Printer Friendly Page Public-Interest Polling, When Americans are Ready to Use Force

2.2.6  “When Americans Favor the Use of Force”

 by Alan F. Kay


 © 1999, 5/16/99, (mod. 2/9/00)


International Journal of Public Opinion Research,

Vol. 12, No.2, Summer 2000, pp. 182-190WAPOR,

World Assn. of Public Opinion Research

Abstract -- There are six screens or tests that have proven useful and robust in understanding when the American people would approve the use of force. These tests are general statements, independent of many of the specific circumstances of a deteriorating international situation. The validity of these tests have been verified by repeated in-depth survey research. The tests are independent of whether the situation may lead to a minor military incursion or a large regional war, independent of the region or countries involved, and valid since the end of the Cold War and probably earlier. Subject to important conditions explained in the text, there will be no majorities for the use of force without passage of all of the first three tests, making them necessary conditions for majority support for the use of force. To the extent that the remaining three tests are also passed, the percent of Americans favoring the use of force rises. With sufficiently high support for these three tests in various combinations, support for the use of force reaches high consensus levels, over 90%.  US political leaders are generally aware of the six screens but tend to dismiss or misunderstand some of them and their importance, leading to military involvements that were improperly pursued (not as successful as they might have been) or perhaps should not have been undertaken at all, and more recently seem to have made the world a more dangerous place for all.
If confronted with a deteriorating international situation, how do the American people form opinions on when, if ever, to favor the use of force? Too often, unfortunately, the public is not privy to the structural and political context which brings a nation to the brink of such military confrontations.

For example, in the NATO action in bombing Yugoslavia after Slobodan Milosovic's genocidal "ethnic cleansing" in Kosovo, few average citizens could have been aware of all the options over the preceding five years which might have prevented Milosovic from pursuing his aims. These structural options range from (1) the US paying up its dues and arrears to the United Nations and supporting its restructuring efforts to strengthen the Security Council, (2) supporting instead of weakening the International Criminal Court (which if it had been functioning might have authorized Interpol to arrest Milosovic for war crimes in Bosnia and put him on trial), (3) rebalancing the funding of US foreign policy between diplomatic versus military spending so as to restore the full potential of preventing and negotiating conflicts to political and strategic options such as the inclusion of Kosovo in the Dayton Accords and the funding of indigenous civic and humanitarian groups to counter Milosovic's use of "hate" radio and TV and other misinformation, rumor-mongering and fomenting ethnic rivalries and conflict. However, as in the case of Kosovo, when precious lead time and proactive options for prevention are bungled or lost, the US public was presented, as is often the case, with a policy fait-accompli.

The consequences have been ominous. NATO, the military organization whose principal function had been coordination of the western allies to counterattack or preempt a Soviet invasion of Europe, was in the process of enlargement by inclusion of the Czech Republic, Poland, and Hungary.  The US led intervention in Kosovo allotted Russia only a secondary role. The two developments engendered a new Russian attitude of ignoring US-led moralizing and taking their security needs into their own hands. The day after the US invasion of Kosovo was launched, the Russian Duma, instead of ratifying Salt II as expected, rejected the treaty. The treaty called for a 50% cut in US and Russian strategic nuclear weapons.  Instead, despite all logic to the contrary, both countries now have as many strategic nuclear weapons on hair-trigger alert ten years after the end of the Cold War as at their Cold War peak. India and Pakistan have joined the nuclear club, and the world lives with the reality that any head of state or some missile launch officers in a stupid act could kill hundreds of millions of people in the world’s major cities in half an hour. In the last year Russian independence has not been only in nuclear matters. On the “conventional” weapons front the Russians took action too. Based on undisclosed evidence for its necessity and contrary to Western importuning, they reinvaded and brutally destroyed Chechnya in early 2000.

Circumstances and conditions under which the public in any country is ready to use force has been a broad and basic question of survey research that is widely believed to have remained murky if not unresolved for many years. This is not entirely correct.

