Gallup Finalizing Election 2012 Review

Gallup, as I've discussed previously, has been undergoing an extensive review of our 2012 election polling for a number of months. We are now in the process of finishing the review of all methodological issues relating to the pre-election polling. The process involves a blue ribbon group of outside experts, including Dr. Michael Traugott of the University of Michigan, and our own Gallup team of methodologists and statisticians. We will be reporting our findings in early June in Washington.

Our review includes a significant amount of new research, including fielding new research experiments focused on six specific areas of pre-election polling, along with simulations and re-examination of every component of the pre-election polling process. All of this research, some of it focused on the future of election polling, is one reason why the review process is taking several months. 

A number of changes have already been implemented in our ongoing tracking, and others will continue to be tested, including experimental tests involving voter sampling and likely voter models.

I'll be discussing more on this process on this blog after the June event.

Few Signs Yet That Sequestration Has Affected Average American

The sequestration that went into effect on March 1 does not appear to be a major issue for most Americans. At least, not according to the available evidence measuring public attitudes that I have seen.

We have tracked a set of two questions four times since March 1 -- one asking if the sequestration has been a good thing or a bad thing for the country and another asking the same question about “you personally.” Both questions include the stated alternative: “...or don’t you know enough to say?” 

There has been remarkably little change over time on these two questions. As you’ll see from the two graphs below, all three alternatives essentially straight line.

Our headline is that a majority of Americans still say they can’t judge the impact of sequestration on either the country or on themselves personally, particularly the latter. Another approch would be to focus on the negative side:   30% of Americans say sequestration has been bad for the country, and 24% say it has been bad for themselves personally. Those are clearly minority numbers, particularly so if you subtract the 17% and 11% respectively who say that sequestration has been a good thing for the country and for themselves. 

Hard to make a case from these numbers that Americans have perceived sequestration as a deadly blow to the nation or to themselves. (Keep in mind that the latest update on these numbers came from an April 29-30 survey, conducted after the news focus on delays in air traffic blamed on air traffic controller furloughs imposed by the FAA as a result of sequestration.)  

We also have seen no decline in economic confidence in the country. As a matter of fact, our Economic Confidence Index monthly average for April was -13, better than -16 in March, and back to where it was in February. And as I write this on Friday, May 3rd, our 3-day economic confidence index has tied its high for the year. No major hit there.

Our analysis also shows that federal government employees' views that their employer is letting people go have improved -- after taking a mild hit in March. We interpreted this as indicating that federal employees thought there might be a impact of sequestration in March, but by April didn’t see it materializing. (Although federal employees still report much more negative views than private sector employees.)

I should make note of the results of a New York Times/CBS News poll question that was reported this week. The question was phased: “As you may know, automatic spending cuts to military and domestic programs are in effect because President Obama and Republicans in Congress could not agree on a plan to reduce the federal budget deficit. These automatic spending cuts are commonly referred to as the Sequester. Do you think these cuts will help the economy, hurt the economy, or will they not have much effect on the economy one way or the other?" The results showed 46% said “will hurt,” 10% said “will help,” 35% said they will not have much effect, and 9% said they didn’t know.

This question shows a 16-percentage-point higher “negative” response on the impact of the sequester on the country than the Gallup question. But there are differences in the question phrasing. For one thing, the New York Times/CBS News question explicitly mentions that the sequester resulted from a failure of the president and Congress to agree on a plan to reduce the deficit.  This is in and of itself adds a negative tone to the question -- since Americans want their elected officials to cooperate.  This phrasing also reminds people that the deficit was not reduced. 

Most importantly, the question uses the verb “will,” making this a projective question rather than an assessment of what has happened so far. In fact when we asked in late February if the economy will get worse if the budget sequestration goes into effect, we found 56% saying yes. So, I think it is reasonable that people may still project a negative impact, but that they simply don’t see that negative impact yet.

In fact, the New York Times/CBS News poll followed the first question outlined above with a question that was not projective, but instead contemporaneous. This question asked about personal impact: "Have you personally been affected by the spending cuts in the Sequester, or not?" If yes: "Have you been affected a great deal, or only somewhat?" The results showed 27% said they had been affected (8% a great deal, 19% only somewhat) while 69% said they had not.  This result is in line with the 24% “bad thing for you personally” that our polling found. (This question didn’t specify whether "affected" meant in a negative way, but presumably the majority of those saying "yes" did assume the question was asking about a negative impact.) 

Bottom line:  While Americans may still feel that sequestration will have negative effects for the country and themselves in the future, most in the U.S. don’t seem to see that effect so far. 

Senate Defeats Background Check Measure, Which 91% of Public Supports. Why?

