Articles and Analysis


NYT/Pew Graphic: Party by Generation

Topics: 2006 , Party Weighing , The 2006 Race

A staple of modern American politics is the article or column that inevitably follows any landslide election speculating whether this particular victory heralds a national realignment that will reshape politics for decades. It rarely does. Make what you will of the fact that this year's version, by David Kirkpatrick in today's New York Times, appears three weeks before the election, but whatever you think of the timing of this piece, do not overlook the remarkable, must-click graphic that accompanies it.


The chart plots data on the current party identification of Americans compiled from more than 23,000 Americans surveyed by the Pew Research Center during 2006. With that many cases, the Pew pollsters were able to tabulate a result for party identification (including independent "leaners") for each birth year and plot the results. What results is a remarkable picture of the politics in play as each age group emerged into adulthood. The most Republican cohorts are those who came of age during the administrations of popular Republicans: Dwight Eisenhower, Ronald Reagan and George H.W. Bush. In contrast, the current crop of 18-24 year olds, according to the Pew Research center data, is the most Democratic leaning group in the population.

The chart provides graphical evidence of the slow rolling realignment that is always at work as new young voters gradually replace their elders. Political scientists generally agree that young people tend to acquire political beliefs, including their partisan attachments, in their 20s. As Kirkpatrick writes, "voters typically develop a party preference based on the political atmosphere at the time they come of age and grow more attached to that party over the course of their lives." Once acquired, a true sense of party changes rarely changes, although some voters are less attached to political parties than others (as I speculated on Friday, some will shift back and forth on surveys depending on the politics of the moment or the wording or context of the survey question).

The gradual shifts toward Democrats since 2005, and the small recent spike I wrote about Friday, are almost certainly not harbingers of some coming realignment. They will likely be as fleeting as the shifts toward the Republicans just prior to the 2004 elections. The bigger question is what these changes portend about the still outcome of the elections next month. The implications for Republicans certainy seem dire, but we will not know for certain until November 8. 



I found the Times article incredibly grating, becuase in each of the previous shifts, there was not only disenchantment with one party, but an accompanying strong alternative story told by the other party. This article would have you believe that FDR and Reagan were basically in the right place at the right time, and completely fails to mention that each man articulated an alternative vision which purported to solve the problems the other party had caused. There is no such vision coming from the Dems right now; it is hard to picture how a realignment can happen without one, and yet, the article just happily glosses over that.

[This is not to say that I won't be voting basically straight-ticket Democrat in November, but... jeez, it isn't like I (or anyone I know) is actually excited about that, like I imagine people must have been for FDR and Reagan.]



Fascinating graphic. But I'd like to see it frozen in time, for example what the self-descriptive percentages among twenty-somethings was in 1960, 1990, etc. I'm sure the basic trends from era to era would be evident, but don't young people tend to be more liberal? As a Democrat I'm not exactly confident the 52-37 edge will fully hold up, when the same block identifies itself 10 or 20 years from now.

Otherwise, I agree with Luis Villa in the first comment. And it's a theme I've stressed many times, that superiority is more lasting and meaningful than ineptitute.

Notice it said of Reagan, "By winning over millions of white working-class Democrats..." That became a category, Reagan Democrats. Every cycle you hear Republicans evaluate their candidates with, "is he another Reagan?"

Democrats haven't won over anybody. They've inherited voters and perhaps a midterm verdict, essentially via forfeit. I seldom concur with Michael Barone, but he nails it late in the NYT article: "If the Republicans suffer losses, it is a negative verdict on competence rather than ideology."

And from Professor Jack Pitney Jr. of Claremont McKenna College, also late on page 2 of the article: "Franklin Roosevelt had a clear answer, a very specific program, the New Deal,� he said. �Do the Democrats have the same kind of clear answer to Iraq? It is damaging to Republicans, but it is hard to see how it is constructive to Democrats.�"

Reminds me a little bit of the difference between Ohio and Virginia. Very likely in this cycle the Democrats fare much better in Ohio, winning the gov race and senate race while Virginia looks like it will go to Allen. But longterm I'm much more bullish on Democratic futures in Virginia. The northern counties are clearly moving in demographic favor of Democrats and at projected pace it will flip the state in the near future. Nothing similar in Ohio. Democrats will win this year, and Kerry nearly prevailed in 2004, due to GOP failures and corruption. But that has no lasting impact. Once Democrats are in office they bear the burden of expectation, with no demographic movement in our favor, and the state returns to base instinct of 2-4 points red.

The "Emerging Democratic Majority" theories mentioned in that Times artlcle were certainly valid, but delayed by 9/11 since it impacted 2002 and 2004 voting patterns among women and Hispanics. I don't think we'll know until at least post-2008 if those blocks have returned to previous tendencies.


Gary Kilbride:

The rambling comment above is mine. For some reason the nametag didn't show up.


Ben Ross:

One small comment about this fascinating, fascinating chart. The chart shows who was president when each age cohort turned 20. It looks to me like the dominant influence is the political atmosphere when people are around 15-16.

