Online Appendix for 'Winners and Losers: The Psychology of Foreign Trade'

Table of Contents

Chapter 2

Chapter 3

Chapter 4

Chapter 5

Chapter 6

Chapter 7

Chapter 8

Chapter 9

Chapter 10

Chapter 11

Chapter 12

Chapter 13

Appendix Content

Table A2.1 Respondents’ Understanding of Their Position in the Global Economy Based on Their Job

Believe employer importsBelieve employer part of supply chainBelieve employer outsources
Believe employer exports0.5030.4890.492
(n)(1605)(1531)(1621)
Believe employer imports0.4920.365
(n)(1496)(1576)
Believe employer part of supply chain0.423
(n)(1520)

Source: 2004 NAES Election Survey.

Note: Entries are correlation coefficients based on Cohen’s Kappa.

Table A2.2 Economic Knowledge Does Not Account for Education’s Impact on Trade Support

Model 1Model 1Model 1Model 1Model 2Model 2Model 2Model 2
Coeff.SEt-valuep-valueCoeff.SEt-valuep-value
White‒0.0460.012‒3.7930.000‒0.0480.012‒3.9300.000
Family income0.0670.0222.9830.0030.0620.0222.7490.006
Age0.0280.0151.9000.0580.0260.0151.7400.082
Democrat0.0700.0272.6420.0080.0710.0272.6690.008
Republican0.0400.0271.4820.1380.0400.0271.4970.135
Empathy‒0.0500.033‒1.5340.125‒0.0530.033‒1.6290.103
National superiority‒0.0730.023‒3.1790.002‒0.0700.023‒3.0600.002
Risk avoidance‒0.0300.028‒1.0670.286‒0.0280.028‒1.0110.312
Social dominance orientation‒0.1800.029‒6.1520.000‒0.1800.029‒6.1820.000
Unemployed0.0200.0191.0540.2920.0220.0191.1560.248
Perceived strength of social welfare system0.1110.0205.7020.0000.1090.0195.5790.000
Female‒0.0400.010‒3.9190.000‒0.0380.010‒3.7010.000
Education0.1320.0177.8850.0000.1250.0177.3270.000
Economic knowledge0.0160.0062.6730.008
Constant0.5420.04412.2650.0000.5360.04412.1150.000

Source: US Trade Survey-Experiment, 2013–2014 (GfK Ltd.)

Note: All variables have been rescaled to range from 0 to 1.

Table A2.3 Effects of the Perceived Influence of Trade on Home Country and Trading Partner Countries on Trade Policy Support

Model 1Model 1Model 1Model 1Model 2Model 2Model 2Model 2
Coeff.SEt-valuep-valueCoeff.SEt-valuep-value
Index of perceived effects on US0.1890.00540.1810.0000.1790.00537.7960.000
Index of perceived effects on trading partners0.0590.00610.0860.0000.0610.00610.5840.000
White‒0.0180.007‒2.5890.010
Household income0.0300.0132.3470.019
Age0.0180.0082.2110.027
Democrat0.0480.0143.4500.001
Republican‒0.0070.014‒0.4690.639
Unemployed‒0.0030.010‒0.2720.786
Female0.0080.0061.3840.167
Education0.0690.0107.1560.000
Constant‒0.1350.018‒7.3200.000‒0.1980.024‒8.3080.000
(n)(3042)(3042)
R-square0.440.470

Source: US Trade Survey-Experiment, 2013–2014 (GfK Ltd.).

Note: All variables have been rescaled to range from 0 to 1.

Table A3.1 Effect of Party-in-Power on Change in Trade Support, Four Wave Panel, October 2016–June 2019

Coef.Std. err.tp-value
August 20170.5660.03814.9300.000
October 20181.0910.04226.2800.000
June 20190.9970.04323.4100.000
August 2017 × Democrat‒0.4990.050‒10.0200.000
October 2018 × Democrat‒0.9110.055‒16.6300.000
June 2019 × Democrat‒0.8200.057‒14.4800.000
Constant2.9980.016189.2800.000

Source: ISCAP Panel Survey (Amerispeak/NORC).

Note: Analyses show fixed effects coefficients with October 2016 as the baseline wave.

