|
The editors have not reproduced the figures that represent the author’s findings.
Introduction
Since the presidency of John F. Kennedy, there has been
a decline in American confidence in the federal government
(Walker, 2000). The importance of confidence in the American
federal government is immense. Political participation can be
defined as “Any activity that attempts to influence public policy
or the selection of government officials” (Austin Community
College). One hopes that Americans do not lack confidence in
the federal government because they dislike this style of government.
Democracy for many years has been trumpeted and
hailed by many as the best way a government can truly reflect,
respect and represent her citizens.
Having a great deal of confidence and participating in
governance can create a more patriotic spirit in citizens, which
in turn might lead to more citizens participating politically,
thereby making the country more democratic. It can be a propelling
force behind a nation’s success, affecting all aspects of
her citizens’ lives: education, business, government, families,
and community organizations (McLagan & Nel, 1997). Since
the 1970’s, the level of confidence in the federal government
of America has drastically decreased. To illustrate the decline
in confidence, the Cable News Network (CNN) Broadcasting
Company recently conducted polls on confidence of the
American public. Among their results were that “4 out of 5
[American adults] lack confidence in government” (Kyoko,
1997). This is approximately 153,600,000 million Americans!
According to the CNN polls, “when asked to name two or
three government successes of the past 30 years, 42 percent
of the respondents could not even name one” (Kyoko, 1997).
To account for this decline in confidence, the first research
question of this study is to see “What proportion of American
adults have no confidence in the federal government from
1973 to 1998?”
Many believe that “perhaps the most visible indication of
that disengagement is the steady decline in voter turnout”
(Berresford, 2000). The presumption is that, if people do not
have confidence in the federal government, then most likely
they would not vote in the presidential election. According to
Susan V. Berresford, the President of the Ford Foundation,
voter turnout is an important indicator of declining confidence
in the federal government. Therefore the study’s second research
question asks, “Is voter turnout decreasing? If so, is this
associated with the decrease in confidence?” One short-term
study perceives that voter turnout has significantly decreased
(Berresford, 2000).
The last of the study’s research questions is to find out
“Demographically, who has no confidence in the federal
government?” Demographic categories investigated include
religion, race, education, sex, and age. Sociologists believe that
those who are religiously involved and have liberal religious beliefs
have a higher level of political participation and therefore
a higher confidence in federal government (Brady, 1989). More
specifically, “the closer a person is to a church, the more likely
they are to support the political authorities” (Hoffman, 1985).
Therefore, this study also compares the confidence level in the
federal government among those who are religiously involved,
specifically, Christians, and those who are not. Also, regarding
race, previous studies suggest that, contrary to what fellow
citizens might think, “There are no interracial differences in
sociopolitical participation after controlling for differences in
socioeconomic status” (Bobo & Gilliam, 1988). Furthermore,
these analyses were done to solidify a concept suggested by
Reverend Jesse Jackson. According to him, “It’s not a matter of
race or color of face, but one of moral deficiencies.”
When considering the relationship between confidence in
the federal government and education, it is generally believed
that the more educated one is, the less confidence he or she
has in the federal government. As one obtains a higher degree
of education, he or she becomes more liberal in thinking.
Furthermore, due to education, one learns about the Electoral
College and that when a vote is cast at the presidential elections,
the voter does not really vote for the candidate of his
or her choice but for the Electoral College. Sociologists have
found that there is very little difference in the level of political
participation among the sexes, hence little difference in their
confidence levels (Beckwith, 1986). Finally, when examining
different age groups, sociologists have found out that among
young adults (i.e., among those that are 18-20 and 21-24 years
old), voting turnout out is fearfully low. Sociologists believe
that the decline in confidence in the federal government could
be a result of many demographic factors, including education,
age, income and ethnicity (Austin Community College).
All these factors are very important and ought to be considered
when analyzing an illustration of political participation
such as voting, which, according to Susan V. Berresford, is one
of the most important indicators of confidence in the federal
government.
Purpose
Since most of the studies done on the relationship of
public confidence and voter turnout have been short-term
studies, this study re-examines this relationship in context
of the trend over an extended period of time. This is a
longitudinal study from 1972 to 1998 with a purpose that is
threefold: (a) To see what proportions of Americans have
no confidence in the federal government, (b) to scrutinize
the assumption or commonly held belief that there is a
relationship between the level of confidence in the federal
government and the ebbing voter turnout for presidential
elections, and to account for this trend if any relationship is
found, and (c) to search out, demographically, which group of
people the government ought to try harder to reach out to, so
as to entice them to have confidence and vote, thereby making
America a better democratic nation.
