While research on remittance inflows and their positive contribution to economic growth in recipient countries is extensive, studies on remittance outflows (RMO) and their impact on sending countries' economic growth are relatively scarce, particularly focusing on the Gulf Cooperation Council (GCC) nations. This study addresses this gap by focusing on Saudi Arabia, a major sender of remittances. The prevailing view is that RMO negatively impacts a country's economic growth because it represents an outflow of income. However, the contribution of expatriates to the host country's GDP is often overlooked. Saudi Arabia, with a significant expatriate workforce, presents a compelling case to investigate this complex relationship. The study utilizes data from 1985 to 2019 to analyze the relationship between RMO and per capita GDP in Saudi Arabia. The research hypotheses posit that remittance outflows impede economic growth (null hypothesis) versus the alternative hypothesis that they do not. The study will employ the non-linear ARDL model, cointegrating regression techniques, and the vector error correction (VEC) Granger causality test to examine the relationship while controlling for trade openness, physical capital, human capital, and labor force.

Existing literature on the RMO-growth nexus is limited, primarily focusing on GCC countries. Studies show mixed results: Alkhathlan (2013) found a negative short-run impact but no significant long-run effect in Saudi Arabia. Kaabi (2016) found a negative impact only in Saudi Arabia among GCC countries. Hathroubi and Aloui (2016) found a positive short-run association between RMO and real GDP growth. Rahmouni and Debbiche (2017) found no significant impact of RMO on Saudi Arabia's EG. Khan et al. (2019) provided inconsistent results across GCC countries. Alsamara (2022) found a negative impact of RMO on Qatar's EG using NARDL but not with a simple ARDL model. The literature also shows inconsistent findings regarding the impact of trade, capital formation, human capital (measured by enrollment ratios), and labor force on economic growth. This study aims to contribute to the existing body of knowledge by utilizing a more advanced NARDL model to account for potential non-linear relationships and by including causality analysis, which many previous studies lacked.

The study utilizes time-series data from 1985 to 2019 obtained from the World Bank's World Development Indicators and Migration and Remittances Data. The dependent variable is per capita GDP (LnGDPC), representing economic growth. Independent variables include remittance outflow (LnRMO), total trade (LnTR), gross capital formation (LnGCF), tertiary enrollment ratio (LnTEN), and labor force (LnLF). To capture the asymmetric effects of RMO, the study decomposes LnRMO into positive (LnRMO_P) and negative (LnRMO_N) shocks using a partial sum procedure. The study employs a non-linear ARDL (NARDL) model to examine the asymmetric relationship between RMO and economic growth. The NARDL model is chosen because it can handle variables of different integration orders and is suitable for small sample sizes. A bounds test is conducted to confirm the existence of a long-run relationship among the variables. The study also employs canonical cointegrating regression (CCR) and fully modified ordinary least squares (FMOLS) to check the robustness of the NARDL results. Finally, a vector error correction (VEC) Granger causality test is performed to investigate the direction of causality among the variables.

The NARDL bounds test confirmed a long-run cointegrating relationship among the variables. The long-run results revealed that positive shocks of RMO do not significantly affect EG, while negative shocks (declines in RMO) positively impact EG. This suggests that a lower level of remittance outflow is beneficial for Saudi Arabia's economic growth. Trade openness (LnTR) had a positive and significant effect on EG, consistent with the country's historical reliance on trade. The labor force (LnLF) also significantly and positively contributed to EG. However, neither physical capital (LnGCF) nor human capital (LnTEN) significantly influenced EG. The long-run results with lag variables further confirmed the asymmetric impact of RMO on EG. The short-run NARDL model indicated that positive RMO shocks do not influence EG, while negative shocks have a mixed impact, with the overall effect likely being negative. The error correction term (ECT) in the short-run model was negative and significant, confirming the long-run relationship. Cointegrating regression (CCR and FMOLS) results supported the NARDL findings, confirming their robustness. The VEC Granger causality test showed unidirectional causality from EG to negative RMO shocks and unidirectional causality from trade to EG. Bidirectional causality existed between the labor force and EG.

The findings highlight the complex relationship between RMO and economic growth in Saudi Arabia. The positive effect of negative RMO shocks on long-run EG suggests that policies aimed at reducing RMO, such as attracting expatriate savings into domestic investment, could benefit the economy. The positive impact of trade and labor force on EG reinforces the importance of trade liberalization and human capital development. The insignificant effect of physical and human capital raises concerns about the efficiency of the Saudi labor force and the quality of education. The study's findings have significant policy implications for Saudi Arabia, suggesting a need for strategies to attract expatriate investment, improve labor productivity, and enhance education quality. These strategies should focus on improving the skills and efficiency of the Saudi labor force to reduce reliance on expatriates and consequently reduce remittance outflows.

This study reveals that reducing remittance outflows could positively influence Saudi Arabia's long-run economic growth. Trade openness and a productive labor force are crucial drivers of economic growth. However, the insignificant impact of physical and human capital highlights the need for substantial improvements in education and workforce development. Policy recommendations include creating an environment attractive to expatriate investment, promoting skill development among Saudi nationals, and improving the quality of education. Future research could benefit from using data disaggregated by nationality of the workforce, potentially allowing for the exploration of different impact channels.

A primary limitation is the use of aggregate labor force data, which includes both Saudi and expatriate workers. Disaggregated data would allow for a more precise analysis of the Saudi workforce's contribution to economic growth. Further research could explore the impact of specific policies aimed at retaining expatriate savings and improving the skills of the Saudi workforce. The study's findings are specific to Saudi Arabia and might not be generalizable to other countries with different economic structures and labor market dynamics.