Kayla N. Jordan, UT Austin

Former New York City mayor, Michael Bloomberg, is a recent addition to the Democratic debate stage (and Democratic party). This week was only his second appearance in the debates as he is currently self-funding his campaign and made the decision to not run in the first four primary states. To begin to catch up with the analysis done on the other candidates, this post focuses on where Bloomberg stands on three important metrics covered in past posts: thinking style, clout, and authenticity.

Thinking Style

In the very first post of this primary season, I discussed that the way candidates use words like pronouns, articles, and prepositions is a marker of analytic versus narrative thinking style. Some people naturally lean more toward an analytic, formal thinking style while others lean more toward an intuitive, informal style. From the data thus far, Bloomberg looks like the most informal of the bunch (though Klobuchar scores a close 2nd). His linguistic patterns are simple, intuitive, and not overly concerned with details and connections between ideas. There is little in the way of hierarchical, logical arguments. Rather, he focuses more on making statements and detailing his record (e.g. “I’ve trained for this job for a long time and when I get it, I’m going to do something rather than just talk about it.”).



In another post, I examined the metric of clout (or confidence). Past work shows that people high in status or confident in their position tend to talk more about others while those low in status or lacking confidence talk more about themselves. With Buttigieg, Bloomberg is one of the least confident of the remaining candidates potentially due to his newness to the national Democratic party establishment. Bloomberg is often defensive or tentative in staking his place on the stage (e.g. “I don’t remember what they were, so I assume if it bothered them, I was wrong, and I apologize.”, “I’ve met with Black leaders to try to get an understanding of how I can better position myself”).



In the first post of the season, I also analyzed the candidates’ authenticity. We know from past work that when people are being honest and straightforward, they tend to talk more about themselves and their present circumstances. Politicians are often not the most authentic or honest individuals, so the differences between them can be interesting. Once again, Bloomberg stands out as the most authentic candidate (with Buttigieg a close second). For example, while his answers may not satisfy his opponents or Democratic voters, he didn’t evade questions of his past behavior (e.g. “We let it get out of control. And when I realized that I cut it back by 95%. And I’ve apologized and asked for forgiveness”, “But right now, I’m sorry if she heard what she thought she heard or whatever happened. I didn’t take any pleasure in that.”).


An Interesting Comparison

Taken together, Bloomberg’s linguistic style is quite like one former primary candidate: Donald Trump. Like Bloomberg, Trump during the 2016 Republican primary debates had a very informal thinking style, was lower in clout than his opponents, and came across as highly authentic (for a politician). Beyond their shared business background and alleged past improprieties, Bloomberg and Trump seem to share a similar style of thinking about and communicating ideas. While Bloomberg might not win the Democratic nomination, his similarity to Trump makes him an interesting figure in the process.

Contact: kaylajordan@utexas.edu


Kayla N. Jordan, UT Austin

In the Las Vegas debate ahead of the Nevada caucus, the 2020 Democratic presidential candidates clashed in arguably the most heated debate so far. Michael Bloomberg, a late comer to the 2020 primary and the Democratic party, hit the debate stage for the first time. Joe Biden struggled with disappointing finishes in Iowa and New Hampshire while Warren, Buttigieg, and Klobuchar fought to unseat Sanders as the new front-runner. Amidst these complex dynamics, the candidates lost much of their usual congeniality attacking each other on both policy and personality.

While I usually analyze the candidates across debates, here I want to focus solely on this most recent debate. Rather than say something about the traits of the candidates, this post will consider how the candidates are reacting and interacting to each other and the ups-and-downs of the primary process. To capture the fieriness of the last debate, I examine the extent to which the candidates used positive and negative emotion. To capture the recent contentiousness of the primaries, I analyze how socially focused the candidates were as a proxy for how much the candidates may value (or not value) unity and teamwork.


Many methods have been devised to measure emotion from language. Here I keep with the method used in previous posts and use the positive and negative emotion dictionaries from LIWC2015. Positive emotionality is measured by words such as excellent, benefit, and respect (620 words in the dictionary). Negative emotionality is indicated by words such as fault, doubt, and unsuccessful (744 words in the dictionary). The graph below shows the percentage of each candidate’s total words that fell into one of these categories, and there are three main points of interest.

1. The least emotional candidates were Warren and Klobuchar which bucks the traditional stereotype of women being more emotional than men. Linguistically, the female candidates kept their cool last night more than their male counterparts.
2. Despite being the target of several attacks, the most positive candidate was Bernie Sanders. Klobuchar and Buttigieg were also on the more positive side. All three of these candidates used over twice as many positive emotion words versus negative emotion words. Despite defending against and instigating attacks, these candidates talked more about their own positive qualities and not just about the opponent’s bad ones.
3. Warren was the only candidate to use more negative (versus positive) emotion words. Leading off the night with attacks on Bloomberg’s record on women, Warren did not hesitate to call out issues with other candidates’ past actions or policy proposals. What is interesting is that while she spent the most time on the attack, Warren was not overly emotional in her attacks. Overall, she used the fewest emotion-laden words, and generally came across as pointing out her opponent’s weaknesses in a relatively matter-of-fact manner.

Emotion Words

Social Focus

Like emotion, the extent to which people are concerned with their social relationships can be illuminated by the words they use. With LIWC2015, social concern is measured by words such as help, talk, friend, and colleague (756 words in dictionary). Despite the general combative mood in the debate, candidates did differ in how socially focused they were and fell in three groups.

