September 1989

The Politics of Evasion:

Democrats and the Presidency

By

William Galston
University of Maryland

Elaine Ciulla Kamarck

Progressive Policy Institute

Progressive Policy Institute

www.ppionline.org

The Politics of Evasion:

Democrats and the Presidency
Introduction
The Democratic Party's 1988 presidential defeat demonstrated that the party's problems would not
disappear, as many had hoped, once Ronald Reagan left the White House. Without a charismatic
president to blame for their ills, Democrats must now come face to face with reality: too many
Americans have come to see the party as inattentive to their economic interests, indifferent if not
hostile to their moral sentiments and ineffective in defense of their national security.
Nor have matters improved for Democrats since the presidential election. On a variety of
measures, from party identification to confidence in dealing with the economy and national security,
the Democratic Party has experienced a dramatic loss of confidence among voters. A recent survey
shows that only 57 percent of Democrats have a favorable image of their own party.1
Democrats have ignored their fundamental problems. Instead of facing reality they have
embraced the politics of evasion. They have focused on fundraising and technology, media and
momentum, personality and tactics. Worse, they have manufactured excuses for their presidential
disasters -- excuses built on faulty data and false assumptions, excuses designed to avoid tough
questions. In place of reality they have offered wishful thinking; in place of analysis, myth.
This systematic denial of reality -- the politics of evasion -- continues unabated today, years after
the collapse of the liberal majority and the New Deal alignment. Its central purpose is the avoidance
of meaningful change. It reflects the convictions of groups who believed that it is somehow immoral
for a political party to pay attention to public opinion. It reflects the interests of those who would
rather be the majority in a minority party than risk being the minority in a majority party.
This paper is an exploration of three pervasive themes in the politics of evasion. The first is the
belief that Democrats have failed because they have strayed from the true and pure faith of their
ancestors -- we call this the myth of Liberal Fundamentalism. The second is the belief that
Democrats need not alter public perceptions of their party but can regain the presidency by getting
current nonparticipants to vote -- we call this the Myth of Mobilization. The third is the belief that
there is nothing fundamentally wrong with the Democratic Party: there is no realignment going on,
and the proof is that Democrats still control the majority of offices below the presidency. We call
this the Myth of the Congressional Bastion.

Progressive Policy Institute

www.ppionline.org

The Myth of Liberal Fundamentalism

The oldest of these myths is that Democrats have lost presidential elections because they have
strayed from traditional liberal orthodoxy. The perpetrators of this myth greet any deviation from
liberal dogma, any attempt at innovation with the refrain "We don't need two Republican Parties."
Liberal fundamentalists argue that the party's presidential problems stem from insufficiently
liberal Democratic candidates who have failed to rally the party's faithful. The facts, however, do
not sustain this allegation. Losing candidates Michael Dukakis and Walter Mondale were very
successful, in fact in most instances more successful, than 1976 winner Jimmy Carter, in winning
over the ideological (and racial) base of the Democratic Party. According to CBS/New York Times
exit polls, Dukakis got 82 percent of the liberal vote and 89 percent of the black vote. This is better
than Carter, who received 74 percent of the liberal vote and 83 percent of the black vote in 1976.
Mondale's loss was so big that he did less well than Carter in most groups, but he still received 71
percent of the liberal vote and fully 91 percent of the black vote.
The real problem is not insufficient liberalism on the part of the Democratic nominees; it is
rather the fact that during the last two decades, most Democratic nominees have come to be seen as
unacceptably liberal. Fully 36 percent of the electorate told ABC exit pollers last November that
Dukakis' views were "more liberal" than their own. In contrast, just 22 percent thought George
Bush's views were more conservative than their own. In 1976, CBS/NYT exit polls showed that
Carter was able to win the support of 30 percent of the self-identified conservatives and 48 percent
of the independent voters. Dukakis won over only 19 percent of self-identified conservatives and 43
percent of independents and Mondale won only 18 percent of conservatives and 36 percent of
independents.
Because there have consistently been many more conservative identifiers than liberal identifiers
in the electorate, the perception that recent Democratic nominees are "too liberal" has worked to the
advantage of the Republicans.2 The drop in conservative support accounts for more than half of the
five-point decline in overall support form Carter's 50 percent to Dukakis' 45 percent.
To understand these trends, it is vital to recall that liberalism has played an honorable and
productive role in twentieth-century American politics. For three decades, liberalism provided the
principles and programs for a governing party that led our country to unprecedented achievements at
home and abroad. In this period, liberalism was innovative and inclusive. And whatever the issue of
the moment might be, it consistently based its policies on the sentiments and interests of the
overwhelming majority of the American people.
But in the past two decades, liberalism has been transformed. The politics of innovation has
been replaced by programmatic rigidity; the politics of inclusion has been superseded by ideological
litmus tests. Worst of all, while insisting that they represent the popular will, contemporary liberals
have lost touch with the American people. It is this transformed liberalism that we call "liberal
fundamentalism," on which the electorate has rendered a series of negative judgments.
Since the late 1960s, the public has come to associate liberalism with tax and spending policies
that contradict the interests of average families; with welfare policies that foster dependence rather
3

Progressive Policy Institute

www.ppionline.org

than self-reliance; with softness toward the perpetrators of crime and indifference toward its
victims; with ambivalence toward the assertion of American values and interests abroad; and with
an adversarial stance toward mainstream moral and cultural values.
This is not to say that today's voters are opposed to all "liberal programs." Public opinion polls
consistently show wide support for increased government activity in such areas as health and child
care. But these programs must be shaped and defended within an inhospitable ideological climate,
and they cannot by themselves remedy the electorate's broader antipathy to contemporary liberalism.
This shift in public attitudes has been mirrored in the conduct of political campaigns. Ronald
Reagan's 1966 gubernatorial campaign against Pat Brown was the harbinger of things to come, and
Richard Nixon's 1968 presidential campaign established the anti-liberal paradigm that has
dominated American national politics ever since.
The campaign of 1988 was waged squarely within this framework, and it dramatically confirmed
continuing public antipathy to liberal fundamentalism. According to ABC exit polls, nearly onequarter of the voters felt defense and foreign affairs were important in making their choice; Bush
won them 88-12. Bush held a 4-1 margin among voters who stressed the Pledge of Allegiance. On
taxes and crime, Bush won 72-27 and 73-27 respectively. Of the 27 percent who named the death
penalty as important; 75 percent backed Bush as opposed to only 24 percent for Dukakis.
Liberal fundamentalism refuses to adjust to changing circumstances by adopting new means to
achieve traditional ends. Instead, it enshrines the policies of the past two decades as sacrosanct and
greets proposals for change with moral outrage. Whether the issue is the working poor, racial
justice, educational excellence, or national defense, the liberal fundamentalist prescription is always
the same; pursue the politics of the past. The result, predictably, has been programmatic stagnation
and political defeat.
During its heyday, the liberal governing coalition brought together white working-class voters
and minorities with a smattering of professionals and reformers. Over the past two decades,
however, liberal fundamentalism has meant a coalition increasingly dominated by minority groups
and while elites -- a coalition viewed by the middle class as unsympathetic to its interests and its
values. The inescapable fact is that the national Democratic Party is losing touch with the middle
class, without whose solid support it cannot hope to rebuild a presidential majority. Jimmy Carter
forged his 1976 victory with the help of a majority of middle-income voters, while Michael Dukakis
was able to win only 43 percent of this vital group.
In fairness to those Democrats who run and win in spite of their presidential party, liberal
fundamentalism is not pervasive at the states and local levels. But it tends to dominate two
important, defining arenas for the Democrats; the institutional party and the presidential nominating
process. In every presidential election this decade, the losing Democratic nominee has been charged
by his intraparty foes with insufficient liberalism. Senator Ted Kennedy's 1980 campaign to unseat a
Democratic president crystallized liberal fundamentalism as the party's reigning dogma, enforced
through ideological litmus tests. His insurgency rested in large measure on the proposition that
Carter was unfit to continue as president because he was not liberal enough and that he would make
a poor candidate because his conservatism would not allow him to rally the faithful.
4

Progressive Policy Institute

www.ppionline.org

Jesse Jackson's 1984 and 1988 campaigns -- which featured vigorous critiques of Mondale and
Dukakis -- were the purest version of liberal fundamentalism. These attacks persisted in spite of the
fact that the nominees were unwilling or unable to separate themselves adequately from liberal
fundamentalism. No doubt the constant pressure they experienced contributed to this failure. But
also (as we shall see), the politics of evasion has meant that Democratic nominees have been unable
to break clearly with liberal fundamentalism because they and their advisers continue to embrace
myths about the electorate that cannot with stand either empirical analysis or political combat.

