Public Opinion and the Vietnam War: How Race, Gender, and Partisanship Affect Views Toward Military Conflicts

  [1] Throughout the Vietnam War, the Women Strike for Peace movement demonstrated against US involvement in the conflict.[2] Violence erupted outside of the 1968 Democratic Convention in Chicago when anti-war protestors clashed with police, and tension over the party’s position on the war threatened to tear the party apart. The next year, Republican president Richard Nixon asked for the support of the “silent majority” of Americans who supported US efforts in Vietnam.[3]

Each of these examples illustrates two things: the controversial nature of the Vietnam War, and the importance of different aspects of identity – including race, gender, and political party affiliation – to citizens’ views on the war. While it isn’t likely to be their only concern, in an effective democracy, leaders faced with decisions about national security and military involvement abroad should consider citizens’ attitudes. But how do people form opinions about war? What underlying characteristics and aspects of identity impact how they think about these issues?

Public opinion on the use of military force is understandably complex. Existing research has focused on how demographic and political factors impact the way citizens think about foreign policy decisions. For example, in his book In Time of War, Adam Berinsky shows how people tend to rely on cues from partisan elites to form opinions about US military involvement abroad, much like they do for domestic issues.[4] Many scholars, including Pamela Conover[5], Virginia Sapiro[6], and Richard Eichenberg[7] have done important work revealing gender differences in views toward the use of force and have put forward possible explanations for these differences. And in Rising Wind, Brenda Gayle Plummer examines the engagement of Black Americans in mid-20^th century foreign policy.[8]

However, it is also important to consider how these characteristics interact and intersect, and we shouldn’t necessarily expect the impact of particular aspects of a someone’s identity to be constant. Instead, the impact may increase or decrease based on social and political circumstances. For example, if partisanship influences people’s views on foreign policy, then we might expect people to change their views on a war when the political context transforms, possibly when party control of Congress or the White House changes hands. People often have to weigh a variety of concerns when deciding how they feel about a particular issue, and they may weigh these concerns differently based on context.

We can explore some of these dynamics using public opinion data regarding the Vietnam War. Because control of the White House changed from Democrat Lyndon Johnson to Republican Richard Nixon as the war was ongoing, Vietnam is a useful case for examining how changes in the political environment affected citizens’ views. Additionally, the controversial nature of the war led to appeals to many aspects of identity, as Martin Luther King’s Speech and the Women Strike for Peace movement illustrate.

 The analysis below uses data from Gallup polls conducted between 1964 and 1973 and available through the Roper Center, combined with survey data from the American National Election Studies collected between 1964 and 1972. Each survey asks participants whether they thought the US involvement in the Vietnam War was a mistake, offering insights into how people were thinking about the conflict and how these views changed over time. Merging these sources results in a dataset of more than 40,000 survey responses that spans almost a decade, allowing for an expansive and inclusive look at public opinion toward US involvement in Vietnam.

One important benefit of such a large sample size is that it allows for a more precise analysis of differences in attitudes by race. Unfortunately, each individual survey contains a small number of nonwhite respondents. However, combining all of the polls from this period provides a much larger number of responses from this group, allowing for a better understanding of the attitudes of Black Americans and other racial minorities.

            To begin to understand the underlying factors impacting views toward the Vietnam War, it is helpful to first analyze the entire sample together. When looking at data from throughout the conflict, what characteristics were associated with more versus less support for the war? Figure 1 below shows the results of a multiple regression analysis. Positive estimates show that a particular characteristic is associated with increased opposition to the war, after controlling for the other variables in the model. The value of the estimate, when multiplied by 100, gives the expected percentage-point change from that variable. For the variables measuring partisanship, the reference group is independents. Red bars surrounding each estimate represent 95 percent confidence intervals, so variables are statistically significant when the bar is non-overlapping with the zero line.

Figure 1. Regression Analysis – Opposition to Vietnam War

Note: Dependent variable is whether the respondent stated that it was a mistake for the US to send troops to Vietnam. Point estimates are regression coefficients from a linear probability model. Ribbons represent 95 percent confidence intervals. In addition to the variables in the plot, the regression model controls for age and education, along with a control for the “other” racial group and fixed effects for year and polling organization.

