Appendix A: Arguments Used
Some whites claim to be victims of affirmative action programs. Nonsense! White Americans have long benefited from a society biased toward white interests, so any current preferences for minorities are only fair. There are no innocent victims of affirmative action. Therefore, we should all support affirmative action programs.
The largest group of Americans to benefit from affirmative action thus far are women. Before 1964, women were excluded from many higher paying occupations and professions based on stereotype, custom and law. There were virtually no women police officers, lawyers, or doctors, for example. Progress has been made, but women still need affirmative action programs.
Nothing in the Constitution prohibits affirmative action. In fact, the Supreme Court upheld affirmative action programs in education in a landmark case. In this case, the Court explicitly stated that "affirmative action is consistent with the Constitution."
When a company with a history of past discrimination passes over a white man and hires a qualified minority or woman instead, that isn't "reverse discrimination." When black professional athletes were first hired, breaking the "color barrier" in sports, some white ballplayers lost job opportunities. But that was not "reverse discrimination," it was a first step toward ending discrimination.
In the historic words of one AfricanAmerican leader, "America has given the Negro people a bad check marked insufficient funds." It is about time that America makes good on its promise of opportunity for all. Affirmative action programs are a necessary first step toward racial equality in America.
In 1990, the average black male worker earned just $731 for every $1,000 earned by a white male worker in a comparable position. Moreover, though white males make up only 43% of the workforce, they occupy 97% of America's top executive positions. After decades of discrimination, only tough affirmative action programs can level the playing field.
Affirmative action programs are very effective. A study from the Clinton administration shows that the percentage of blacks entering the fields of law and medicine has increased from less than 2% to over 10% in the past 20 years. Affirmative action is working.
Who says racism is dead in America? Far from it. Surveys show that a majority of white Americans still believe that African and Latino Americans are less intelligent, less hard working and less patriotic than whites. Affirmative action programs are an important step toward changing these racist attitudes.
Affirmative action plans treat people based on race, not past or present circumstances. Middle class blacks are given preferences while lower class whites are not! This is unfair reverse discrimination and is itself a form of racism. Affirmative action programs must stop.
Many of the victims of affirmative action are AsianAmericans who have been excluded from top schools due to racial quotas. But they had no role at all in the country's history of discrimination against blacks and they are truly innocent victims! Affirmative action programs are doing more harm than good.
According to a prominent AfricanAmerican economist, under affirmative action, blacks often get admitted into schools and programs even though they have worse credentials than most white applicants. As a result, their dropout rate is higher. Affirmative action plans harm both blacks and whites and should be stopped.
The Constitution absolutely prohibits racial discrimination, including affirmative action. As one landmark case declared, "our Constitution is colorblind, and neither knows nor tolerates classes among citizens." Therefore, affirmative action plans are unconstitutional.
The preeminent AfricanAmerican leader of all time put it best: "Men should be judged by the content of their character, not the color of their skin." Clearly this statement recognizes the injustice of any form of racial preferences. In other words, even the most famous black leader in American history opposed affirmative action!
Merit has always been the most important factor determining success in this country. People of all races and classes can get ahead if they are willing to work. Unfortunately, some Americans expect to be handed a free lunch. Opportunities exist for all, but you have to be willing to pull your weight. Affirmative action violates the merit principle and should be ended.
In a recent national poll, 50% of Americans said they oppose affirmative action. It seems that most of our laws these days favor minorities, and Americans are getting fed up. If a majority of American citizens believe that affirmative action programs are unfair, then why have these laws not been repealed? End affirmative action now!
Affirmative action programs at American universities "stigmatize" African Americans and other minority students who are assumed to be incompetent because they were admitted based on color, not on merit. Individuals, whether black or white, are far more likely to be successful if they prove their abilities in equal competition rather than receiving unfair and unearned advantages. Affirmative action works to the disadvantage of minorities.
Appendix B: Additional Analyses for Study 2
The question of causality is best dealt with by the experiment in study 3. Nonetheless, below we discuss additional analyses run on study 2 data to examine the plausibility of alternate causal routes.
