1. Kessler 2011; Yglesias 2011.
2. Reich 2007, p. 50.
3. See chapter 2.
4. See chapter 4.
5. Ibid.
6. Graetz and Mashaw 1999; Chetty and Finkelstein 2013.
7. Graetz and Mashaw 1999; Barr 2001; Moss 2002; Quiggin 2007.
8. See chapter 3.
9. I made this title up, assuming there was no such thing as a “Progressives’ Handbook.” It turns out there is.
10. Gangl 2006.
11. Duncan, Huston, and Weisner 2007.
12. Polanyi 1944.
13. Yglesias 2013.
14. See chapter 4.
15. Atkinson 1999; Lindert 2004.
16. Przeworski 1985; Esping-Andersen 1990; Berman 2006; Sejersted 2011.
17. Esping-Andersen 1990.
18. Esping-Andersen et al. 2002; Huo, Nelson, and Stephens 2008;
Kenworthy 2008a, 2010b; Kristensen and Lilja 2011;
Pontusson 2011; Hemerijck 2012; Morel, Palier, and Palme 2012.
19. Baumol, Litan, and Schramm 2007.
20. See chapter 4.
21. See Kenworthy 2011c, ch. 5.
22. Due to the coming retirement of the baby boom generation and projected increases in healthcare costs, Medicare and Social Security costs are likely to rise by 2–4 percent of GDP.
23. See chapter 4.
24. See chapter 4.
25. Kenworthy 2011c, ch. 8.
26. This conclusion is similar to that of Wilensky 1975, 2002.
27. Prominent exceptions include expansion of Social Security in the 1950s and indexation of its benefits to inflation in the early 1970s, creation of the EITC in 1975 and its expansion in the 1980s, designation of food stamps as a federal entitlement in the mid-1970s and expansion of food stamp eligibility in the early 2000s, increases in Medicaid coverage and benefits in the 1980s, and the extension of Medicare coverage to prescription drugs in 2004.
28. See Heclo 1974.
1. Lebergott 1976; Cox and Alm 1999; DeLong 2009; Fischer 2010, ch. 2.
2. Moss 2002; Castles et al. 2010. These laws and programs offered, in Franklin Roosevelt’s words, a “safeguard against misfortunes which cannot be wholly eliminated” (speech to Congress, June 1934).
3. For a contrary view, see Winship 2012.
4. Osterman 1999.
5. Uchitelle 2006.
6. Hacker 2006. See also Hacker 2011.
7. Gosselin 2008.
8. Hacker, Rehm, and Schlesinger 2010, figure 4.
9. Osterman 1999; Baumol, Blinder, and Wolff 2003; Fligstein and Shin 2003; Uchitelle 2006; Blinder 2009a; Farber 2010; Kalleberg 2011.
10. Western, Bloome, Sosnaud, and Tach 2012.
11. Congressional Budget Office 2010. The standard indicator of low income is the poverty rate, which tells us the share of people in households with incomes below a particular level—the poverty line. But there is little agreement about where the line should be drawn, so skeptics rightly worry that if the line is a little bit lower or higher, the estimate of how many are poor will change significantly. I prefer to instead look directly at the incomes of the least well-off. See Kenworthy 2011b, 2011c; also Burtless and Smeeding 2001.
12. Rector 2007; Eberstadt 2008.
13. Edin and Lein 1997; Ehrenreich 2001; DeParle 2004.
14. Caner and Wolff 2004.
15. Mayer and Jencks 1993; Edin and Lein 1997.
16. Rank 2004, table 4.2, using Panel Study of Income Dynamics (PSID) data.
17. Ibid.
18. Boushey, Brocht, Gundersen, and Bernstein 2001. See also Gould, Wething, Sabadish, and Finio 2013.
19. Wider Opportunities for Women 2011.
20. Census Bureau 2011, table B-2.
21. Wolff 2011, table 4.1.
22. Center for Retirement Security 2009, using data from the Survey of Consumer Finances. The numbers sum to more than 100 percent because some people are enrolled in both types of plan.
23. Another potential advantage is that a defined-contribution plan isn’t vulnerable to a firm going out of business or declaring bankruptcy. Such an occurrence can leave those in a traditional company pension plan out in the cold, though most such pensions are guaranteed by the federal government-backed Pension Guarantee Benefit Corporation.
24. Economist 2008; Ghilarducci 2008; Wolff 2011; Fletcher 2013.
25. My calculations using OECD data.
26. Layard 2005.
27. Rose and Winship 2009, figure 6.
28. Rose and Winship 2009, figure 2. For additional estimates based on the PSID data, see Gosselin and Zimmerman 2008; Hacker and Jacobs 2008; Jensen and Shore 2008; Dynan 2010; Dynan, Elmendorf, and Sichel 2012.
29. Winship 2012, figure 1. For additional estimates based on the SIPP data, see Acs, Loprest, and Nichols 2009.
30. Congressional Budget Office 2008, figure 5, p. 10.
31. See economicsecurityindex.org. Accessed January 15, 2013. These numbers are for people who lost one-fourth of their income, not including out-of-pocket medical costs or repayment of debt.
32. Mark Rank and Thomas Hirschl have done a related calculation, also using PSID data. They ask what share of Americans experience five or more total years of income below the poverty line, of welfare receipt, or of unemployment between the ages of 25 and 60. The answer is 35 percent. See Rank and Hirschl, forthcoming.
