Young Voters Will Decide the Election

The single most important television event of this election cycle is not the DNC, the RNC or the debates. Instead, it was the MTV Video Music Awards this past Sunday when, during the OutKast performance in the final ten minutes of the show, the concert hall erupted into a mock-convention complete with state delegations and campaign placards. This is because in this election, the most important voting demographic is 18-29 year-olds. In fact, they are so important that no other demographic even comes close.

If John Kerry wins this election, it is becoming increasingly clear that it will be on the strength of the major demographic (18-29 year-olds made up 17% of the 2000 vote) that has swung more dramatically against Bush since 2000 than any other. According to 2000 exit polls, 18-29 year olds broke 48% for Gore, 46% for Bush, and 6% other. This is only slightly different from the nation as a whole, which went 48.38% Gore, 47.87% Bush, and 3.75% other. However, Bush's current numbers among 18-29 year-olds are far lower than his numbers with the nation as a whole:

Bush Support in Recent Polls
	 18-29	 Total	  Date
Pew	  35	  45	  8/10
Newsweek   32	  42	  7/30 
There is also this piece of information form a recent Washington Post article (emphasis mine): In the latest Post-ABC News poll, taken immediately after the Democratic National Convention, Kerry led Bush 2 to 1 among registered voters younger than 30. Among older voters, the race was virtually tied. About 1 in 6 voters in 2000 was between 18 and 29 years old. [Kerry led 52-45 nationally among RVs in that poll] (...)

Bush's problems with younger voters began long before the Democratic convention, Post-ABC polls suggest. The last time Bush and Kerry were tied among the under-30 crowd was in April. In the five surveys since then, Bush has trailed Kerry by an average of 18 percentage points. [In those five polls Kerry led by an average of 5 points nationally]

This race is static except for voters under 30. They are the persuadables, the swingers, the undecideds. They are not polarized like the rest of the nation.

In 2000, one of the main reasons Gore was unable to put Bush away is because he lost the younger voters Clinton had won for Democrats for the first time since the 1970's. In fact, that is the only significant pro-Clinton demographic where Gore noticeably under-performed relative to the national popular vote. Clinton won victories far in excess his national popular victories in this demographic in 1996, and in 1992 Bush barely defeated Perot among the 18-29 set. Had Gore equaled the 10-12% partisan advantage Clinton held among 18-29 year olds, he would have won the popular vote by 2.2-2.5% instead of 0.5%. That is enough to take Florida, probably New Hampshire, and maybe even Nevada.

It is easy for pundits to dismiss young voters both because no pundit is between 18-29 and because young people generally do not turn out to vote at significant rates. However, even though turnout was low among this group in 2000, it was still 17% of the vote! Further, as the same WP article quoted above notes, 18-29 year olds are as volatile a swing group as you can find:

Dole lost to President Bill Clinton by 53 percent to 34 percent among 18-to-29-year-olds. Bush's father split the young vote in 1988 and lost to Clinton by nine points in 1992. The Reagan era marked the recent high-water mark for the GOP with younger voters, who gave the Gipper his biggest victory margin of any age group in 1984. Any voting demographic that is large is important (17% of the vote in 2000). Any voting group that is volatile and prone to large swings is important (voted overwhelmingly for Reagan and Clinton; swung ten points from 1996 to 2000). Any voting group that turns out in wildly varying degrees is important (over 50% turnout in 1992, under 40% turnout in 2000). Any voting group that has changed significantly while the rest of the country remains polarized and static is important (10-15 point swing since 2000). The 18-29 year old demographic has every single criteria that makes a demographic important in this election. This cannot be said about any other demographic to the same degree.

If Kerry is going to win in November, he must continue his success in reaching young voters. Strike that--he must continue to do even more to reach them, even to excite them. Gore ignored young voters in 2000 and paid the price. Clinton openly wooed them and governed for eight years.

99% of Americans Underrepresented in Congress and Electoral College

Brendan Loy notes an obvious flaw in American democracy:In case anyone doubts that the Electoral College has its problems, and that those problems tend to favor Republicans in the current political climate, I would like to point out a few statistics that I uncovered while researching this topic over the weekend to prove a point to Becky's brother, Casey.

First off, take a look at the 11 "red states" that make up the heart of the solid Republican Mountain West and Great Plains: Nevada, Idaho, Utah, Montana, Colorado, Wyoming, North Dakota, South Dakota, Nebraska, Kansas and Oklahoma. Combined, they have a population of 18,671,566.

Now take a look at New York state. It has a population of 18,976,457. Almost identical -- actually just over 300,000 more.

Well, guess what, folks? New York has 33 electoral votes. Those eleven states have a combined total of 52.