Liberal Fundamentalism and the Nominating Process

Liberal fundamentalism is reinforced by the dynamics of the nomination process. Primary and
caucus goers are a small portion of the electorate in both parties, but they tend to be the true
believers. This is a big problem for the Democrats who, as their national fortunes have faltered,
have attracted fewer and few participants in their nominating contests, leaving the liberal wing of
the party even more in control in some critical states. Where comparable data are available, 1988
ABC exit polls show a pattern: in many states, liberals have increased as a percentage of total
primary participants while the conservative share has diminished.
The prime engine for this shift has been a decline in overall Democratic primary participation. In
Florida for example, 20.2 percent of the voting-age population participated in the 1976 Democratic
primary won by Jimmy Carter, who also carried Florida in the general election. By 1988, only 13.2
percent of the voting-age population participated in a primary won by Dukakis who even at the
height of his popularity with voters nationally never had a realistic chance of beating Bush in
Florida.
Similar declines in Democratic presidential primary turnouts have been at work throughout the
South and elsewhere. Georgia had an 11 percent drop in turnout between 1984 and 1988; during that
same period Alabama had a 13 percent drop, North Carolina had a 29 percent drop and Illinois had a
10 percent drop.3
Who is abandoning the Democratic primary process? One explanation is suggested in "The
Mystery of the Vanishing Democrats.4 (See chart next page.)
Based on exit polls from recent primaries, the data in this chart show that lower middle-class voters
participated in the Democratic primary process in far smaller numbers in 1988 than they did in
1984. Given that the candidacy of Jesse Jackson assured continuing high rates of black primary
participation, the most plausible explanation for the large decline in lower-income voter
participation between 1984 and 1988 is that working-class whites were deserting the Democratic
nomination process in droves. During this same period, the percentage of Democratic primary
participants with incomes of $50,000 or more doubled in a number of key states.

Progressive Policy Institute

www.ppionline.org

The Mystery of the Vanishing Democrats

Exit polls suggest that in key Democratic presidential primaries this year, lower-income
Americans -- considered a critical voting bloc for the Democrats have been vanishing. An
examination of exit polls conducted by CBS The New York Times in Alabama, Georgia, Illinois,
Indiana, Maryland, Massachusetts, New York, North Carolina, Ohio and Pennsylvania reveal a
consistent trend: Democrats with Annual family incomes of less that $25,000 make up a smaller
share of the 1988 party primary electorate than was the case in 1984.
Percent of voters in Democratic primaries
Family
Below
Income*
$12.5k
Alabama
1984
34%
1988
22
Georgia
1984
23%
1988
13
Illinois
1984
26%
1988
15
Indiana
1984
25%
1988
18
Maryland
1984
14%
1988
9
Massachusetts
1984
12%
1988
10
New York
1984
19%
1988
10
N. Carolina
1984
24%
1988
13
Ohio
1984
23%
1988
14
Pennsylvania
1984
25%
1988
17

$12.5k-$25k

$25k-35k

$35k-$50k

$50k+

Turnout change

Primary type

24%
21

18%
17%

10%
17

5%
12

-22,641

Open

26%
22

19%
21

15%
20

9%
18

-61,789

Open

27%
20

21%
17

12%
22

5%
15

-158,497

Open

32%
25

19%
20

9%
22

3%
8

-71,247

Open

23%
18

21%
21

22%
24

13%
24

+3,084

Closed

26%
15

23%
16

21%
26

9%
24

+81,893

Democrats,
independents

27%
19

19%
17

16%
17

12%
28

+187,236

Closed

27%
25

19%
24

13%
18

9%
13

-280,899

Closed

30%
26

20%
20

12%
18

5%
12

-70,384

Open

30%
26

20%
19

10%
14

3%
10

-140,947

Closed

* income in year before election; because some voters declined to state income, percentages do not add up to 100.
Copyright 1988, National Journal. Reprinted by permission.

There is a vicious circle at work here. As the increasing role of upscale liberals in the
nominating process reinforces the party's emphasis on the kinds of issues that tend to antagonize
working-class voters, these voters fall out of the party's nominating process, making it even more
6

Progressive Policy Institute

www.ppionline.org

likely that Democrats will nominate a candidate without significant appeal to the demographic and
political center. The failure of "Super Tuesday," an attempt by Southern elected officials to move
Democratic presidential candidates in a more moderate direction by forcing them to spend time in
the South, lay in their failure to anticipate that the Southern primary electorate would be as small
and unrepresentative of the general election electorate as it ultimately was.

The Myth of Mobilization

The second pervasive theme in the politics of evasion is the Myth of Mobilization. The argument
goes as follows: the Democratic party need not alter its program or message, because it can regain
the presidency by getting current nonparticipants to vote.
The most general form of this argument is that higher turnout across the board is the solution.
The facts do not support this contention. According to a CBS/NYT poll of nonvoters taken shortly
after the 1988 election, if everyone had voted Bush would still have won by a larger margin.5
There are three reasons why general mobilization will not do the job. First, the large lead in
party identification that Democrats have enjoyed since the New Deal has nearly disappeared. In
1976, the last time the Democrats won the presidency, they enjoyed a 15-point advantage over
Republicans among those who identified with a political party. By 1988, 37 percent of the voters
identified themselves as Democrats and 35 percent identified themselves as Republicans, leaving
the democrats with a mere two-point advantage.6 Had party identification been the same in 1988 as
it was in 1976, the percentages of Democrats, Republicans, and Independents that Dukakis actually
won would have been enough to give him the presidency.7
The second reason why general mobilization will not work for Democrats is rooted in the
changing nature of "peripheral" voters -- persons whose attachment to the political process is relatively weak and who tend to vote in only high-intensity elections. During the heyday of the New
Deal coalition, these voters tended to be Democrats, and it was a truism of party politics that higher
turnout tended to be correlated with Democratic successes.
During the past decade, however, evidence began to mount that peripheral voters were not
necessarily Democrats, a thesis urged by James De Nardo.8 While political scientists were debating
his statistics and methodology, political developments were lending support to his key conclusion.9
The elections of 1986 had the lowest midterm turnout since 1942, with only 33.4 percent of the
voting-age population participating.10 Nevertheless, Democrats prevailed in nearly all the closely
contested races and regained control of the Senate, provoking Republican activists to fault their
tacticians for not emphasizing the kinds of high-intensity issues that had turned out voters in 1980
and 1984. Two years later, with a more strongly ideological contest and the higher turnout
characteristic of presidential contests, George Bush achieved a victory almost as stunning as
Reagan's 1984 triumph.
Forward-looking Republicans have drawn the obvious conclusion that increased registration and
turnout are now at least as much in their party's interests as in the Democrats'. Therefore, in a switch
from the traditional Republican position, Congressman Newt Gingrich, the Republican Whip,
supported a proposal that would make it much easier for Americans to register to vote. His position
7

