Date: Wed, 16 Nov 1994 00:50:35 -0600
Sender: H-Net Political History discussion list
Subject: ELECTION '94: Roundtable - "Theses on the 1994 Election" -
Burnham
***************
Theses on the 1994 Election
Walter Dean Burnham
University of Texas at Austin
Copyright, 1994, Walter Dean Burnham, all rights reserved
NOTE FROM PROFESSOR BURNHAM: "This essay is in rough-draft,
unedited form and will probably be extended somewhat prior to
publication. I should like to request that in any professionl
reference to the presentations in Tables 1A and 1B, users contain
an explicit citation of this source."
A TRANSFORMING ELECTION: 1994 IN CONTEXT
1. The 1994 election is very widely perceived as an excep-
tional turning point in the electoral, party and policy history
of the USA. This perception is quite correct. The situation is
also "unusual" in another dimension. The vast majority of the
time, off-year elections modify or confirm an underlying state of
affairs, but they do not appear to inaugurate a turning-point in
the organization and articulation of the political system.
Historical cases that come to mind of elections with system-
transforming potential include only 1854, 1894, (possibly) 1930
-- and 1994 -- three or four cases in all, out of 52 off-year
congressional and state elections. Certain other elections --
1838, 1874, 1910, 1918, 1946, 1958 and 1974 (eight in all) --
involved considerable electoral and policy shifts, but fell below
the "mega-event" range. The other 40 cases were for the most
part articulations of "politics as (then-) usual" with marginal-
to-minimal impact on the system as a whole. Of all these elec-
tions, the structure of the 1946 election in many ways most
closely resembles that of 1994. This should remind us of some-
thing. 1946 was a very large event, temporarily recreating the
electoral structure of the 1920s; and it was significant in its
policy consequences (notably in labor law -- the Taft-Hartley act
-- but also otherwise). But it electorally did not lead
anywhere; it was a "sport," promptly repealed in 1948. The pos-
sibility should not be excluded that 1994 will resemble 1946 in
this regard as well, but I rather think not.
2. Is 1994, then, Really Realignment? It would of course help
to know the results of the next four or five congressional elec-
tions; but since that knowledge is not given to us, I can only
express the provisional view that it leads toward a marked
acceleration of crisis/conflict politics in the USA and that,
both electorally and in policy terms, some of its consequences
are not likely to be reversible in the foreseeable future.
3. Let us recall the immediate past, and then contrast it with
what happened in 1994:
a.) Conventional wisdom had had it that the decline of party-
in-the-electorate had led, in a candidate-centered world, to huge
and bilaterally-distributed district advantages for incumbents of
both parties. Along with the near-disappearance of open seats,
this process reached its apogee in the 1988 election, in which
competitive district outcomes also nearly vanished. In turn, of
course, it has lead to a congealing of the electoral process
(tiny swing ratios, very low partisan and personnel turnover)
that formed the perfect background for a populist drive for term
limits. What had been produced was a near-perfect re-election
machine, with Democrats congressionally dominant for 40 years in
the House (and 30 of 40 years in the Senate), while since 1968
Republicans have normally won presidential elections. This
formed a kind of "division of labor" in a deadlock or interregnum
state, with (cf. Cantrell and Roll, 1967) "ideological con-
servatism" represented in the presidency and "operational
liberalism" (also then at the majority mode of public opinion)
represented in Congress. It also meant that there were in effect
_two_ majority parties in this period rather than _one_: follow-
ing Lowi (1964) an ideal condition for making the system as a
whole unresponsive to growing demands for change.
i.) These trends began showing signs of reversal in 1990 and,
even more, in 1992; while I should have been more alert to their
implications than I was, they were still movements in minor key.
1994 was an abrupt, brutal and fundamental change, in which
Democrats lost 56 USHR seats (35 incumbents and 21 open seats) to
for open-seat losses for the Republicans. In sheer volume of
turnover, 1994 ranks fifth of 21 off-year elections since 1918 --
just a shade behind 1946. But 1946, like all other of the top
seven -- and also chronologically the last of these seven -- was
held long before the development of candidate-dominated elections
and massive incumbent-insulation effects.
ii.) 1994 is striking in another respect, to which we shall
return when we attempt a synthetic overview of what is going on.