As a lifelong resident of large states (New York and Pennsylvania), I have always thought that the Electoral College and Congress were anything but examples of equal representation. Delaware, half the size of Philadelphia and smaller than each of the three suburban counties adjacent to Philadelphia, gets two Senators, while the Philadelphia area gets around 2/3's of one (and considering the way our voting habits differ from the rest of Pennsylvania, right now we get zero). And yet, despite the obvious representational bias in favor of residents of small states, we crow about our great democracy to the rest of the world? Bah! Tell that to me when I have equal representation.

However, as Kuff notes, our system is not just railroading large states like New York and Pennsylvania with what borders on a three-fifths compromise. In fact, every single state is getting screwed except the four smallest:

Q . Which state is most over-represented in the Electoral College?

The answer is Wyoming, whose 3 electoral votes cover just over 500,000 people, or about 167,000 per person. California, with over 35 million people and 53 electoral votes, has a ratio of one EV to nearly 670,000 people.

They're not the most screwed in terms of Congressional representation, though. That dubious honor falls on Montana, whose population of 917,000 is nearly double Wyoming's, but they both have one solitary member in the House. Delaware, South Dakota, Utah, and Mississippi all have over 700,000 people per representative as of the 2000 reallocation.

Kuff goes on to provide a spreadsheet that shows how many members of Congress each state would have if they were all treated equally. Here in Pennsylvania, we would have six more. In New York, there would be nine more. In California, there would be sixteen more. And its not just Gore states either. In fact, if population determined Congressional representation rather than land area, both Bush and Gore states would gain a total 67 new representatives each in the House alone. The 99% of the country that does not live in the four smallest states is not being treated equally.

I do not think we should abolish states, but population should determine Congressional representation and Presidential votes, not land area. It should be one person, one vote. After all, that is the way democracy works.

Kerry's Likely Voter Gap

Polling Report's cool new summary table of monthly two-way trial heats offers a quick glimpse at the difference in the state of the campaign as measured by likely voter models and registered voter models. In short, the likely voter model leans Bush:
      Registered Voters   Likely Voters
Gallup 2   Bush +1		Bush +3 
ICR	  Kerry +5	Even
Time	  Kerry +8	Kerry +7
Fox	  Kerry +3	Kerry +5
ABC	  Kerry +7	Kerry +1
Gallup 1   Even 		Bush +4
On average, Bush does three points better under likely voter models than he does under registered voter models. Considering this discrepancy, successful GOTV efforts of traditionally low voting, though Democratic leaning, groups such as minorities and people under thirty will be essential to Kerry's efforts this year.

Based on repeated polls showing a higher level of voter engagement this year, I believe that turnout this year will be higher than normal. Hopefully that will comes to pass, because not only would it be good for Kerry and other Democrats if election results looked more like registered voter models than likely voter models, but it would also be great for democracy.

Whose Base is Bigger?

Poll weighting has become a frequent topic of discussion among political junkies in the Blogosphere. Arguments over likely voter models compared to registered voter models, whether or not Nader should be included in poll questioning, estimates of racial demographics among the electorate and the breakdown of party affiliation among the electorate are the most common areas of discussion within this topic. Considering that my previous post discussed the state of the race among independents, and that I have at least briefly touched on all the other areas of poll weighting before, a post about party affiliation and the comparative sizes of party bases is in order.

There have been two major studies on party identification within the past year, one released by Harris in late February and the other conducted by the National Annenberg Election Survey during the first three weeks of July. Combined, the two surveys interviewed 10,000 people, with Harris weighing in at over 6,000 and NAES at 3,715 registered voters out of 4,275 interviews. Both surveys had very low margins for error and weighted their results to reflect the national demographics as a whole. Their results were as follows:

Harris	33   28   24   15
NAES	34   32   24   10
This could possibly be read as a gain for the GOP in party ID over the past few months, but considering the Pew poll of Party ID from November of 2003, that is unlikely. The more likely cause would be different methods or questions between the two surveys. The press release for the Harris poll itself pointed out its difference from other polls, noting that "unlike some of the other polls, The Harris Poll finds that the Democrats still retain a small lead over the Republicans in party identification."

Democrats apparently have a small, though possibly significant, lead in Party ID among the electorate. If the electorate really is as polarized as we are constantly told it is, a small, 2-5% lead can make a huge difference, especially if self-identified Demcorats and Republicans favor Kerry and Bush by similar margins. Looking once again at the same three polls from the previous post, it would appear that is the case:

    %Reps for Bush    %Dems for Kerry
ARG	   87		 87
Fox	   88		 83
Newsweek*   90		 86
* = In a three way race including Nader In these three polls, Bush clocks in at an average of 88.33, while Kerry cones in exactly three points lower. This three-point difference is not enough to overcome the 2-5 point difference among party ID throughout the nation, since Bush is receiving 88.33 percent of either 28 or 32 percent of the nation while Kerry is receiving 85.33 percent of either 33 or 34 percent of the nation. Bush closes one point, but Kerry maintains an even smaller, though still significant, 1-4 point lead in base size.

