13. See Mayer 2016. For information on our calculations of billionaires’ contributions, see note 62 to chap. 2.
14. As noted in chap. 2, our searches had to cover a number of years in order to avoid missing relevant statements. We watched carefully for any changes in individuals’ stands over the ten- year period but did not detect any in these policy areas.
15. The marriage search terms included “marriage equality,” “same- sex marriage,” and “gay marriage.” Since different actors use different terms to signal their divergent attitudes, it is necessary to consider more than one term on this issue.
16. Given the large number of states that have passed highly restrictive abortion laws, we consider the position that abortion should be a “state issue” to be antiabortion.
17. Newport 2016; Saad 2014.
18. Kramer 2014.
19. Mehta and Mishak 2010.
20. Sarkar 2015.
21. Baker 2010.
22. Lowery 2014.
23. Mundy 2012.
24. Mishra and Ganguli 2009.
25. CBS New York 2012.
26. On social liberalism among the top 4 percent or so of income earners, see Page and Hennessy 2010.
27. See chap. 2, table 2.6.
28. For total statements, visibility b = 0.43, adjusted r- squared = 0.24; for total actions, visibility b = 0.13, adjusted r- squared = 0.30 .
172
notes to pages 90–95
29. As noted earlier, our “heir” measure might more accurately be described as “nonentrepreneur.” Rather than attempting to untangle the complicated issues of exactly how much each billionaire got from his or her parents by gift or inheritance, and when he or she did so, we simply coded as an heir anyone who did not participate in starting his or her own business.
30. A dichotomous consumer- exposure variable was measured by identifying whether or not the primary business through which a billionaire acquired his or her wealth markets products directly to public consumers.
31. A complex regression analysis involving nine independent variables produced an adjusted r- squared of 0.26, and highly significant coefficients for making at least one relevant statement and for the interactions of statement- making with wealth level, being an heir, and level of visibility. But only eight billionaires actually took actions; most were coded zero as a middle category between pro and con.
The independent variables were highly collinear.
32. These results suggest that the meaning of wealth effects is different for economic issues (where greater wealth entails increased liberalism) than for social issues (where wealth is mainly relevant as a factor that increases prominence and available resources).
33. Regression analyses that “control” for posttreatment variables, i.e., for variables that are caused by x and influence y, will ordinarily produce biased estimates of the effect of x on y (Rosenbaum 1984). If the posttreatment variable captures the main causal pathway by which x causes y, then the causal inference will be biased toward zero.
34. Brenner 1992.
35. But most Americans do support a pathway to citizenship (contingent on meeting certain requirements) for those already in the US illegally (Jones 2016; Smeltz et al. 2016, 18). See Pew 2015. On attitudes concerning high- vs. low- skilled immigrants, see Hainmueller and Hiscox 2010.
36. Page and Bouton 2006, 41, 50, 185. A hefty 74 percent said “very important”
in 1994; majorities continued to do so in four other surveys through 2004.
37. Fully 72 percent of Americans saw large numbers of immigrants as a “critical threat” in economically stressed 1994, but majorities continued to perceive a threat in every survey in which this question was asked until 2012.
38. Sides, Tesler, and Vavreck 2017.
39. See Ferguson and Page 2017.
40. In one analysis of 2002 CCGA data, for example, Americans’ judgments
about the importance of controlling and reducing illegal immigration were significantly and independently predicted by their assessments of terrorism as a critical threat; by the importance they attributed to the goals of protecting American jobs or reducing the trade deficit; and by their feelings (controlling for those security and economic factors) about Mexico (Page and Bouton 2006, 189).
41. Bound, Khanna, and Morales 2017.
notes to pages 95–102
173
42. The clearest case is probably that of wealthy farmers or owners of agribusi-nesses, who profit from cheap immigrant farm laborers. But very few if any owners of agricultural businesses have made it into the ranks of the very wealthiest billionaires, so we have not studied them.