Progressive Policy Institute

www.ppionline.org

was based in part on data indicating that one of the largest groups of potential new voters -- young
people -- has turned increasingly Republican.11
The third reason why general mobilization by itself cannot get the job done lies in the changing
voting patterns in the heart of the electorate. According to CBS/NYT exit polls, the heart of the
electorate (forty percent of the total) is made up of middle-class voters with family incomes between
$20,000 and $50,000 per year. In 1976 Jimmy Carter carried this group with 51 percent of the vote - by no coincidence his overall national margin. In 1988 Michael Dukakis received only 43 percent
of their vote.
The fact is that the Republican Party is no longer the party of the rich. It is also -- and
increasingly -- the party of those who work for a living. Bush beat Dukakis 55-44 among lower
middle-class white voters with annual family incomes between $10,000 and $20,000. According to
CBS/NYT exit polls, if only voters with family incomes of under $50,000 per year had participated
in the 1988 election, George Bush still would have won.
The Democrats' "disappearing middle" can be documented along dimensions other than income.
For example, ABC exit polls indicate that the Democratic presidential vote now comes from the
most and the least educated strata of the electorate, while Republicans claim everything in between.
Voters with less than a high school education (7 percent of the total) went for Dukakis 60-40, while
voters with postgraduate education (15 percent of the electorate) supported Dukakis 50-49.
Everyone else (78 percent of the electorate) went for Bush 56-44.
Religion is another key dimension in the decline of the Democratic middle. The collapse of
support for national Democrats among white evangelical Protestants is well known. Equally critical,
however is the erosion of the former bedrock of the New Deal coalition outside the South: ethnic
Catholics of the Northeast and Midwest. Jimmy Carter (notoriously Protestant and evangelical)
received 55 percent of the Catholic vote in 1976. But Michael Dukakis, son of immigrants and
Greek Orthodox, received only 47 percent of the Catholic vote according to a CBS/NYT exit polls.
Since 1976 the erosion of Democratic support among Catholics has been more pronounced than
among white Protestants -- by some measures twice as large. According to an ABC analysis, it was
Dukakis' failure to do better among ethnic Catholics that doomed his chances in key Midwestern
and Northeastern states.12

Selective Mobilization
There is, in short, little evidence to support the proposition that a general mobilization of the
electorate would remedy the Democrat's presidential woes. But a very popular variant of the politics
of evasion argues that selective mobilization of groups that strongly support Democratic candidates,
especially minorities and the poor, would get the job done for Democratic presidential candidates.
In approaching this question it is important to distinguish quantitative and empirical questions
from moral and historical issues. The exclusion of minorities and the disadvantaged from full voting
rights was a grave injustice, the rectification of which is a continuing moral imperative of our times.
But we are here concerned simply with the empirical question -- the effect of fuller participation on
8

Progressive Policy Institute

www.ppionline.org

presidential outcomes.
In an important article Ruy Teixeira demonstrated that even if black and Hispanic turnout had
exceeded white turnout by 10 percentage points, Dukakis still would have lost the election by 2.5
million votes. If turnout among adults in families making less than $12,500 a year had exceeded
turnout among wealthy Americans by 10 percent, Dukakis would have lost the election by more
than 3.3 million votes. And even if these race- and class-based voting upsurges had occurred
simultaneously, they would not have been enough to close the gap with Bush. Teixeira's overall
conclusion is irresistible: Democrats lost the presidential election "because they didn't have enough
support in the nation as a whole, not because enough of their people failed to show up at the polls.13

Table #1
How Increases in Black Voter Turnout Would Affect
The 1988 Presidential Election, By State
(see Appendix A for actual numbers)

States moving from Bush to

Dukakis if blacks had voted at
a turnout rate of 52% of VAP.

Illinois
Maryland

States moving from Bush to

Dukakis if blacks had voted at
a turnout rate of 62% of VAP.

Illinois
Maryland

States moving from Bush to

Dukakis if blacks had voted at
a turnout rate of 68% of VAP.

Illinois
Louisiana
Maryland

The selective mobilization thesis, which fails so dismally at the national aggregate level fares no
better at the state level, Table #1 shows the states that would shift into the Dukakis column under
three scenarios: black turnout at national white turnout levels (52 percent), black turnout 10 percent
above white levels (62 percent), and black turnout at 68 percent.14 It should be noted that the last
two scenarios involve the highly improbable assumption that large increases in black turnout can be
achieved without correspondingly large increases in white turnout. Nevertheless, even under these
scenarios, the increase in black votes is not enough to put many new states in the Democratic
column.
The mobilization argument is made for Hispanics as well as for blacks. Hispanic turnout is a
small percentage of voting-age population (15 percent in California and 21 percent in Texas,
according to ABC exit polls) because many Hispanics are not yet United States citizens and are
9

Progressive Policy Institute

www.ppionline.org

therefore unable to vote. In California, increasing Hispanic turnout from 15 percent to 52 percent
would have put that state in the Dukakis column. But in Texas, even that large an increase in turnout
would not have given the state to Dukakis.15
As turnout among Hispanics catches up with the turnout in the rest of the population Democrats
should be advantaged even though they no longer win overwhelming margins among Hispanics, as
they do among blacks. But because so many Hispanics are not yet eligible to vote,
Democratic gains are unlikely to be realized in the near future.
Much of the support for the mobilization thesis originated with a study from The Joint Center
for Political Studies which was then widely quoted by Jesse Jackson. This study gives Dukakis 146
more electoral votes by assuming a turnout rate among blacks of 68 percent -- 18 percent higher
than the population as a whole, by assuming that 100 percent of these newly mobilized voters would
cast their ballots for Dukakis, and by assuming that these events have no impact on white turnout.16
But in practice, it doesn't work that way. In state after state, especially in the South, mobilization
among black voters has been at least matched by mobilization among white voters.17 This counter
mobilization was particularly evident in the 1984 Senate race between Jim Hunt and Jesse Helms
where black registration efforts were eclipsed by a counter mobilization of conservative whites. The
problem with selective mobilization is that in politics, as in business, it's not the gross that counts,
it's the net.
The myth of selective mobilization gained its greatest currency after the midterm elections of
1986 when (it was alleged) massive minority turnout was responsible for the return of the Senate to
Democratic control. Again, this thesis does not survive empirical examination. In the three states for
which exit polls are available, the evidence indicates that Walter Mondale did as well or better
among black voters in his 1984 loss than did the winning Senate candidates two years later. The
difference in each case was that the Senate candidate ran significantly better than Mondale among
white voters.18
These data suggest that successful Senate candidates prevailed by holding onto strong black
support and bringing a substantial percentage of white voters back into the fold -- precisely the kind
of biracial coalition that wins for Democrats outside the South as well. In 1984, for example,
Michigan's Carl Levin and Illinois' Paul Simon ran about even with Mondale among the black
voters of their states, but they retained their seats in spite of the Reagan tidal wave because they
were able to run far ahead of Mondale among white voters.
The other argument that is made to buttress the selective mobilization thesis is that women as a
Democratic voting bloc -- the famous gender gap can bring Democratic candidates to victory. But
contrary to conventional belief, the gender gap has not worked in favor of Democratic presidential
nominees.
Dukakis was supported by a slightly lower percentage of women than was Carter in 1976.19 By
contrast, Dukakis' support among men was fully nine points lower than Carter's (42 percent versus
51 percent) according to CBS/NYT exit polls. The gender gap that has opened up in the past twelve
years is not the product of a surge of Democratic support among women, but rather the erosion of
Democratic support among men.20
10

Progressive Policy Institute

www.ppionline.org

The overall implication of these data is clear. Yes, intensified mobilization among groups that
have stood loyally with the Democratic party is politically and morally essential. But the gains from
such an effort cannot by themselves compensate for the broad erosion of support the party has
experienced in other sectors of the electorate. There is no alternative: if the Democratic Party wants
to rebuild a presidential majority, it must regain competitiveness among voters it has lost.

The California Dream

One final element of the myth of mobilization is what we call "The California Dream." The thesis is
that rising strength in the West can counterbalance the collapse of Southern support for the party's
presidential candidates and that Democrats therefore don't have to work hard at regaining
competitiveness in the South.
This exercise in the politics of evasion fails the test of basic arithmetic. Non-Southern gains
cannot fully compensate for a Southern wipeout. If Dukakis had prevailed in all the Western states
where he had a chance, carried the heartland states he narrowly lost, and won all the Eastern states
within reach, he still would not have assembled enough electoral votes to win.
The underlying logic of the electoral college shows why. There are 155 electoral votes in the
Southern and border states, 41 in the Plains and Rocky Mountain states with impregnable
Republican majorities, and 23 more in reliably Republican states of the Midwest and Northeast. If
the South is conceded to the Republican presidential nominee, he begins with a base of 219 electoral
votes and needs only 51 more. Michigan, Ohio and New Jersey are enough to put him over the top -and George Bush carried them handily, with margins of 8 to 14 points.
The electoral college arithmetic only gets worse in 1992. According to projections from
preliminary Census estimates, reapportionment will net the states in the Republican base 12
additional electoral votes for a total of 231. New Jersey and Ohio would be just about enough to
give Bush a victory even if he loses California and a host of other states he carried last time. If
Democrats are only competitive in states with 310 electoral votes, the odds against their nominee
attaining 270 are dauntingly high. The Republican nominee will start with two pairs while his
Democratic opponent would have to draw to an inside straight.21
The conclusion is unavoidable: just as Democrats must regain competitiveness with large
segments of the electorate that they have lost, they must also regain competitiveness in every region
of the country. The biggest surprise of 1988 was not that Dukakis was trounced in Dixie, but that he
failed to prevail in the heartland states such as Illinois, Pennsylvania and Michigan where the costs
of Reaganomics have been high and where class and ethnic identification should have worked in his
favor. The Democratic Party has more than a Southern problem, and it needs a truly national
remedy.

