Indexed in: KCI, ESCI, SCOPUS
search for


Can Digital Authoritarianism Deter Political Freedom?: Innovation in Digital Technology and Democratization
The Korean Journal of International Studies 20-1 (April 2022), 21-53
Published online April 30, 2022
© 2022 The Korean Association of International Studies.

Yongjae Lee [Bio-Data]
Received January 31, 2022; Revised March 13, 2022; Accepted March 21, 2022.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License ( which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Current works on the relationship between digital technologies and democracy focus on the rate of the internet or mobile phone users as an independent variable. These variables are not enough for reflecting on the innovativeness in digital technologies. This paper explores whether or not the creation of digital technologies in authoritarian states prevents democratization by using patent data. This study expects that innovative non-democratic states in digital technologies successfully deter democratic movements because digital technologies enable governments to achieve economic development, detection of anti-governmental activities, manipulation of information, decrease of indiscriminate suppression, and application of imported technologies. However, empirical analyses found that only innovative authoritarian states with strong censorship, that generate a greater number of patents of digital technologies can successfully deter democratization.
Keywords : Digital Technology, Democratization, Patent, China, Authoritarianism

With the development of ICT (information communication technology), its influence in politics are very disputable. According to conventional perspectives on technological innovation, improvement in communication technologies promotes democratization by spreading information and accelerating collective behaviors. However, technologically innovative authoritarian states, such as China and Russia are reinforcing political control toward their citizens with innovative new technologies, such as artificial intelligence and a facial recognition technology these days. This research project attempts to answer whether technological innovation hinders democratization and reinforces an authoritarian regime’s control by analyzing patent data measuring technological innovativeness.

Previous work on the relationship between technology and democracy concentrating more on the increase in use of social media and the internet than the creation of technologies did not fully reflect technological innovativeness of authoritarian states. This assumes that the innovativeness of digital technologies is more essential in political liberalization than the mere usage of technologies. Non-democratic governments having original technologies can more easily and effectively use them in controlling public freedom like China and Russia without foreign dependence and sanctions. However, authoritarian states importing digital technologies may suffer from various technological difficulties in usage and operation of the technologies for censorship and surveillance. Empirical results found that innovative authoritarian states generating a greater number of patents (log) of digital technologies with a stronger censorship can limit collective actions for democratization. Fixed effect analyses report that non-democratic states inventing new technologies with greater GDP per capita (log) suffer from political liberalization. This paper consists of the five sections below. The first section addresses previous literature about the relationship between ICT and democratic movements. The second section theoretically addresses how innovativeness in digital technologies disturb democratization in authoritarian states. The third section explains research methods this research project has applied. The fourth section reports on empirical results. The last section discusses the conclusion.


Although there is plentiful literature discussing the effect of technological improvement on democratization, most of them concentrate on case studies without a systematic empirical analysis. A few have attempted to conduct empirical analyses to demonstrate how technologies influence political liberalization.

Several works examined social and digital media’s role in political activities. Bulovsky (2019) introduced national leaders’ tendency in use of social media in democratic and non-democratic states. Democratic leaders have a greater incentive to use social media than authoritarian leaders because they need to communicate with voters for advertising their policies during an election period (23-24). For example, “in general, the more democratic a country, the more their leader demonstrates a general willingness to engage with other users by replying to their tweets” (30). Statistical analyses reported a positive significant relationship between the average number of tweets its leader’s account tweets per day and the level of political freedom (29). Christensen and Groshek (2020) analyzed the relationship among emerging media, protests against a government, and repression on political opposition. They argued that “countries with higher levels of emerging media are more likely to have higher levels of anti-government demonstrations, regardless of regime status as democratic or autocratic” (691). Moreover, it also mentioned that countries with higher levels of emerging media are more likely to repress political opposition, particularly in authoritarian states as opposed to democratic states. The empirical works covering 162 countries over an 18-year proved three facts below (691). The first one is that democratic countries have a higher level of emerging media than authoritarian states; second, the higher level of emerging media in democratic and authoritarian countries increases anti-government protests; the third finding is that governmental repression on anti-government protests increase in both democracy and autocracy (695, 697, 698). Boulianne (2020) examined whether or not the development of digital media including online news sources and social networking sites increase offline participation in civil and political activities, such as voting, volunteer activity, and protest through a meta-analysis (948). Digital media has been used as a tool to share information in the service of offline civil and political activities (953). The results of the meta-analysis using a statistical synthesis of data found that digital media is associated with offline participations in civic and political life. That is, “the trend is explained by the rise of social networking sites, more interactive websites, and the rise of online tools to facilitate political participation” (948-949).

A few studies focused on how internet access affects political movements and governmental suppression. Rød and Weidmann (2015) explored whether or not the use of the internet promotes democratization. They argue that “autocrats are likely aware of the tremendous potential this technology has for creating and maintaining a tightly controlled sphere of public opinion” (338). The authors found positive significant relationships between press censorship and internet penetration by conducting large-N analyses on authoritarian countries from 1993 to 2010. That is, an authoritarian regime with tougher press censorship expands internet access across an entire country more than a regime with lighter press censorship (344). The work concluded that the expansion of the internet does not lead to democratization (345). Ruijgrok (2017) examined the relationship between information access and protests in non-democratic countries. He or she expected that the popularization of the internet increases protests by decreasing communication costs, instigating attitudinal changes, reducing the uncertainty of information for potential protesters, and spreading dramatic videos and images (499). The empirical analyses proved that “internet use has a positive direct effect on the number of public protests, and is significant using a 99% confidence interval” (509). Xu (2021) conducted a very innovative empirical study analyzing the impact of the Golden Shield Project, the Chinese internet censorship system, on welfare and targeted suppression against opponents. As improved ICT technologies help China detect anti-government figures more easily, the Chinese regime has less incentive to provide welfare toward the public for soothing their dissatisfaction and a smaller motive to conduct indiscriminate suppression against opponents (313). Xu measured the Golden Shield Project’s effect on welfare and suppression by using data for welfare provision, agriculture investments, provision of other public goods, local public security expenditure, and the number of political prisoners at the county level. The author found that the Golden Shield Project decreased welfare provision and increased targeted suppression local governments conducted (318, 320). Shirazi (2008) conducted empirical research to explore the relationship between ICT expansion and freedom in ten Middle Eastern countries from 1995 to 2003. The author anticipated that expansion of ICT and education will improve liberty by improving citizens’ power to intervene in the decision making. Furthermore, a government’s regulation of the flow of information will restrict freedom (13). Empirical results also supported the original arguments (17). Stoycheff, Burgess, and Martucci (2020) analyzed whether online censorship and surveillance deter democratization and disruptive political movement. They expected that online censorship and surveillance limit democratic movement and disruptive political participation (478). Statistical results verify that internet surveillance has a negative significant impact on democratization, but internet censorship could not deter democratization. Interaction terms of online censorship and surveillance showed a negative significant impact on democratization (480). In terms of disruptive political participation, online censorship decreased disruptive political participation, but states with stronger internet surveillance suffer from a greater number of disruptive political participation (482). Johnson and Kolko (2010) tried to examine “whether e-Government has increased transparency, accountability and/or responsiveness to citizens” at the national, regional, and city levels in three Central Asian states, such as Kazakhstan, Kyrgyzstan, and Uzbekistan. They assessed three components of e-Government in the three Central Asian countries, such as information access, services/ interaction, and audience/agenda setting by analyzing government websites at national, regional, and city levels (17). They found that e-Government websites at the national-level do not improve the transparency and accountability of government institutions and agencies, while at the city and regional level, e-Government websites achieve transparency and accountability to the public (17).

