“The export-economic growth nexus: The case of Saudi Arabia”

Export strategies are the means by which the country can dispose of its domestic production surpluses to bridge successive deficits in national balances of payments and achieve economic growth. These strategies are of particular importance to the economy of Saudi Arabia, as it has opted in the last decade to diversify its economy and migrate gradually away from an economy heavily reliant on oil exports. Given the importance of diversifying the economy, this study aims to examine the relationship be-tween exports and economic growth in the Saudi Arabian economy. The multivariate Granger Causality Test and cointegration, which is the most common model, was used in examining the short-term and long-term patterns of exports, non-oil exports, GDP, GDP per capita, and government spending from 1991 to 2016. The findings support a long-standing connection involving Saudi exports and the country’s rate of economic expansion. Unidirectional causality exists between exports, non-oil exports


INTRODUCTION
Export-led economic growth is one of the strategies many economies adopt, especially emerging ones. Recognizing this reality, relevant economic policies require measures to ensure the proper allocation of resources, technological transformation, accumulation of foreign exchange, generation of government revenues, capital formation, and employment. Developing non-oil exports is a vital requirement often underscored by most countries producing and exporting primary oilbased materials. The importance of this strategy comes from providing the foreign reserves necessary to finance the development of many programs (e.g., paying the costs of needed imports of products and services, increasing investments, and reducing unemployment), especially when there is a decline in the revenue of raw materials. Furthermore, this strategy represents a tool by which the countries can utilize the surpluses in their domestic production, which will lead to the expansion of the market, achieving the economic production level.
Regarding oil resources, outputs, exports, and refining capabilities, Saudi Arabia is regarded as the most powerful petroleum-producing nation in the world. It accounts for above 20% of international oil sales, 12% of international oil production, 19% of international oil reserves, and more than five million barrels per day of internal and external refining capability (SM Energy, 2021). The blueprint of Saudi Vision 2030 (Government of Saudi Arabia, 2016) highlights and demands a national strategy for developing non-oil exports in various sectors and segments (e.g., investment sector, employment, and foreign franchises) to enable Saudi enterprises to enter foreign markets and contributes to its economic growth. Thus, it is necessary to conduct a study that analyzes the causal relationship between the increase in non-oil exports and the improvement of some macroeconomic variables. With the succession of global economic crises, diversification from an oil-driven economy is vital for economic growth and sustainability.