33. See also Dynan, Elmendorf, and Sichel 2012.
34. Elizabeth Jacobs, personal communication.
35. Families USA 2009.
36. Mendes 2012, using Gallup data.
37. Jacoby 2012.
38. Pew Research Center. For more, see Page and Jacobs 2009; McCall 2013.
39. Sen 1999; Nussbaum 2011.
40. Noah 2012.
41. The share of women aged 25 to 29 who have completed four or more years of college has risen steadily over the past half century. In fact, by the early 1990s women had caught up with men in college graduation, and in the ensuing decades they raced ahead. By 2010, 36 percent of women aged 25 to 29 had completed college, compared to just 28 percent of men. This gap in favor of women was as large as the gap favoring men was in 1960. (Census Bureau, “Educational Attainment: CPS Historical Time Series Tables,” table A-2, using Current Population Survey data.)
Women have made significant strides in employment as well. According to OECD data, in 1960 just 41 percent of prime-working-age (25 to 54) women were employed. By 2010 that had jumped to 73 percent. Among men in this age group, the employment rate fell from 92 percent to 87 percent.
Median earnings among women employed full-time and year round likewise have risen steadily, from (in 2010 dollars) $21,000 in 1960 to $28,000 in 1979 to $38,000 in 2010 (Census Bureau, “Historical Income Tables,” table P-21.). This is an advance in an absolute sense and also relative to men. In the 1960s and 1970s, women’s median earnings were about 60 percent of men’s. In the 1980s women began gaining ground, and by 2010 the ratio was 77 percent. In fact, if we narrow the comparison to those aged 25 to 34, the ratio is 91 percent. (Bureau of Labor Statistics 2011. The female-to-male earnings ratio then decreases for older age groups, partly because women are much more likely to leave work temporarily for family reasons. Part of the gap is accounted for by work hours (men are more likely to work overtime) and employee benefits (women are more likely to hold jobs with better benefits but lower pay). See CONSAD Research Corp 2009.)
As of 2012 women held 52 percent of managerial, professional, and related positions (Bureau of Labor Statistics 2011). At the top of the corporate hierarchy there is still a large gender gap; yet here, too, progress has been significant. Women represent only 18 percent of senior officers (vice president or higher) in Fortune 500 firms, but that’s up from 9 percent in 1995. Women hold only 3 percent of Fortune 500 CEO slots, yet that’s up from 1 percent in 2000. At the very top, on the boards of directors of Fortune 500 companies, women hold 16 percent of the seats, up from 1 percent in 1988. (These data are from Catalyst, catalyst.org.)
All told, women’s capabilities have expanded markedly. Noteworthy gaps with men persist in various aspects of labor market success, but progress has been substantial.
42. In 1960, only 40 percent of African Americans aged 25 to 29 had completed high school. By the end of the 1970s, the share had increased to 77 percent, and in 2010 it reached 89 percent (this includes those with a GED). The advance was not only absolute, but also relative to whites. White high school completion in this age group was 67 percent in 1960, nearly double the black share. In 2010 the white high school completion rate was 95 percent, just 6 percentage points higher than among blacks. (Census Bureau, “Educational Attainment: CPS Historical Time Series Tables,” table A-2, using Current Population Survey data.)
College graduation among African Americans has risen substantially. In the early 1960s, only 5 percent of 25-to-29-year-old blacks had a four-year college degree. By 2010, 20 percent did. On the other hand, we might have hoped for a larger increase. College completion among whites jumped from 15 percent to nearly 40 percent during that half century, and college completion among women rose from 10 percent to 35 percent.
Reliable employment data for African Americans date from the early 1970s (Bureau of Labor Statistics, “Labor Force Statistics (CPS): Household Data,” using Current Population Survey data). Between 1973 and 2007 (business-cycle peak years), the employment rate among blacks rose by about 5 percentage points. The same was true for whites. For both groups the 2008–09 recession brought a significant decline.
In earnings, African Americans have experienced less advance (Census Bureau, “Historical Income Tables,” table P-38, using Current Population Survey data). Among those employed full-time and year-round, median earnings rose steadily in the 1960s and 1970s. But then the improvement ceased, apart from a few years in the late 1990s. Whites experienced a similar fate, though, which suggests the barrier was changes in the economy rather than obstacles specific to African Americans.
The same is true for household income (Census Bureau, “Historical Income Tables,” table H-5 and Census Bureau 2011, using CPS data). The late 1990s is the only period in the past several decades in which we observe improvement in either median household income or the share of African Americans with above-poverty-level income. Again, though, the same holds for whites.
In the past generation, incarceration has become a relatively common feature in the life course of African American men. Spending time in prison and having a prison record significantly diminishes the likelihood of employment during young adulthood, and it reduces the earnings of those who do find employment (Western 2006). We don’t yet know the degree to which this effect continues through people’s working career, but it seems likely to act as a permanent brake on labor market opportunity for a considerable number of black men.
Overall, opportunity for many African Americans has improved, though in some important areas both absolute and relative progress has been limited.
43. The following discussion draws on Wilson 1978, 1987, 1996, 2009; Jencks and Mayer 1990; Fischer et al. 1996; Duncan and Brooks-Gunn 1999; Mayer 1999; Lareau 2003; MacLeod 2009; Currie 2011; Duncan and Murnane 2011; Kenworthy 2012b.
44. Economic Mobility Project 2012.
There is additional evidence of unequal opportunity by family background. Begin with educational attainment. Caroline Ratcliff and Signe-Mary McKernan (2010, figure 5, using PISD data) examined American children who grew up in the 1970s and 1980s and found those with low-income parents less likely to graduate from high school. Among those whose household never was poor during their childhood (never had an income below the official poverty line), 96 percent got a high school diploma. Among those whose household was poor for half or more of their childhood years, only 77 percent did.
As adults, children of low-income parents are less likely to be employed and more likely to have low earnings and incomes. Ratcliff and McKernan found that among males whose family was poor in half or more of their childhood years, only 34 percent were employed consistently as young adults (from ages 25 to 30). Among those never poor during childhood, 76 percent were consistently employed.
