54. The adjusted R- squared for the full regression in table 2.1 was 0.2984; the R- squared for a bivariate regression including only level of wealth was 0.2722.
55. We simply added up a billionaire’s scores on each of our five specific tax subissues, with +1 for favoring more taxes and – 1 for favoring less, so that the final scale can vary from – 5 (favoring lower taxes of all five types) to +5 (favoring more revenue from all five).
56. In this case, the three highly pro- tax cases, as a group, may have played an important part in the regression results, although none of them had a particularly notable Cook’s distance— that is, none was individually very influential.
57. Verba, Schlozman, and Brady 1995.
58. Overby 2007: Even back in 2004, a bundler had to raise $200,000 to qualify as one of George W. Bush’s “Rangers.” In 2008, when Barack Obama and John McCain voluntarily listed their bundlers, they numbered only 558 and 536, respectively. Center for Responsive Politics 2009.
59. Cook, Page, and Moskowitz 2014, 389– 90.
60. This last figure is probably understated, due to unreported dark- money contributions.
notes to pages 43–47
165
61. Smith et al. 2009.
62. Our calculations of billionaires’ contributions use data from the National Institute on Money in State Politics’ online database, which compiles federal, state, and local campaign finance data. Their 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 figures utilize federal data from 2010– 12 and state data from 2001– 12. Average annual contributions were calculated by adding together average annual amounts contributed to national and state politics. Although some small 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 variants of billionaires’ names. Any gaps (with the exception of dark money) in the data most likely involve small contributions that would amount to rounding errors in our overall calculations.
63. The figures in this paragraph include candidate, party, ballot initiative, and outside group contributions. Not all ballot initiative contributions could be coded in partisan and/or ideological terms, which is why the Republican and Democratic figures combined do not equal the total average annual contribution figure.
64. See Page and Jacobs 2009, chap. 4. Poll data on taxes should be interpreted with caution. Survey questions that inquire about “repeal” of the estate tax, for example, tend to elicit replies favoring repeal, but queries about the preferred rate of taxation on large ($100 million) estates yield median responses well above zero.
See Bartels 2016, chap. 6; Page and Jacobs 2009, 92.
65. This analysis uses a linear probability model rather than a logit or probit model. The linear probability model usually produces similar estimates of marginal effects as maximum- likelihood models, but requires somewhat fewer assumptions or (often arbitrary) modeling decisions (Angrist and Pischke 2009, 102– 7).
No results in this chapter are noticeably changed by using a logit or probit model in place of the linear probability model.
66. Knowles 2015; Voorhees 2015; Gold 2014; Washington Post 2014; Center for Responsive Politics 2014; MacColl 2010.
67. Bertoni 2012.
68. Mayer 2016, 57– 59 and chaps. 6– 14. In gripping detail, Jane Mayer chron-icles the Koch brothers’ subsequent quiet but increasingly active and important roles in funding and shaping conservative Republican politics. See also Nancy MacLean’s (2017) eye- opening archival research on the Kochs’ long- term funding of the libertarian intellectual network centered at George Mason University. On the contemporary Koch funding network, see Skocpol and Hertel- Fernandez, forthcoming. For three decades or more, the Koch brothers succeeded in staying below the radar of the general public and most political observers.
69. Buffett 2011.
166
notes to pages 48–55
70. Fox Business Channel 2011.
71. Bloomberg 2008; see also Talk of the Nation 2012. It is not surprising that Bloomberg, a former political candidate and mayor of New York, has taken a number of specific policy stands. But many of those stands have concerned issues unrelated to city government, on which neither election campaigns nor his role as mayor required him to speak out.
72. Citizens United v. FEC (2010) did not substantially change the legal status of dark money (Bump 2015). Dark money existed prior to that decision and is allowed by the IRS for 501(c)(4) “social welfare” groups and 501(c)(6) “business leagues” (Prokop 2015). However, the use of dark money has increased markedly since Citizens United, suggesting that the Supreme Court’s generally favorable out-look on “money as speech” may have encouraged the use of dark money. Earlier Supreme Court decisions, from Buckley v. Valeo (1976) onward, have severely limited the regulation of political money.
