Polls, Publics and Pipelines:
Mapping Public Opinion toward the
Keystone XL Pipeline in the United States and the
Northern Gateway Pipeline in Canada
Timothy B. Gravelle
Principal Scientist & Director, Insights Lab
AAPOR
May 16–19, 2013
Boston, MA
© 2000-2013 PriceMetrix Inc. Patents granted and pending.
Background
• Proposed new crude oil pipelines are a current (and contentious)
environmental policy issue in the United States and Canada.
- Keystone XL pipeline (US)
- Northern Gateway pipeline (Canada)
• The politics of pipelines in North America are tied up in debates
about energy self-sufficiency, economic development,
environmental risks, climate change, and Canada–US relations.
• Despite cultural similarities, the US and Canada have long
maintained (and continue to maintain) distinct political cultures
and social values (Lipset 1990; Adams 2003).
2
…but sometimes, we really don't look
that different from one another…
3
Research questions
• What factors shape Americans' and Canadians' attitudes toward
new, high-profile oil pipelines?
- Political variables (party identification, ideology)
- Attitudes toward the economy
- Proximity to the pipeline
- How do how to these factors interact?
4
Hypotheses: Political factors
• Political debates about pipelines have taken on a partisan
dimension in both the US and Canada (political right in favor;
center and left opposed).
• Mass public opinion ought to mirror these cleavages in elite
opinion (Zaller 1992).
5
Hypotheses: Economic attitudes
• A soft rational choice approach would argue that perceptions
economic gains from the pipelines should increase support.
Perceptions of potential economic harm should decrease support
(cf. Downs 1957; Riker 1995).
6
Hypotheses: Proximity
• Different strands of social scientific inquiry suggest that proximity
and distance will play a role in shaping attitudes.
- Durkheim (1899/1978, 1909/1978): social morphology
- NIMBYism and public attitudes toward environmental hazards
(Ansolabehere and Konisky 2009; Kraft and Cleary 1991)
- Proximity effects on attitudes in other policy areas – e.g.,
US immigration (Branton et al. 2007; Dunaway et al. 2010);
attitudes toward the European Union (Berezin and Díez
Medrano 2008; Kuhn 2011)
7
...but we still tend to treat public opinion
as though it comes from here
8
Data
Data: US
• Pew Research Center for the People and the Press
- n = 1,501 (900 landline, 600 cell phone)
- English and Spanish
- 50 states (including Alaska and Hawaii)
- Field period: 8–12 February 2012
10
Data: Canada
• EKOS Research Associates (for WWF–Canada)
- n = 1,500 using a probability-recruited panel (landline and cell
phone)
- Multi-mode (online and phone); phone component is excluded
due to key questions appearing in the online questionnaire
only (n = 1,181)
- English and French
- 10 provinces and 3 territories
- Field period: 20–27 November 2012
11
Data: Pew Center (US)
"How much, if anything, have you heard about the Keystone XL
pipeline that would transport oil from Canada's oil sands region
through the Midwest to refineries in Texas? Have you heard..."
50%
39%
40%
30%
Percent
37%
24%
20%
10%
1%
0%
A lot
A little
Nothing at all
DK/Ref
n = 1,501
12
Data: Pew Center (US)
[If heard a lot/heard a little]:
"Do you think the government should or should not approve the
building of this pipeline?"
70%
66%
60%
50%
Percent
40%
30%
23%
20%
11%
10%
0%
Government should
approve the pipeline
Government should not
approve the pipeline
DK/Ref
n = 1,072
13
Data: EKOS Research (Canada)
"The Enbridge Northern Gateway Pipelines Project is a proposed
pipeline that would export crude oil from Alberta to the west. From
what you know, do you support or oppose the building of this
pipeline?"
40%
28%
30%
Percent
20%
23%
19%
18%
12%
10%
0%
Strongly
support
Somewhat Somewhat
support
oppose
Strongly
oppose
DK/Ref
n = 1,181
14
Method
NGOs have plotted the pipeline
routes in Google Earth using
publicly-available geodetic data
16
Geocoding and distance calculations
• US data: Performed using SAS PROC GEOCODE (ZIP code
geocoding). Secondary matching on county FIPS code was done
using Census Bureau 2010 TIGER/Line shapefiles (county
centroids).
• Canadian data: Performed by matching the forward sortation
area (FSA) to latitude–longitude coordinates created from
Statistics Canada Census 2011 map shapefiles (FSA centroids).
Secondary matching on area code was done using the
AreaCodeWorld Gold database.
• Distances between respondents and pipeline latitude–longitude
coordinates were calculated using the SAS GEODIST function.
17
Regression, diagnostics, multiple imputation,
survey data analysis
• Binary logit (Pew) and ordinal logit (EKOS) models were fit.
• Controls for demographics and other attitudinal variables are
included in the models.
• Regression diagnostics were performed to identify outlying, highinfluence and high-leverage cases.
• Missing data were multiply imputed using IVEware version 0.1 for
SAS (ISR, University of Michigan).
• Multiply imputed datasets were analyzed using SAS PROC
SURVEYLOGISTIC with estimates then being combined using
SAS PROC MIANALYZE.
18
Results
Results: Pew Center (US)
Model 1.1
Model 1.2
b (s.e.) Sig.
b (s.e.) Sig.