Markus Jäntti and colleagues (2006, using National Longitudinal Survey of Youth data) examined the adult earnings of boys who grew up in the 1960s and 1970s. Of those whose father was in the lowest fifth of earnings, 42 percent were themselves in the lowest fifth as adults, compared to just 10 percent of those whose father was in the top fifth.
A similar pattern holds for household income. David Harding, Christopher Jencks, Leonard Lopoo, and Susan Mayer (2005, table 3.4) examined the adult incomes of various generations of American children. Of those whose parents’ income was in the bottom quarter, 40–45 percent were themselves in the bottom quarter as adults. Of those whose parents’ income was in the top quarter, only 10 percent were in the bottom quarter as adults. Ratcliff and McKernan found that among people whose household income was below the poverty line in more than half of their childhood years, 32 percent ended up themselves poor as young adults. The share among those who grew up in a never-poor home was just 1 percent. (“Nonpoor” as a young adult is defined as having a household income above the official poverty line in two or fewer of the six years from ages 25 to 30.)
45. Hauser et al. 2000.
46. McLanahan and Sandefur 1994; McLanahan 2001.
47. Ellwood and Jencks 2004; Murray 2012a; Parlapiano 2012.
48. Lareau 2003; Duncan and Murnane 2011.
49. Reardon 2011.
50. Jencks 2009; Bailey and Dynarski 2011.
51. Western 2006.
52. Wilson 1987, 1996; Wright and Dwyer 2003; Autor 2010; Western and Rosenfeld 2011; Blinder 2009a.
53. Schwartz and Mare 2005.
54. Some recent studies find no indication of an increase in inequality of opportunity. See Harding, Jencks, Lopoo, and Mayer 2005, using data from the General Social Survey (GSS) and the PSID; Lee and Solon 2009, using data from the PSID; Winship 2013, using data from the National Longitudinal Survey of Young Men and the National Longitudinal Survey of Youth. Others conclude that there likely has been an increase. See Aaronson and Mazumder 2008, using estimates from Census data; Bloome and Western 2011, using data from the National Longitudinal Survey of Young Men and the National Longitudinal Survey of Youth.
55. Rawls 1971; Friedman 2005; Kenworthy 2011c.
56. These data are created by merging information from the Current Population Survey of households with tax records from the IRS. Each of these two sources has a strength and a weakness. Household surveys provide good information for the bulk of the population, but not for those at the top. Tax records give us much better information on those at the top, but are not very good for those at the bottom. By merging the two, the CBO has created the most accurate and complete data series available. Unfortunately, it only goes back to 1979.
57. See, for instance, Samuelson 1995; Cox and Alm 1999; Easterbrook 2003; Rose 2010; Burkhauser, Larrimore, and Simon 2011; Meyer and Sullivan 2011.
58. Kenworthy 2012a.
59. Actually, it’s worse than the trend for all families. That’s because median income among the retired has been growing at a slightly faster pace than median income among prime-working-age families, due to rising Social Security benefit levels. The data are at Census Bureau, “Historical Income Tables,” table F-11.
60. Census Bureau, “Foreign Born.”
61. The data are at Census Bureau, “Historical Income Tables,” table F-5.
62. Johnson 2004; Meyer and Sullivan 2011. See also Attanasio, Hurst, and Pistaferri 2012.
63. See Wolff 2012.
64. Isaacson 2011.
65. The data in the following paragraphs come from a variety of sources, including Lebergott 1976; Cox and Alm 1999; Easterbrook 2003; Fischer 2010.
66. For an argument that the pace of innovation has been less rapid since the mid-1970s, see Cowen 2011.
67. See in particular Warren and Tyagi 2003; Lesmerises 2007; Blank 2010.
68. Census Bureau, “Historical Income Tables,” table F-12.
69. Census Bureau, “Historical Income Tables,” table P-38.
70. Census Bureau, “Historical Income Tables,” table F-12.
71. Jacobs and Gerson 2004.
72. Presser 2003.
73. McKenzie and Rapino 2011.
74. Some suggest that the cost of goods and services consumed by low-income households has risen less rapidly than that of the bundle used to adjust for inflation. See Broda and Romalis 2008. Others argue that this is wrong. See Meyer and Sullivan 2011.
75. Lesmerises 2007, figure 12, using data from Warren and Tyagi 2003. See also DeLong 2012.
76. College Board, “Trends in College Pricing 2006,” table 4a.
77. Allegretto 2011, table 10, using Survey of Consumer Finances data.
78. Warren and Tyagi 2003; Weller 2012.
1. Census Bureau 2011, table 8.
2. The 2010 Affordable Care Act may accelerate this shift. See Hacker 2011.
3. Seidman 2013.
4. For additional cost-reduction ideas, see Emanuel et al. 2012; Gawande 2011, 2012; Soltas 2012; Aaron 2013.
5. Klein 2007; Reid 2009; Davis, Schoen, and Stremekis 2010; OECD 2011.
6. Heymann et al. 2009.
7. Bureau of Labor Statistics 2012b, using data from the March 2012 National Compensation Survey.
8. Waldfogel 2006, ch. 2. See also Kenworthy 2009c.
9. Ferrarini and Duvander 2010.
10. Anrig 2006.
11. House Ways and Means Committee, Green Book, various editions.
12. Kletzer and Litan 2001; LaLonde 2007; Kling 2008.
13. Kenworthy 2009b.
14. Murray 1984; Ellwood 1988; Marmor, Mashaw, and Harvey 1990; Jencks 1992; Skocpol 1992; Blank 1997; Gilens 1999; Haskins and Sawhill 2009; Thiebaud 2012.
15. This logic could also be applied to existing disability policy. See Autor and Duggan 2010; Burkhauser and Daly 2012; Reno and Ekman 2012.
16. Hays 2003; DeParle 2004, ch. 14; Morgen, Acker, and Weigt 2010.
17. Edin and Lein 1997; Newman 1999, 2006; DeParle 2004; Edin and Kefalas 2005; England and Edin 2009; Nelson and Edin 2013.