Unfortunately, poll internals rarely reveal how Party ID was weighted. Since the convention, only two polls, ABC and ARG, have provided this information:

   Dem	GOP  Ind
ABC 39	 29   26
ARG 37	 35   28
The ABC numbers definitely stick out compared to the studies I just discussed, and are probably the reason that Kerry led 52-45 in the two-way, registered voter trial heat compared to ARG's 49-46 two-way, registered voter trial heat. However, ABC did offer this justification: There was some movement in political party identification in this poll: Among registered voters, 39 percent say they're Democrats, 29 percent Republicans and 26 percent independents (among likely voters, who account for 55 percent of adults in this poll, it's 40 percent-32 percent-24 percent). That's more Democratic, and less Republican, than usual; it was 34 percent-33 percent-29 percent among registered voters in the last ABC/Post poll. Moving loyalty, of course, is precisely what conventions are all about. Not only does this poll reveal just how much weighting takes place in all polls (especially those with likely voters), it also reveals a very real Democratic Party bounce from the convention. Both candidates are doing very well among their bases, but the Democratic Party base is slightly larger and might be growing. If on Election Day 39% of voters are Democrats and 29% are Republicans, not only will Kerry sweep into office, but almost all of the downticket Republican gains from 1992-1996 will be wiped out.

Deep Blue Sea

With the apparent transformation of New Hampshire and Pennsylvania this year, D.C. and the eleven states that form the northeastern part of the Unites States fill a sea of blue where nary a Republican is safe. However, it wasn't always that way:

Northeast Partisan Index

     1988	      2000	  Swing
CT: DNC 3.6   DNC 17.0	 DNC 13.4
DE: GOP 4.7   DNC 12.6	 DNC 17.3
ME: GOP 3.7   DNC 4.6	 DNC 8.3
MD: DNC 4.7   DNC 15.9	 DNC 11.2
MA: DNC 15.6  DNC 26.8	 DNC 11.2
NH: GOP 18.4  GOP 1.8	 DNC 16.6
NJ: GOP 5.9   DNC 15.3	 DNC 21.2
NY: DNC 11.8  DNC 24.5	 DNC 12.7
PA: DNC 5.4   DNC 3.7	 GOP 1.7
RI: DNC 19.4  DNC 28.6	 DNC 9.2
VT: DNC 4.2   DNC 9.4	 DNC 5.2
In 1988, only DC, MA, NY and RI were safe Democratic states in this region (although not nearly as strong as they are now). Several, including DE, ME, NH and NJ were lean Republican states (NH was safe Republican). Now, only NH, PA and Maine's 2nd CD remain in the "swing" column. However, recent polls from those areas indicate that they might now also be considered safe Democratic states. Pennsylvania in particular has had strikingly pro-Kerry poll results over the last two months (+12, +8, +10 and 37-59% independent disapproval of Bush in the last four non-partisan PA polls). Further, with the exception of Pennsylvania, Nader received more than his national average in every single one of these states in 2000.

In the Senate, the northeast delegation includes 14 Democrats, 7 Republicans and 1 Indy who caucuses with Democrats. Four of the Republicans, Chafee, Collins, Snow and Specter, are pro-choice, and often draw the ire of the their Republican colleagues. Only Santorum and the New Hampshire duo are palatable to the conservative wing of the GOP. The Democrats in this group include many of the most liberal, such as Reed, Corzine, Kennedy and Sabarnes--four of Progressive Punch's top five. In Congress, the GOP once again comes in at less than 40%, with 35 of the 92 representatives from these 11 states. Six of the twenty most serious challenges being mounted against GOP held seats are coming from this region, while only one Democratic held seat, the PA-13, is under threat to switch GOP.

The northeast is not as populous as the south, a region that has been trending strong Republican since 1980. However, the northeast plus Florida is as large as the South plus Oklahoma, Kentucky and West Virginia. Florida has been trending strongly Democratic since 1988 (DNC +14.1). Considering long-term trends, Virginia may also soon become a swing state.

For every southern voter Republicans have swung with the "culture wars," they have lost a voter from the northeast (even if that voter now lives in Florida). The end result has been to make the Northeast so liberal that it almost makes the West Coast look conservative. It sure feels good to be swimming in a deep blue sea.


Advertise Blogads