One of the more common refrains from both the media and Democratic establishment is about how the country is increasingly craving bi-partisanship and ideological moderation. The Los Angeles Times featured a story this morning that is an excellent example of this refrain. Based entirely on campaign rhetoric and a few anecdotes that centered mainly on about stem cells and nuclear non-proliferation, the article stated, as fact, that the country had reached a centrist consensus, moderate politicians are thriving, and people are sick of both partisanship and anything except middle-of-the-road ideology. Here is the passage, which somehow missed ending up in the op-ed section:
The centrist consensus on many issues underscores an important dynamic in the 2008 political climate: At a time of growing frustration with gridlock and partisan bickering in Washington, politicians with a pragmatic, middle-course tack are thriving.
Perhaps the problem that pundits, journalists and politicians face when they frequently declare the country as growing tired of ideology and partisanship and as having reached a centrist, bi-partisan consensus, is that these pundits, journalists and politicians don't actually know that there are quantitative, macro-measurements of public mood in America. Rather than just basing arguments about the public mood on anecdotes and selective reading of polls and election results, their arguments would be better served if, at their core, they were supported by quantitative studies that demonstrated the central thesis of the article. You know, the sort of writing people were taught to do during their first year of college.
If you are a pundit or a politician, and you intend to claim that the country is sick of partisanship and ideology, and is demanding bi-partisan moderates on a large scale, let me help you out. Before making such a claim, in order to avoid running afoul of reality, you can simply check to see if any of the following six, macro-political trends are actually taking place:
Are the number of Americans who do not self-identify with either major party increasing as a percentage of the population?
Are the number of Americans who self-identify as moderates increasing as a percentage of the population?
Is the national policy mood of the public becoming more moderate?
Is third-party performance improving in general elections?
Are moderates challengers of both parties defeating incumbent non-moderates in primary elections?
Are moderates challengers of both parties defeating non-moderates incumbents of the opposing party in general elections?
All of these trends can be measured using quantitative, scientific studies of public opinion, voting records, and election results. If four or more of these six trends are taking place, then it can indeed be safely stated that the country is becoming less partisan and more moderate. If either one or zero of these trends is currently underway, then it can be safely said that the country is not moving toward increasing moderation and non-bi-anti-post-partisanship. If two or three are the case, then the overall trend is debatable.
For the first part of this series I look through the best available source of quantitative data to determine the current viability of each of the first three trends. Those trends all focus on measuring public opinion, while the second three measure election results. I will finish the series tomorrow, and part one in the extended entry.
Here is a quantitative look at each of the first three questions listed above:
Are the number of Americans who do not self-identify with either major party increasing as a percentage of the population? Answer: Yes, but only among Republicans
According to exit polls from presidential elections, the number of self-identified independents has remained static since 1984. In that election, as in 1988, 2000 and 2004, exactly 26% of the electorate self-identified as independent. In 1992, the number was 27%, and in 1996, it was 22%.
According to Pew, the number of self-identified independents in the electorate has remained static at 15% since 1990. When Independent voters are not pushed to choose a party toward which they lean, there has been a recent decline in Republican self-identification, but not in Democratic self-identification.
Harris also picks up the decline in Republican self-identification, but no drop off in Democratic self-identification since 1994
So, there appears to be a short-term drop in Republican self-identification, and also a long-term trend where people self-identify with a major party, but increasingly describe themselves as "leaning independent." However, this is not clearly a move toward bi-or non-partisanship, since only Republicans have faced self-identification declines since at least 1994, and since there is also an increasing sense of "strong" partisanship. Overall, weak partisan Republicans might be more prone to self-identify as independent or leaning independent, while weak partisan Democrats might be more likely to refer to themselves as strong partisans. Thus, this is a trend where more Republicans are self-identifying as Independent, not a broad move toward non-partisanship across all demographic groups.
Are the number of Americans who self-identify as moderates increasing as a percentage of the population? Answer: No Compared to the previous trends, this one is easy:
Harris shows "moderate" self-identification to be entirely static from the 1970's (40%), 1980's (40%), 1990's (41%) and 2000's (40%). Further, Harris also shows the combined number of people who self-identify as either liberal or conservative actually increasing from 50% in the 1970's, to 54% in the 1980's, to 56% in the 1990's, to 54% in the 2000's so far. While the 2000's as a whole are actually back down to 1980's levels of ideological self-identification, the previous two years, 2006 and 2007, have shown the lowest level of moderate self-identification, 37%, recorded since 1972.
ANES shows the percentage of the population who identifies as moderate, slightly liberal, or slightly conservative to all be entirely static from 1972-2004. Liberalism has seen an increase since 2000, and conservatism saw an increase starting in about 1994.
The General Social Survey also finds moderate self-identification to be entirely static across decades. Moderates self-identified as 39.4% of the population in the 1970's, and as 39.1% of the population so far in the 2000's.
Overall, the country is becoming more identified with non-moderate ideologies, not less. At best for moderates, it is possible that the country might not be turning less moderate.
Is the national policy mood of the public becoming more moderate? Answer: No Again, this one is easy. The national "policy mood," is a measure of actual liberal vs. conservative sentiment in the population developed by political scientist James Stimson. The measure is important, because it goes beyond mere ideological self-identification, and measur ethe way the country actually feels on a battery of policy questions. According to this measurement, the country has never been in a more liberal mood:
That graph was produced in 2004, before the last four years that have seen a big swing away from Republicans.
One thing is for sure based on the graph: the country can't be at its most liberal point ever and also be desiring moderate public policy. If the country was in a moderate mood, it would not be peaking in one direction or the other.
Across all three broad measures of public opinion, the country is not moving toward increasing moderation and / or anti-bi-non-post-partisanship. Moderate self-identification is either static or declining. The country is currently at a peak of liberal policy preferences, not a centrist node. The country is becoming less Republican, and has not experienced a decline in Democratic self-identification since at least 1994 (and possibly since 1984). What these three trends point to, if anything, is the country becoming more progressive, liberal, leftist, or whatever you want to call it. Of course, to state such quantifiable statistics in the media or Democratic establishment would be heresy, so instead we are treated to a series of anecdotes that counter the far more provable thesis.