Party: Republican
0.58 (0.32)
0.60 (0.32)
Party: Independent
0.27 (0.45)
0.33 (0.46)
Ideology (Liberal)
-0.83 (0.16) ***
Economy is fair
0.25 (0.32)
Economy is poor
0.70 (0.34) *
ln Distance to Pipeline (km)
-0.84 (0.16) ***
-0.03 (0.36)
-0.29 (0.13) *
0.47 (0.37)
-1.62 (0.45) ***
ln Distance × Economy is fair
1.62 (0.47) ***
ln Distance × Economy is poor
1.21 (0.51) *
n
d.f.
Nagelkerke pseudo-R2
Model X2
X2 difference
935
935
18
20
0.43
0.45
317.38
–
***
330.36
***
12.98
**
Note: Binary logit models with controls for sex, age (years logged), education, race/ethnicity, and
attitude toward environmental protection and attitude toward regulation of the oil and gas industry
20
Results: Pew Center (US)
Predicted
Probability
21
Results: EKOS Research (Canada)
Model 2.1
Model 2.2
b (s.e.) Sig.
Party: Conservative
b (s.e.) Sig.
0.89 (0.22) ***
0.90 (0.22) ***
Party: NDP
-0.85 (0.20) ***
-0.85 (0.20) ***
Party: Green
-1.50 (0.29) ***
-1.49 (0.30) ***
Party: Bloc Québécois
-0.83 (0.36) *
-0.84 (0.35) *
Ideology (Liberal)
-0.30 (0.05) ***
-0.29 (0.05) ***
Economy (Index)
0.37 (0.07) ***
0.32 (0.07) ***
ln Distance to Pipeline (km)
0.02 (0.06)
0.05 (0.07)
ln Distance × Economy (Index)
n
d.f.
Nagelkerke pseudo-R2
Model X2
X2 difference
-0.16 (0.08) *
1,008
1,008
14
15
0.38
0.38
442.02
–
***
447.50
5.48
*
Note: Ordinal logit models with controls for sex, age (years logged), education, country of birth
(born in Canada) and visible minority status.
22
Results: EKOS Research (Canada)
Economy = 2.333 (10th percentile)
Cumulative
Probability
Economy = 4 (median)
Cumulative
Probability
Economy = 5.333 (90th percentile)
Cumulative
Probability
23
Key Findings
• In both the US and Canada, attitudes toward oil pipelines are
shaped by political factors and attitudes toward the economy.
• One does not consistently find direct effects of proximity to the
pipeline (occurs in the US but not in Canada).
• Rather, one finds in both countries that the effects of attitudes
toward the economy are contingent upon (moderated by)
proximity to the pipeline.
• Survey data (and our respondents) come from somewhere. In
studying public opinion and policy attitudes, we ignore this at our
peril.
24
The politics of pipelines are sure to remain
current for some time to come….
Toronto Star online, 26 April 2013.
http://www.thestar.com/news/canada/2013/04/26/alberta_explores_possibility_of_oil_pipeline_to_canadas_arctic.html
25
Thank you!
Timothy B. Gravelle
Principal Scientist & Director, Insights Lab
[email protected]
References
Adams, Michael (2003) Fire and Ice: The United States, Canada and the Myth of Converging
Values. Toronto: Penguin.
Ansolabehere, Stephen and David M. Konisky (2009) “Public Attitudes Toward Construction of
New Power Plants.” Public Opinion Quarterly 73(3): 566-577.
Berezin, Mabel and Juan Díez Medrano (2008) “Distance Matters: Place, Political Legitimacy
and Popular Support for European Integration.” Comparative European Politics 6(1): 1–32.
Branton, Regina, Gavin Dillingham, Johanna Dunaway and Beth Miller (2007) “Anglo Voting on
Nativist Ballot Initiatives: The Partisan Impact of Spatial Proximity to the U.S.–Mexico
Border.” Social Science Quarterly 88(3): 882–897.
Downs, Anthony (1957) An Economic Theory of Democracy. New York: Harper.
Dunaway, Johanna, Regina P. Branton and Marisa A. Abrajano (2010) “Agenda Setting, Public
Opinion, and the Issue of Immigration Reform.” Social Science Quarterly 91(2): 359–378.
Durkheim, Émile (1978) “Note on Social Morphology.” Émile Durkheim on Institutional Analysis.
Ed. and trans. Mark Traugott. Chicago: University of Chicago Press.
Durkheim, Émile (1978) “Sociology and the Social Sciences.” Émile Durkheim on Institutional
Analysis. Ed. and trans. Mark Traugott. Chicago: University of Chicago Press.
27
References
Durkheim, Émile (1909) “Sociologie et sciences sociales.” De la méthode dans les sciences.
Paris: Alcan.
Durkheim, Émile (1899) “Morphologie sociale.” Année Sociologique. 2: 520.
Kraft, Michael E. and Bruce B. Cleary (1991) “Citizen Participation and the Nimby Syndrome:
Public Response to Radioactive Waste Disposal.” Political Research Quarterly 44(2): 299328.
Kuhn, Theresa (2011) “Europa ante portas: Border Residence, Transnational Interaction and
Euroscepticism in Germany and France.” European Union Politics 13(1): 94–117.
Lipset, Seymour Martin (1990) Continental Divide: The Values and Institutions of the United
States and Canada. London: Routlege.
Riker, William H. (1995) “The Political Psychology of Rational Choice Theory.” Political
Psychology 16(1): 23-44.
Zaller, John R. (1992) The Nature and Origins of Mass Opinion. Cambridge: Cambridge
University Press.
28