73. Sargent 2015, Lichtblau 2015.
74. West 2014, 11– 15.
75. A few prominent donors like the Kochs and Adelson have been spotlighted by investigative journalists and scholars looking into dark money, but other, less scrutinized billionaires may well have been contributing large amounts of dark money without being detected.
76. Gilens 2012; Gilens and Page 2014; Page and Gilens 2017.
77. Perhaps more important than the very occasional, random appearance of a multimillionaire in national samples is the possibility that the policy preferences of the affluent act as statistical proxies for the preferences of the truly wealthy, which tend to differ from those of average citizens in similar— but sharper— ways than the preferences of the affluent do.
78. Across Gilens’s 1,779 issues, the correlation between the preferences of “economic elites” (affluent Americans) and the alignments of business interest groups was nonsignificant— essentially zero— at – .02. Gilens and Page 2014, 571.
79. Gilens and Page 2014, 575. Page and Gilens 2017, chaps. 4– 5.
80. See Altman and Kingson 2015. On Obama: Wallsten, Montgomery, and
Wilson 2012. On Clinton: Gilon 2008.
81. Graetz and Shapiro 2005. But see Bartels 2016, chap. 6.
82. Wolters 2017; Huang and Cho 2017.
83. Hacker and Pierson 2010, especially chaps. 5 and 8.
Chapter Three: Four Billionaires Up Close
1. See Collier, Brady, and Seawright 2004; Lieberman 2005; and Seawright 2016.
2. This included using search terms not specific to the policy area of interest, and weeding through the large number of resulting false positives to check whether
notes to pages 55–61
167
relevant material emerged that was missed in the systematic searching. Thus, for example, when conducting the case study of Warren Buffett, we searched using terms like “Buffett AND politics,” “Buffett AND government,” or “Buffett AND
Washington.” These terms produce results that overlapped with the data from structured searching, as well as irrelevant articles and web pages about the Buffett Institute for Global Studies at Northwestern University, the political activities of
“Margaritaville” singer Jimmy Buffett, and Howard Graham Buffett’s foundation dedicated to ending world hunger. The case study process also included reading all Berkshire Hathaway shareholder letters written by Buffett in search of political content, as well as working through websites of collected Buffett quotes (e.g., http://www.suredividend.com/warren- buffett- quotes/ and https://en.wikiquote.org
/wiki/ Warren_Buffett) and perusing edited volumes of Buffett statements (Lowe 2007; Andrews 2012).
3. Random numbers were generated using data on atmospheric radio noise
from random.org.
4. Seawright 2016; and Seawright and Gerring 2008.
5. Forbes 201
3 (US), 32, 125.
6. Forbes 2016b (US), 129. As of early 2017 Buffett had regained his #2 position among the wealthiest Americans, with a fortune of $75.6 billion ( Forbes 2017a
[global], 138). But by October 2017 Bezos had overtaken Buffett again ( Forbes 2017b [US], 86).
7. Schroeder 2008, 38– 43, 61, 79.
8. Schroeder 2008, 88– 96, 113– 114, 146, 147, 178– 180.
9. Schroeder 2008. The “snowball” image came to Buffett as a child when he read One Thousand Ways to Make $1,000 and learned about compound interest (Schroeder 2009, 59– 61).
10. Schroeder 2009, 540, 592, 667; De la Merced 2009.
11. Quick 2016; Cabural 2015.
12. Berkshire Hathaway 1994. This letter discussed the year 1993 but was issued on March 1, 1994.
13. NBC 2007.
14. CNBC Squawk Box 2011a.
15. CNBC Squawk Box 2011b .
16. National Economic Council 2012.
17. Kavoussi 2012.
18. Berkshire Hathaway 2003.
19. Drawbaugh 2007.
20. White 2005.
21. Isidore and Harlow 2015.
22. Larson 2015.
23. Storch 2003.
24. Lambert 2016.
168
notes to pages 62–70
25. Forbes 2013 (US), 148.
26. Forbes 2016b (US), 142. Early the following year, Forbes (2017a [global], 140) estimated Menard’s fortune at $10.2 billion; later in 2017, Forbes put it at $9.9 billion (2017b [US], 92).
