Research Article |
Corresponding author: Osama D. Sweidan ( osweidan@uaeu.ac.ae ) © 2024 Non-profit partnership “Voprosy Ekonomiki”.
This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY-NC-ND 4.0), which permits to copy and distribute the article for non-commercial purposes, provided that the article is not altered or modified and the original author and source are credited.
Citation:
Sweidan OD (2024) The geopolitics of technology: Evidence from the interaction between the United States and China. Russian Journal of Economics 10(2): 130-150. https://doi.org/10.32609/j.ruje.10.118505
|
Recently researchers performed empirical economic studies to investigate how geopolitical risk impacts diverse economic sectors. We take a fresh perspective by exploring whether advancements in the U.S. IT sector can account for fluctuations in China’s geopolitical risk. The conflict between China and the United States regarding semiconductors revolves around technological supremacy, economic dominance, and national security concerns. China has been striving to become self-sufficient in semiconductor production to reduce reliance on foreign suppliers, particularly the United States. However, the United States has imposed restrictions on semiconductor exports to China. Our study constructs a theoretical framework and utilizes the bounds testing approach for cointegration to estimate the parameters of the Autoregressive Distributed Lag model. We use monthly data from January 1993 to November 2023. The findings reveal that the U.S. IT sector significantly and positively influences China’s geopolitical risk. From a policy implication perspective, the race to lead the global IT sector may emerge as the primary source of economic and political instability unless rival nations reach a compromise.
geopolitical risk, geopolitics of technology, information technology, ARDL model, United States, China.
Merriam-Webster dictionary defines Information Technology (IT) as “the technology involving the development, maintenance, and use of computer systems, software, and networks for the processing and distributing data.” IT has been widely used in all economic sectors. It helps these sectors to build, grow, and generate maximum output by enhancing total factor productivity (TFP). These benefits can be achieved by automating their processes, reducing work inefficiencies, provide accurate information and data about various commercial transactions. Many scholars, i.e.,
Nowadays, the emergence of new technologies, such as artificial intelligence, 5G telecommunications, and big data, is critical to economies, mainly to the financial services industry, supporting new products and services, and changing operational processes. More specifically, these new evolving technologies can help firms promote investment strategies and loan valuation, enhance human capital achievements, improve product development and customer relations via controlling data, strengthen cybersecurity and fraud detection, and expand mobile capabilities and services.
Currently, the IT international market has substantial competition between American companies, i.e., Apple, Alphabet, and Microsoft, and Chinese companies, i.e., Huawei, Tencent, and JD.com, to dominate the international IT market. This kind of rivalry between the two largest economies in the world may increase the geopolitical risk between the two nations. The famous incident of arresting Huawei CFO in Canada on December 1, 2018, is an excellent example of this rivalry and tension between the two countries because of the IT industry. The arrest was based on a legal charge made by the United States authority, who claimed that Huawei’s CFO is violating the United States sanctions against Iran and its technology, which is a real threat to U.S. national security. A few days later, China retaliated and arrested two Canadian citizens, a businessman and diplomat.
In the same vein,
The current study is motivated mainly by the fact that there is rivalry between the United States and China. We believe exploring this topic can enrich knowledge of geopolitical risk. There are minimal empirical studies conducted so far on the geopolitics of technology or using the IT sector as a determinant of geopolitics evolution. Thus, the current research aims to contribute to this new research strand and find evidence on whether the development of the U.S. IT sector can explain the fluctuations in China’s geopolitical risk. The general aim of the current research is to examine the factors that initiate it. In a recent work,
Throughout the history, scientific and technological collaboration has been vital to the U.S.– China economic relations. Following the Cold War, the process of globalization facilitated the integration of China’s abundant low-skilled labor into the operations of the United States and Western firms. Consequently, they shifted the lower and middle segments of industrial value chains to the Chinese mainland while maintaining strict controls over technology exports to China. This strategy facilitated China’s integration into the global economic framework and industrial networks to influence its political trajectory according to their preferences (
Against the abovementioned desire, China built a strong economy, including a significant IT sector. It converted China into a geopolitical competitor rather than a partner in the United States hegemonic plan (
Recently, the United States has announced a series of restrictions on tech exports, mainly chips and semiconductors, to China.
