Framing the Debates

There was much to dislike about the three presidential and one vice-presidential debates, but one objection that was near the top of everyone’s list was the narrow range of topics about which the candidates were questioned. Adam Johnson has tabulated the number of questions asked by the moderators about each of the 22 issues they brought up, along with 10 issues that were not included.

questionsubjects

Of course, candidates could have brought up issues that were not specifically targeted by the questions. Johnson’s second chart tabulates the number of mentions of each of 33 issues.

finaldebatementions

Russia, terrorism and taxes were the moderators’ favorites, and Donald Trump’s taxes and Hillary Clinton’s emails received more attention than such issues as climate change, poverty or campaign finance. Johnson describes the framing of the issues as “center-right in nature,” and offers some examples to support his case, i.e., Elaine Quijano’s question, “Do we ask too much of police officers in this country?”

I subsequently ran across an article by Alexander Podkul and Elaine Kamarck of the Brookings Institute. As part of the 2016 Primaries Project, they tabulated the issue positions, taken from their campaign websites, of over 1700 Congressional primary candidates. They found that candidates in the two parties are not talking about the same issues. Here are the top five issues mentioned by Republican and Democratic hopefuls. Aside from their common focus on the Affordable Care Act, there is little overlap.

gs_20161020_primaries-project-issues

In the debates, there was one question about Obamacare. With regard to the other top Republican issues, there were four questions about taxes, three about the debt, two about immigation and one about gun control, for a total of eleven questions about Republican issues. The Democrats did not do as well. There were two questions about social security, but the framing suggested it needed to be “reformed” rather than expanded as some Democrats maintain. Since there were no questions about climate change, education or the minimum wage, the Democrats scored a total of three questions. It appears that the debate moderators (or their corporate media bosses) shared the views of Republican candidates about which issues are more important.

Tabulation the number of mentions of each issue yields a similar result. There were 241 mentions of the five Republican issues and 90 mentions of the Democratic issues. (The 45 mentions of Obamacare account for half of the comments about Democratic issues.) Unfortunately, Johnson does not tabulate mentions of the minimum wage, but even if we assume that it was referred to all ten times that poverty came up for discussion, that would still bring the Democratic issue mentions up to only 100.

Of course, these mentions were largely triggered by the debate questions. However, Secretary Clinton could have raised some of the Democrats’ issues more often than she did. Thus a second interpretation of these data is that the Democratic candidate approaches the upcoming election from a more Republican point of view than their Congressional candidates.

You may also be interested in reading:

Are the Terrorists Getting What They Want?

The World According to the Donald

Trumping Bernie

Racial Profiling in Preschool

Data from the U. S. Department of Justice, Office of Civil Rights, shows that African-American children, especially boys, are suspended or expelled from preschools at a higher rate than White children. For example, while 20% of preschool boys are Black, 45% of the boys suspended are Black. However, this is not proof of racial discrimination, since a skeptic could argue that, even at this young age, Black children are more likely to misbehave.

A new study by Walter Gilliam and his colleagues at the Yale University Child Study Center takes an experimental approach to this issue by holding the behavior of Black and White children constant and observing how teachers respond. Participants were 132 prechool teachers recruited at an annual conference. Sixty-seven percent of the teachers were White and 22% were Black. They took part in two studies.

In the first study, participants were shown a 6-minute video of four preschool children—a Black boy, a Black girl, a White boy and a White girl—seated around a table. The teachers were asked to watch for “challenging behavior,” but in fact the video did not contain any misbehavior. A computerized eye-tracking device was used to measure the amount of time the teachers spent watching each child. At the conclusion, the teachers were asked to report which of the four children required the most attention.

The eye tracking results showed that the participants spent more time looking at boys than girls, and more time looking at Black children than White children. In addition, the time spent gazing at the Black boy was significantly greater than would have been expected on the basis of his combined race and gender. The race of the teacher made no difference in this study.

The title of the paper frames the research as a study of implicit bias, and media reports of the study have followed suit. The authors define implicit bias as the “automatic and unconscious stereotypes that drive people to behave and make decisions in certain ways.” However, the teachers’ conscious appraisal of which child they paid the most attention to appeared to match the eye-tracking results fairly closely, as shown in the chart below. Apparently the teachers were well aware that they were paying more attention to the Black boy.

yale_implicit_bias_infographic_v07

I mention this because the term “implicit bias” is sometimes used to deny personal responsibility for one’s own and others’ discriminatory behavior on the grounds that it is unconscious. By labeling this as a study of implicit bias, the authors may have given their teacher-participants less blame for their behavior than they actually deserved.