In the past 76 years there have been 10 (of 21) off-year elec-
tions in which the presidential party had a gross loss of at
least 40 seats. All of them except 1994 are associated with
abnormal short-term contextual stresses for the presidential in-
party. In rank order (though some elections were a mix of more
than one stress), these stress factors were: (1) economic slump;
(2) the impact of war or reconversion/stabilization stress
immediately following war; (3) (for 1974 only, and in addition to
economic stress en route to deep recession following the first
oil price shock of 1973), basic presidential scandal (Nixon's
resignation and his pardon by his successor). NOT ONE OF THESE
CONDITIONS WAS PRESENT IN 1994.
iii.) Associated with this, there is something else: the
widespread "c.w." view that presidential parties take their big-
gest hits in the sixth year (or later) rather than in the second
year of a newly incumbent administration. In partisan-control-
duration terms, the only historic cases of a second-year "bath"
endured by the presidential party were in 1914 and 1922, prior to
1994. But the 1914 result was a political artifact (massive
entry of Progressive candidates in 1912 siphoning off the
Republican vote and leaving Democrats with 295 seats to defend,
followed by partial collapse of those third entrants in 1914 and
their voting streams rejoining the "normal vote" of the time);
and can therefore be set aside. In 1922, Republicans had 303
seats to defend following the greatest landslide in modern GOP
history (1920); and moreover with effects of a very deep post-
World War I recession. Adding the two factors together accounts
for most of the 80-seat GOP drop in 1922, which even at that left
the party with an overall majority of nine in the 1923-25 House.
In 1994, by contrast, Democrats had 256 seats to defend (not much
different from the 244 in 1946), and this was no artifact of
landslide in the preceding presidential-election cycle (on the
contrary, the Democrats lost ten seats net in 1992, and another
two in 1994 before the election). A loss of this magnitude in
the second year of a new presidential party is thus substantively
UNIQUE so far as the historical record is concerned.
b.) As Tip O'Neill used to say, "all politics is local." A
standard theme of much congressional literature has been that of
effective localization by incumbents of the terms of their elec-
tion, such that (largely by effectively claiming "delivery of the
goods" from the Washington policy machine for the benefit of
their districts) they could and did easily survive adverse
presidential tides in those years. In 1984, all Democratic con-
gressional incumbents had to do to breast the Reagan tide was to
run more than 19 points (on average) ahead of Walter Mondale in
their districts; and Reagan actually won more districts returning
Democrats at the same time than Republicans.
i.) The Republicans in 1994 chose to run a campaign which, to
the maximum possible extent, sought to "nationalize" the elec-
tion. They were resoundingly successful in, essentially, over-
coming enough of the orthogonality between the two opinion modes
of "ideological conservatism" and "operational liberalism" to win
power and taking rather thin but I think policy-decisive control
of both Houses of Congress. Certain other off-year elections
have had this characteristic; but at least in post-1968 terms,
the only one that one can point to is 1974. In both 1974 and
1994, we should add, very clear signals of upheaval were given
long in advance by the presidential party's consistent loss in a
string of bye-elections.
c.) "The party's over" -- or so we've heard for the past
twenty and more years. The congelation of congressional election
results, the entrenchment of incumbents and the usual "all-
politics-is-local" syndrome all reflect a steep decay of party as
the dominant entrepreneurial organizer of the electoral market.
As usual when talking about any aspect of electoral politics
(including critical realignments or anything else), we deal here
not with "everybody" but with politically decisive minorities
(and with considerable residual strength, after all, in party ID
measures, for example).
i.) 1994, in contrast, has many (though by no means all)
characteristics of an old-fashioned PARTISAN ELECTION. One of
the most striking of its features is that it reflected an
abnormally large and consistent Republican surge at all major
levels: whether one is looking at elections to the U.S. Senate,
the U.S. House, state governors or (both houses of ) state legis-
latures, the pattern of Republican surge carries across the
board. It also does so geographically. Much is said about the
South, and for good reason -- since after 1994 (and for the first
time since 1872), the Republicans have a majority (64 to 61) of
the 125 seats in the former CSA. This has fundamental long-term
structural implications. But in terms of swing, the South in the
main by no means stands out. Apart from New England, every
region of the country showed strong Republican swing and
Democratic losses, with the Midwest in particular leading the
way. And, of course, Spokane, Washington, where House Speaker
Tom Foley lost his bid for a 16th term in the House, forms no
part of the old Confederacy. (Indeed, both at the state-office
level -- governor and judges apart -- state-legislative contests,
and in the congressional races, Texas in the end showed less
Democratic erosion than most states, not only in the South but in
the rest of the country.)