Table A3.2 Effect of Party-in-Power on Change in Trade Support, Controlling for Perceptions of Change in the National Economy, Four Wave Panel, October 2016–June 2019

Coef.Std. Err.t-valuep-value
August 20170.3970.03910.0900.000
October 20180.8310.04518.4700.000
June 20190.7520.04616.5200.000
August 2017 × Democrat‒0.2390.053‒4.5400.000
October 2018 × Democrat‒0.5540.060‒9.2500.000
June 2019 × Democrat‒0.4330.062‒6.9400.000
National Economy0.1590.01213.6800.000
Constant2.5140.03964.9600.000

Note: Analyses show fixed effects coefficients with October 2016 as the baseline wave.

Source: ISCAP Panel Survey (Amerispeak/NORC).

Table A3.3 Effects of Party-in-Power on Change in Trade Support, Controlling for Perceptions of Change in the National Economy and Change in Trust in Government, Four Wave Panel, October 2016–June 2019

Coef.Std. errt-valuep-value
August 20170.3970.03910.0900.000
October 20180.8260.04518.3300.000
June 20190.7520.04516.5300.000
August 2017 × Democrat‒0.2310.053‒4.3700.000
October 2018 × Democrat‒0.5380.060‒8.9400.000
June 2019 × Democrat‒0.4250.063‒6.7900.000
National economy0.1540.01213.0500.000
Trust in government0.2020.0892.2700.023
Constant2.4690.04356.8500.000

Note: Analyses show fixed effects coefficients with October 2016 is the baseline wave.

Source: ISCAP Panel Survey (Amerispeak/NORC).

Table A3.4 Effect of Party-in-Power on Change in Trade Support, Three Wave Panel, 2008–2012–2014

Coef.Std. Err.t-valuep-value
20120.0250.0241.0300.302
20140.0410.0281.4500.147
2012 × Democrat0.1280.0323.9500.000
2014 × Democrat0.1410.0383.6700.000
Constant2.3340.005510.8500.000

Note: Entries represent fixed effects regression coefficients with robust standard errors. October 2008 is the baseline wave.

Source: GfK Panel Surveys.

Table A3.5 Effect of Party-in-Power on Change in Trade Support, Controlling for Change in Perceptions of National Economy, Three Wave Panel, 2008–2012–2014

Coef.Std. err.t-valuep-value
20120.0030.0240.1100.916
2014‒0.0010.030‒0.0200.981
2012 × Democrat0.0510.0361.3900.164
2014 × Democrat0.0840.0402.0800.038
National economy0.0460.0104.5700.000
Constant2.2720.014158.8000.000

Note: Entries represent fixed effects regression coefficients with robust standard errors. October 2008 is the baseline wave.

Source: GfK Panel Surveys.

Table A3.6 Tests of Presidential Opinion Leadership on Trade: Change in Support for Trade by Party, Voting Intentions, and Trump Support, 2016–2020

Model 1: PartyModel 2: VotingModel 3: Support
Estimates (S.E.)Estimates (S.E.)Estimates (S.E.)
Wave (baseline=2016)
2017		0.017**
(0.007)		0.021***
(0.007)		0.033***
(0.006
20180.044***
(0.008)		0.059***
(0.008)		0.085***
(0.007)
20190.046***
(0.008)		0.058***
(0.008)		0.077***
(0.007)
20200.051***
(0.008)		0.066***
(0.008)		0.087***
(0.007)
Party × Wave
Republican × 2017		0.154***
(0.012)
Republican × 20180.255***
(0.013)
Republican × 20190.227***
(0.013)
Republican × 20200.244***
(0.013)
Trump voter × Wave
Trump voter × 2017		0.168***
(0.013)
Trump voter × 20180.283***
(0.014)
Trump voter × 20190.256***
(0.015)
Trump voter × 20200.275***
(0.015)
Rated Trump high × Wave
Rated Trump high × 2017		0.142***
(0.014)
Rated Trump high × 20180.246***
(0.015)
Rated Trump high × 20190.230***
(0.015)
Rated Trump high × 20200.231***
(0.016)
Constant0.494***
(0.004)		0.500***
(0.004)		0.517***
(0.004)
Sigma_u0.2200.2100.216
Sigma_e0.1940.1970.196
Rho0.5600.5310.548
N5,5893,1836,229

Note: Entries represent fixed effects regression coefficients with robust standard errors in parentheses. *p<.05, **p<.01, ***p<.001. Support for Trade is continuous from 0 to 1, with higher values connoting higher support. Independent variables are all binary variables coded 0 or 1, where 1 indicates being in the group named in the interaction term. Baseline reference wave is October 2016.