43. Broockman, Ferenstein, and Malhotra 2017.
44. Enos (2014) finds that contact with Hispanics can induce a sense of group threat, in turn driving down support for immigration. Therefore, more sheltered lifestyles among the wealthy, with lower levels of intergroup contact, may result in more pro- immigration attitudes. Change in contact may be key.
45. In the SESA survey of multimillionaires, only 10 percent of respondents wanted to reduce legal immigration from current levels, whereas a total of 80 percent wanted to increase it (35 percent) or keep it at its current levels (45 percent).
46. CCGA survey data on the immigration views of the American public and
foreign policy leaders have regularly found especially large “gaps” between the opinions of the two groups on economic policies. The gap on immigration may have reached a peak in the mid- 1990s (under a Democratic administration), when 57 percent of the public wanted to decrease levels of legal immigration but only 8 percent of foreign policy leaders agreed (Page and Bouton 2006, 212– 13). But large gaps have appeared every time the Chicago Council has looked.
47. Kaufmann 2016; Ferguson and Page 2017; Inglehart and Norris 2016.
48. Page and Gilens 2017, 129– 30.
49. Even as late as 2016, the Republican Party Platform continued to oppose same- sex marriage (Naylor 2016).
50. For example, billionaires Paul Singer and Seth Klarman have devoted considerable resources to advocating for stronger LGBT rights (Palmer and Sherman 2014).
51. See Bartels 2016; Page and Gilens 2017.
Chapter Five: Reshaping State and Local Politics
1. US Office of Management and Budget, n.d.; Urban Institute 2017.
2. National Federation of Independent Business v. Sebelius, 567 US 1 (2012).
3. Hertel- Fernandez 2016.
4. Schattschneider 1960, chap. 2.
5. McConnell 1967, especially chaps. 4, 6.
6. Klarner 2015.
7. Robbins 2017.
8. Center for Responsive Politics, n.d.
9. Kranish 2015.
10. National Conference of State Legislatures, n.d.
11. Nebraska’s unicameral legislature is technically nonpartisan, but Rep ub lican-affiliated politicians make up a strong majority of it and are led by Republican
174
notes to pages 102–106
Pete Ricketts, a member of the billionaire Ricketts family. See Stoddard and Nohr 2016.
12. National Conference of State Legislatures, n.d.
13. Kron 2012.
14. See Page and Gilens 2017, chap. 6. At the federal level it is generally easy for wealthy contributors to stop new legislation they oppose, but much harder hard to win new legislation they favor.
15. Weir, Zubak- Skees, and Wieder 2017.
16. Blumenthal 2015a.
17. Blumenthal 2015a; Harris 2012.
18. Blumenthal 2015a.
19. Byrne 2014.
20. Blumenthal 2015b.
21. Jeffers 2015; Blumenthal 2015c.
22. Blumenthal 2015c.
23. The Koch network has a complex structure. It includes many organizations under the Americans for Prosperity rubric (see Skocpol and Hertel- Fernandez, forthcoming.)
24. Skocpol and Hertel- Fernandez 2016; Sclar et al. 2016.
25. Gold 2015 .
> 26. Hertel- Fernandez 2016 .
27. Debenedetti 2017 .
28. For details on Mercer’s key role in Trump’s campaign and in installing Steve Bannon as head of the campaign, see Ferguson, Jorgensen, and Chen 2017 .
