(June 2006, MIT Press)
Nolan McCarty, Keith T. Poole, and Howard Rosenthal
Abstract
Political polarization, income inequality, and immigration have all
increased dramatically in the United States over the past three
decades. The increases have followed an equally dramatic decline in
these three social indicators over the first seven decades of the
twentieth century. The pattern in the social indicators has been
matched by a pattern in public policies with regard to taxation of
high incomes and estates and with regard to minimum wage policy. We
seek to identify the forces that have led to this observation of a
social turn about in American society, with a primary focus on
political polarization.
Our primary evidence of political polarization comes from analysis of
the voting patterns of members of the U.S. House of Representatives
and Senate. Based on estimates of legislator ideal points (Poole and
Rosenthal 1997 and McCarty, Poole, and Rosenthal 1997), we find that
the average positions of Democratic and Republican legislators have
diverged markedly since the mid-1970s. This increased polarization
took place following a fifty-year blurring of partisan divisions.
This turning point occurs almost exactly the same time that income
inequality begins to grow after a long decline and the full effects of
immigration policy liberalization are beginning to be felt.
Some direct causes of polarization can be ruled out rather quickly.
The consequences of “one person, one vote” decisions and redistricting
can be ruled out since the Senate, as well as the House of
Representatives, has polarized. The shift to a Republican South can
be ruled out since the North has also polarized. Primary elections
can be ruled out since polarization actually decreased once primaries
became widespread.
It is more difficult to find the causes of polarization than to reject
them because social, economic, and political phenomena are mutually
causal. For example, immigration might lead to policies that increase
economic inequality if immigrants are at the bottom of the income
distribution and do not have the right to vote. We document an upward
shift in the income distribution of voting citizens. In turn,
dispersal in income might lead to polarization. It also might lead to
laxity toward immigration if inexpensive immigrant labor in the form
of domestic and service workers is a complement to the human capital
of the wealthy.
In additional to our focus on the polarization of elected
office-holders, we look at patterns of polarization among economic
elites. By examining campaign contributions, we find very high levels
of polarized giving. While some billionaires clearly spread their
contributions to both parties to buy access, increasing numbers
concentrate their largess on the ideological extremes. This polarized
campaign giving, coupled with the emergence of the soft money loophole
has arguably contributed to the ideological extremism of political
parties and elected officials.
Finally, we also examine polarization among the electorate. While it
is fairly clear that the views of most citizens have not become more
extreme, those with strong partisan identifications have (DiMaggio, et
al., Fiorina). Consistent with other findings (King, Jacobson), we
find that partisans are more likely to apply ideological labels to
themselves and a declining number of them call themselves moderate.
Strong party identifiers are the most likely to define politics and
ideological terms while the differences in the ideological
self-placements of Republicans and Democrats have grown dramatically
since the 1980s. Given Bartels’ findings that partisanship has become
a better predict of vote choice, this polarization of partisans has
contributed to much more ideological voting behavior.
We also find that the polarization of the electorate has increasingly
taken place along economic or class lines. Unlike the patterns of the
1950s and 1960s, upper income citizens are more likely to identify
with and vote for Republicans than are lower income voters. However,
we find that class polarization is most likely a result of the
ideological shift of the Republican Party towards a more economic
libertarian position. This shift to the right was aided by a number
of social, political, and economic factors. First, as American
society has become wealthier on average, a larger segment of society
prefers to self-insure rather than depend on government social
programs. Such voters have become more attracted to the Republicans
and their agenda for an “Ownership Society.” Second, due to patterns
of immigration and incarceration, members of lower income groups are
less likely to be part of the electorate. This has the effect of
moving the median income voter closer to the mean income citizen,
reducing the demand for redistribution (Romer, Meltzer and Richard).
Third, middle-income voters in the so-called “Red states” increasingly
sympathize with Republican positions on social, cultural, and
religious issues (e.g. Franks). The Republican advantage on these
issues has mitigated any loss of votes that might have been associated
with their shift on economic issues. Finally, the emergence of a
class-based, two-party system in the United States has benefited the
Republicans and mirrored the patterns of economic polarization found
in other regions.
Finally, we examine the policy consequences of the fall and rise of
political polarization. The separation of powers makes it difficult
to generate coalitions large enough to produce policy change even when
opinion shifts. We exploit this observation to get some leverage in
disentangling the effects of political, economic, and social policies.