The Democratic peak among people now in their late 70s reflects the World War II years. The Republican peak that is shown on the chart matching Eisenhower's presidence actually matches up with Truman's second term and Eisenhower's first (the McCarthy years). The steady rise in Democratic support matches 1956-1972, a much better timing for the youth movement of the "sixties" than 1960-1976. The next Republican peak then matches up with the Reagan years.

And just to complicate things, let me throw in an alternative explanation that the late Jim Chapin advocated. He argued that the 30-year waves of left-right alternation that you see in American history are caused by parents handing their partisan identification down to their children. He predicted through the seventies and eighties that there would be another wave of leftism in the nineties. Chapin turned out to be wrong about the direction of government policy, but his theory was really about voter self-identification, and this chart suggests his prediction was right to that extent.


Jeff Z:

A very interesting chart, but of limited value compared to how a single voter's party id changes over the years. I would have been listed as a D if I were in the youngest cohort, but shifted to R a long time ago. What are the long-term trends among voter shifts? The country seems much more R now than when I was younger.


MP readers rightly point out some significant problems with the (very interesting) NYT analysis ? with a single snapshot one can not distinguish life-cycle effects from cohort effects. But we now have 50 years of survey data to look at. By pooling cross sectional data over time, it would be possible to distinguish these two sets of effects. I seem to recall that Philip Converse looked at this in the mid 1970s, and with data from the National Elections Study, it would be possible (though not easy) to tease these effects out. But I will leave that to someone else?


Icarus Ascending:

I agree and disagree with the commenters who think that voters have moved away from Republicans but not necessarily toward Democrats.

While it's certainly true that Democrats have yet to lock-in these newly unaffiliated voters, they also haven't had much of a chance. It's not like they can effect major change as a minority party; even when they propose bold measures, the media barely reports on it (if at all) because said measures don't have the votes to become law.

If the Dems retake Congress this year, they will again become relevant. The media will report on their beliefs, their policies, and their plans. They will have the opportunity to make their case for a couple of years before the major debate about our future is had in 2008. (There was no major debate in 2004 partly because the Dems' candidate was so weak, and partly because there was never any chance that the election would be anything other than a referendum on the policies of George W. Bush.)

I guess my argument is that the Dems' first real opportunity to articulate a vision is yet to come. If they blow it, we're doomed to become a nation of "independents." If they blow it bad enough, Republican defectors will come home, and the electorate will remain largely as it is (and possibly a bit more Republican-friendly). If they do a good job, however, it should be enough to cement a Democratic-leaning electorate for at least 15-20 years.


emcee fleshy:

So, over the long term, Truman was the worst for his party, and Ford was just as good as, or better than, Reagan?

I'm skeptical.



It's just plain silly to expect the dems to come up with a full blown political agenda for this election.

1) Bush will veto anything they propose.

2) To have a chance at taking the senate, they have to give some red state dems plenty of manuevering room.

The time for the "vision" is 2008- when there will be a presidential candidate to articulate it. The only goal this time is to stop Bush from totally ruining the country- and people understand that.



I notice most of (all of?) the posters are men, as usual (sigh). I want to take issue with the idea that people become more conservative as they grow older.

People grow more conservative as they grow RICHER, not as they grow older.

For most men, that is a steady linear process--toward wealth, away from caring about people without wealth (the definition of modern-day "conservatism," which is really quite radical). So it makes sense that most men, who earn more as they age, get more conservative (more concerned about conserving their money and the privileges it brings) as they get older.

But women don't get richer as we get older. We get "richer" by marrying wealth or by not having children. This we can decide to do at any age. There are young, fabulously wealthy women who are wealthy because they married it (or slept with it). A typical woman's paid economic life is full of starts and stops influenced by her other fulltime, nonpaid work as nurturer of the next generation. She doesn't always make more money as she gets older. She may go part-time when her children are born, or quit paid work entirely. When she goes back, it will probably be at a far lower rate of pay than when she left.

Men don't DO this. They go out into the work perfectly expecting to make progress in their jobs and their incomes. They know they've got a great deal getting women to do almost ALL the nonpaid labor essential to keeping the human race going (which, I grant you, may not be a good thing...)

So women don't grow more conservative, because at the end of the day, most women don't have much (money) to conserve. What they DO have are problems that require collective solutions, because they can't afford to implement individual ones.

The split here is a gender split. The more women get involved in political life, and the more men drop out, the more "liberal" we will say our society is becoming. I think we've let the boys be in charge for far too long and frankly, they've ****ed it up bigtime. Male values of competition and dominance are simply NOT the way humanity was wired to be.



I like your attitude Deb: men have ****ed everything up. Congratulations, you are SEXIST! If comptetition and dominance (as you say) are male values, how are they not part of humanity? Are you implying that men are not part of humanity? You make some good points, but you should try to get over your sexism before you post. This graph has little to do with male/female voting patterns. Please try to stay on topic in the future.


Post a comment

Please be patient while your comment posts - sometimes it takes a minute or two. To check your comment, please wait 60 seconds and click your browser's refresh button. Note that comments with three or more hyperlinks will be held for approval.