Printer Friendly Page Americans Ready to Give Up Sovereignty - The Full Story

2.2.7a Giving Up Sovereignty to Pooled Control
by Multinational Agreements

Publication as scheduled as Chapter  12 [i]
The Search for Consensus  
Original Author Alan F. Kay, ATI
"The Issue of Belief" Edited by Willem E. Saris, U. of Amsterdam and Paul Sniderman,
Stanford University, Princeton U. Press, expected 2002

Chapter 12 was edited by Prof. Saris to conform to the standards for this book.

Public-interest polling has been defined (Kay 1998) as survey research that uses the best techniques to determine what the general public wants for governance. It differs from political polling fundamentally only in the intention of the sponsor. As part of his/her intentions, the sponsor must be committed to ongoing polling (to ask follow-up questions) and must have the determination, competence and resources to do so. To act truly in the public interest, the sponsor must ensure that surveys are written, conducted and analyzed according to best practices, with the aim of determining what the public wants for governance accurately, reliably, and in an unbiased fashion. Moreover , the sponsor must report and promote the results in such a way that interested observers can confirm the use of the best practices. The sponsor must demonstrate a willingness to test support for governance proposals that cover a range of the political spectrum- generally much broader and deeper than political leaders and mainstream news media consider. Covering a broader range of proposals, particularly including some that the sponsor may personally or institutionally oppose (or prefer to ignore), is one of the most convincing ways of demonstrating lack of bias. Other ways of reducing bias and increasing the accuracy and reliability of the findings are the debate format, presenting expert knowledge to the respondents to see the effect of this information and inter-survey serial polling, using a series of questions on an issue in different polls to develop a better insight in the opinion of the people. We will discuss these two features below. Before doing so we want to draw the attention to another aspect of public-interest polling which is locating consensus.

Locating consensus

So far the description of public-interest polling could also be applied to academic survey research which is, in general, also not special interest polling. However an aspect that makes a difference between academic survey research and public-interest polling is the effort in public-interest polling to find a consensus for policies. This is normally not the case, although there are also academic studies which go in this direction (Sniderman and Carmines,1997).  By a consensus we mean that a proposal gets support of more than 2/3 of the population. This usage of the term is in line with the approach of Yankelovich and Smoke (1988). Nothing in the definition requires public-interest polling to include searching for consensus. Yet, public-interest polling often leads to finding relatively unknown policy proposals that the public strongly and reliably supports. The words “Locating Consensus” are in the title of this paper because, in fact , the activity of locating consensus is one of the most interesting, important, natural, and fascinating aspects of public-interest polling. I believe locating consensus will inevitably come out of any major public-interest polling effort, even if the sponsor were not originally attracted to that pursuit.

As will be shown below, policy proposals are easily found with large public support in every issue area. These proposals must be tested in depth, exploring why and how they are supported. Does the high support hold up if we vary the features of the policy, vary the language used to describe the policy, and re-test the support after  making the respondents consider the effect of arguments for and against the policy ? All such efforts are requirements for the public interest.

Organizations that undertake this effort in an area where no in-depth research has been done previously will find it time-consuming and expensive. Multiple surveys will generally be required. Such testing is not practical to pursue unless all signs continue to point to the fact that some version of the original policy, or some quite different policy, emerges that persists and survives all tests with a high level of public support. If multiple organizations work in the same area and share results, the time and costs may be shared and the burden is far less onerous, particularly if each organization, as is likely, has somewhat different methods of pursuing this process, which adds to the assurance that the findings are robust.

The process would not be worthwhile for policy formation if the support that ultimately emerges is only a fragile 51%. The higher the support, the more worthwhile the effort. The “Americans Talk Issues” (ATI) foundation has found a few such proposals supported by over 90% of the public, but 70% and 80% are more common and still valuable (Kay, 1998). The size of that percentage support is only one of several factors that makes the effort more valuable: another is its intensity. If two proposals both have 80% support - one of them with 60% strong support and the other with 20% strong support, all other things being equal - the former is clearly a more valuable policy.

Many other aspects remain to be considered. If a policy has leverage, i.e. it solves two or more governance problems previously treated as independent problems, it is clearly more valuable. If the benefit-to-cost ratio is very high for some or somewhat high for many, and can be expected to remain high for a long time, then the proposal becomes clearly more valuable on any of these counts. Seeking policy proposals with large benefits vs. costs, including large, persistent and consistent public support, is what ATI has tried to do in searching for new proposals to test and in performing further tests for previously considered proposals.

We call this process “Locating Consensus” and we use the term consensus referring to policy proposals rather loosely because of the many variables that affect the value of the proposal and the strength, persistence, and consistency of the support.

It is important to remember that the stronger the public support of a policy, the more valuable the policy for the political process and governance. If support for a proposal holds up under testing at, for example 80+%, it is a consensus proposal, but it is much more than that. Just as only in close elections the exact vote count matters, the statistical errors in data showing 80+% support are less critical, leading to more confidence in the validity of the support. Support at this level invariably means that all major demographic groups and almost all states and congressional districts also support the proposal. In general, analysis of demographic data shows where support is coming from that can assist in coalition building. In the case of an 80+% consensus, the coalition is all major demographic sectors. For a summary of the benefits of 80+% support policies we refer to Table 1.