In my title, I described these results as similar to racial profiling. Racial profiling targets people based on stereotypes about their race, as when the police stop and frisk Black teenagers having no evidence that they are committing crimes. Like the police, these teachers were scanning for misbehavior, and they responded by giving special attention to African-American boys. (An editorial writer for the New York Times drew this same analogy.)

These same participants also took part in a second experiment. In this study, they were asked to read a vignette describing a preschool child who repeatedly engaged in disruptive behavior. The child’s race and gender were manipulated by changing the child’s name (DeShawn, Jake, Latoya or Emily). Half the participants in each race and gender condition also read background information suggesting that the child lived with a single mother who was under a great deal of stress. The others were not given background information. The teachers were then asked to rate the severity of the child’s behavior and to recommend whether the child should be suspended or expelled.

The following results were found for ratings of the severity of the behavior.

  • The same behavior was rated as more seriously disruptive when the child was White than when he or she was Black.
  • Giving teachers background information increased the ratings of the severity of the behavior.
  • The Black teachers rated the behavior as more serious than the White teachers.
  • The background information increased the perceived severity of the behavior when the teacher was of a different race than the child, but the teachers responded more sympathetically to it when the teacher and the child were of the same race.

With regard to suspension or expulsion, the only finding was that Black teachers were more likely to recommend these options.

The results of the second study are not a good fit with the Department of Justice data, since the teachers appear to be discriminating against the White children. The researchers explain this by suggesting that these teachers expected the Black children to be disruptive, but held the White children to a higher standard. Therefore, the identical behavior was seen as more serious when attributed to a White child.

My guess is that had the same behavior been rated more disruptive when when the child was Black, the results would have been interpreted in a straightforward manner as discrimination against African-Americans. However, since the results were unexpected, a more complex explanation was presented. This explanation may be correct, of course. There is some evidence for “shifting standards” with respect to race. However, the authors could have strengthened their argument with a followup study measuring teachers’ expectations about the misbehavior of Black and White children and the extent to which the behavior described in their vignette violated those expectations.

Since the Black teachers were stricter overall, it appears that increasing the representation of Black teachers will not by itself reduce the number of suspensions and expulsions.

Some additional perspective on this issue is provided by a set of two experiments by Jason Okonofua and Jennifer Eberhardt. Their participants, grade school teachers, read a desription of either a White or Black boy in middle school who committed two infractions—one class disruption and one act of insubordination. After each incident, they were asked how severely the child should be disciplined.

There was no difference in the punishment recommended for Black and White boys after the first infraction. As shown in the table, the recommended disciplinary action increased in severity after the second infraction, but it did so more for the Black boy than for the White boy. (In the table, “feeling troubled” refers to a combined measure the the severity of the misbehavior and the extent to which it would hinder and irritate a teacher.)

Apparently, the teachers were more likely to infer a disposition to misbehave from two bad actions when the child was African-American than when he was White.

You may also be interested in reading:

White Prejudice Affects Black Death Rates

Outrage

Asian-American Achievement as a Self-Fulfilling Prophecy

White Prejudice Affects Black Death Rates

Dr. Jordan Leitner of the University of California at Berkeley and three colleagues have published a study of the relationship between White racial attitudes and the health outcomes of Black Americans. Here are some things we already know:

  • African-Americans have a higher death rate from cardiovascular diseases (e.g., heart attacks, strokes) than White Americans. (Other diseases as well, of course.)
  • The perception by Blacks (and others) that they are being discriminated against (e.g., being followed by store employees, being pulled over by the police for a minor offense) is associated with physiological stress responses known to cause circulatory problems, and with increased mortality. However, since these studies measure perceived rather than actual discrimination, a skeptic could argue that Blacks only imagined that Whites were biased against them.
  • African-Americans have higher death rates in locations where national surveys show that anti-Black attitudes are greater. But since these surveys include both Black and White respondents, it could be argued that the results were influenced by the attitudes of Black people who hate themselves.

Social psychologists distinguish between two types of prejudice. Implicit bias refers to automatic responses that are unintentional, and of which people may not be aware. Implicit bias was not related to any of the outcomed measured in this study. Explicit bias refers to responses that are deliberate and intentional. In this study, explicit bias was defined as the difference between how warmly (on a 10-point scale) participants said they felt toward White and Black Americans.

