Are surveys of electoral preferences of active-duty military cause for concern?

In a democracy that wants and needs a non-partisan military, shouldn’t active-duty servicemembers avoid creating, in advance of a divisive and hotly-contested election (or any election), the impression that the armed forces has a favorite candidate?  Shouldn’t organizations avoid polling them and using the results?  Is a widely-publicized poll that allegedly reflects active-duty military voting preferences concerning?

The ethic and appearance of non-partisanship among the armed forces as a whole is vital.  Sure, military members have varied political views (that’s good thing in a diverse country that’s a democracy), but to poll and promote electoral preferences can serve to fracture the cohesiveness of the forces.  Moreover, the collective information can be used in an attempt to establish and sway political preferences – something in which the armed forces should have no part.

As I contend below, there are technical, philosophical and legal reasons – including an obscure statute that prohibits polling military members about voting preferences – why wariness of surveys of this sort is in order.


In late August, the Military Times reported results of a poll conducted “in partnership with the Institute for Veterans and Military Families (IVMF) at Syracuse University.” The article stated that the poll showed a “continued decline in active-duty service members’ views of President Donald Trump and a slight but significant preference for former Vice President Joe Biden in the upcoming November election among troops surveyed.”

(Military Times/IVMF has just released another poll, this one saying Trump backed by majority of veterans, but not younger ones.”  That poll is not discussed in this post, but since it involves veterans as opposed to those still serving, it doesn’t raise many of the concerns a survey of active-duty voting preferences does.)

The poll of active-duty troops has been cited widely by the media and by pundits (see e.g., here, here, and here).  Let’s discuss some aspects you may want to consider as you decide for yourself about it.

Can we be sure all the Military Times/IVMF poll participants were actually active-duty servicemembers? 

It seems Military Times emailed the survey to 132,185 of its subscribers, and about 3% responded.  Although it isn’t entirely clear, it appears the pollsters assumed responses were from “active-duty” persons if the respondents claimed that status, and if their response came from an email with a “.mil” extension.  Is that enough for you to conclude the respondents are active-duty military members?

Furthermore, decide for yourself: is such a small sampling of only those who self-identify as military members who happen to be subscribers to this particular publication truly representative of the armed forces as a whole?

Even assuming all the participants were active duty military, how accurate is a poll where 99.99% of troops did not participate?

Ms. Rosalinda Maury, director of applied research at IVMF, admits the poll “is not a perfect snapshot of public opinion within the entire military force.”  Among other things, the survey was based on just 1,018 responses from those who said they were in the military (there are 1.3 million troops serving on active duty).  If all the responses came from those currently serving, this means more than 99.99% of active duty troops did not participate in the survey.

Additionally, the Military Times concedes that the “average age of poll respondents was nearly 39 years.”  Is that important?  You decide: about two-thirds of the active duty force is age 30 or less, significantly younger than the average age of poll participants.

Are online polls accurate?

The Military Times/IVMF conducted its poll entirely online.  Experts suggest that such a methodology can be vulnerable to inaccuracies.  A 2018 essay in The Conversation (Online polls are everywhere: here’s why we should be wary before trusting them) explains how bias, polarization, and other factors can cloud results.  In addition, we live in a world where hostile nation-state actors with powerful cyber capabilities are determined to manipulate the election.  (Intelligence officials suggest Russia prefers Trump, but China and Iran “are against him.”).

Consider this history: when the Military Times and IVMF conducted a similar poll in 2016, the results were headlined: This poll of the U.S. military has Gary Johnson tied with Donald Trump in the race for president.”  Obviously, that proved to be inconsistent with the election’s actual outcome.  To his credit, Military Times journalist Leo Shane III said in reporting the recent results:

In the 2016 Military Times poll conducted in October, about 34 percent of troops surveyed said they planned to vote for a third-party candidate instead of either Trump or Clinton. Exit polls after the election showed that only about 5 percent of veterans and military members cast votes for candidates who weren’t one of the two major party nominees. (Emphasis added.)

Some additional information about IVMF

Military Times’ partner in the poll, IVMF, has a mission to “empower service members, veterans, and their families through actionable research, innovative programs, and insightful analytics.”  It does some great work: in 2017 I wrote (here) glowingly about its Entrepreneurship Bootcamp for Veterans with Disabilities (EBV), in which my alma mater, St. Joseph’s University, participates.

