Values Did Count
John C. Green and
Last November, as is customary in the wake of a national election,
journalists rushed onto the field of battle while the smoke was still
clearing to search out the meaning of it all. What greeted them was a
national exit poll showing that a small plurality of Americans chose “moral
values” as “the one issue that mattered most” to their presidential vote.
This datum seemed to fit the red state/blue-state maps omnipresent on
election night, suggesting that once again the country had narrowly picked
the conservative side in a geographically organized culture war.
“Take an afternoon off for recriminations, a morning for whining, then race
through the Kubler-Ross stages of grief and get back to work,” wrote
Boston Globe syndicated columnist Ellen Goodman November 6. “Job No. 1:
After an initial avalanche of like-minded commentary emphasizing the crucial
impact of moral-values voters, however, the small size of their actual
plurality began to register. The best measure of the exit poll data shows
that 23.7 percent of voters selected “moral values” as their most important
issue, just ahead of the 21.1 percent who selected “the economy and jobs,”
who in turn barely beat out the 20.3 percent who named “terrorism,” who were
themselves not far in front of the 15.7 percent who pointed to “Iraq.” The
remaining 19.2 percent gave priority to other economic concerns—health care,
taxes, and education.)1
Almost immediately, a cascade of revisionist punditry declared moral values
to be a bogus hermeneutic for comprehending the re-election of the
president. Whatever “moral values” meant—and couldn’t it mean different
things to different people?—the exit poll numbers showed that the American
electorate in 2004 had diverse issue priorities.
Religious conservatives, so the new interpretation went, played a less
central role in Bush’s victory than first advertised—and perhaps a very
modest one. As Washington Post columnist E.
J. Dionne put it on November 9, “John Kerry was not defeated by the
religious right. He was beaten by moderates who went—reluctantly in many
cases—for President Bush.”
Although the revisionist line had something to be said for it, like the
original interpretation, it erred in seeking a single phenomenon to explain,
across the board, the presidential result. Thanks to the Electoral College,
however, we choose our presidents not by national referendum but one state
at a time. What if, for example, Americans in some parts of the country
selected “moral values” in much higher proportions than those in other
places? Might that not mean that moral-values voters were more important to
the president’s victory than the national totals imply?
regional breakdown of the exit poll results suggests that such was indeed
As Table 1 shows, Bush and Kerry each won four regions of the country. As
defined by Greenberg Center’s Religion by Region project, Bush carried the
Southern Crossroads, Mountain West, and South handily and just squeaked by
Meanwhile, Kerry won big in the Pacific, Middle Atlantic, and New England,
and prevailed narrowly in the Pacific Northwest.
The second column of Table 1 shows that one-quarter or more of the
electorate in each of the Bush regions cited “moral values” as the issue
that mattered most, outstripping the aggregate national figure. By contrast,
less than one-quarter did in each of the Kerry regions.4 In fact,
“moral values” came in first place in all four of the Bush regions, and in
none of the Kerry regions.
Nationally, more than four out of five moral-values voters cast their
ballots for the president, a pattern with only modest variation by region.
These voters accounted for 37 percent of total Bush support nationwide
(third column in Table 1), but this proportion did vary geographically to a
In the Bush regions, moral values voters made up two-fifths or more of the
president’s backers, while in the Kerry regions, they provided less than
two-fifths—and in three out of four cases, less than one-third—of the Bush
vote. In other words, the president won where his voters cared most about
The final column in Table 1 reports Bush’s moral-values voters as a
percentage of the total vote cast in each region (and not just as a
proportion of the Bush vote). This absolute value of the president’s
moral-values constituency amounted to more than one-fifth of the vote in the
South and Midwest, and more than one-quarter in the Southern Crossroads and
Mountain West. But it was much smaller in the Kerry states, ranging from
one-sixth to less than one-eighth of the total vote cast.
These are not trivial figures: The 21.8 percent of the vote cast by
Midwestern Bush moral-values voters was ten times larger than the
president’s winning margin in the region (about two percent). Just as
importantly, this constituency was also larger than his margin of victory in
his three strongest regions.