Leitner and his colleagues used a data base from Project Implicit consisting of the scores of about 1.4 million White Americans on the Implicit Association Test (IAT), a measure of implicit bias, collected between 2003 and 2013. When filling out the IAT, the participants indicated their race, age and gender, and completed the measure of explicit bias. The county in which their computer was located was determined from their Internet protocol address. Although it is large, this is not a representative sample of Americans, since the participants were younger than the average resident of their county. To correct this bias, the researchers weighted the responses of older participants more heavily. The results were the same with or without this correction.

In Study 1, racial bias was correlated with data from a 2012 telephone survey by the Centers for Disease Control (CDC), in which both race and county of residence were identified. Two questions were of interest. Access to affordable health care was measured by asking respondents whether they had ever, in the past year, needed to see a doctor but did not because of the cost. Coronary disease diagnosis was indicated by whether they reported being told by a health professional that they had a heart attack or heart disease.

In Study 2, racial bias was related to county-level statistics, also from the CDC, indicating the age-adjusted death rates from circulatory diseases of Blacks and Whites from 2003 through 2013. To control for alternative explanations, the data analyses of both studies statistically eliminated the effects of the following county-level characteristics: population, education, income, residential segregation, housing density and geographical mobility.

Below are scatterplots showing the outcomes of the two studies. Each dot represents a county and the lines indicate the statistical averages.

  1. Blacks overall reported less access to affordable medical care. More importantly, as explicit racial bias among the county’s Whites increased, Blacks had less access to affordable medical care. Explicit bias did not affect Whites’ access to medical care.
  2. However, explicit bias had no significant effect on coronary disease diagnosis among either Blacks or Whites.
  3. In the second study, they found that the higher the explicit racial bias among Whites, the more likely both Blacks and Whites were to die of circulatory diseases. However, this relationship was stronger for Blacks than it was for Whites. For example, among counties in which Whites were high in explicit racial bias, the difference between Blacks’ and Whites’ death rates from circulatory diseases was 62 per 100,000. Among counties low in explicit bias, the difference was 35 deaths per 100,000.

According to the authors, this is the first large-scale study to demonstrate that White prejudice increases the death rate due to coronary disease of African-Americans living in the same counties. However, racial bias did not affect Black death rates due to cancer. Thus, physiological stress due to discrimination and its effects on the cardiovascular system appears to be critical in producing this effect.

The results of Study 1 imply that these increased deaths were also due in part to Blacks’ reduced access to affordable medical care. The failure of prejudice to affect coronary disease diagnosis among Blacks could be related to their difficulties in obtaining health care. Diagnosis and treatment require doctor visits, but death does not.

The fact that explicit racial bias predicted Black outcomes but implicit bias did not suggests that these health outcomes were an result of conscious bias on the part of the White majorities in these counties. Failure to provide adequate health care for poor people and minorities is an outcome of social policy decisions made by politicians and by corporate executives such as the managements of hospitals and clinics. Although the present data were collected prior to the Affordable Care Act, it would not be surprising if many of these same counties were located in states that failed to take advantage of the federal government’s offer to expand Medicaid in 2014.

I suspect that White prejudice at the community level has many other effects on the lives of African-Americans in addition to limiting access to health care. Black-White wage inequality and criminal justice policies affecting Blacks would seem to be obvious topics for future research.

You may also be interested in reading:

Outrage

The Implicit Association Test: Racial Bias on Cruise Control

Old-Fashioned Racism

The Cost of Climate Inaction

A recent headline says that climate change will cost the millennial generation $8.8 trillion. But from where does this number come? The trail leads to a 2015 study by Marshall Burke of Stanford University and two colleagues from the University of California at Berkeley in which they attempted to measure the relationship between temperature and economic productivity.

We know that global temperatures are increasing, and we can estimate how much they will increase if nothing is done to mitigate climate change (the “business-as-usual” scenario). How can you measure the relationship between temperature and economic productivity? You can’t do it simply by comparing the economies of warmer and cooler countries, since there are many cultural and environmental differences between, for example, Sweden and Nigeria. But if you compare the productivity of each country during warmer- and cooler-than-usual years, each country serves as its own control group.

However, other variables that influence the economy may take on different values during warmer and cooler years. For example, a global trade agreement may have increased productivity in certain countries in certain years, and those years may also have happened to be warmer (or cooler). These confounding variables have to be measured and statistically removed from the data.