A massive project begun in 1987 by the non-profit Americans Talk Issues Foundation (ATI) focused on finding what people want for governance (policy, legislation, regulations), by high quality, in-depth survey research. ATI developed a concept of using small balanced teams of issue and polling experts to design and supervise the surveys. The teams were required to consider a wide range of policy choices, including and going beyond what the two minority parties, the Republicans and the Democrats, offer as they routinely tussle with each other in Congress and in the Administration to control the national agenda. Reliable techniques for finding and confirming the percent of Americans supporting various policies were developed and tested in many circumstances over the years, leading to an approach to polling, called public-interest polling, described in the book, Locating Consensus for Democracy -- a Ten-Year US Experiment [1]. I and many others, beginning with the founding fathers of random sample polling in the '30s, have helped to define and refine what today is called public-interest polling. In a decade, ATI conducted about twenty telephone surveys, sample size a thousand plus, covering the subject of national and international security in many different ways, which among other things explored the question of when Americans support the use of force.

In the first twelve surveys, when questions on when to use force were tested by ATI, little was found of use except brief specific descriptions of situations from which no reliable, or particularly useful, generalizations emerged, as explained on pp. 77-78 of [1]. Some ATI research results based on scenario building, the examination of past uses of force, possible future uses of force, defense of allies, and responding to attacks by others, appears on pp. 78-80, in individual questions where the factual data presented to describe the situation in each case is too sparse and too specific to uncover what ATI later learned.

The breakthrough began during the five months of operation Desert Shield, when allied forces confronted the forces of Iraq. Americans, expecting war day after day, were focusing on this situation and ready to choose from tough batteries with a wide range of policy offerings, and take responsibility for their choices in follow-up questions.

At the successful conclusion of the Gulf War, President Bush told us that we were entering a New World Order.  He never defined what that was, but many others did.  Former Secretary of State, Henry Kissinger, and CIA head, Robert Gates, are examples. ATI explored support for these and other policies for a New World Order in surveys #15, #16, and #17. Strong support was found for the US and the UN (or a broad group of allies sharing the costs and the risks) together handling acts of aggression by rogue dictators engaged in a variety of heinous crimes.  When ATI asked whether the US and the UN should take clear action to stop this behavior, in some cases, 80 to 90% support was found for doing so.  It was amazing to discover that if the next question posited that the only clear action that could succeed at stopping the behavior eventually turned out to be military force, perhaps of enormous size, support did not drop below 70% in a number of cases.  This remarkable result was explored in depth in ensuing surveys.  Testing this finding in hypothetical situations in several parts of the world did not reduce support very much.  The concept was called "the People's Military Policy" (see [2]), and ATI continued to test it further.  Somewhat embarrassed by this finding that nobody wanted, ATI attempted to poke holes in it in different ways, some of which turned out not to be legitimate. Moreover a variation of the People's Military Policy, was uncovered later in ATI surveys #21, #23, #25, and #28, when consensus-level support was found for the UN (this time alone, without the US) and (very important) with due process to arrest and bring to justice leaders who were clearly responsible for more-or-less the same heinous crimes tested earlier.  To my knowledge, these findings have not previously been pulled together for a short paper on when Americans are ready to use force.

In a wide variety of deteriorating international situations, ATI research established that, based on what they know and, over time, learn of the situation, Americans apply what can be thought of in a collective, statistical sense as six tests, screens, standards, or filters (here usually called "screens" and described later in this paper) to produce a level of support for intervention that is at consensus levels (70% to 90%) if the first four have been fully met.  If also the fifth and/or sixth are widely perceived to be met, support for the use of force will rise to the upper consensus levels of 80-95%.  If any one of the first three screens are not met, support for the use of force will be found only among a minority of adults.  If the first three are fully met, support is still at a minority level unless the fourth is also fully met when it will attain a consensus level (70-90%). If either the fifth or sixth screen, or both, are also met, support again rises into the upper ranges of consensus at 80-95%. In some cases described in the 4th screen, support may reach a sub-consensus majority if the fourth screen is partially met.

Twelve years of public-interest polling has taught ATI that these six screens are virtually universal regardless of the specific situation, regardless of the part of the world where the conflict is developing or the countries involved, and whether we are considering the situations of today or others of the past ten years and in many instances even earlier than that.  They apply to use of force, ranging from a swat team foray to all-out war. In a given situation support levels can be more for the former than for the latter.  It is probably true that not every possible future situation will be judged by these same six screens, but a concept with nearly universal validity for such a long period is worth knowing and testing against new situations as they develop.