The Myth of the Congressional Bastion

The final element in the politics of evasion is what we call the Myth of the Congressional Bastion. It
goes like this: there's nothing fundamentally wrong with the Democratic Party; there's no
11

Progressive Policy Institute

www.ppionline.org

realignment going on; the proof is that Democrats still control Congress and a majority of state and
local offices as well.
This line of reasoning stems from the 1932 experience in which changes occurred simultaneously in the presidency, both houses of Congress, and hundreds of state and local offices. Ever
since, we have discounted the existence of realignment unless it is as dramatic and comprehensive
as in 1932. Political scientists have invented phrases such as "split-level realignment" in an effort to
characterize alleged voter preference for divided government.
In our judgement, the notion of split-level realignment as an enduring feature of the American
political landscape is blind to the underlying dynamics of contemporary politics. It also defies
common sense. We are witnessing instead a slow-motion, trickle-down realignment in which, over
time, Republican presidential strength is inexorably eroding Democratic congressional, state, and
local strength.
A key leading indicator of voting behavior is party identification. Here a strong Republican tide
is running. The 15-point democratic edge in national party identification at the beginning of this
decade shrank to only two points by the 1988 elections. This tide can be seen most clearly in the
South, which is now a bastion of Republican presidential strength. In the 1988 election Bush
retained more of Reagan's 1984 strength in the South than in any other region. According to ABC
exit polls, nearly as many Southern voters called themselves Republicans (40 percent) as Democrats
(41 percent), an improvement over four years ago when Democrats still enjoyed a five point edge
and a big improvement over 1980 when Democrats held a 25-point lead. The trend is most
pronounced in Florida, where Republican voters now outnumber Democrats 46 to 41 percent.
The Republican surge is not confined to the South. ABC exit polls show that in Ohio a 15-point
Democratic edge in 1980 was cut to only two points by 1988. During this same period, an 11-point
Democratic advantage in Michigan was transformed into a one point Republican edge, and a 14point Democratic advantage in Illinois turned into a similar one point Republican edge. In New
Jersey, a surge in Republican Party identification between 1984 and 1988 turned a four point
Democratic lead into a three point Republican advantage.
The trend is likely to continue. For example, a county-by-county survey by the Institute for
Southern Studies showed that higher turnout was positively correlated with higher Bush majorities
and that the Bush counties are growing twice as fast as the counties carried by Dukakis.22
While realignment at the presidential level has been dramatic, it has been slower at other levels.
The South is the strongest region for Republican presidential candidates but it is also the basis for
Democratic congressional power. In the 12 Southern states, Democrats still hold 16 of 24 U.S.
Senate seats, 78 out of 120 House seats and most state and local offices. But one aspect of American
politics since the 1932 realignment that is relatively new and very powerful is the ability of
members of congress and other incumbents to protect themselves from national ideological trends.
Incumbency thus guarantees that the dramatic realignment at the presidential level will be slow to
appear at other levels.23 In 1988, despite the Republican presidential sweep, a staggering 98 percent
of incumbents gained re-election to the House, emphasizing the fact that in modern politics,
incumbency is a far more powerful force than party.
12

Progressive Policy Institute

www.ppionline.org

Incumbency is in fact the chief obstacle to realignment in the South and in other parts of the
country. Earl and Merle Black, leading experts on Southern politics, point out that "Democratic
incumbency constrained Republican senatorial gains between 1966 and 1984. The
irony of nonpresidential southern Republicanism is that the setting most conducive to Republican
gains has appeared more often in the office that is comparatively isolated from national political
influences [Governorships], while the office that is more susceptible to pro-Republican national
influence has frequently been immunized through Democratic incumbency."24
But there is evidence, both empirical and anecdotal, to the effect that once incumbency is taken
into account, slow, trickle-down realignment is taking place. Nevertheless, this trend makes barely a
ripple in the overall congressional alignment because of the relatively few seats that change hands
even when aggregated over a decade.25 The role of incumbency as a powerful bulwark against
realignment is a fact of which the Republican Party (under the leadership of Congressman Newt
Gingrich, Republican Party Chairman Lee Atwater and Republican Campaign Committee Director
Ed Rollins) is keenly aware. Republicans have embarked on a determined effort to nationalize
House races along the same kinds of ideological lines that have proven effective for their party in
other arenas. They have embraced campaign finance reform and PAC reform because they realized
that current laws help Democratic incumbents much more than they help Republican challengers.26
Finally, recent events in Congress will place Congress under pressure to enact ethics laws trimming
back electoral advantages enjoyed by current office-holders.
The Democratic grip on the Senate, where individual results are linked more closely to national
trends and where the power of incumbency is weakened somewhat by the ability of challengers to
attract free media, is even less secure. The 1986 mid-term election swept away most of the Reagan
Senate class of 1980. But with weakness once again at the top of the Democratic ticket, 1992 could
easily be a rerun of 1980. In 1992 Democrats will defend 20 of the 34 seats up, 11 of which are held
by senators elected for the first time in 1986. Ten of the new senators won with 55 percent or less of
the popular vote, eight with under 52 percent and four squeaked by with only 50 percent. Of the 11
freshman, 10 are from the South and West where the damage done by the presidential party to the
congressional party tends to be most severe.27
The effects of trickle-down realignment are evident in two other ways: attitudes of young people
towards the Republican Party, and the related erosion of Democratic strength among youth.
According to ABC exit polls, Reagan won the 18 to 24 year-old age group by only one percentage
point in 1980; four years later he won that same age group by 19 points. While less attractive to the
youngest voters than Reagan, Bush still prevailed among them by five percentage points.
By some measures, the tendency for the young to identify with the Republican Party is actually
growing. In a 1982 survey Gallup asked 13 to 17 year olds which political party they were more
likely to vote for; 45 percent said Democrats versus 33 percent who named the Republicans. Five
years later, the same survey found the numbers practically reversed; among 13 to 17 year olds 33
percent were likely to vote Democratic and 48 percent Republican.28 Finally, a recent New York
Times poll showed that among 18 to 29 year olds, Republicans led 52 to 38 percent, a change of 10
percentage points in less than a decade, a finding echoed by a recent Gallup poll which shows the
GOP with a significant edge among those under 30 years of age on all measures of party strength.29
13

Progressive Policy Institute

www.ppionline.org

The data on age and party is best summed up in a quote from the Republican pollster Bob
Teeter, who opens up discussions of this issue with the following: "The bad news is that there are
still more Democrats in the electorate, the good news is that they're dying off." Indeed, if
Democratic voting strength is concentrated increasingly in the older age cohorts and Republican
strength among young people continues to rise, a realignment cannot be postponed indefinitely.
The other place where evidence of realignment is prevalent is among those young people who
are involved in politics and from whose ranks future talented candidates are likely to arise. In the
post-1988 election period, Republicans have been the beneficiaries of an epidemic of party
switching by state and local officials.30 Few of these switches have occurred at the congressional
level (after all, there are real and powerful advantages to being in the majority party in Congress).
More worrisome to Democrats in the long run, however, is that these switches have occurred at the
state legislative and county levels among those elected officials who are most likely to become
congressional candidates.31
Among an even younger group of elites -- politically active college students -- it is clear that
there has been a resurgence of political activism among Republicans while Democrats have been all
but moribund. A comment by a Louisiana State University senior spells real trouble for Democrats
if it is, as some think widely held. "Democrats," according to Rod Dreher, "seem to be too bound to
the solutions of the past. All the creative thinking -- for better and for worse -- is coming from the
32
right."
Eventually, the massive political realignment at the top of the ticket will affect races at the
bottom of the ticket. Southern politicians know this better than anyone and they are worried. In their
lifetimes they have seen a Republican Party that could not even fill their slots on the ballot turn into
a party where presidential successes have contributed to renewed enthusiasm and competition at the
grass roots level.33 But this will affect Democrats in other places as well. A resurgence of
Republican strength in two bastions of Democratic liberalism -- New York City and Massachusetts - could very well scare Democrats out of their complacency. Whatever else happens, unless
Democrats can regain credibility with entry-level voters, the passage of time and the movement of
young people who now lean Republican into the electorate will assure the completion of this trickledown realignment.