Burke and his colleagues gathered data from 166 countries over the 50-year span between 1960 and 2010. They used multiple regression to calculate the relationship between temperature and productivity, while eliminating the effects of “common contemporaneous shocks,” such as global price changes or technological innovations, “country-specific . . . trends in growth rates,” such as those produced by changing political institutions or economic policies, and the lagged effects of previous years’ temperature and rainfall. Their final curve is an average of the impact of temperature on productivity in the 166 countries, weighted by the countries’ population size.

They found that the relationship between temperature and productivity is a curve which peaks at 55 degrees Fahrenheit (13 degrees Celsius). That is, countries are most productive when their average annual temperature is 55 degrees, and their productivity declines the more the average deviates from that temperature in either direction. The curve is shown below, along with the average yearly temperatures of selected countries. The blue shaded area represents the 90% confidence interval around their best estimate. At right are separate breakdowns for rich and poor countries, years of measurement, and agricultural and non-agricultural productivity.

figure2

Next, they used this relationship to calculate the effects of expected future climate change, assuming business-as-usual, on future global income and the incomes of each country. The model predicts that global productivity will decline approximately 23% by 2100, as compared to the same future without global warming. While some cooler-than-average countries, such as Canada and Russia, will see their economies improve, the majority (77%) will see declines in income, especially those countries near the Equator. Since the countries that can anticipate the worst effects are already poorer than average, the result will be an increase in global inequality. Here is a brief presentation of their findings by Dr. Burke.

How can these results be explained? The authors found that agricultural productivity peaks at around the same temperature (see the chart above). They also mention increased energy costs and declines in health at warm and cool temperatures. Finally, they cite research showing that human cognitive errors and interpersonal conflicts increase at warmer temperatures.

Can we trust these predictions? An optimist might note that there is a danger of overestimating the damage climate change will cause if the peak in productivity at 55 degrees is actually due to confounding variables unrelated to temperature that are not controlled in their analysis. However, it’s difficult to think of phenomena not caused by temperature that would still produce a productivity curve peaking at 55 degrees.

The authors also point out that between 1960 and 2010 annual temperatures fluctuated fairly randomly. This provided little incentive for people to adapt to warmer or cooler temperatures. However, future temperatures are expected to increase consistently, which may instigate successful efforts to adapt to these warmer temperatures.

Optimists might also argue that the assumption of no climate action at all between now and 2100 is unrealistic. To the extent that effective action is taken to mitigate climate change, the loss of productivity will not be as great.

On the other hand, a pessimist could think of reasons why their analysis might underestimate climate change’s damage to the economy. The authors note that their model focuses only on the effects of temperature and those other phenomena that are directly influenced by temperature. But climate change will affect other things besides temperature, such as sea level rise and extreme weather events. If these other effects reduce productivity, the harm due to climate change will be greater than they predict.

They also note that their model predicts the effects of annual temperatures only within the range that they have been observed between 1960 and 2010. But if global temperatures increase substantially, the future may not be predictable from the past. For example, if temperature increases cause sustained droughts over large areas, the cumulative effects on agricultural productivity may be much greater than the effects of any known previous droughts. In reality, we probably have little idea of what future catastrophes await us.

We can now return to the effect of climate change on the incomes of millennials. Two nonprofits, Demos and NextGen Climate, have published an analysis of the lifetime cost of climate change to American millennials, using the data from Burke and his colleagues. The Burke analysis predicts that, in the absence of climate action, the United States economy will shrink 5% by 2050 and 36% by 2100—slightly more than the global average of 23%.

Millennials are typically defined as people born between the early 1980s and the early 2000s. The Demos/NGC paper calculated the lifetime earnings lost by Americans who turned 21 in 2015 (born in 1994) and those born in 2015. This is simply a matter of arithmetic, and the formulas are given in their appendix. Using these formulas, you can calculate the cost of climate change to any birth cohort. Obviously, the later the birth year, the greater the cost. The $8.8 trillion figure is the aggregated cost to all millenials.

The chart below illustrates the average cost of climate change to Americans turning 21 in 2015, calculated separately for college graduates and non-graduates.

nextgen-figure-3

The second chart compares wealth lost by 2015 college graduates due to climate change to two other drains on the income of their generation—college debt and the lingering effects of the Great Recession.

lifetime-lost-wealth

Of course, the accuracy of these figures depends entirely on the validity of the analysis by Burke and his colleagues.

You may also be interested in reading:

Snow Job

Deep Background

Crime in Slow Motion

The research I’m about to present resonates with a personal experience of mine. Three years ago, I served on a jury that acquitted Cheswick, PA, councilman Jonathan Skedel on a charge of assaulting Joe Ferrero, president of the Cheswick Volunteer Fire Department. (I was stunned when the prosecutor allowed a retired college professor whose field is social psychology to sit on the jury.) The charge resulted from a fistfight between the two men in which Ferrero suffered facial injuries requiring dental surgery. The fight took place in the parking lot of a physical therapy clinic and the entire episode was captured by one of our ubiquitous surveillance cameras.