Critical raw materials are the materials that are vital to economic growth, job creation, and improving quality of life. Besides, there is a high risk associated with their supply because it is concentrated in particular countries. As a subset of critical raw materials, rare earth materials are a relatively abundant group of seventeen elements and are considered substantial to technological advances (
Country | 2016 | 2017 | 2018 | 2019 | 2020 | 2021 | 2022 |
Australia | 15.0 | 19.0 | 21.0 | 20.0 | 21.0 | 24.0 | 18.0 |
Brazil | 2.2 | 1.7 | 1.1 | 0.7 | 0.6 | 0.5 | 0.08 |
Burma | NA | 5.0 | 19.0 | 25.0 | 31.0 | 35.0 | 12.0 |
China | 105.0 | 105.0 | 120.0 | 132.0 | 140.0 | 168.0 | 210.0 |
India | 1.5 | 1.8 | 2.9 | 2.9 | 2.9 | 2.9 | 2.9 |
Russia | 2.8 | 2.6 | 2.7 | 2.7 | 2.7 | 2.6 | 2.6 |
United States | NA | NA | 18.0 | 28.0 | 39.0 | 42.0 | 43.0 |
World production | 129.0 | 132.0 | 190.0 | 220.0 | 240.0 | 290.0 | 300.0 |
Moreover, due to the dynamic feature of the international market on technology advances, some raw materials that are currently not identified as critical might become critical when new data or demand becomes available. Recently,
Arsenic | Beryllium | Borate | Cadmium | Cerium |
Cobalt | Cooper | Europium | Fluorspar | Gadolinium |
Gallium | Gold | Hafnium | Indium | Iridium |
Lanthanum | Lutetium | Manganese | Natural rubber | Neodymium |
Palladium | Platinum | Praseodymium | Rhodium | Ruthenium |
Samarium | Selenium | Silicon metal | Silver | Terbium |
Tin | Tungsten | Vanadium |
As stated above, the geopolitics of IT is a relatively new research area. There are no solid theoretical or empirical works that extensively cover this topic. This part, therefore, introduces the definition of geopolitical risk, its importance, and why IT can generate such a risk. The available literature stresses the critical role of IT in the future of various nations.
Recently, a growing body of literature has focused on investigating the influence of geopolitical risk on various economic sectors and indicators. For instance, several empirical works concluded that geopolitical risk has a significant impact on the energy sector (
In their work,
Literature shows a critical gap in the current empirical works of understanding the determinants of geopolitical risk. Most of the empirical works in this research area looked at geopolitical risk as an exogenous variable, and focused on exploring its influence on various economic sectors and financial indicators. Very few empirical works explored the opposite research direction. In a recent study, Sweidan (2023d) examined whether the United States major macroeconomic indicators influence international geopolitical risk. He found a statistically significant effect of the United States macroeconomic variables on international geopolitical risk. Khan et al. (2022) inspected the causality between international geopolitical risk and technology growth. They concluded a bidirectional causality between the two variables.