Source: ISCAP Panel Survey (Amerispeak/NORC).

Table A3.7 Tests of Presidential Opinion Leadership on China: Change in Support for China by Party, Voting Intentions, and Trump Support, 2016–2020

Model 1: PartyModel 2: VotingModel 3: Support
Estimates (S.E.)Estimates (S.E.)Estimates (S.E.)
Wave (baseline=2016)
2017		0.031***
(0.010)		0.031***
(0.009)		0.029***
(0.008)
20180.020**
(0.010)		0.026***
(0.010)		0.013
(0.009)
20190.049***
(0.010)		0.041***
(0.010)		0.032***
(0.009)
20200.049
(0.011)		0.010
(0.011)		0.009
(0.009)
Party × Wave
Republican × 2017		0.030*
(0.016)
Republican × 20180.007
(0.016)
Republican × 2019‒0.016
(0.017)
Republican × 20200.025
(0.017)
Trump voter × Wave
Trump voter × 2017		0.031*
(0.017)
Trump voter × 20180.003
(0.018)
Trump voter × 2019‒0.012
(0.019)
Trump voter × 20200.009
(0.019)
Rated Trump high × Wave
Rated Trump high × 2		0.051***
(0.018)
Rated Trump high × 30.022
(0.019)
Rated Trump high × 40.029
(0.020)
Rated Trump high × 50.038*
(0.020)
Constant0.404***
(0.006)		0.407***
(0.005)		0.409***
(0.006)
Sigma_u0.2800.2640.279
Sigma_e0.2550.2510.256
Rho0.5460.5250.544
N5,6183,1996,261

Note: Entries represent fixed effects regression coefficients with robust standard errors in parentheses. Support for Support for China is continuous from 0 to 1, with higher values connoting higher support. . *p<.05, **p<.01, ***p<.001. Independent variables are all binary variables coded 0 or 1, where 1 indicates being in the group named in the interaction term.

Source: ISCAP Panel Survey (Amerispeak/NORC).

Table A3.8 Test of Candidate Opinion Leadership on Trade: Change in Republican Support for Trade by Primary Candidate Preference, January 2016–October 2016

Coef.set-valuep-value
Wave‒0.2480.043‒5.7000.000
Trump supporter × Wave‒0.0090.071‒0.1300.900
Constant2.2300.022102.9800.000
sigma_u0.705
sigma_e0.545
rho0.626
n=687 respondents,
2 observations per respondent

Note: Entries represent fixed effects regression coefficients with robust standard errors.

Source: ISCAP Panel Survey (Amerispeak/NORC).

Table A3.9. What Predicts Individual-level Increases in Trade Support from 2016 to 2020?

t-valuep-value
Coef.Std. errort-valuep-value
Republican0.4980.0945.2870.000
Belief in value of competition0.7550.1644.6130.000
Social dominance orientation0.4630.1982.3350.020
Trump thermometer rating0.8010.1405.7160.000
Education0.1750.1211.4380.151
constant‒0.6190.147‒4.2120.000

Note: Dependent variable is change in Trade Preference from 2016 to 2020 among panel respondents, where high scores indicate shifting in a more positive direction. Independent variables represent values as assessed in 2016, with the exception of Belief in the value of competition, which was asked in 2020. N=1300.

Source: ISCAP Panel Survey (Amerispeak/NORC).

Table A6.1 Trade Support by Race of Assumed Major Trading Partners, Combined Sample and by Race

Combined SampleCombined SampleWhitesWhitesNon-whitesNon-whites
Coeff.(se)Coeff.(se)Coeff.(se)
Proportion
dark-skinned countries		‒.10(.02)***‒.11(.02)***‒.05(.04)
Accuracy.03(.01)***.03(.01)***.04(.01)***
Education.03(.01)***.04(.01)***.01(.01)
White‒.02(.01)––––
constant.45(.02)***.42(.03)***.47(.04)***
N28022057744

Note: *** p ≤ .001.

Source: ISCAP Panel Survey (Amerispeak/NORC).