1. Warren and Biden used the most social words of the six candidates. Despite any animosity between the candidates, Warren and Biden both spent time focusing on how they need to work together to accomplish the goals they all agree on (e.g. ‘we have to change the tax code’, ‘We need to make an investment to level the playing field.’).
2. Sanders and Bloomberg were in the middle somewhat socially focused but generally focused more on conflicts between them. This can be seen in some of the interactions they had with each other. After agreeing that the rich needed to pay more tax, they devolved into personal attacks against each other (e.g. ‘the best known socialist in the country happens to be a millionaire with three houses’, ‘maybe we should also ask how Mayor Bloomberg in 2004 supported George W. Bush for president”).
3. Klobuchar and Buttigieg used the fewest social words. Rather than unity and shared goals, they tended to focus more on why they themselves were voters’ best option. This can be seen in the extended back and forth they had pointing out each other’s flaws (e.g. “You have been unusual among Democrats”, “Are you trying to say that I’m dumb? Are you mocking me here, Pete?”).

Social Words

Looking Forward

Unlike past primary election cycles, the outcome of the Democratic primary is far from clear, and the probability of no candidate securing the nomination is unusually high. With so many candidates still in the race and the relatively high agreement between them on many issues, determining what does separate the candidates, like their communication styles and psychological traits, is potentially more important than ever. The analysis in this post demonstrates that one way the candidates differ is in how they handle conflict. It is hard to say which style would make the best candidate or best president, but hopefully the analysis can help voters determine which candidate might best match their ideal leader.

Contact: kaylajordan@utexas.edu

Kayla N. Jordan, UT Austin

While I have been focused on the Democratic candidates lately, this seemed like a good time to check in on the (presumptive) Republican candidate, Donald Trump. This week Donald Trump became the second president in modern history to give a State of the Union (SOTU) address during an impeachment trial. Unlike Bill Clinton, however, Trump is facing reelection. In this post, I analyzed Trump’s recent SOTU address compared to his first three as well as Bill Clinton’s at the time of his impeachment. Beyond exploring possible psychological impacts of impeachment, the analysis gives a peek into the mindset of Donald Trump as he gears up for his reelection bid.

In facing potential removal from office, two linguistic measures are potentially interesting: emotional tone and I-words. As discussed in previous posts, emotional tone gives a sense of a person’s general outlook by comparing their use of positive emotion words (e.g. happy, respect, love) to their use of negative emotion words (e.g. anger, hate, death). People high in emotional tone come across as upbeat and optimistic while those low in emotional tone seem pessimistic and dark. Presidents facing impeachment might be expected to become more pessimistic given their circumstances.

The use of I-words (e.g. I, my) can be a bit of a double-edged sword. On one hand, greater use of I-words is associated with honesty and authenticity. On the other hand, greater use of I-words is also associated with depression and insecurity. In their SOTU addresses, presidents, particularly those being impeachment, likely want to sound as honest and sincere as possible, but may at the same time be conveying a sense of insecurity. So what can we learn from Trump’s (and Clinton’s) SOTU addresses?

The graph below shows how Trump’s tone has shifted across his four SOTU addresses. For comparison, Clinton’s first SOTU address after he was impeached by the House as well as his three preceding SOTU addresses are included. Both Trump and Clinton show a similar pattern across their addresses with a decreasing positivity over time and an increased optimism in the SOTU address following their impeachment. A couple of possible explanations exist for this counter-intuitive optimism. First, both Clinton and Trump may have simply been optimistic about their chances of surviving the impeachment process. The second explanation is strategy; even in the face of the vote against them, they believed they were still good for the country and wanted to remind people of the positive things they’ve done (e.g. “our economy is the best it has ever been”) and plan to do (e.g. “making sure that every young American gets a great education”).


Similar patterns are also found in Trump and Clinton’s use of I-words. For both presidents, I-words increase in the SOTU address following their impeachment. Both presidents are generally thought of as ‘authentic’, but the increase in I-words after their impeachments could further signal insecurity and a lack of confidence following the Congressional rebuke. While never directly mentioning the impeachment processes, both presidents made sure to highlight their contributions and goals (e.g. “I will send to Congress a plan… [Clinton]”, “I moved rapidly to revive the U.S. economy… [Trump]”). Such a self-focus may reflect a need to bolster their confidence after being hit with impeachment.


Despite showing similar patterns in their responses to impeachment, Trump is in a unique position. Like Clinton, he is very, very unlikely to actually be removed from office. However, even if acquitted, Trump faces re-election, so it is interesting to consider how impeachment may impact his campaign style. In the 2016 election, he was relatively optimistic in the primaries, but much darker and more pessimistic in the general election. Consistent with his latest address, he tended to use I-words at high rates. So on the campaign trail, we will likely see a return of the authentic (but perhaps insecure) Trump, but will the sense of optimism he brought to the SOTU remain or will he return to the doom-and-gloom candidate of the 2016 general election?

Contact: kaylajordan@utexas.edu

Kayla N. Jordan, UT Austin

In the seventh and most recent debate, six of the remaining Democratic primary candidates battled it out as the primary season finally ramps up to the Iowa caucus. While not much new substance was learned during this debate, the candidates’ use of language can illuminate insights into their psychological states. In this post, I examine where each of the candidate’s attention is and what implications that has by looking at seemingly unimportant words: pronouns (specifically I and We).

When people use I (or me or my), they are demonstrating a greater focus on themselves and their own circumstances. On the negative side, research has shown greater use of I-words to be associated with depression and insecurity. However, greater I-word use has also been found to be related to honesty and authenticity. Hence, political candidates showing such self-focus through the use of I may be seen positively by voters as straightforward and personal. At the same time, they may be also seen as lacking confidence.

