significant inverse association exists between RTM and GDP growth in the country. This finding is consistent with earlier studies [30,48,49,50,51,52,53,95] that have also reported similar results, highlighting the detrimental impact of RTM on economic growth. These studies have demonstrated that road traffic accidents (RTAs) not only result in the loss of human lives and disabilities but also hinder the performance of economies. In Saudi Arabia, RTAs are considered an epidemic issue [96], with one of the highest death rates caused by road accidents [97]. The consequences of RTAs include injuries, deaths, property damage, congestion, disruptions, and delays to public transportation systems [98]. Globally, RTAs result in the deaths of approximately 130,000 people each year, which accounts for about 5% of total deaths in Saudi Arabia. The RTM (road traffic mortality) in 2019 was 3.9 per 100,000 people [45]. The impact of RTAs also leads to increased hospitalization and healthcare expenses, significantly affecting the Saudi economy and contributing to health problems.

The healthcare expenditure of a country or region is crucial in determining its GDP growth rate. Health-related expenditure is on the rise worldwide, posing a challenge to the stability of national health systems’ GDP growth even in high-income countries. In Saudi Arabia, the healthcare sector is a top priority, and there are significant opportunities for growth in this high-potential occupational sector [91]. The relationship between healthcare expenditure and GDP growth rate is an important concern [56]. At the national level, healthcare expenditure contributes to multi-factor productivity [99], indicators of labor productivity [100], personal spending [101], GDP [102,103], and other factors. Most previous studies have shown that an increase in healthcare expenditure has a positive relationship with the GDP growth rate [56,104,105]. However, this study identified a significant negative relationship between healthcare expenditure and GDP growth in Saudi Arabia. The study findings did not provide support for the fourth hypothesis (H4) that posited a significant positive relationship between healthcare expenditure (HCE) and GDP growth in Saudi Arabia. This finding contradicts the results of earlier studies conducted in other countries, which reported a positive association between HCE and GDP growth. Halıcı-Tülüce et al. [106] found that higher healthcare expenditure had a statistically significant negative effect on GDP growth.

The negative association between healthcare expenditure and GDP growth in Saudi Arabia may be due to several factors. One possible reason is the high mortality rate and the need to become a healthy nation by 2030, which require increased healthcare spending and ultimately decrease GDP growth. Saudi Arabia invests a large portion of the government’s budget in healthcare, with less funding available for infrastructure development or education, which could have a more direct impact on GDP growth. A second potential reason could be inefficiencies in the healthcare system, such as a lack of coordination between different healthcare providers and excessive use of expensive medical technologies. Inefficient allocation of healthcare funds and a lack of effective policies to optimize healthcare spending can have a negative impact on GDP growth. Another potential reason could be overinvestment in healthcare relative to actual demand, leading to a negative impact on GDP growth. Saudi Arabia has a relatively young population, which means that the demand for healthcare services may not be as high as in other countries with older populations. Additionally, inadequate regulation of the healthcare sector, low productivity, and limited technological advancement can have a negative effect on the economy. Furthermore, Saudi Arabia’s economy is heavily dependent on oil exports, which means that fluctuations in oil prices can have a significant impact on the country’s GDP. Healthcare expenditures may not have as direct an impact on the economy as the oil industry and thus may not contribute as much to overall GDP growth. Finally, the COVID-19 pandemic has had a significant impact on healthcare expenditure in Saudi Arabia, which has affected the country’s economic growth. It is worth noting that there may be other factors at play as well, and the relationship between healthcare expenditure and GDP growth can be complex and multifaceted. However, these are some potential reasons why healthcare spending may have a negative effect on GDP growth in Saudi Arabia. Overall, the negative effect of healthcare expenditure on GDP growth in Saudi Arabia is likely the result of a combination of factors.

6. Practical Implications

The study’s findings have important policy implications for achieving sustainable GDP growth in Saudi Arabia. Based on the study results, the following policy recommendations are suggested: i. The government should develop plans and strategies to reduce infant, maternal, and road traffic mortalities, which in turn can decrease healthcare expenditure and contribute to GDP growth. ii. The government and its agencies should collaborate with the United Nations to reduce infant and maternal mortalities and decrease healthcare expenditure, which can positively impact GDP growth. iii. The private sector should be encouraged to support the government in implementing policies aimed at reducing infant, maternal, and road traffic mortalities to promote GDP growth. iv. The government should implement effective regulations and awareness programs to reduce road traffic mortality. By implementing these recommendations, the government can create an environment that promotes sustainable GDP growth while also improving the overall health and well-being of its citizens.

7. Conclusions

This study aims to investigate the impact of health factors on GDP growth in Saudi Arabia using annual time series data from 1990 to 2019. The study seeks to identify and analyze whether there is a long-run equilibrium relationship, a short-run dynamic relationship, or a causal relationship between Saudi Arabia’s GDP growth and health factors over time. To achieve this, the Johansen cointegration test, VEC model, and Granger causality test were employed. The study’s findings uncover several significant relationships between key factors and the GDP growth of Saudi Arabia. Firstly, infant mortality, road traffic mortality, and healthcare expenditure exhibit a notable negative association with GDP growth. In contrast, maternal mortality demonstrates a positive association with GDP growth. Additionally, all of these factors are identified to have a long-run equilibrium relationship with GDP growth. Furthermore, the study highlights a significant unidirectional causality running from the infant mortality rate and road traffic mortality to healthcare expenditure.

These findings contrast with the existing literature and contribute new insights to the field. Specifically, the study reveals a positive relationship between maternal mortality and GDP growth as well as a negative relationship between healthcare expenditures and GDP growth. These new findings contradict the initial hypotheses of the study. These findings emphasize the importance of prioritizing healthcare investment to reduce infant mortality and road traffic mortality rates in Saudi Arabia. By addressing these issues, policymakers can promote sustainable GDP growth. The implications of this study extend to policymakers in the health sector and potential investors, highlighting the need for strategic decision-making and investment in healthcare. Overall, this study provides valuable insights into the relationships between key factors and GDP growth in Saudi Arabia, offering guidance for policymakers and investors in the pursuit of sustainable economic development.

8. Limitations and Future Directions of the Study

This study has some limitations that may require further investigation in future studies. Firstly, the study was conducted solely in the context of Saudi Arabia, and the results may not be generalizable to other developing countries. Therefore, future research should be conducted in other developing countries using longer time series data and more advanced analytical methods to validate the conclusions. Secondly, this study relied solely on secondary data sources. Therefore, future studies could consider gathering primary data to investigate the effects of health factors on GDP growth. Lastly, this study only considered a limited set of health factors. Future studies could explore the impact of additional health factors, such as mental health and lifestyle factors, on GDP growth.