Table 1.  Summary of the benefits of 80+% Consensus  (Table 14,2 of Kay (1998).)




When embodied in legislation, consensus support means the whole country is united behind the law and this mass incentive to make the law work can be awesome, as compared to legislation that a majority opposes, which leads to foot-dragging enforcement and widespread evasion.


Support for the proposal is significantly less vulnerable to long-term public opinion drift, e.g., if of those originally favorable, exactly 10% change their minds, a reasonable expectation for drift, still a large majority remains favorable.


Support is so high that it would still be at consensus levels including the maximum effects of sampling and other errors working against its support.


Support of all demographic groups must be favorable to the proposal (e.g., with a real landslide vote of 80% for national office, as a practical matter every state would have to have voted with the winner).  This fact is observed repeatedly in polling data.[ii]


From sampling theory, a sample size as small as n=25 will confirm that if the support of the whole population is 80%, then at least 68% of the sample would be supportive.  (In this case the sampling error is ±12%.) Thus, if many organizations each poll small samples of 25 on the question, they will find 80% support on average with a scatter and expected minimum of about 68%.  The full marginal cost of testing one extra question in a sample size of 25 is about $250.  This fact is extremely important for the practicality of prospective independent confirmation of the validity of consensus positions.




Elected officials need not slavishly vote with the majority on every issue. Such figures as Congressman Bernie Sanders and Senator Jesse Helms, at opposite ends of the spectrum, have won election after election in part because they are admired for their independence and the courage of their convictions. But this phenomenon has limits.

When a law is opposed by large segments of people in a free society, cooperation with it is half-hearted or withheld. Laws on complex issues affecting the general public enacted without strong majority support can prove ineffective and disruptive, and further reduce public confidence in government. In contrast, legislation supported by an 80+% majority, after enactment, gets solid cooperation from the entire nation and a kind of support in practice that is the essence of successful policy. It can be the glue that holds a country together. 

Special Interest Polling

In contrast to (and much more common than) public-interest polling, almost all other political polling is “special interest “ polling, which is survey research designed to advance the cause of special interests. Kay (1998) defines special interests as groups that are not large and diverse enough to include a majority of the public. Often in fact, they are just a tiny but wealthy and powerful minority whose principal interest is to advance their own cause or seek a political result with little or no regard for what those outside these groups want. Sponsors who engage in special interest polling are participating in a time-honored and entirely legal activity, beating the system from within the system. The sponsors of such polls usually make as large an effort as they can afford to get wide distribution since their intention is to interest, inform, educate, and/or activate the public (including elites) and to affect public policy. Consumers are not special interests because we are all consumers. But those who favor a particular piece of pro-consumer legislation are special interests if they are not in the majority. We are also all environmentalists, at least in the sense that virtually all of us want an environment that does not kill us. (When asked in surveys, 60% to 80% of the public say they are “environmentalists”.) So environmentalists, altogether, should not be thought of as a special interest. Particular environmental groups may be special interests, for example, “ environmental activists”. Those who favor (or oppose) any particular piece of legislation are not special interests if they are truly in the majority. And in a functioning democracy, of course, the majority rules vote by vote, while legal protection of demographic minorities, corporations, and individuals is respected.

Most political polling is special interest polling for two reasons. Special interest polling very rarely asks what the public wants for governance – only when occasionally seeking to find out how the public reacts to what the particular special interest, or its adversaries, are pursuing. Political polling with few exceptions is sponsored by those who have a political interest in the outcome and seldom meets the sponsor-intention requirement imposed on public-interest polling. Sponsors of political polls that focus on getting candidates elected seldom try to deny their bias and typically make public a few of the question results, if they think it can help their candidate to victory. They do not make their poll public otherwise. Polling is expensive for most candidates but does produce valuable, privileged information. Keeping the findings within the candidate’s campaign offers an important competitive advantage.

Most of the polls that the public becomes aware of are those sponsored by the major news media themselves. The New York Times, The Wall Street Journal , the Washington Post, USA Today, and others, often teamed with a major TV network,  each conduct on the order of 20 polls a year, sometimes more. The findings are very well covered by the sponsors’ and their syndicates’ own media and, thus, have a substantial impact on the public and the political leadership. “Public polls” can seldom be considered  “public-interest polls”. Although many are very good and have little bias, they seldom cover what the public wants for governance. When they do, rarely do they examine any, let alone a wide range of proposals, outside of what political leaders are proposing or promoting. What the public wants for governance, unless it is under active consideration by top political leaders in Congress or the Administration, is not only ignored by the major news media but it is not considered to be news as suggested by the polling director of CBS (Kathy Frankovic) who said “We do not do social science research” (Kay 1998, page 12). The media do not consider the public’s own desires for governance in the world’s greatest democracy to be news.

Does it make a difference?