To implement these questions, the following research
hypotheses were used: (a) The proportion of Americans that
have no confidence in the federal government from 1973 to
1998 is significantly greater than the proportion of Americans
that have confidence in the federal government, (b) lack
of confidence in the federal government is associated with
voter turnout, and (c) the population who are Christians are
more likely to have a great deal of confidence in the federal
government than the population of Americans who are, for
the purposes of this study, considered atheists.
Research Methodology
Sample
The data for this study were collected from 1972 to 1992
as part of the General Social Survey. The National Opinion
Research Center (NORC), which since 1972 has surveyed more
than 35,000 respondents with over 2,500 different questions,
conducted this survey. Each of the respondents for this survey
was independently drawn using some sample parameters.
These parameters were that subjects were English-speaking,
citizens of the United States of America, at least 18 years old
and non-institutionalized. For each of the variables in the
study’s research questions, respondents were selected at random
and were interviewed for about one and a half hours. The
research sample for this study consisted of a variety of cases,
depending on which year and on the research question that was
being analyzed.
Measures
Several variables were taken from the 1972 to 1998 General
Social Survey for use in this study. The first of these was the
independent variable. This variable was produced by merging
two other variables called CONLEGIS and CONFED. The
questions posed to the respondents by these variables were
“As far as the people running the Congress are concerned,
would you say you have a great deal of confidence, only some
confidence, or hardly any confidence at all in them?” and
“As far as the people running the executive branch of the
federal government are concerned, would you say you have
a great deal of confidence, only some confidence, or hardly
any confidence at all in them?” To measure the respondents’
confidence in these questions, the following code was devised
to indicate confidence levels: 1= A great deal, 2= Only some,
3= Hardly Any, 8= Don’t Know, 9= No Answer or BK= Not
Applicable. The resulting variable created from the merging of
these two variables was called CONF_GOV. This new variable
was later recoded from the original variables CONLEGIS and
CONFED to measure the confidence levels of respondents by
1= A great deal, 2= Only some, or 3= Hardly Any. Due to the
recoding, the respondent sample dropped from 38,116 cases
to 17,151 cases. Although there was a sample drop of 55%,
nevertheless there were still sufficient cases to prove the validity
of the measurement. This placed 12.31% of respondents as
those with a great deal of confidence, 63.33% of respondents
as those with only some confidence and 24.36% of respondents
as those with hardly any or no confidence in the federal
government.
Other variables were VOTE72, VOTE76, VOTE84,
VOTE88 and VOTE92. These variables measured the voter
turnout at each election in 1972, 1976, 1984, 1988 and 1992.
The common question asked in each of these variables was
“Do you remember for sure whether or not you voted in
that election?” Respondents’ responses were measured by
1= Voted, 2= Did not Vote, 3= Ineligible, 4= Refused, 6=
Religious non-voter, 8= Don’t know/Don’t remember, 9= No
answer or BK= Not Applicable. After recoding these initial
variables to filter out responses that are not applicable to this
study, the following new variables were created, elect_72,
elect_76, elect_84, elect_88, and elect_92. These new variables
measured the respondents’ response to whether they voted at
each election by 1= Voted or 2= Did not vote (which included
the response ‘refused’). Due to recoding, the initial number of
cases for these variables in 1972, 1976, 1984, 1988 and 1992,
respectively dropped from 38,116 to a total figure of 35,238
cases -- a 7.6% drop.
Sample descriptors used for this study were RACE, RELIG,
RELITEN, and DEGREE. RACE was used as an indicator
to check what race the respondent considered him- or herself.
The question posed was “What race do you consider your-
self?” Respondents’ responses were measured by 1= White,
2= Black, 3= Other. At first the number of cases recorded for
the variable RACE was 6,336, but after comparing these cases
to the responses of those that expressed their level of confi-
dence in the federal government, the sample size dropped to
5,128—a 19% drop.
