Given the trends over time, it is worth considering whether any particular region, education, or cohort group is responsible for the declines in acceptance of these explanations.³  On the question of discrimination, it is interesting that neither education nor cohort is consistently a significant predictor of holding this belief.  In studies of racial attitudes more generally, southerners typically hold more conservative racial attitudes than non-southerners (Schuman et al. 1997; Tuch and Martin 1997).  This was certainly true in earlier years:  southerners were more likely to deny discrimination as a cause of racial inequality than non-southerners. While both regions of the country show declining endorsement of this explanation, interestingly it is the non-southerners who have declined more substantially.  Indeed, by 2004, there is no longer a statistically significant difference between southerners and non-southerners in endorsing this explanation; this convergence came about because non-southerners came to look more like southerners rather than the other way around.   
 

Regional differences have also disappeared for the question of whether differences in in-born ability to learn causes racial inequality.  But level of education differences persist on this belief, with most of the overall decline due mainly to changes among the least educated (who also had the most room to move):  from 1977 to 2004, the percentage of those with less than a high school degree who endorsed this belief dropped from just over 4 in 10 to just over 1 in 10.  The other two education levels also show a decline, though less dramatically, since the levels in 1977 were so much smaller to begin with.  In 2004, there remains an education difference: 4 percent of those with some college endorsed this belief, while about 12 percent of those with less education did.   
 

Lack of access to education as an explanation for inequality declined in the overall time trends, as reported in Figure 5.  The pattern for this question follows a similar pattern to the discrimination explanation with respect to its relationship to levels of education.  In the most recent survey, there were no differences between those with more and less education—all three groups showed declines in endorsing a lack of access to education as a reason, in comparison to their 1977 levels.  However, the decline was steeper for those with some college, where it dropped from 62 percent to just 42 percent in 2004.  Those with less than a high school education declined from 44 percent to 38 percent over this same time period, and those with a high school degree dropped from 46 percent to 36 percent.  In short, the pace of change has been slightly different across education levels, and has lead to a convergence across education levels.  
 

Finally, endorsing the belief that blacks have less motivation has, as noted above, shown declining support over the past several decades, though it has and continues to be the modal response.  Here there are few differences by region or education: all groups have shown about the same levels of declines.  But education and region continue to be significantly related to this response:  those with more education and those outside the south are somewhat less likely to select it than the less educated and the southerners.   
 

Of the four explanations for inequality, only this question on motivation shows a cohort difference.  Whites from the pre-Civil Rights cohort are consistently more likely to endorse this reason than younger cohorts.  To some extent, then, the overall declines in selecting this reason are driven in part by increasing levels of education and cohort replacement as the oldest generations who most strongly held this view are replaced by younger and more educated cohorts.   
 

Perceptions of Discrimination.
The slow but steady decline among whites for the discrimination explanation of inequality is of significance, given its central role in shaping public opinion about racial policies.  If whites do not perceive blacks as facing discrimination, then racial policies that are predicated on the notion that blacks continue to experience discrimination will likely fail in the court of public opinion.  A series of new questions provided by the Gallup Poll tap white perceptions of discrimination in general (as opposed to its role in explaining racial inequality).  Results from 1997 to 2004 show both little change over time, and relatively low levels of recognition that blacks are “treated less fairly” than whites in a range of venues, including on the job, in neighborhood shops, downtown, in restaurants and by the police.  The percentage of whites who acknowledge discrimination against African Americans in public settings and employment generally ranges in 2004 from 10 percent (in restaurants) to 15 percent (downtown).  The only venue where whites in greater numbers perceive unfair treatment is by the police.  In 2004, just over 1 in 3 whites reported they thought blacks were treated less fairly by the police.   Between 1997 and 2004, there has been a steady but slow decline in recognition of unfair treatment across all of the venues included in the Gallup surveys except for the police, where the figures have stayed relatively constant. 
 

Endorsement of Racial Stereotypes. 
Underlying some of the explanations of inequality reviewed above (lack of motivation, less in-born ability) are the beliefs that people hold about the characteristics of racial groups—what are commonly referred to as racial stereotypes.  Only since 1990 have national surveys again regularly included measures of this important dimension of racial attitudes.  In part, the earlier disappearance of these kinds of questions from surveys was due to increasing social pressures against reporting negative racial stereotypes.  Thus, survey researchers have worked to create more subtle measures of stereotyping and here we report one approach.  Survey respondents are presented with a seven-point scale anchored on one end by a negative trait (poor, lazy, and unintelligent) and the other by a positive trait (rich, hardworking, and intelligent).  Respondents are first asked to rate whites as a group on this seven-point scale, and then in a separate question asked to rate blacks as a group.  One can then calculate for each respondent a difference score—a simple subtraction of the score they gave to whites and the score they gave to blacks.  In the tables, we report the percentage of respondents who rate whites more positively, who rate no difference between whites and blacks, and who rate blacks more positively.