Table A7.1 Ingroup Favoritism: Trade Support by Who Wins versus Loses

Partial SSdfF
Model660.7711646.56***
US
(vs. Canada)		1.49811.69
Trading partner loses / Home country gains
(vs. Trading partner gains / Home country loses)		424.1071478.15***
US
× Trading partner loses / Home country gains		87.828199.02***
Magnitude of trade's impact23.16664.35***
Magnitude of impact
× Trading partner loses / Home country gains		75.340614.16***
Pre-experiment trade preference14.248116.06***
Residual2477.3392793
Total3138.1102809
Number of Observations2810

Source: US Trade Survey-Experiment, 2013‒2014 and Canadian Trade Survey-Experiment 2015.

Note: The table presents the results of an analysis of variance. ***p<.001.

Table A7.2 Social Dominance Orientation Exacerbates Ingroup Favoritism

Pooled SamplePartial SSdfF
Model556.8801835.38***
US (vs. Canada)2.86213.27
Trading partner loses / Home country gains
(vs. Trading partner gains / Home country loses)		346.9341396.77***
US × Trading partner loses / Home country gains66.405175.94***
High Social Dominance Orientation (vs. Low SDO)22.328125.53***
High SDO
× Trading partner loses / Home country gains		11.060112.65***
Magnitude of trade's impact23.55364.49***
Magnitude of impact
× Trading partner gains / Home country gains		73.855614.08***
Pre-experiment trade preference20.676123.65***
Residual2102.0552,404
Total2658.9342,422
Number of observations2423

Source: US Trade Survey-Experiment, 2013‒2014 and Canadian Trade Survey-Experiment 2015.

Note: The table presents the results of an analysis of variance. ***p<.001.

Table A7.3 Zero-Sum Perceptions Exacerbate Ingroup Favoritism

Pooled samplePartial SSdfF
Model707.2351845.15***
US (vs. Canada)0.27810.32
Trading partner loses / Home country gains
(vs. Trading partner gains / Home country loses)		455.6221523.54***
US × Trading partner loses / Home country gains57.032165.53***
Zero-sum perceptions
(vs. No zero-sum perception)		13.236115.21***
Zero-sum perceptions
× Trading partner loses / Home country gains		34.848140.04***
Magnitude of trade's impact23.85364.57***
Magnitude of impact
× Trading partner gains / Home country gains		74.378614.24***
Pre-experiment trade preference7.69618.84**
Residual2424.5562,786
Total3131.7912,804
Number of Observations2805

Source: US Trade Survey-Experiment, 2013–2014 and Canadian Trade Survey-Experiment 2015.

Note: The table presents the results of an analysis of variance. **p <.01, ***p<.001.

Table A7.4 Outgroup Indifference: Trade Support by Whether the Trading Partner Also Wins

Partial SSdfF
Model252.5041619.17***
US
(vs. Canada)		1.17511.43
Trading partner gains / Home country gains
(vs. Trading partner loses / Home country gains)		44.943154.59***
US
× Trading partner gains / Home country gains		57.742170.14***
Magnitude of trade's impact95.264619.29***
Magnitude of impact
× Trading partner gains / Home country gains		13.16362.66*
Pre-experiment trade preference35.683143.35***
Residual2351.9632857
Total2604.4672873
Number of Observations2874

Source: US Trade Survey-Experiment, 2013–2014 and Canadian Trade Survey-Experiment 2015.

Note: The table presents the results of an analysis of variance. *p<.05, ***p<.001.

Table A7.5 Social Dominance Orientation Exacerbates Outgroup Indifference

Pooled SamplePartial SSdfF
Model275.0201818.81***
US (vs. Canada)0.17410.21
Trading partner gains / Home country gains
(vs. Trading partner loses / Home country gains)		44.418154.67***
US × Trading partner gains / Home country gains40.828150.25***
High Social Dominance Orientation (vs. Low SDO)9.596111.81***
High SDO
× Trading partner gains / Home country gains		26.746132.92***
Magnitude of trade's impact81.549616.73***
Magnitude of impact
× Trading partner gains / Home country gains		17.04963.5**
Pre-experiment trade preference36.951145.48***
Residual1995.4592,456
Total2270.4782,474
Number of observations2475

Source: US Trade Survey-Experiment, 2013–2014 and Canadian Trade Survey-Experiment 2015.

Note: The table presents the results of an analysis of variance. **p<.01, ***p<.001.