Current literature above could not exactly address how technological innovativeness that authoritarian states have disturb democratization because they mainly focused on the impact of the use of media and the internet on democratic movements instead of the creation of the digital technologies to reflect technological innovativeness.


Digital technological improvements can limit the democratization process in several ways, such as economic growth (provision of welfare), detection of anti-governmental activities (censorship and surveillance), manipulation of information, decrease of indiscriminate suppression, and application of imported new technologies. Table 1 reveals the list of the number of patents in digital technologies in 2018 by ranking. China, Russia, Singapore, and Saudi Arabia are high-ranked innovative states generating many digital technologies.

Economic growth

Authoritarian regimes need enduring economic growth that ensures welfare for the masses to prevent social disruption stimulating a revolutionary movement because lower economic productivity can result in public backlash and mobilization (Nafziger and Auvinen 2002; Sullivan 2016). Dictators who fear frequent protests disturbing economic production, investment, and normal functioning of a government and bureaucrats have a great incentive to provide welfare to potential protesters as a preemptive means (Pan 2015, 13-16). Furthermore, “along with the industrial development and the improvement in the knowledge society, the usage of knowledge had become the new economic important resource, which changes the approach towards performance and competitiveness completely” (Gherghina 2020, 4).

Having high technologies is central for continuing economic growth and maintaining authoritarian control because they retain global competitiveness1 relying on progress, innovation and on the ability to self-change and improve (Porter 1998). In addition, new technologies that involve new forms of management and production organization, methods, and skills are essential in the improvement of productivity (Bulturbayevich and Jurayevich 2020, 5; Khong 2019, 122). Especially, digital technologies developing digital networking and communication infrastructures and facilities improve economic productivity and efficiency through integration of economies in the world by enabling cooperation, communication, and information exchange between entities (Bulturbayevich and Jurayevich 2020, 5). According to Figure 1, there is a positive relationship between the logged number of patents for digital technologies and logged GDP per capita. Innovative states in digital technology are richer than states generating a smaller number of patents for digital technology. As the development of digital technologies helps authoritarian governments to maintain their competitiveness in the global economy, they can keep their autocratic governance stably by providing welfare.

H1a: An innovative authoritarian regime with a higher economic growth is likely to curb democratization.

H1b: An innovative authoritarian regime with a more advanced economy is likely to curb democratization.

Detection of anti-governmental activities

The second way of preventing democratization of innovative authoritarian states is controlling information flow through censorship and surveillance. Digital technologies and equipment, such as CCTVs (Closed-circuit Television), smart phone applications, face recognition technology, and AI (artificial intelligence) help a dictator regime detect anti-governmental activities and information both online and offline. For example, China considering digital technologies as a driving force in economic development has showed how to use digital technologies for detecting citizens’ activities. China has developed online censorship and surveillance dramatically in the past two decades (Polyakova and Meserole 2019, 2-3): the internet in China has been monitored by 60 agencies under the Cyberspace Administration (Roberts 2018, 106-107); dissidents and human rights activists who posted their opinions on social media websites such as Weibo and WeChat are frequently arrested2; 9,000 mobile applications and 700 websites violating the Chinese government’s requirements were shut down in early 2019 (He 2019). China has monitored offline activities of Chinese citizens through surveillance systems. 800,000 surveillance cameras were established in Beijing in 2010, and more than 20 million cameras were installed in the entire country (Langfitt 2013). The Ministry of Public Security (MPS) and the Ministry of Industry and Information Technology (MIIT) jointly developed SkyNet, a program to install a national network of CCTV (Closed-circuit Television) feeds in 2005 to fully control the nationwide surveillance system (Zhang 2012).

H2: An innovative authoritarian regime with a stronger censorship is likely to curb democratization.

Manipulation of information

According to the Oxford University report in 2019, more than 56 countries have conducted cyber military activities on Facebook to manipulate information (Bradshaw and Howard 2019, i). Advanced digital technologies lead authoritarian regimes to manipulate information to energize supporters or disorient opponents. Social media that shape ideas and redesign society has become one of the key means including deceptive news in controlling of public opinions (Chen, Chen, and Xia 2022, 2). Authoritarian regimes can spread pro-regime disinformation through algorithms using bot and troll armies or key social-media influencers. The Chinese, Iranian, and Russian governments have generated progovernment messages toward huge numbers of people by employing their own social-media experts (Forelle et al 2015, 1). “For example, social-media platforms use content-curation algorithms to drive users toward certain articles—and keep them addicted to their social-media feeds” (Feldstein 2019, 46). As the government engages in social media contents, pro-government information has increased in Turkey. (Yeşil, Sözeri, and Khazraee 2017, 17). The Justice and Development Party, a ruling party, in Turkey has been alleged to disseminate pro-regime information by using 18,000 troll armies and has formulated ‘white hat’ hackers aimed to support security agencies to hack dissident users (Al-Rawi 2021, 142; Poyrazlar 2014; Topak 2019). Saudi Arabia having a million Twitter users (40% of entire Twitter users in the MENA) is the most connected Twitter state in the MENA (Middle East and North Africa). The Saudi Arabian government has been alleged to generate a tremendous number of tweets to “lionize the Saudi government, or praise Saudi’s efforts and intervention in Yemen, pointing to pro-Saudi propaganda”, and create sectarian and hate-inciting agendas (Jones 2016). Authoritarian governments have spread pro-regime disinformation through social-media platforms and distributed automated and hyperpersonalized disinformation targeting specific anti-regime figures or groups made by AI (artificial intelligence). AI technologies, such as deep-fake technology can generate and spread fabricated video and audio as pro-regime disinformation to attack dissidents. The first AI system generate fake video and audio, the second AI system tries to uncover them the first AI system made. The second AI system’s feedbacks play as a source to create more realistic forgeries for the first AI system. As these repeated processes generate more sophisticated forgeries AI systems cannot detect. These forgeries that AI systems create are applied as a tool to delegitimize and discredit anti-regime figures by describing them as a person making inflammatory remarks or engaging in vile acts (Feldstein 2019, 46).