We should be careful not to over interpret this. State-level
regressions among gubernatorial, senatorial, and congressional
percentages Republican of the two-party vote still produce R
squares insignificantly different from zero (a post-1968 develop-
ment; earlier in the century, values ranging from .75 to .95 were
common enough). Nor, of course, has the distribution by party of
CD outcomes suddenly "snapped back" to full approximation of
1940s levels. But they have moved in that direction.
* * *
There are various ways of attempting to size up the relative
deviation of an election result from what could be called
"expected." Ideally, as we have said, we should like to use
something like iterated t tests or R squares, or discriminant
analysis of the sort employed, for example, in statistical
geography. But alas, while we can do this for the past and thus
identify the temporal location of critical-realignment upheavals
and midpoint crisis "moments," and while the geographers always
have the whole territory to analyze, we cannot see into the
future. The next best recourse is to evaluate the 1994 shift in
two-party percentages of the vote and the seats (the latter both
for the USHR and state legislatures), and then measure the next-
year deviation from the mean of the past five elections, dividing
this figure by the five-election standard deviation. In the
overwhelming majority of cases, this will produce a number
smaller than 2.00 (29 of 34 so far as votes are concerned, or
85.3% of cases from 1918 to 1994; and 30 of 34 so far as seats
are concerned, or 88.2%). When one iterates across an entire
field of data in this way, "turning-points" of various absolute
and relative magnitudes duly emerge (see Tables 1A and 1B).
TABLE 1A: CUTTING-POINT ELECTIONS: US HOUSE
Election- %Dem 2-Party Next Value Dev. fr. dev./ Non-South:
year band Mean s.d. Year M s.d. dev./s.d.
--------- ---- ------- ---- ----- ------- ----- ----------
1884-92 V 52.2 1.58 1894 V 44.3 -7.9 -5.00 -7.21
S 58.6 9.56 S 29.8 -28.8 -3.01 -3.14
1922-30 V 43.4 2.27 1932 V 56.2 +12.8 +5.65 +4.93
S 44.7 4.58 S 72.2 +27.5 +3.64 +6.62
1948-56 V 51.6 1.58 1958 V 56.7 +5.1 +3.24 +3.01
S 54.2 4.12 S 65.0 +10.8 +2.62 +2.98
1964-72 V 52.8 1.81 1974 V 59.2 +6.4 +3.55 +2.08
S 59.0 5.00 S 66.9 +7.9 +1.58 +2.08
1984-92 V 53.8 4.19 1994 V 46.3 -7.2 -6.29 -5.67
S 59.7 1.32 S 47.0 -12.7 -9.66 -7.43
TABLE 1B: CUTTING-POINT ELECTIONS: STATE LEGISLATURES
1884-92 Sen 53.8 4.19 1894 Sen 42.3 -11.5 -2.75
HR 52.8 4.45 HR 36.2 -16.6 -3.73
1922-30 Sen 45.3 1.67 1932 Sen 63.7 +18.4 +11.00
HR 46.1 4.28 HR 64.4 +18.3 +4.28
1948-56 Sen 55.2 3.01 1958 Sen 66.2 +11.0 +3.65
HR 54.5 3.08 HR 65.4 +10.9 +3.54
1964-72 Sen 60.4 4.70 1974 Sen 67.5 +7.1 +1.51
HR 60.6 3.46 HR 67.9 +7.3 +1.84
1984-92 Sen 60.8 0.89 1994 Sen 53.3 -7.5 -8.45
HR 59.5 1.08 HR 52.6 -6.9 -6.40
------------------------
V = Votes
S = Seats
Sen = Senates
HR = Houses of Representatives
________________________
While the crucial dimension of durability cannot of course be
tested in this way, the extraordinarily high values of the
dev/s.d. figure achieved in 1994 are certainly worthy of comment.