Above we have indicated that special interests will make public only those results of research which serve their goals. On the other hand we have argued that “public polls” will only give attention to a limited set of policies out of all possibilities. This means that the public will get biased and incomplete information about its own preferences.  There is a lot of literature which has shown that one can obtain very different results from survey research depending on the formulation of the question. Schuman and Presser (1981) give an overview of different formulation effects studied by survey methodologists. In the political science literature this topic has been discussed under the name of framing experiments (Smith, 1987; Rasinski, 1989; Gamson and Modigliani, 1987; Sniderman and Theriault, chapter IV in this book). The problem can be illustrated by an example which will be used later on as well, namely public opinion about the United Nations. Kay (1998, page 367) mentions a series of possible questions which will generate considerably different results. The following questions are suggested:

Q1: Do you approve or disapprove of the United Nations?

This question will generate probably a considerable consensus close to 80%. For the next modified question this will already be less:

Q2: Do you approve or disapprove of the job being done by the United Nations?

The support will drop because some who could approve in principle, could not find themselves ready to approve “the job being done”. A third version
Q3: Do you think the United Nation is doing a good job or a poor job trying to solve the problems it has to face?
would find still less support than Q2 as some respondents reason that while they approve of the job being done, they still can not call it a “good job”. A fourth version finds still less support than Q3:
Q4: Do you think the United Nations is succeeding at ‘reducing hunger in the world ?”
Here, the distinction is that “succeeding” is more supportive than “doing a good job”. If one were to ask:
Q5:Do you think the United Nations has made progress in eliminating hunger in the world?
there would be still less agreement; and for:
Q6: Do you think the United Nations has succeeded at eliminating hunger in the world ?
there would be almost no agreement.

These six questions are an example of a series where the wording is gradually changed through the series so that starting from a statement that has overwhelming support, one proceeds with gradual changes in wording to end with a statement with almost no support. If we would submit these six questions to six different survey samples, we would find some generally decreasing progression of the support percentages. Similar wording experiments can be done in many other fields and with many policy issues.

Given this phenomenon that one can get any amount of support or lack of support a special interest group nearly always has the opportunity to report a survey result that shows the strong support or the lack of support for any policy or any organization. A concrete example comes from Kay (1998, page 303) and concerns a policy with respect to campaign contributions. Ross Perot has reported a support of 80% for a policy which was asked as follows:

Should laws be passed to eliminate all possibilities of special interests giving huge sums of money to candidates?
The bias of this question, the lack of any negative statement about Perot’s proposal in the question or in a preamble, was rightly challenged by a highly regarded professional pollster, Warren Mitofsky, Executive Director of the organization supplying exit polling for four TV networks in campaign ’92. He reported a support for the same policy of only 40% to a question which he saw as comparable but more balanced:

Should laws be passed to prohibit interest groups from contributing to campaigns or do groups have a right to contribute to the candidates they support ?

This real life example shows how large the differences can be. This will not be clear to the public and the policy makers if the results of such different formulations are not presented to the public but only the one result that supports the interest of the special interest group. These effects can be used purposely by special interest groups by choosing a formulation for which the result is most favorable with respect to their goals.

It is, however also possible that research companies unconsciously create such an effect. A nice example was reported by Slot and Saris (1999) related to a referendum in the Netherlands. During the campaign for a new housing project one polling institute reported that a majority of the population was in favor of the project and another institute that the majority was against while the questions they used were the same. The explanation was that these studies had different sponsors and that the research company for an environmental protection society thought that this sponsor would like questions about the environmental problems before the vote question while the research company for the local government thought that this sponsor probably would like questions about housing problems before the vote question. So even without real manipulation of the questions by the sponsors large differences can be realized in special interest polling.

Besides arbitrary or biased results, society is also confronted with incomplete information. We have argued that public polls done by the media do not cover the complete range of policy proposals but only the proposals under discussion by the political parties. But that means that potential consensus policies might be ignored in favor of partisan policies.

So the question is how public-interest polling can overcome these problems using the same tools as the special interest polls and the public polls if they are performed with respect to sampling and field work. The difference is made by two basic principles which have been applied in public-interest polling and are ignored in the other two kinds of polls. These principles are the use of a debate format and the use of serial polling. These principles will be discussed in the next two sections.

The debate format in public-interest polling

When a proposal appears to have considerable support, but before too much is made of it, it should be tested by what in public-interest polling is called the debate format (Kay, 1998, 53). In the basic version of this format, a question about support for a policy is followed by a sequence of balanced pro and con arguments, each phrased as a question in some reasonable way, such as, “Have you ever heard this argument before?” or “ How convincing is this argument?” or even deeper questions. After the pro and con arguments have been evaluated by respondents, the initial proposal question is re-asked (called the “second asking”). This sequence of questions constitutes the basic debate format and is indicated schematically by: policy, balanced pro and con arguments, policy. There are many variations of this format, including a half sample design where half the respondents are not initially asked about the policy proposal. In this way this approach is very similar to the “choice questionnaire” (Saris et al. 1984, Neijens, 1987, Neijens et al. 1996, see also chapter 10 and 11 in this book) but the kind of arguments used are different as we will show now by an example.