27. Forbes 2013 (US), 148.
28. Van de Kamp Nohl 2007.
29. Murphy 2013.
30. Isikoff 2015.
31. Aronson 2011.
32. Menard’s reportable donations have been much smaller than his dark-
money contributions, totaling an annual average of around $9,000 (86 percent of which went to Republicans or conservative groups), according to the data we collected from the National Institute on Money in State Politics.
33. Lueders 2015.
34. Lueders 2016.
35. Geoghegan 2004; Lichtenstein 2013.
36. Nolan 2016 .
37. Forbes 2013 (US), 137. A remarkable biography of Icahn— which includes the intriguing story of how it came to be written— is Stevens 2014 [1993], which takes Icahn through his 1980s– 1990s struggle with USX.
38. GuruFocus 2016.
39. Icahn’s net worth dropped sharply, from $23.5 billion in February 2015 to $15.7 billion in fall 2016; it recovered only a bit, to $16.6 billion by February 2017
and $16.7 billion by October of that year. Forbes 2015 (global), 151; 2016b, 138; Forbes 2017a (global),138; Forbes 2017b (US), 91.
40. Dodds 2015; Stevens 2014, chap. 9.
41. On one engagement in the USX battle, see http://www.nytimes.com/1990
/05/08/business/icahn- seen- as- loser- in- usx- vote.html.
42. Gupta 2013.
43. Harwell 2016.
44. Gass 2015.
45. Lipton 2017.
46. Keefe 2017.
47. See http://carlicahn.com/. Accessed 14 June 2017 .
48. Icahn 2013.
49. Cox 2016.
50. CNBC Fast Money Halftime Report 2015.
51. Icahn 2015a.
52. Icahn 2015b.
53. Icahn 2016.
54. As we have noted— and as we pursue further in chap. 4— among US bil-
lionaires social issues do not fit smoothly with the economic liberal/conservative
notes to pages 70–76
169
dimension either. Many billionaires are libertarians, conservative on economics but liberal on social matters. On social issues, they generally disagree less with average Americans and have less reason to pursue stealthy political strategies.
55. Forbes 2013 (US), 125.
56. Forbes 2016b (US), 132; Forbes 2017b (US), 86– 87.
57. Mooney 2015.
58. Leonard 2013.
59. Leonard 2013.
60. See the remarkable research by Nancy MacLean (2017).
61. Sclar et al. 2016. This amount was reduced and redirected mainly to Senate races after Donald Trump was nominated as Republican presidential candidate.
62. Mayer 2016; Skocpol and Hertel- Fernandez 2016 and forthcoming.
63. See Mayer 2010 and 2016; Skocpol and Hertel- Fernandez, forthcoming;
Washington Post 2014.
64. In New York City, David Koch is chiefly known as a major philanthropist and patron of the arts, a role that tends to smooth his social relations and shield his conservative political activities from criticism in that liberal bastion. See Schulman 2014, chap. 15.
65. Fang 2014.
66. Mayer 2016.
67. Skocpol and Hertel- Fernandez 2016.
68. Mayer 2016, especially chap. 8; Skocpol and Hertel- Fernandez 2016.
69. Mayer 2016, chap. 4.
70. Mayer 2016, 56– 58; Schulman 2014, 109– 14.
71. Koch 1980.
72. Koch 2016.
73. The only other billionaire we are aware of who resembles Icahn in his unusual mix of policy stands and his energetic backing of Donald Trump is Robert Mercer, who played a key role in installing Steve Bannon as Trump’s campaign chairman. Sheldon Adelson also supported Trump in the autumn 2016 campaign, but apparently for different reasons (Mayer 2017; Addady 2016.) See Ferguson, Jorgensen, and Chen 2018.
74. Lee Drutman (2015, 72– 79) reports evidence that corporations often (perhaps 40 percent of the time) lobby for narrowly targeted distributive benefits for themselves, but even more often (60 percent of the time) seek industry- wide benefits.