This is to some extent an artifact of the very low standard
deviations found in the 1984-92 period. But this can be read
_also_ reflecting some sense of the magnitude of the 1994 break
from most recent history, marked as that history has been by all
the factors that led to the remarkable damping-down of interelec-
tion swings culminating in the 1988 nothing-but-actuarial-change
election. And considering that very large parts of the
"candidate-dominated" election profile continue in 1994 as they
have over the past generation, even the size of the absolute
deviations from the previous five-election means registered this
year is remarkable.
Should this be read as "realignment"? The reader (or listener)
can take his or her choice. All I can say is that, the more
closely I scrutinize the 1994 outcome, the bigger it looms in my
own mind. Before turning to a more interpretative or expressive
mode of discourse, I should note that at one point the Southern
transformation looms structurally quite large in the probably
future partisan history of the of the House of Representatives.
Once, Democrats automatically received a 100-seat bonus from this
then one-party region. Thorough a series of pro-Republican
"ratchets," this majority was scaled back to the 29-seat majority
of recent elections. But the 1992 balance of 77 D, 48 R, was
transformed in 1994 to 61 D, 64 R. Nor, in all probability, is
this the end: at least another dozen seats may well fall to the
Republicans before some relatively stable balance is reached.
Without speculating in detail, we may note that the huge
Democratic majorities following 1964 and 1974-1976 were
importantly based on a Southern Democratic "bulge" that has com-
pletely disappeared, and may be replaced eventually by a
Republican "bulge" of perhaps 25 to 30. These "ratchets,"
including 1994, point in one direction. There is little reason
to anticipate a major reversal in the foreseeable future.
There are two important implications involved, particularly when
one considers both intrapartisan and interpartisan congressional
dynamics in the near future. The first is that, barring a pro-
Democratic national punctuational-change flipover at some point
in the future, any Democratic House majorities that may emerge
will be much smaller than has been the norm since 1933. And one
could expect that partisan turnover of majorities in the House
will become much more frequent in the years/decades ahead than
over the past two generations. Secondly, 1994 was an election
where, in a significant number of cases, Democrats who claimed to
be conservative-to-moderate were defeated (particularly in the
South and the border states). The electorates there opted for
the real thing and will of course get it. The growth of the
Southern congressional Republican party has been almost
exclusively concentrated on the far-to-extreme right -- a prime
reason, along with perhaps less marked developments in the party
north of the Mason-Dixon line, why Newt Gingrich rose to command-
ing influence in the House GOP well before the nominal minority
leader, Bob Michel, retired in 1994; and why, in January 1995,
Newt Gingrich will become the next Speaker of the House. On the
Democratic side, the disappearance or defection of the party's
right wing will inevitably mean that the party's congressional
caucus will find its center of gravity shifted to what passes for
the left in this country. Growing interparty polarization --
quite likely extreme polarization -- seems inevitable in con-
sequence.
END PART I
*********
Part II
COPYRIGHT, 1994, Walter Dean Burnham, all rights reserved.
WHY DID IT HAPPEN, AND WHAT DOES IT PORTEND?
Some causes appear to be the following:
1. The long-term and continuing deterioration in the living
standards of "middle America," not merely among the working class
but among large parts of the middle class as well. It is
probably for this reason that in October of 1994 (_Newsweek_ sur-
vey), 59% of Americans believed that the country was still in a
recession. There certainly has been a recovery since early 1992,
but largely "jobless" in character. So far, no quarter has shown
growth of much above 3.5%, while occasional quarterly growth
rates of 5 or 6% are needed to create conditions for "classic"
expansion. And the Fed -- having to worry about the bond market
and international currency traders -- will continue to ratchet up
the rate of interest to head off any trend toward inflation (if
it can), thus producing a growth ceiling. For very many
Americans, serious anxieties not only about one's own economic
future but those of one's children remain at high levels.
2. It has not helped that Clinton devoted so much domestic cap-
ital to a "Rube-Goldberg" 1,200+ page plan for universal health
care. This caused economics to disappear entirely from view
(stupid), while opening wide the gates for Republican/industry
seizure of the agenda's "high ground" by attacking the threat of
yet another vast, intrusive and expensive government program.