The example to be discussed here concerns campaign contributions, the same topic we have discussed above but a quite different policy will be explored (Kay 1998, 207).  This policy was discussed using the following debate format in January, 1994:

Do you favor a proposal to …

Require candidates for Congress to raise, at least, one-half of their campaign funds from  individual voters in their own district? (61% in favor)

Tell me if you think these arguments for or against the proposal are very, moderate , a little or not at all convincing.

Supporters say that…..

  1. Many candidates for Congress now receive well over half their campaign money from people and special interest groups outside their district. This practice distracts them away from representing the interest of the people in their district. (56% very convincing).

  2. By forcing candidates to raise most of their money from people and groups inside their districts, they will pay more attention to what is important to people in their own districts. (58% very convinced)

  3. This requirement would lessen the enormous money advantages that incumbents have over new people running for office if the incumbents couldn’t accept so much money from the outside interest groups. (43% very convinced)

Opponents say that …..

  1. Voters all over the country have a stake in the entire composition of Congress and should be allowed to help finance candidates nationwide. Although Congress Members must represent their district, they cannot know the best interests of their districts unless they also know what the national interest is , too. (23% very convinced) 

  2. People ought to have the freedom to choose to contribute to anyone they want as long as they observe contribution limits. This proposal would limit people’s ability to give to the candidates they want in other districts – such as women and minority candidates and candidates they particularly agree with on the key issues – if those candidates are over the limits in outside contributions. ( 24% very convincing) 

  3. There are wealthy and poor districts and urban and rural districts. The proposal would be  unfair to candidates in poor and rural districts who would have a difficult time raising half the money they need from inside their districts ( 46% very convincing)

Now that you have considered all these arguments, rate the original proposal again:  ( 70% in favor)  

Do you favor or oppose making this proposal into law? (68% in favor)

The respondents are provided with pro and con arguments and it seems that their positive reaction to the proposal has not been reduced by the con arguments but a bit strengthened by the pro arguments. This is not necessary always the case. Kay (1998, 210/11) also gives an example where the opposite occurs. It seems that the opinion of the people on the campaign contributions is rather strong and can not easily be changed by the given arguments. In such a case one can have more confidence in the consensus result which has been obtained.

Serial polling in public-interest polling

Instead of looking for consensus policies within one survey one can also look for consensus policies by a series of surveys which we have called “serial polling”. Sometimes, responses come in from a survey that seem surprising (majority gave an unexpected answer), confusing, or ambiguous (did the respondents understand what we thought we were asking ?), or not fully confirming previous results in the same or earlier surveys. In such situations we did what people do if they are curious when similarly reacting to someone’s conversational remarks. They ask follow-up questions to clear up the ambiguity, the confusion, the contradiction, or, to shed some light on why they were surprised. In doing so we were mostly able to clarify our confusion and get a better picture of public opinion.

In order to illustrate this approach we will discuss a serial polling process concerning the relationship between the US and the UN. Nation-states everywhere are under pressure for devolution, to let decisions be made internally at the lowest political jurisdiction which has the information needed to make them.  The US, for example, tries to give more authority to the states.  Many countries have regional or scattered ethnic or religious minorities that are struggling for independence and in some countries have achieved it.  In the US, militias have wanted to declare independence in West Texas and Montana.  National governments feel this pressure from below at the same time globalization is creating problems that no nation can solve without cooperation from others, requiring international agreements for resolution. National governments are in a vise between forces from below and forces from above.  For generations most nation-states have been the focus for bonding patriotic feelings and for fervent nationalism, which seem in the last decade or so to be waning despite heroic talk that this can never be allowed, sovereignty can never be given up.  National governments have traditionally aggrandized all powers to settle issues by force, i.e., to wage war.  All of these issues will take years to work out.  The heart of the matter is when, where, and how much sovereignty will a nation-state give up.  On this issue ATI tried for years to determine the view of the people of the US.  The high-resolution view of this issue that ultimately emerged was worth the effort.

This research line of serial polling started after we received the answers to the following question with eleven items, which were tested in March ’91, using a battery under this frame:

President Bush says that the war in the Persian Gulf is the beginning of a NEW WORLD ORDER  – new ways for the United States and other  countries to behave.  I am going to read you some ways of doing things that might be part of the New World Order, and I’d like you to tell me whether you strongly agree, agree, disagree, or strongly disagree.  [Items randomized]

The top-scoring item was:

ATI#15, Q58: The US should use its position to get other countries to join together to take action against world environmental problems.                                            Agree:  93%.

This very highly supported result practically cried out, “There is a red spot – a summit – out there somewhere.”  If we explored in depth, either we would find a willingness of the American people to “give up” some aspects of national sovereignty or else we would catch the public in an inconsistency whose roots could be tracked down in a way similar to what we found in the case of foreign aid (“The Longest Running Disconnect”, Kay (1998, pp. 248-249)).  No sovereignty issue was raised directly in Q58.  But as a practical matter it is hard to imagine a group of countries taking effective action against “world environmental problems,” as Q58 describes it, without either the consent of all countries involved or some breach of national sovereignty for some countries, which could turn out to include the US itself.  The enormous support, 93% (69% strongly), made the urge to explore irresistible, even raising the possibility of a consensus of strong supporters alone!  What’s the significance of such unbelievably strong support here?  If we could pursue a line of reasonable and realistic questions leading in this direction, where would the American people come out?  Would the people opt out with a knee-jerk, super-patriot reply like, “Others have to give up their sovereignty, but not the US of A”?  Or would they find solid ground on which to support some pooling of sovereignty?