So, why is the IT sector important? Traditional economic theories, i.e., neoclassical growth and endogenous growth theories, explain technology’s importance in generating economic growth. Technology and knowledge accumulation are significant components to boost production because they can increase labor skills and productivity, thus enhancing human capital and economic growth. Many well-known studies empirically proved this conclusion, i.e.,
The regulatory models of the IT system in the United States and China are different. In the former, the government’s role is secondary compared to the leading role of the private sector. Thus, the government’s mission is to encourage IT innovations rather than restrict them. In the latter, the IT system works in an environment where domestic companies dominate in a silo internet environment, adopt a mass-surveillance policy, and the government serves as the crucial organizing node (
According to Abishur Prakash,
Within these significant functions of IT and its new generations, governments are expected to launch new lines and public policies to enable data sovereignty. Alternatively, information gathered online by IT companies will be subject to the country’s laws where it is collected or processed and remains within its borders. Undoubtedly, those who own the advanced technology and can develop new generations of this technology, will dominate the world’s economic and military powers and the earth’s atmosphere. Based on these facts, we expect fierce rivalry among nations leading the world of technology, generating permanent geopolitical risks among the competing nations.
The current study’s empirical model assumes that there are four factors that can influence China’s geopolitical risk index (GPRCt). These four explanatory factors cannot be controlled by the Chinese government, implying that they are exogenous to the Chinese economy. The first and core independent variable is S&P U.S. Information Technology Index (USITt). It tracks the stock performance of the largest 500 companies listed in the United States stock exchanges. Alternatively, it shows the investments in the United States information technology sector. We think that USITt is a relevant and significant indictor to the Chinese, as rivals, and also to the rest of the world to know more about progress in the U. S. IT sector, and hence react to these developments accordingly. The other three determinants are extracted from the literature and previous studies as potential determinants of China’s geopolitical risk and are as follows: crude oil prices (OPt), the international geopolitical risk index (IGPRt) and the relative importance of the United States’ competitiveness index to China’s competitiveness index (CIRt).
GPRCt = F (USITt, OPt, CIRt, IGPRt). (1)
Equation (1) can be written in a linear regression form as follows:
GPRCt = α0 + α1USITt + α2OPt + α3CIRt + α4 IGPRt + Ut, (2)
where α0, α1, α2, α3 and α4 are the model’s coefficients; Ut is a white noise error term. Logically, the signs of the three parameters (α0, α1, α2, α3, α4) are expected to be positive. Fig.
Our study extracts the data from four primary sources. GPRCt is used from
The variables’ descriptive statistics and correlation coefficients (N = 371).
Variable | LnGPRCt | LnUSITt | LnOPt | LnCIRt | LnIGPRt |
Mean | –0.89 | 6.16 | 3.79 | 100.22 | 4.53 |
Std. dev. | 0.53 | 0.90 | 0.63 | 3.17 | 0.35 |
Min | –2.47 | 4.30 | 2.42 | 94.89 | 3.67 |
Max | 0.60 | 8.07 | 4.90 | 108.91 | 6.24 |
LnGPRCt | 1.00 | ||||
LnUSITt | 0.65 | 1.00 | |||
LnOPt | 0.48 | 0.52 | 1.00 | ||
LnCIRt | –0.36 | –0.52 | –0.66 | 1.00 | |
LnIGPRt | 0.38 | 0.04 | 0.14 | 0.12 | 1.00 |
Variable | LnUSITt | LnOPt | LnCIRt | LnIGPRt |
VIF | 1.49 | 2.06 | 2.06 | 1.11 |
At this point, this study began estimating the ARDL model by investigating potential structural breaks within China’s geopolitical risk, visually represented in Fig.
Sequential F-statistic determined breaks: 0 Significant F-statistic largest breaks: 0 |
|||
Break test | F-statistic | Scaled F-statistic | Critical value** |
0 vs. 1* | 2.08 | 27.01 | 27.03 |
1 vs. 2 | 2.01 | 26.01 | 29.24 |
2 vs. 3 | 1.04 | 13.54 | 30.45 |
3 vs. 4 | 1.20 | 15.54 | 31.45 |
4 vs. 5 | 0.00 | 0.00 | 32.12 |
Estimated break dates: 1: 2001M09 2: 2001M09, 2016M11 3: 1998M06, 2003M12, 2016M12 4: 2001M09, 2006M10, 2011M07, 2016M12 5: 1998M06, 2002M11, 2007M06, 2011M11, 2016M12 |
The current work uses the bounds testing technique for cointegration as an instrument to approximate the Autoregressive Distributed Lag (ARDL) model parameters. This approximation process was developed by Pesaran et al. (2001). It inspects the presence of a long-run relationship among the model’s variables. It estimates short-run and long-run coefficients, and the speed of adjustment or the error correction term (ECMt) toward the long-run equilibrium. This estimation method can be applied regardless of the variables’ integration orders. It is applicable if the variables of I (0) or I (1) or a grouping of them, but not I (2). In addition, the ARDL technique performs well on small sample size.