18. Ben-Galim and Dal 2009.
19. Gottschalk 1998.
20. DeParle 2012; Shaefer and Ybarra 2012.
21. Diamond and Orszag 2004; Ruffing 2011; Aaron 2013.
22. Graetz and Mashaw 1999; Munnel 2012; US Senate Committee on Health, Education, Labor, and Pensions 2012.
23. Currie 2006; OECD 2008, ch. 9; Esping-Andersen and Myles 2009; Garfinkel, Rainwater, and Smeeding 2010; Kenworthy 2011c, ch. 7.
24. OECD 2008, ch. 7.
25. Heckman 2008; Ermisch, Jäntti, and Smeeding 2012, p. 465; Duncan and Murnane 2011, p. 9.
26. Downey, von Hippel, and Broh 2004; Alexander, Entwisle, and Olson 2007.
27. See Bowen, Chingos, and McPherson 2009; Hout 2009; Jencks 2009; Arum and Roksa 2011; Dynarski and Scott-Clayton 2013; National Commission on Higher Education Attainment 2013.
28. Jencks 2009, p. A7.
29. OECD 2010, table B5.1; OECD 2012.
30. Murray 2009; Vedder et al. 2010.
31. This excludes those who get a certificate of high school equivalency by passing the General Educational Development tests (GED). Heckman and LaFontaine 2007.
32. Kozol 1991.
33. Hoxby 2003.
34. Gleason et al. 2010; Center for Research on Education Outcomes 2013.
35. DeLuca and Rosenblatt 2010.
36. Hanushek 2010; Chetty, Friedman, and Rockoff 2011.
37. Gordon, Kane, and Staiger 2008; Auguste, Kihn, and Miller 2010; Kristof 2012; Wilson and Laurent 2012.
38. Esping-Andersen 2004, 2011; Smeeding, Erickson, and Jäntti 2011; Ermisch, Jäntti, and Smeeding 2011; Esping-Andersen and Wagner 2012.
39. Heckman 2008; Ruhm and Waldfogel 2011.
40. Gornick and Meyers 2003; OECD 2006; Esping-Andersen 2009.
41. Vandell and Wolfe 2000; Waldfogel 2006. Among three- and four-year-olds, approximately 28 percent are enrolled in public prekindergarten and 52 percent in any type of prekindergarten school. See Barnett et al. 2011, table 4; National Center for Education Statistics 2012, table A-1-1.
42. Heckman 2008; Barnett 2013.
43. Kenworthy 2012b.
44. Duncan, Ziol-Guest, and Kalil 2010.
45. Waldfogel 2009, p. 52.
46. Western 2006.
47. Western 2012.
48. Bowen and Bok 1998; Reskin 1998.
49. Kahlenberg 1995; Murray 2012b.
50. For more detail, see Kenworthy 2011c.
51. For a more detailed discussion, see Kenworthy 2011i.
52. Mishel and Shierholz 2013.
53. Bernstein and Baker 2003.
54. Kenworthy 2008a, 2013a.
55. Kenworthy 2008a.
56. For more discussion, see Kenworthy 2011g.
57. Rising employment is particularly important for those at the low end of the labor market. Here, too, the 2000s upturn was a disappointment. In working-age households in the bottom quarter of income, average employment hours failed to rise at all. See Kenworthy 2011c, ch. 3.
58. Gordon 2010.
59. Blinder 2009a; Pierce and Schott 2012.
60. Brynjolfsson and McAfee 2011; Jaimovich and Siu 2012.
61. With the ebbing of the outsourcing and offshoring craze, some manufacturing and service jobs have begun to move back to the United States. But the number is, and will be, relatively small. See Fishman 2012; Economist 2013b.
62. Ben-Galim and Dal 2009.
63. Katz 1998; Gottschalk 1998. A good candidate here is green jobs; see Apollo Alliance 2008; Block 2011.
64. The Bureau of Labor Statistics projects that in 2020 approximately one third of US jobs will be in occupations with a median wage of $25,000 or less. See Bureau of Labor Statistics 2012a.
65. For similar sentiment, see Gans 2011.
66. Baumol 2012; Carlin 2012.
67. My suggestion is similar in spirit to the idea of “inequality insurance” proposed by Robert Shiller; see Shiller 2003, ch. 11. See also Reich 2010.
68. Average compensation tends to rise in sync with GDP per capita. See Pessoa and Van Reenen 2012.
69. Note that to be effective, a rising earnings subsidy will need to be coupled with a rising minimum wage. Otherwise, the subsidy may lead to reductions in low-end wage levels, which will offset the improvement in income achieved by the subsidy. This can happen in two ways. First, if the subsidy succeeds in pulling more people into work, the increase in competition for jobs will put downward pressure on wages. Second, regardless of labor supply, employers will be tempted to incorporate the value of the subsidy into the wages they offer. For more discussion see Kenworthy 2011c, ch. 5.
1. Kenworthy 2008a, ch. 8; 2011, ch. 8.
2. Kenworthy 2009a.
3. OECD 2008, ch. 4; Kenworthy 2011c, ch. 8.
4. Mahler and Jesuit 2006; Kenworthy 2011c, ch. 8; Wang and Caminada 2011.
5. This means households with incomes below approximately $250,000.
6. Kenworthy 2011a.
7. Our GDP in 2007 was $14 trillion. Ten percent of that is $1.4 trillion. That year the top 5 percent of Americans, 5.9 million households, had an average pretax income of $611,200, according to the Congressional Budget Office (CBO), making their total pretax income $3.6 trillion (Congressional Budget Office 2010). The $1.4 trillion needed to boost tax revenues by 10 percent of GDP amounts to 38.8 percent of that $3.6 trillion. Thus, the effective tax rate on the incomes of the top 5 percent of households would need to be increased by 38.8 percentage points.
