Elsevier Science International Encyclopedia of the Social & Behavioral Sciences Vol. C, Books, Books eng, books NON ...
[ Pobierz całość w formacie PDF ]
C
Campaigning: Political
1986, Kazee 1994). Generally the rather obvious
benefits far outweigh the costs of winning for those
with the requisite political ambition; but costs must be
weighed—loss of privacy, loss of time with family,
frequently loss of income. Similarly, the evident costs
of losing are almost always greater than the benefits;
but some benefits do accrue to losers—making a name
for oneself, building up good will with party o
cials.
Beyond these broad generalizations, little is known
about how specific candidates make individual de-
cisions given the dazzling array of di
cult-to-estimate
variables (Maisel and Stone 1997).
At the microlevel, political campaigning involves an
individual candidate attempting to win political o
ce.
At the macrolevel, political parties seek to gain seats in
government for their partisans and ultimately to gain
power or at least to share power in the relevant
governmental arena. At each level decisions must be
made about candidacies, organizations must be built,
messages must be refined, strategies must be set, and
the public must be convinced to support those waging
a particular campaign. Only certain aspects of political
campaigns are viewed by the public. Frequently, the
actions taken away from public view are more de-
terminative of success and failure.
1.2 Parties’ Choices of Candidates
The ways in which political parties choose their
candidates for o
ce varies significantly from political
system to political system and from party to party. The
key variable in this aspect of the process relates to the
extent to which party organization controls places on
the ballot contrasted with voters acting independently.
In the former, the process is often described as
candidate recruitment, in the latter, candidate emerg-
ence.
At one extreme are those systems in which political
parties slate their candidates for o
ce, ranking cand-
idates or placing them in safe constituencies. Most
parties operating in electoral systems with propor-
tional representation follow this procedure. It is also
followed in systems with strong party systems, in
which citizens vote more for parties than they do for
individual candidates. Many European systems fit this
description (Hix and Lord 1997).
At the other extreme are systems, of which the
United States is prototypical, in which voters choose
party candidates in primary elections. The shape of the
electorate in primary elections varies considerably
from polity to polity. In theory, primary elections are
open only to members of one political party or
another, but in actual practice how membership is
defined varies according to local rules and can be quite
restrictive or totally open. In either case, voters in
the primary elections determine the party nominee,
even if that nominee would not have been the choice of
party o
cials (Maisel 1999).
Many systems are hybrids, with party o
cials and
the voters sharing power. The US presidential nom-
inating system stands as perhaps the most complex
means of choosing party candidates. Delegates elected
to national party conventions make the actual nom-
1. The Choice of Candidates
Decisions about who will run for oce involve
decisions by potential candidates as to whether or not
they will run and decisions by political parties, acting
under a variety of rules, concerning who their candi-
dates will be.
1.1 The Decision to Run
Individuals who run for political oce must exhibit
political ambition beyond the desire to simply serve
their community. Campaigning for and serving in
political oce involves personal costs and public
exposure that deter many from seeking oce. How-
ever, even those ambitious for elective oce must make
decisions about what oces to seek and when to seek
those oces (Jacobson and Kernell 1981). These
decisions are based on a complex, often-implicit cost-
benefit analysis. Potential candidates weigh the costs
and benefits of both winning and losing. Their decision
calculus can be summarized:
p
p
(
B
C
)
p
(
B
C
)
r
w
l
w
w
l
l
where
p
probability of running;
p
probability of
r
w
winning;
p
probability of losing;
B
benefit of
l
w
winning;
C
cost of winning;
B
benefit of losing;
w
l
C
cost of losing.
The di
culty that potential candidates face in
making these decisions relates to the uncertainty
regarding virtually all of the relevant variables (Maisel
l
1433
Campaigning: Political
inations. Some delegates are selected because of their
positions within the party or as elected o
cials with
party a
liations; others are chosen at party meetings
(or caucuses) held in local communities; still others are
chosen by the voters in primary elections, based on
lists of delegate candidates approved by the presi-
dential campaigns they seek to support (Polsby and
Wildavsky 2000, Wayne 2000).
specialize in media advertising, polling, direct mail,
telephone banks, fundraising, and virtually every other
aspect of campaigning; other firms are structured to
take over all or many aspects of a campaign. Some
work in particular geographic areas. Others are ready
to sell their services anywhere throughout an entire
nation or even internationally.
3. Defining Messages and Setting Strategy
A political campaign requires getting a message to
voters. To do that, a candidate must define what the
message is and must devise a strategy to reach the
voters who will be swayed by that message. Again,
how this is to be done varies with whether the
campaign in question is party-centered or candidate-
centered.
For party-centered campaigns, the message is the
party platform, and the voters to be reached at those in
the party’s core constituency and others to whom the
message might appeal in a particular year. Neither the
message nor the strategy vary much from year to year,
though marginal changes are made as the context
changes—and these marginal changes might well spell
the difference between victory and defeat.
For candidate-centered campaigns these decisions
are among the most crucial. Candidates must decide
whether to stick with the party message or to devise
one of their own; they must decide what particular
issues will play best to their voters; and they must
figure out who precisely those voters are. Campaign
strategists divide the electorate in a variety of
ways—by demographic characteristics, by economic
interests, by geographic locations. The goal is to use
research and polling data to determine what appeals
will work with which audiences and to have the
candidate address those audiences appropriately. For
incumbents seeking re-election, the strategic con-
siderationofteninvolvesrelyinguponstrengthsdemon-
strated earlier. For challengers, the strategy must be
to find an opponent’s weakness and to emphasize
one’s own strengths.
Effective campaign strategies remain more art than
science. Professional political campaigners know what
has worked in the past. But many strategists learn
those lessons. The successful strategists are those who
can see how to change the approach as the context
changes, as is often said in the military, ‘not to refight
the last war but to plan for the next one.’