LITERATURE REVIEW AND HYPOTHESES DEVELOPMENT
The Saudi economy has survived a number of transformations throughout the last two decades. The last is the Saudi Vision 2030, representing a major economic transformation phase and a new launch of tomorrow's economy (Government of Saudi Arabia, 2016). During the last five years, the Saudi government has implemented many structural reforms and programs in several sectors to diversify the Saudi economy's production base. These reforms are in consonant with one of the main objectives of the Saudi Vision 2030: enhancing the Saudi economy to be among the top 15 largest economies in the world after being ranked 19th and increasing the localization of oil and gas sectors from 40% to 75%. The vision also aims to diversify the Saudi economy by expanding investment in the non-oil sector, enhancing the potential of promising economic sectors, and privatizing several government services. The Saudi Vision 2030 embraces a number of economic and social strategic goals, most relevant to this study are: • raising the contribution of foreign direct investment in GDP from 3.5% to 5.7%; • increasing the contribution of the private sector to GDP from 40% to 65%, and • increasing the government's non-oil revenues from USD 45 billion to USD 270 billion.
There are specific initiatives of the Saudi government, such as the Saudi Fund for Development, which has contributed to enhancing, financing, and ensuring exports, along with diversifying sources of national income, primarily through the advancement of non-oil exports and improving their competitiveness in line with Vision 2030. Another initiative is the Saudi Export Development Authority, which plays a pivotal role in conducting studies and developing plans that reduce the exporters' challenges. In addition, the Authority participates in international events and commercial missions to market Saudi national products. It provides support by organizing workshops to improve Saudi organizations' abilities and their experiences in the field of export. Lastly, the National Policy for the Advancement of Nonoil Exports is conducting development programs in collaboration with the relevant authorities. As a result of these initiatives in the past five years, the percentage of non-oil exports exceeded 20% of the total exports (SAMA, 2020). The Saudi Vision 2030, through its various programs, works toward increasing non-oil exports to 50% of non-oil GDP (Government of Saudi Arabia, 2016).
Studies investigating the relationship between trade openness and economic growth were conducted specifically along the nexus between exports and growth, the nexus between imports and growth, and the nexus between trade and growth (Agrawal,  Munir and Javed (2018) examined how the diversification of exports affected the economic development of South Asian nations using the COP Douglas function. In this study, the diversification of exports is classified into horizontal and vertical diversification. Accordingly, the Hervindal Index demonstrated an inverted U-shaped link between economic growth and inflation. Although export diversification has a significant influence on economic growth above a certain threshold, export diversification initially resulted in stronger economic growth. The initial economic growth benefits of horizontal export diversification are minimal. Nevertheless, once the threshold level has been reached, adding a new industry boosts economic growth in South Asian nations.
The trade-led growth hypothesis was extensively studied (Çoban et al., 2020;Rahman & Mamun, 2016). The BRICS economies could be seen as the world's first commercial bloc in one sense and as the strongest developing economies in another. Due to this, Raghutla and Chittedi (2020) investigated the trade-led growth, exports-led growth, growth-led imports, and the import-led growth hypotheses employing the Granger Causality Assessment for causality direction and the Johansen cointegration methodology for causality in the long term among BRICS economies. Accordingly, the output revealed that the export-led growth hypothesis was relevant for Brazil and Russia. In contrast, the growth-led exports hypothesis applies to China, India, and South Africa. As for South Africa, Brazil, China, and India, the growth-led imports hypothesis seems more relevant, whereas the import-led growth hypothesis is more applicable in the context of Russia.
In the same context, Oyelade (2019) investigated the trade-led growth and the export-led growth hypotheses in selected African countries (Nigeria, Gambia, Ghana, Guinea, Liberia and Sierra Leone). The estimated findings supported the import-led growth hypothesis for Nigeria, Guinea, and Liberia, the trade-led growth hypothesis for Gambia and Sierra Leone, and the export-led growth hypothesis for Ghana. The study emphasized imports as the most important factor influencing economic growth over exports.
Studies investigating the export-growth nexus in Saudi Arabia are scarce. Among them, Islam (2021) used time-series annual data for 1985-2019 to assess the trade-led growth theory in the Saudi Arabia context. The ARDL technique and Toda-Yamamoto Granger Causality Test have been used to analyze the data. The trade-led growth hypothesis is confirmed for Saudi Arabia based on the results of the ARDL estimation, which show that trade openness fosters productivity development in both the long and the short term. Trade openness is a cause of gross fixed capital formation.
Furthermore, the labor force increases commerce volume and economic growth. Similarly, Alshahrani and Alsadiq (2014) confirmed that government spending stimulates growth in the long run in Saudi Arabia. In addition, it was noted that openness to trade and spending in the housing sector can also boost short-run production. Waheed et al. (2020) found that non-oil exports positively affect the economic growth of Saudi Arabia. They proposed that increasing non-petroleum exports may be a sound plan for long-term expansion and as a substitute for petroleum-based products. This result was previously proposed by Ouassaf and Kouidri (2007).
In the same context, a panel geographic autoregressive model for 77 trade partners through the period of 2000-2016 is employed by Gouider et al. (2020) to investigate the possible regional diversity of industrial goods exports in Saudi Arabia. Observationally, the findings showed a geographical association between the exports of Saudi manufactured goods and those external factors, which include GDP, per capita GDP, freedom of trade, a trade intensity indicator, and the bilateral exchange ratio. Based on that, Cherikh and Karagiannis (2019) To summarize, many empirical models have been utilized. These models include several explanatory variables, in addition to exports, that are believed to influence the economic growth level, such as imports, technology, human capital, and physical capital.
Therefore, the purpose of this study is to investigate the existing relationship between exports and economic growth in the Saudi economy. The paper first tests the unidirectional causality between exports, non-oil exports, and economic growth expansion (GDP, GDPC, and GS), which means the growth rate rises as exports grow. Second, it assesses the bidirectional causality between the variables using the multivariate Granger causality test and cointegration to examine the long-term and short-term patterns of exports (EXP), non-oil exports (NoEXP), GDP, GDP per capita (GDPC), and government spending (GS) from 1991 to 2016.
Based on the literature review, the following are hypothesized: H1: There is a causal association between the development of exports and the growth of Saudi economics.
H1.1: A long-term causal relationship is directed from the improvement of exports to the Saudi economic growth.
H1.2: There is a short-term causal relationship between GDP and the total exports to GDP and non-oil exports to total exports.