Decide for yourself, however, if this is important: The founder and Director of IVMF is Dr. Michael Haynie.  According to IVMF’s biography of him, he had a connection with the Obama/Biden administration.  Specifically, IVMF says:

In 2013, Haynie was appointed by the Obama Administration to serve as the Chairman of the U.S. Department of Labor’s Federal Advisory Committee on Veterans’ Employment and Training.  In 2015, the White House asked Haynie to serve again, this time as the vice chairman of a newly created Presidential Task Force, chartered to set the course for long-term reform at the U.S. Department of Veteran Affairs.

Furthermore, IVMF lists as a “strategic partner the Schultz Family Foundation, which IVMF says has donated $7.5 million to its coffers. The Foundation’s co-founder is Howard Schultz, a billionaire who ended his own campaign for the Democratic Party presidential nomination in September of 2019, reportedly because he didn’t want to be a spoiler if Joe Biden becomes the nominee.”  A few months later in April of this year, Mr. Schultz said this:

Joe Biden is a very good friend of mine over the years. And certainly, I’m going to be enthusiastic for his campaign and hope that he becomes the next president of the United States.

In September, Mr. Schultz formally endorsed Mr. Biden.

Are there legal issues with polling active-duty military via government emails?

Is using the military’s official “.mil” email system to convey a personal, partisan political opinion about a presidential candidate to a commercial media outlet (along with a non-governmental organization) improper?  The Department of Justice says “employees may not use the Internet or any other government equipment to engage in partisan political activities.”  There seems to be no reason to think that such a prohibition would not likewise apply to the Department of Defense.

U.S. law also says this:

18 U.S. Code § 596. Polling armed forces

Whoever, within or without the Armed Forces of the United States, polls any member of such forces, either within or without the United States, either before or after he executes any ballot under any Federal or State law, with reference to his choice of or his vote for any candidate, or states, publishes, or releases any result of any purported poll taken from or among the members of the Armed Forces of the United States or including within it the statement of choice for such candidate or of such votes cast by any member of the Armed Forces of the United States, shall be fined under this title or imprisoned for not more than one year, or both.

The word “poll” means any request for information, verbal or written, which by its language or form of expression requires or implies the necessity of an answer, where the request is made with the intent of compiling the result of the answers obtained, either for the personal use of the person making the request, or for the purpose of reporting the same to any other person, persons, political party, unincorporated association or corporation, or for the purpose of publishing the same orally, by radio, or in written or printed form.

In a recent article in the Vanderbilt Law Review (VLR) a law student theorized that the statute only prohibits compelled responses.  The author claims:

Because it only prohibits polling that “requires or implies the necessity of an answer,” it likely does not apply to standard volunteer-based surveys, and these days, there is no shortage of data on how servicemembers plan to vote in the lead up to presidential elections.

Opinions can vary, but I think that’s a misreading of the law.  In the first place, the notion that, as the author seems to suggest, ‘everybody does it’ hardly means such behavior comports to the law.  Studies show, for example, that at least “70 percent of American adults have committed a crime that could lead to imprisonment,” but that potentially widespread culpability doesn’t mean the underlying offenses are a nullity.

Secondly, and perhaps even more importantly, the VLR article’s conceptualization of the definitional section just doesn’t square with the plain text of the statute’s prohibitory language.  It uses the terms “whoever” and “any” poll, cutting against an interpretation that limits criminality solely to situations where participation is coerced.

The author speculates the statute is meant to stop military leaders who might “force subordinates to reveal political preferences.”  Maybe so, but that doesn’t mean other actors are excluded.  Indeed, how could limiting the application of the law to military leaders as the author hypothesizes fit with a broad word like “whoever” that the statute uses?  How can that be reconciled with the statutory language imposing liability on persons “within or without the Armed Forces”?

As to the definitional section, the words “requires or implies the necessity of an answer,” simply describes the “any request” that constitutes a “poll,” thus differentiating it from other rhetoric that may inadvertently elicit a response.  It shouldn’t be misconstrued as a deliberate effort to limit the proscription to compelled responses.  This may have been done to narrow the meaning of what inquiries and/or statements were prohibited so as to avoid becoming overbroad or otherwise violative of the First Amendment.  But that kind of tailoring doesn’t mean the statute only applies to coerced participation.

Thirdly,  there is no authority – and the student cites none (the footnote refers to the text of § 596) – which gives the statute such a limited interpretation.  What does exist instead militates the opposite way.  For example, the 1946 House report declared, without qualification or limitation, that under this law, “polls of members of the armed forces as to their choice of any candidate are prohibited.”  That certainly doesn’t support the notion that this provision has a limited application only to coerced polls.