Could we find how the public would handle the dilemma?  If it could be found, where would the high ground stand?  Wouldn’t it be fun to look?  At this point the expedition seemed impossible, and the idea was relegated to the pile where a handful of other quest ideas lie untested.  It took five more surveys and another three years, but ATI did pursue the quest and found the right questions to ask and the right way to ask them that provided the answer.  Best of all, the answer made the quest worthwhile and makes one proud to be an American member of the human family. 

So the first logical step was a debate format, described in Kay (1998, Chapter 5). The first asking of the policy question in the debate format, Q57 in ATI #16, was a confirmation [iii] of the ATI #15 result that started the search, although with the preamble changed a little (see table 2). We had taken George Bush and his New World Order out of it to avoid contamination of the response by possible support based on the offering of an authority figure.  Since the results of the two first askings were almost identical, the President apparently was no longer effectively an authority figure for this issue.  The President had not made use of his own creation, the concept of a New World Order, and it had become almost immaterial to policy discussions.  What a great illustration of the importance of leadership and vision. 

After this policy question the pro and con questions followed. After four pro and con arguments, the second asking of the debate format, was Q62 on p. 118 of Kay (1998).

ATI #16, Q57.  Please tell me whether you agree or disagree with the following statement.  The United States should use its position to get other nations to join together to take action against world environmental problems.  Is that strongly or somewhat agree/disagree?

Table 2.  Solving the Global Environmental Problems






Original Q

ATI #15, Q58

First Asking

ATI #16, Q57

Second Asking

ATI #16, Q62





Strongly agree








Somewhat agree








Somewhat disagree








Strongly disagree
















Here are the pro and con arguments: 

Let me read you statements that OPPONENTS/SUPPORTERS have made against/for the proposal you just heard.  Please tell me whether you find the opponents’/supporters’ arguments convincing or unconvincing.

First/How about...(first pair from opponents, second from supporters, pairs ROTATED)

Do you find this argument convincing or unconvincing?  Would that be very or somewhat?

Q58.  Joint action with other countries could shift priorities to global environmental problems, like global warming, that are in the distant future when there are more immediate environmental threats close to home.

Very convincing:  10%

Q59.  Joint action could lead to other countries and governments developing regulations that hurt America’s economy and are not appropriate for our society.

Very convincing:  13%

Q60.  There are serious environmental threats like global warming and ozone depletion that require immediate attention.

Very convincing:  45%

Q61.  Many environmental pollution problems go beyond any country’s borders and can only be addressed effectively by all nations acting together.

Very convincing:  59%.

Q62.  Let me ask you an earlier question again. [Repeat Q57.]

Note how few find the opponents’ arguments very convincing (10% and 13%) compared to the supporters’ (45% and 59%).  This is the kind of difference among the “very convincing” that generally appears to be required for highly supported policies in the first asking to remain highly supported in the second asking.  Conclusion:  ATI #15, Q58 is a solid consensus proposal, but when the American people realize that they are fooling around with national sovereignty, how will they come out on this?  Won’t the devil be in the details?  ATI has plenty more work to do before being fully satisfied with this result.

In what seemed like an unrelated matter at first, ATI #16 found wide support for the US participating in “some new international organizations that may be started by private groups and aided by the United Nations and a number of countries.”  The preamble gave these new organizations a good pedigree and only asks about support for the participation of the US in a general way with no up-front commitment.  Still there is an implication that if there was agreement on the details, these proposals would be adopted.  Three were tested with support emerging in the consensus range of the low 80s:


Q48.  A bank that would finance environmental clean-up around the world and joint-venture with poorer countries to transfer environmentally sound technologies.

82% approve.

Q49.  An international organization like the Red Cross to take the lead in cleaning up environmental disasters, such as Chernobyl, the Exxon Valdez oil spill or the Persian Gulf oil spill and fires.

81% approve.

Q50.  A bank to invest in all countries, to develop the use of more efficient energy systems in agriculture, industry, housing, and transportation.

80% approve.

Here we have consensus-level support for three specific environmental proposals, quite daring ones too by modern American standards, proposals that no political leader or major news media would seriously suggest.  Support was about ten points below the ATI #15, Q57 result, but still remarkably high.  We seem to have found a sizable ridge.  To continue the ascent, what still should be asked? 

In ATI #16, Q75 was the sovereignty question flat out.  Pre-testing indicated many people did not know what the word sovereignty meant.  So we stated what the idea is in plain English.  We asked the identical question again as Q57A seven months later in ATI #17 in half-sample A and a different version in Q57B.  Table 3 shows that only a non-consensus majority supported the first version either in Q75 or in Q57A (half-sample A) and a majority was opposed in the second version of half-sample B where the concept of giving up sovereignty was made even clearer. 