Roughly nine in 10 Americans (91%) say they would vote for a law requiring criminal background checks for all U.S. gun purchasers. The U.S. Senate considered legislation last week that would have expanded background checks for gun purchases: 54 senators voted yes; 46 voted no. Because 60 votes were needed to advance the background check bill, the bill failed. Although the legislation the Senate considered last week was much more complicated (more on this below) than the concept embodied in a simple survey question, the fact remains that the Senate did not engage in voting behavior that seemingly would be in line with the views of 90% of the U.S. public.

As New York Times columnist Maureen Dowd asked on Sunday, in reference to President Obama’s strong push to pass such legislation: “How is it that the president won the argument on gun safety with the public and lost the vote in the Senate?”

Dowd focuses in her column on how the president could have engaged in more efforts to cajole, push, pull, seduce, or threaten senators to vote for the bill (or, in her references, how Obama could have mimicked the behaviors of fictional president Andrew Shepherd in the Rob Reiner movie "The American President").

The situation raises other interesting questions, however. If the public is so overwhelmingly in favor, why did the Senate end up not passing it? And what does this tell us about the way in which representative democracy works in this country?

I am in favor of elected representatives listening to the voice of the people when they make policy; generally speaking, I think the collective views of the public are centered and wise, particularly on broad issues and directions in which the country can move. I think Americans believe they are smarter than the men and women they elect as well.

In fact, a recent review of research by Columbia University's Robert Shapiro shows that as a rule, policy decisions in fact do tend to follow the broad scope of public opinion. Shapiro concludes that there is, as he puts it, "a substantial connection, overall, between public opinion and policymaking in the United States." The connection is fairly stable and often substantial.

Which leads us to ask what happened in this situation in which this connection was not so substantial?

Commentators have brought up several points on this question -- in addition to Dowd’s column, which focuses on the inability of gun control proponents to twist arms in order to get legislation passed.

One issue, of course, is the fact that the Senate by design is not representative of the U.S. population. The juxtaposition of the 90% (national adult public opinion) figure with the 54% (Senate vote) figure assumes there should be a direct relationship between the two. That was not the Founding Fathers' intention. As part of the compromise that led to the development and passage of the Constitution, the upper legislative body gives each state equal representation.

In theory, if both senators from each of the smallest 23 states voted "no" on the background legislation, they would be representing about 13% of the nation’s population with those votes. If both senators from each of the largest 27 states voted "yes," they would be representing 87% of the nation’s population. In other words, in theory, 46% of the Senate vote could reflect the public opinion of only 13% of the population, and 54% of the Senate vote could reflect the public opinion of 87% of the nation’s population.

In fact, both senators from the nation’s smallest state, Wyoming (Barrasso and Enzi) did vote “nay.” The votes of these two senators represented a little more than half a million people. Both senators from the nation’s largest state, California (Boxer and Feinstein) did vote “yea.” Their votes represented more than 38 million people. In a national public opinion poll, the residents of Wyoming are about 1.5% of the representation of the residents of California. Phrased differently, California residents represent about 12% of the nation’s population, while Wyoming residents represent about 0.2% of the nation’s population.Yet both states’ senators get the same representation as far as votes are concerned. California’s residents’ views on policy legislation are much, much more important statistically in a national survey than are Wyoming residents’ opinions. But the opinions of the states’ residents are equalized in terms of the Senate vote.

These calculations are all hypothetical, of course. In fact, the vote on the background check law was not directly related to state population size. Nor, of course, is public opinion in each state monolithic, such that residents in small states are all opposed to background checks and large states all in favor.

In fact, the very high level of national support for background checks statistically means that it is likely that majorities of residents in every state support background checks, just at different levels.

Political scientist Brian Schaffner did an analysis (posted on themonkeycage.org) of support in each state for stricter gun control laws and the votes of the two senators from each state on the assault weapons ban bill that was also voted on Wednesday. The results show a very high relationship in general between the two variables. In other words, using a measure of public opinion on gun control at the state level and a vote by the state’s senators on a gun control bill, Schaffner shows that there is a strong correlation. The senators in this instance were in fact generally representing their constituencies.

Now, as noted, the 90% overall public support for background checks makes this a little different situation that support for stricter gun laws. But the concept is important. It’s certainly possible for there to be a situation in which each state’s senators faithfully do follow public opinion of their state's residents, but with a result in which the percentages of the final Senate vote end up being out of whack with the overall percentages of the U.S. population.

Secondly, there is the question of the link between senators and those they represent. The focus has been on the disjuncture with public opinion and the Senate. But senators are not elected by the total population in their state. Senators are elected first by going through a nomination process involving a small percentage of voters in one party, and then by a statewide election involving a larger, but still fractional, percent of the adult population. So it is possible to have a state in which the majority of residents favor a policy such as background checks, but in which the smaller group of voters in a state’s primary oppose the policy. For a senator facing re-election at some point, it is this group of residents in the primary and then in the general election who matter.