Fourthly, and even more significantly, the Supreme Court has not given the statute the same reading the VLR article does.  In Greer v. Spock, the Court addressed a situation where the “commanding officer of Fort Dix rejected the request of respondent candidates for President and Vice-President to distribute campaign literature and hold a political meeting on the post.”  The Court upheld the Constitutionality of Army regulations that permitted the commander to do so.

In coming to that conclusion, the Court stated:

What the record shows, therefore, is a considered Fort Dix policy, objectively and evenhandedly applied, of keeping official military activities there wholly free of entanglement with partisan political campaigns of any kind.  Under such a policy, members of the Armed Forces stationed at Fort Dix are wholly free as individuals to attend political rallies, out of uniform and off base.  But the military as such is insulated from both the reality and the appearance of acting as a handmaiden for partisan political causes or candidates.

Such a policy is wholly consistent with the American constitutional tradition of a politically neutral military establishment under civilian control.  It is a policy that has been reflected in numerous laws and military regulations throughout our history. [Footnote 12]  And it is a policy that the military authorities at Fort Dix were constitutionally free to pursue. (Emphasis added.)

In Footnote 12 (mentioned above), the Court cited several statutes, including 18 U.S.C. § 596.  In explaining that provision of law the Court said simply but unequivocally that “[m]embers of the Armed Forces may not be polled by any person or political party to determine their choice among candidates for elective office.”  (Emphasis added.)

Obviously, the inclusion of “political party” as an example of a prohibited pollster makes it clear that the restrictive interpretation of the statute the VLR article’s author gives it is not shared by the Court.  What’s more, the context of the footnote is such that it reflects “the American constitutional tradition of a politically neutral military establishment under civilian control.”  The student’s blinkered interpretation would not serve that end.

Still, do poll participants face legal liability?  Very unlikely.  The statute itself does not extend liability to those who merely participate in the poll.  Nevertheless, military law does have concepts – the law of principals, accessory, and conspiracy among them – that can theoretically expand liability in various ways.

Concluding thoughts 

Allow me to be clear: I don’t advocate – and would not support – any prosecution of pollsters or participants with respect to any past polls, or any other litigation about them.

In fairness, these sorts of polls (and not just by Military Times/IVMF) seem to have proliferated in recent years, as the VLR author observed.  For its part, the statutory prohibition has lain dormant (no prosecutions to my knowledge) and it may have fallen legally into desuetude.

Why then should we be concerned now?  In the past the polls were mostly just interesting, but relatively benign footnotes on electoral machinations.  Unfortunately, in today’s profoundly-polarized and super-connected world, they are increasingly becoming powerful, partisan political cudgels to a degree they never were before.  We need to rethink their implications for the future.

Instead of legal action, let’s look ahead and do a better job at explaining why, in an era of such bitter partisanship, it is so important for the armed forces to remain, as the Court in Spock says, “wholly free of entanglement with partisan political campaigns of any kind.”

Spock is important as well because of its explicit statements as to the foundational importance of a politically-neutral military.  The Court recognized that polling by political parties and others is the kind of action that could erode the military’s ‘insulation’ against “both the reality and the appearance of acting as a handmaiden for partisan political…candidates” that the “American constitutional tradition” seeks to avoid.

One party may currently be celebrating the recent survey results, but that fact is that in the future another party might very well stand to benefit.  This is not to suggest that unofficial polls of military members on other topics ought never be taken, even though they may inevitably have some degree of political overtones (e.g., here).  There is merit to having the public know what its military is thinking on many topics.

Rather, the line ought to be drawn with respect to the fundamental act of any democracy: electoral preferences.   Americans should not have to even think about the possible implications for themselves of a survey of active duty military personnel that shows a preference by the nation’s most physically powerful institution for a candidate different from their own.

Such a situation will surely undermine the extraordinary, bipartisan public confidence the military currently enjoys.  Experts point out that by “remain[ing] above the partisan political fray” the military has managed to avoid the debilitation that has “consumed many once-trusted cornerstones of American public life.”

We should do everything we can to avoid having the public perceive its armed forces as anything but a non-partisan servant of all the people.  The poll in which all military members ought to participate is the official one on Election Day. 

Still, remember what we like to say on Lawfire®: gather the facts, examine the law, evaluate the arguments – and then decide for yourself!



You may also like...