Consequences of the Politics of Evasion

The set of myths which constitute the core of the politics of evasion lulls Democrats into a false
complacency. It prevents them from engaging in the kind of comprehensive, thematic and policy
review that could revitalize their party.
Statistical evidence confirms what political experience suggests: that presidential voters are
moved by three broad "baskets" of concerns -- the economy, defense and foreign policy, and social
issues. Each of these baskets, moreover, is framed by basic values that provide context and meaning
for specific policy issues.
The politics of evasion allows the Democratic Party establishment to sidestep these essential
14

Progressive Policy Institute

www.ppionline.org

facts. Instead of facing up to the need for fundamental re-examination, these Democrats explain
their failures in two ways that serve to stifle debate and avoid change. The first is a tendency
towards economic reductionism. It can be summed up in the proposition "It's all economics," and it
radically downplays the disastrous impact of the party's stance on national defense and social issues.
The second is a tendency towards racial reductionism ("It's all race"). It works to thwart reevaluation of Democratic positions on serious issues that have racial dimensions.
The first excuse goes as follows: Democrats encountered economic bad luck in the 1970s, and a
Republican Party willing to purchase unsustainable prosperity with hot checks in the 1980s. But this
cannot last indefinitely. Some combination of hard times and a refurbished economic message that
takes on the Republican tilt towards the wealthy would be enough to restore Democratic presidential
dominance. Other issues (social policy, crime, national security) that bulk so large in campaign
rhetoric are in fact negligible in their effect.
To begin with, this thesis allows Democrats to avoid confronting the fact that they have lost the
economic base that they once enjoyed among people who work for a living. (See the Myth of
Mobilization). Voters have lost confidence in the Democrats' ability to manage the economy, a
traditional strength dating back to the days of the New Deal. According to the July 26, 1989 Gallup
poll, Republicans are favored 51 percent to 30 percent on the question of which political party will
do a better job of keeping the country prosperous.
In addition, this thesis overlooks the profound impact of noneconomic issues on presidential
elections. To be sure, there is a substantial body of data supporting the common-sense view that
economic conditions, coupled with public perception of the party's economic competence, are very
important in determining the outcome of presidential races.34 Still, the manner in which the
economy affects presidential races varies with specific economic circumstances. Postwar history
suggests that if the economy has been growing vigorously (3 percent or more: 1964 and 1984), the
incumbent party has a powerful advantage; that if the economy has been sluggish or recessionary (2
percent growth or less: 1960 and 1980), that challenger has the edge; and that if the economy has
just muddled along in the 2 to 3 percent range (as frequently happens), the impact of economic
factors will be roughly neutral and the election will be decided by noneconomic considerations.35
The 1988 election fell into this last category. A recent review of presidential election forecasting
models concluded that, considered in isolation, 1988 economic conditions would have been
translated into a vote of just 50.4 percent for Bush.36
From this standpoint (and from others as well), the Bush campaign acted prudently in promoting
crime, national strength, and patriotic values to at least a coequal status with economic issues. Even
though confidence in the ability of Democrats did better among those voters who felt that economic
issues were most important than among voters who gave priority to other kinds of issues. According
to ABC exit polls, Bush beat Dukakis by only 5 percentage points among the 41 percent of the
electorate that felt economic issues were most important. But Bush won 88 percent of those who felt
that defense and foreign policy issues were among the most important issues in the race, while
Dukakis won only 12 percent of that group.37
Among voters who stressed values questions such as the Pledge of Allegiance, Bush's margins
were massive.38 And on one of the central social issues of our time drugs and violent crime -- the
15

Progressive Policy Institute

www.ppionline.org

situation was no brighter for Democrats. Nearly one-third of the voters in the ABC exit poll listed
the death penalty as a very important issue in choosing their candidate; Bush won 75 percent of
these voters to 24 percent for Dukakis. The July 26, 1989 Gallup poll found that among those who
felt that drugs are the most important problem, Republicans had a 40 percent to 28 percent
advantage as the party better able to deal with the issue.
Democratic Party vulnerability on social issues goes far beyond crime. In a CBS/NYT poll last
year, 73 percent thought that the United States had experienced a severe breakdown in moral
standards over the past 20 years; only 22 percent disagreed.39 But for many Americans whose
support is essential, Democrats are part of the problem, not the solution. In their eyes, Democrats
have become the party of individual rights but not individual responsibility; the party of selfexpression but not moral accountability.
"It's all economics" is thus a very powerful tactic in the politics of evasion. It allows Democrats
to avoid dealing with problems of vulnerability on national defense and social issues -- especially
crime -- issues that assume a greater importance when the economic picture is neither dramatically
bad nor good.
The second excuse used to avoid confronting the need for a comprehensive review of the
policies of the Democratic Party is "It's all race." According to this thesis, the major themes of the
past two decades, which Republicans have exploited so effectively, are all products of -- and codes
for -- racial divisions. Whatever the ostensible issue crime, public safety, the death penalty, jobs -- the real issues is race. Because the Democratic Party has embraced the right but unpopular position
on racial justice, it has paid a heavy price among voters who do not share this view.
Nothing can be done about this, continues the argument. "Repositioning" is out of the question
because it would come at the expense of the party's moral integrity. Democrats' duty then, is to stand
fast, bear witness, take their lumps -- and hope that the American people will eventually agree with
them.40
No one should doubt the continuing power of racial conflict in American politics. In the South,
erosion of the Democratic Party began with the Dixiecrat revolt against Hubert Humphrey's 1948
convention speech and accelerated with Lyndon Johnson's focus on civil rights. Controversy over
busing, affirmative action, and the general principle of race-based entitlements further exacerbated
white flight from the party during the past two decades.41
But it is one thing to say that race matters, and quite another to say that it dominates everything.
A white Willie Horton may not have struck the same degree of terror as did a black Willie Horton -but violent rapists are frightening regardless of color. By concentrating on race alone, Democrats
avoid confronting the fact that for years they have been perceived as the party that is weaker on
crime and more concerned about criminals than about victims.
The institutional tendency of the Democratic Party to be out of sync with mainstream values exists
on other issues as well. For example, according to the ABC exit poll, Bush won about as much
support for his stand on the Pledge of Allegiance as for his emphasis on prison furloughs. Whatever
the racial content of the Willie Horton issue, the Pledge was surely not a racial issue. It was a values
16