2. The Campaign Organization
In polities with strong party systems, campaign organ-
izations tend to be party-centered. In these cases, the
voters cast their ballots for candidates as represent-
atives of a party. The party defines the campaign
message and communicates it to the voters. The party
raises the money to fund the campaign, does the
polling necessary to understand what the electorate is
thinking, and structures the campaign strategy and
tactics. Typically full-time, year-round party workers
perform all these functions or they are contracted out
to professionals.
Where strong party systems do not exist, individual
candidates must build their own organizations. In
these cases, candidates are most concerned with their
own elections, not that of those who share this party
label. Candidates choose campaign managers whose
sole goal is victory for that one candidate. They build
personal organizations to handle the important func-
tions of campaigning—doing research to develop issue
positions and to counter opponents’ proposals; polling
the public to ascertain perceptions of the candidate
and reactions to strategies; writing speeches; com-
municating with the press; developing and placing
advertisements; sending out direct mail appeals, tele-
phoning and leafleting the district; scheduling and
preparing for candidate appearances; and perhaps
most important, raising money. For larger and more
expensive campaigns, all of these functions are per-
formed by paid workers or volunteers in a candidate’s
own organization, or bought with campaign cont-
ributions.
Recent campaign organizations have been less able
to handle all of these functions, and two means of
coping have emerged. For campaigns in smaller
constituencies, the candidate often performs all of
these functions him or herself. They are done more or
less well depending on the skill of the candidate, the
time available, the level of competition, and the
sophistication that is expected in campaigning for
o
ce (Maisel 1999).
Even those with relatively high budget campaigns
often find that they cannot hire staff to perform all of
the necessary campaign activities. These candidates
frequently turn to paid political consultants, experts
from outside the campaign who specialize in certain
aspects of the campaign process and sell their services
to many campaigns during any election cycle. Firms
4. Communicating with the Voters
To be effective, a political campaign must be able to
send the message it has devised to the audience it has
targeted. Campaign messages are communicated in
two ways, through free media and paid media.
1434
Campbell, Donald Thomas (1916–96)
4.1 Free Media Co
erage
Political campaigns seek free media coverage in
whatever form they can, whenever they can. Candi-
dates want to be reported on in newspapers and
newsmagazines; they want to grant interviews on
radio; they want their events covered on television.
Free media coverage is advantageous not only because
there is no cost but also because the viewer sees news
coverage of a candidate as giving that candidate a
sense of legitimacy; the candidate is seen as part of the
day’s news, not as part of a paid advertisement.
However, free media is uncontrolled exposure. Camp-
aign strategists can try to structure what topics will be
covered—and campaigns are exploring increasingly
sophisticated means to do this—but too frequently the
message delivered is very different from the message
the candidate seeks to convey.
Bibliography
Ansolabehere S A, Iyengar S 1995
Going Negati
e: How Attack
Ads Shrink and Polarize the Electorate
. Free Press, New York
Hix S, Lord C 1997
Political Parties in the European Union
. St.
Martin’s Press, New York
Jacobson G C, Kernell S 1981
Strategy and Choice in Cong
-
ressional Elections
. Yale University Press, New Haven, CT
Kazee T A (ed.) 1994
Who Runs for Congress? Ambition,
Context, and Candidate Emergence
. Congressional Quarterly,
Washington, DC
Maisel L S 1986
From Obscurity to Obli
ion: Running in the
Congressional Primary
, rev. edn. University of Tennessee,
Knoxville, TN
Maisel L S 1999
Parties and Elections in America: The Electoral
Process
, 3rd edn. Rowman & Littlefield, Lanham, MD
Maisel L S, Stone W J 1997 Determinants of candidate emer-
gence in U. S. House elections: an exploratory study.
Legislati
e Studies Quarterly
22
: 79–96: Current work
from the Candidate Emergence Project can be found at
http:
socsci.colorado.cdu
CES
home.html
Polsby N W, Wildavsky A 2000
Presidential Elections: Strat-
egies and Structures of American Politics
, 10th edn. Chatham
House Publishers, New York
Wayne S J 2000
The Road to the White House 2000
. St. Martin’s
Press, New York
4.2 Paid Media Ad
ertising
The opposites apply to paid advertising. Paid medi-
a—whether in the form of television, radio, or print
advertisements, of Internet sites, or of direct mail—is
costly and is often viewed by the prospective voter as
the slanted message it in fact is. However, these paid
media also have the distinct advantage of allowing a
candidate to emphasize exactly the message desired.
These messages can be narrowcast, that is designed for
and directed to specific target audiences. These ads can
create images, discuss positions, or—increasingly in
recent years—attack an opponent, all in precisely the
way strategists feel will be most effective for a specified
audience.
The critical concerns about political campaigns
today relate to paid media. Many fear that effective
political messages cannot be conveyed in 30-second
advertisements, the medium preferred by advertising
executives. Others complain that the emphasis on
negative, attack advertising has poisoned the atmo-
sphere that surrounds political life, keeping some of
the best candidates from running (Ansolabehere and
Iyengar 1995). Virtually everyone bemoans the exor-
bitant costs of campaigns, costs spurred by reliance on
paid media. However, political campaigns will cont-
inue to rely on paid media so long as they are effective
in communicating a candidate’s message to the target
audience; for that is the means to electoral succ-
ess—the ultimate message of a campaign’s effect-
iveness.
