Body Language in 2004 Presidential Debates
President George W. Bush and Senator John F. Kerry
First Debate—September 30, 2004
University of Miami, Miami, Florida
By J. J. Tecce, Boston College
Body language can reflect positive and negative emotional states. Frequency of blinking is a particularly sensitive measure of feelings. For example, increased blink frequency is associated with unpleasant feelings and decreased blink frequency is associated with pleasant ones. Stress is one type of negative emotional experience that can produce rapid blinking. Since nationally televised debates involve a high level of stress, it would be interesting to determine the patterns of blinking shown by President George W. Bush and Senator John F. Kerry during their first debate. Important advantages of these debates are their highly structured format and their occurrence in a well-controlled environment. In this way, the debate simulates a laboratory study.
I. Body Language
We usually communicate through the use of words whether spoken or written. However, there are limitations to how words can express feelings. For example, we learn to hold back the verbal expression of some emotions, such as intense feelings of anger or intimacy. And there are some emotional experiences which are so strong that words cannot fully describe them. Words also can be controlled and crafted to mask true feelings. Given these limitations, it is not surprising that we often reveal feelings through nonverbal expressions, which typically are not subject to voluntary control. Consequently, body language is potentially more significant than words in providing clues to the true experience of emotions.
While most parts of the body can be used to express feelings, the face seems especially revealing. A smile or eye contact can indicate interest and approval while a grimace or stare can reflect pain or anger. The eyes in particular have been called "windows of the soul" and have played an especially important role in nonverbal communications. For instance, folklore has it that not looking a person in the eye during conversation is a sign of weakness or lack of trustworthiness. And in Roman times, the Emperor Gaius screened gladiators for their courage on the basis of whether they blinked in the face of danger. Only two of 20,000 gladiators passed this test, since it is natural for humans to blink rather than stare without blinking in the face of a menacing gesture. In modern psychology, it is how fast or slow people blink that is one of the most revealing facial indicators of emotional distress.
II. The Blink-Feeling Connection
In the late 1960s, a number of us studying electroencephalography (EEG) were having difficulty with the occurrence of eyeblinks, which were getting in the way of accurate measurement of brain waves. In the early 1970s, it occurred to me that blinks might be important in and of themselves and were not just a pesky artifact that ruined our brain wave measurements. This interest in blinks eventually led to a body of research findings from my laboratory showing that blink frequency was increased during difficult stressful attention tasks where participants were responsible for multitasking but was decreased when pleasant feelings were produced by drugs.
Based on this laboratory research and work published by other investigators, I described in 1976 the following hypothesis (an educated guess about the association between two factors), which was two years later named the "hedonia-eyeblink frequency hypothesis" (Tecce et al., 1978). It is now called the eyeblink-hedonia hypothesis (Tecce, 1992).

“One hypothesis is that positive or pleasurable states (such as those produced by amphetamine) accompanying elevated arousal are usually associated with lowered eyeblink frequency, whereas negative experiences (such as annoyance, fatigue, fear, and intimidation) are usually associated with increased eyeblink frequency.” (Tecce & Cole, 1976, p. 198).