The general specification form of the ARDL (p, q) is as follows:
, (3)
where Qt denotes the dependent variable; Rt represents a list of independent or explanatory variables; δ, α, and β are the model’s assessed parameters; εt is the random disturbance. Equation (2) can be re-written to match the ARDL model in Equation (3) and as follows:
, (4)
where ∆ stands for the first difference operator. The coefficients α1 to α5 indicate the short-run parameters in Equation (4), while the long-run coefficients are denoted by the coefficients β2 and β5 normalized by the parameter β1. Through the process of estimating Equation (4) parameters, this study imposes a maximum of 12 lags via the automatic selection option. Besides, it conducts diagnostic and stability tests to inspect the robustness of the results. It also employs two methods of the ARDL methodology to investigate the existence of a cointegration relationship among the variables. First, it computes and compares the upper and lower critical F-values of Pesaran et al. (2001). If the computed F-statistic is less than the lower bound critical values, the null hypothesis of no cointegration cannot be rejected. On the contrary, if the calculated F-statistic is higher than the upper bound critical values, the null hypothesis can be rejected, and the long-run relationship exists. The results are indecisive if the computed F-statistic is in-between the upper and lower bound critical values. Second, ECMt is estimated and placed in the model instead of the long-run variables. If the computed coefficient of ECMt is statistically significant, negative, and less than one, then the long-run relationship among the model’s variables exists.
The prevalence of unit root challenges in economic time series data is widely acknowledged. To ensure precision, we precisely scrutinized our data for any indication of unit root presence before employing the ARDL bounds testing method for cointegration. This crucial step aids in selecting the appropriate time series model and diminishes the likelihood of erroneous regression. Our examination involved four unit root tests — two standard tests: Augmented Dickey–Fuller (ADF; Dickey, Fuller, 1981) and Phillips–Perron (PP; Phillips, Perron, 1988), and two tests designed for accommodating structural breaks (ADF and PP). Across all tests, the null hypothesis implies that a unit root impacts the series. The findings in Table
Variable | Standard | Structural Breaks | |||||||||
ADF | PP | ADF | PP | ||||||||
Level | 1st diff. | Level | 1st diff. | Level | 1st diff. | Level | 1st diff. | ||||
LnGPRCt | –8.03*** | – | –12.24*** | – | –12.34*** | – | –12.22*** | – | |||
LnUSITt | –1.54 | –15.04*** | –1.65 | –15.18*** | –3.41 | –15.71*** | –3.41 | –7.15*** | |||
LnOPt | –2.89 | –14.82*** | –2.52 | –14.37*** | –5.48*** | – | –5.48** | – | |||
LnCIRt | –1.74 | –18.39*** | –1.71 | –18.38*** | –3.62 | –41.25*** | –3.68 | –18.46*** | |||
LnIGPRt | –6.60*** | – | –6.26*** | – | –7.90*** | – | –5.93*** | – |
The current study estimates an ARDL model over the main sample (01/1993–11/2023) and two sub-samples (01/2000–11/2023 and 01/1993–11/2017).