L. S. Maisel
Campbell, Donald Thomas (1916–96)
Donald Thomas Campbell, born November 20, 1916
in Grass Lake, Michigan, died May 6, 1996 in
Bethlehem, Pennsylvania, thus ending a career marked
by an array of superlatives. Prolific scholar and author,
original thinker, ebullient teacher with a notable
twinkle in his eye, and generous colleague, Campbell
was widely regarded as the most important social
science methodologist of the twentieth Century. Sev-
eral of Campbell’s articles vie with one another as
among the most widely cited pieces of social science
scholarship. Repeatedly he coined phrases any one of
which most scholars would be proud to have con-
ceived, for example, ‘quasi-experiments,’ ‘unobtrusive
measures,’ ‘internal and external validity,’ and ‘plaus-
ible rival hypotheses.’ His concepts are incorporated
as fundamental in several fields—psychology, soci-
ology, anthropology, organization and management
sciences, public policy, evaluation, education, and
philosophy—and in common use by scholars unaware
of their inventor. Moreover, Campbell’s welcoming
and self-critical personal style, modeled on his schol-
arship, led him to embrace and examine all ‘plausible
rival hypotheses’ and endeared him to students and
colleagues alike.
Campbell received his B.A. (1939) and Ph.D. (1947),
both in Psychology, from the University of California
(Berkeley), and taught at several institutions, including
See also
: Advertising: Effects; Electoral Systems;
Electronic Democracy; Mass Media, Political Econ-
omy of; Media and Social Movements; Media Effects;
Media, Uses of; Party Identification; Political Mach-
ines; Political Money and Party Finance; Political
Parties; Polling; Primary Elections; Voting, Sociology
of; Women’s Suffrage
1435
Campbell, Donald Thomas (1916–96)
Ohio State University (1947–50), the University of
Chicago (1950–53), Syracuse University (as N.Y. State
Board of Regents Albert Schweitzer Professor, 1979–
82) and Lehigh University (as Distinguished Uni-
versity Professor of Sociology-Anthropology, Psy-
chology and Education, 1983–96), but the majority of
his scholarly years were spent as professor of psy-
chology at Northwestern University (1953–79) with
which Campbell’s name and reputation are ‘perman-
ently and inextricably co-identified’ (Campbell 1981).
Campbell’s honors include a Fulbright Visiting
Professorship in Social Psychology at Oxford (1968),
the Distinguished Scientific Contribution Award from
the American Psychological Association (1969), elec-
tion to the National Academy of Sciences and the
American Academy of Arts & Sciences (1973), the Kurt
Lewin Memorial Award from the Society for the
Psychological Study of Social Issues (1974), presidency
of the American Psychological Association (1975), the
Williams James Lectureship at Harvard (1977), the
Myrdal Prize in Science (1977), and the Distinguished
Contribution Award from the American Educational
Research Association (1981). In addition, two annual
awards are given in his name: The Donald Campbell
Award for Significant Research in Social Psychology
(1982–) and the Donald Campbell Award from the
Policy Studies Organization, to an ‘outstanding meth-
odological innovator in public policy studies’ (1983–).
Campbell received honorary degrees from numerous
universities, including Oslo, Michigan, Chicago, and
Southern California.
But Campbell felt most honored by the vast array of
books dedicated to him, a list that grew steadily in
several disciplines before and after his death.
tion of the physical world and mechanisms for
compensating for these errors. (Campbell later de-
clared profound ‘ambivalence’ toward the use of
disguised measures; see Kidder and Campbell 1970.)
Campbell explored the sources and loci of bias and
developed and refined methods for illustrating and for
minimizing these biases. For example in Jacobs and
Campbell (1961) he demonstrated how consensus in
the ‘reality’ of an arbitrarily invented but shared group
norm could persist over generations of experimental
subjects. In a series of crosscultural explorations
Campbell and co-workers (e.g., Segall et al. 1966)
explored sources of validity and invalidity in percep-
tions of ‘in-groups’ and ‘out-groups,’ and the differ-
ential susceptibility to perceptual illusions between
European and non-European cultures.
Campbell and Fiske (1959) published ‘Convergent
and Discriminant Validation by the Multitrait, Multi-
method Matrix,’ a complex and frequently cited
analysis of the need for multiple measures of under-
lying constructs (traits) and of the need for the
demonstrated capacity of methods to distinguish
among traits if one is to minimize irrelevant measure-
ment artifacts. Since every measure is partially invalid,
multiple and distinctive methods are called for to yield
a ‘heterogeneity of irrelevancies.’ Campbell and Fiske
explored the value of simultaneously deploying maxi-
mally different measures of an underlying trait (in
pursuit of convergent validation) and of different
traits measured the same way (in pursuit of discrimi-
nant validation). Only by demonstrating that an
underlying trait can be measured in various distinct
ways and that the chosen measurement strategies
distinguish among measured traits, can one minimize
the impact of misleading artifacts.
By the early 1960s Campbell had established an
enviable reputation for the vigor and reach of his
campaign for advancing the methods and theory
associated with issues of validity. His reputation was
further advanced with the publication with (Campbell
and Stanley 1963) of ‘Experimental and Quasi-exper-
imental Designs for Research,’ a widely circulated work
(elaborated as Cook and Campbell 1979) in which
they popularized the terms ‘internal’ and ‘external’
validity:
1. Major Contributions
When he died, Campbell’s resume
listed more than 230
published books, monographs, and articles. Any brief
discussion of his contributions must necessarily focus
on a few. Best known as a methodologist, with
considerable scholarly work in the areas of exper-
imental design, measurement, and social experimen-
tation, Campbell is perhaps remembered especially for
his explorations of the concept of validity. He was also
a well-regarded epistemologist with a keen interest in
the sociology of science. To each domain, Campbell
brought his unique mind to bear on the problem of
knowledge production. In a remarkable array of
works, Campbell explored from several perspectives
the inevitable fallibilities inherent in both observers
and methods in accurately portraying the world.
In his earliest publication, ‘The Indirect Assessment
of Social Attitudes’ (1950), Campbell evidenced an
aspect of the interests that would become the overall
focus of his scholarly life: the imperfections introduced
by human observers and methods, including the
scientific method, in the search for veridical descrip-
Internal validity is the basic minimum without which any
experiment is uninterpretable: did in fact the experimental
treatment make a difference in this specific experimental
instance? External validity asks the question of generalizabi-
lity: To what populations, setting, treatment variables, and
measurement
variables
can
this
effect
be
generalized?