This is the point where intensive lobbying by groups such as the NRA come in. One of the most significant weapons that lobbying groups with a big grass-roots membership can wield is the ability to affect enough actual voters in a state to sway a Senate election. That in turn takes advantage of differences in the intensity with which groups of voters hold their views on key issues. The side with the most intense followers generally has a real advantage when it comes to voter turnout.

It is also possible that senators viewed this legislation as more than just the same type of straightforward, simple up-or-down vote on background checks as measured by the survey question. In fact, the legislation per se is complicated and, as is true of most bills, written in a very complex way to cover many possibilities. Some senators could have objected to specifics in the bill rather than its overall purpose.

It’s possible that senators viewed voting for a specific gun control measure as indicative of a general gun control philosophy, even if constituents favored the specific measure. In other words, voting for background checks could have been seen as the "nose of the camel in the tent," or as legislation that would lead to more than background checks, or as legislation that was in fact symbolic of gun control legislation more generally. While 91% of Americans support background checks, 58% say gun control laws need to be made stricter in general. So there is some disjuncture there in Americans', as well as perhaps in senators', minds.

And there is the issue of priority, which I reviewed here. The American public does not give guns and government efforts to control gun violence through legislation the highest priority at this time.

Finally, there is the fact that many elected representatives don't necessarily believe it is their purpose to slavishly reflect public opinion on the key issues of the day, but rather to give their constituencies the benefit of their thoughtful wisdom and deliberation. Thus, the response of these representatives (and other observers) is that a disjuncture between public opinion and the vote of elected representatives is not unusual or something to be disdained. Typically, of course, this old "delegate versus trustee" argument is used by individuals or groups when it is to their advantage. If public opinion is in favor of one's position, one evokes public opinion as the rationale for representatives voting in that direction. If public opinion goes against one's position, then one argues that the public is not aware of all the circumstances and implications and context, and one praises the independence of the representatives in voting the other way.

Still, that 91% figure -- the percentage of Americans who favor criminal background checks -- is extremely large. Rarely do we find nine in 10 Americans agreeing on anything. Despite all of the possible reasons why the disjuncture between public opinion and the Senate vote occurred, this is one of the more striking examples of what is a fairly unusual disconnect between public opinion and the vote of the representatives elected to serve them.

Prioritizing Gun Violence Reduction Measures

Only 4% of Americans say that gun violence or gun issues constitute the most important problem facing the country today, based on our April 4-7 monthly update of the "most important problem" measure. This puts guns in the same 4% category as immigration issues, education, and the situation with North Korea.

This also puts guns -- on this measure -- well below a number of other economic and governmental issues (we’ll have a full discussion of these results on Monday at gallup.com).

Trend wise, the mention of gun issues is about where it has been since December, when it jumped to 4% from virtually no mentions in November. This increase was, no doubt, a direct result of the impact of the Newtown, Conn., elementary school shootings. The mention of guns as the most important problem has stayed at about this level since December -- at 4% in January, 6% in February, and 4% in March and April.

Does the very low level of mentions of gun violence as the nation’s top problem mean that Americans don’t want Congress to pass gun legislation? No, I don’t think this is the case. This type of top-of-mind question is, by its nature, exclusionary. We actually allow respondents to name more than one issue, and many do. But by its nature, this question focuses the respondents in on “the” most important problem. And just because few Americans volunteer that guns are the most important problem off the top of their heads doesn’t mean that this isn’t considered to be an important priority when measured in different ways.

However, a Pew Research poll conducted Jan. 9-13 of this year asked Americans to rate 21 issues in terms of being top priorities for the president and Congress. The results showed a range from 86% who said that strengthening the economy was a top priority to 28% who said that dealing with global warming was a top priority.

“Strengthening gun laws” came in fourth from the bottom, with 37% saying that it was a top priority.

In short, both an open-ended question asking about the nation’s top problems and a closed list of questions asking Americans to rate issues as priorities show the same thing.  Gun issues have a relatively low priority.  

A January Gallup survey approached this topic in a somewhat different way and found that 51% of Americans were dissatisfied with the nation's gun laws, while 43% were satisfied. This 51% dissatisfaction rating put gun laws at about the middle of the list of those issues measured.  Other elements of American life where more than half were dissatisfied included the nation’s policies on crime, the quality of public education, healthcare, immigration, the amount Americans pay in taxes, the nation’s efforts to deal with poverty, and in particular the state of the nation’s economy (79% dissatisfied).