Table 3: The public opinion on the role of the UN and the US

ATI #17, Q57A (same as ATI #16, Q75). United Nations resolutions should have THE FORCE OF LAW and should rule over the actions and laws of individual countries, including the United States, where necessary to fulfill essential United Nation’s functions.

ATI #17, Q57B. United Nations resolutions should rule over the actions and laws of individual countries, where necessary to fulfill essential United Nations functions, including ruling over US laws even when our laws are different.


ATI #16, Q75


ATI #17, Q57A


ATI #17, Q57B






Strongly favor

Somewhat favor




} 59%




} 50%




} 38%

Somewhat oppose

Strongly oppose




} 39%




} 42%




} 54%






An explanation of why Q75 is more favored by nine points than Q57A is the greater due process implied by addition of the phrase “force of law” than is implied by “rule over” without any qualification to assure that the ruling will not be dictatorial.

In these three questions we had gone beyond the issue of world environmental problems to a much broader mandate for the UN. This was clearly to much for a consensus. The country was and is nowhere near ready to give up sovereignty across the board.

We had found strong support only in the issue area of the environment and not the whole issue of environment either, only “world” environmental problems.  So we went back to that, and finally hit the terms of a proposal that meant consensus-level support for some measure of giving up national sovereignty as can be seen in table 4.

Here in Q22, limited to two global environmental problems, which were selected to be, and surely are, surrogates for any truly global environmental problems, we have a clear willingness to give up sovereignty.  Compared to ATI #17, Q57B, this proposal is stronger with the tougher clause, “even the laws of the United States when our environmental protection laws are weaker.”

Table 4 The role of the US and the UN with respect to environmental policies

ATI #21, Q22. Here is a proposal that some people say would allow the UN to do a better job, and I want you to tell me whether you favor or oppose it.

In order to protect and preserve the world’s environment, United Nations resolutions on polluting the atmosphere and dumping toxic wastes in the ocean should have the FORCE OF LAW and rule over the actions and laws of individual countries with weaker environmental laws, even the laws of the United States when our environmental protection laws are weaker.







Strongly favor

Somewhat favor




} 72%




Somewhat oppose

Strongly oppose




} 26%





There is much more to say about this issue. For more details we refer to chapter 14 in Kay (1998). Here we only wanted to illustrate the importance of the serial poll in public-interest polling. This procedure is especially important in cases like the above mentioned where one gets results that are far from expected. In such cases it is important not only to check the robustness of the result against counter arguments but also the effect of different formulations and its stability through time. Only if all these requirements are satisfied and the support remains as high as 2/3 of the population one can speak of a consensus in the sense of public-interest polling.   


This chapter illustrates very clearly the way in which public-interest polling differs from good quality special interest polling. If both use good quality survey procedure with respect to sampling , fieldwork and questionnaire design the two still differ with respect to four very important aspects.

1.       Public-interest polling is looking for consensus policies within a wide variety of options while special interest polling only looks for support for a policy of special interest group.

2.       Public-interest polling checks its results by use of a debate format where the respondents are confronted with pro and con arguments with respect to the policies studied. This does not happen normally in special interest polling.

3.       Public-interest polling uses serial polling to evaluate the robustness of the consensus under different formulations and conditions and through time. This is not done by special interest polling.

4.       Public-interest polling makes all its results publicly available so that anybody interested can check the procedures used to derive a certain conclusion (see appendix 1 for an overview of the ATI reports). Also in this respect the special interest polling has quite a different approach as has been discussed above.

Even though ATI in its public opinion polling used high requirements, mentioned above, before one would speak of a consensual policy, nevertheless many consensual policies have been found as can be seen in Table 5. This is in contrast with political reality where consensual policies hardly ever are discussed. The influence of special interest polling and public polling on the media reporting of survey research might be due to this. The many asterisked policies of Table 5 have seldom if ever been mentioned by the mainstream news media or the political leaders of either party.  It is hard to give an exact count of them, because a successful consensus search often finds a number of related or similar, but technically different, consensus proposals.   More than just a family of proposals, what consensus searches sometimes reveal are ways of thinking about issues that a consensus of the public has acquired, which can be formulated as a number of different consensus proposals.

An example of a consensus, which in principle could have led to an infinity of different consensus proposals, is the “Arms Control Conjecture,” first uncovered by ATI in its early surveys in 1988, referred to by Kay (1998, p. 72 and p. 100).  Based on considerable evidence, this conjecture suggested that any arms control proposal that met each of four criteria would have consensus (80+%) support.  Tests also showed that that support would drop substantially if any one of the criteria was unmet. 