(Campbell and Stanley 1963).
Another widely cited treatise,
Unobtrusi
e Mea-
sures: Non-reacti
e Research in the Social Sciences
(Webb et al. 1966) evolved during affable lunchtime
conversations with colleagues from different depart-
ments. Over several years these colleagues sought to
identify relatively unbiased, nonreactive ways of mea-
1436
Campbell, Donald Thomas (1916–96)
suring behavior. A frequently cited example describes
ways a Chicago museum tried to identify the popu-
larity of its exhibits. ‘Obtrusive measures’ included
questioning visitors leaving the museum; an ‘unob-
trusive measure’ was the frequency with which mu-
seum staff had to replace worn tiles in front of exhibits.
By that index, the exhibit of baby chicks hatching live
and wet from quivering eggs was by far the most
popular.
Campbell’s participation in an interdisciplinary
conference spawned his brilliant foray into the so-
ciology of science entitled ‘Ethnocentrism of Disci-
plines: A Fishscale Model of Omniscience’ (1969a). On
reading a paper ‘in an area of high relevance’ to his
own work by esteemed fellow social psychologist
William McGuire, Campbell realized and then ack-
nowledged that he had not ‘read or read-at even half of
McGuire’s citations, and was not at all aware of the
existence of another sizeable proportion’ (Campbell
1969a). While many scholars might react with chagrin,
Campbell instead reflected on how the ‘highly ar-
bitrary’ organization of universities and academic
fields skewed knowledge production:
Thus anthropology is a hodgepodge of all novelties venturing
into exotic lands—a hodgepodge of skin color, physical
stature, agricultural practices, weapons, religious beliefs,
kinship systems, history, archeology, and paleontology …
Thus psychology is a hodgepodge of sensitive subjective
biography, or brain operations, or school achievement
testing, of factor analysis, of Markov process mathematics, of
schizophrenic families, of laboratory experiments on group
structure in which persons are anonymous … Thus econo-
mics is a hodgepodge of mathematics without data, of history
of economic institutions without mathematics or theory, of
an ideal model of psychological man … (Campbell 1969a)
Campbell took a breathtaking leap when he con-
ceived the ‘myth of unidisciplinary competence,’ dem-
onstrating how social, administrative, structural, and
political supports which accrue over time around
arbitrary disciplines ultimately create incentives that
reinforce these arrangements and punish others of
equal epistemological worth. Campbell detailed the
inevitable consequence—biased knowledge produc-
tion—and offered a model of ‘interdisciplinary
narrowness’ for mitigating this ‘ethnocentrism of
disciplines.’
By the late 1960s, following Vietnam War protests,
campus unrest, and a national focus on urban inequi-
ties, Campbell began trying conscientiously to in-
fluence the design, implementation, and evaluation of
social policies. In ‘Reforms as Experiments’ (1969b)
Campbell introduced a range of real-world quasi-
experiments (attempts to approach the standards of
experimentation even where random assignment of
subjects to conditions is unachievable)—each tested
against standards of ‘internal and external vali-
dity’—to the practical world of social reform. That
article frequently is cited as the most important single
work in the field now known as ‘program evaluation.’
In a brilliant tongue-in-cheek section of this article
Campbell advised ‘trapped administrators whose pol-
itical predicament will not allow the risk of failure’
how to capitalize on ‘threats to validity’ to assure
positive results (Campbell 1969b). For example, were
they to accept the occasional ‘grateful testimonial’ as if
it were a representative outcome, to reserve interven-
tions for carefully selected subpopulations most likely
to succeed, and to eliminate from their analyses all
those who prematurely quit the program, ‘successful’
program outcomes could be virtually assured.
Campbell continued in various fora to insist that
social scientists have both obligations and opportun-
ities to test the relevance of their theories and methods
in the service of public good. In this, Campbell’s
unswerving concern was with the biases affecting the
generation of knowledge. For example, in his too-
infrequently cited ‘Assessing the Impact of Planned
Social Change’ (1975) Campbell again displayed re-
markable methodological and sociological insight in
generalizing the problem of data validity in the policy
arena in the following way:
The more any quantitative social indicator is used for social
decision-making, the more subject it will be to corruption
pressures, and the more apt it will be to distort and corrupt
the social processes it is intended to monitor (Campbell 1975).
Corrupted indicators remain a persistent and only
vaguely explored source of real-world invalidity (e.g.,
Cochran et al. 1980).
Another highly influential, widely circulated piece
went unpublished for nearly 20 years: in ‘Methods for
theExperimentingSociety’Campbellattemptedspecifi-
cally to design a strategy for a society willing to test
innovations and their intended effects. He argues for a
society that would
vigorously try out possible solutions to recurrent problems
and would make hard-headed, multidimensional evaluations
of outcomes, and when the evaluation of one reform showed
it to have been ineffective or harmful, would move on to try
other alternatives (Campbell, in Overman 1988).
Surveying the multitude of societal reforms under-
way at the time, many of which he endorsed (e.g.,
Alexander
Dubcek
in
Czechoslovakia,
Salvador
Allende, in Chile), Campbell argued:
There is no such (experimenting society, committed to reality
testing, self-criticism, to avoiding self-deception) anywhere
today. While all nations are engaged in trying out innovative
reforms, none of them are yet organized to adequately
evaluate the outcomes of these reforms (Overman 1988).