Progressive Policy Institute

www.ppionline.org

issue. It played on voters' doubts about the Democratic Party's patriotism -- doubts tracing back to
the 1968 and 1972 conventions.
The emphasis on racial reductionism masks an equally serious problem, the post 1968 intraparty
conflict between lower middle-class voters and the white liberal elites who increasingly dominate
national party and presidential politics. This clash, beginning with the influx of upscale anti-war
activists into the "reformed" 1972 convention and continuing to this day, has been in many ways as
pervasive and significant as the clash between whites and blacks.42 The shrinking influence of lower
middle-class Democrats and the concomitant rise of higher socioeconomic status Democrats who
hold liberal views on social issues is a continuing source of unresolved conflicts in the party.43
If the white working-class felt morally and culturally isolated from those who took over the
party in the early seventies; they were to feel economically isolated when white elites turned to nogrowth policies in the mid-seventies just as the economy was beginning to grind to a halt. This new
economic isolation was reflected in the popular culture by items such as bumper stickers that read
"If you're out of work and hungry, eat an environmentalist." It was reflected in the tax rebellion -- a
result of the non-indexed income tax during a period of soaring inflation that pushed average
families into higher marginal brackets while their real incomes stagnated. It was also reflected in the
inability of organized labor to deliver substantial proportions of their membership for the
Democratic ticket and in the rise of "Reagan Democrats," working-class voters who abandoned the
Democratic Party for the cultural and moral affinity provided by Ronald Reagan.44
The overall effect of racial reductionism is to chill honest discussion of key issues within the
Democratic Party -- that is, to thwart sober reflection on the relation between means and ends. The
Democratic Party's commitment to racial justice is -- and should be -- unswerving. It does not
follow, however, that every policy adopted during the past quarter-century to promote this goal need
be preserved unchanged, let alone transformed into a litmus test of moral purity.
Affirmative action is a good example. Christopher Edley Jr. and Gene Sperling have made an
eloquent plea for flexibility and innovation -- and for an end to dogmatism, litmus-testing, and
finger-pointing that have dominated discussion of this issue.45 They strongly endorse "carefully
constructed race-conscious remedies to reach the most entrenched forms of discrimination" while at
the same time recognizing that all-out advocates of affirmative action "give scant weight to the
problem of innocent bystanders whose jobs, or chances of a job, may be at stake."46
And yet examinations such as those by Edley and Sperling are all too rare. It is hard to escape
the conclusion that Democrats are afraid even to probe questions such as affirmative action, crime,
and policies to alleviate poverty. After the 1984 presidential election, the Democratic National
Committee commissioned a poll that delineated the problems the party faced among white middleclass voters. Once completed, however, the poll was suppressed on the ground that it was too
controversial.47
Reluctance to examine the established orthodoxies of the Democratic Party has reinforced the
power of litmus tests on a wide range of issues. For example, among Democratic identifiers as a
whole there is a profound division on the abortion question.48 Nevertheless, it is virtually
unthinkable that a serious candidate for the Democratic presidential nomination would deviate far
17

Progressive Policy Institute

www.ppionline.org

from the strict pro-choice position.

The most serious effect of the politics of evasion, however, is that it tends to repress the
consideration of new ideas. Walter Mondale began his quest for the presidency with a highly public
commitment to rethink established orthodoxy. But the dynamic of the nominating process (coupled
with the deep 1981-1982 recession, which rekindled the classic Democratic desire to rerun the
campaign against Herbert Hoover) led Mondale to reaffirm most aspects of the conventional
wisdom and to use Gary Hart's mild and occasional deviations from it as evidence of unacceptable
heresy. Even today, suggestions that the traditional Democratic goals -- for example, improving the
well-being of the working poor -- may require untraditional means are greeted with moral outrage.

Conclusion: The Road Ahead

What is to be done?
The Democratic Party must choose between two basic strategies. The first is to hunker down,
change nothing, and wait for some catastrophe deep recession, failed war, or a breach of the
Constitution -- to deliver victory. This strategy has the disadvantage of placing the party entirely at
the mercy of events. It puts the party in the position of tacitly hoping for bad news -- a stance the
electorate can smell and doesn't like. And it is a formula for purposeless, ineffective governance.
The other strategy, active rather than passive, is to address the party's weaknesses directly. Thus
the next nominee must be fully credible as commander-in-chief of our armed forces and as the
prime steward of our foreign policy; he must squarely reflect the moral sentiments of average
Americans; and he must offer a progressive economic message, based on the values of upward
mobility and individual effort, that can unite the interests of those already in the middle class with
those snuggling to get there. Finally, he must recast the basic commitments of the Democratic Party
in themes and programs that can bring support from a sustainable majority.
There is almost certainly a powerful constituency for such a message. A wealth of data suggests
that the American people are uneasy about the place of our economy in the world, that they favor a
diverse and tolerant society, that they are troubled by the consequences of the increasing gap
between the most and least advantaged sectors of our population, and that they believe our strength
abroad depends on economic and social progress at home. They want leadership that addresses real
challenges and meets real needs.
But all too often the American people do not respond to a progressive economic message, even
when Democrats try to offer it, because the party's presidential candidates fail to win their
confidence in other key areas such as defense, foreign policy, and social values. Credibility on
these issues is the ticket that will get Democratic candidates in the door to make their affirmative
economic case. But if they don't hold that ticket, they won't even get a hearing.
Above all, the next Democratic nominee must convey a clear understanding of, and identification with, the social values and moral sentiments of average Americans. The firm embrace of
programs, such as national service, that link rights to responsibilities and effort to reward, would be
a good start. The consistent use of middle-class values -- individual responsibility, hard work, equal
18

Progressive Policy Institute

www.ppionline.org

opportunity -- rather than the language of compensation would also help. And finally, the American
people overwhelmingly believe that the central purpose of criminal punishment is to punish -- to
express our moral outrage against acts that injure our community. The next Democratic nominee
cannot appear indifferent to the victims of violent crime.
This is not a hopeless task. The Republicans have not solidified their hold on a governing
majority. For all their successes at the presidential level, success at other levels is coming only
gradually. For all the positive opinions that voters have of Republicans, they are still perceived as
the party of the rich. For all the gains they have made in party identification, there are still large
numbers of independent voters in the electorate. The Democratic Party can recapture the middle
without losing its soul.
It should be noted that Republicans are beset with their own ideological purists -- arch conservatives whose instincts and policies are not popular with most of the electorate. But in recent
years the Republican Party and Republican nominees have been better able to put party fights and
party rhetoric behind them and craft a message that appeals to a majority of the electorate.
In contrast, the leadership of the Democratic Party has proven unable to shake the images
formed by its liberal fundamentalist wing and has been prone to take the rhetoric of the primaries
into the general election, with the predictable negative results. The politics of evasion contributes
significantly to this failure by leading its proponents to believe things about the electorate that do
not stand up to empirical tests.
How can the Democratic Party recapture the center? This past spring the British Labour Party
decided that it was tired of losing, dumped some of its extreme left stance, and moved towards the
political center. In the summer of 1989, the Japanese Socialist Party took a similar step. But
American political parties are loose federations that cannot change course through a centrally
designated body. The process of change in the Democratic Party must be as decentralized as the
party system itself. Political leaders at all levels must take a new interest in the party and its
nomination process. Influential Democrats, including candidates for the party's presidential
nomination, must have the courage to challenge entrenched orthodoxies and to articulate new
visions.
This will require an end to litmus tests that have for so long throttled debate. And most
importantly, it will require an end to the conspiracy of silence, to the perpetuation of myths that
have so weakened the Democratic Party. Only conflict and controversy over basic economic, social,
and defense issues are likely to attract the attention needed to convince the public that the party still
has something to offer the great middle of the American electorate. The restrained pace of political
realignment indicates that many voters do not want to call themselves Republicans yet; Democrats
need to give them a reason to retain (or re-establish) their traditional affiliation.
Without active public controversy that begins today, led by Democrats who are able to move
beyond the politics of evasion, Republicans will be able to convince the electorate that the
Democratic Party of 1992 is the same as the Democratic Party of 1972. And if they do, they will win
in 1992 as they have so often in the past two decades.

19

Progressive Policy Institute

www.ppionline.org

The Republican Party was transformed into a governing party during the 1970s because it was
willing to endure a frank internal debate on political fundamentals. If Democrats hope to turn
around their fortunes in the 1990s, they must set aside the politics of evasion and embark upon a
comparable course.