When people use we (or us or our), they are demonstrating a focus on others and the groups they belong to. Additionally, greater use of we-words can signal leadership and confidence. Political candidates who use we more may be perceived as more attuned to and concerned with the needs and attitudes of the nation (and in the case of the primary, members of their party). Like I-words, the use of we-words has potential drawbacks. An excessive use of we may come across and inauthentic and dishonest. So, where is the focus of the remaining 2020 Democratic primary candidates?

The Me Candidates. Biden and Klobuchar stand out as the most self focused candidates. They (especially Klobuchar) use significantly more I-words than we-words. In the debates, these candidates have focused more on building a case as to why they should be the candidate. They want voters to know what they’ve done (e.g. ‘I’ve been in the U.S. Senate for over 12 years.’, ‘I led that effort.’) and what they think are the best solutions to today’s problems (e.g. ‘I have a plan for that.’, ‘I lay out how I’d pay for that.’). While their high use of I-words likely contributes to voter’s perceiving them as authentic, it also may betray insecurity. Biden, even though he is leading in most polls, continues to struggle to rise decisively above the other candidates, and Klobuchar seems to have little to no chance of victory. Such uncertainties seem to bleeding through in their debate language.

The We Candidates. Three candidates stand out for their group-focus: Warren, Sanders, and Buttigieg. They are focused on the concerns of the group (both the nation and the Democratic party) and what can be done to address these concerns. While they are making the decisions, they see themselves as part of a collective and confidently assert their leadership credentials to solve the country’s problems. They talk about the problems and solutions of the nation as affecting and involving everyone (e.g. ‘We have a problem’, ‘our campaign has the strongest grassroots movement’, ‘we could be using those dollars for something else’). Though these candidates might come across as more confident leaders, they also run the risk of inauthenticity. For example, all three of these candidates have been questioned on their ability to understand and relate to the concerns of African American voters perhaps making their use of we less than inclusive.

The Other Candidate. One candidate does not fall neatly in either of these categories. Tom Steyer, who has not made it into any posts so far but continues to qualify for debates, is equally self and other focused using I- and we-words at nearly identical rates. Steyer pushes his own experiences and ideas (e.g. ‘I understand how America interacts with other countries’, ‘I proposed a wealth tax’), but also talks about the concerns and responsibilities of the whole (e.g. ‘I know we can do it.’). Steyer seems to balance the best of both worlds conveying a sense of confidence and leadership along with a connection to the collective. While he still lags in the polls, perhaps this style will help him connect with voters as the primaries progress.

I and We words

Note. The y-axis depicts the average percentage of the words in the candidate’s debate language that are either I-words (e.g. me, mine, myself) or we-words (e.g. we, us, our). Averages were computed across all seven debates since June 2019. LIWC2015 was used to analyze the debate language. 

Summary. The remaining Democratic frontrunners are split between the me-candidates (Biden and Klobuchar), the we-candidates (Warren, Sanders, and Buttigieg), and the in-between candidate (Steyer). Biden and Klobuchar’s use of I suggest honesty but insecurity, while Warren, Sanders, and Buttigieg’s use of we suggest confidence but inauthenticity. Steyer’s use of both may help him avoid the pitfalls of a focus on one or the other. The primary contest is far from decided, and hopefully this new psychological look at the candidates helps readers gain a better understanding of these individuals as they decide who to vote for.

 Contact: kaylajordan@utexas.edu

Kayla N. Jordan, UT Austin

We all know that one person who is willing to put it all on the line. What makes that person willing to take any risk for a potential pay-off, compared to others who conservatively weigh potential benefits against potential risks? Given the implications of such differences in how people make decisions in the political domain, this post examines how the 2020 Democratic candidates approach risk and reward and what implications that might have for their leadership styles.

In psychology, a long line of research has examined risk versus reward focus. Proposed by Higgins and others, regulatory focus theory provides insight as to the implications of being more focused on reward or more focused on risk. Reward-focused individuals are often risk-takers seeking accomplishment and able to make big, progressive choices. They consider the benefits of an action, but might not really consider all the risks. Risk-focused individuals, on the other hand, tend to be more cautious seeking safety and security. They might consider the benefits of an action, but more heavily weigh the risk in order to prevent negative outcomes.

Linguistically, we can get a sense of where a person’s focus might be by looking at the words they use with LIWC2015, the psychological text analysis program used in previous posts. People who are reward-focused tend to use more words like achieve, optimistic, wager, and better. People who are risk-focused tend to use more words like avoid, defense, danger, and disadvantage. In the world of political leadership, differences in risk-reward orientation can have enormous implications for how leaders make decisions, ranging from how they handle a crisis to what policies they pursue. So which of the current candidates are risk-takers, risk-averse, or somewhere in between?

The Risk-Takers

Perhaps unsurprisingly, two of the candidates who have been most reward-focused are Sanders and Warren. As two of the most liberal candidates, they push big ideas that carry the potential for enormous benefits (but also potentially severe risks). From their health care plans to their tax policies, they focus on the potential positive outcomes of their proposals glossing over the potential negative consequences. Take for example the attacks Warren faced on her health care plan. While her opponents tried to focus on a potential negative outcome (i.e. a tax increase), Warren stuck to the positive outcome of decreasing costs (e.g. “what I have committed to, is costs will go down for hardworking, middle-class families.”).

More surprising, might be that Klobuchar also falls into this group. While as a moderate she tends to have more modest plans, she looks to the benefits of her proposals. Often talking about her work in the Senate, Klobuchar focused on the benefits of practical legislation. For example, in the debate over tech monopolies, while many other candidates discussed the potential negative outcomes of either action or inaction, Klobuchar put a positive spin on her proposal (e.g. “Start talking about this as a pro-competition issue”).

