Increasing partisanship since the 1990s: more evidence

Share on FacebookShare on Google+Email this to someoneTweet about this on Twitter

In the book Red State, Blue State, Rich State, Poor State (see Razib’s review here and the book’s blog here), the authors note that the two major political parties have become more polarized in various ways since the 1990s, even though the average voter hasn’t changed much. Also, the key message of the book is that the red state – blue state culture war is mostly restricted to high-income, and to a lesser extent middle-income voters.

They searched some mainstream media outlets for the words “polarizing / polarization,” as well as buzzwords for the cultural split like “NASCAR dad” and “soccer mom,” and found that they either show up for the first time or increase during the early/mid-1990s and remain as high today. I’ve searched the NYT for “partisan,” as well as a variety of newspapers for the pejorative “partisan hack,” and they show the same pattern.

Here are the graphs:

For the first graph, I took the number of articles with “partisan” and standardized this by dividing by the number of articles with “the” — basically, all articles. (The 2008 point is an estimate based on the year so far.) Aside from 1984, when there was a huge divide between the two presidential candidates, there is nearly no change from 1981 to 1991. However, in 1992, when the culture war begins to take center stage, the frequency increases to about twice as high as during the 1980s.

For the second graph, I did a Lexis-Nexis search for “partisan hack,” a common culture war swear-word for what the other guy is. I included the 12 newspapers with the highest counts, and that covered most of the major papers as well as some lesser known ones (see full list below). Not being able to search the database for “the,” I couldn’t standardize these data, but they show the same pattern as above, so I doubt the year-to-year variation in total output explains it. Here is the total output per year for the NYT, for comparison. Again, the 2008 point is for the year so far.

Aside from a few jabs from The Imblerian in the early 1990s, the first time this phrase shows up is in 1994, and it spreads to an order of magnitude larger by the 2000s. Outside of newspapers, Lexis-Nexis returns a result from 1984 where a politician is quoted as calling another a partisan hack. So the term must have been invented before the 1990s, but surged during the culture war.

These data agree with the larger picture in the book: the topic of partisanship has become much more talked about since the 1990s, and the specific slander “partisan hack” has increased noticeably during the same time.

List of newspapers included in the Lexis-Nexis results: New York Times, Washington Post, Washington Times, Boston Globe, Arkansas Democrat-Gazette, The Star-Ledger, Richmond Times, Palm Beach Post, St. Petersburg Times, Pittsburgh Post-Gazette, Atlanta Journal-Constitution, and The Oregonian.

Labels: ,


  1. the problem is that words themselves go in and out of fashion without respect to real events. this is especially true of political words: 
    community organizer 
    pork barrel spending 
    reach across the aisle (i guarantee you that this *bipartisan* phrase has become more common) 
    etc. etc.

  2. Well, “partisan” isn’t a very fashionable or trendy word like “NASCAR dad” — has existed for a long time, and appeared at a decent frequency even before the 1990s. But suddenly it jumps in the 1990s. There doesn’t appear to be a fashion cycle, just a logistic type growth. By analogy, the carrying capacity increased because the political climate was more partisan, so there was more need to use it. 
    For “partisan hack,” it’s the same — it shows up as early as 1984 in a quote (not in a newspaper though), but it’s not until the mid-1990s that it really surges. It must be that its growth was more favored then than in the mid-1980s.

  3. When I say there was no fashion cycle for “partisan,” I’m referring to the fact that it’s prevalence is basically flat from 1981 to 1981, aside from a blip in a hot election year. If there was a fashion cycle, you’d expect to see it decreasing steadily during this time, before increasing in the 1990s.

  4. sure, but perhaps the dynamic is not cycles but spikes when words suddenly become popular. i’d bet that “vetting” was pretty modest until these last two years.  
    it’s also worth noting that vanilla/everyday words can become shallow catchphrases when used incessantly by the press.

  5. I’m sure that the “NASCAR DAD” kind of pop-soc demographics appeared first in marketing and then was applied to politics. In each case it started as internally-held information guiding decsion-making, and then bled out into journalism and pop lit. Once it was out there in the mainstream, you started getting self-referential feedback, where people started thinking of themselves or people they knew in accordance with pop-soc.  
    On top of that, a lot of commentators are equal-opportunity partisans hired to give “balance” to editorial boards. This almost always means a Republican, since the big media, even as late as 1982 or even 1984, were overwhelmingly centrist or center-left Democrats. The Republicans actually won on this, because their guys were hired to be Republicans, whereas the Democrats are natural Democrats who think of themselves as objective. (David Brooks is a maestro of pop soc, but also a Republican partisan.) 
    Since the realignment of 1968, Democrats in the House have become more liberal, and Republicans more conservative. Both are more partisan than their voters. (The Senate and Presidnecy are entirely different stories.) The Republicans have been picking off Southern Democrats and also (in the primaries) moderate Republicans. Parties are more ideological than they were as late as 1966, when a lot of the dividing lines were geographical. 
    The term “partisan hack” is used against media people who are knee-jerk partisan, by the partisan hacks of each party against everyone in the other party, and by centrist wheeler dealers in each party against the more ideological members of their own party. 
    My own opinion is that partisanship is often a good thing and really shouldn’t be a smear term — the parties are functional parts of our system, like the houses of Congress. I feel the same way about “liberal” and even “radical”. You don’t want representatives to be a wind-socks on the issues.  
    Even conservativism can be a good thing, though in my completely partisan opinion, too many of the people who call themselves conservatives today are Armageddonist loonies and bigots, influence-peddlers, and free-market fanatics.