Table A7.6 Zero-Sum Perceptions Exacerbate Outgroup Indifference

Pooled SamplePartial SSdfF
Model261.1871817.69***
US (vs. Canada)1.23911.51
Trading partner gains / Home country gains
(vs. Trading partner loses / Home country gains)		31.769138.72***
US × Trading partner gains / Home country gains42.793152.16***
Zero-sum perceptions
(vs. No zero-sum perceptions)		0.58910.72
Zero-Sum Perceptions
× Trading partner gains / Home country gains		11.820114.41***
Magnitude of trade's impact94.609619.22***
Magnitude of impact
× Trading partner gains / Home country gains		12.63962.57*
Pre-experiment trade preference31.731138.68***
Residual2337.3572,849
Total2598.5442,867
Number of observations2868

Source: US Trade Survey-Experiment, 2013–2014 and Canadian Trade Survey-Experiment 2015.

Note: The table presents the results of an analysis of variance. *p<.05, ***p<.001.

Table A7.7 Predictors of Support for International Trade, by Country

CanadaCanadaCanadaCanadaUnited StatesUnited StatesUnited StatesUnited States
BStd. errort-valuep-valueBStd. errort-valuep-value
Age‒0.0050.015‒0.3520.7250.0130.0140.9920.321
Female‒0.0190.009‒2.1600.031‒0.0410.009‒4.4270.000
Education0.0240.0161.4840.1380.1130.0167.1800.000
Household income0.0640.0183.6610.0000.0500.0212.3860.017
Strength of social welfare system0.1220.0196.4530.0000.0830.0184.5660.000
National superiority0.0760.0223.5350.000‒0.0490.021‒2.3140.021
Social dominance orientation‒0.0580.024‒2.3800.017‒0.1130.025‒4.4480.000
Perceive trade as zero-sum‒0.0990.010‒10.2080.000‒0.0670.009‒7.4360.000
Non-economic Isolationism‒0.2480.024‒10.5410.000‒0.2940.021‒13.8320.000
(Constant)0.6490.02822.7970.0000.7150.03520.3470.000
Conservative Party0.0540.0124.5360.000Democrat0.0350.0251.4250.154
Liberal Party0.0440.0123.8240.000Republican0.0180.0250.7250.468
NDP‒0.0340.014‒2.3710.018White‒0.0420.011‒3.7300.000
Parti Quebecois‒0.0550.030‒1.8540.064
Green Party‒0.0430.027‒1.5960.111
R-square0.2700.260
(n)14141673

Source: US Trade Survey-Experiment, 2013‒2014 and Canadian Trade Survey-Experiment 2015.

Note: All variables have been standardized to vary between 0 and 1.

Table A8.1 Effects of the Tone of Newspaper and Television Coverage of Trade on Perceptions of Trade’s Impact on the Nation

OLS modelsOLS modelsOLS modelsInstrumental variables regressionInstrumental variables regressionInstrumental variables regression
Model 1Model 1Model 2Model 2Model 3Model 3Model 4Model 4
Coef. (se)Coef. (se)Coef. (se)Coef. (se)Coef. (se)Coef. (se)Coef. (se)Coef. (se)
Education‒0.022(0.037)‒0.025(0.036)
Export-oriented sector‒.136(.171)‒0.163(0.165)
Non-tradable sector0.183(0.121)0.150(0.124)
Tariff rate‒.037(0.023)‒0.031(0.024)
Manufacturing industry‒0.565(0.539)‒0.586(0.532)‒0.560(0.737)‒0.592(0.721)
Information industry5.628*(2.520)5.749*(2.566)5.571*(2.511)5.712*(2.560)
Union member‒0.238**(0.071)‒0.233**(0.072)‒0.232*(0.107)‒0.228*(0.107)
Unemployed‒0.241(0.141)‒0.241(0.145)‒0.241(0.205)‒0.246(0.206)
Partisanship0.057***(0.015)0.058***(0.015)0.062**(0.022)0.063**(0.022)
TV slant0.102*(0.045)0.103*(0.044)0.106*(0.040)0.108**(0.039)
Newspaper slant0.204**(0.072)0.208**(0.072)0.196**(0.071)0.200**(0.070)
Perceived effects of trade on self0.670***(0.035)0.670***(0.035)0.657*(0.287)0.647*(0.279)
Male0.106(0.069)0.102(0.067)
Age-0.002(0.002)‒0.002(0.002)‒0.002(0.002)‒0.002(0.002)
Income0.016(0.027)0.016(0.027)0.013(0.030)0.013(0.030)
Constant0.502*(0.241)0.510*(0.235)0.563(0.830)0.584(0.803)
R-square0.3150.3140.3130.312
(n)1071107110711071

Source: Data are from NAES 2004 Election Study.