Decrease of indiscriminate suppression

Developed detecting technologies can reduce indiscriminate suppression on citizens. A stable authoritarian regime is reluctant to impose universal repression on public because it can ignite citizens’ discontent to facilitate mass mobilization (Xu 2021, 2). Selective suppression on radical dissidents is a more suitable way to curb mass mobilization than indiscriminative repression (Ritter and Conrad 2016). However, as people rarely express their anti-regime sentiments under an authoritarian governance without the freedom of expression, an authoritarian leader with an incomplete information problem has a hard time knowing public opinions in detail (Kuran 1991). Advanced digital technologies, especially surveillance technologies (employing a high-resolution camera, a facial recognition system, and a big data processing system) let dictators accurately identify radical dissidents by analyzing communication and shared information among citizens in digital formats. For example, the Chinese, Iranian, and Syrian regimes tried to monitor citizens’ online and offline anti-regime activities by employing digital surveillance facilities in a timely manner (Gohdes 2014; Gunitsky 2015; Liu and Wang 2017; Qin, Stromberg, and Wu 2017). After all, authoritarian regimes with innovative ICT can punish targeted radical dissidents and increase the likelihood of the regime survival at a lower cost (Xu 2021, 2).

Application of imported technologies

High-tech authoritarian countries, such as China and Russia have exported digital technologies for censorship and surveillance toward 18 to 36 authoritarian states. China has exported these digital technologies to monitor citizens to dictator states participating in the Belt and Road Initiative and educated elites in these states how to use these technologies in controlling their citizens. Russia has transferred surveillance technologies to former Soviet Union states and countries in the Middle East (Polyakova and Meserole 2019, 1, 6). Countries having expanded fundamental infrastructures and facilities can apply imported high technologies more easily and effectively. For example, a digital surveillance system can operate under basic information and communication facilities, such as wired and wireless networks. Compared to African countries with a lower share of internet users, such as South Sudan, Togo, Somalia, and Congo, the Middle East countries with a higher share of internet users can more comprehensively use digital surveillance technologies imported from high-tech authoritarian countries, such as China and Russia. Saudi Arabia generating the fifth largest number of digital technology patents among non-democratic states has used imported surveillance technologies (a malicious mobile phone spyware and hacking spyware generated by the Israeli company, the NSO Group and Quadream) and effectively applied them in the surveillance on anti-government activists, such as Jamal Khashoggi and Omar Abdulaziz (Fatafta 2021, 43; Marczak et al. 2018; Megiddo 2021).

H3a: An innovative authoritarian regime with a higher internet user rate is likely to curb democratization.

H3b: An innovative authoritarian regime with a higher mobile phone user rate is likely to curb democratization.

Figure 2 reveals a relationship between innovation and democracy. The x-axis is the logged number of patents for digital technologies given authoritarian states create, and the y-axis is the democracy index (polity 2 index). The higher democracy indices concentrate on the minimum number of patents, while the greater numbers of patents stay at the lowest democracy indices. According to the figure, the most innovative dictatorships maintain full authoritarianism. It demonstrates digital technologies help dictatorships curb democratization.


Unit of analysis

This research project is based on state level and annual based data. The main target of the research is authoritarian nation states, and annually collected data has been applied. A criterion for selecting authoritarian countries is whether the DEMOC index of the Polity V project is lower than 6. 91 non-democratic states were filtered as such. The time range is from 2000 to 2018. The total sample size is 1,729.

Dependent variable

This research project tried to find the impact of innovativeness in digital technologies on democratization. The dependent variable is democratization of authoritarian states. A democratic transition in authoritarian states was measured by the polity 2 index from the Polity V project (Center for Systemic Peace 2020). The polity 2 “is computed by subtracting the AUTOC score from the DEMOC score; the resulting unified polity scale ranges from +10 (strongly democratic) to -10 (strongly autocratic)” (Marshall and Gurr 2020, 16).


Independent variables comprised patent, logged patent, internet user, mobile phone user, censorship, GDP growth, and GDP/capita (log). Patents are the best variable measuring national innovativeness. More innovative states generate more patents than less innovative countries. The number of patents and the logged number of patents successfully reflect innovativeness. WIPO (World Intellectual Property Organization) (2021) statistics database provides patent data in each country. This study filtered patents concerning digital technologies (including electrical machinery, apparatus, energy, computer, IT methods for management, semiconductors, telecommunications, digital and basic communication, audio-visual, micro-structural and nano, and optics technologies) from the patent data, and then summed them. The patent data are classified in patent publications3 and patent grants4, and the patent variable is the combined number of patent publications and patent grants concerning digital technologies mentioned above. Patents reflecting technology production is a variable concerning the level of innovativeness for technologies. They cannot fully reflect usage of digital technologies in disturbing democratization. This work uses internet users and mobile phone users that represent usage and basic conditions for ICT infrastructures in authoritarian countries. These variables sufficiently reflect usage and fundamental environments of digital technologies in information and communication. Data for mobile cellular subscriptions per 100 people and individuals using the internet per population were collected from the World Bank (2021) database. Censorship is one of key factors to curb political liberalization. Because digital authoritarian states use various digital equipment, such as AI (Artificial Intelligence) technologies, smart phone applications, and internet censorship programs in search engines as a means of censorship, variables reflecting digital censorship tools should be applied. However, current data concerning censorship is limited: there is no comprehensive data including the usage of AI (Artificial Intelligence) technologies, smart phone applications, and internet censorship programs in all non-democratic countries. For this reason, this research project relies on the freedom press score that the Freedom House project (2021) provides. This index reflects a censorship level by comprising legal, political, economic, and repressive control on media content. Current digital authoritarian countries have censored contents and expressions not only through digital technologies and equipment but also legal procedures, such as highly restrictive speech and expression laws (Morgus 2018, 86, 88). The freedom press score including legal control on media contents can partially reflect present digital non-democratic states’ censorship. Better economic situations are an important driving force to maintain an authoritarian regime’s stable control. GDP growth and GDP per capita (log) are indicators of economic conditions. Data for these indices were collected from the World Bank (2021) database.