So gun laws were not high on the list of concerns using this measure either, although the finding that slightly more than half were dissatisfied indicates some level of concern. (But a follow-up question asked of those who were dissatisfied with the nation’s gun laws showed that not all of this dissatisfaction was because the laws should be made stronger; some thought the laws should be made less strong, while others didn’t know or couldn’t say.)

This is not to say that Americans are not in favor of stronger gun violence control measures. As my colleague Lydia Saad summarized in her review of the gun violence control measures proposed by President Obama a couple of months ago:

“Given the chance to vote 'for' or 'against' each of nine key proposals included in President Barack Obama's plan to reduce gun violence, Americans back all nine. Americans are most likely to be in favor of requiring background checks for all gun sales (91%), increasing funding for mental health programs aimed at youth (82%), increasing funding for programs to train law enforcement and schools in responding to active armed attacks (79%), and increasing criminal penalties for people who buy guns for others -- so-called straw purchasers (75%).”

So Americans favor laws that would be aimed at controlling gun violence, but the data don’t suggest that passing such laws is their highest priority at this point in time.

None of this speaks to the issue of intense focus on this issue among niche groups of Americans.  That issue is of course of significant concern to Representatives and Senators seeking re-election.  Smaller groups of gun activists -- on either side of the issue -- can make a difference in a close elections. So elected officials in Washington face a situation in which they recognize that the average American favors passing gun violence control measures (particularly measures like background checks), but one in which the average American does not consider these measures to be of the highest priority.  But, a situation in which the officials may well recognize that smaller groups of constituents in his or her district or state may care very deeply about gun measures, making their decisions on how to handle pending gun legislation complicated indeed.

Korea, 63 Years Later

As we recently reported, a snapshot of American public opinion last week showed that the majority of Americans favor the U.S. coming to the aid of South Korea should North Korea decide to invade again -- as it did in 1950. This comes as no surprise in a general sense, given the evidence showing that a) most Americans have an unfavorable opinion of North Korea, and b) most Americans are worried about North Korea as a threat.

Here’s the trend on views toward foreign countries, including North Korea. I’m not sure who the 12% of Americans are who say they have a favorable opinion of North Korea, but it’s possible there is some confusion between the two countries occupying the Korean peninsula. One modestly surprising finding is the slightly more positive view of North Korea held by Americans back in early 2001.

Below is the table showing that 83% of Americans perceive the development of nuclear weapons by North Korea as a “critical” threat, about the same as say this about the development of nuclear weapons by Iran, another not-very-favorably viewed country. This, too, may help explain why Americans would be willing to come to the aid of South Korea.

Of course, this country has a history of coming to the aid of South Korea. Only Americans who are about 75 years and older will remember the Korean War directly, but others may have been introduced to that war by Robert Altman’s movie M.A.S.H. or the popular TV show of the same name, which spun off of the movie (the movie itself came from the excellent novel by Richard Hooker, a pseudonym for a doctor who served in Korea).

A review of public opinion data from the Korean War era shows that, as is usually the case when the U.S. gets into a war, initial reaction to the U.S. intervention in Korea in 1950 was positive, with only 20% saying it was a “mistake” to send U.S. troops there. That opinion soured pretty quickly. By early 1951, almost half said it was a mistake. Opinions varied from that point on, and by early 1953, with both sides of the conflict continuing at that point to meet at the negotiation table in an effort to hammer out a peace agreement, the mistake percentage was down to 36%.

We did ask this mistake question one more time, in 2000, and found a similar response, with 34% saying it was a mistake and 47% saying it was not. This more positive attitude may too help explain why Americans are positive now about helping South Korea in the eventuality that North Korea invades. If it worked the first time, it can work again.

I came across another fascinating finding about the Korean war. When Gallup interviewed Americans in September 1950, just a few months after the U.S. had become involved, about seven in 10 Americans were convinced that the war would be over within a year or so. Only 14% said that it would last more than a year (the rest had no opinion, which was probably the most rational response).

In fact, it wasn’t until the summer of 1953, of course, before the peace agreement was signed. And by that time well over 30,000 Americans had died in combat, along with estimates of millions of deaths of Koreans, Chinese, and soldiers from other countries that got involved. The disparity between expectations of a quick war and the reality of a protracted war may have been one of the reasons why Americans became less positive about it as it went on.

One other very interesting finding. Over half of Americans in November 1950 thought that the U.S. was now involved in “World War III.”

Of course, there isn’t any official definition of what constitutes a “world” war is as opposed to other wars, but certainly Korea never opened up to be the same type of global conflict as World War II -- with fighting that was taking place in multiple points around the globe at once. The World War II conflict stretched from the China-Burma-India theater eastward across Southeast Asia, throughout the vast western and southwestern Pacific, and across most of Europe, the Mediterranean region, and Russia. The Korean War involved a number of different nations, but was confined to the relatively small Korean peninsula.