Although also a sample descriptor, RELIG was also utilized
as an independent variable, a means of demographically
analyzing the population that has no confidence in the federal
government. RELIG measured what the respondent’s religious
affiliation was. The question posed was “What is your religious
preference? Is it Protestant, Catholic, Jewish, some other
religion, or no religion?” The response to this question was
measured by 1= Protestant, 2= Catholic, 3= Jewish, 4= None,
5= Other, 8= Don’t know and 9= No answer. This variable
was recoded to REL_AFIL. It was measured by 1= Christian
(which merged Catholic and Protestant religions, in which
members typically go to church), 2= Jews, 3= Atheists and 4=
Other. Due to the recoding, the original sample dropped from
4,852 to 3,696—a 23.8% drop.
The variable RELITEN measured how religiously strong
the respondent was. The relevant direction was to “‘Ask everyone
with any religious preference named in RELIG’…‘Would
you call yourself a strong or a not very strong [member of
your denomination]?’” The response was measured by 1=
Strong, 2= Not very strong, 3= Somewhat strong, 4= No
religion, 8= Don’t know, 9= No answer and BK= Not applicable.
It was later recoded into the variable REL_INT, by
transforming the response ‘4= No religion’ to ‘4= Atheist’ and
merging the fields ‘8= Don’t know’, ‘9= No answer’, and ‘BK=
Not applicable’ as system missing. Therefore, the variable
REL_INT was measured by 1= Strong, 2= Not very strong,
3= Somewhat strong and 4= Atheist. Due to this recoding, the
original sample dropped from 4,852 cases to 3,696 cases—a
23.8% drop.
The descriptor/variable DEGREE measured the respondent’s
level of education. The General Social Survey asked
the questions, “Did you ever get a high school diploma or a
GED certificate?” and “Do you have any college degrees?”
Responses were measured by 0= Less than high school, 1=
High school, 2= Associate/Junior college, 3= Bachelor’s, 4=
Graduate, 8= Don’t know or 9= No answer. DEGREE was
recoded into DEG_ED. It was measured by 1= <HS (less than
High School), 2= HS (High School) or 3= >HS (greater than
High School). After considering the respondents who indicated
their degree of education and also indicated their level of con-
fidence in the federal government, there was an overall sample
drop from 4,852 to 3,696—a 23.8% drop.
Findings
The first research question in this study was “What
proportion of American adults have no confidence in the
federal government from 1973 to 1998?” For a tabular
representation of this relationship, please refer to Appendix B.
In 1973, 15.77% of respondents had a great deal of confi-
dence in the federal government, while 21.33% of respondents
had hardly any confidence. Although the values fluctuated
over time for these two attributes, the proportion of respondents
with hardly any confidence has always been significantly
greater than the proportion with a great deal of confidence. In
1998, 5.36% of respondents had a great deal of confidence,
while 42.85% of respondents had hardly any confidence in the
federal government. The chi-square value for this relationship
is 1229.10, with significance far below the 0.05 mark. In fact,
the relationship between confidence in the federal government
and year is significant at approximately 0.00. The phi value
is at 0.21, indicating that the relationship between these two
variables is weak.
The hypothesis for this relationship, which stated, “The
proportion of Americans that have no confidence in the
federal government from 1973 to 1998 is significantly greater
than the proportion of Americans that have confidence in the
federal government,” is supported by the data. These findings
confirm those of Walker and other researchers, who suggest
that the confidence of Americans in the federal government
since the 1970s is drastically low and still decreasing!
The second question for this study was “Is voter turnout
decreasing? If so, is this associated with the decrease in confi-
dence?” Since this is actually a two-part question, the data were
analyzed accordingly. The first part of this question addresses
the relationship between voter turnout and years. For a tabular
representation of this relationship, please refer to Appendix B.
In 1972, 12.81% of respondents voted, and although this
value fluctuated for the following election years, the fluctuation
was reasonable (i.e., a fluctuation of 7.8%); in 1992, turnout
reached its highest peak when 17.93% of respondents voted.
The second part of this question examines the association
between the voter turnout and confidence throughout the
years. For a tabular representation of this association, please
refer to Appendix B.
In 1972, 32.80% of respondents had hardly any confidence
in the federal government, while 12.81% voted. These two attributes
seem to fluctuate throughout the following years, but
for each of the years, the proportion of respondents who have
hardly any confidence in the federal government is greater than
the proportion that vote. There seems to be a logical explanation:
Those who have hardly any confidence do not vote. Apparent
proof of this can be seen in the following correlations.