The way these results have been obtained and confirmed by survey research has been carefully documented in [2].  There has been no survey data that I have seen since the book was published in 1998 that does not conform to "the People's Military Policy." To appreciate this point, bear in mind that it is valid polling that determines whether the public has passed through a screen, not the reported opinion of pundits or elites. The two may be quite different (see [3]).

People are not ready to go to war lightly.  A lot has to happen before developments reach a point where any one of the six tests has been passed.  In the modern world, news stories and commentary on breaking developments can get the thinking of large numbers of Americans to pass through any of these screens, even all of them, very rapidly.  Remember Pearl Harbor. Much more commonly, a situation that ends in a military intervention is years in the making with numerous twists and turns.

Posing a situation to arise wherein a foreign military force -- not responding to an attack by the US, but deliberately and apparently out of the blue -- attacks our military forces or our own territory and population, then support for a military response from the US rises to a majority, even high consensus level, sometimes over 90%.  However, this overwhelming majority is split as to whether our counterattack should be all-out or proportional in magnitude to the initial attack.  If the foreign attack is major, then a significant fraction, but still only a minority wants a counterattack that, even without using weapons of mass destruction, is far in excess of proportional.  Compared to numerous situations in recent years wherein the question of initiating the use of force has been confronted, the possibility of an unprovoked military attack against us has been so rare that we do not in this paper consider further this case, which has its own distinct, and more belligerent patterns of public responses.

Each of the six screens described here is passed when high quality, unbiased polling, particularly public-interest polling, shows that to be so. An effort by political leaders and mainstream news media to present and frame the developing situation with the intention of persuading people that their presentations are correct, balanced and fair, is central to the process of coming to judgment by the public on the use of force.  At the same time, some influential leaders and information sources may describe the situation as not meeting the tests of these screens. Particularly when this is the case, the public tends to be dubious of both the leaders and the media, but even when this is not the case, the public, often burnt in the past by biased news, remains dubious of the developing conventional wisdom by its leaders and media.

Political leaders are generally well aware of the process of getting people through stages of agreement on the need for using force and they also know some of the substance of the screens.  As explained in the Kosovo example at the beginning, leaders have not yet categorized and defined the screens and apparently are not aware of all of them. In their efforts to persuade people to agree with their views, they often omit key steps, and lose support. They do not get the public shepherded through the screens.

Another factor is that the President, to some degree even alone and certainly with the support of Congress, can override at one stage or another the lack of majority support for the use of force.  If the country arrives at a point where force is being used and troops are, or soon likely to be, put in harm's way, then that very fact may be enough for people to accept the de facto situation.  Majority support for military intervention may thereby emerge.  The failure to bring people through all the screens jeopardizes the stability of that acceptance and invites internal opposition from other leaders. If leaders kept in mind the desirability of shepherding people through all of the screens, or whenever this was not happening, of postponing or finally deciding against the use of force, then the country would be most unified and the President and supporting leaders would be in the best possible situation.

The Six Screens

1. Rogue Leaders

The leader (or key leaders) of a country (or faction) against which we are ready to consider taking military action must be perceived as contemptible and villainous. Often, they are considered rogue dictators, like Saddam Hussein, Manuel Noriega, and Slobodan Milosovic. The importance of this first screen is this. If a whole country, or leaderless mobs, or looting and pillaging ethnic (or other) factions, are committing despicable acts, even genocide, but no leaders are known or can be identified, the first test is not met. A majority of Americans will not favor use of force in this case. Meeting test #1 is clearly a necessary condition.

Commentary.  How the "ordinary" people of the country, not the few close to the leader, are perceived is largely irrelevant.  The people are held harmless and often perceived by many in the US as victims of their leaders, even when they appear to be obeying or politically supporting them. There is a "sympathy factor" as we too may feel victimized by these rogues. When there is a decision to use force against a country, sometimes not just their leaders but all citizens of the country quickly become demonized in the minds of the US public. Demonization largely comes after the war has begun, stimulated by a propaganda campaign (or pro-US mainstream news media coverage), mostly in the case of all-out war, such as WWII.