75. See Mayer 2016; Skocpol and Hertel- Fernandez, forthcoming; and Dom-
hoff 2014, chaps. 4– 5. Domhoff outlines a comprehensive theory of how wealthy individuals operate through a “policy- planning network” and an “opinion- shaping network,” each of which involves foundations, think tanks, universities, policy discussion groups, the media, and other institutions.
76. Citizens United v. Federal Election Commission 558 U.S. 310, 2010. Citizens United, of course, was only one in a series of Supreme Court decisions that have eviscerated efforts to regulate the role of money in American politics.
170
notes to pages 76–80
77. Downs 1957, 94. Downs left unanswered the question whether— even if he was right— the extent of political inequality might be significantly altered through human choices.
Chapter Four: Keeping Quiet on Social Issues
1. See, for example, Chauncey 2004; McVeigh and Diaz 2009.
2. Gerring 1998; Karol 2009; Noel 2013.
3. Jelen and Wilcox 2003, 490– 91; Saad 2016.
4. Baunach 2012. For more- recent data— including evidence that in 2013, for the first time, a majority of Americans approved of same- sex marriage—see http://
www.pewforum.org/2016/05/12/changing-attitudes-on-gay-marriage/; McCarthy 2016.
5. Dubner 2016. Chick- Fil- A’s late founder, S. Truett Cathy, was included among the 100 wealthiest billionaires we studied and was one of only four to take action against same- sex marriage. Cathy passed away shortly after the 2013 Forbes list was compiled. His fortune was divided between his two sons: current Chick-Fil- A CEO Dan Cathy, and Dan’s brother, Bubba. As of autumn 2017, Dan and Bubba Cathy each had a net worth of $4.6 billion, which tied them at #144 on the Forbes 400- wealthiest list ( Forbes 2017b [US], 106) .
6. See Page, Bartels, and Seawright 2011. As we have noted, even the Fed-
eral Reserve Board’s authoritative Survey of Consumer Finances, which oversamples affluent and wealthy Americans, does not attempt to reach the Forbes 400- wealthiest list (Bricker et al. 2014, 38).
7. For our earlier research on taxes and Social Security, we initially did both Google News/general web and LexisNexis Academic searches. As noted in chap. 2, Google and LexisNexis provide somewhat complementary resources. Google produces a very large number of potentially
relevant web pages and, helpfully, includes links to videos of interviews with our subjects. Google’s search results, however, are sometimes noisy and include numerous websites of dubious authority.
LexisNexis produces a smaller number of results and does not include video links, but it draws exclusively from mainstream journalistic and academic sources. After collecting data for approximately one- quarter of our billionaires, we discovered that LexisNexis searches did not uncover any relevant political statements missed by Google searches. After this discovery, we decided to use Google exclusively. We continued to do so on social issues.
8. Quotation marks, which instruct search engines to look for exact matches to the words or phrases included within them, were generally not used. In most cases, experimental searches that included quotation marks produced very similar results to those that did not. But the use of quotation marks sometimes leads to the exclusion of relevant texts. For example, a search for “defense of marriage” would fail to locate texts using the phrase “marriage defense.” Quotation marks were
notes to pages 80–90
171
occasionally used when the initial results from a particular search were unusually noisy. For example, quotation marks were placed around “Paul Allen,” after initial searches without quotation marks mostly returned results pertaining to people named either Paul or Allen, but not to the Paul Allen who cofounded Microsoft.
9. The terms “liberal” and “conservative” in US political discourse are con-tested, may be in flux, and— when applied simultaneously to both the economic and the social realms— seem to us not to be logically coherent. (Indeed, efforts to compress both into a single dimension have increasingly become a major source of tension within the political parties.) For convenience in analyzing the social issues, however, we refer to positions relatively favorable to abortion, same- sex marriage, and immigration as “liberal” positions.
10. Verba, Schlozman, and Brady 1995.
11. Cook, Page, and Moskowitz 2014, 389.
12. Only 18 percent of Americans reported making campaign contributions in a 2008 general- population survey (Smith et al. 2009.) In the SESA survey of multimillionaires, 60 percent reported contributing an average of $4,633 (Cook, Page, and Moskowitz 2014, 389, 390).
Billionaires and Stealth Politics Page 26