PRINCETON, NJ -- Barack Obama leads John McCain by a significant margin among voters with the most education, but trails the likely Republican nominee among voters with the least formal education.
Thus, it appears that Obama's in the Democratic nomination contest would extend to the general election as well.
These results are based on aggregated data from Â鶹´«Ã½AV's daily tracking survey, spanning March 31 to April 6 interviewing. Obama and McCain were tied at 45% for the general election among all registered voters in these data.
Obama's weaker performance among voters with less education is inconsistent with the typical pattern seen in general-election contests, in which the Democratic candidate tends to do better than the Republican among this group.
But Obama makes up for his poorer performance among voters with the least formal education by doing well against McCain among college graduates -- particularly those with postgraduate education (who tend to vote Democratic) but also on a relative basis among those with a four-year college degree only, a reliable Republican group. Among four-year college graduates, Obama and McCain run even.
The education patterns for the Obama-McCain trial heats seem fairly well established thus far -- McCain has bested Obama among voters without a college education each of the four full weeks that Â鶹´«Ã½AV has tracked general-election voting preferences, and Obama has had a consistently large lead among postgraduates. The two candidates have been competitive among the two middle education groups -- voters with some college education and four-year college graduates only.
As , the would look different in a McCain-Obama matchup compared with a McCain-Clinton matchup.
In a hypothetical Clinton vs. McCain race, the vote by education patterns is typical for a general-election contest. Clinton holds an advantage over McCain among voters with a high school education or less, and McCain runs strongly among those with a bachelor's degree only. Even so, McCain maintains considerable appeal to voters with less education against Clinton, and he edges her out among postgraduates.
McCain's large advantage versus Clinton among voters with only a bachelor's degree has been apparent each of the four weeks of tracking, as has his smaller advantage among those who attended college but did not graduate. Clinton has typically held an edge among those with a high school education or less, while the two have generally been competitive among postgraduates.
The relationships between education and the vote persist even when taking into account a voter's age (since age and education tend to be correlated). Within different age categories, for the most part, support for Obama versus McCain tends to increase at higher levels of voter educational attainment, and support for Clinton versus McCain decreases at higher education levels.
Interestingly, the vote-by-education effects are primarily evident among Democrats and independents in an Obama-McCain matchup. Republicans solidly back McCain regardless of their education. But in a Clinton-McCain contest, the education effects largely disappear for all party groups once party is taken into account.
Implications
Â鶹´«Ã½AV analysis has demonstrated that in the Democratic nomination contest, Clinton's strength is currently among voters who are high school graduates or less and Obama's is with college graduates and postgraduates. But even in a general-election contest, educational differences are still evident.
If Obama is the Democratic nominee, McCain would appear to have a good chance of making inroads among voters with a high school education or less, who normally vote Democratic, but he would find himself in a tougher-than-usual fight for those who have a bachelor's degree only.
A Clinton-McCain election would be fought on more familiar partisan turf, with Clinton faring better among voters with the least and the most formal education, and McCain receiving solid support from those with a bachelor's degree.
The fact that in a Clinton-McCain contest, the education differences largely disappear when taking into account voters' party affiliation adds weight to the notion that such a contest would largely be fought along party lines, while an Obama-McCain matchup could change some of the normal Democratic and Republican political coalitions.
Survey Methods
Results are based on telephone interviews with 6,158 registered voters, aged 18 and older, conducted March 31-April 6, 2008, as part of Â鶹´«Ã½AV Poll Daily tracking. For results based on the total sample of registered voters, one can say with 95% confidence that the maximum margin of sampling error is ±1 percentage point.
For the sample of 1,440 registered voters with a high school education or less, the maximum margin of sampling error is ±3 percentage points.
For the sample of 1,936 registered voters who have some college education, the maximum margin of sampling error is ±2 percentage points.
For the sample of 1,388 registered voters with a four-year college degree but no postgraduate education, the maximum margin of sampling error is ±3 percentage points.
For the sample of 1,350 registered voters with some postgraduate education, the maximum margin of sampling error is ±3 percentage points.
Interviews are conducted with respondents on land-line telephones (for respondents with a land-line telephone) and cellular phones (for respondents who are cell-phone only).
In addition to sampling error, question wording and practical difficulties in conducting surveys can introduce error or bias into the findings of public opinion polls.