8. Kenworthy 2008a, figure 8.12.
9. This estimate is based on information in Krueger 2009; Toder and Rosenberg 2010; Barro 2011; Campbell 2011.
10. In fact, a consumption tax can be made progressive. See Frank 2008.
11. Bartlett 2012b.
12. Kenworthy 2011c, ch. 8.
13. Kenworthy 2011h.
14. Surowiecki 2010.
15. Carasso, Reynolds, and Steurle 2008; Toder et al. 2010.
16. Irons 2010.
17. The total payroll (Social Security and Medicare) tax rate is about 15 percent, and since the mid-1980s it has consistently collected 6.6 percent to 7.0 percent of GDP.
18. I use the shares of pretax income going to each quintile as of 2007. See Congressional Budget Office 2010.
19. Huang 2012; Bartlett 2012a; Saez, Slemrod, and Giertz 2012.
20. The list of relevant citations is lengthy. A good start is Polanyi 1944; Stiglitz 1989; Gough 1996; Lindert 2004; Madrick 2009.
21. Jones 1995.
22. This line of reasoning follows Stokey and Rebelo 1995; Myles 2000.
23. Tanzi 2011.
24. For more discussion, see Kenworthy 2008a, ch. 4.
25. Slemrod and Bakija 2008; Myles 2009; Bakija 2012. As with almost any phenomenon, it’s possible, with enough tweaking, to find an association that conforms to one’s prior beliefs. For instance, some studies find a negative association between taxation and economic growth by adding a group of moderate-tax high-growth countries or a particular set of control variables. See Fölster and Henrekson 2001; Bergh and Karlsson 2010; Bergh and Henrekson 2011.
When comparing across countries, our best bet, from an analytical perspective, is to compare changes over time rather than levels. (This is a strategy known as “differences in differences”; see Kenworthy 2011d.) This means looking at the cross-country association between changes in the size of government and changes in economic growth. But this requires a good bit of variation among the countries in the degree of change in the hypothesized cause, and we don’t have that here, unfortunately. During this period, government revenues rose by a few percentage points of GDP in nearly all countries. See Kenworthy 2011d, figure 2.
26. Hall and Soskice 2001.
27. Coherence applies both within and across economic spheres. A country’s institutional mix is deemed more coherent to the extent that its institutions within each sphere are closer to one or the other of the two poles (liberal market or coordinated market) rather than in between and its institutions are consistent across spheres. Incoherence can be a product of being in the middle within each sphere or having liberal market institutions in some spheres and coordinated market institutions in others.
28. There have been other attempts to measure institutional coherence: Amable 2003; Hall and Gingerich 2004; Kenworthy 2006; Schneider and Paunescu 2012.
29. Hall and Gingerich have several control variables in their regressions, including inflation, change in the country’s terms of trade (export prices divided by import prices) weighted by its degree of trade openness, and the share of the population younger than age 15 or older than age 64. However, none of these is related to economic growth in their analyses. A fourth is the average growth rate among the group of countries as a whole weighted by the degree of trade openness in each nation. In an analysis with yearly data, this is useful to control for business cycle effects, but it is unnecessary in an analysis that covers one or more complete business cycles. I estimated a series of regressions with various combinations of three other controls: education (average years of schooling completed), real interest rates, tax revenues (as a share of GDP). This did not yield a positive association between institutional coherence and economic growth.
30. See also Campbell and Pedersen 2007; Kristensen and Lilja 2011.
31. Acemoglu, Robinson, and Verdier 2012.
32. Heller 2001.
33. Atkinson, Piketty, and Saez 2011, figure 10.
34. Schultze 1977; Anderson 1978; Simon 1978; Wanniski 1978; Friedman and Friedman 1979; Weidenbaum et al. 1980; Gilder 1981; Murray 1984.
35. Magaziner and Reich 1982; Reich 1983; Zysman 1983; Piore and Sabel 1984; Thurow 1984; Osborne 1987; Dertouzos et al. 1989; Kuttner 1989; Florida and Kenney 1990; Womack, Jones, and Roos 1990; Harrison and Kelley 1991; Tyson 1992.
36. Reich 1991, 1999; Clinton 1995; Sperling 2005.
37. Boix 2000.
38. Giddens 1998, 2000.
39. Both years were business-cycle peaks.
40. These historical GDP per capita data are from Angus Maddison, www.ggdc.net/maddison/historical_statistics/vertical-file_02-2010.xls.
41. Kenworthy 2010a.
42. Kenworthy 2012c.
43. Baumol 1967.
44. Mandel 2000, 2012; Baumol 2012.
45. Cowen 2011; Gordon 2012. Part of Gordon and Cowen’s pessimism owes to the fact that the benefits of innovation no longer reach all households in the form of income growth. This is an important problem, as I emphasized in chapter 2. But it’s a problem of distribution, not production. See Kenworthy 2011f.
46. St. Louis FRED, employment-population ratio (LNU02300000), using Bureau of Labor Statistics data.
47. North 1990.
48. These employment rates are for persons age 16–64.
49. Prescott 2004; Ohanian, Raffo, and Rogerson 2007.
50. Blanchard 2004; Alesina, Glaeser, and Sacerdote 2005; Immervoll and Barber 2006; Ray and Schmitt 2007; Boeri and Cahuc 2008; Causa 2008; Kenworthy 2008a.
51. Burgoon and Baxandall 2004.
52. Scharpf 2000; Kemmerling 2009; OECD 2007; Kenworthy 2008a, 2011.
53. Gornick and Meyers 2003; Huo, Nelson, and Stephens 2008; Kenworthy 2010b.
54. Some include Ireland and the UK in the “weak labor” group and Spain and Portugal in the “traditional family roles” group, but doing so doesn’t alter the conclusion.