Cointegration hypotheses | Time interval | F-statistics |
LnGPRCt = F (LnUSITt, LnOPt, LnCIRt, LnIGPRt) | 01/1993–11/2023 | 13.139*** |
LnGPRCt = F (LnUSITt, LnOPt, LnCIRt, LnIGPRt) | 01/2000–11/2023 | 9.379*** |
LnGPRCt = F (LnUSITt, LnOPt, LnCIRt, LnIGPRt) | 01/1993–11/2017 | 7.473*** |
Throughout estimating the ARDL models and the cointegration tests, this study imposes a maximum of 12 lags. The ARDL model’s results of the main sample are presented in Table
Variable | Coefficients | Standard errors |
(A) Short-run parameters | ||
Constant | –2.302** | 0.931 |
∆LnGPRCt–1 | –0.120** | 0.048 |
∆LnUSITt | 0.865*** | 0.298 |
∆LnOPt | 0.082** | 0.040 |
∆LnCIRt–2 | 0.081*** | 0.031 |
∆LnCIRt–4 | 0.059* | 0.031 |
∆LnCIRt–9 | 0.091*** | 0.031 |
∆LnIGPRt | 0.680*** | 0.078 |
(B) Long-run parameters | ||
Constant | –4.693*** | 1.800 |
LnUSITt–1 | 0.293*** | 0.050 |
LnOPt–1 | 0.167** | 0.080 |
LnCIRt–1 | –0.009 | 0.016 |
LnIGPRt–1 | 0.495*** | 0.110 |
ECMt– 1 | –0.491*** | 0.060 |
(C) Diagnostics tests | Probability | |
Adj. R2 | 0.421 | |
Jarque–Bera | 1.291 | 0.616 |
LM – Stat. (BG-test), F (12, 330) | 0.834 | 0.939 |
Heteroskedasticity (BPG-test), F (18, 342) | 0.675 | 0.836 |
Heteroskedasticity (ARCH-test), F (1, 358) | 0.300 | 0.584 |
Ramsey RESET (F-test), F (5, 337) | 1.099 | 0.361 |
CUSUM | Stable | |
CUCUMSQ | Stable |
Figs
A sample of the CUSUM test from the main sample, 01/1993–11/2023.
Source: Author’s calculations.
A sample of the CUSUMSQ test from the sub-sample, 01/2000–11/2023.
Source: Author’s calculations.
In the short run, the results of Table
Variable | Coefficients | Standard errors |
(A) Short-run parameters | ||
Constant | –2.783*** | 1.020 |
∆LnGPRCt–1 | –0.175*** | 0.052 |
∆LnUSITt | 1.158*** | 0.342 |
∆LnOPt | 0.102** | 0.051 |
∆LnCIRt | 0.266*** | 0.084 |
∆LnCIRt–1 | –0.164** | 0.083 |
∆LnIGPRt | 0.690*** | 0.081 |
(B) Long-run parameters | ||
Constant | –6.194*** | 2.113 |
LnUSITt–1 | 0.318*** | 0.058 |
LnOPt–1 | 0.227** | 0.109 |
LnCIRt–1 | 0.001 | 0.017 |
LnIGPRt–1 | 0.512*** | 0.127 |
ECMt– 1 | –0.449*** | 0.059 |
(C) Diagnostics tests | Probability | |
Adj. R2 | 0.451 | |
Jarque–Bera | 0.647 | 0.724 |
LM – Stat. (BG-test), F (12, 264) | 0.781 | 0.670 |
Heteroskedasticity (BPG-test), F (10, 276) | 0.871 | 0.561 |
Heteroskedasticity (ARCH-test), F (1, 284) | 0.105 | 0.747 |
Ramsey RESET (F-test), F (5, 271) | 0.510 | 0.769 |
CUSUM | Stable | |
CUCUMSQ | Stable |
Variable | Coefficients | Standard errors |
A) Short-run parameters | ||
Constant | –2.150** | 1.050 |
∆LnGPRCt–4 | –0.090* | 0.054 |
∆LnUSITt | 0.926*** | 0.329 |
∆LnOPt | 0.089** | 0.043 |
∆LnCIRt–2 | 0.078** | 0.032 |
∆LnCIRt–4 | 0.066** | 0.032 |
∆LnIGPRt | 0.686*** | 0.087 |
(B) Long-run parameters | ||
Constant | –4.176*** | 2.113 |
LnUSITt–1 | 0.237*** | 0.074 |
LnOPt–1 | 0.173** | 0.079 |
LnCIRt–1 | –0.009 | 0.016 |
LnIGPRt–1 | 0.445*** | 0.111 |
ECMt– 1 | –0.515*** | 0.076 |
(C) Diagnostics tests | Probability | |
Adj. R2 | 0.426 | |
Jarque–Bera | 0.1.271 | 0.530 |
LM – Stat. (BG-test), F (12, 265) | 0.722 | 0.730 |
Heteroskedasticity (BPG-test), F (16, 277) | 0.800 | 0.686 |
Heteroskedasticity (ARCH-test), F (1, 291) | 0.026 | 0.873 |
Ramsey RESET (F-test), F (5, 272) | 1.354 | 0.242 |
CUSUM | Stable | |
CUCUMSQ | Stable |
In summary, our study demonstrates that the evolution of the U.S. IT sector poses a geopolitical risk for China. This result particularly aligns with the findings of the limited empirical works in this research strand. For instance,
Moreover, Tables