The most plausible alternate explanation is that some personal characteristic is acting as a confounding variable, such that it causes people to both select into more heterogeneous networks and seek out more information. If people are selecting into more heterogeneous networks in general on the basis of some such personal characteristic that also governs information-seeking (e.g. curiosity), then this should be the case not just for affirmative action, but for other political attitudes as well. Yet, we find that network attitudinal composition with respect to another politically salient issue (gun control) was not significantly correlated with network composition on affirmative action (r = 0.12, n.s.).
Specific potential confounding variables include need for cognition (high in people who like to think deeply), and political interest. Other potentially useful control variables include global processing speed (which we estimate using the average time spent reading two instruction screens about the information board), social desirability (using the self-monitoring scale to assess the extent to which people try to fulfill social expectations; Snyder 1974), and need for closure (which assesses a desire for order and correctness over ambiguity; Kruglanski & Webster 1996).Were-ran the main models in Tables 2 and 3 and included all of these useful or potentially confounding variables as controls.
As seen in Tables 2a and 3a below, our results remain substantively unaltered and network heterogeneity still significantly predicts engagement with the information seeking process, and spending more time reading counter-attitudinal arguments. Interestingly, the alternate network heterogeneity variable also has independent effects. In the case of opting out (Table 2a), this might be indicative of some third variable additionally at work (although not accounting for our results), or possibly of effects of heterogeneity at the party id level. This explanation is inadequate for explain time spent reading messages, however, because these effects (Table 3a) are actually in the opposing direction. While this deserves further attention in future work, no other potential confound approached significance. Combined, these findings provide reassurance about the robustness and direction of causality in these data.
Table 2a. Probit Estimates of Network Heterogeneity on Opting Out of Information Board
β / SEConstant / 2.517 / 2.190
Network Heterogeneity / -2.047* / 1.191
Party ID / -0.013 / 0.763
Gender / -0.699 / 0.458
Need Cognition / 0.838 / 1.587
Network Heterogeneity (Gun Control) / -2.194* / 1.153
Processing Speed / -0.001 / 0.005
Need Closure / -3.063 / 2.411
Political Interest / 0.448 / 0.866
Self Monitoring / 0.393 / 1.004
Pseudo R2 / 0.1775
N / 51
* when p0.05 on a one-tailed test
Table 3a. Effects of Network Heterogeneity on Time Spent Viewing Info Boxes
Fixed Effects / β / SEIntercept / 13.74 / 6.96
Pro-attitudinal / 0.28 / 0.72
Counter-attitudinal / 2.33 / 1.70
Network Heterogeneity / 9.31 / 7.28
Pro-attitudinal * Network Het / 7.36 / 4.67
Counter-attitudinal * Network Het / 19.38* / 10.51
Network Heterogeneity (Gun Control) / -2.96 / 6.34
Pro-attitudinal*Network Het (Gun Control) / -7.07* / 4.22
Counter-attitudinal*Network Het (Gun Control) / -6.29 / 9.23
Party ID / 3.08 / 2.82
Gender / -3.13** / 1.53
Need Cognition / -0.24 / 5.60
Processing Speed / 0.01 / 0.02
Need Closure / -8.53 / 7.73
Political Interest / -1.56 / 2.94
Self Monitoring / 2.62 / 3.46
Random Effects / SD
Intercept / 7.14
Pro-attitudinal / 2.08
Counter-attitudinal / 10.74
Residual / 3.37
N (obs) / 168
N (groups) / 49
Deviance / 1069.8
** when p0.05 on a two-tailed test, and * when p0.05 on a one-tailed test
DV is the average time spent per info box in seconds
References
Kruglanski, A. W., & Webster, D.M. 1996. “Motivated closing of the mind: ‘Seizing’ and ‘freezing.’” Psychological Review 103(2): 263–283.
Snyder, M. 1974. “Self-monitoring of expressive behavior.” Journal of Personality and Social Psychology 30(4): 526–537.
1