After specifying requirements of such an ‘experiment-
ing society’ Campbell attempted in various ways to
determine conditions under which these results might
be achievable. For example, after Allende was assas-
sinated in Chile, Campbell invited Chilean social
scientist Ricardo Zun
iga to Northwestern University
1437
[ Pobierz całość w formacie PDF ]
zanotowane.pl doc.pisz.pl pdf.pisz.pl upanicza.keep.pl
C
Campaigning: Political
1986, Kazee 1994). Generally the rather obvious
benefits far outweigh the costs of winning for those
with the requisite political ambition; but costs must be
weighed—loss of privacy, loss of time with family,
frequently loss of income. Similarly, the evident costs
of losing are almost always greater than the benefits;
but some benefits do accrue to losers—making a name
for oneself, building up good will with party o
cials.
Beyond these broad generalizations, little is known
about how specific candidates make individual de-
cisions given the dazzling array of di
cult-to-estimate
variables (Maisel and Stone 1997).
At the microlevel, political campaigning involves an
individual candidate attempting to win political o
ce.
At the macrolevel, political parties seek to gain seats in
government for their partisans and ultimately to gain
power or at least to share power in the relevant
governmental arena. At each level decisions must be
made about candidacies, organizations must be built,
messages must be refined, strategies must be set, and
the public must be convinced to support those waging
a particular campaign. Only certain aspects of political
campaigns are viewed by the public. Frequently, the
actions taken away from public view are more de-
terminative of success and failure.
1.2 Parties’ Choices of Candidates
The ways in which political parties choose their
candidates for o
ce varies significantly from political
system to political system and from party to party. The
key variable in this aspect of the process relates to the
extent to which party organization controls places on
the ballot contrasted with voters acting independently.
In the former, the process is often described as
candidate recruitment, in the latter, candidate emerg-
ence.
At one extreme are those systems in which political
parties slate their candidates for o
ce, ranking cand-
idates or placing them in safe constituencies. Most
parties operating in electoral systems with propor-
tional representation follow this procedure. It is also
followed in systems with strong party systems, in
which citizens vote more for parties than they do for
individual candidates. Many European systems fit this
description (Hix and Lord 1997).
At the other extreme are systems, of which the
United States is prototypical, in which voters choose
party candidates in primary elections. The shape of the
electorate in primary elections varies considerably
from polity to polity. In theory, primary elections are
open only to members of one political party or
another, but in actual practice how membership is
defined varies according to local rules and can be quite
restrictive or totally open. In either case, voters in
the primary elections determine the party nominee,
even if that nominee would not have been the choice of
party o
cials (Maisel 1999).
Many systems are hybrids, with party o
cials and
the voters sharing power. The US presidential nom-
inating system stands as perhaps the most complex
means of choosing party candidates. Delegates elected
to national party conventions make the actual nom-
1. The Choice of Candidates
Decisions about who will run for oce involve
decisions by potential candidates as to whether or not
they will run and decisions by political parties, acting
under a variety of rules, concerning who their candi-
dates will be.
1.1 The Decision to Run
Individuals who run for political oce must exhibit
political ambition beyond the desire to simply serve
their community. Campaigning for and serving in
political oce involves personal costs and public
exposure that deter many from seeking oce. How-
ever, even those ambitious for elective oce must make
decisions about what oces to seek and when to seek
those oces (Jacobson and Kernell 1981). These
decisions are based on a complex, often-implicit cost-
benefit analysis. Potential candidates weigh the costs
and benefits of both winning and losing. Their decision
calculus can be summarized:
p
p
(
B
C
)
p
(
B
C
)
r
w
l
w
w
l
l
where
p
probability of running;
p
probability of
r
w
winning;
p
probability of losing;
B
benefit of
l
w
winning;
C
cost of winning;
B
benefit of losing;
w
l
C
cost of losing.
The di
culty that potential candidates face in
making these decisions relates to the uncertainty
regarding virtually all of the relevant variables (Maisel
l
1433
Campaigning: Political
inations. Some delegates are selected because of their
positions within the party or as elected o
cials with
party a
liations; others are chosen at party meetings
(or caucuses) held in local communities; still others are
chosen by the voters in primary elections, based on
lists of delegate candidates approved by the presi-
dential campaigns they seek to support (Polsby and
Wildavsky 2000, Wayne 2000).
specialize in media advertising, polling, direct mail,
telephone banks, fundraising, and virtually every other
aspect of campaigning; other firms are structured to
take over all or many aspects of a campaign. Some
work in particular geographic areas. Others are ready
to sell their services anywhere throughout an entire
nation or even internationally.
3. Defining Messages and Setting Strategy
A political campaign requires getting a message to
voters. To do that, a candidate must define what the
message is and must devise a strategy to reach the
voters who will be swayed by that message. Again,
how this is to be done varies with whether the
campaign in question is party-centered or candidate-
centered.
For party-centered campaigns, the message is the
party platform, and the voters to be reached at those in
the party’s core constituency and others to whom the
message might appeal in a particular year. Neither the
message nor the strategy vary much from year to year,
though marginal changes are made as the context
changes—and these marginal changes might well spell
the difference between victory and defeat.
For candidate-centered campaigns these decisions
are among the most crucial. Candidates must decide
whether to stick with the party message or to devise
one of their own; they must decide what particular
issues will play best to their voters; and they must
figure out who precisely those voters are. Campaign
strategists divide the electorate in a variety of
ways—by demographic characteristics, by economic
interests, by geographic locations. The goal is to use
research and polling data to determine what appeals
will work with which audiences and to have the
candidate address those audiences appropriately. For
incumbents seeking re-election, the strategic con-
siderationofteninvolvesrelyinguponstrengthsdemon-
strated earlier. For challengers, the strategy must be
to find an opponent’s weakness and to emphasize
one’s own strengths.
Effective campaign strategies remain more art than
science. Professional political campaigners know what
has worked in the past. But many strategists learn
those lessons. The successful strategists are those who
can see how to change the approach as the context
changes, as is often said in the military, ‘not to refight
the last war but to plan for the next one.’