From 1972 to 1976, as hardly any confidence increased, voter
turnout decreased. In 1976 to 1984, as hardly any confidence
decreased, voter turnout increased. But, in 1984 to 1988, as
hardly any confidence decreased, it did not seem to have any
effect on voter turnout, as voter turnout also decreased. From
1988 to 1992, hardly any confidence increased, but voter turnout
also increased. Therefore, the relationship between confi-
dence and voter turnout proved to be insignificant.
This result disproved the causal relationship held by many
that lack of confidence in the federal government is related
to low voter turnout. This also disproves the notion held by
Berresford that voter turnout is the most visible indication of
confidence in the federal government.
The final question in this study was “Demographically, who
has no confidence in the federal government?” As a means of
analyzing the demographics of the population that have no
confidence, this study used religious affiliation to examine the
relationship between voter confidence and religion. The nature
of this relationship is presented in Table 1.
| Table 1. Confidence by Religious Affiliation |
| |
Christians |
Atheist |
| A great deal |
7.68 |
4.54 |
| Hardly any |
36.38 |
47.31 |
| Chi-square=118.69,p<.05,phi=.066 |
We can observe in Table 1 that the proportion of atheists
who had a great deal of confidence in the federal government
was 4.54%, while the proportion of Christians that had a great
deal of confidence was 7.68%. The proportion of Christians
with hardly any confidence in the federal government was
36.38%, while the proportion of atheists with hardly any confi-
dence was 47.31%.
The hypothesis for the relationship between confidence and
religious affiliation, which stated that “The population who are
Christians are more likely to have a great deal of confidence in
the federal government than the population of Americans who
are atheists” is, in fact, supported by the data. The chi-square
value for this relationship was 118.69, with significance below
the 0.05 mark. However, the phi value for this relationship is
0.066, indicating that the relationship between religious affiliation
and confidence is negligible.
Conclusion and Discussion
The focus was in three parts: (a) to examine what proportion
of Americans has no confidence in the federal government,
(b) to compare this proportion to the voter turnout
trend, and (c) to search out which group(s) of people have no
confidence in the federal government (i.e., to define the alienated).
Analysis of data from the General Social Survey (GSS)
from the years 1992 to 1998 concluded that (a) the proportion
of Americans with no confidence in the federal government
from 1973 to 1998 is significantly greater than the proportion
of Americans with confidence in the federal government, (b)
lack of confidence in the federal government is not associated
with voter turnout, and (c) the proportion of the population
that is Christian is more likely to have a great deal of confi-
dence in the federal government than the proportion of the
population of Americans that is atheist.
In the early seventies, there was a drastic decline in the
confidence of Americans in their government. This can be
explained by the series of scandals that saturated the American
environment, scandals such as the Watergate scandal, in which
President Nixon was indicted for party corruption. During
the time-span of 1976 to 1984, the proportion of those with
hardly any confidence decreased, which meant more people
had confidence. This can be explained by the strong political
leadership that was displayed in the 1980s under the early
Reagan administration. During this period, America underwent
a period of economic and political prosperity, and issues such
as the Iranian hostage crisis were resolved. Again, during the
period of 1984 to 1988, there was still a decline in the proportion
of respondents with hardly any confidence. This was also
due to the economic and political prosperity of this era. But
from 1988 to 1992, there was a steep increase in the proportion
of the population with hardly any confidence. This may
be blamed on the economic recession during the Bush Administration.
For the public, the recession entailed higher unemployment
rates, which acted as a catalyst for failed businesses,
lower real estate prices and a general low morale regarding the
American government. Another speculation is that this trend
was also due to the rising amount of ‘soft money’ used by politicians
for campaigning, which, due to media attention, created
a negative connotation in the mind of the public.
When the level of confidence is compared with the proportion
of the population that voted at the various elections, the
results are very interesting. There is an inconsistency in the
trends of the two variables, therefore eliminating any causal
relationships between them. For instance, in 1972 to 1976, a
greater proportion of people had hardly any confidence, and a
lower proportion of people voted. The proportion of people
who voted might not have been related to the low confidence
level in the government. Instead, the many Americans who
were out of the country during the Vietnam War might have
affected the low voter turnout. Or the turnout may have been
due to the language barrier, which at that time was a problem
(Constantine, 1999). In 1976 through 1984, a lower proportion
of respondents had hardly any confidence in the federal
government, and a higher proportion of people voted. Here
again, the higher confidence level might not have been related
to the higher voter turnout. This turnout might have been
because the people did not want another Nixon and, therefore,
voted for whom they hated the least, which happened to be
Ford. In 1984 to 1988, a lower proportion of respondents had
hardly any confidence, meaning more people had confidence
than the previous year. Yet, a lower proportion of respondents
voted. Clearly, this illustrates the lack of relationship between
these two variables. Here again, the low voter turnout might
have been caused by the problem of the language barrier (Constantine,
1999). Also, people are more apt to vote when the
economy or political affairs are bad than when things are good.