29. Vogel and Peters 2017.
30. Skocpol and Hertel- Fernandez 2016; Hertel- Fernandez, Skocpol, and Lynch 2016 .
31. Kaiser Commission on Medicaid and the Uninsured 2017.
32. Hertel- Fernandez 2014.
33. Scola 2012; Hertel- Fernandez 2014.
34. Sclar et al. 2014.
35. Mayer 2016, 345, 347– 48.
36. Phillips 2016; http://www.ncsl.org/Portals/1/Documents/Elections/Legis _Control_2017_August_4th_10am_26973.pdf.
37. Confessore 2014.
38. Mayer 2016.
39. Fausset 2015.
40. Hertel- Fernandez 2016 .
41. Skocpol and Hertel- Fernandez 2016; Hertel- Fernandez, Skocpol, and Lynch 2016 .
42. West 2014.
43. Confessore 2015b. On Rauner’s income and wealth, see Grimm 2016.
notes to pages 107–113
175
44. Garcia 2017; Rhodes 2017.
45. Forbes 2017b (US); Janssen 2017.
46. Ehrenfreund 2015; Smith and Bosman 2017.
47. Samuels 2015; Umhoefer 2016.
48. Mayer 2016, chap. 13.
49. Straus 2016.
50. Kranish 2015.
51. MacLean 2017, 230; Fischer 2016.
52. Blumenthal 2015a.
53. Byrne and Perez 2016.
54. Keefe 2013; Green 2015.
55. Cunningham- Cook and Sirota 2015.
56. Blumenthal 2015a.
57. Lipton 2014 .
58. West 2014.
59. Gibson (2013) analyzed how— in federal, democratic political systems—
dominant subnational political parties often work on their national levels to keep as many political issues and powers as possible off the national agenda and within their own subnational jurisdictions, while simultaneously working to establish and maintain one- party rule within those regional, provincial, or local jurisdictions.
That is, they work to control the subnational level and keep the federal government out— by creating “boundaries” around the areas they rule. Our theory of boundary control as a campaign contribution strategy differs from Gibson’s theory in many respects, but it is similar in that a symbiotic relationship between national and subnational politics is used to restrict central authority and gain power and benefits from subnational governments.
60. Mayer 2016, 346.
61. Stolberg and McIntire 2013; Goldfarb 2013.
62. A possible example of this occurred during the 2004 presidential election, when President Bush criticized the “Swift Boat” attack advertisements that deni-grated John Kerry’s military service. (Of course, we cannot be sure how sincere Bush’s criticism was.) The ads were financed by an outside group that drew much of its funding from billionaire Harold Simmons, who is discussed later in this chapter. Milbank 2004; Fitzsimmons 2013.
63. Most of the minority of billionaires who are liberal to moderate favor some sort of vigorous action by the federal government, if only to guarantee federal protection of civil rights and civil liberties that they see as threatened by social conservatives.
64. On industrial sectors and party affiliations, see Ferguson 1995.
65. We examined the costs that firms with numerous locations impose on each location. A billionaire’s firms were coded as having high net costs if they imposed such costs on at least one of the places where they were located.
176
notes to pages 113–115
66. A supplementary appendix listing the scores for each industry and the sources used to code them can be obtained by contacting the authors.
67. Selecting an individual billionaire’s highest- scoring industry most accurately captures that billionaire’s incentive to use the strategy in at least one place.
68. For previously noted reasons, using campaign contributions to infer the ideological positions of donors is potentially problematic. But contributions data are useful for constructing a general measure of partisan alignment roughly equivalent to party identification (which could be ascertained with complete confidence only through survey data that are not available on billionaires). We created this dichotomous variable by adding together a billionaire’s contributions to each party and each party’s candidates, and identifying which party received more money.
69. We also tried using a slightly stricter standard— eight or more years of unified government. The number of cases scoring 1 (“boundary controller”) on the dependent variable did not change.
70. We used data from the online Follow the Money database of the National Institute on Money in State Politics. The database includes federal data beginning in 2010 and state data beginning in 2000 (with the exception of ballot initiative data, which begins in 2003). Our calculations thus utilize federal data from 2010 to 2012 and state data from 2001 to 2012. Although some gaps may exist in the data, we believe we have captured nearly all— if not all— of the contributions made (to federal- and state- level candidates and parties, state ballot initiatives, and outside groups) by the billionaires under study for the years covered by the data. Care was taken to capture various versions of billionaires’ names. When possible, this data was combined with data collected by journalists tracing unreported contributions back to certain individuals. Inclusion of this second type of data affected only one case of boundary control: that of David Koch, whose national- level contributions are understated in FEC- reported data.