The Risk-Avoiders

Three candidates stood out as being risk-focused: Booker, O’Rourke, and Yang. The language of these three candidates suggest that they are focusing on avoiding negative outcomes. When weighing proposals and decisions, they may be more likely to err on the side of caution and lean toward the safe choices. Notably, the risks these candidates identified were about the other candidates’ policies or behaviors. For example, Booker repeatedly mentioned the risks of infighting amongst the candidates (e.g. “we’re not talking about the clear and existential threat in America”), O’Rourke attacked Warren’s tax policy (e.g. “Senator Warren is more focused …pitting some part of the country against the other”), and Yang tried to focus the jobs debate on the dangers of automation (e.g. “Saying this is a rules problem is ignoring the reality”).

The Middle Grounders

The remaining four candidates (barring Gabbard and Steyer who may have made the debate stage but not this post) are somewhere in the middle. Three candidates, Castro, Harris, and Biden, were lower than average on both risk and reward focus. For these three candidates, the potential outcomes of their plans or proposal were seemingly less important, and instead they relied more on their knowledge and experiences. For example, Harris often simply laid out facts (e.g. “There are states that have passed laws that will virtually prevent women from having access to reproductive health care.”) or relied on anecdotes and personal experience (e.g. “as a former prosecutor, I know a confession when I see it.”). Joe Biden took a similar approach while also spending a lot of time defending himself and his past decisions (e.g. “Look, my son did nothing wrong. I did nothing wrong.”).

Pete Buttigieg may be the most interesting case as he was above average on both risk and reward focus. Unlike the other candidates, he seems to more completely balance risk and reward. Ensuring positive outcomes is important, but potential negative outcomes should be considered and mitigated. For example, in responding to the plans of other candidates, he argued that grand plans are good but they should be tempered with what is safe and practical (e.g. “Yet there are some here on this stage who say it doesn’t count unless we go even further…We have an opportunity to do the biggest things we’ve done”). Potentially more than the other candidates, Buttigieg recognizes the importance of dreaming big, but also the importance of recognizing what actually has a reasonable chance of success.



The top Democratic candidates differ widely in how they approach decision making. Some, like Warren and Sanders, are risk-takers seeking to maximize benefits. Others, like Booker and O’Rourke, are more circumspect seeking to minimize bad outcomes. Still others, like Biden and Harris, look more to their own experience rather than the potential outcomes of their decisions. Only one candidate, Buttigieg, seems to really balance the costs and benefits of policies. While your own risk-reward orientation may make one of the candidates more attractive than others, hopefully this post gives you a better sense of how the candidates might approach decisions if elected.

Kayla N. Jordan, UT Austin

The 2020 Democratic candidates for president met again on the debate stage this week for the third time. Unlike the previous debates, this debate was a single night with just the top 10 candidates. As the field has begun to be winnowed, now seems like a good time to talk about the candidate’s confidence. Here I am going to consider two facets of confidence: clout and certainty.

Clout and Status

Clout gives a sense of how candidates may be feeling about their status and power in the party. The linguistic metric for clout was developed from both experimental and real-world studies of leaders. When someone is a leader or high status, they tend to use we-words and social words (like ally, friend, group). Followers or those with lower status tend to use I-words more along with negations (like no, not, never) and swear words. Of the 10 front-runners, O’Rourke and Yang have been highest in clout with Warren a close third. Sanders, Castro, and Buttigieg were lowest in clout.

Consider the differences in how O’Rourke and Castro answered a question about the racial divide in the country. O’Rourke responded decisively and confidently framing the solution as a collective responsibility, “We have to be able to answer this challenge…But we will also call out the fact that we have a white supremacist in the White House and he poses a mortal threat to people of color all across this country.” Alternatively, Castro was less confident and more self-focused in his response, “I’m proud that I put forward a plan to disarm hate … We need to root out racism, and I believe that we can do that.” These examples show a difference in a leader bringing a group together to solve a problem and a leader who is thinking what they can do themselves. One caveat to note here is that being low in clout in the primaries is not necessarily bad for a candidate. Evidence from other recent primaries suggest that being somewhat humble may be more attractive to primary voters.

Note. Clout scores range from 0 (least confident) to 100 (most confident). In a political context where everyone is relatively high status, the effective range is generally 60 – 100. Individual scores in this graph are averages across the three debates so far.

Certainty and Processing

Certainty gives a sense of how sure candidates are about their ideas and positions. It is clear that some candidates are still processing their thoughts and positions about issues using words like should, could, think, know, and believe. Other candidates know exactly where they stand on issues and clearly state what their positions and plans are. The graph below shows where the front-runners stand on this metric. The most certain candidates are Harris and O’Rourke (note low scores are more certain in this case). The least certain is Buttigieg with Biden and Castro right behind.

Consider how Harris versus Buttigieg handled questions about trade. Harris very clearly and confidently laid out her policy, “My trade policy, under a Harris administration, is always going to be about saying, we need to export American products, not American jobs…Look, we need to sell our stuff…That means we need trade policies that allow that to happen.” Buttigieg hedged more providing a less than clear alternative to current policy, “I would have a strategy that would include the tariffs as leverage, but it’s not about the tariffs…when what we know is that the tariffs are coming down on us more than anybody else and there’s a lack of a bigger strategy.” Candidates still have time to iron out their ideas, but some, like Harris, already stand out as having clear, decisive positions.

Note. High scores in this case indicate greater processing and uncertainty while low scores indicate more certainty. Scores represent the percent of words in the text indicative of cognitive processing average across the three debates.