55. Nor are tax levels associated with change in employment hours from the 1980s to the 2000s. See Kenworthy 2011e.
56. A good discussion is Tanzi 2011, ch. 1.
57. Baker 2011, ch. 10.
58. Weeden 2002; Carpenter et al. 2012.
59. Avent 2011; Glaeser 2011; Yglesias 2012.
60. Baker and Moss 2009; Stiglitz 2009; Baker 2011, ch. 9; Zingales 2012.
61. Zingales 2012, p. 6. See Friedman 1962; Friedman and Friedman 1979.
62. Alesina and Angeletos 2005, p. 1241.
63. Rothstein 2011.
64. Using other measures of the quality of government, such as Transparency International’s corruption perceptions index, does not change the story.
65. Teles 2012.
66. IRS Taxpayer Advocate Service, 2010 Annual Report to Congress, vol. 1, cited in Teles 2012.
67. “Full Transcript of the Romney Secret Video,” Mother Jones, September 2012.
68. Eberstadt 2012. See also Murray 1984, 2012a.
69. Though they represented just one in three employees, gains from collective bargaining spilled over to other firms whose management was keen to preempt unionization efforts.
70. Some worry about these developments because they believe stable couple families engender better behavior, thereby improving communities and individual happiness. See, for instance, Murray 2012a. For my purposes, family decline is a source of concern insofar as it has contributed to rising economic insecurity, to inadequate opportunity, and to the slow growth of living standards for America’s lower half.
71. This updates figure 1.1 in Ellwood and Jencks 2004.
72. McLanahan and Sandefur 1994; McLanahan 2001.
73. Ellwood and Jencks 2004, p. 9. According to Charles Murray (2012a, p. 158): “No matter what the outcome being examined—the quality of the mother-infant relationship, externalizing behavior in childhood (aggression, delinquency, and hyperactivity), delinquency in adolescence/criminality as adults, illness and injury in childhood, early mortality, sexual decision making in adolescence, school problems and dropping out, emotional health, or any other measure of how well or poorly children do in life—the family structure that produces the best outcomes for children, on average, are two biological parents who remain married. Divorced parents produce the next-best outcomes. Whether the parents remarry or remain single while the children are growing up makes little difference. Never-married women produce the worst outcomes. All of these statements apply after controlling for the family’s socioeconomic status.”
74. Census Bureau, “Historical Income Tables,” www.census.gov/hhes/www/income/data/historical, tables F-12 and H-12.
75. Douthat and Salam 2008, p. 53.
76. Edin and Kefalas 2005; Cherlin 2009, ch. 7; England and Edin 2009.
77. Ellwood and Jencks 2004; Edin and Kefalas 2005; Cherlin 2009, ch. 7; England and Edin 2009; Haskins and Sawhill 2009, ch. 10; Wilson 2009, ch. 4; Conger, Conger, and Martin 2010; Isen and Stevenson 2010; England, McClintock, and Shafer 2012; England, Wu, and Shafer 2012; Murray 2012a, ch. 8; Nelson and Edin 2013.
78. Ellwood and Jencks 2004, figure 1.13. See also England and Edin 2009.
79. Douthat and Salam 2008, ch. 7; Haskins and Sawhill 2009, ch. 10.
80. Improving the financial incentive for marriage was a key aim of the mid-1990s welfare reform. See DeParle 2004. But it hasn’t worked. AFDC-TANF benefit levels have been falling steadily and substantially since the 1970s, with no impact on the trend in family dissolution among less-educated Americans. See Aber et al. 1994; Moffitt and Scholz 2009.
An advertising and messaging campaign might help a bit, but I doubt it will yield much progress. Yes, similar campaigns helped reduce smoking and teen births. But marriage is different. We have strong evidence that smoking and teen parenthood are bad for the decision maker. The evidence is less overwhelming that avoiding or exiting marriage is bad for the person doing so. It tends to be bad for children, but it may or may not be bad for the adults. Also, smoking is an (addictive) activity, whereas marriage is a relationship. A partner or spouse who treats you poorly or cheats on you may be less tolerable than a daily routine without cigarettes.
The Healthy Marriages Initiative, which ran from 2005 to 2010, provided marital counseling and support for vulnerable couples. A thorough review found no noteworthy beneficial impact. See Wood et al. 2012; also Furstenberg 2008.
81. Cherlin 2009.
82. Ibid.; England and Edin 2009, chs. 1, 3, 6; Wilson 2009, ch. 4.
83. Sawhill 2002.
84. Wilson 1987, 1996, 2009.
85. Edin and Kefalas 2005.
86. Douthat and Salam 2008.
87. OECD 2001; Rønsen 2001; Rønsen and Sundstrom 2002; Morgan and Zippel 2003; Gangl and Ziefle 2012.
88. Edin and Kefalas 2005.
89. In Sweden the home-care allowance is used much more frequently by immigrant mothers, who because of language and education deficits have weaker labor market prospects than native-born women. See Earles 2011.
90. See chapter 2.
91. Mayer 1999.
92. Tocqueville 1840; Putnam 2000; Skocpol 2003; Brooks 2012b; Murray 2012a.
93. Putnam 1993a, 1993b, 2000.
94. Putnam 2000, figure 8.
95. Anderson et al. 2006.
96. Skocpol 2003.
97. Esping-Andersen 1990.
98. E.g., Esping-Andersen 1999.
99. The social democratic world is sometimes called “socialist” or “universalist.” The conservative world is sometimes called “corporatist” or “social insurance.” The liberal world is sometimes called “residual.”
100. Hicks and Kenworthy 2003. See also Esping-Andersen 2003. These findings hold if the elements fed into the statistical analysis are expanded beyond traditional social insurance programs to include labor market policies and work-family policies.