Whether rightly or wrongly, most Americans (and especially most
people who actually voted in 1994) are probably satisfied enough
with the health care they have, and fear unexplained innovations
that can be tapped into their modal ideological conservatism.
3.) There is the sheer brute fact that Clinton is so widely
unpopular (one survey indicates that 20% of respondents hate
Clinton, and other 25% dislike him). At bottom, there is some-
thing not easily explicable about the extent of this dislike,
though obviously the links between the RNC and key talk-show
people like Rush Limbaugh help account for this (a real
"mobilization of bias" on the airwaves, in Schattschneider's
sense). But there are other factors.
a.) I think he never recovered from beginning his administration
with the gays-in-the-military issue. Strategically concentrated
as they may be in certain fields like the media, the arts and
higher education, homosexuals are a small-to-tiny-minority of the
American public and are intensely disliked (especially in the
culturally conservative South). But connected with this is a
whole range of other empowering efforts undertaken by Clinton in
Washington (and I also suspect Ann Richards in Austin) on behalf
of other minorities and -- from the main-street/middle-America
point of view -- marginal, disliked outgroups.
b.) Other publicized issues feed into this, and also into the
"religio-cultural wars." In addition to the standard Protestant
religious right, Clinton from the beginning has been on the
poorest of terms with the Catholic church and with the military
establishment. Certain "lightning-rod" figures like Surgeon-gen-
eral Joycelyn Elders have served to give additional personal con-
centration to this fury, and may help account for the report we
hear that Republican turnout was mobilized in 1994 (and
Democratic turnout demobilized) for the first time since 1970.
c.) More generally, the Democratic party since McGovern's time
has acquired the image of being in favor of all sorts of
minorities and social-activist interest groups against the
cultural, but also the economic, interests of the white middle-
class male population. The activities, preferences and public
images of the Clinton administration not only reactivated this
image but -- precisely because it now _was_ the presidential
party -- attached the stigma to Democrats as a whole in 1994. It
is striking that white males in the 30-44 age group showed among
the largest of all demographic group swings to the Republicans in
1994. In a sense, the once-famous "quota" case of _Bakke v. Bd.
Of Regents_ has become universalized and has come home to roost.
And, what could perhaps be accepted in the economic environment
of the 1970s is perhaps no longer tolerable in the continually
diminishing, corporate-downsizing world of the 1990s: jobs as
well as values are at stake.
4. Further exigency is given on the cultural front by
apparently manifest signs of the growing rotting of the social
fabric; focusing chiefly on the crime issue (particularly in
dramatic cases, lavishly covered by the media, of various
atrocities like husbands allegedly killing wives [O.J. Simpson]
or mothers killing their own children [Mrs. Smith of S.C.], and
the spread of violence and killing among subteenage children; and
also on the perennial education and welfare issues.
REVITALIZATION is called for, in a context (complete with
ideologically polarized interest groups) uncannily similar to the
pre-revolutionary "cultural distortion" Phase 3 in Chalmers
Johnson's (1966) model of _Revolutionary Change_. When large
numbers of people are stressed and find the present intolerable,
they seek saviors with time machines. Since there is a growing
market, such saviors regularly present themselves. Whether the
dials on these machines are pointed toward the future or the past
will depend entirely on the kind of society it is and the time in
which this impetus occurs.
5. In the case of the modern USA, all revitalization movements
will be firmly aimed at return to a past apparent state of bliss.
This is because there is no organized left in the USA, and also
because the only known incentive in a propertied-liberal
hegemonic culture that could produce a desire for some leap into
the future would be economic catastrophe equal to or exceeding
that of the Depression period.