III. The Blink-Stress Connection
Stress is a discomforting experience. Consequently, as predicted by the blink-hedonia hypothesis, when people feel stressed they tend to blink more frequently. The average blink rate for someone relaxed is 10 to 20 blinks per minute (bpm). Talking increases the rate to about 20 to 25 bpm. When someone appears before a television camera, blink frequency increases to 30 to 50 times a minute because of "audience stress." As a blink rate climbs above 50, and especially when it gets above 70, it indicates an increased level of stress.
In this context, an interesting historical event occurred on August 8, 1974. Former President Richard M. Nixon gave his resignation speech on television and appeared calm, cool, and collected. But he showed episodic bursts of blinking above the normal 50 bpm cut-off. This rapid blinking during stress is called the "Nixon effect." This observation suggested that it might be worthwhile to study eyeblinks and stress in political situations, such as debates by candidates for the U. S. presidency.
IV. Previous Presidential Debates and Elections
As can be seen in Table 1, of the seven presidential elections for which televised debates are available, in six elections the candidate who blinked the most frequently during debates eventually lost the election. In the seventh election, that of 2000, the faster blinker (George W. Bush) lost the popular vote but won the electoral vote. This accuracy of blink rate in the prediction of the eventual winner outright was compromised in the 2000 election by other nonverbal expressions of candidate then Vice President Albert A. Gore, Bush's opponent. Gore showed behaviors that might have projected a negative image to viewer-voters, for example, looking down, double blinks, and blink bursts of three per second or more.
V. Presidential Debate of 2004
The first presidential debate was held on September 30, 2004 at the University of Miami, Miami, Florida. The participants were presidential candidates, U. S. President George W. Bush (Republican and U. S. Senator John F. Kerry (Democrat), and Moderator Jim Lehrer of PBS. In all but the final few minutes of each debate, candidates answered a series of questions. The question-and-answer period (Q&A) lasted from 9 PM to 10:25 PM. Then, a concluding statement of approximately two minutes was given by each candidate. Blink rate is expressed in blinks per minute (bpm). The normal range for blink rate on camera is 30 to 50 bpm. Consequently, a number over 50 indicates the strong likelihood that the candidate is experiencing discomfort and any number below 30 indicates reflects a relatively comfortable state.

A. Results
These findings are based on all observable blinks for the Q&A periods (which comprise 95% of each candidate's speaking time) and the concluding statements (each of which was slightly less than two minutes long). Table 2 shows blink rates for the Q&A periods and Table 3 presents the highest and lowest blink rates for answers given in response to specific questions by the moderator.
1. Blink Rate
As can be seen in Table 2, Bush blinked less frequently than Kerry during Q&A with a rate of 40 bpm, which is at the midpoint of the normal 30 - 50 range. Kerry had a rate of 51 bpm, which places him just outside the normal range. However, the story is quite different for the two-minute concluding statements. Kerry blinked only 48 times a minute compared to a rate of 109 bpm for Bush.
Table 3 gives peaks and nadirs for blink responses to questions asked directly by the moderator. As can be seen, the fastest blink response by Bush was 54 bpm, which occurred when asked whether he thought the Iraq war was worth the losses in human lives. The fastest blink response for Kerry was 58 bpm, which occurred when asked about what he considered to be mistakes made by Bush. The lowest blink rate for Bush occurred when asked about his assessment of the character of Kerry. The lowest blink rate for Kerry occurred when he described his agreement with the policy of pre-emptive war.
2. Other Body Language
In addition to blink frequency, other eye movements may have disclosed the psychological states of the candidates. Both Bush and Kerry showed gaze aversion (looked down), double blinks, and blink bursts of three or more per second. Bush looked down more than Kerry, while Kerry displayed more half-blinks (eyelids do not touch and do not completely cover the pupil), than Bush. Bush showed more spontaneous head movements than Kerry and showed more shifts in voice inflection. Kerry stood more erect and more still.
It is interesting to note that Kerry addressed the moderator in most of his responses to questions, hardly ever looking into the camera to address television viewers. Bush, on the other hand, divided his attention between the moderator and the camera in an attempt to address the television audience.

B. Comments
Bush had the advantage of blinking less frequently than Kerry for 95% of the debate, that is, during the Question-and-Answer format. Here, Bush's blink rate of 40 is within the normal range of 30-50 bpm while Kerry's 51 was slightly above this normal range. However, during the final two-minute statements, Bush blinked at a rapid rate and ended up with a high of 109 bpm in his concluding remarks. On the other hand, Kerry's blink rate dipped slightly from 51 to 48 during his two-minute conclusion.
The last minute surge in blinking by Bush in the present debate is reminiscent of a similar event in the third debate against Gore in 2000.