101. Pension payments are a significant portion of government transfers in all rich countries. In one interpretation, counting public pensions in a measure of targeting-universalism or redistribution is misleading, because pension programs are best conceptualized as forced saving. The government requires employed citizens to put money away during their working years and then returns it to them (with interest) in their retirement years. In retirement, many people have no income from employment, so the pension they receive appears in the calculations as though it is going to a very poor household. According to this view, the measures therefore overstate the degree of targeting and the degree of redistribution achieved by transfers. Peter Whiteford (2008, 2009) has attempted to address this concern by calculating targeting-universalism and redistribution using households’ position in the income distribution after transfers are added and taxes subtracted, rather than before. If a retired couple’s income consists solely of a public pension payment, they will be at the very bottom of the distribution according to the calculations in figure 4.19. In Whiteford’s calculations they instead might be at the twentieth percentile or even higher, depending on the size of their pension. In these calculations, Australia remains the most targeting-heavy of the rich nations, but it ranks higher on redistribution than in figure 4.19.
102. Korpi and Palme 1998; Kim 2000; Pontusson 2005.
103. Kenworthy 2011c, ch. 6; Marx et al. 2012.
104. Adema 2001; Adema and Ladaique; 2009; Fishback 2010. See also Adema 1997; Howard 1997, 2007; Hacker 2002; Garfinkel, Rainwater, and Smeeding 2010; Gilbert 2010; Mettler 2011.
105. Kenworthy 2011c, ch. 9; Morgan 2013.
106. Adema and Ladaique 2009, table 5.5; Fishback 2010, table 5.
107. One important tax benefit for low-income households is the EITC, but it is already included in the standard OECD data on government social expenditures. Another is the Child Tax Credit, but it is only partially refundable and thus of limited value to low-income households, many of whom don’t owe any federal income tax.
108. They also tend to cost more, due to higher administrative costs and management fees. Graetz and Mashaw 1999; Attewell 2012.
109. LIS 2012.
110. Consumption tax rates are higher in the Nordic countries than in the United States. But these are incorporated in the purchasing power parities (PPPs) used to convert incomes to a common currency, so the income numbers in the third row of figure 4.21 are adjusted for differences in consumption taxes.
111. Kenworthy 2011c, ch. 2.
112. See also Kenworthy 2011c, ch. 4.
113. Graetz and Mashaw 1999; Kim et al. 2007.
114. For more discussion, see Hacker 2002; Hills 2011; Morgan and Campbell 2011; Konczal 2012.
115. Blair 2010, p. 569.
116. Emanuel 2013.
117. Judt 2010.
118. Kaus 1992.
119. Gosselin 2008, ch. 10; Kahneman 2011; Kliff 2012.
120. And those who would like additional coverage are free to purchase it.
121. Milanovic 2011.
122. Sen 1999, ch. 4; Sperling 2005; Stiglitz 2006, ch. 3; Collier 2007, ch. 10; Galbraith 2007; Krugman 2007; Rodrick 2007, ch. 9; Blinder 2009b; DeLong 2008; Kristof 2009.
123. Kenworthy 2008b.
124. Freeman and Medoff 1984.
125. In the public sector the unionization rate is 37 percent.
126. Kruse, Freeman, and Blasi 2008.
127. Weitzman 1984.
128. Edelman 2012, p. 32.
129. Household income tends to be shared (even if not necessarily equally), so it, much more than individual earnings, determines consumption and living standards. Moreover, household income appears to have a stronger effect on happiness than personal earnings; see Firebaugh and Schroeder 2009.
130. Immervoll and Pearson 2009.
131. Herzenberg, Alic, and Wial 1998.
132. E.g., Hall and Soskice 2001; Hall and Gingerich 2004; Barth and Moene 2009.
133. Pontusson 2011.
134. Rothstein 1998.
135. Goodin et al. 1999.
136. Kenworthy 2008a, ch. 6.
137. Kenworthy 2008a.
138. Baumol 1967.
139. OECD 2008, ch. 9; Esping-Andersen and Myles 2009; Paulus, Sutherland, and Tsakloglou 2009; Garfinkel, Rainwater, and Smeeding 2010; Kenworthy 2011c, ch. 7.
140. According to Duncan Gallie, among those responding to a mid-1990s survey in various European Union countries, semi-skilled and non-skilled workers were less likely than higher-skilled workers to report being in jobs in which they could “definitely learn new things or exercise significant influence over the way things are done” (Gallie 2002, p. 100). Surveys from a variety of affluent countries suggest an increase in work intensity and work effort in the 1980s and 1990s (Gallie 2002; Green 2006; Kalleberg 2011). Interestingly, in the mid-1990s EU survey there was no difference across skill groupings in the likelihood of experiencing heightened work pressure. But Gallie points out that “there is now a wide body of research that points to the fact that the long-term health effects of increased pressure are likely to be particularly severe among the low-skilled. This is because the impact of work pressure is mediated by the degree of control that employees can exercise over the work task. Where people are allowed initiative to take decisions themselves about how to plan and carry out their work, they prove to be substantially more resilient in the face of high levels of work pressure. It is jobs that combine high demand with low control that pose the highest health risks” (Gallie 2002, p. 105–106).
141. O’Toole and Lawler 2006.
142. Gallie 2003. Although not to quite the same extent, Finland’s government and interest group organizations also have encouraged greater attention to these issues among employers. Francis Green has examined survey data on employee discretion and influence over their work tasks in the United Kingdom and Finland, the only two countries for which there are comparable data over a number of years. He finds evidence of decreased worker discretion in the UK but increased discretion in Finland. See Green 2006, ch. 5.