Control variable

There are several control variables, such as FDI (foreign direct investment) inflow (log), trade dependency, civil violence, tertiary education, and urban population. FDI (foreign direct investment) and trade are a route facilitating flow of firm-specific assets at an international level, such as production technologies, managerial and organizational skills, and trademarked brands (Pandya 2014, 477). FDI and trade transferring high technologies toward authoritarian states are a quite important source for curbing democratization. This paper applied logged FDI inflow and the ratio of trade in GDP obtained from the World Bank (2021) database. Domestic instability affects a regime’s control of democratization. For example, a civil war may weaken a government’s capability to deter democratic movements and to develop high technologies (Rød and Weidmann 2015, 343). Major Episodes of Political Violence (MEPV) (Marshall 2019a) in the COW (Correlates of War) project provide data involving all types of major armed conflict at domestic and international level. CIVTOT index reflecting total summed magnitudes of all societal armed conflict, such as civil violence (CIVVIOL), civil war (CIVWAR), ethnic violence (ETHVIOL), and ethnic war (ETHWAR) was used for measuring civil violence. Each score of sub index is from 0 (lowest) to 10 (highest) (Marshall 2019, 4). Education is a variable influencing both democratization and technological improvement. Not only is technological innovation based on higher education (Audretsch and Vivarelli 1996) but education also stimulates political participation. For example, “levels of education are significant (although not powerful) predictors of individual voter registration and voting” (Mc Ginn 1996, 346). This project used a tertiary school enrollment rate collected from the World Bank (2021) database as a variable to represent an education level. Finally, urban population should be controlled because “a large urban population may pose a greater threat to autocratic governments (e.g. through protests or riots) than a dispersed and spatially remote rural population” (Rød and Weidmann 2015, 344). Data concerning the percentage of urban population were acquired from the World Bank (2021) database. All variables are summarized in Table 2 below.


Table 3 reports baseline results for democratization, the dependent variable. Model I and III are OLS (Ordinary Least Squares) model, and Model II and IV are OLS model with a robustness test. The number of patents reflecting innovativeness had negative relationships with the Polity 2 index in Model I and II, but these relations are not statistically significant. Patents that authoritarian states generate do not affect democratization. These results demonstrate that generated digital technologies do not help authoritarian regimes control democratic movements more effectively. However, logged patents are positively associated with democracy, and it is statistically significant in Model I and II. That is, more innovative states enjoy more political freedom.

Internet and mobile phone users representing the expansion of infrastructures revealed different effects on democratization. The ratio of internet users was negatively associated with the democracy index, and there is a statistical significant relationship in Model III and IV, while the proportion of mobile phone users had positive significant relationships with democratization in all four models. The expansion of the internet acts as a tool for surveillance, propaganda, and disinformation campaign to disturb democratic collective behaviors (Brooking and Singer 2016; King et al. 2017; Marwick and Lewis 2017). However, mobile phone users have positive significant relationships with the democracy index. Authoritarian states with the popular use of mobile phones suffered from democratization than non-democratic countries with less prevalence in the use of mobile phones. Mobile phones that facilitate collective action are a tool for liberation by sharing information, such as text massages, videos, and images (Bimber et al. 2005; Garrett 2006; Howard and Hussain 2013; Shirky 2011). Censorship on media was negatively related with democratization, and relationships are statistically significant. An autocracy imposing a stronger censorship preventing information exchange on media effectively disturb democratic movements.

In terms of economic conditions, GDP growth did not have an important impact on democratization, while the logged GDP per capita included negative significant relationships with the democracy index. Rapid economic improvement does not democratize non-democratic states. However, richer autocracies ensuring the well-being of citizens could effectively curb democratization and maintain their dictatorship in a stable manner. FDI and trade are routes to obtain new technologies. However, statistical results reported that both variables (FDI inflow (log) and trade dependency) are negatively related with democracy, but only a relationship between FDI and democracy is statistically significant. Non-democratic states can obtain digital technologies facilitating surveillance and censorship from foreign corporations investing in these countries (Fatafta 2021, 43; Marczak et al. 2018; Megiddo 2021).

Civil instability showed positive significant impacts on democratization. Because civil disputes have an anti-government tendency or can develop into anti-government movements, civil violence, such as a civil war is a strong signal for resisting current dictatorship control. Education level was positively associated with democratization, and it is statistically significant. These results proved that higher education level facilitates participation in political activities. Finally, the proportion of urban population had negative significant impacts on democratic transition. According to a conventional viewpoint, compared with a dispersed rural population, a crowded urban population lead to collective actions more easily than usual (Rød and Weidmann 2015, 344). However, the result was not consistent with common belief. As the national and global economy has focused on urban productivity and urban renewal, people in cities are better off than persons in rural areas (Stéphanie and Zérah 2011, 1). Thus, an urban population has less motive to fight against a dictator.

Table 4 shows results for fixed (year) effect analyses on democratization, the dependent variable. Model I and III are fixed (year) effect analysis models, and Model II and IV are fixed (year) effect analysis models with a robustness test. The number of patents is positively related with democracy, and it is statistically significant in Model I and II. These results prove that innovativeness of authoritarian states improve the level of democracy. However, logged patents are positively associated with the polity2 index, while they are not statistically significant. Logged patents do not facilitate political freedom. These results are not consistent with results in Table 3 that revealed positive significant relationships between the logged numbers of patents and democracy.