Final Thoughts

It is still very earlier in the primary season. Buttigieg and Castro may be struggling with confidence in their status and ideas now, but they still have time to change. Likewise, O’Rourke and Yang are confident now, but if they fail to improve in the polls, they may lose that confidence. As the primary season progresses and we get closer to the Iowa caucus, we will check back in to see how the candidates’ confidence changes as they try to win the nomination.

Kayla N. Jordan, UT Austin

This week the 2020 Democratic presidential candidates met for the second round of primary debates. In this post, I turn to a different psychological process: motivation. One of the most well-known theories of human motivation was proposed David McClelland, who posited three main drives of human behavior: power, affiliation, and achievement. Importantly, McClelland’s motivation theory has been extensively applied to the political realm by David Winter and others. These motivations not only describe how individuals (including political leaders) orient to the world around them, but also predict a variety of outcomes including presidential greatness ratings and behavior during international crises.

It is important to note that these motivations are not independent (a candidate can be both power motivated and affiliation motivated). For this reason, I will go through the motivations one at a time explaining what each motivation means, the implications of each motivation for leadership, and what debate language of the 2020 front-runners (note Andrew Yang has replaced Kirsten Gillibrand as a front-runner for this post) reveals about their potential motivations.

The Power Motive

Individuals motivated by power are highly attuned to status and influence. The power-motivated individual wants to know who’s in control and how their status compares to the status of others. Linguistically, those motivated by power tend to use words like leader, follower, demand, and weak. Politically, research on American presidents has shown that power motivation has both strengths and weaknesses. Power-motivated presidents are more likely to be judged by historians to be great, effective presidents, but they are also more likely to enter a war (e.g. LBJ and FDR).

The figure below shows how power-motivated each candidate is compared to the others based on their language in the debates so far. The most power-motivated candidates are Harris, Sanders, and O’Rourke while Biden, Yang, Buttigieg, and Warren are least power-motivated. One interesting way the power-motivation was seen was in how candidates referred to one another. For example, Harris nearly always referred to other candidates by their titles (e.g. “Senator Bennet’s plan”, “When Vice President Biden…”, “agree with Governor Inslee”). Biden, on the other hand, often addressed other candidates directly and used their first names (e.g. “Bernie acknowledges it”, “I found that Julian”).

Note. Y-axis represents average percentage (across the two debates) of words used that are related to power.

The Affiliation Motive

Individuals motivated by affiliation are highly attuned to social relationships. The affiliation-motivated are looking for friends and allies and want to develop close relationships. Linguistically, those motivated by affiliation tend to use words like help, family, ally, and we. Politically, researchers have found affiliation-motivated presidents can be better negotiators. However, they also seem more susceptible to scandal and corruption. Richard Nixon and George W. Bush are two recent affiliation-motivated presidents who cared greatly about personal relationship even if they weren’t always good at them.

The most affiliation-motivated candidates were O’Rourke and Yang as shown below. Both of these men used the word ‘we’ liberally throughout the debate often focusing on unity and coalition building (e.g. “Whatever our differences, we know that, before we are anything else, we are Americans first, and we will ensure that each one of us is well enough and educated enough and paid enough to realize our full potential [O’Rourke].”; “I’m building a coalition of disaffected Trump voters…conservatives, as well as democrats and progressives [Yang].”).

The candidates least motivated by affiliation were Harris, Castro, Klobuchar, and Sanders (potentially because unlike relatively newcomers, O’Rourke and Yang, they already have a base of supporters). Of these four, Harris stands out. Whereas other front-runners often focused on their policies and finding common ground, Harris directly clashed with other candidates at several points (e.g. “Vice President Biden, you’re just simply inaccurate”, “with all due respect”, “I’m done with the conversation.”).

Note. Y-axis represents average percentage (across the two debates) of words used that are related to affiliation.

The Achievement Motive

Individual motivated by achievement are highly attuned to success and failure. The achievement-motivated are ambitious and driven, striving for accomplishment and success. Linguistically, those motivated by achievement tend to use words like win, lose, excellence, and earn. Researchers have found that achievement-motivated presidents are often highly idealistic, but at the same time, often end up being ineffective lacking political skill and emotional maturity. While usually not useful in a presidential context, achievement-motivation has been found to be effective in many business settings (which may make achievement-motivated candidates seem attractive to voters).

The candidates highest in achievement motivation were Klobuchar, Booker, and O’Rourke while Sanders, Buttigieg, and Yang were relatively low in achievement motivation. Consider when the debate turned to the topic of electability. Where Sanders had an abstract, philosophical answer (e.g. “we need to have a campaign of energy and excitement and of vision”), Klobuchar focused on success (e.g. “I think how we win an election is to bring everyone with us. And, yes, I have won in a state every single time statewide. I have won those congressional districts that Donald Trump won by over 20 points.).


Note. Y-axis represents average percentage (across the two debates) of words used that are related to achievement.

Trends and Takeaways

So what does this all mean? Well, there are a few key trends worth noting. First, some candidates like Biden and Castro didn’t score highly on any motivation. They, along with Buttigieg and Warren, are low to average for each motivation. Unlike others, these candidates seem to have more balanced motives rather than a single, strong motivation. A balance between two or more motivations can be useful. Some past work suggests that a balance between affiliation and power motivations can mitigate the downsides that individuals of high in one or the other can struggle with.

One candidate stands out as being high on all three motivations: Beto O’Rourke. Throughout the debate, O’Rourke relied on different motivations often depending on what the topic was. He seemed to be motivated by affiliation when discussing electability, power when discussing race issues (e.g. the wealth that we have built, the way we became the greatest country … on the backs of those who were kidnapped and brought here by force”), and achievement when discussing education (e.g. “earn that associate’s degree, realize your full potential…of being able to better yourself so that you can better this country”). O’Rourke demonstrates that, for some, their motivation can shift depending on the context.

