Autocratization in An Age of Digitalization

My book project examines how elite access to digital media in authoritarian countries paradoxically empowers elites while undermining ruling party discipline and loyalty, destabilizing the very foundations of authoritarian rule.  

Authoritarian regimes rely on cohesive ruling parties to monopolize elite loyalty and suppress challenges to ensure regime survival. Yet history reveals elite loyalty is perpetually fragile as most transitions from authoritarian rule originate in splits within the regime's own ruling colation and elite. Thus, autocrats’ paramount task is preventing defections and coups, neutralizing challengers, and maintaining strong party cohesion. My book project directly engages with this evolving challenge, investigating how elite access to new communication technologies reshapes intra-party contention.

I argue that digital platforms equip elites with independent mobilization capabilities that run outside the party structure, enabling them to ascend within the ranks and assert autonomy. This fundamentally challenges the party's unity and the autocrat's control in two distinct and perilous ways: First, digitally influential elites displace traditional elites lacking an online presence, reshuffling the internal hierarchy and weakening established bonds of loyalty. More insidiously, it signals to the party rank-and-file that loyalty and traditional party work are no longer the primary paths to advancement, eroding the institutional norms that sustain discipline. Second, empowered by their digital followings, these elites increasingly adopt political behavior diverging from the party line, directly fracturing cohesion and creating new avenues for dissent within the ruling coalition.

My book project, therefore, provides unique insights into the durability of authoritarian governance and its implications for political liberalization.

Using causal inference, machine learning methods, qualitative evidence and survey experiments, I empirically test these arguments in Turkey and provide evidence supporting my theory. While my book project offers an in-depth case study of Turkey, it also provides evidence presenting the relevance of this theory for other electoral autocracies, like Gambia, Morocco and Malaysia. 

Below is an overview of data sources my book project utilizes. 

Low- and High-Level Politician Career Dataset

Through searches of local newspapers, party websites, and elites' digital media posts, I generated the following datasets:

1) Province Branch Leader Dataset: This dataset tracks the career advancements of 244 unique province branch leaders of the AK Party for each election between 2014 and 2024, totaling to 838 observations. It includes information on their digital media behavior, socio-economic characteristics, and political career trajectories. 

2) MP Dataset: I also collected career advancements of AK Party MPs from 2002 to 2023 for 1,194 MPs throughout each general election since 2002. This dataset includes information on their digital media behavior, socio-economic characteristics, and political career trajectories.  

Low- and High-Level Politicians Digital Media Use

To examine politicians' digital media use, I collaborated with Meta to collect Facebook and Instagram posts from both low- and high-level politicians. 

The dataset includes approximately 460,000 posts from low-level politicians and around 3 million posts from high-level politicians.

Legislative Speech Dataset

Aside from politicians' career trejectories, it is important to understand how their political behavior change as they establish their mobilization capacities in digital media. 

Therefore, I rely on a dataset on parliamentary speeches, encompassing all sessions held between 2002 and 2021, totaling 2112 sessions. This dataset comprises a total of 1,562,921 elite speeches spanning six parliamentary terms, from the 22nd to the 27th Parliaments. 

Online Experiments

My book project also explores how elites leverage digital media to build independent mobilization capacities. To investigate this, I conducted two online experiments. The first experiment (n = 800) examines how digital media facilitates the distribution of clientelistic goods. The second is a conjoint experiment (n = 520),  investigating the effect of elite activities in digital media on their mobilization capabilities.