2. The Campaign Organization
In polities with strong party systems, campaign organ-
izations tend to be party-centered. In these cases, the
voters cast their ballots for candidates as represent-
atives of a party. The party defines the campaign
message and communicates it to the voters. The party
raises the money to fund the campaign, does the
polling necessary to understand what the electorate is
thinking, and structures the campaign strategy and
tactics. Typically full-time, year-round party workers
perform all these functions or they are contracted out
to professionals.
Where strong party systems do not exist, individual
candidates must build their own organizations. In
these cases, candidates are most concerned with their
own elections, not that of those who share this party
label. Candidates choose campaign managers whose
sole goal is victory for that one candidate. They build
personal organizations to handle the important func-
tions of campaigning—doing research to develop issue
positions and to counter opponents’ proposals; polling
the public to ascertain perceptions of the candidate
and reactions to strategies; writing speeches; com-
municating with the press; developing and placing
advertisements; sending out direct mail appeals, tele-
phoning and leafleting the district; scheduling and
preparing for candidate appearances; and perhaps
most important, raising money. For larger and more
expensive campaigns, all of these functions are per-
formed by paid workers or volunteers in a candidate’s
own organization, or bought with campaign cont-
ributions.
Recent campaign organizations have been less able
to handle all of these functions, and two means of
coping have emerged. For campaigns in smaller
constituencies, the candidate often performs all of
these functions him or herself. They are done more or
less well depending on the skill of the candidate, the
time available, the level of competition, and the
sophistication that is expected in campaigning for
o
ce (Maisel 1999).
Even those with relatively high budget campaigns
often find that they cannot hire staff to perform all of
the necessary campaign activities. These candidates
frequently turn to paid political consultants, experts
from outside the campaign who specialize in certain
aspects of the campaign process and sell their services
to many campaigns during any election cycle. Firms
4. Communicating with the Voters
To be effective, a political campaign must be able to
send the message it has devised to the audience it has
targeted. Campaign messages are communicated in
two ways, through free media and paid media.
1434
Campbell, Donald Thomas (1916–96)
4.1 Free Media Co
erage
Political campaigns seek free media coverage in
whatever form they can, whenever they can. Candi-
dates want to be reported on in newspapers and
newsmagazines; they want to grant interviews on
radio; they want their events covered on television.
Free media coverage is advantageous not only because
there is no cost but also because the viewer sees news
coverage of a candidate as giving that candidate a
sense of legitimacy; the candidate is seen as part of the
day’s news, not as part of a paid advertisement.
However, free media is uncontrolled exposure. Camp-
aign strategists can try to structure what topics will be
covered—and campaigns are exploring increasingly
sophisticated means to do this—but too frequently the
message delivered is very different from the message
the candidate seeks to convey.
Bibliography
Ansolabehere S A, Iyengar S 1995
Going Negati
e: How Attack
Ads Shrink and Polarize the Electorate
. Free Press, New York
Hix S, Lord C 1997
Political Parties in the European Union
. St.
Martin’s Press, New York
Jacobson G C, Kernell S 1981
Strategy and Choice in Cong
-
ressional Elections
. Yale University Press, New Haven, CT
Kazee T A (ed.) 1994
Who Runs for Congress? Ambition,
Context, and Candidate Emergence
. Congressional Quarterly,
Washington, DC
Maisel L S 1986
From Obscurity to Obli
ion: Running in the
Congressional Primary
, rev. edn. University of Tennessee,
Knoxville, TN
Maisel L S 1999
Parties and Elections in America: The Electoral
Process
, 3rd edn. Rowman & Littlefield, Lanham, MD
Maisel L S, Stone W J 1997 Determinants of candidate emer-
gence in U. S. House elections: an exploratory study.
Legislati
e Studies Quarterly
22
: 79–96: Current work
from the Candidate Emergence Project can be found at
http:
socsci.colorado.cdu
CES
home.html
Polsby N W, Wildavsky A 2000
Presidential Elections: Strat-
egies and Structures of American Politics
, 10th edn. Chatham
House Publishers, New York
Wayne S J 2000
The Road to the White House 2000
. St. Martin’s
Press, New York
4.2 Paid Media Ad
ertising
The opposites apply to paid advertising. Paid medi-
a—whether in the form of television, radio, or print
advertisements, of Internet sites, or of direct mail—is
costly and is often viewed by the prospective voter as
the slanted message it in fact is. However, these paid
media also have the distinct advantage of allowing a
candidate to emphasize exactly the message desired.
These messages can be narrowcast, that is designed for
and directed to specific target audiences. These ads can
create images, discuss positions, or—increasingly in
recent years—attack an opponent, all in precisely the
way strategists feel will be most effective for a specified
audience.
The critical concerns about political campaigns
today relate to paid media. Many fear that effective
political messages cannot be conveyed in 30-second
advertisements, the medium preferred by advertising
executives. Others complain that the emphasis on
negative, attack advertising has poisoned the atmo-
sphere that surrounds political life, keeping some of
the best candidates from running (Ansolabehere and
Iyengar 1995). Virtually everyone bemoans the exor-
bitant costs of campaigns, costs spurred by reliance on
paid media. However, political campaigns will cont-
inue to rely on paid media so long as they are effective
in communicating a candidate’s message to the target
audience; for that is the means to electoral succ-
ess—the ultimate message of a campaign’s effect-
iveness.