Finally there are the candidates that primarily show a single motivation. For Sanders and Harris, that motivation is power. Both candidates through their words demonstrate a focus on who has power, who doesn’t, and how they can change it. For Sanders, that’s Wall Street and the 1%; for Harris, it’s Trump and the Republicans. For Yang, his motivation is affiliation. As a political outsider, he is focused on making a base for himself by gathering friends and allies to further his cause. For Booker and Klobuchar, the motivation achievement. They just want to get things done and rack up the successes. The more singular focus of these five candidates might be useful in understanding how they would approach the office of the presidency.

For those of you in a hurry, here are the bullet points:

• Sanders and Harris are power-motivated. They are ambitious and concerned with influence and prestige.
• Yang is affiliation-motivated. He is looking for friends and is concerned with social relationships.
• Booker and Klobuchar are achievement-motivated. They want to get things done and are concerned with success.
• Everyone else (Biden, Buttigieg, O’Rourke, Castro, and Warren) show some combinations of motives. They balance concerns simultaneously or switch between, depending on the context.

Hopefully, the above analysis has helped understand the 2020 front-runners just a bit better. It is still early in the primary season. Candidates still have time to grow and change how they are thinking about and communicating their political messages, and I will continue tracking them as the election progress to see how they develop. Check in after the next round of debates (Sep. 12 & 13) for another look at the presidential hopeful through their words.

Contact Info: kaylajordan@utexas.edu

Kayla N. Jordan
University of Texas at Austin

While candidates for the 2020 Democratic presidential nomination have already been campaigning for months, the primary season hit a new phase this week with the first debates between the candidates. Here, in the first of several blog posts about the 2020 election, I use computer-based text analysis methods to start to get a sense of the psychology of the candidates. Rather than their positions or policies, the goal of these posts will be to understand the candidates as people by considering questions like how are they thinking, how are they relating to other people, and how are they communicating their ideas.

The system the majority of the analyses will rely on is a program developed in the Pennebaker Lab at UT Austin called Linguistic Inquiry and Word Count, or LIWC2015 (available for academic research at liwc.net or for commercial application at Receptivit.com). LIWC can analyze any text to determine the percentage of words in the text indicative of negative emotion and cognitive processing and 80+ other dimensions. LIWC has been used in hundreds of studies in multiple disciplines ranging from psychology to business to medicine to political science to computer science. To learn more about some of these studies, check out this link.

So what can LIWC tell us about the 2020 presidential candidates? As the election season unfolds I will be looking at many dimensions including motivations, confidence, and time orientation, but for now I want to look at three central psychological dimensions: thinking style, emotional tone, and authenticity. Also, rather than cramming in all 20 candidates, I am going to focus on the ten candidates who, at the moment at least, seem to have the best chances of securing the nomination: Joe Biden, Bernie Sanders, Elizabeth Warren, Beto O’Rourke, Cory Booker, Kamala Harris, Julian Castro, Pete Buttigieg, Amy Klobuchar, and Kristen Gillibrand.

Thinking Style

People naturally differ in the ways that they think and communicate ideas. On one end of the spectrum, there are people who are very analytic, logical thinkers. Analytic thinkers organize their ideas in formal, hierarchical ways focusing on concepts and ideas. On the other end of the spectrum are people who think in narrative, intuitive ways. Narrative thinkers organize their ideas more informally, often telling stories and focusing on people and actions. Linguistically, analytic thinkers use more articles and prepositions while narrative thinkers use more pronouns, auxiliary verbs, adverbs, negations, and conjunctions. Check out this link or this link to learn more about the research on analytic thinking. So where do the Democratic candidates fall on this spectrum?

The Analytic Thinkers: Of the 10 candidates, 3 stand-out as the most analytic: Buttigieg, Gillibrand, and Sanders. These candidates talked about their policies and ideas in logical ways. They relied more on facts and figures focusing on concepts and details rather than stories and anecdotes. For example, take Bernie Sanders. When talking about his health care plans, he laid out the problem, gave statistics, and stated his proposals rarely making his positions personal.

The Narrative Thinkers: On the narrative end of the spectrum, Klobuchar, Biden, and Castro stand out. Unlike their analytic counterparts, they tell stories and anecdotes focusing on people and actions more than concepts and abstractions. Rather than making structured arguments, these candidates communicates their ideas in looser, more informal ways. Compare Biden’s discussion of health care to Sanders’. Where Sanders gave impersonal arguments, Biden connected his policy plans to his families’ personal experiences with cancer.

The In-Between Thinkers: Rather than falling on either end of the spectrum, four candidates fall somewhere in the middle: Harris, Booker, O’Rourke, and Warren. Drawing on both styles, these four candidates use both formal, logical structures as well as stories and personal experience. For example, when Elizabeth Warren talked about gun violence she used not only statistics and structured policy proposals but also anecdotes from her time on the campaign trail talking to voters.


Note. Analytic thinking scores are standardized composite scores ranging from 0 (most narrative) to 100 (most analytic).

Emotional Tone

Emotion has become a central feature of political campaigns and has been studied in a variety of ways such as fear appeals and negative advertising. Here, however, I focus on what the use of emotional language might say about a person’s general outlook. A more optimistic, upbeat outlook is indicated by positive emotional words such as love, respect, and happy. A more pessimistic outlook is indicated by the use of negative emotional words such as anger, death, and hurt. What were the candidates’ emotional outlooks in the first debates?

The Optimists: Three candidates were high in positivity: Gillibrand, Harris, and Klobuchar. These three women candidates all presented optimistic, upbeat messages. Gillibrand and Klobuchar, in particular, come across as affable during the debates occasionally mixing in humor and provide hopeful views of the future.