Ratios of internet and mobile phone users have different impacts on the polity2 index. The ratio of internet users has negative insignificant relations with the political freedom, and these relationships are not statistical significant. The expansion of the internet does not influence democratization. Internet usage does not act as means to curb democratization. However, the mobile phone user rate has a positive significant relationship with democracy in Model I. These empirical results prove that mobile phones are an instrument for liberation (Bimber et al. 2005; Garrett 2006; Howard and Hussain 2013; Shirky 2011). Censorship has negative significant relations with democracy. Dictators censoring citizens’ behaviors and opinions successfully prevent democratic movements.

Regarding economic situations, the GDP growth and the GDP per capita (log) did not have a substantial impact on political liberalization. The GDP growth has positive relations with democracy, whereas the logged GDP per capita involved negative relationships with democracy. However, both are not statistically significant. FDI and trade are channels for obtaining new technologies. The logged FDI inflow has negative significant relationships with polity2 in Model I and III. That is, authoritarian countries obtaining digital technologies through FDI can deter democratization effectively (Fatafta 2021, 43; Marczak et al. 2018; Megiddo 2021). On the contrary, the ratio trade in GDP has positive significant relations with democracy. Non-democratic countries that rely more on trade suffer from the democratization process more than closed dictatorships.

Domestic instability (civil violence) revealed positive significant effects on political liberalization. The results demonstrate that civil violence (such as a civil war) involving an anti-government tendency seriously weakens dictatorship’s control. The education level was negatively associated with democratization, and it is statistically significant. These results did not prove that a higher education level facilitates participation in political activities. It is possible that an authoritarian regime’s nationalistic education could break the desire for democracy by propagandizing citizens. Finally, the ratio of urban population had negative significant relationships with democratic transition. The result was not consistent with the conventional viewpoint that crowded urban population lead to collective actions more easily (Rød and Weidmann 2015, 344). A wealthier urban population may have less motive to fight against a dictator than people in rural areas.

Table 5 reports statistical results for relations between interaction terms including the logged number of patents of digital technologies and democratization. Model I, III, V, VII, and IX are OLS (Ordinary Least Squares) model, and Model II, IV, VI, VIII, and X are OLS model with a robustness test. Interaction terms of patents (log) and economic conditions were analyzed from Model I through IV. Interaction terms of logged patents and GDP growth had negative relationships with democratization, but they are not statistically significant. Innovative authoritarian states with greater economic growth did not affect political freedom. Interaction terms of logged patents and GDP per capita have a similar effect of interaction terms of logged patents and GDP growth: there are negative non-significant relationships with the polity2 index. The innovativeness of non-democratic countries with an advanced economy does not have a significant impact on their democratization. These results rejected the hypothesis H1a and H1b. In terms of censorship, interaction terms of logged patents and censorship have negative significant relations with the polity 2 index. Innovative dictatorships implementing stronger censorship effectively control democratic movements by detecting anti-regime activities. The results support the hypothesis H2.

In terms of ICT infrastructures, the ratio of internet users and the rate of mobile phone users do not have an important effect on democratic movements in targeted authoritarian states. Interaction terms of patents and the internet user rate had negative insignificant relationships with the polity 2 index, whereas interaction terms of patents and the mobile phone user rate also had positive insignificant relationships on political freedom. They all are not statistically significant. The hypotheses H3a and H3b were not supported.

With regard to the control variable, GDP growth had negative significant relationships with democracy. Rapid economic growth disturbs democratization by curbing public discontent with welfare. The logged GDP per capita is negatively associated with political freedom. An advanced economy limits political liberalization in dictatorships by providing ample welfare toward the public. Economic globalization variables, FDI inflow (log) and the share of trade in GDP, do not have substantial effects on democratic movement. Both variables did not have a statistically significant relationship with polity2 index.

The civil violence variable reflecting domestic instability reveals positive relations with democracy, but they are not statistically significant. Civil violence does not involve a meaningful impact on democratization. The statistical results for the tertiary education rate showed positive significant relationships with democracy. The proliferation of higher education enhances political development by facilitating willingness for political participation. Finally, the proportion of an urban population to the total population was negatively related with democracy, and the relationship was statistically significant. These results are opposite to the previous research arguing a dense urban population has a greater likelihood of facilitating collective behaviors than a dispersed rural population (Rød and Weidmann 2015, 344). As the national and global economy depends on urban productivity and urban renewal, an urban population lives a richer life than rural people (Stéphanie and Zérah 2011, 1). From this logic, people in urban areas do not have greater motive to participate in collective political actions against the current dictatorship.

Table 6 shows statistical results for fixed (year) effect analyses. Model I and II showed results for interaction terms of patents and economic conditions. The interaction term of logged patents and GDP growth had a positive insignificant relationship with democratization. The interaction term of patents and GDP per capita were positively associated with the democracy index, and it is statistically significant. Although innovative authoritarian states with higher economic growth do not affect democratic movements, wealthier innovative authoritarian regimes suffer from democratization. The advanced economy promoting a development of a middle class and a spread of education is one key driving force for democratization (Bollen 1979; Bollen 1983; Bollen and Jackman, 1985; Brunk, Caldeira, and Lewis-Beck 1987; Dahl 1989; Jackman 1973; Lipset 1959). These results reject the hypotheses H1a and H1b. In terms of press freedom, the interaction term of logged patents and censorship were negatively related with the polity 2 index, and this relationship is not statistically significant. Innovative dictatorships with stronger censorship do not effectively disturb democratization. Hypothesis H2 was not supported.

Regarding infrastructures, the internet user rate and the ratio of mobile phone user had similar impacts. The interaction term of the logged number of patents and the internet user rate had a negative relationship with the polity 2 index, but it is not statistically significant. Ditatorships spreading the internet could not keep their control. The interaction term of the logged number of patents and the mobile phone user rate had a positive insignificant relation with democratization. According to the empirical results, authoritarian states having digital infrastructures across an entire country cannot prevent collective actions for democratization. These results rejected hypotheses H3a and H3b.