2. Guilty of Heinous Crimes

The second screen is that the leader(s) must not only be identifiable rogues, they must clearly have committed the most heinous of crimes. Crimes perceived to be in the most heinous category may change from time to time, but they have not changed much in the past 12 years.  They have included: sponsoring terrorism, acquiring weapons of mass destruction, grossly violating human rights including but not limited to genocide, invading and occupying by force a neighboring country, and perhaps a few others. Many but perhaps not all possibilities have been tested by survey research and many do not make the "most heinous" list. The list, for example, does not include "refusing to sell vital materials, like oil, to the US and other developed countries".

3. Exhaust Non-Military Means

The third screen is that we first use all measures short of military force -- whatever seems possible under the circumstances, in a thorough and reasonable way -- aimed at stopping the commission of the crimes perceived as heinous. This is true even if the public is not aware of all the long-term structural options, such as mentioned in the Kosovo case.  Before the use of force, the public feels that we must first be convinced that we have thoroughly exhausted diplomatic, negotiating, and sanction efforts, with no signs of success, before a majority of Americans will favor supporting a military intervention.  A devastating and unexpected attack, like Pearl Harbor, alone exhausts all such efforts, but in every other situation in the 20th century, there has been considerable opportunity for the utilization of means short of war, often several years of opportunity.  This pre-intervention process must include (a) the approval of Congress or an effort by the Administration to obtain Congressional approval for the short-of-war process being followed, (b) readying our armed forces, our people, and our economy for whatever may be necessary for a military intervention and, most important, (c) seeking the support and cooperation of allies as needed for negotiations, diplomatic efforts and particularly sanctions and, in the process, seeking some degree of support, acquiescence or neutrality from countries not accepting the ally role. It is important that leaders make clear that they are proceeding with all deliberate speed to pursue these non-military means and are not stalling out of fear or doubt.

4. Military Allies

If the US does not find allies to share the costs and the risks as described in #3 above, then only a minority of Americans will support a US military intervention unless the first three screens are fully met, in which case supporters may become a majority of Americans, but NOT at consensus levels. An alternative to support of allies is that the use of force is supported by all US leaders, or the top leaders in both parties, Republican and Democratic, particularly the top military advisors, like the Chairman of the Joint Chiefs, the Secretary of Defense, or in whomever the public has the most confidence in military matters, if it is not the President. To the extent that we are weak in getting allied support, this alternative support may substitute. At best it will produce a sub-consensus majority for the use of force. Working with the United Nations has been the preferred approach for securing the diplomatic efforts, negotiations, and sanctions needed before permission to use military force is finally and reluctantly granted by the American people. The UN is also the preferred umbrella under which those forces are committed when force is necessary. NATO may also serve in this role. A broad group of allies sharing the risks and the costs is the desired criterion.

Commentary. The costs and the risks do not have to be shared in equal measure by each ally. If an ally shares the costs and if its own security interest seems to people as remote from being threatened directly by the developing situation, then it need not share the risks. If on the other hand, it is a "frontline" state or its security seems otherwise directly threatened by the situation, it must share the risks and does not need to share much or any of the costs.

5. A Visionary Objective

The fifth screen is of a different nature. Although the objective of the intervention is stopping the heinous or rogue behavior of leaders, if that objective can be plausibly embedded into a more fundamental, visionary, or high-minded goal and recast as going beyond the initial objective, then support for military action, assuming the first four screens are fully passed, reaches the upper range of consensus levels, 80-95%. The visionary goal must be reasonably perceivable as resolving the underlying problems causing the deteriorating situation, such as overcoming the lack of democratic roots in key countries involved in the situation, or turning enemies into allies, or bringing long-term peace to the entire region of the conflict by resolving the key underlying antagonisms.