The video was played several times during the trial, both at real speed and in slow motion. In his summation, the prosecutor paused the video just before Mr. Skedel delivered the punch which injured Mr. Ferrero, and stated that Mr. Skedel could have stopped the fight at that point, but instead decided to assault Mr. Ferrero.

During the jury’s deliberations, I was disturbed to discover that some of my fellow jurors accepted the prosecutor’s definition of the situation. I tried my best to argue—with limited success—that pausing the video was an artificial intervention in what was, in reality, a continuous episode that provided little opportunity for conscious deliberation by either man. The jury eventually acquitted Mr. Skedel, but this was probably due to the majority’s belief that both men had acted equally badly, and it was unfair to single out one of them for prosecution.

Playing crime scene videos in slow motion, or pausing them at critical points, is common practice in jury trials and their effects should be investigated. The former of these issues was the subject of four experiments by Dr. Eugene Caruso of the University of Chicago and his colleagues. They compared the effects of watching a video either in slow motion or at regular speed. Their slow motion was 2.25 times slower than regular speed. The researchers measured participants’ estimates of how much time had passed, and their judgments of the intentionality of the defendant’s behavior.

Three of these experiments used a surveillance video from a Philadelphia trial in which the defendant, John Lewis, was convicted of first degree murder for  shooting a man during a convenience store robbery. Here it is (in slow motion).

They measured the intentionality of the act because the real jury had to decide whether the defendant was guilty of first degree murder, which is premeditated, or second degree murder, which is not.

Study 1 showed that participants in the slow motion condition estimated that more time had passed than those in the real time condition, and saw the defendant’s behavior as more intentional. Further analysis showed that their judgments of intention were mediated by their estimates of how much time had passed. The researchers refer to this effect of slow motion on perceived intentionality as the intentionality bias. It occurs because the participants mistakenly infer that the defendant had more time to think before acting than he actually had. Study 2 replicated this finding with a video of a professional football tackle involving violent contact. (You might want to remember this the next time you watch a slow motion replay during a sports event.)

Mr. Lewis’s lawyers argued on appeal that showing the slow motion video had biased the jurors, causing them to see his actions as more intentional than they actually were. The judges rejected this argument because, they said, the jurors were shown the video at regular speed as well as slow motion, and because the amount of elapsed time was stamped on the video.

The researchers effectively demolished both of these arguments. Study 3 added a “time salient” condition in which participants were reminded that they could see how much time had elapsed from the time stamp on the videotape (which was present in all conditions). This reduced the amount of intentionality bias produced by slow motion, but did not eliminate it. Finally, Study 4 included a condition in which participants were shown the video twice, first at regular speed and again in slow motion. This too reduced the magnitude of the intentionality bias but did not eliminate it.

Summarizing the data, the researchers calculated that, prior to deliberation, juries randomly composed of Study 1 participants would be almost four times as likely to unanimously believe that the killing was premeditated in the slow motion condition.

Unfortunately, Dr. Caruso and his colleagues did not include a condition in which the video was paused immediately before the critical action took place. My guess is that such a condition would have further increased the intentionality bias, since it stretches the length of the presentation.

The use of slow motion is often justified on the grounds that it provides a “better” look at an event, and this may be true in some instances. However, when intentionality is at issue, slow motion also produces a biased causal attribution for the event. These studies are probably too late to help Mr. Lewis, who was sentenced to death and is awaiting execution.

You may also be interested in reading:

A Downside of Police Body Cameras

Francesca

We have a new family member, a Norwich terrier mix whom we’ve named Francesca. She will be 3 years old in August and weighs only 19 pounds. Like most terriers, she has a high energy level, especially when there are squirrels in the vicinity.

Francesca
                                                         Francesca

We adopted her from the same place we got her “brother,” Django. Tracy’s Dogs is a San Antonio nonprofit that rescues dogs from high-kill shelters in Texas. On May 21, they transported 63 dogs to Monroeville, PA, all of whom that had been adopted by folks in the area.

It is probably not unreasonable to refer to dogs as family members. A recent series of five studies by David Rubin and his students looked at misnaming, in which a person incorrectly calls a familiar other person by someone else’s name. Approximately 1700 participants were asked to recall instances in which they were misnamed or had misnamed someone else. In most cases, people are misnamed with the name of someone else in the same semantic category, i.e., a family member misnamed as another family member, a friend as another friend.