L. S. Maisel
Campbell, Donald Thomas (1916–96)
Donald Thomas Campbell, born November 20, 1916
in Grass Lake, Michigan, died May 6, 1996 in
Bethlehem, Pennsylvania, thus ending a career marked
by an array of superlatives. Prolific scholar and author,
original thinker, ebullient teacher with a notable
twinkle in his eye, and generous colleague, Campbell
was widely regarded as the most important social
science methodologist of the twentieth Century. Sev-
eral of Campbell’s articles vie with one another as
among the most widely cited pieces of social science
scholarship. Repeatedly he coined phrases any one of
which most scholars would be proud to have con-
ceived, for example, ‘quasi-experiments,’ ‘unobtrusive
measures,’ ‘internal and external validity,’ and ‘plaus-
ible rival hypotheses.’ His concepts are incorporated
as fundamental in several fields—psychology, soci-
ology, anthropology, organization and management
sciences, public policy, evaluation, education, and
philosophy—and in common use by scholars unaware
of their inventor. Moreover, Campbell’s welcoming
and self-critical personal style, modeled on his schol-
arship, led him to embrace and examine all ‘plausible
rival hypotheses’ and endeared him to students and
colleagues alike.
Campbell received his B.A. (1939) and Ph.D. (1947),
both in Psychology, from the University of California
(Berkeley), and taught at several institutions, including
See also
: Advertising: Effects; Electoral Systems;
Electronic Democracy; Mass Media, Political Econ-
omy of; Media and Social Movements; Media Effects;
Media, Uses of; Party Identification; Political Mach-
ines; Political Money and Party Finance; Political
Parties; Polling; Primary Elections; Voting, Sociology
of; Women’s Suffrage
1435
Campbell, Donald Thomas (1916–96)
Ohio State University (1947–50), the University of
Chicago (1950–53), Syracuse University (as N.Y. State
Board of Regents Albert Schweitzer Professor, 1979–
82) and Lehigh University (as Distinguished Uni-
versity Professor of Sociology-Anthropology, Psy-
chology and Education, 1983–96), but the majority of
his scholarly years were spent as professor of psy-
chology at Northwestern University (1953–79) with
which Campbell’s name and reputation are ‘perman-
ently and inextricably co-identified’ (Campbell 1981).
Campbell’s honors include a Fulbright Visiting
Professorship in Social Psychology at Oxford (1968),
the Distinguished Scientific Contribution Award from
the American Psychological Association (1969), elec-
tion to the National Academy of Sciences and the
American Academy of Arts & Sciences (1973), the Kurt
Lewin Memorial Award from the Society for the
Psychological Study of Social Issues (1974), presidency
of the American Psychological Association (1975), the
Williams James Lectureship at Harvard (1977), the
Myrdal Prize in Science (1977), and the Distinguished
Contribution Award from the American Educational
Research Association (1981). In addition, two annual
awards are given in his name: The Donald Campbell
Award for Significant Research in Social Psychology
(1982–) and the Donald Campbell Award from the
Policy Studies Organization, to an ‘outstanding meth-
odological innovator in public policy studies’ (1983–).
Campbell received honorary degrees from numerous
universities, including Oslo, Michigan, Chicago, and
Southern California.
But Campbell felt most honored by the vast array of
books dedicated to him, a list that grew steadily in
several disciplines before and after his death.
tion of the physical world and mechanisms for
compensating for these errors. (Campbell later de-
clared profound ‘ambivalence’ toward the use of
disguised measures; see Kidder and Campbell 1970.)
Campbell explored the sources and loci of bias and
developed and refined methods for illustrating and for
minimizing these biases. For example in Jacobs and
Campbell (1961) he demonstrated how consensus in
the ‘reality’ of an arbitrarily invented but shared group
norm could persist over generations of experimental
subjects. In a series of crosscultural explorations
Campbell and co-workers (e.g., Segall et al. 1966)
explored sources of validity and invalidity in percep-
tions of ‘in-groups’ and ‘out-groups,’ and the differ-
ential susceptibility to perceptual illusions between
European and non-European cultures.
Campbell and Fiske (1959) published ‘Convergent
and Discriminant Validation by the Multitrait, Multi-
method Matrix,’ a complex and frequently cited
analysis of the need for multiple measures of under-
lying constructs (traits) and of the need for the
demonstrated capacity of methods to distinguish
among traits if one is to minimize irrelevant measure-
ment artifacts. Since every measure is partially invalid,
multiple and distinctive methods are called for to yield
a ‘heterogeneity of irrelevancies.’ Campbell and Fiske
explored the value of simultaneously deploying maxi-
mally different measures of an underlying trait (in
pursuit of convergent validation) and of different
traits measured the same way (in pursuit of discrimi-
nant validation). Only by demonstrating that an
underlying trait can be measured in various distinct
ways and that the chosen measurement strategies
distinguish among measured traits, can one minimize
the impact of misleading artifacts.
By the early 1960s Campbell had established an
enviable reputation for the vigor and reach of his
campaign for advancing the methods and theory
associated with issues of validity. His reputation was
further advanced with the publication with (Campbell
and Stanley 1963) of ‘Experimental and Quasi-exper-
imental Designs for Research,’ a widely circulated work
(elaborated as Cook and Campbell 1979) in which
they popularized the terms ‘internal’ and ‘external’
validity:
1. Major Contributions
When he died, Campbell’s resume
listed more than 230
published books, monographs, and articles. Any brief
discussion of his contributions must necessarily focus
on a few. Best known as a methodologist, with
considerable scholarly work in the areas of exper-
imental design, measurement, and social experimen-
tation, Campbell is perhaps remembered especially for
his explorations of the concept of validity. He was also
a well-regarded epistemologist with a keen interest in
the sociology of science. To each domain, Campbell
brought his unique mind to bear on the problem of
knowledge production. In a remarkable array of
works, Campbell explored from several perspectives
the inevitable fallibilities inherent in both observers
and methods in accurately portraying the world.
In his earliest publication, ‘The Indirect Assessment
of Social Attitudes’ (1950), Campbell evidenced an
aspect of the interests that would become the overall
focus of his scholarly life: the imperfections introduced
by human observers and methods, including the
scientific method, in the search for veridical descrip-
Internal validity is the basic minimum without which any
experiment is uninterpretable: did in fact the experimental
treatment make a difference in this specific experimental
instance? External validity asks the question of generalizabi-
lity: To what populations, setting, treatment variables, and
measurement
variables
can
this
effect
be
generalized?