The Pessimists: Four candidates were on the negative side: Buttigieg, Sanders, Biden, and Booker. These candidates presented less positive images painting more pessimistic views of the future. Bernie Sanders was the clearest representation of negativity. For Sanders, there are many serious problems that must be addressed immediately and decisively to avoid a future filled with doom and gloom.

The Realists: The three remaining candidates fell in the middle: Warren, Castro, and O’Rourke. For these three candidates, their outlook is mixed with serious, urgent problems needing solutions, but with a hopeful outlook that such problems are solvable.


Note. Tone scores are standardized scores ranging from 0 (most negative) to 100 (most positive).


Politicians are often portrayed as less than honest, hence a final important dimension to consider is authenticity. Individual high on authenticity come across as honest and straightforward while those low on authenticity come across as evasive and impersonal. Linguistically, studies have found that authentic individuals tend to use more I-words, present-tense verbs, and relativity words (e.g. old, far, here) and fewer she-he words and discrepancy words (e.g. could, should). How authentic are the Democratic candidates?

The Most Authentic: The most authentic candidate was Pete Buttigieg closely followed Sanders and Castro with Booker is a somewhat distant fourth. In their debates, Buttigieg and Castro, in particular, came across open and personal giving straightforward, clear answers to questions. Warren, Klobuchar, Biden, and Gillibrand fall in the middle of pack, but were slightly closer to the authentic candidates than to the inauthentic candidates. These four generally came across as straight-forward and personal but with moments of distance and evasion.

The Least Authentic: The least authentic candidate was Beto O’Rourke with Kamala Harris in a close second. O’Rourke was particularly striking in the debates. Despite the large amount of attention he received in his 2018 Senate Run, O’Rourke during the first debate came across as distant and impersonal often giving responses seeming robotic and rehearsed.


Note. Authenticity scores are composite standardized scores ranging from 0 (least authentic) to 100 (most authentic). In political contexts, the range is generally limited with scores effectively ranging from 0 to 50.

Going Forward

These are just a few initial insights into the 2020 Democratic candidates, and as we gather more data throughout the primary season, we will gradually gain a clearer sense of who these candidates are and how they might behave as leaders. That said, there is one broad takeaway from this first look at the 2020 candidates. While many of the candidates have similar (or even identical) policy positions, the analysis presented here shows the candidates have very different personalities and communication styles. The ideological similarities between candidates in a primary election can make choosing between them difficult, and I hope that the psychological views on the candidates that I will be presenting in these blog posts can provide additional information for voters who want another perspective of the candidates.

Check back later for further insights of the 2020 Democratic presidential candidates as I analyze the debates and other (linguistically) interesting campaign events this election cycle. For more information on this project, contact Kayla Jordan (kaylajordan@utexas.edu). For more information about LIWC, check out this link.

Kayla N. Jordan
University of Texas at Austin
Kavita Vedhara and Pamela Pepper
The University of Nottingham

Less than 2 months ago the UK’s prime minister, Theresa May, made the unexpected decision to hold a snap election. The elections will take place this week and has ostensibly been called in response to division in Westminster that she believes threaten future Brexit negotiations. As the UK begins the process of leaving the EU, the voters are expected to elect a leader from one of the two main parties in the country: the Conservatives led by Theresa May and the Labour party led by Jeremy Corbyn. After the surprising Brexit vote and U.S. election of Donald Trump, elections around the world have drawn attention for growing populism and isolationism. So as election day grows closer, we look at what the language of May and Corbyn reveals about them and the types of leaders they may be.

While Theresa May refused to take part in traditional debates, both May and Corbyn participated in three Q&A events interacting with both media figures and regular voters. These took place between May 22 and June 2 and will be the basis for our analysis. Furthermore, to put this in context we compare these Q&A events to pre-election interviews by recent prime ministers, Tony Blair and David Cameron, as well as to the 2016 general election debates between Donald Trump and Hillary Clinton. Specifically, we look at how these candidates compare on three dimensions: thinking style, clout, and authenticity.

Thinking Style

When trying to better understand political candidates, one interesting facet to consider is their thinking style. Do they approach problems in an analytic, logical way or are they more intuitive and narrative in their style? In recent American politics, Donald Trump has stood out as being an exceptionally intuitive, informal thinker both as a candidate and as a president. So where do the current British candidates stand on this measure?

Between the two candidates, Theresa May is less analytic than Jeremy Corbyn. However, both are more analytic than recent British prime ministers, Tony Blair and David Cameron, as well as recent U.S. presidential candidates, Donald Trump and Hillary Clinton. While May lays out her ideas in a somewhat simpler fashion compared to Corbyn, no matter who wins the election, the next British prime minister is likely to have a different thinking style than recent UK prime ministers and a vastly different style to the current American president, Donald Trump.

analytic graph

Confidence and Clout

Another interesting facet of political leadership is how confident they are. We know from past research that those who confident and high status tend to use more we-words (e.g. we, our) and social words (e.g. friend, ally, group) while using fewer I-words (e.g. my, mine), negations (e.g. no, not, never), and swear words. Since calling for the election when she had relatively high levels of support, Theresa May has seen the election become very close with some polls suggest that the gap between Labour and Conservative is shrinking. So has the narrowing of the polls impacted either May or Corbyn’s confidence?

