![[Adobe Stock]](https://sp-ao.shortpixel.ai/client/to_webp,q_glossy,ret_img,w_750,h_375/https://www.psypost.org/wp-content/uploads/2025/09/twitter-x-750x375.jpg)
Federal legislators in the United States actively curate their digital footprints to project a specific professional identity. A new analysis reveals that these officials frequently remove social media posts that mention their private lives or name specific colleagues. But they tend to preserve posts that criticize policies or opponents. The research was published in the journal Computers in Human Behavior.
The digital age has transformed how elected officials communicate with voters. Social media platforms allow politicians to broadcast their views instantly. However, this speed also blurs the traditional boundaries between public performance and private thought.
Sociologist Erving Goffman described this dynamic as impression management. This concept suggests that individuals constantly perform to control how others perceive them. They attempt to keep their visible “front-stage” behavior consistent with a desired public image.
In the political arena, maintaining a consistent image is essential for securing votes and support. A single misstep on a platform like X, formerly known as Twitter, can damage a reputation instantly. Researchers wanted to understand how this pressure influences what politicians choose to hide. They sought to identify which specific characteristics prompt a legislator to hit the delete button.
The study was led by Siyuan Ma from the Department of Communication at the University of Macau. Ma worked alongside Junyi Han from the Leibniz-Institut für Wissensmedien in Germany and Wanrong Li from the University of Macau. They aimed to quantify the effort legislators put into managing their online impressions. They also wanted to see if the deletion of content followed a predictable pattern based on political strategy.
To investigate this, the team collected a massive dataset covering the 116th United States Congress. This session ran from January 2019 to September 2020. The researchers utilized a tool called Politwoops to retrieve data on deleted posts. This third-party platform archives tweets removed by public officials to ensure transparency. The dataset included nearly 30,000 deleted tweets and over 800,000 publicly available tweets from the same timeframe.
The researchers analyzed a random sample of these messages to ensure accuracy. Human coders reviewed the content to categorize the topics discussed. They looked for specific variables such as mentions of private life or policy statements. They also tracked mentions of other politicians and instances of criticism. This allowed the team to compare the content of deleted messages against those that remained online.
The timing of deletions offered early insights into political behavior. The data showed a sharp rise in the number of deleted tweets beginning in late 2019. This increase coincided with the start of the presidential impeachment inquiry. The high-stakes environment likely prompted legislators to be more cautious about their digital history.
The onset of the COVID-19 pandemic also shifted online behavior. As the health crisis unfolded, the total volume of tweets from legislators increased dramatically. Despite the higher volume of posts, the proportion of deleted messages remained elevated. This suggests that during periods of national crisis, the pressure to manage one’s public image intensifies.
When the researchers examined the content of the tweets, distinct patterns emerged. One of the strongest predictors for deletion was the mention of private life. Legislators were statistically more likely to remove posts about their families, hobbies, or vacations. This contradicts some political theories that suggest showing a “human side” helps build connections with voters.
Instead, the findings point toward a strategy of strict professionalism. By scrubbing personal details, politicians appear to be focusing the public’s attention on their official duties. They seem to use the platform as a space for serious legislative work rather than social intimacy. The data indicates that looking professional is prioritized over looking relatable.
Another major trigger for deletion was the mention of specific colleagues. Tweets that named other politicians were frequently removed from the public record. This behavior may be a strategic move to minimize liability. Mentioning a colleague who later becomes involved in a scandal can be damaging by association. Deleting these mentions keeps a legislator’s timeline clean of potential future embarrassments.
In contrast, the study found that criticism is rarely deleted. Legislators were likely to keep tweets that attacked opposing policies or ideologies visible. This suggests that being critical is viewed as a standard and acceptable part of a politician’s role. It signals to voters that the official is actively fighting for their interests.
The study also evaluated the accuracy of the information shared by these officials. Popular narratives often suggest that social media is flooded with false information from all sides. However, the analysis showed that legislators rarely posted demonstrably false claims. This adherence to factual information was consistent across both deleted and public tweets.
Party loyalty acted as a powerful constraint on behavior. The researchers found almost no instances of legislators posting content that violated their party’s stance. This was true even among the deleted tweets. The lack of dissent suggests an intense pressure to maintain a united front. Deviating from the party line appears to be a risk that few elected officials are willing to take.
The status of the legislator also influenced their deletion habits. The study compared members of the House of Representatives with members of the Senate. The results showed that Representatives were more likely to delete tweets than Senators. This difference likely stems from the varying political pressures they face.
Senators serve six-year terms and represent entire states. They typically have greater name recognition and more secure political resources. This security may give them the confidence to leave their statements on the public record. They feel less need to constantly micromanage their online presence.
Representatives, however, face re-election every two years. They often represent smaller, more volatile districts where a small shift in opinion can cost them their seat. This constant campaign mode creates a higher sensitivity to public perception. Consequently, they appear to scrub their social media accounts more aggressively to avoid potential controversies.
The findings illustrate that social media management is not random. It is a calculated extension of a politician’s broader communication strategy. The platform is used to construct an image that is professional, critical of opponents, and fiercely loyal to the party. The removal of personal content serves to harden this professional shell.
There are limitations to the study that the authors acknowledge. The analysis relied on a random sample rather than the full set of nearly one million tweets. While statistically valid, this approach might miss rare but important deviations in behavior. Funding constraints prevented the use of more expensive analysis methods on the full dataset.
The study also did not account for the specific political geography of each legislator. Factors such as gerrymandering could influence how safe a politician feels in their seat. A representative in a heavily gerrymandered district might behave differently than one in a swing district. The current study did not measure how these external pressures impact deletion rates.
Future research could address these gaps by using advanced technology. The authors propose using machine learning algorithms to classify the entire dataset of tweets. This would allow for a more granular analysis of political behavior on a massive scale. It would also help researchers understand if these patterns hold true over longer periods.
Understanding these behaviors is important for the voting public. The curated nature of social media means that voters are seeing a filtered version of their representatives. The emphasis on criticism and the removal of personal nuance contributes to a polarized online environment. By recognizing these strategies, citizens can better evaluate the digital performance of the people they elect.
The study, “More criticisms, less mention of politicians, and rare party violations: A comparison of deleted tweets and publicly available tweets of U.S. legislators,” was authored by Siyuan Ma, Junyi Han, and Wanrong Li.