A large body of growing empirical works investigated the effect of geopolitical risk on various economic sectors and financial variables over the past six years. These empirical studies are motivated by the geopolitical risk index established by
The preliminary indicators and theoretical background show that some superpower countries, mainly the United States and China, have a severe rivalry in the IT sector. Over the past three decades, the Chinese economy has experienced remarkable real economic growth, reaching a rate of 8.75%, compared to 2.45% for the United States economy. These figures underscore China’s rapid progress and potential to emerge as a major competitor to the United States economy. Additionally, China has made significant strides in accessing and excelling in the IT sector. Strategically, China is the largest supplier of critical raw materials essential for electronics manufacturing. Data reveals that China possesses five times more rare earth materials than the United States, which are crucial for producing essential electronic components like chips. However, while the United States leads in the production of these components, commanding a 48% international market share compared to China’s 8%, this gap might be more significant if we consider the United States allies, such as South Korea, Taiwan, and Europe. This economic dynamic has generated geopolitical tensions and economic rivalry as both nations vie for control over the international IT sector and the global economy.
Technology is a significant core component of the supply-side economy (production functions), as explained in the various economic growth theories. Nowadays, the nation that owns, produces, and innovates new generations of IT will dominate the international economy and military power, and control the earth’s atmosphere. Undoubtedly, our traditional life has become dependent on technology, mainly with the COVID-19 pandemic. These changes generate the concept of Big Data as another new strategic element that the superpower countries have control over.
The current study estimates an ARDL model over three samples to guarantee robust results. The current study successfully extracts empirical proof of the geopolitics of technology between the United States and China. It confirms that the progress in the U.S. IT sector can explain the movement of China’s geopolitical risk. The three estimated ARDL models show that this effect is statistically significant and positive either in the short or long run. This conclusion bears a crucial policy implication: the arena of global geopolitics among superpowers is influenced by the IT sector and other macroeconomic indicators. Cultivating a robust IT sector is pivotal for asserting dominance in the global atmosphere and economy. Undoubtedly, there is fierce competition among global superpowers to lead the next wave of IT advancements. It could become a significant source of economic and political instability if compromises are not reached. Our study anticipates a substantial surge in global geopolitical risk due to the heightened tensions among nations worldwide. It is motivated by the resurrection of an international multipolar system and the intensified race in the IT sector.
The author would like to thank the editor of the Russian Journal of Economics and three anonymous referees for their valuable and helpful comments. The author is responsible for any remaining errors.
The geopolitics of technology
Data type: Table
Explanation note: This data is used to conduct the study. A description of the data exists in the study.