Despite the slides in the polls, May has remained more confident than Corbyn. Over time, May’s confidence has remained stable while Corbyn became more confident as election day draws closer (70.6 to 75.5). However, it is interesting to note that compared to previous UK prime ministers, both May and Corbyn are more confident in talking about the issues they face. May, in particular, resembles recent American candidates more than past British PMs.

clout graph


A final dimension worth looking at is authenticity. When confronted with questions, are the candidates sincere and straightforward or evasive and impersonal? Authentic individuals tend to use more I-words, present-tense verbs, and relativity words (e.g. old, far, here) and fewer she-he words and discrepancies (e.g. could, should). So how do May and Corbyn compare?

Both May and Corbyn are relatively inauthentic particularly compared to recent PMs, Blair and Cameron. When faced with difficult questions such as broken promises by May or IRA connections with Corbyn, both candidates have sounded evasive and distant. Interestingly, May and Corbyn are quite similar to Donald Trump when it comes to authenticity.

Throughout his campaign and after his election, Trump generally came across as a authentic and personal (even if he was often objectively incorrect). However, during the general election, following the scandal with the leaked Access Hollywood tape, Trump was less authentic in the later debates. While not so extreme, both May and Corbyn also have criticisms they have to endure and these appear to have impacted their authenticity during these events.

authenticity graph

The Big Picture

So what do we know about the two major candidates, Theresa May and Jeremy Corbyn? May is a more intuitive thinker and more confident in her position but relatively inauthentic. Corbyn is also relatively inauthentic, but is much more of an analytic thinker; and while he is showing less confidence than may, this is growing. While in U.S. elections intuitive and confident individuals tend to do better in elections, the situation and electorate in the UK could favor a different type of candidate so we shall have to wait to see whether the more confident, intuitive May will be victorious or the more analytic, increasingly confident Corbyn will prevail.

No matter which candidate ends up as prime minister it is interesting to note that the next British prime minister will be a departure from other recent PMs. Compared to Blair and Cameron, May and Corbyn are both more analytic, more confident, and less authentic than their predecessors. In fact when it comes to confidence and authenticity, May and Corbyn are more like their recent American counterparts, Donald Trump and Hillary Clinton, than recent British PMs. While there are obvious difference between May and Corbyn, they both are likely to be different types of leaders than their recent predecessors.

Helpful Resources:

Newman, M. L., Pennebaker, J. W., Berry, D. S., & Richards, J. M. (2003). Lying words: Predicting deception from linguistic styles. Personality and social psychology bulletin, 29(5), 665-675.

Kacewicz, E., Pennebaker, J. W., Davis, M., Jeon, M., & Graesser, A. C. (2014). Pronoun use reflects standings in social hierarchies. Journal of Language and Social Psychology, 33(2), 125-143.

Pennebaker, J. W. “The secret life of pronouns: How our words reflect who we are.” New York: Bloomsbury (2011).

Pennebaker, J. W., Boyd, R. L., Jordan, K., & Blackburn, K. (2015). The development and psychometric properties of LIWC2015. liwc.net

Pennebaker, J. W., Chung, C. K., Frazee, J., Lavergne, G. M., & Beaver, D. I. (2014). When small words foretell academic success: The case of college admissions essays. PloS one, 9(12), e115844.

Kayla N. Jordan and James W. Pennebaker
University of Texas at Austin

After just over a month in office, Donald Trump gave his first address to Congress. The content of Trump’s address was similar to past presidents’ State of the Union (SOTU) addresses with a focus on recent accomplishments and plans for the upcoming year. But what about the style of the speech?

Unlike his inaugural address which was linguistically similar to Trump’s typical language, his recent address was more analytic and less authentic than normal. Similar to his RNC acceptance speech, Trump’s first SOTU was heavily shaped by a speechwriter. Although the content of the address overlapped considerably with Steve Bannon’s recurring themes of fear of outsiders, the linguistic markers were quite similar to the language of Stephen Miller in a recent interview on Face the Nation.  Despite Miller’s probable role in the address, Trump’s latest speech is useful in understanding overall trends in the presidency and where Trump fits in.

Decline in Analytic Thinking

Similar to trends in inaugural addresses, SOTU speeches are generally highly analytic and formal, but have been becoming less so over time. Starting with Woodrow Wilson and Franklin Roosevelt, modern presidents have adopted a more informal, narrative style in their annual addresses to Congress. Trump’s first address continued this trend with his level of analytic thinking on par with the last 5 presidents.

The decline in analytic thinking signals a shift in how presidents are thinking about problems and presenting their ideas. Trump, like other recent presidents, laid out his ideas in a simpler, more straightforward way than past presidents. Going forward, Trump will likely rely more and more on offering simple, intuitive solutions and ideas to the problems he faces.


Rise in Confidence

The language presidents use can show how confident and self-certain they are as leaders. Confidence or clout is indicated by more we-words and social words and fewer I-words, negations (e.g. no, not), and swear words.

Whereas analytic thinking has decreased over the last century, clout has increased. Around the same time presidents began becoming less analytic, they also started to exude more confidence. Presidents have increasingly approached these addresses to Congress with confidence and certainty. Trump is the most confident so far, but is still similar to recent presidents. Trump and other modern presidents are decisive and confident in their plans and proposals.


The Big Picture

In their SOTU addresses, presidents have been becoming more confident and less analytic. These trends show that presidents are changing how they are thinking and interacting with lawmakers and the American people. Administration after administration, the yearly SOTU addresses are laying out simpler and less nuanced world views with bolder more decisive proposals. Faced with complex, hard-to-solve problems, clear and easy solutions are likely more appealing to present to an increasingly polarized Congress (and electorate).

While Trump is often seen as a significant departure from presidential norms, in many ways, he isn’t all that different than other modern presidents. Rather than being an extreme outlier, Trump is part of long-term trends. He is a more confident, intuitive thinker, but Obama and Bush were as well. The content of what Trump is saying may be abnormal, but the style is typical of recent presidents.