Another example is the “People’s Military Policy,” discussed at length in Kay (1998, chapter V, pp. 119-130) and rediscovered from a somewhat different point of view with an entirely different line of questions in Chapter 10, pp. 228-229.  This policy is also a whole set of policies related through principles that the research has established.  On pp. 129-130, we speculated that the People’s Military Policy would remain largely valid for many years, and we revisited this aspect later.  On pp. 349-350, Chapter 15, we explored examples of responses to ten-year-old questions on military interventions which still conform to principles established years earlier.  My paper, “When Americans Favor the Use of Force”, submitted for publication in the WAPOR-Oxford University Press publication, the International Journal of Public Opinion Research in 1999, expands on the People’s Military Policy and its relevance to US military policy in the post-Cold War era, and earlier, including the Kosovo NATO action.

In the process of doing these studies one also uncovers insights into the reasons for the difference between the support for different versions of similar or nearly the same questions, most of the time based on substantive arguments. These consensual policies are often not very easy to change by adding arguments because they turn out to be well established on strong arguments. This does not mean that individual people can not change their opinion by hearing the new arguments but often approximately as many people move in the direction of support as opposition so that the percentages of support remain the same.These results are in line with the results of Page and Shapiro who also report very stable public opinions through time. On the other hand these results seem to contradict results reported by others (Converse,1964, Schuman and Presser, 1981, Smith, 1987; Rasinski, 1989; Gamson and Modigliani, 1987) who report a lot of volatility in opinions through time on individual level. The consensual opinions discovered by Public- Interest Polling are more stable partially because they have been derived by a series of studies looking for a high consensus while in the other studies a specific policy has been evaluated which often received less support. As a consequence the fluctuation in the expressed opinions can also be larger. Whether this is the only source for the reduced volatility requires further research.

Table 5 Overview of consensus proposals mentioned in Kay (1998 Table 14.1), with reference information.




Ch. 3

p. 55, 66

INF Treaty


p. 56*

Six win-win joint US/USSR undertakings


p. 65

Initial support for nuclear bombing of Japan


p. 66

Nuclear freeze


p. 72, 100*

Balanced, verifiable, understandable arms control agreements – with due process




Ch. 4

p. 96

Changes in Soviet policy


p. 99

Support for START Treaty


p. 99

Nuclear Test Ban


p. 102*

Nine goals of the Gulf War


p. 102*

When support for war is a consensus


p. 109*

Support for six proposals that would have avoided Gulf War




Ch. 5

p. 114*

Three long-term military commitments in the Middle East


p. 118*

Five New World Order proposals


p. 118*

Monitor and tax international arms sales


p. 119*

Three international organizations/banks for environmental and energy support


p. 122*

People’s Military Policy


p. 126-127*

Stopping rogue behavior of dictators


p. 131*

Regulation of globalization activities destroying environment, manufacturing with cheap labor




Ch. 6

p. 138*

Renewable energy sources


p. 141*

Stopping pollution/flight of industries to safe havens


p. 142*

Reinstating CAFE standards


p. 154*





Ch. 8

p. 174*

Several government reforms, including Quality-of-Life Indicators and Questionnaire in tax forms




Ch. 9

p. 212

Term Limits


p. 209*

Quality of Life Indicators


p. 212*

In-District Campaign Financing




Ch. 10

p. 229*

US support and submit to an International Criminal Court under UN with teeth


p. 229*

Support for People’s Military Policy


p. 232*

Pooling sovereignty in UN for regulating selected, truly global environmental matters




Ch. 11

p. 243*

Importance of UN nuclear inspections, humanitarian aid, global pollution control


p. 250*

Make allies pay when our military interventions help them


p. 253

UN peacekeeping training at Ft. Dix


p. 255*

Six criteria for UN interventions




Ch. 12

p. 271*

Populist view of government reform


p. 280*

UN take the lead in stopping aggression




Ch. 14

p. 313-321*

Consensus and sovereignty




Ch. 15

p. 346

Health care


p. 347





Note: * indicates political leaders and mainstream news media, when the issue was relevant, ignored this consensus and for issues still relevant, still continue to ignore them.


Converse P.E. 1964. "The Nature of Belief Systems." In: D.E. Apter. Ideology and Discontent. New York: Free Press. Pp. 206-61.

Gamson W. A., and A. Modigliani. 1987. “The changing culture of affirmative action.” Research in Political Sociology. 3:139-77.

Kay, Alan F., “Locating Consensus for Democracy – a Ten-Year US Experiment,”, Americans Talk Issues Foundation, (ATI), 1998

Popkin, Samuel L.  The Reasoning Voter, University of Chicago Press, 1991.

Page Benjamin I. & Robert Y. Shapiro, The Rational Public: Fifty Years of Trends in America’s Policy Preferences, University of Chicago Press, 1992.

Neijens, P., 1987. The Choice Questionnaire. Design and evaluation of an instrument for collecting informed opinions of a population. Amsterdam: Free University Press.

Neijens, P.C., M. Minkman, J.A. de Ridder, W.E. Saris & J.J.M. Slot (1996). A decision aid in a referendum, International Journal of Public Opinion Research, 8 (1), 83-90.     

Rasinski K. A. 1989. "The effect of question wording on public support for government spending." Public Opinion Quarterly 53(3):388-94. 

Saris, W.E., P.C. Neijens & J.A. de Ridder (1984). Kernenergie: ja of nee? [Nuclear power; yes or no?] Amsterdam: SSO. 