During this time, there was an economic boom; therefore, in
order to maintain the status quo, many people might not have
voted. Many people also might not have voted because candidates
in the 1990s relied more heavily than ever before on
media such as television, rather than on personal contact, to
relay their messages and gain the people’s trust (Gerritt, 1997).
Therefore, their messages might not have been accessible or
believable to many. In 1988 to 1992, the proportion of respondents
who had hardly any confidence in the federal government
rose to its highest peak, and voter turnout also rose to its
highest peak. There have many explanations proposed for this
result. For instance, although people have hardly confidence
in the federal government, they might still vote for the candidate
they hate the least. Alternatively, such an increase might
have been caused by a higher percentage of elections done in
different languages so as to try and eliminate the language or
other structural barriers. Others might have voted because of
the people that the president brings into office. For instance,
the president appoints Supreme Court judges. This is a lifetime
position and so very, very important. Therefore, people with
hardly any confidence in the federal government might still
vote for a president because of the type of people he proposes
to appoint to certain positions such as to the post of Supreme
Court judge.
In Table 1, when confidence in the federal government was
compared with Christianity and atheism, the data showed that
a larger proportion of Christians than atheists had a great deal
of confidence, and a lower proportion had hardly any confi-
dence. This proves that those who are religiously involved and
have strong religious beliefs have a higher level of political
participation and a higher confidence in federal government
(Brady, 1989). More specifically, “the closer a person is to a
church, the more likely they are to support the political authorities”
(Hoffman, 1985). The logical explanation of this stance
taken by sociologists, and the results presented by this study’s
data, is that Christian doctrine and the Bible teach Christians to
be respectful and loyal to governing authority: “ ‘I urge, then,
first of all, that requests, prayers, intercession and thanksgiving
be made for everyone – for kings and all those in authority’” (1
Timothy 2:1-2). As Table 1 indicates, to increase public con-
fidence, the federal government needs to aim more programs
toward the group identified within this study as atheists.
When the phenomena are analyzed on a long-term scale,
such as this study’s, one finds that the proportion of Americans
with no confidence in the federal government from 1973
to 1998 is significantly greater than the proportion of Americans
with confidence in the federal government, that lack of
confidence in the federal fovernment is not associated with
voter turnout, and that the proportion of the population that
is Christian is more likely to have greater confidence in the
federal government than the proportion of the population that
this study identified as atheist.
Future studies could eliminate some of the limitations that
restricted the current study. For this study, data was obtained
from the General Social Survey. This study was done under
certain respondent parameters that propose structural barriers
that hinder and impede a valid representation of American
adults. For instance, the General Social Survey only surveyed
Americans that were English-speaking. This parameter automatically
excluded the approximately 14% of Americans who
are non-English-speaking (census.gov, 1990).
Also, since voter turnout is not an accurate indication of
confidence in the federal government, future longitudinal studies
ought to be done to determine what other factors, apart
from confidence levels in the federal government, might influ-
ence voter turnout.
Finally, future studies should be done which investigate the
relationship between confidence in the federal government,
voter turnout, and the income, age, and party affiliation of
respondents. Sociologists at the Keystone Research Organization
and at Austin Community College propose a relationship
amongst these variables. Such studies should be done in order
to better evaluate the factors influencing confidence in the
federal government and voter turnout so as to make America a
more democratic nation.
Appendix A.
FILE HANDLE GSSIN NAME=’C:\Program Files\spss\datasets\
gss7298.sys’.
FILE HANDLE GSSOUT NAME=’c:\MyFiles\confidence.sys’.
GET FILE GSSIN.
SAVE OUTFILE=GSSOUT /map/compressed/
KEEP=YEAR, RACE, RELIG, RELITEN, DEGREE,
VOTE72, VOTE76, VOTE84, VOTE88, VOTE92, CONLEGIS,
CONFED
FINISH.
Appendix B.