  6. BTW, the fact that pop soc arose in marketing sounds like it means that it’s not really science, but the truth might be the opposite. If marketers were able to increase sales and make money with the help of sociological data, it’s an experimental or engineering validation of sociology.  
    One of the strengths of the Republican party over the last two decades has been a state-of-the-art electorate database, combined with top-notch polling. This shows especially in the last two weeks, when Republicans are able to tell where they’re safe and where they need votes, and if they need votes they know the best places to go to find them, right down to the precinct and neighborhood level.  
    They finally got burned in 2006 and probably will get burned worse this year, but for a long, long time they were able to parley 30%-35% ideological support into real national political power (total control 2002-2006).  
    Liberals tend to favor pure science over applications, and credentialed over uncredentialed knowledge, but with one undergrad year of schooling, over the years Rove whipped the asses of dozens and dozens of PhDs.  
    And “applications” are an important test of science.

  7. One of the most interesting—because counter-intuitive—of the RED STATE/BLUE STATE divides 
    was a map showing state-by-state contributions to federal income: in other words, which states were net tax-PAYERS and which were net tax-EATERS. I’m not really sure, but that map might have been on this very site in the past few months. 
    What it showed was that the PAYERS are BLUE and the EATERS are RED; Texas, which is RED on the political map, was about even-steven. 
    My guess is that the imbalance is due mostly to the less-densely-populated red states requiring more (comparatively, to their populations of taxpayers) expenditures on road-building and maintenance and, possibly, to receipt of more funds spent on military sites and payroll (due to such installations being sited in those less-populated areas of more available land). 
    But I think that map would still surprise most red-staters.

  8. We certainly don’t need the non-partisan unity of a one party state. A left that uses financial panic-mongering, through media predominance, should lose massively. They put power-greed ahead of concern for the collateral damages

  9. JSBolton: 
    In that respect, the left hasn’t changed a bit. They’ve always elevated the advance of socialism above other causes, including even the security of the country. Rush Limbaugh exxagerates not at all when he says that the left welcomes economic malaise (for much of which their policies are directly resposible) simply for the opportunity it affords to spin it as the result of unchecked 

  10. Gene, I found something like it here – from 2003. That is interesting. One thing though is that it’s only logical that the richer the state the less federal largesse. After all, what’s the point of a rich state sending money to Washington only to just get it back? (Obviously the point is politics – it allows politicians to spend money while pretending someone else is paying for it). And the correlation of this Federal Expenditure to median income by state is -.5. 
    And since we know that richer states (if not richer people) vote more heavily Democratic, this isn’t that surprising. What intrigued me was which states are getting more than their fair share – which states are getting funded disproportionately to their income? 
    What I did was compare each states net Federal benefits to their income. I added together the z-score for their income to the z-score for their Federal expenditures. If the poorest state got the most money, each proportionately to the average, then that state would get a zero. States with highest scores are the ones getting the most out of the federal system. Not surprisingly, Alaska comes out on top. Perhaps surprisingly, Texas is dead last. 
    The correlation of income to % vote for Bush in 2004 is -.45. The correlation with net Federal Expenditures is +.45. But the correlation of % vote for Bush and my relative federal benefit statistic is 0.01. So there doesn’t seem to be much there as far as which states manage to game the Federal system and party in power. 
    Also, these expenditure data might be a bit distorted by including Social Security. 
    Data can be found here.

  11. It is amazing, how some can postulate good management by the Republicans. It is hilarious. McCain mismanaged his campaign early, and the party has not a clue, as to why they have lost, they can only make excuses. 
    The Democrats, want to save everything, but without any oversight or forethought. All the while statistics are thrown together with the primary components left out to spin their point of view.  
    Simply astonishing. It is no wonder we need a paradigm shift, to wake people up. It is the fraud that is perpetrated daily by these people, more than the subject itself.

  12. Yes, the accusations of partisanship have increased, but I think that has been used to mask an actual increase in bipartisanship. Once the Republicans got rid of the Gingrich crowd who were interested in a strongly ideological conservatism that approached libertarianism (please note that one of the members of that group was Bob Barr, now the Libertarian Party candidate for President) in order to try to keep power, they acted just like the Democrats in most things, especially in economics. And, despite the rhetoric, the Democrats are almost identical to the Republicans on social issues. Even on things like abortion, they are really arguing about the nuances of when it is appropriate to have an abortion, not if it is appropriate. The policies put in place by both parties following 9-11 and the recent bailout make it even clearer that the Republicans and the Democrats are almost identical in their desire for the government to seize power throughout the country. And whether it is Obama with his factually-inaccurate understanding of economics or McCain with his admitted ignorance of economics, the outcome was the same: both voted for the bailout bill, which has resulted in the partial nationalization of many banks (and before the bailout, the U.S. government already nationalized Freddie Mac and Fannie Mae, thus bringing back into the fold two huge mistakes which will remain mistakes, since nobody will get rid of them). The elevated rhetoric has only worked to mask these facts. I don’t think that’s an accident, either.

  13. Re: red states being net tax eaters, the problem there is the aggregation of voters into states. I remember doing this analysis a few years ago.  
    If you disaggregate into voters, it comes out that high income individuals are tax payers (obviously) and more likely to vote Republican (though ultra high income voters are more likely to vote Democrat).  
    In other words, the recipients in the red states are disproportionately blue voters and the donors in the blue states are disproportionately red voters (where the def of “disproportionately” is TBD — you’d have to see whether the correct denominator is the immediate surrounding environment or the national environment…)