Articles and Analysis


The Curious Case of the Liberal-Leaning Youth Vote

Over the last few months, I've focused a great deal of attention on researching the youth vote, digging into the factors that influence partisan identification and voting behavior among these young voters. As a Republican pollster, I'm primarily interested in diagnosing "what went wrong?" in terms of young voters in the last few cycles. Some of the possible answers seem obvious - the charisma of Obama, the age of McCain, the disillusionment with the GOP, the Iraq war, President Bush, the GOP's differences with young voters on issues like the environment and "moral issues" - and in the coming weeks, I hope to examine these factors. But first, I'm interested in starting a discussion with Pollster.com readers about the topic.

As a starting point, let's take a look at the electorate from 1972* through 2004. This comes from media exit poll data and the universe for "overall" in this instance is those respondents who provided their age. Additionally, "independent" does not include "other/something else". I am currently waiting on the data set for 2008, and thus I'm not yet able to break down 2008 by age, so it has been omitted from this chart. (Democrats did better in terms of partisan identification in 2008 than they did in 2004.)
Thesis Figures.003.jpg

This chart shows that prior to 1984, partisan identification was more fluid, with Watergate turning massive quantities of Republicans into Independents, particularly among young voters. Yet once 1984 rolls around, young voters behave quite similarly to voters overall in terms of partisan identification. With the exception of the 1996 election (featuring the Republican nomination of Sen. Bob Dole), trends in partisan identification among young voters look quite a bit like they do for the electorate overall.

Yet ideologically, a very different story is told about elections since 1984. While partisan identification among young voters has very closely followed that of the overall electorate since that time, the youth vote has been ideologically very different ever since the 1992 election of Bill Clinton. In this case, I use 1984 as the starting point because media exit polls prior to 1984 ask their ideology question in a different manner, omitting "moderate" and instead using "somewhere in between" as a response option. For the sake of comparing apples to apples, I've chosen to begin the chart where the question wording becomes consistent.

Thesis Figures.004.jpg
In 1984 and 1998, the 18-29 cohort contained more "conservative" identifiers than "liberal" identifiers. Yet in 1992, this changes and more 18-29 year olds identified as "liberal" than they did "conservative". Yet among voters overall, conservative still found many more identifiers than liberal. This trend has continued since Clinton's election; while the overall electorate has more conservatives than liberals, the young electorate has more liberals than conservatives. Something happened in 1992 (perhaps the election of the young governor of Arkansas?) that changed the ideological makeup of the young electorate. It will be extremely interesting to see what the ideological makeup of the 2008 electorate is in terms of young voters.

So what does this mean for the Republican Party moving forward? In 2004, young voters looked a lot like voters overall in terms of party ID, but looked further to the left in terms of ideology. What does this party ID/ideology discrepancy mean?

Once we get a chance to dig into the 2008 data, we will know more about how the youth vote changed in the environment that was nothing short of toxic for Republicans. But even the ideological post-Clinton shift that endured into the Bush years should be a fact that the GOP considers as they craft a strategy moving forward to rebuild a long-term majority coalition.

*In the interest of disclosure, 1972 is not an arbitrary starting point, it is rather the oldest data set I currently have, and I would welcome the input of any scholars or readers with data from media exit polls prior to 1972).



"In 2004, young voters looked a lot like voters overall in terms of party ID, but looked further to the left in terms of ideology. What does this party ID/ideology discrepancy mean?"

it looks like it means that conservative youth is largely centrist, and democratic youth is left leaning.

which makes me wonder if either party knows how to prepare for the future, between Obama's centrist rhetoric and Republicans recoiling to the right.



One thought - if you're going to look for some kind of temporally linked cause of this, I wouldn't look in 1992 when you start to see the effects of it, because its not as if these kids have been voting for awhile and then switched parties. Rather, I would look somewhere around 1972 when this cohort was born, or shortly thereafter.

My personal hunch, and it is just a hunch (though I'm 30, so basically in the cohort we're discussing), is that people my age and younger have no real memory of the Cold War, the '60's, or Jimmy Carter. So the guilt-by-association tactics used to keep people from identifying as liberals seem bizarre and out-of-touch.



You might want to look to the American National Election Studies for pre-1972 data. These data are based on pre- and post-election interviews in the presidential election years. If their 1972-2004 data compares favorably with your exit poll data then the ANES could provide pre-1972 data points.



@ striatic - Actually, Obama's strategy sounds pretty solid - by appealing to centrists, he can probably be able to start grabbing more youth Party ID. If Republicans keep moving towards the right (or if just the perception of them does so), the left-leaning youth are already in the bag, after all.


William Ockham:

This type of analysis will doom the Republican party. You're looking at the wrong things. First, neither party id nor self-described ideology is very useful in analyzing voting behavior in Presidential elections. What matters is simply which party's candidate do people vote for. Second, people get 4 years older for each Presidential election. I know that sounds obvious, but because an age cohort's relative political split stays roughly constant over their lifetime, comparing the voting behavior of 18-29 year olds in different election years obscures as much as it illuminates.

I'll tell you the important fact that your data is hiding from you. People born since 1978 have voted 2-1 for Democratic Presidential candidates in every election they've voted in. This was first measurable in 2000. It was true in 2004 and it was crystal clear in 2008 since the entire 18-29 year old group belongs to that generation now. It will undoubtedly be true in 2012. The thing that should be scaring the bejeebers out of a Republican pollster is that by 2016, this cohort will dominate the political landscape their sheer size (much larger than the Republican-leaning cohort that started coming of age during the Reagan years) and because they will be reaching the age of maximum political participation.


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.