8
NARAL Pro-Choice America Micro-targeting Pro-Choice Women Who Voted For President Obama in 2008 But Are Not Strong Supporters in 2012

Micro-targeting Pro-Choice Women Who Voted For President Obama in 2008 But Are Not Strong Supporters in 2012

Embed Size (px)

Citation preview

Page 1: Micro-targeting Pro-Choice Women Who Voted For President Obama in 2008 But Are Not Strong Supporters in 2012

NARAL Pro-Choice America

Micro-targeting Pro-Choice Women Who Voted For President Obama in 2008 But Are Not Strong Supporters in 2012

Page 2: Micro-targeting Pro-Choice Women Who Voted For President Obama in 2008 But Are Not Strong Supporters in 2012

What We Did: Modeling Pro-Choice Obama Defector Women

PROJECT GOAL: The identification of a set of women in battleground states most likely to have voted for President Obama in 2008, but who are not strongly supporting him now.

During June and July 2012, GQRR completed more than 20,000 IVR (interactive voice response) interviews. We then cross-referenced our survey data with consumer data and other variables from Catalist. Using advanced analytic techniques, we built three mathematical models:

1. Choice. Updates NARAL Pro-Choice America’s model predicting the likelihood of whether a voter is pro-choice or anti-choice.

2. Persuasion Obama Defectors. Identifies the people most likely to have voted for President Obama in 2008 but are either supporting Romney, are undecided, or are only soft Obama supporters now.

3. Turnout Obama Defectors. Identifies the people most likely to have voted for President Obama in 2008 but are least likely to turn out in 2012.

© Greenberg Quinlan Rosner 2

Page 3: Micro-targeting Pro-Choice Women Who Voted For President Obama in 2008 But Are Not Strong Supporters in 2012

What We Did: Building a List

The models were then applied to the entire Catalist voter file, giving every voter on the file a score of between 0 and 100 on both choice and defection:

• Choice. A score of 100 means that voter is most likely to be pro-choice, while a score of 0 means the voter is most likely to be anti-choice.

• Obama Defectors (both persuasion and turnout). A score of 100 means that voter is most likely to be an Obama defector, while a score of 0 means the voter is least likely to be a defector.

The choice and persuasion models were then crossed to find the women voters with the highest scores on both.

© Greenberg Quinlan Rosner 3

Page 4: Micro-targeting Pro-Choice Women Who Voted For President Obama in 2008 But Are Not Strong Supporters in 2012

4© Greenberg Quinlan Rosner

Obama Underperforming Among Women Relative to 2008

Obama McCain Obama Romney Obama McCain Obama Romney

5652

4946

4347 48 49

+5 +1 -3+13Women Men

Exit Polling 2008 Sept. 2012Sept. 2012 Exit Polling

2008

Page 5: Micro-targeting Pro-Choice Women Who Voted For President Obama in 2008 But Are Not Strong Supporters in 2012

5

National Battleground States Battleground Counties0

1

2

3

4

5

6

7

5,145,542

1,219,281338,020

Number of Women Most Likely to Be Pro-Choice Obama Defectors

In M

illio

ns338,000 Women in the Battleground Counties Are Most Likely to Fit the Pro-Choice Obama Defector Profile

© Greenberg Quinlan Rosner

Page 6: Micro-targeting Pro-Choice Women Who Voted For President Obama in 2008 But Are Not Strong Supporters in 2012

A Majority of the Most Likely Pro-Choice Obama Defector Women Are Independent and Younger

Most Likely Pro-Choice Defector Women

Democrat 29

Independent/Other/Unknown 56

Republican 1

18-29 26

30-39 27

40-49 17

50-64 20

65+ 10

White 67

Black 18

Hispanic 8

Other 6

© Greenberg Quinlan Rosner 6

Page 7: Micro-targeting Pro-Choice Women Who Voted For President Obama in 2008 But Are Not Strong Supporters in 2012

The Most Likely Pro-Choice Obama Defector Women by County

State County # of Likely Pro-Choice Defectors

CO Adams 11,190

CO Arapahoe 2,906

CO Denver 10,927

CO Jefferson 3,296

FL Hillsborough 32,035

FL Orange 5,173

FL Pinellas 27,590

IA Johnson 8,017

IA Polk 3,410

IA Story 2,202

NC Mecklenberg 5,667

NC Wake 9,744

State County # of Likely Pro-Choice Defectors

NH Hillsborough 3,433

NH Rockingham 2,537

NV Clark 13,779

OH Cuyahoga 43,067

OH Franklin 43,616

OH Hamilton 20,432

VA Fairfax 49,661

VA Henrico 3,590

VA Loudoun 12,942

VA Prince William 3,965

WI Kenosha 993

WI Milwaukee 15,855

WI Racine 1,993

© Greenberg Quinlan Rosner 7

Page 8: Micro-targeting Pro-Choice Women Who Voted For President Obama in 2008 But Are Not Strong Supporters in 2012

What It Means

The pro-choice Obama defector model—exclusive to NARAL Pro-Choice America—allows us to maximize efficiency in targeting this key voting bloc at the individual level.

• NARAL Pro-Choice America now owns the only Obama defector women model and the most accurate choice model in the country.

• The model eliminates guesswork. – NARAL Pro-Choice America no longer wastes resources communicating with someone who is anti-choice,

nor do we “miss” voters who are pro-choice. We can now communicate more often and more directly with prime targets without incurring any additional cost.

– It allows an organization like NARAL Pro-Choice America to put a specific pro-choice message in front of a specific voter.

• The model will help guide NARAL Pro-Choice America’s 2012 political outreach efforts. NARAL Pro-Choice America will not work directly with the Obama campaign, but because we know exactly who these women are and what messages they respond to, NARAL Pro-Choice America can maximize the value added to the Obama campaign in a focused, strategic way.

• The model can also be shared with pro-choice candidates and campaigns around the country to help them improve their communications.

© Greenberg Quinlan Rosner 8