      Similar to what other researchers have found, women are more likely to oppose the war than men.[9] However, the difference is relatively modest. These results imply that, all else equal, women are only 1.5 percentage-points more likely than men to oppose US involvement in Vietnam. This may be due to the interaction of gender with other variables such as partisanship (more on this below). By contrast, differences by race are substantially larger. Black respondents are much more likely to oppose the war than other racial groups – 9.5 percentage points more likely than white respondents.

Turning to the impact of partisanship, this analysis shows that members of the two parties held different views on the war. While the Republican party tends to have the reputation of being the more hawkish party on foreign policy, Republicans were actually more likely to voice opposition to the Vietnam War than Democrats. This may be because the majority of the surveys were conducted between 1964 and 1968, when Lyndon Johnson, a Democrat, was president. Democratic voters may have been reticent to oppose a president of their own party, while some Republicans may have voiced opposition to the war due to their political differences with the president that was overseeing the conflict.

We can further explore this last point by running separate analyses for each presidential administration. In Figure 2, the plot on the left includes data from surveys conducted while Lyndon Johnson was president, and the plot on the right shows the results using data from Richard Nixon’s presidency.

Figure 2. Regression Analysis by Presidential Administration

Note: Dependent variable is whether the respondent stated that it was a mistake for the US to send troops to Vietnam. Point estimates are regression coefficients from a linear probability model. Ribbons represent 95 percent confidence intervals. In addition to the variables in the plot, the regression model controls for age and education, along with a control for the “other” racial group and fixed effects for year and polling organization.

Here we see that women were more likely to oppose the war than men early in the conflict, when Johnson was president, but during the Nixon administration, differences by gender largely disappear and are no longer statistically significant. However, differences between Black and white respondents remain stark during both presidential administrations.

Perhaps the largest difference between the Johnson and Nixon administrations is how members of each party view the war. During the Johnson administration, Democrats were much less likely to oppose the war than Republicans, but during the Nixon Administration the gap between Democrats and Republicans evaporated. This highlights how people often rely on their party identification as a “shortcut” for deciding how to think about political issues.[10]

Finally, we can analyze the survey responses separately for Democrats and Republicans to see if gender identity or race seem to impact people’s views differently based on how they interact with partisanship. As I mentioned above, given that most previous research finds a large and pervasive gender gap in views toward the use of military force, it is interesting that the differences by gender in this data are fairly small. Figure 3 below shows the results of the analysis by party.

Figure 3. Regression Analysis by Party Affiliation

Note: Dependent variable is whether the respondent stated that it was a mistake for the US to send troops to Vietnam. Point estimates are regression coefficients from a linear probability model. Ribbons represent 95 percent confidence intervals. In addition to the variables in the plot, the regression model controls for age and education, along with a control for the “other” racial group and fixed effects for year and polling organization.

Interestingly, the gender gap in views toward the Vietnam War is present among Democrats but not Republicans. This may be due to the turmoil within the Democratic party over how to view the war – without clear cues from party leaders, Democrats may instead rely on other aspects of their identity, including gender, to form attitudes toward the war.  

Similarly, racial identity has a large impact on views toward the war for Democrats, but it’s not clear that this is true for Republicans. However, there are less than 300 Black Republicans in the dataset, so any analysis of how views differ by race among Republicans should be interpreted with caution due to the small sample size available. As is still the case today, most Black respondents in these surveys identified as Democrats.

There is much more to explore, but the results presented here begin to reveal the complex interaction between different aspects of identity that takes place when citizens form opinions toward military conflicts. This work is part of a larger project I am undertaking with Cornell Professor Douglas Kriner on this topic. Our ongoing research is focused on how gender, race, and party identification have impacted opinion formation toward the use of military force in the post WWII-era, from the beginning of the Korean War in 1950 to the 2021 withdrawal from Afghanistan.

By Roper Center Kohut Fellow Aaron Childree.