Table A8.2 Effects of Change in Exposure to Unemployment News on Change in Perceived Effects of Trade on the U.S. Economy

Model 1Model 2Model 3Model 4
Number of stories on unemployment (in previous 6 months)‒0.010*
(0.004)		‒0.009*
(0.004)		‒0.008*
(0.004)		‒0.008*
(0.004)
Personal unemployment‒0.274*
(0.133)		‒0.172
(0.126)		‒0.178
(0.126)
Perceived effect of trade on personal finances0.402***
(0.039)		0.397***
(0.039)
Republican (2007) by wave‒0.200*
(0.087)
Wave‒0.246***
(0.066)		‒0.231***
(0.067)		‒0.147*
(0.064)		‒0.098
(0.067)
Constant2.891***
(0.042)		2.900***
(0.042)		1.705***
(0.122)		1.719***
(0.122)
R20.0860.0910.1890.193
(n)921921920920

Source: Data are from 2007–2009 Mansfield and Mutz Panel study.

Note: Entries are fixed effects regression estimates with standard errors in parentheses. *p< 0.05, **p<0.01, ***p<0.001.

Table A9.1 Results for Pilot Experiment Using Mechanical Turk Sample

Job loss due to trade(se)Job loss due to automation(se)p-value for mean difference
Negative emotions.63(.02).58(.02).062
Support for trade.54(.02).59(.02).044
Believe jobs could return.69(.03).61(.03).015
Acceptability of asserting American superiority.32(.02).24(.02).001
Trustworthiness of US versus China.51(.02).46(.02).012
Trustworthy US.67(.02).55(.02).000
Trustworthy China.47(.02).45(.02).280
Trade versus Automation at fault.45(.02).36(.02).000
Sample n125126

Note: Entries are means and standard errors by experimental condition. All variables are on a 0 to 1 scale. p-values are for directional predictions (one-tailed tests).

Source: Pilot data from Mechanical Turk.

Table A9.2 Effects of Job Loss Due to Trade versus Automation

Model 1Model 1Model 2Model 2Model 3Model 3Model 4Model 4
Negative emotions(se)Support for trade(se)Believe jobs could return(se)Asserting American superiority(se)
Due to automation‒.011(.017).047**(.016)‒.042#(.024)‒.027#(.017)
Due to trade.087 ***(.017)‒.011(.016).019(.024).039*(.017)
Constant.588***(.012).592***(.012).533***.017.235***(.012)
Sample n1007100810091009

Source: Attributing the Causes of Job Loss, (TESS, Amerispeak/NORC).

Note: Entries are OLS regression coefficients representing random assignment to experimental attribution treatments. ***p<.001, **p<.01, *p<.05, #p<.10.

Table A9.3 Effects of Automation and Trade Attributions Relative to Control Condition, by College Education

Support for tradeSupport for tradeSupport for tradeNegative emotionsNegative emotionsNegative emotionsBelief that jobs can come backBelief that jobs can come backBelief that jobs can come backBelief in American superiorityBelief in American superiorityBelief in American superiority
Coef.sep-valueCoef.sep-valueCoef.sep-valueCoef.sep-value
Automation0.0630.0200.0020.0170.0210.415‒0.0570.0290.051‒0.0540.0210.010
Trade0.0130.0200.5270.0850.0210.000‒0.0170.0300.5670.0420.0210.051
College Education0.1480.0230.000‒0.1030.0240.000‒0.1810.0340.000‒0.0680.0250.006
Automation × College‒0.0470.0320.144‒0.0690.0340.0420.0490.0480.3060.0740.0350.033
Trade × College‒0.0660.0330.0460.0060.0340.8550.0990.0490.043‒0.0060.0350.862
(Constant)0.5370.0140.0000.6260.0150.0000.6000.0210.0000.2600.0150.000

Source: Attributing the Causes of Job Loss, (TESS, Amerispeak/NORC).

Note: Entries are OLS regression coefficients representing random assignment to experimental attribution treatments.

Table A10.1 Trade Policy Attitudes as Predicted by Perceived Effects of Trade on Family, on Local Community, and on Nation as a Whole

Coef.SEBetat-valuep-value
Perceived effects on national economy0.2820.0190.33114.9150.000
Perceived effects on local community0.0940.0220.0924.1980.000
Perceived effects on family0.0480.0250.0421.8810.060
Constant1.8140.1909.5320.000

Source: ISCAP Panel Surveys.