For much of the period when polarization fell, immigration policy was
restrictive and unchanged while income and estate taxes, defined in
nominal terms, became more onerous. For the period since the onset of
renewed polarization, we find strong evidence that “gridlock” has
resulted in a less activist federal government. The passage of new
laws has been curtailed due to the increasing difficulty of generating
the requisite bipartisan coalitions. The effects on social and tax
policy have been especially dramatic as real minimum wages have
fallen, welfare devolved to the states, and tax rates have diminished.
We also show how polarized politics has affected administrative and
judicial politics.
Party Polarization: 1879 - 2009
Updated 4 January 2010
Below is a graph of the difference between the Republican and Democratic Party
means on the first DW-NOMINATE dimension from the end of Reconstruction through
the 110th Congress. This difference in first dimension means is a
good measure of the level of political polarization. By
this measure polarization is now at a post-Reconstruction high in the House and Senate.
[Above are three more graphs that show the relationships
between political polarization, income inequality, and immigration. These
three graphs are from the first chapter of our (Nolan McCarty, Keith T. Poole,
and Howard Rosenthal) forthcoming book
Polarized
America: The Dance of Ideology and Unequal Riches (MIT Press,
spring 2006). An abstract of our book is shown above.]
With few exceptions, roll call voting throughout American history has been
simply structured. Only two dimensions are required to account for the great
bulk of roll call voting. The primary dimension is the basic issue of the
role of the government in the economy, in modern terms liberal-moderate-conservative.
The second dimension picked up
regional differences with the United States -- first slavery, then bimetalism,
and after 1937, Civil Rights for African-Americans. With the passage of
the 1964 Civil Rights Act, the 1965 Voting
Rights Act, and the 1968 Open Housing Act, this second dimension slowly
declined in importance and is now almost totally absent. Race related
issues - affirmative action, welfare, Medicaid, subsidized housing, etc. -
are now questions of redistribution. Voting on race related issues
now largely takes place along the liberal-conservative dimension and the
old split in the Democratic Party between North and South has largely
disappeared. Voting in Congress is now almost purely one-dimensional -
a single dimension accounts for about 93 percent of roll call voting choices
in the 110th House and Senate - and the two parties are increasingly
polarized.
Polarization declined in both chambers from roughly the
beginning of the 20th Century until World War II. It was then fairly stable until the late
1970s and has been increasing steadily over the past 20 years. Our
(Poole and Rosenthal, 1997) original D-NOMINATE
estimation ended with the 99th Congress. Interestingly, Congresses 100-
110, if anything, mark an acceleration of the trend (especially in the House). Note,
however, that the acceleration is smooth and does not show a particular jump in
polarization induced by the large Republican freshman class elected in 1994.
Polarization in the House and Senate is now at the highest level since the end of
Reconstruction.
In addition, the percentage of moderate Representatives and Senators continues to plummet. In
the House the percentage of moderates (-.25 to +.25 on the first
DW-NOMINATE dimension) has declined to about 10 Percent in both
Chambers.
The files below were used to construct the figure. Each file has 12 variables.
In order these are: Congress #; first year of Congress; Chamber Mean 1st dimension;
Chamber Mean 2nd dimension; Democratic Party Mean 1st dimension; Democratic Party
Mean 2nd dimension; Republican Party Mean 1st dimension; Republican Party Mean
2nd dimension; Northern Democrat Mean 1st dimension; Northern Democrat Mean
2nd dimension; Southern Democrat Mean 1st dimension; Southern Democrat Mean
2nd dimension.
Note that all 2nd dimension coordinates have been weighted. See the
DW-NOMINATE page for an explanation.
Political Party Means 46th to 111th Houses
Political Party Means 46th to 111th Senates
Below are graphs of party unity scores for the House and Senate. A Party Unity
vote is one that pits a majority of one party against a majority of the opposite
party. The Proportions shown in the plots are the proportions of the members
voting with their party on Party Unity votes. These figures tend to track the
polarization figures.
In the file below the first column is the congress number, the second is the first
year of the Congress, the third is the number of roll call votes in the House,
the fourth is the number of party unity roll calls (at least 50% of one party
against 50% of the opposite party), the fifth colum is the percentage of party
unity votes, the sixth is the proportion of Republicans voting with the majority
of their party, the seventh is the proportion of Democrats voting with the
majority of their party, the eighth is the number of roll call votes in the
Senate, the ninth is the number of party unity roll calls in the Senate,
the tenth is the percentage of party unity votes, the eleventh is the
proportion of Republicans voting with the majority of their party, and the
twelve is the proportion of Democrats voting with the majority of their party.
Political Party Unity Scores 46th to 110th Houses and Senates
Political Party Unity Scores for individual members of the House and Senate for the 46th
to 110th Congresses
View Site Map