Francesca and Django
                            Francesca and Django

As an aside, the researchers reported 42 instances of misnaming involving pets. All but four of the pets were dogs. In 41 of these cases, the dogs were misnamed as family members, or vice versa. Although participants reported living with cats almost as often as dogs, only dogs were involved in misnamings, suggesting that dogs are regarded as family members while other pets are not.

 

You may also be interested in reading:

Our New Family Member, Django

Wasted Opportunities

This post is for those of us who are bothered by all the time we waste in the doctor’s waiting room. Our annoyance is explained by the economic concept of opportunity cost. Opportunity cost refers to the other more productive and/or enjoyable things we could be doing with this lost time.

Kristin Ray and her colleagues attempted to measure the opportunity cost of a doctor visit. Their two studies (here and here) are summarized in this video.

https://www.youtube.com/watch?feature=player_embedded&v=7Ot-LpNEWJY%20

Wouldn’t it be great if more researchers posted short videos like this one summarizing the results of their research?

Here are a couple of loose ends I’d like to clear up:

  • Measuring opportunity cost is simpler for the employed people in the sample, since they were asked to report their hourly wages. For those not employed, demographic variables (age, sex, education, etc.) were used to estimate their hourly wages, which were then adjusted for the probability that someone in their demographic category would be employed. Bottom line: The researchers assume that the time of non-employed people is less valuable. (Some of us might want to contest that assumption.)
  • If you’re puzzled by their estimate of $32 as the average cost of a doctor visit, note that this is the out-of-pocket cost. The average real cost was $279, but most of it was paid by insurance.
  • The extra 25 minutes spent by minorities and unemployed people was not explained by length of time spent face-to-face with the doctor, and their travel time was only slightly longer. Most of it was extra time spent in the waiting room. This suggests that these folks go to doctor’s offices that are more crowded or that schedule their patients less efficiently.

I’m not optimistic about the potential to remedy this situation because, with the possible exception of those who cater to the wealthy, I don’t see that physicians have much incentive to make their services more user-friendly.

You may also be interested in reading:

Don’t Worry, Be Happy?

Outrage

Self-Censorship

Suppose you were completing an online survey and encountered the following warning:

The next section of the survey asks for your honest opinions about some controversial political issues. While we make every attempt to ensure your opinions are kept confidential, it is important to keep in mind that the National Security Agency does monitor the online activities of individual citizens, and these actions are beyond the study’s control.

That statement is absolutely true, but how often do we think about it? And if we do think about it, will it make any difference in our responses to the survey?

Social psychologists have been studying conformity for 80 years. Conformity refers to a change in a person’s attitude or behavior due to real or imagined pressure from another person or group. In the 1940s, using a perceptual task, Solomon Asch demonstrated how many people conform to the incorrect judgments of others. His research was followed a series of studies documenting many factors that affect the conformity rate, i.e., we conform more if we think the other group members are experts, if we like them, etc.

In the Asch conformity experiments, participants were asked which line, A, B or C, is equal to the line at the left.

There are two main reasons for conformity. People use other group members as sources of information and sometimes conclude that the others are right and they are wrong. In other words, they privately accept to group’s judgment. Social psychologists call this informational social influence. In other cases, people privately believe their own judgment is correct, but they conform to gain the approval (or avoid the disapproval) of other group members. In this case, there is public compliance without private acceptance. This is called normative social influence. One way to disentangle these two forces is to do an experiment in which some participants respond publicly, while others think they are responding privately. Conformity will usually be greater in the public condition, and the difference between the two conditions represents normative social influence. However, conformity also will be greater than zero in the private condition, and this represents informational social influence.

The spiral of silence theory of German political scientist Elisabeth Noelle-Neumann described the consequences of conformity for society. Due to fear of social isolation, people who think they hold minority positions will tend to remain silent in social situations, while people who think they are in the majority will be confident and vocal. Over time, the perceived consensus will drift toward the majority position, as everyone overestimates the percentage of people who hold that view. For example, if atheists are reluctant to reveal their beliefs to survey researchers, the public will underestimate the number of atheists and atheists will feel more marginalized. In extreme cases, there is evidence of pluralistic ignorance, in which the majority privately rejects a position, but mistakenly assumes that others accept it, and therefore go along with it.