In terms of empirical results for control variables, GDP growth had a positive significant relationship with the polity 2 index in Model IV. Fast economic development of affluent dictatorships facilitate democratization successfully. The logged GDP per capita does not have an important impact on democratization. It has a negative insignificant relationship with democracy. Greater wealth does not restrain or promote political liberalization in dictatorships. Logged FDI inflow and trade dependency reflecting economic globalization had different effects on democratization. FDI inflow had negaive relationships with the democracy index, while the ratio of trade to GDP was positively related with democratization. Both variables have a statistically significant relationship with democracy. Non-democratic countries can deter democratization effectively by obtaining digital technologies through FDI (Fatafta 2021, 43; Marczak et al. 2018; Megiddo 2021). Dictatorships depending more on trade face democratization movements compard to closed authoritarian countries.

According to the statistical results, the civil violence variable has negative significant relationships with democratization. It is possible that a democratic movement is suppressed by an authoritarian regime by using civil conflict as an excuse. The results for the tertiary education rate revealed negative significant relationships with democratization. Tertiary education disturbs democratization because an authoritarian regime’s nationalistic education might break willingness for democracy by propagandizing citizens. Lastly, the ratio of the urban population in the total population was negatively associated with democratic movements, and it was statistically significant. These results were not consistent with the previous research arguing crowded urban population has a greater likelihood to promote collective actions than dispersed rural population (Rød and Weidmann 2015, 344). As the national and global economy are operated by urban productivity and urban renewal, people in urban areas enjoy more affluent life conditions than rural people (Stéphanie and Zérah 2011, 1). The urban population does not have greater incentive to participate in collective political movement against a current dictator regime.

As Figure 3 showed marginal effects of independent variables on democracy in fixed effect (year) models with a 95% confidence interval. The horizontal axis is values of each independent variable, and the vertical axis indicates the predicted level of democracy. The average Polity2 score for each state in each year is -0.49. As the number of patents increases from 0 to 577,000, the polity2 score decreases from -0.49 to -2.30. Moreover, as the logged number of patents increases from 0 to 14, the polity2 score decreases from -0.23 to -3.18. The innovativeness in digital technologies significantly decreases the level of democracy. The margin effect of the logged patent variable on the level of democracy is not consistent with empirical results in Table 3 and 4. These results verified dual-effects of digital technologies: only innovativeness in digital technologies does not facilitate political liberalization and rather disturbs democratic movements. As for economic situations, such as GDP growth and the logged GDP per capita increase from -65% to 150% and from 3 to 12, the democratic level increases from -2.85 to 4.93 and decreases from 1.58 to -2.22. The faster economic development lets dictatorships suffer from democratization, but wealthier economies help authoritarian regimes maintain their non-democratic control. When the internet user rate increases from 0 to 100, the polity 2 score decreases from 0.09 to -3.27. As the ratio of mobile phone users increases from 0 to 220, the polity 2 score increases from -0.74 to 0.25. These results prove that mobile phones are a more significant key instrument for democratization than the internet. The increase of level of the freedom in press (censorship) from 22 to 100 decreases the polity 2 score from 7.84 to -6.86. Censorship acts as a means for curbing democratic movements.

In terms of impacts of innovativeness and usage of technologies, increase and decrease of polity2 based on variations of the number of patents, the logged number of patents, the internet user rate, and the ratio of mobile phone users are presented in Table 7 below. According to Table 7, the increase of the internet user rate has the greater impact on decrease of the democracy level than patents (and logged patents). However, although the mobile phone user rate has the smallest impact on democracy, it contributes to democratization. In short, the expansion of the internet is the most important single factor curbing democracy.


The dual use of digital technologies that authoritarian states have is a controversial topic in the study of the impact of digital technology on political liberalization. Recently, digital authoritarian states are reinforcing their controlling capability on citizens by using innovative digital technologies, such as AI and facial recognition technology. By using patent data, this study examined whether authoritarian states creating innovative digital technologies restrain democratization. Because previous literatures on the impact of digital technologies on democratic movement concentrate on just the usage of digital technologies (the ratio of using the internet and mobile phones) instead of the creation of digital technologies, they could not elaborate on the impact of the innovativeness of authoritarian states on democratization accurately. This work using the patent data attempted to resolve this problem. This work argued that innovative authoritarian countries could deter democratization for several reasons, such as the provision of welfare with higher economic growth, the detection of citizens’ activities, the manipulation of information to disorient opponents, the reduction of universal repression on the public, and more effective application of technologies with digital infrastructures. The empirical analyses did not support the original argument above. Statistical results showed that the increase of the number of patents for digital technologies (including logged patent) improve political freedom in given authoritarian states. The spread of innovative digital technologies gives citizens much more opportunities for participation in political activities online and offline. Although a few digital authoritarian states, such as China and Russia have successfully controlled democratic movements, digital innovation has led the wave of democratization. Empirical results for interaction terms showed that only dictatorships generating more innovative digital technologies with stronger censorship disturb democratization by detecting opponents’ activities. On the contrary, fixed effect analyses proved that innovative authoritarian countries with a more advanced economy (higher GDP per capita (log)) undergo democratization.


1 “The capacity of producing proper goods and services of eligible quality, at the right price and at the right time” (Cousins, 2018) or “the capacity of companies to compete, to develop and to increase profits” (Wignaraja 2003).

2 “China web users arrested over posts on Sina Weibo,” BBC News, August 22, 2013,

3 “A patent publication is a published utility patent application. A patent publication is not a patent. While a published patent application may eventually issue into a patent, the patent publication consists of only the application itself, namely, the drawings and written specification. The patent publication does not provide information about events subsequent to the publication date, e.g., whether the application was approved. To determine the subsequent file history, you have to look up the application number at USPTO Public Pair” (Lin).

4 “A patent grant gives an inventor a property right in his invention, allowing him to ask others to stop using, making, and selling his invention for a limited period of time. That said, it is not the patent office's responsibility to stop others from using the patent holder's invention” (Adam 2019).