(Campbell and Stanley 1963).
Another widely cited treatise,
Unobtrusi
e Mea-
sures: Non-reacti
e Research in the Social Sciences
(Webb et al. 1966) evolved during affable lunchtime
conversations with colleagues from different depart-
ments. Over several years these colleagues sought to
identify relatively unbiased, nonreactive ways of mea-
1436
Campbell, Donald Thomas (1916–96)
suring behavior. A frequently cited example describes
ways a Chicago museum tried to identify the popu-
larity of its exhibits. ‘Obtrusive measures’ included
questioning visitors leaving the museum; an ‘unob-
trusive measure’ was the frequency with which mu-
seum staff had to replace worn tiles in front of exhibits.
By that index, the exhibit of baby chicks hatching live
and wet from quivering eggs was by far the most
popular.
Campbell’s participation in an interdisciplinary
conference spawned his brilliant foray into the so-
ciology of science entitled ‘Ethnocentrism of Disci-
plines: A Fishscale Model of Omniscience’ (1969a). On
reading a paper ‘in an area of high relevance’ to his
own work by esteemed fellow social psychologist
William McGuire, Campbell realized and then ack-
nowledged that he had not ‘read or read-at even half of
McGuire’s citations, and was not at all aware of the
existence of another sizeable proportion’ (Campbell
1969a). While many scholars might react with chagrin,
Campbell instead reflected on how the ‘highly ar-
bitrary’ organization of universities and academic
fields skewed knowledge production:
Thus anthropology is a hodgepodge of all novelties venturing
into exotic lands—a hodgepodge of skin color, physical
stature, agricultural practices, weapons, religious beliefs,
kinship systems, history, archeology, and paleontology …
Thus psychology is a hodgepodge of sensitive subjective
biography, or brain operations, or school achievement
testing, of factor analysis, of Markov process mathematics, of
schizophrenic families, of laboratory experiments on group
structure in which persons are anonymous … Thus econo-
mics is a hodgepodge of mathematics without data, of history
of economic institutions without mathematics or theory, of
an ideal model of psychological man … (Campbell 1969a)
Campbell took a breathtaking leap when he con-
ceived the ‘myth of unidisciplinary competence,’ dem-
onstrating how social, administrative, structural, and
political supports which accrue over time around
arbitrary disciplines ultimately create incentives that
reinforce these arrangements and punish others of
equal epistemological worth. Campbell detailed the
inevitable consequence—biased knowledge produc-
tion—and offered a model of ‘interdisciplinary
narrowness’ for mitigating this ‘ethnocentrism of
disciplines.’
By the late 1960s, following Vietnam War protests,
campus unrest, and a national focus on urban inequi-
ties, Campbell began trying conscientiously to in-
fluence the design, implementation, and evaluation of
social policies. In ‘Reforms as Experiments’ (1969b)
Campbell introduced a range of real-world quasi-
experiments (attempts to approach the standards of
experimentation even where random assignment of
subjects to conditions is unachievable)—each tested
against standards of ‘internal and external vali-
dity’—to the practical world of social reform. That
article frequently is cited as the most important single
work in the field now known as ‘program evaluation.’
In a brilliant tongue-in-cheek section of this article
Campbell advised ‘trapped administrators whose pol-
itical predicament will not allow the risk of failure’
how to capitalize on ‘threats to validity’ to assure
positive results (Campbell 1969b). For example, were
they to accept the occasional ‘grateful testimonial’ as if
it were a representative outcome, to reserve interven-
tions for carefully selected subpopulations most likely
to succeed, and to eliminate from their analyses all
those who prematurely quit the program, ‘successful’
program outcomes could be virtually assured.
Campbell continued in various fora to insist that
social scientists have both obligations and opportun-
ities to test the relevance of their theories and methods
in the service of public good. In this, Campbell’s
unswerving concern was with the biases affecting the
generation of knowledge. For example, in his too-
infrequently cited ‘Assessing the Impact of Planned
Social Change’ (1975) Campbell again displayed re-
markable methodological and sociological insight in
generalizing the problem of data validity in the policy
arena in the following way:
The more any quantitative social indicator is used for social
decision-making, the more subject it will be to corruption
pressures, and the more apt it will be to distort and corrupt
the social processes it is intended to monitor (Campbell 1975).
Corrupted indicators remain a persistent and only
vaguely explored source of real-world invalidity (e.g.,
Cochran et al. 1980).
Another highly influential, widely circulated piece
went unpublished for nearly 20 years: in ‘Methods for
theExperimentingSociety’Campbellattemptedspecifi-
cally to design a strategy for a society willing to test
innovations and their intended effects. He argues for a
society that would
vigorously try out possible solutions to recurrent problems
and would make hard-headed, multidimensional evaluations
of outcomes, and when the evaluation of one reform showed
it to have been ineffective or harmful, would move on to try
other alternatives (Campbell, in Overman 1988).
Surveying the multitude of societal reforms under-
way at the time, many of which he endorsed (e.g.,
Alexander
Dubcek
in
Czechoslovakia,
Salvador
Allende, in Chile), Campbell argued:
There is no such (experimenting society, committed to reality
testing, self-criticism, to avoiding self-deception) anywhere
today. While all nations are engaged in trying out innovative
reforms, none of them are yet organized to adequately
evaluate the outcomes of these reforms (Overman 1988).
After specifying requirements of such an ‘experiment-
ing society’ Campbell attempted in various ways to
determine conditions under which these results might
be achievable. For example, after Allende was assas-
sinated in Chile, Campbell invited Chilean social
scientist Ricardo Zun
iga to Northwestern University
1437
[ Pobierz całość w formacie PDF ]