By contrast, the Bush moral-values constituency was smaller than Kerry’s
margin of victory in his three strongest regions. Only in the
did this constituency outstrip Kerry’s margin—but not by quite enough to put
Bush over the top. In sum, the moral values vote was critical in every part
of the country except for the
New England, Middle Atlantic, and Pacific
It is important to recognize that the two issues that rivaled moral values
in importance nationally, terrorism and the economy, did not vary nearly as
much geographically. Even though economic conditions and the apparent threat
from terrorists varied a good deal from region to region, “security moms”
were about as likely to vote for Bush in Los Angeles as in Atlanta, and
“factory dads” backed Kerry at about the same rate in Pittsburgh and
The only other issue that varied significantly by region was Iraq, where the
geographic pattern was the opposite of moral values. In the Bush regions, an
average of 12 percent named Iraq as the most important issue, while in the
Kerry regions the average was 22 percent, led by a whopping 31 percent in
In nearly inverse proportion to the Bush moral-values voters, Iraq voters
gave about three-quarters of their ballots to Kerry.
One way to look at the 2004 election is this: Terrorism and the economy
provided each presidential candidate with a strong and roughly equal voter
base in every region, but it was the interplay of moral values and Iraq that
made the difference. In all the regions Bush won, moral values
trumped Iraq by margins of 2-to-1 or more, while in the Kerry regions the
two issues were closely balanced, except for the
where Iraq had a 3-to-2 margin over moral values.
What explains the regional variation in issues priorities? On the
moral-values front, the prime factor was religion. The first four columns in
Table 2 report the percentage of the Bush moral-values voters in the three
largest Christian traditions and in all the other religious groups combined.5
Nationally, evangelical Protestants provided 48 percent of the Bush
moral-values vote, Catholics 19 percent, and mainline Protestants 15 percent
(and the other groups the remaining 18 percent).
Evangelicals were the backbone of the Bush moral-values vote in all regions,
with the Latter-day Saints (found in the “other” column here) making a
significant contribution in the Mountain West and Pacific, and Catholics
doing the same in the Middle Atlantic and New England.
The final column in Table 2 looks at regular worship attenders (once a week
or more) regardless of religious affiliation. Nationally, 71 percent of the
Bush moral-values voters were regular attenders. These figures were
comparable or greater for the Bush regions, and lower for the Kerry regions,
but still substantially above 50 percent.
As another perspective on attendance and issue priorities, 35 percent of
regular worshipers nationally selected moral values, as opposed to only 12
percent who selected
Among only those who said they attend worship more than once a week, the
numbers were 42 percent and 10 percent, respectively. This basic pattern
held across regions. In short, whatever moral questions were raised by the
Iraq war in the campaign—including, perhaps, prisoner abuse—they did not
register strongly with the most religious segment of the American
Table 3 is concerned with the bottom line, so to speak: the number of Bush
moral-values voters provided by these religious categories as a percentage
of the total vote on a region-by-region basis.
In the Bush regions, the contribution from evangelicals averaged 11 percent,
while in the Kerry regions, it averaged six percent. This roughly 2-to-1
ratio applied among mainliners (average of 3.2 percent versus 1.8 percent)
and “others” (average of 6.3 percent to 3.1 percent).
Only the Catholic moral-values voters didn’t fit the pattern. They made the
biggest contribution in the
Midwest (5.1 percent) and virtually none at all in the
Pacific Northwest (0.2 percent). In the South and Southern Crossroads they
supplied about as large a proportion of the Bush moral-values vote as they
did in New England and the Middle Atlantic.
Looking across the board at regular worship attenders of all religious
stripes who selected moral values, these constituted 13.7 percent of the
national vote. But in all the Bush regions together, the average
contribution was 18.2 percent, while in all the Kerry regions it was 8.5
percent. The bottom line is that the Bush moral-values constituency had its
biggest impact in those parts of the country where there are the most
evangelicals and the largest number of regular attenders (many of whom are,
of course, evangelicals).
To understand why evangelicals were more likely—twice as likely as
their share of the population—to select “moral values” on the exit poll, we
need to recognize that the evangelical subculture has long used both words
to characterize its public concerns.
“Moral” (as in Moral Majority) has been a shibboleth ever since the
religious right emerged onto the American scene a quarter century ago. And
in the 2004 campaign, the Southern Baptist Convention—largest of the
evangelical denominations—created a website that stated the message in no
“What are your core values as we approach Election Day 2004? Would your list
include Jobs? The Economy? Health-Care? Education? National Security? As
important as those issues are, think about what your core values should be
as a follower of Jesus.” (These days, such values have to do above all with
opposition to abortion and gay marriage.)
Non-evangelical religious communities are far less tied to moral-values
rhetoric, and thus even the most religiously committed among them were less
likely to choose that expression as their most important issue on the exit
poll. But regional culture matters: These other religious folks were more
likely to choose moral values in the Bush regions than in the Kerry regions.