Fig. 1. Digital Technologies (log) and GDP/capita (log) in 2018
Fig. 2. Innovation and Democracy
Fig. 3. Marginal Effect of Independent Variables on Polity2 Index in Fixed Effect Model
Table. 1. The Number of Patents for Digital Technology in 2018 (WIPO 2021)
Table. 2. Summary of Variables
Table. 3. Baseline Results for Democratization (the Dependent Variable)
Table. 4. Fixed Effect (Year) Analyses for Democratization (the Dependent Variable)
Table. 5. Results for Relationships between Interaction Terms (logged patent) and Democratization (the Dependent Variable)
Table. 6. Fixed (Year) Effect Analyses for Interaction Terms
Table. 7. Marginal Effect of Technology Usage and Innovation on Democracy
Table. 8. Table
  1. Adam, Noah. 2019. “What Does Patent Grant Mean?” PatentRebel (September 22). Accessed at (August 26, 2021).
  2. Al-Rawi, Ahmed. 2021. “Disinformation under a networked authoritarian state: Saudi trolls’ credibility attacks against Jamal Khashoggi.” Open Information Science 5(1), 140-162.
  3. Audretsch, David B., and Marco Vivarelli. 1996. “Firms size and R&D spillovers: Evidence from Italy.” Small Business Economics 8(3), 249-258.
  4. Bimber, Bruce, Andrew J. Flanagin, and Cynthia Stohl. 2005. “Reconceptualizing collective action in the contemporary media environment." Communication theory 15(4), 365-388.
  5. Bollen, Kenneth. 1979. “Political Democracy and the Timing of Development.” American Sociological Review , 44(4), 572-587.
  6. Bollen, Kenneth. 1983. “World System Position, Depen- dency, and Democracy: The Cross-National Evidence.” American Sociological Review , 48(4), 468-479.
  7. Bollen, Kenneth and Robert W. Jackman. 1985. “Political Democracy and the Size Distribution of Income.” American Sociological Review , 50(4), 438-457.
  8. Brunk, Gregory, Gregory A. Caldeira, and Michael S. Lewis-Beck. 1987. “Capitalism, Socialism, and Democracy: An Empirical Inquiry.” European Journal of Political Research 15(4), 459-470.
  9. Boulianne, Shelley. 2020. “Twenty years of digital media effects on civic and political participation.” Communication research 47(7), 947-966.
  10. Bradshaw, Samantha and Philip N. Howard. 2019. “The global disinformation order: 2019 global inventory of organised social media manipulation.” Project on computational Propaganda.
  11. Brooking, Emerson and P. W. Singer. 2016. “War goes viral.” The Atlantic (November). Accessed at 11/war-goes-viral/501125/ (January 3, 2019).
  12. Bulovsky, Andrew. 2019. “Authoritarian communication on social media: The relationship between democracy and leaders’ digital communicative practices.” International Communication Gazette 81(1), 20-45.
  13. Bulturbayevich, Mullabayev Baxtiyarjon, and Mahmudov Baxriddin Jurayevich. 2020. “The impact of the digital economy on economic growth.” International Journal of Business, Law, and Education 1(1), 4-7.
  14. BBC. 2013. “China web users arrested over posts on Sina Weibo,” BBC News (August 22). Accessed at (October 3, 2021).
  15. Center for Systemic Peace. 2020. “Polity V Project, Political Regime Characteristics and Transitions, 1800-2018.” Accessed at (October 11, 2021).
  16. Chen, Long, Jianguo Chen, and Chunhe Xia. 2022. “Social network behavior and public opinion manipulation.” Journal of Information Security and Applications 64, 1-15.
  17. Christensen, Britt, and Jacob Groshek. 2020. “Emerging media, political protests, and government repression in autocracies and democracies from 1995 to 2012.” International Communication Gazette 82(8), 685-704.
  18. Cousins, John. 2018. MBA ASAP Marketing 2.0: Principles and Practice in the Digital Age Independently published.
  19. Dahl, Robert. 1989. Democracy and its Critics , New Haven: Yale University press.
  20. Fatafta, Marwa. 2021. “Transnational Digital Repression in the MENA Region.” In Marc Lynch ed. Digital Activism and Authoritarian Adaptation in the Middle East , 41-47.
  21. Feldstein, Steven. 2019. “The road to digital unfreedom: How artificial intelligence is reshaping repression.” Journal of Democracy 30(1), 40-52.
  22. Forelle, Michelle, et al. 2015. “Political bots and the manipulation of public opinion in Venezuela.” 1-7.
  23. Freedom House. 2021. “Freedom of the Press (FOTP) Data: Editions 1980-2017”, Accessed at (October 15, 2021)
  24. Garrett, Kelly. 2006. “Protest in an information society: A review of literature on social movements and new ICTs.” Information, communication & society 9(2), 202-224.
  25. Gherghina, Ștefan Cristian, et al. 2020. “Small and medium-sized enterprises (SMEs): The engine of economic growth through investments and innovation.” Sustainability 12(1), 1-22.
  26. Gohdes, Anita R. 2014. “Repression in the Digital Age: Communication Technology and the Politics of State Violence.” PhD dissertation, Universität Mannheim.
  27. Gunitsky, Seva. 2015. “Corrupting the Cyber-Commons: Social Media as a Tool of Autocratic Stability.” Perspectives on Politics 13(1), 42-54.
  28. He, Cathy. 2019. “China Purges 9,000 Mobile Apps, 700 Websites in Internet Crackdown,” The Epoch Times (January 23) Accessed at (October 20, 2021).
  29. Howard, Philip N., and Muzammil M. Hussain. 2013. Democracy's fourth wave?: digital media and the Arab Spring . Oxford: Oxford University Press.
  30. Jackman, Robert. 1973. “On the relation of economic development to democratic performance.” American Journal of Political Science , 17(3), 611-621.
  31. Johnson, Erica, and Beth Kolko. 2010. “e-Government and transparency in authoritarian regimes: Comparison of national-and city-level e-government web sites in Central Asia.” Digital Icons: Studies in Russian, Eurasian and Central European New Media 3(3), 15-48.
  32. Jones, Marc Owen. 2016. “Automated sectarianism and pro-Saudi propaganda on Twitter.” Exposing the Invisible (January 19). Accessed at (December 12, 2021).
  33. King, Gary, Jennifer Pan, and Margaret E. Roberts. 2017. “How the Chinese government fabricates social media posts for strategic distraction, not engaged argument.” American political science review 111(3), 484-501.