While Bush had a phlegmatic 19 bpm rate for almost the entire debate (Q&A), he spiked to 116 bpm in the final two minutes (concluding statement). This eleventh hour surge in blinking could be important. There is a phenomenon known as the recency effect: people tend to remember events that are recent. Thus, as far as blinking is concerned, Bush's swing-and-a-miss in the “ninth inning” of the third debate could leave a lasting impression on viewer-voters (Tecce, 2000).

In the current debate, Bush again had the advantage of a lower blink rate for the entire debate until closing statements. Then, Bush went from a rate of 19 bpm (Question-and-Answer segments) to 116 bpm in the final two minutes (Tecce, 2000). The fact that on two occasions Bush showed a marked increase in blink frequency when shifting between debate formats presents a paradox that is not easy to understand. One possible explanation may involve the separate but related psychological concepts of trait and state, that is, both personality and situational factors. It is highly likely that he is more comfortable with an informal conversational format, such as, during questions and answers. Hence, he had a normal blink rate of 40 bpm. By the same token, he is probably uncomfortable in giving the type of mini-speech that characterizes closing remarks. Add to this the abrupt transition that occurs after one hour and 25 minutes - going from a conversational mode to a speech mode - and we have a doubly stressful situation for Bush.
While Bush had an informal conversational style in answering questions, Kerry seemed more like a tutor in giving the moderator mini-lectures for each question. Kerry did not try to talk to the television audience. Hence his transition from nine 2-minute mini-lectures to a final 2-minute mini-lecture might have been an easy one for him to make. The down side of Bush's eleventh hour blink rush is that it was the lasting impression of body language for viewer-voters. While first impressions are psychologically important, the most recent impression often sticks in the observer's mind and can have a lasting effect.
An additional disadvantage in the body language of Bush is his tendency to look down more than Kerry. Studying notes and writing them are useful reasons to look down and both candidates did that. However, looking down also can indicate a need for gaze avoidance that serves as a break from increased tension and discomfort during visual engagement. Bush appeared to need this type of break in visual contact more than Kerry.
The fact that Kerry had a dialogue with the moderator for most of the debate suggested that he may have had a favorable rapport with him and was more comfortable engaging with him one-to-one rather than looking into the camera and addressing the television audience. Similarly, Bush may have had a less favorable rapport with the moderator and felt more comfortable looking into the camera and talking to the television audience. The latter interpretation certainly fits his frequent attempts to persuade the American people of his views.
Kerry showed fewer head movements than did Bush and stood somewhat immobile at the lectern. He gave the impression of being devoid of strong emotion and more controlled in his presentation. Bush show more spontaneous head movements and shifts in voice inflection than did Kerry, projected more emotion in his remarks, and appeared less formal than Kerry.
The peak and nadir blink rates in Table 3 suggest that some questions evoked more stress than others. For Bush, the question about whether the Iraq war was worth the loss of lives produced a peak response, 54 bpm. Accompanying this elevated blink rate were words of remorse, suggesting that he felt anguish about the loss of life. For Kerry, the fastest blinking (58 bpm) occurred when asked about his criticism of decisions made by Bush. Kerry might have been uncomfortable in having to make these comments publicly or had negative reactions to what he saw as unnecessary mistakes by Bush.
The slowest blink rate for Kerry was 39 bpm when he discussed his agreement with the principle of pre-emptive military action. It appeared that Kerry is comfortable talking about the policy of pre-emptive war and about his favoring it. The slowest blink rate for Bush was when he stated that Kerry was a patriotic soldier and a good dad, but that Kerry would be an inferior president and commander-in-chief. It seemed that Bush was comfortable in both his praise and criticism of Kerry.
While not directly related to eye movements, it should be noted that at the outset of the debate when the candidates met at center stage and shook hands, Kerry prolonged the handshake and delayed Bush from walking to the podium. This action might have been a cordial gesture that reflected Kerry's need to sustain a friendly gesture towards Bush or it might have been an aggressive attempt to intimidate Bush by establishing control and dominance for the evening. The action by Kerry is reminiscent of Gore's apparent attempt to intimidate Bush in the third debate of 2000:

Gore was also involved in another unusual expression of body language very early in the third debate at the Six-minute mark. While Bush was finishing up his Comments on patients' rights to medical care, Gore Walked to within a few feet of Bush and looked directly at him. Bush nodded in acknowledgment(the audience laughed) and continued his response. But Gore stayed close to Bush for 35 seconds. When Bush finished talking, Gore asked him directly … “What about the Dingell-Norwood bill?” (Tecce, 2000)

A final note.
The question arises as to whether any of this body language has an effect on a viewer's voting preference. The whole premise of nationally televised debates is to expose viewer-voters to the candidates' messages that related to their views and policies. What viewers hear during the debates is intended by the candidates to be the primary determinant of their voting preference. However, what viewers see may also be important, especially if they believe the spoken messages to be contrived and if the messages do not sufficiently differentiate the candidates. The mind abhors a vacuum and given a Bush-Kerry stalemate on the spoken word, viewer-voters might turn to nonverbal indicators of personal attributes. Research has shown that eye movements are a valuable source for this type of information. Furthermore, to the extent that body language of the candidates observed during the debates is representative of their behavior in general, the findings reported here may also offer a clue as to what voter perceptions might be on the campaign trail.

Tecce, J. J. (1992). Psychology, physiological and experimental. In McGraw-Hill Yearbook of Science & Technology (6th ed.) (pp. 375-377). New York: McGraw-Hill.
Tecce, J. J. (2000). Blink Behavior of George W. Bush and Albert A Gore: Presidential Debates of October 3, 11, and 17, 2000. Unpublished report.
Tecce, J. J., & Cole, J. O. (1976). The distraction-arousal hypothesis, CNV, and schizophrenia. In D. I. Mostofsky (Ed.), Behavior control and modification of physiological activity (pp. 162-219). Englewood Cliffs, N. J.: Prentice-Hall.
Tecce, J. J., Savignano-Bowman, J., & Cole, J. O. (1978). Drug effects on contingent negative variation and eyeblinks: The distraction-arousal hypothesis. In M. A. Lipton, A. DiMascio, & K. F. Killam (Eds.), Psychopharmacology: A generation of progress (pp. 745-758). New York: Raven Press.

Table 1

Blink Rates of U.S. Presidential Candidates and Subsequent Election Results
YearCandidatesFaster Blink RateLoser
1960Nixon / KennedyNixonNixon
1964Johnson / GoldwaterNo information availableGoldwater
1968Nixon / HumphreyNo information availableHumphrey
1972Nixon / McGovernNo information availableMcGovern
1976Carter / FordNo information availableFord
1980Carter / ReaganCarterCarter
1984Mondale / ReaganMondaleMondale
1988Dukakis / BushDukakisDukakis
1992Bush / Clinton / PerotBushBush
1996Dole / ClintonDoleDole
2000Bush / GoreBushGore *

* Bush was the faster blinker and lost the popular vote, but he won the electoral vote.

Table 2
Blink Rates (bpm) for President George W. Bush and Senator John F. Kerry:
The First 2004 U.S. Presidential Debate—September 30, 2004

Type of Format
CandidatesQuestion and Answer *Closing Statement
George W. Bush40109
John F. Kerry5148

* Based on 1,352 blinks by Bush in 34 minutes and 1,777 blinks by Kerry in 35 minutes.

Table 3
Questions Eliciting Highest and Lowest Blink Rates (bpm) for President George W. Bush and Senator John F. Kerry:
The First 2004 U.S. Presidential Debate—September 30, 2004

George W. Bush54—loss of lives in Iraq war28—character of Kerry
John F. Kerry58—criticism of decisions made by Bush39—policy of pre-emptive war