20

Progressive Policy Institute

www.ppionline.org

Appendix A
How increases in Black Voter Turnout would affect the 1988 Presidential Election, by State
State &
Black
VAP (1)
AL
697,000
AR
240,000
CA
1,576,000
DE
80,000
Fl.,
1,133,000
GA
1,138,000
IN
305,000
IL
1,213,000
KY
188,000
LA
873,000
MD
833,000
MI
833,000
MS
587,000
MO
370,000
NJ
756,000
NC
993,000
OH
799,000
PA
783,000
SC
693,000
TN
538,000
TX
1,369,000
VA
808,000

Actual
T/O (2)

% of Black
Vote for
Dukakis

Net Gain
at 52% (3)

Net Gain
at 62% (4)

Net Gain
at 68% (5)

Bush
Margins (6)

40%

90%

70,263

126,720

160,595

266,070

45%

91%

14,099

33,799

45,587

117,341

38%

80%

138,041

234,177

291,858

352,684

31%

89%

12,957

19,197

22,941

30,992

30%

77%

132,432

193,605

230,314

960,746

30%

91%

205,775

300,229

356,902

366,539

50%

81%

5,301

29,091

43,635

437,120

30%

92%

223,465

325,327

386,492

94,999

28%

71%

19,738

28,010

32,973

153,913

39%

96%

104,195

185,384

234,098

166,242

39%

88%

83,800

148,774

187,758

49,863

33%

92%

144,805

220,069

265,227

289,703

40%

92%

61,504

111,399

141,336

193,969

45%

92%

19,455

48,315

65,631

83,334

29%

79%

137,694

196,662

232,042

423,063

26%

87%

197,779

273,247

318,528

347,091

66%

92%

XXXXX

XXXXX

12,550

476,920

52%

89%

XXXXX

60,986

98,100

105,143

28%

92%

137,053

195,265

230,192

235,889

36%

91%

69,230

113,884

140,677

267,439

32%

93%

339,658

458,761

530,223

684,081

41%

88%

72,221

136,053

174,352

449,363

21

Progressive Policy Institute

www.ppionline.org

Appendix Sources
(1) Voting Age Population ("VAP") consists of the estimated black population aged 18 and over, as
reported by the U.S. Bureau of the Census, Current Population Reports, series P-25, No. 1019. As
Published in the Statistical Abstract of the United States, 1989.
(2) Actual turnout ("T/O") is calculated by:
a) multiplying the entire Presidential vote by the percentage of black voters in the electorate.
Presidential vote is taken from 1988 CQ Almanac. Percentage of black voters is taken from ABC
exit polls, as published in ABC News, The '88 Vote.
b) dividing the number of black voters by black voting-age population in the particular state.
Black vote for Dukakis is also taken from ABC exit polls.
3) "Net Gain at 52 percent" signifies the net gain that Michael Dukakis would have received had
black voter turnout been the same as white voter turnout, nationwide. To calculate this, the
percentage of blacks voting for Dukakis and George Bush were taken on a state-by-state basis in the
following states: AL, AR, CA, FL, GA, IL, KY, LA, MD, MI, MS, NC, PA, SC, TN, TX, and VA.
The other states were assumed to vote according to the national average (88 percent-10 percent).
Once the newly estimated vote totals are calculated, the net increase in voters is figured by
subtracting Bush's gain from Dukakis' gain. The net gain is determined by subtracting the old
Dukakis-Bush margin among black voters from the new Dukakis-Bush margin among black voters.
(4) "Net Gain at 62 percent" signifies the net gain that Michael Dukakis would have received had
black voter turnout been ten points higher than white voter turnout. The same method of calculation
that was used above is applied here as well.
(5) "Net Gain at 68 percent" signifies the net gain that Michael Dukakis would have received had
black voter turnout been at the level suggested by the Joint Center for Political Studies. The same
method of calculation that was used in (3) is applied here as well.
(6) "Bush Margin" is the number of votes that separated George Bush and Michael Dukakis in the
particular state. All numbers were taken from the 1988 CQ Almanac, which were compiled from the
various Secretaries of State.

22

Progressive Policy Institute

www.ppionline.org

Endnotes
1. Tubby Harrison's analysis, released June 11, 1989, of data from a survey commissioned by
Democrats for the 90's also found that voters held more positive views of the GOP than of the
Democratic Party: that 43% trust the Republican Party to lead the country versus only 34% who
trust the Democratic Party; that the Republican edge includes the economy as well as national
defense; and that the electorate connects a range of traditional values -- family, religion and the rule
of law -- more often with Republicans than with Democrats.
Similarly, a July 26, 1989 Gallup poll found that on all three barometers of party image -- peace,
prosperity and handling of the nation's problems Americans viewed Republicans as more capable
than Democrats. The poll also showed a continued steady increase in the number of Americans who
think of themselves as Republicans (34%) and Democratic affiliation at only 38% -- a 31 year low
reached in 1985 as well.
2. The ideological makeup of the electorate has not changed very much during the last twelve years.
According to CBS/NYT exit polls, 20% of the 1976 electorate called themselves liberals, 16% of
the 1984 electorate called themselves liberals and 18% of the 1988 electorate called themselves
liberals. There has been similar stability among conservatives, who constituted 31% of the electorate
in 1976, 33% of the electorate in 1984 and 33% of the electorate in 1988. Using a somewhat
different question, a more recent Democrats for the 90's poll (see Note #1) found the same relative
strength of liberals versus conservatives; liberals constituted 27% of the electorate and conservatives
42%.
3. See Elaine Ciulla Kamarck, "Where Have All the Voters Gone?," Newsday, September 12, 1988.
4. This chart originally appeared in the National Journal Convention Preview, June 20, 1988.
5. See E.J. Dionne Jr., "If Nonvoters Had Voted: Same Winner, But Bigger," New York Times,
November 21, 1989.
6. Data taken from CBS/NYT exit polls. The difference is that in 1976 only 22% identified themselves as Republicans and 41% as independents. By 1988 the number identifying themselves as
Republicans had increased to 35% and the number of independents had decreased to 26%.
7. According to CBS/NYT exit polls, Bush won with 92% of Republicans and 57% of independents.
8. See James De Nardo, "Turnout and the Vote: The Joke's on the Democrats," American Political
Science Review, 74:406-20.
9. See Harvey J. Tucker and Arnold Vedlity, "Does Heavy Turnout Help Democrats in Presidential
Elections?," American Political Science Review, 80:1291-98.
10. Norman Ornstein, Thomas E. Mann and Michael J. Malbin, Vital Statistics on Congress, 19871988 (Washington, D.C.: American Enterprise Institute/Congressional Quarterly, 1987), p. 46.
23

Progressive Policy Institute

www.ppionline.org

11. See E.J. Dionne Jr., "Voter Sign Up Bill Gains in Congress," New York Times, May 7, 1989,
and Peter A. Brown, "Democratic Plan to Increase Voter Registration Could be GOP Boost,"
Scripps-Howard News Service, May 20, 1989.
12. In Pennsylvania, Ohio, Michigan and Illinois, Bush took half of this traditionally Democratic
bloc; in New Jersey Bush won 61% of the Catholics. (Data from exit polls in ABC News, The '88
Vote.)
13. Ruy Teixeira, "Registration and Turnout," Public Opinion, January/February, 1989.
14. To confirm our results, we performed the same analysis using an alternative source of data, the
November, 1988 Current Population Survey (U.S. Bureau of the Census), which has larger state
level sample sizes than exit polls. We found that using Census data did not change the results; in
fact, it made them stronger. Fewer states with fewer electoral votes go to Dukakis when this data
source is used.
15. Increasing Hispanic turnout in California to 52% of the Hispanic voting-age population and
assuming that 64% of the Hispanics voted for Dukakis (ABC exit poll) gives Dukakis a net increase
of 1,067,731 votes, enough to win the state. In Texas increasing Hispanic turnout to 52% (assuming
a 75% Democratic vote) gives Dukakis only 602,643 extra votes, not enough to beat Bush. Ruy
Teixeira has performed a similar analysis of state level selective mobilization using slightly different
data (The New Republic, April 3, 1989). He looked at every state Bush won and calculated the
effects of up to 20 percentage point increases in, not only black and Hispanic turnout, but also
turnout among the white poor. His results are very consistent with ours only a small number of
states go for Dukakis, far short of enough to swing the election.
16. See Dr. Linda F. Williams, "'88 Election Results: Problems and Prospects for Black Politics,"
Joint Center for Political Studies, Washington, D.C., p.11.
17. Louisiana is a good example. Democratic registration among blacks in that state increased by
77,779 voters in the eight years of the Reagan presidency but Democratic registration among white
voters decreased by 161,230 more than twice as much -- while Republican registration increased by
176,477.
18. According to ABC exit polls, in Florida, Senate winner Bob Graham ran worse among blacks
than did Mondale but ran 24 points ahead of Mondale among whites, who made up 92% of the
electorate in 1986 and 91% in 1984. In North Carolina, losing 1984 Senate candidate Jim Hunt
(who ran far ahead of Mondale) and winning 1986 Senate candidate Terry Sanford ran equally
strongly among black voters, but Sanford ran a critical 4% better among whites, who comprised
86% of the electorate in both years. In Alabama, according to CBS exit polls, black turnout was
higher in 1984 than in 1986 (24% and 21%, respectively), and Mondale won a higher percentage of
the black vote (93%) than did the 1986 Senate winner, Richard Shelby (88%). But Shelby ran 17
points better than Mondale among whites and eked out a close victory.
19. CBS/NYT and ABC exit polls show Dukakis with 49% of women's votes. In 1976, according
24

Progressive Policy Institute

www.ppionline.org

to CBS/NYT exit polls, Carter received 51% of women's votes.