A recent study by Elizabeth Stoycheff of Wayne State University measures normative social influence on the internet. It contrasts a condition in which participants are reminded that they are under surveillance with one in which they are not. The 255 participants are a representative sample of Americans recruited through Survey Systems International in January 2015. They are told they are participating in a study of attitudes toward the ongoing American bombing of Iraq and Syria. After answering questions about their political attitudes, personality traits and media use, half of them were given the above message reminding them of the possibility of government surveillance. They other half didn’t receive the message. Next they encountered a Facebook post which passed along, without comment, a fictitious, neutral ABC News article about American bombing of ISIS targets in Iraq.

The measure of self-censorship was they average of their responses to four questions: their likelihood of sharing the post, liking the post, commenting on the post, or writing a new post on the same topic. They were asked whether they approved of U. S. airstrikes in Iraq, and to estimate the average American’s attitude toward the airstrikes. The difference between these two measures determined what the author called their climate of opinion—how much they thought their attitude deviated from the majority viewpoint. Participants were also asked whether they thought government surveillance of the internet was justified.

There would be evidence of self-censorship if those participants who were reminded of surveillance were more likely to speak out when they thought the climate of opinion was friendly and less likely to speak when they thought it was hostile. Although some secondary sources have implied that this is what Stoycheff found, the actual results are more complicated than that. She divided people into three groups depending on their attitude toward surveillance: Those who thought it was justified, those who merely tolerated it, and those who thought it was unjustified. The results are shown below.

329F806A00000578-3513034-image-a-88_1459202220023Those who thought surveillance was unjustified showed no evidence of self-censorship. They were slightly less likely to speak when under surveillance, but their likelihood of speaking was unaffected by the climate of opinion. Those who believe that government spying on citizens is unacceptable apparently refuse to be silenced even when they know the opinion climate is hostile to their views and they are reminded that they are under surveillance. Stoycheff reports that these people are also higher in political interest than the other participants.

However, those who tolerated surveillance, and especially those who thought it was justified (“because [they] have nothing to hide,”] showed evidence of self-censorship. They were more likely to speak out when they thought they were in the majority, and less likely to speak out when they thought they were in the minority. They conform in two ways. First, they acquiesce to government spying, and secondly, they censor their opinions by telling other people only what they think they want to hear.

Conformists cheat the group or society by withholding whatever information or good judgment they possess. But as Stoycheff notes, “Democracy thrives on a diversity of ideas, and self-censorship starves it.” Better outcomes will come to a group or society that creates incentives for people to reveal dissenting information. The First Amendment is an important safeguard when conformity is demanded by the government, but freedom of speech may not be sufficient if people decide that they have nothing to say.

You may also be interested in reading:

Chomsky, Greenwald and Snowden on Privacy

Are Terrorists Getting What They Want?

Are Terrorists Getting What They Want?

When terrorists attacked the Brussels International Airport and a metro station on March 22, killing 31 people and injuring 340, the response in this country was predictable. The corporate media provided blanket coverage of the attack, but failed to address its causes. The presidential candidates called for more of our current policies—on steroids. Donald Trump advocated revising international law to allow waterboarding and other unspecified forms of torture. Ted Cruz suggested “patrol(ing) and secur(ing) Muslim neighborhoods before they become radicalized.” Hillary Clinton made a vague call for increased cooperation between the technology community and government. While she did not spell out the surveillance implications of this cooperation, it can only mean that she accepts the Obama administration position that there can be no telephone or internet communication between American citizens that can’t be accessed by the federal government.

What do we know about the effects of terrorist actions such as the Brussels bombings on public attitudes?

On July 7, 2005, a small group of terrorists affiliated with al Qaeda carried out an orchestrated set of attacks on the London subway and bus system, killing 52 people and injuring 770. By a fortunate coincidence, a group of researchers headed by Julie van Dyver at the University of Kent had conducted survey measuring intergroup prejudice among a nationally representative sample of about 1000 U. K. residents six weeks before the July 7 attacks. They repeated the survey with an equivalent group of British people four weeks after the attacks.

The two surveys measured negative attitudes toward Muslims and toward immigrants, and political orientation—that is, whether the participant favored the political left (Labour party) or the political right (Conservative party). They predicted that the effect of the bombings would be to increase negative attitudes toward Muslims and immigrants of all nationalities, but that not everyone would be equally affected. Based on what they called the reactive liberals hypothesis, they expected the shift to be greater among liberals than conserva- tives, since conservatives already held negative attitudes toward Muslims and immigrants before the bombings.

Here are the results for prejudice toward Muslims.