  34. Khong, Yuen Foong. 2019. “Power as prestige in world politics.” International Affairs 95(1), 119-142.
  35. Kuran, Timur. 1991. “Now out of Never: The Element of Surprise in the East European Revolution of 1989.” World Politics 44(1), 7-48.
  36. Langfitt, Frank. 2013. “In China, Beware: A Camera May Be Watching You,” NPR (January 23) Accessed at (November 7, 2021).
  37. Lin, Vic. “Patent publication = published patent application.” Patent Trademark Blog. Accessed at (August 25, 2021).
  38. Lipset, Seymour. 1959. “Some social requisites of democracy: Economic development and political legitimacy1.” American political science review , 53(1), 69-105.
  39. Liu, Joyce, and Xiqing Wang. 2017. “In Your Face: China’s AllSeeing State.” BBC News (December 10) Accessed at (August 10, 2021).
  40. Marczak, Bill, John Scott-Railton, Ron Deibert, Adam Senft, and Bahr Abdul Razzak. 2018. “The Kingdom Came To Canada: How Saudi-Linked Digital Espionage Reached Canadian Soil”. The Citizen Lab. Accessed at (December 12, 2021).
  41. Marshall, Monty. 2019. “Codebook: Major Episodes of Political Violence (MEPVv2018),” Center for Systemic Peace Accessed at (October 10, 2021).
  42. Marshall, Monty. 2019a. “Major Episodes of Political Violence (MEPV) and Conflict Regions, 1946-2018,” Center for Systemic Peace Accessed at (October 10, 2021).
  43. Marshall, Monty and Ted Robert Gurr. 2020. “POLITY5: Dataset Users’ Manual,” Center for Systemic Peace Accessed at (October 10, 2021).
  44. Marwick, Alice and Rebecca Lewis. 2017. “Media manipulation and disinformation online,” Data & Society Report (May 15). Accessed at (January 3, 2019).
  45. Mc Ginn, Noel F. 1996. “Education, democratization, and globalization: A challenge for comparative education.” Comparative Education Review 40(4), 341-357.
  46. Medeiros, Evan S. 2019. “The changing fundamentals of US-China relations.” The Washington Quarterly 42(3), 93-119.
  47. Megiddo, Gur. 2021. “Secretive Israeli Cyber Firm Selling Spy-Tech To Saudi Arabia,” (June 8). Accessed at (December 12, 2021).
  48. Morgus, Robert. 2018. “The spread of russia’s digital authoritarianism.” In Nicholas D. Wright ed., AI, China, Russia, and the Global Order: Technological, Political, Global, and Creative Perspective , 85-93.
  49. Nafziger, E. Wayne, and Juha Auvinen. 2002. “Economic Development, Inequality, War, and State Violence.” World Development 30(2), 153-63.
  50. Pan, Jennifer Jie. 2015. “Buying Inertia: Preempting Social Disorder with Selective Welfare Provision in Urban China.” Ph.D Dissertation, Harvard University.
  51. Pandya, Sonal. 2014. “Democratization and Foreign Direct Investment Liberalization, 1970-2000.” International Studies Quarterly 58(3), 475- 488.
  52. Polyakova, Alina, and Chris Meserole. 2019. “Exporting digital authoritarianism: The Russian and Chinese models.” Policy Brief, Democracy and Disorder ,1-22.
  53. Porter, Michael. 1985. Competitive Advantage: Creating and Sustaining Superior Performance , New York: Free Press.
  54. Poyrazlar, Elcin. 2014. “Turkey’s Leader Bans His Own Twitter Bot Army,” Vocativ (March 26). Accessed at (December 13, 2021).
  55. Qin, Bei, David Stromberg, and Yanhui Wu. 2017. “Why Does ¨ China Allow Freer Social Media? Protests versus Surveillance and Propaganda.” Journal of Economic Perspectives 31(1), 117-40.
  56. Ritter, Emily Hencken, and Courtenay R. Conrad. 2016. “Preventing and Responding to Dissent: The Observational Challenges of Explaining Strategic Repression.” American Political Science Review 110(1), 85-99.
  57. Roberts, Margaret, and Margaret E. Roberts. 2018. Censored , Princeton: Princeton University Press.
  58. Ruijgrok, Kris. 2017. “From the web to the streets: internet and protests under authoritarian regimes.” Democratization 24(3), 498-520.
  59. Rød, Espen Geelmuyden, and Nils B. Weidmann. 2015. “Empowering activists or autocrats? The Internet in authoritarian regimes.” Journal of Peace Research 52(3), 338-351.
  60. Shirazi, Farid. 2008. “The contribution of ICT to freedom and democracy: An empirical analysis of archival data on the Middle East.” The Electronic Journal of Information Systems in Developing Countries 35(1), 1-24.
  61. Shirky, Clay. 2011. “The political power of social media.” Foreign Affairs (January/February). Accessed at clay-shirky/the-political-power-of-social-media (December 30, 2020).
  62. Stoycheff, Elizabeth, G. Scott Burgess, and Maria Clara Martucci. 2020. “Online censorship and digital surveillance: the relationship between suppression technologies and democratization across countries.” Information, Communication & Society 23(4), 474-490.
  63. Sullivan, Christopher M. 2016. “Political Repression and the Destruction of Dissident Organizations: Evidence from the Archives of the Guatemalan National Police.” World Politics 68(4), 645-76.
  64. Tawa Lama-Rewal, Stéphanie, and Marie-Hélène Zérah. 2011. “Introduction. Urban Democracy: A South Asian Perspective.” South Asia Multidisciplinary Academic Journal 5, 1-17.
  65. Topak, Özgün . 2019. “The authoritarian surveillant assemblage: Authoritarian state surveillance in Turkey.” Security Dialogue 50(5), 454-472.
  66. Wignaraja, Ganeshan. 2003. Competitiveness strategy in developing countries: A manual for policy analysis . London: Routledge.
  67. WIPO (World Intellectual Property Organization). 2021. “WIPO statistics database”, Accessed at (August 1, 2021).
  68. World Bank. 2021. “World Development Indicators”, DataBank. Accessed at preview/on (November 1, 2021)
  69. Xu, Xu. 2021. “To Repress or to Co‐opt? Authoritarian Control in the Age of Digital Surveillance.” American Journal of Political Science 65(2), 309-325.
  70. Yesil, Bilge, Efe Kerem Sözeri, and Emad Khazraee. 2017. “Turkey’s Internet policy after the coup attempt: The emergence of a distributed network of online suppression and surveillance.” Internet Policy Observatory (February 28) Accessed at (December 14, 2021).
  71. Zhang, Zihan. 2012. “Beijing’s guardian angels?,” Global Times (October 10) Accessed at (October 6, 2021).

22-1 (April 2024)