H1.3:
There is a short-term causal association between GDP per capita and the total exports to GDP and non-oil exports to overall exports.
H1.4: There is a short-term causal association between government spending and the total exports to GDP and non-oil exports to overall exports.

METHODOLOGY
This study used the export-led growth model to choose the best variables. Table 1 shows the variables utilized in this study, their description, and their operationalization. GDP is the total monetary or market value of all the finished goods and services produced within a country's borders in a specific time period. As a broad measure of overall domestic production, it functions as a comprehensive scorecard of a given country's economic health.

Gross Domestic Product Per Capita (GDPC)
The GDPC is the Gross domestic product divided by population. LN GDPC Government Spending (GS) Government spending refers to the money that the government spends on buying goods and provision of services such as education, healthcare, social protection, ...

Independent variables
Total Exports to GDP (Exp) Input of total exports to GDP Exp/GDP Non-oil Exports to Overall Exports (NoExp) Input of non-oil exports to total exports NoExp/Exp

Descriptive and correlation analysis
The empirical analysis started with displaying the descriptive values of all the variables. Based on  Table 3 shows the correlation coefficient matrix among variables. It is observed that the dependent variables (GDP, GDPC, and GS) are at most positively correlated and also statistically significant with the independent variables (EXP and NoEXP). Note: * indicates the significant level at 0.05 (2-tailed); ** indicates the significant level at 0.01 (2-tailed).

Augmented Dickey-Fuller test
This test proposes a modification to the test with some extra mutations so that the dependent variable cancels out the autocorrelation. How long the decomposition is in all three conditions is marked by the Schwartz Bayesian Criterion (SBC), the Akaika Information Criterion (AIC), or Lagrange Multiplier. These three possible conditions have the equations below: • Test for a unit root: • Test for a unit root with constant and deterministic time trend: where y t -is the variable, t -is the time index, Δis the first factor of difference (change), u t -is the error term, α -is the constant, β -is the Marginal propensity.

Phillips-Perron test
The suggestion given by the Augmented Dicky-Fuller Test is that the mistakes possibility is independent and has a constant variation. So, by relating to the Augmented Dicky-Fuller, it is necessary to ensure that the possibility of the mistakes is unrelated and has a constant variation. However, the Philips-Perron strategy permits autocorrelation at the error limit. The strategy of Philip Peron is to adjust the t-Dicky-Fuller statistics to contemplate how limited the errors are. As shown in Table 4, the ADF and PP tests reveal that the variables are non-stationary. After the first difference, they were made stationary.

Pairwise Granger causality test
The pairwise Granger causality is used to test a causal relationship in the short term between two variables. It depends on the vector auto-regression (VAR) model for the first difference between the two variables, which was needed to test their causality. In the case of the two variables, EXP and GDP, the following equations are given: ( )  (5), ε -the random change.
The null hypothesis in Granger causality test is: (H_0) there is no causality between two variables; therefore, the rejection of (H_0) implies the causality between the two variables. The results reject the null hypothesis that exports do no Granger cause economic growth at a 1% significance level.  Table 5 shows a causality association between the two variables as the p-value of the F-statistic is less than 0.05. Therefore, there is a positive and statistically significant association between EXP → GS (Unidirectional Causality), GDP→ GS (Unidirectional Causality), GDPC → GS (Unidirectional Causality), and NoEXP→ GS (Unidirectional Causality).