Confidence by Year
A great deal Only some Hardly any
1973 15.77 62.89 21.33
1974 6.91 48.84 44.26
1975 5.96 54.46 39.58
1976 5.69 58.75 35.56
1977 13.71 63.96 22.33
1978 7.12 60.75 32.14
1980 4.97 48.99 46.04
1982 7.42 58.76 33.82
1983 4.14 58.28 37.58
1984 6.98 58.41 34.60
1986 9.07 60.17 30.76
1987 8.02 58.31 33.66
1988 7.30 59.47 33.23
1989 10.62 59.28 30.10
1990 10.53 57.89 31.58
1991 11.59 58.15 30.26
1993 3.37 47.08 49.55
1994 4.15 45.83 50.03
1996 3.67 40.82 55.51
1998 5.36 51.79 42.85
Chi-square=1229.10;p<0.5;phi=0.21
Voter Turnout by Year
Voted
1972 12.81
1976 10.37
1984 13.36
1988 10.13
1992 17.93
Voter Turnout by Confidence
Voted Hardly Any
1972 12.81 32.80
1976 10.37 35.56
1984 13.36 34.60
1988 10.13 33.23
1992 17.93 39.91
Chi-Square=8.626;p>0.5;phi=0.05
Appendix C.
CONLEGIS: “As far as the people running the Congress are concerned,
would you say you have a great deal of confidence, only some
confidence, or hardly any confidence at all in them?
CONFED: “As far as the people running the executive branch of the
federal government are concerned, would you say you have a great
deal of confidence, only some confidence, or hardly any confidence
at all in them?”
VOTE72, VOTE76, VOTE84, VOTE88, VOTE92: “Do you remember
for sure whether or not you voted in that election?”
RACE: “What race do you consider yourself?”
Ihuegbu 89
Honors College Review 2001
RELIG: “What is your religious preference? Is it Protestant, Catholic,
Jewish, some other religion, or no religion?”
RELITEN: “ ‘Ask everyone with any religious preference named in
RELIG: ‘Would you call yourself a strong or a not very strong […]?’”
DEGREE: “Did you ever get a high school diploma or a GED certificate?”
and “Do you have any college degrees?”
Bibliography
Austin Community College. (no date). Political participation.
Outlines [On-line]. Available: http://www2.w.austin.
cc.tx.usplane/cr/outlines/opart.html
Beckwith, K. (1986). American women and political participa
tion. Westport, CT: Greenwood Press. Available: http://
info.greenwood.com/books/0313245 031324507x.html
Berresford, S. (2000, February 9). Finding trust in unexpected
places. John C. Whitehead Forum [On-line]. Available:
http://www.excelgov.org/whitehead/walker_speech.htm
Bobo, L., & Gilliam, F. D. (2000, February 4). Race, sociopolitical
participation and black empowerment. (Topical
Report #16) [On-line]. Available: http://www.icpsr.umich.
edu/gss/report/t-report/topic16.htm
Brady, H. E. (1989). Religious conservatism and political
participation. (Bib00313). Paper presented to the Midwest
Political Science Association. Chicago, IL.
Constantine, S. (1999, February 15). Vote: Low turnout traced
to language, disaffection. MassLive News [On-line]. Avail
able: http://www.masslive.com/news/se/feed/daily/
ho215hix.html
Gerrit, J. (1997, October 13). Some blame rain, apathy, confu
sion for low voter turnout. Detroit Free Press [On-line].
Available: http://www.freep.com/news/politics
qnovote13.htm
Hoffman, T. J. (1985, August). Religion and political change:
The impacts of institutional connectedness and religious
imagery. (Bib01414). Paper presented to the American
Political Science Association. New Orleans, LA.
Keystone Research Center. http://keystoneresearch.org/
90 Ihuegbu
Honors College Review 2001
press.releases/pademocracy.html
Kyoko, A. (1997, March 21). Four of 5 Americans lack confi
dence in government. Cable News Network [On-line]. Available:
http://www.cnn.com/us/9703/21/govt.poll/
McLagan, P., & Nel, C. (1997). San Francisco: Berrett-Koehler
Publishers. ReadersNdex Archive [On-line]. Available:
http://www.readersndex.com/imprint/000001n/00001cj/
0000…/title.htm
Walker, D. (2000, April 25). Improving government perfor
mance and building trust. John C. Whitehead Forum
[On-line]. Available: http://www.excelgov.org/whitehead/
walker_speech.htm |