As Moral Majority founder Jerry Falwell hoped, the coalition of the moral
has expanded beyond evangelicals, but for the most part more in the
evangelical heartland than elsewhere.
Of course, presidential elections are decided not by regions but by states,
and states within the same region often differ a good deal politically.
Nowhere was this more the case in the 2004 election than in the Midwest,
which contained the deep red states of Kansas and Nebraska, the deep blue
state of Illinois, and many shades of purple in between.
The election was decided in magenta
which hard times put within Kerry’s reach but not into his grasp: Bush won
with 51 percent of the vote. Next door was violet Michigan, which stayed
within the Democratic fold despite vigorous Republican entreaties: Kerry won
with 52 percent.
Not only were both states hard hit economically during George W. Bush’s
first administration but in the 2004 election both had ballot initiatives
banning same-sex marriage that passed easily. What explains the difference
in how they voted for president?
As shown in Table 4, 24 percent of Ohio voters named moral values as the
most important issue, in contrast to the 20 percent of Michiganders who
did—a disparity about equal to the difference in the president’s performance
in the two states on November 2. That this disparity was rooted in religion
can be seen in the contribution made by the Bush moral-values voters to the
total vote in each state.
Evangelicals accounted for a larger share of this constituency in Ohio than
in Michigan. So, Buckeye Baptists and Pentecostals helped Bush more than
Wolverine members of Reformed and Nondenominational churches.
Much the same occurred among Catholics: Kerry lost his co-religionists in
Ohio and won them in Michigan. This pattern likely reflects the difference
between traditionally Republican small business German Catholics in the
area and traditionally Democratic unionized Catholics of eastern and
southern European descent living around
Interestingly, mainline Protestants made the same contribution to the Bush
moral-values vote in both states, but by different paths: Kerry lost
Methodists and Episcopalians in
but carried them in
For his part, Bush did a bit better with the other religious groups south of
Lake Erie, reflecting greater support among the Eastern Orthodox and
Overall, regular worship attenders and less regular attenders were both more
likely to be Bush moral values voters in
than in Michigan. Apparently, the moral-values agenda is more deeply
embedded in the Buckeye state.
In a close election such as 2004, many voting blocs can take the credit for having “made the difference.” So,
the extent to which voters based their presidential choice on economic or
national security concerns is not to be ignored. But the issues with
geographic purchase were “moral values” and “Iraq.”
And, at the end of the day, the conservative Christian troops of
psychologist James Dobson controlled more territory than the liberal
anti-warriors of filmmaker Michael Moore.
Geography matters in American politics today above all because the religious
configuration of the country varies considerably from one region to another.
In the final analysis, those first-day stories about moral values—and the
red-and-blue maps that went with them—conveyed something real.
These numbers differ a little from the original exit poll numbers on two
counts. First, the issue priorities here are taken as a percentage of the
respondents who actually gave an answer to the question (92.3 percent of
those polled). Second, the national exit poll data were re-weighted to match
the actual election results by region. Thus, the original figure of 21.7
percent for the moral values voters comes out as 23.7 percent in this
Regions, of course, are not states, and a definitive analysis would consider
each “red” and “blue” state separately. Unfortunately, such an analysis
cannot be conducted for the 2005 election because the number of interviews
per state in the national exit poll is not high enough to give meaningful
results for many states—and not all state exit polls included a “moral
The regions are comprised as follows: New England (ME, NH, VT, MA, CT, RI);
Middle Atlantic (NY, NJ, PA, DE, MD, DC); South (WV, VA, KY, TN, NC, SC, GA,
FL, AL, MS); Midwest (OH, MI, IN, IL, WI, MN, IA, NE, KS, ND, SD); Southern
Crossroads (LA, TX, AR, OK, MO); Mountain West (MT, WY, CO, ID, UT, NM, AR);
Pacific (NV, CA, HI); Pacific Northwest (OR, WA, AK).
Two regions were outliers in this regard: The Mountain West had by far the
largest proportion of moral values voters, largely due to the numerous
Latter-day Saints; New England had more such voters than the other Kerry
regions, due to a large number of evangelicals. These results may reflect
sampling error that appears to have exaggerated the numbers of Latter-day
Saints in the Mountain West and evangelicals in New England. However, it
could be that both groups turned out in especially high numbers in 2004
because of the issue of same-sex marriage. Both groups live in regions with
strong liberal voices on this issue. This was especially the case in New
England where the legalization of civil unions (Vermont) and same-sex
Evangelicals were defined as white Protestants who consider themselves “born
again” and mainline Protestants as white Protestants who do not.