Long-term association between the variables
Two methods are used to test the causality relationship in the long term: the cointegration test and the Toda and Yamamoto causality test. For both tests, there is a need to establish the appropriate number of lags for the VAR model. Table 6 reveals the results of the optimum lags selection.
According to the different methods used to select VAR lag, as shown in Table 6, the suitable number of lags to the VAR model can be obtained by depending on the lag which attains most of the selection criteria; therefore, the appropriate number of lags is 3 (three). Table 7 shows the results of the Johansen cointegration test for the EXP and NoEXP (independent variables) and GDP (dependent variable). The results of the Johansen cointegration test for independent variables EXP and NoEXP and the dependent variable GDPC are shown in Table 8. Finally, Table 9 shows the results of the Johansen cointegration test for independent variables EXP and NoEXP and the dependent variable GS. Note: * All p-values are less than 0.05, indicating a cointegration between the variables used in the model.

Toda and Yamamoto causality test
When using the Toda and Yamamoto Causality with a VAR Model with lags = 3 and stationarity at 1st difference, the Vector Autoregression (VAR) estimates are as shown in Table 10.

VAR Granger Causality/Block Exogeneity Wald tests
According to the data shown in Table 11, the value of the probability of the Chi-square test (p-value) is lower than 0.05, indicating a Bi-directional causality between all variables in the study included in sub-hypothesis H1.1 to H1.4, Therefore, all sub-hypothesis are accepted and consequently, the main hypothesis is accepted.   Moving on to the correlation analysis, it is observed that the dependent variables, namely GDP, GDPC, and GS, are positively correlated. This suggests that increased total exports, non-oil exports, and government spending are associated with higher GDP and GDPC. These findings are consistent with the main hypothesis of the study, which posits a causal relationship between export development and Saudi economic growth.
Furthermore, the correlation matrix reveals that the independent variables, EXP and NoEXP, positively correlate with GDP, GDPC, and GS. This supports the sub-hypotheses that propose shortterm causal relationships between GDP, GDPC, and government spending with both total exports and non-oil exports. These results indicate that an increase in exports, particularly non-oil exports, contributes positively to economic growth, GDP per capita, and government spending. The positive relationship between exports and economic growth has been widely acknowledged in developed and developing countries. In Saudi Arabia's case, the export base's diversification and the emphasis on non-oil exports, as outlined in the Saudi Vision 2030, have the potential to enhance economic growth and reduce dependence on oil revenue.
To summarize, the data analysis results support all the study hypotheses, indicating a causal relationship between export development and Saudi economic growth. The positive correlations between total exports, non-oil exports, GDP, GDPC, and government spending highlight the significance of diversifying the export base and increasing non-oil export contributions to the economy. These findings provide valuable insights for policymakers and stakeholders in shaping economic strategies and promoting sustainable growth in Saudi Arabia.

CONCLUSION
This study aims to assess at the dynamic causal link between Saudi exports and economic development from 1991 to 2016. The empirical findings revealed the presence of a causal association between Saudi exports' development and its economic growth in the long term. There is unidirectional causality between exports, non-oil exports, and government spending. Also, the study revealed unidirectional causality between GDP, GDP per capita, and government spending. Therefore, decision-makers should focus more on reducing dependency on oil export revenues in financing development programs, emphasizing non-oil sector export revenues, strengthening the private sector, and encouraging foreign direct investment in non-oil sectors.
Planning to enhance and improve the non-oil exports needs to have goods and services that include high-added value, evolve the outcomes and competition of the economic base, and with the help of the private sectors, this can grow the openness to foreign investment and trade. Furthermore, these results can stretch out important information for the economic policymakers in Saudi Arabia regarding supporting and improving non-oil exports and their relationship to fulfill economic growth.
As in other studies, this study has some limitations. First, the period of this study included some issues like the 2008 financial crisis and the oil price decline in 2014 and 2015, which may impact the results of this study. Thus, future studies could employ a more extended period to better capture the situation. Second, this study focused only on the impact of total export on GDP and non-oil exports on overall exports. Future studies may examine further factors like national and foreign investments. Finally, a revisited examination of the topic could be useful for future decision-making related to the accomplishment of Saudi Vision 2030.