As predicted, the liberals showed a significant increase in anti-Muslim bias, but the conservatives did not change. In other words, the effect of the terrorist threat was to cause liberals to think more like conservatives. The results for prejudice toward immigrants were nearly identical.

If liberals are more influenced by terrorism than conservatives, can this be explained by changes in their basic values? The moral foundations theory of political ideology proposes that liberals and conservatives hold different values. Liberals place a higher value on harm reduction and fairness, while conservatives place a higher priority on ingroup loyalty and respect for authority. Previous research not only supports these predictions, but it also shows that in-group loyalty and respect for authority are predictive of greater prejudice toward minorities, while harm reduction and fairness are associated with lower prejudice. These results are consistent with the well-established finding that conservatives are more prejudiced than liberals.

The London surveys included items measuring these four values. Liberals showed an increase in in-group loyalty and a decline in concern with fairness as a result of the bombings, while conservatives’ concern for these values was unchanged. (Neither liberals nor conservatives changed their attitudes toward harm reduction or respect for authority.) Finally, the researchers’ statistical analysis showed that these changes in attitudes toward Muslims and immigrants were mediated by the changes in the basic values of in-group loyalty and fairness. (See this previous post for an explanation of how mediational hypotheses are tested.)

Many progressive commentators, beginning with Noam Chomsky in his 2001 book, 9/11 (now in its second edition), warned that the United States and Europe were falling into a trap set by Osama bin Laden. As Tom Engelhardt, Glenn Greenwald, and others have also pointed out, the West is continuing to follow the terrorists’ “playbook.”

The short-term strategy behind 9/11 and subsequent terrorist attacks was to provoke outrage against Muslims among Western populations, in the hope that their governments would overreact by bombing and invading Middle Eastern countries. Their greatest success was George W. Bush’s ill-advised invasion of Iraq, which destabilized the country and led to the establishment of the Islamic State (ISIS). Since the most important predictor of suicide terrorism is the perception by its perpetrators that their homeland is occupied or threatened by foreign military forces, such actions have the effect of recruiting more terrorists.

In fact, as early as 2004, a secret study commissioned by the Defense Department acknowledged that the primary cause of Muslim terrorism was American foreign policy, but knowing that we had no intention of changing our policies, its authors suggested “transforming our strategic communications”–that is, reframing our propaganda directed at Muslims.

A second reason for terrorism, according to this analysis, is to provoke Americans and Europeans into harassing and discriminating against their domestic Muslim populations. If Muslims living in the West are convinced that they can never be assimilated, they will initiate local acts of terrorism, as in San Bernadino, Paris and Brussels. The combined effect of increased military action abroad and repression of Muslims at home is to create a self-perpetuating military machine which recruits many more terrorists than it is able to kill.

The endgame of al Qaida and ISIS is to convince the U. S. and Europe to withdraw completely from the Middle East by drawing us into a series of long, expensive and ultimately unsuccessful ground wars in the Persian Gulf. In this way, they hope to end the West’s economic exploitation and cultural influence on the region.

The study of the London bombings, which its authors entitled “Boosting Belligerence,” suggests that, when Muslim terrorists attack Western countries, the effect on public opinion is exactly what they are hoping for—increased support for right-wing political candidates, an aggressive foreign policy, and repressive domestic policies. It seems to follow from the political analysis of Chomsky and others that ISIS would prefer a Republican to be elected the next president of the United States. Donald Trump is ideally suited to their purposes. Assuming the election is close, ISIS could easily influence its outcome by scheduling a few small-scale terrorist attacks in the weeks leading up to Election Day.

You may also be interested in reading:

Deep Background

On Obama’s Speech

Chomsky, Greenwald and Snowden on Privacy

Chomsky, Greenwald and Snowden on Privacy

Last night, there was a panel discussion entitled “A Conversation on Privacy” at the University of Arizona, featuring Noam Chomsky, Glenn Greenwald and (live from an undisclosed location) Edward Snowden. The discussion was moderated by Nuala O’Connor of the Center for Democracy and Technology. The video is about two hours long. Although it starts off slowly, your patience will be amply rewarded by the end.

No, this isn’t a lobby card from Frankenstein Meets the Wolf Man. It’s actually an image from the poster advertising the panel discussion.

The discussion was broadcast via Livestream, which I was unable to insert directly into the blog. To watch it, click on the link below. If you’re familiar with the participants, you can skip the introductions. The discussions starts about 11 minutes into the video.

http://livestream.com/azpm/events/4958510/videos/116998760