20. It is of course possible that remobilization of support around the abortion issue will change the
dynamics of gender politics in ways that are hard to predict, conceivably to the Democrats'
advantage. But a recent CBS/NYT poll casts some doubt on the ability of the abortion issue to
mobilize large segments of the population. See E.J. Dionne Jr., "Poll Finds Ambivalence on
Abortion Persists in U.S.," New York Times, August 3, 1989. See also James A. Barnes, "Politics
After Webster," National Journal, August 12, 1989, pp.2044-48, and Jack W. Germond and Jules
Whitcover, "Abortion Issue Giving Democrats Big Opening," National Journal, August 12, 1989,
p.2056.
21. The daunting mathematical problems faced by Democrats as they look at the electoral college
and the advantages of the Republicans are described in Jack W. Germond and Jules Whitcover,
Whose Broad Stripes and Bright Stars? (New York: Warner Books, 1989), especially Chapter 29,
"The Dukakis Surge."
22. Bob Hall and Barry Yeoman, "What Happened November 8th," Southern Exposure, Spring
1989.
23. For a very good explanation of the phenomena of incumbency at the congressional level, see
Thomas Mann, Unsafe at any Margin: Interpreting Congressional Elections (Washington, D.C.:
American Enterprise Institute, 1978). In their book Politics and Society in the South (Cambridge:
Harvard University Press, 1987), p.285, Earl and Merle Black point out that the recent trend towards
allowing Southern governors to serve two consecutive terms (nine states changed their laws on this
in the last two decades) has inhibited Republican gains by increasing the number of incumbents.
24. Ibid, p.285
25. In the entire decade of the '70s there were only 76 open congressional seats and 84 incumbent
seats that changed party. Of these only 37% moved from previously Democratic to Republican. In
the '80s there were 132 House seats that changed party. 59% of these seats moved from previously
Democratic to Republican -- a not inconsequential increase from the previous decade.
26. See Elaine Ciulla Kamarck, "Cutting in on Campaign Cash," Newsday, February 12, 1989, Ideas
Section, p. 1.
27. Governor Jim Hunt in North Carolina in 1984 and Congressman Buddy McKay in Florida in
1988 are two examples of strong Senate candidates who were adversely affected by the presidential
candidate at the top of the ticket. Buddy McKay ran 22 points ahead of Michael Dukakis in Florida
and still lost, although narrowly.
28. Public Opinion Magazine, "Opinion Roundup, Gallup Youth Survey," p.22.
29. R.W. Apple Jr., "Public Rates Bush Highly But Sees Mostly Style," New York Times, April 20,
1989, B12. See July 26, 1989 Gallup poll, "Republican Party's Image at a High Point," by Andrew
Kohut and Larry Hugick.
25

Progressive Policy Institute

www.ppionline.org

30. See Thomas Edsall, "Racial Forces Battering Southern Democrats," Washington Post, June 25,
1989, p.A6; James A. Barnes, "Florida Snowball," National Journal, July 29, 1989, p. 1959; Dave
Maraniss, "In Wright's Texas Alliances are Shifting," Washington Post, August 10, 1989, p. A3.
31. Between Bush's inauguration in January, 1989 and August, 1989, 130 elected officials -- most of
them from the South -- had switched from the Democratic to the Republican Party. See Elaine
Ciulla Kamarck, "Pols Who'd Rather Switch Than Fight," Newsday, August 21, 1989. Sec also
James M. Perry, "Republican Campaign Is Paying Off As Conservative Democrats Make Switch in
South," Wall Street Journal, August 14, 1989, p. A10.
32. Susan Feeney, "Hostile Student Views at LSU Have Two Top Demos Rattled," New Orleans
Times Picayune, March 9, 1989.
33. For example, when Democratic Congressman Bill Nelson announced that he would leave to run
for Governor of Florida, no fewer than 14 candidates began jockeying for the Republican
nomination.
34. See Steven Rosenstone, Forecasting Presidential Elections (New Haven: Yale University Press,
1983), and Steven Rosenstone, "Explaining the 1984 Presidential Election," Brookings Review,
Vol. 3, #2, Winter 1985, pp.25-32.
35. Rosenstone, "Explaining the 1984 Election," Table #4.
36. See Robert S. Erikson, "Economic Conditions and the Presidential Vote," American Political
Science Review, 82-3 (June 1989), pp.567-573.
37. Those who felt defense and foreign affairs were most important accounted for 22% of the
electorate according to ABC exit polls. The post-election survey of Americans Talk Security
Project, Boston, Massachusetts, December, 1988, Numbers 10 and 11, corroborates these findings.
On election day, 67% of the voters thought Bush would deal with the Soviets better than Dukakis
would; on arms control negotiations, it was Bush 64, Dukakis 22; maintaining a strong defense, 66
to 22; fighting terrorists, 57 to 26. Even in keeping the country out of war, a traditional Democratic
advantage, Bush held a 47-36 edge. More than a third of the electorate (37%) had reached the
conclusion that Dukakis would actually weaken our national security, while only 8% thought Bush
would.
38. ABC News, The '88 Vote, p.22.
39. The poll cited in footnote #1 showed that Republicans have a comfortable lead over Democrats
on a series of values questions having to do with belief in God, pride in the country and respect for
the flag. The CBS/NYT poll mentioned here is cited and discussed in Samuel L. Popkin, "Outlook
on the Future and Presidential Voting," prepared for delivery at the 1988 meeting of the American
Political Science Association, Washington, D.C.
40. For a quintessential exposition of this argument, see Susan Estrich, "Willie Horton and Me,"
26

Progressive Policy Institute

www.ppionline.org

Washington Post Magazine, April 23, 1989, pp.20-25.

41. For a good review of this history see Michael Oreskes, "Civil Rights Act Leaves Deep Mark
on the American Political Landscape," New York Times, July 2, 1989.
42. The first exposition of this can be found in Jeane J. Kirkpatrick, The New Presidential Elite
(New York: Russell Sage Foundation, 1976). Anthony Lukas' Common Ground (New York: Knopf,
1985) and Jonathan Rieder's Canarsie (Cambridge: Harvard University Press, 1985) offer vivid
depictions of the growing tensions between white working-class Democrats, blacks and white elites.
43. For an explanation of the phenomenon of upper income and educational groups shifting to the
Democratic Party, see Everett Carll Ladd Jr., Where Have all the Voters Gone?, 2nd ed. (New York:
Norton, 1982).
44. Ever since the 1980 election, the Republican Party has managed to win nearly half of the union
vote in spite of very large and expensive union campaigns on behalf of the Democratic nominees.
According to CBS/NYT exit polls, Carter won 62% of union households in 1976; in 1980 he won
only 48% of union households. Mondale managed to get 54% of the union household vote in 1984,
and Dukakis received 57% in 1988.
45. Christopher Edley Jr. and Gene Sperling, "Have We Really 'Done enough' for Civil Rights?,"
Washington Post, June 25, 1989, B1.
46. Ibid.
47. Peter A. Brown, "Democrats Concealed Study of How to Win White Votes," Scripps-Howard
News Service, April 17, 1989.

About the Authors

Dr. William Galston received his Ph.D. from the University of Chicago and is currently
Professor, School of Public Affairs, University of Maryland at College Park, and Senior Research
Scholar at the University's Institute for Philosophy and Public Policy. He is the author of three
books and numerous articles on American politics, public policy, and political theory. He served as
Issues Director in the Mondale for President campaign, 1982-1984.

Dr. Elaine Kamarck received her A.B. degree at Bryn Mawr College and her Ph.D. in political
science at the University of California, Berkeley. She is a veteran of the Democratic National
Committee and three Democratic presidential campaigns. She is currently a Senior Fellow at the
Progressive Policy Institute, a regular contributor to Newsday, and a Visiting Assistant Professor at
Barnard College in New York City.

27