Schuman Howard and Presser Stanley Questions and answers in attitude surveys: experiments on question form , wording and context, Academic Press , New York, 1981. 

Sniderman P.M. and E.G.Carmines (1997) Reaching beyond race. Cambridge, Massachusetts, Harvard University Press.  

Smith, T.W. 1987. "That which we call welfare by any other name would smell sweeter: an analysis of the impact of question wording on response patterns." Public Opinion Quarterly 51(1):75-83. 

Slot J. en W.E.Saris (1999) Verschillen in resultaten tussen verschillende opinie-onderzoek (Differences between opinion polls). In Neijens P. and P.van Praag De Slag om Ijburg. Amsterdam, Het Spinhuis. 

Yankelovich Daniel, Coming to Public Judgment, Syracuse University Press, 1991 

Yankelovich Daniel and Smoke Richard, Foreign Affairs, "America's New Thinking," Aug 5 1988, p. 3. 

 Appendix.   ATI SURVEY REPORTS through 1995






Field Date

Report Date

Report Theme/Title

Each report covers a telephone survey of a random national sample of 1000 or 1500 adults or voters.  Reports #10 and #11 are bound together.  Survey #13 is covered by three reports.








Benchmark Survey on National Security Issues




Post Summit Perceptions; Nuclear Arms Reduction




Affordability of US Defense




Evolving Definitions of National Security




US Policy in Central America




Viewing the World on the Eve of the Moscow Summit




US/Soviet Relations Moving toward Cooperation




Waste in Procurement; US Foreign Aid Policies




American Use of Force




National Security (one month pre-election) and the ‘88 Election




National Security (eve of election) and the ‘88 Election




The Soviets; Use of Force by the US; Nuclear Weapons




The Peace Dividend as the Public Sees It




The US Looks at the Revolutions of 1989




Nuclear War and Weapons




The Use of Force – Showdown in the Gulf, 1990




The New World Order – What the Peace Should Be




The Emerging World Order




Perceptions of Globalization, World Structures, and Security

















What the American People Want in the Federal Budget





The Economy, Energy, Security and the Environment





The Economy, Energy, Security and the Environment








Global Uncertainties




Changing the System to Make the Federal Government Work in the Interest of All the People




Structures for Global Governance




Steps for Democracy – The Many Vs. The Few – First Wave




Steps for Democracy – The Many Vs. The Few – Second Wave




The UN at Fifty – Mandate from the American People




Fighting Poverty in America




Nuclear Proliferation




Who Will Reconnect the People: Republican, Democrats, or... None of the Above?

* A few field dates spanned month’s end. “Month” is when bulk of interviewing took place.
† Report Available from Center for the Study of Policy Attitudes, 202-232-7500 


[i] This chapter is based on Kay (1998) . This chapter is printed with permission of the publisher, the American Talk Issues Foundation, which reserves all rights to Kay (1998). Requests for copies of this chapter should be asked from the publisher the Americans talk issues foundation, P.O.Box 5190, St. Augustine, Florida 32084, USA, Phone 904-826-0984, Fax 904-826-4194.

[ii]  Samuel L. Popkin, The Reasoning Voter, University of Chicago Press, 1991, and The Rational Public: Fifty Years of Trends in America’s Policy Preferences, by Benjamin I. Page & Robert Y. Shapiro, University of Chicago Press, 1992.  The latter studies many time series of public opinion on policy questions.  It confirms that generally public opinion varies slowly except when major events relevant to the issue occur, and then opinion moves, if at all, in the expected direction.  This is a path-breaking book not only for the huge amount of evidence it presents for this phenomenon, but also because the authors show that events that move the total public, usually move all major demographic sectors in the same direction. The trends among different demographic sectors, when plotted as time series graphs, are roughly “parallel.”  (See Chapter 7 in Page and Shapiro.)  An important corollary is that a policy proposal, if favored by a large majority, will almost always be favored by each major demographic sector.  This is especially true of consensus proposals.  This finding is an attractive invitation for politicians to seek and support proposals that are the right thing to do for all.  They are natural coalition builders.

“Major events” in ATI’s experience, such as the ending of the Cold War , were big, multi-day “headline” stories in the mainstream news media.  An important point is that there are other “major” but slow-moving events, like the end of support for slavery in the era 1800-1861 or, in our time, the efforts of Mothers Against Drunk Driving or the feminist and ecology movements that develop because of the enormous efforts of the civil society, whose workings and effects the mainstream news media neither recognizes nor, apparently, understands.  These efforts can produce the same effect on public opinion, only much more slowly.  This is not unexpected.  Few doubt that there is slow drift over the years in the response statistics of those time series questions that can be asked for decades, such as whether the US did the right thing dropping the nuclear bombs on Japan in ’45. My bet on p. 65 (Kay 1998) is a fair bet, because at any time and in a relatively short time that long steady 58% to 68% support for dropping “the” two bombs could move out of the narrow range for reasons not noted in the mainstream press.

[iii]Three months later in time and, as always, with a new sample.

>>> 2.3  Year 2000 National Campaigns

Table of Contents

Return to Top