143. Gallie 2002, pp. 120–122.
144. Presser 2003.
145. Presser 2003, ch. 9.
146. Esping-Andersen 1999, 2009; Esping-Andersen et al. 2002; Andersson, Holzer, and Lane 2005; Fitzgerald 2006.
147. Steensland 2007.
148. Van Parijs 2001; Murray 2006. For more, see Wright 2010; Widerquist 2013.
149. Van Parijs 2001, pp. 3, 19.
150. Moffitt 1981.
151. Kenworthy 2008a.
152. See Galston 2001.
153. Bergmann 2006.
154. Ackerman and Alstott 1999; Sherraden 2007; Boshara 2009; Cramer and Newville 2009.
155. Economist 2013a.
156. Barr 2012.
157. Wooldridge 2013.
158. Alesina and La Ferrara 2005; Alesina, Harnoss, and Rapoport 2013; Kenworthy 2013c. Moreover, the Nordic countries are not as homogenous as is sometimes assumed. For instance, though United States’ foreign-born share is larger than Denmark’s, it is lower than Sweden’s.
159. Lindert 2004; Baumol, Litan, and Schramm 2007.
1. Lipset 1996.
2. Micklethwait and Woolridge 2004, pp. 382, 303.
3. See also Brooks and Manza 2007.
4. Pew Research Center, ABC/Washington Post, CBS/New York Times, Los Angeles Times, reported in Pew Research Center 2011a, pp. 109–110. “Depends” and “don’t know” responses omitted.
5. Gallup 2011. “No opinion” responses omitted.
6. Pew Research Center 2011a, p. 147.
7. National Election Study, sda.berkeley.edu/archive.htm.
8. General Social Survey, sda.berkeley.edu/archive.htm, series conlegis and confed.
9. General Social Survey, sda.berkeley.edu/archive.htm, series natfarey, nateduc, natheal, and natsoc.
10. Reported in the data set for Page and Jacobs 2009, series qhc2.
11. Pew Research Center 2011b, p. 24.
12. Data set for Page and Jacobs 2009, series qtaxl.
13. Data set for Page and Jacobs 2009, series qtaxm.
14. Data set for Page and Jacobs 2009, series qjob4.
15. General Social Survey, sda.berkeley.edu/archive.htm, series natfare.
16. Gilens 1999.
17. Wilentz 2008.
18. Manza, Heerwig, and McCabe 2012; Kenworthy 2013b.
19. Page and Shapiro 1983.
20. Gilens 2012.
21. Erikson, MacKuen, and Stimson 2002.
22. Newman and Jacobs 2010.
23. Ibid., p. 7.
24. Ibid., p. 5.
25. Pierson 1994.
26. Hirschman 1991.
27. “Republican Hopefuls Deny Global Warming,” Guardian, September 14, 2010; “House Repubs Vote That Earth Is Not Warming,” Scientific American, March 16, 2011.
28. Kenworthy, Barringer, Duerr, and Schneider 2007, using General Social Survey data.
29. Larry Bartels finds no decline in the share of whites in the bottom income third that voted Democratic in presidential elections between 1952 and 2004. But in eight of those fourteen elections a majority of this group voted for the Republican candidate, so it can’t really be considered to have been the electoral base of the Democrats. See Bartels 2008; Kenworthy 2010c.
30. Teixeira 2011; Teixeira and Halpin 2012.
31. Inglehart 1977; Inglehart and Abramson 1994.
32. Judis and Teixeira 2002. See also Browne, Halpin, and Teixeira 2011.
33. Hibbs 2009; Bartels 2013.
34. Bartels and Zaller 2000; Fair 2012; Silver 2012; Sides and Vavreck 2013.
35. Hibbs 2012; McGhee 2012.
36. Bai 2012.
37. Center for Responsive Politics, “Outside Spending,” www.opensecrets.org.
38. Bai 2012.
39. McGregor 2012; Starr 2012; Moser 2013.
40. Jacobs et al. 2004; Drutman 2012.
41. Feingold 2013.
42. Keith Poole, “Party Polarization: 1879–2010,” polarizedamerica.com/Polarized_America.htm, 2013.
43. Schattschneider 1960; Ferguson and Rogers 1986; Domhoff 1990; Hacker and Pierson 2010.
44. Ferguson and Rogers 1986; Vogel 1989; Krugman 2007a; Hacker and Pierson 2010.
45. Blank 1997; Garfinkel, Rainwater, and Smeeding 2010; Ben-Shalom, Moffitt, and Scholz 2011; Edelman 2012; Meyer and Sullivan 2012.
46. Gilens 1999.
47. See also Meyer and Sullivan 2012.
48. Huber, Ragin, and Stephens 1993; Tsebelis 1995; Amenta 1998. There is the additional possibility of veto by the judicial branch. Lobbying, too, plays a role in minimizing policy change; see Baumgartner et al. 2009.
49. Baldassarri and Gelman 2008; Bartels 2008.
50. Poole and Hare 2012.
51. On top of ideological purification, the leadership in both parties has taken to using an array of rewards and punishments—from allocation of committee positions to backing of reelection campaigns—to get backbenchers to vote the party line.
52. Mann and Ornstein 2012, pp. 57–58.
53. Keith Poole, “Party Unity Scores,” pooleandrosenthal.com/party_unity.htm, 2013.
54. Grunwald 2012, p. 67.
55. Douthat and Salam 2008; Frum 2008; Brooks 2012; Murray 2012a; Ponnuru 2013.
56. Skocpol and Williamson 2012.
57. See “The Tortoise and the Hare,” one of Aesop’s fables.
1. Emigration from the United States may well increase as the rest of the world develops economically and language barriers ease, though it likely will remain a small-scale phenomenon.
2. Helle Thorning-Schmidt, speech at the conference on Progressive Governance: Towards Growth and Shared Prosperity, organized by Policy Network and Global Progress, Copenhagen, April 11–12, 2013.
3. See, for instance, the 2011 executive order “Improving Regulation and Regulatory Review.”