Note: Control variables also included in this model included age, gender, income, and education along with dummy variables representing Republican and Democratic party identification. All perceived effects questions were asked on similar 5-point scales where high = trade helps.

Table A11.1 Relative Thermometer Advantage of Republican Over Democratic Candidate, 2012–2016, Fixed Effects Analysis of Change in Independent Variable on Change in Relative Thermometer Advantage (R-D)

Coef.Std. Err.tp-value
Party (Democrat)‒0.9090.211‒4.3000.000
Family income‒0.0190.044‒0.4400.663
Looking for work‒0.4220.491‒0.8600.390
Personal finances‒0.0090.130‒0.0700.942
Trade opinion (pro)‒0.0190.074‒0.2600.795
Immigration opinion (pro)‒0.0530.083‒0.6500.519
China opinion (pro)0.1370.0781.7500.081
Relative Distance from candidates (Republican-Democrat)Relative Distance from candidates (Republican-Democrat)Relative Distance from candidates (Republican-Democrat)Relative Distance from candidates (Republican-Democrat)
On China‒0.3040.049‒6.2300.000
On trade‒0.1850.050‒3.7000.000
On immigration‒0.1620.050‒3.2500.001
Social Dominance Orientation0.1400.0702.0000.046
National economy (good)‒0.7320.119‒6.1300.000
Unemployment in area0.0050.0640.0700.943
Manufacturing in area0.0110.0190.5600.577
Incomes in area‒0.0110.006‒1.8300.068
Wave0.0080.6580.0100.991
Constant13.3570.87015.3600.000

Note: All coefficients represent the relationship between change over time in each independent variable on change over time in Relative Thermometer ratings in the two election years, both in October. Fixed effects panel regression.

Source: ISCAP Panel Surveys, 2012-2016 (GfK Ltd.)

Table A11.2 Effects of Issue Opinion Change and Candidate Position Changes on Support for Trump, Among Less and More Sophisticated Voters on Easy and Hard Issues, 2012–2016

Main effects of change on change in thermometersMain effects of change on change in thermometersMain effects of change on change in thermometersMain effects of change on change in thermometersInteractions of change with Sophisticated (1) versus Less sophisticated (0)Interactions of change with Sophisticated (1) versus Less sophisticated (0)Interactions of change with Sophisticated (1) versus Less sophisticated (0)Interactions of change with Sophisticated (1) versus Less sophisticated (0)
Change in:Coef.S.E.t-valuet-valueCoef.S.E.t-valuet-value
Party ID (Democrat)‒0.0460.012‒3.920***0.0060.0220.260
Trade (oppose)0.0170.0270.630‒0.0330.043‒0.780
Relative Republican closeness0.1060.0176.160***0.0030.0280.090
(Democrat-Republican distance)
Healthcare (oppose)0.0530.0301.800‒0.1530.057‒2.680**
Relative Republican closeness0.0740.0184.200***0.0930.0322.890**
(Democrat-Republican distance)
Wave‒0.0350.006‒5.570***
Constant0.6110.02821.950***
Sigma u.192
Sigma e.131
Rho.682
(n)1,194

Note: Coefficients represent the extent to which a change in the independent variable is related to a change in the dependent variable at the individual level. Results are based on fixed effects regression analysis estimating change in Relative Thermometer Ratings as the dependent variable, in relation to change over time in each independent variable. High scores indicate greater Republican candidate preference. Robust standard errors are shown. All variables are recoded to range between 0 and 1.

Source: ISCAP Panel Surveys, 2012-2016 (GfK Ltd.)

Table A13.1 Relationship Between Perceived Impact of Trade on International Relationships and on Job Availability as Predictors of Trade Policy Support

Coeff.(SE)t-valuep-value
Perceived impact of trade on international relationships0.3490.01819.7430.000
Perceived impact of trade on jobs available in the US0.0910.0146.5390.000
Party Identification (R)0.1140.0129.8520.000
Constant0.2680.01617.2310.000

Note: Perceived impact on jobs and on international relationships are both measured on 5-point scales in which high represents a positive impact.

Source: ISCAP Panel Survey, June 2019 (Amerispeak/NORC).

*p<.05, **p<.01, ***p<.001