71. Of the eighty- three billionaires included in this analysis, thirty- two were active in states with dominant parties, twenty- seven met the national- level contribution criterion, thirty- five met the state- level contribution criterion, and twelve met both the national- and state- level criteria.
72. Ten billionaires were active in states that were dominated by Democrats during the period of study, but none of them made contributions that fit the definition of boundary control.
73. Somewhat counterintuitively, a slightly greater proportion of non– boundary controllers’ contributions (vs. boundary controllers’ contributions) go toward state politics. This is very likely a result of the large size of the national- level contributions made by boundary controllers to outside groups. Within each level, the relative balance of candidate- specific vs. general donations among members of each group aligns with expectations (i.e., a much greater proportion of boundary controllers’ national-level dollars relative to non– boundary controllers’ go to outside groups and parties, and the opposite is true on the state level).
74. Firth 1993.
notes to pages 116–121
177
75. The minimum possible score on the regulation variable is 2, referring to an industry that faces only light, non- industry- specific regulation and has no specific regulations on the state level. The maximum possible score is 8, which refers to a highly regulated industry in which all important industry- or firm- specific regulation occurs on the state or local level. The minimum actually observed score in our data was 4 and the maximum actually observed score was 7.
76. Seawright 2016, chaps. 2– 3.
77. Seawright 2016, chap. 4.
78. Random numbers to select among fifteen billionaires who tied as extreme low cases were generated using random.org.
79. Rowling was randomly selected (using random.org) among boundary con-
trollers with the modal score on the summed independent variables.
80. Helman 2013.
81. Langley 2013.
82. Langley 2013.
83. Fitzsimmons 2013.
84. Fitzsimmons 2013.
85. Simmons did contribute to super PACs dedicated to electing presidential candidates Newt Gingrich and Mitt Romney, including $2.8 million to a Romney-specific (though nominally “independent”) super PAC, Restor
e Our Future. But Simmons’s contributions to candidate- specific super PACs were still minuscule compared with his contributions to more general causes.
86. Center for Responsive Politics, n.d.
87. Texas Tribune, n.d. In addition to Bob Perry, Antonio Sanchez and David Dewhurst also donated more money than Simmons. But both Sanchez (who launched a failed gubernatorial campaign in 2002) and Dewhurst (who became Texas’s lieutenant governor) donated exclusively to their own campaigns. Subsequent to our use of the excellent Texas Tribune database, it appears to have disappeared from public view.
88. Harkinson 2011.
89. See Ryan, Lee, and Larson 2007.
90. Wald 2014.
91. Allison 2012.
92. Hunter 2006.
93. Harkinson 2011; Allison 2012.
94. Harkinson 2011.
95. Allison 2012; Campbell 2009.
96. Allison 2012; Thurber 2009.
97. Thurber 2009; Waste Control Specialists, n.d.
98. Harkinson 2011.
99. See Wilder 2013. It is more typical for firms to provide financial assurance through assets that are highly stable and very liquid. By providing stock in one of his other firms as collateral, Simmons was able to avoid the liquidation of any of his assets and shift risk away from himself to the state of Texas.
178
notes to pages 121–126
100. Hunter 2006; Wilder 2013.
101. Soros’s big short of the pound was widely seen as precipitating — if not all by itself causing — the sharp drop in the pound’s value that enabled Soros to buy pounds cheaply, pay them back to his lenders, and make an enormous profit.
102. Schaefer 2015.
103. Confessore and Preston 2016.
104. National Institute on Money in State Politics, n.d. As we have noted elsewhere, however, there are definite limits to Soros’s economic liberalism. Soros warmly appreciates the property rights and free markets that made him rich. He has not expressed a great deal of concern about inequality of income or wealth, or much enthusiasm for heavy taxation of the wealthy. See Mayer 2004.
105. The top- scoring industry in which Rowling has been active— energy and natural resources— scored a very high 14 out of a possible 16 on factors predicting efforts at boundary control. This was the modal score among billionaires we scored as boundary controllers.
Billionaires and Stealth Politics Page 27