6. Early Non-military Intervention

A 6th screen is that if we start the effort to eliminate some heinous behavior(s) of leaders at an early point when the rogues are not yet well-entrenched and still lack solidified support from their own elites, the American people believe that efforts employing non-military means such as trust building, conflict resolution techniques (particularly with the help of a broad group of allies) to stop the heinous behavior will be more likely successful, as was described in the Kosovo example in the beginning of this paper. Support for military action, when and if finally necessary, occurs among an even larger majority of Americans than if this is not the case. If the first four or five of the screens are fully met, along with this sixth screen, again the support for use of force will reach the upper range 80-95% consensus level.

Testing, Modifying, Confirming Screens

Many questions were asked in ATI #14 during Operation Desert Shield on alternative reactions to the situation and to the wide variety of further possible developments. People had given a lot of thought to these possibilities and were ready to answer substantively and consistently with few "Don't Knows". Whenever such a situation has arisen in the intervening eight years since, where the use of force was considered or justified by leaders, such as in Chechnya, Haiti, Rwanda, Somalia, Bosnia, and Kosovo, opportunities to ask important and deeper questions in survey research have further built up the body of knowledge on when Americans are ready to use force. In its complexity, costs, risks, seriousness, and scope, the situation in Kosovo might have produced one of those defining moments when a new paradigm on when to use force emerged.  So far that has not occurred.  The entire lengthy exercise ATI has gone through in over twenty surveys in 12 years that sought in part to answer the question of when do the American people favor the use of force, proves again the surprising result that correct reading of the public's views using public- interest polling might have favorably changed the course of history. If President Clinton and the Congress had really understood the ATI polls and others showing similar results, these "leaders" might have thought through their options on Kosovo quite differently.  A ground operation might not have been ruled out and a more realistic use of air power, missiles and bombing might have avoided many diplomatic disasters and much human misery.  In fact, if the Clinton Administration and Congress had understood the US public's support for the UN and the International Criminal Court, Interpol might have arrested Milosovic several years ago and he might be on trial in the Hague rather than still in Belgrade.

Political polling is a powerful tool and it is important that a substantial amount of it should be done in the public interest.  Political polling to help elect candidates, support the goals of partisan policy organizations, sell newspapers, or provide TV news soundbites does not cover this kind of important information.  Not every pollster has to conduct public-interest polling in the same way that not every newspaper needs to be honest and non-partisan, but some should, and, as is well known, some are.  The distinction between public-interest and typical commercial polling and the important values of both must be recognized.


[1] Locating Consensus for Democracy -- a Ten-Year US Experiment, By Alan F. Kay, Americans Talk Issues Foundation, St. Augustine, Florida, USA, 1998.

[2] Ibid, pages 119-130, and elsewhere throughout the book.  See "People's Military Policy" in the index on p. 415. 

[3] Ibid, p. 394. This is illustrated by the misunderstanding of public opinion by political leaders, which led to the US decision to withdraw ground forces from Somalia in October '93, described in footnote 6.8, as follows:

Speakers on a panel at a conference held at the Washington Marriott Hotel on October 20, 1997, "Misreading the Public, Policymakers, Public Opinion, and Foreign Policy in the '90s" sponsored by the Ford Foundation, and attended by high level staff from the Administration and Congress, policy organization leaders, and news media reporters seemed to reach a consensus on the following explanation of the US debacle in Somalia:

When in October 1993, just after CNN showed the dead bodies of 15 US soldiers killed in Somalia being dragged through the streets of Mogadishu on TV screens in living rooms everywhere, Congress demanded and received an explanation from the Administration.  Many, particularly Republican Members, believed that the explanations of the current strategy and goals of the Administration were so weak and ill-thought out that they made a spot decision to come out for immediate withdrawal, because they believed it likely that by election day the US effort would be widely perceived as a failure, and they needed to get on the right side of the issue immediately, not to be tarred with the failure brush later.  The consensus of the panel members was that it was not actual public opinion (no survey results were then available to them), that was responsible for the haste of withdrawal but the guess on the part of Members where public opinion was likely to have settled on election day, not totally unreasonable, a viewpoint arguably a manifestation of the Vietnam syndrome. If Members were more in tune with public opinion that public-interest polling could provide, such disconnects would be less likely.

>>> 2.2.7  Giving Up Sovereignty - One Example

Table of Contents

Return to Top