This yearned-for state of bliss requires for its achievement a
reinstatement of the older cultural, social and political order
of things so far as it is possible to achieve it. The country,
and particularly the active electorate, is in the main hostile to
government as such, tolerating it only when it delivers concrete
goods and the tax price is not too high. Back in 1800, the tri-
umph of the Jeffersonians involved absolute victory of the
Country party and its ideology over the Federalist Court party
and its ideology. Key elements of the doctrine were: (1) that
government governs best that governs least; (2) state and local
government, where any is needed at all, is to be preferred to
federal; (3) the majority is wiser than the minority. Since the
1930s, and especially since the 1960s, we have abandoned the
ancient truths -- so the argument goes, as framed, inter alia, by
Speaker-designate Gingrich -- and must return to them. In the
meantime, a vastly influential and thoroughly un-American
cultural elite has come to dominate the media and the
universities, inculcating pernicious doctrines that have
profoundly corrupted the Republic.
Essentially -- again, so the argument goes -- Ronald Reagan's
effort at revitalization was all well and good so far as it went,
but it did not go anywhere near far enough. What is required is
a genuine (counter-) revolution that, root and branch, digs out
and destroys (or extirpates) the bad old system that got us into
this mess and return to ancient truths. The way ahead in achiev-
ing this goal requires the dismantling of most of the federal
government. It will also require, at some point, attending to
the strategic role of secular-humanist "cultural elites" in the
institutions which they control. At the end of the day, we will
achieve this unity which so many ordinary Americans want, replac-
ing and overcoming interest-group fragmentation and ideological
polarization that characterized the old order.
Have we heard this sort of thing somewhere before in this
century? If you think I am making this up, by all means read and
ponder what Mr. Gingrich has been saying since the election, not
to mention his effusions before. Once, not very long ago, he
could be regarded as just another right-wing blowhard. Now he
will occupy what many regard as the second most-powerful office
of state. It is doubtful, to say the least, that he will be able
to realize his maximum program, at least before 1997, on so
fragile a base; and Senate Republicans, from Bob Dole downwards,
will have their own and very different ideas of what the times
require. Nevertheless, as the old slogan of 1968 put it, you
don't have to be a weatherman to know which way the wind is blow-
ing.
In sum, 1994 appears to have marked a major accelerating point in
the political crisis which has been increasingly engulfing the
United States in the past 20-30 years. We are in for explosively
polarizing times. Whether 1994 turns out to be a "sport," like
1946, or the dawning of a new era is of course impossible to say
with certainty. It may be nothing more in the end than an
unusually empathic closure of the latest midpoint crisis in an
ongoing regime order, though I think it is rather more than that.
The public continues to grope for a satisfactory revitalization
tool. Newt Gingrich, in his demand that the whole Republican
"contract" be carried out and form the basis for further develop-
ments, has noted that if the Republicans in their turn fail to
achieve this, the stage will be set for a public turning to third
parties (or candidates). In this we may well agree with him.
Events will decide whether the growing contradictions in American
economic, political and social life can be contained or channeled
through revitalization strategies that Gingrich and his allies
propose. If not, then among many other things the liquidation of
the traditional two-party system may indeed follow.
Whatever happens along these dimensions, the political jolt in
1995 will be tremendous. Forty years is a very long time; it is
one-fifth of the history of the American republic and well over
half the average life expectancy for individuals. It is as
though a great sledgehammer blow was struck against a Capitol
dome which appeared to be of iron but was actually of glass. It
was shattered into bits. Just this sort of thing has happened in
earlier American cycles of punctuated-change upheaval. The crea-
tive potential of electoral politics not to be eternally repeti-
tive is again confirmed. As for myself, my memory of things
political extends back to the 1940 election. In these more than
fifty years, I have never experienced an off-year election of a
transformative impact that remotely equals 1994, and only one or
two presidential elections that are in the same range. In this
half-century, both the Republican and Democratic parties have
undergone profound transformations. We have not seen the House
Republicans in a majority position since most of that transforma-
tion occurred. For one thing, the distance between Speaker
Joseph W. Martin and Speaker Newton Gingrich is measured politi-
cally in light-years. The sequel seems bound to be not merely
interesting, but dramatic in the extreme.
At the end of the day, the anxieties I have long felt about the
future of democratic government in this country have been con-
siderably sharpened in the wake of this election. And I should
hasten to add that this heightened anxiety has nothing to do with
the fact of Republican victory as such. But it has everything to
do with the entire _mis-en-scene_ I have inadequately tried to
describe in this discussion, one in which this Republican victory
and this new cast of policy-dominant characters duly take their
place.
* * *