The Gender Wage Gap: Extent, Trends, and Explanations

• p. 77 Table 7, Fortin (2008) sample

  US NELS 1972 and 1988

• p. 20 Section 3.3

  the National Longitudinal Study of the High School Class of 1972-NLS 72-

Judge's note: The Table 7 image (p.77) reads "US NELS 1972 and 1988 Cohorts: 1979, 1986 and 2000". The text on p.20 names the two surveys as NLS-72 and NELS 88. The table label wrongly applies the NELS name to the 1972 cohort.

Second check: another model agreed

• p. 7 Section 2, specification description

  of 14 industry and 20 occupation dummy variables

• p. 29 Section 3.7

  industry (14 categories and government employment) at 18 percent

• p. 7 Section 2, specification

  14 industry and 20 occupation dummy variables

• p. 29 Section 3.7

  based on only 21 occupations

Only the authors can say what is right here.

Judge's note: p7 and p29 page images checked. If the 20 occupation dummies stand for 21 occupations, then 14 industry dummies would stand for 15 industries. The paper gives no rule that explains the difference (for example, no omitted industry category). Only the authors can say which count is right.

Second check: another model agreed

• p. 28 Section 3.7, first paragraph

  actually increased from 27% of the 1980 gap to 49% of the much smaller 2010 gap

• p. 9 Section 2.1, Table 4 Panel B discussion

  or 20% of the gender pay gap

• p. 39 Section 4.1

  occupation and industry differences accounted for about 51% of the gender pay gap

• p. 28 Section 3.7, first paragraph

  increased from 27% of the 1980 gap to 49% of the much smaller 2010 gap

Only the authors can say what is right here.

Judge's note: In Table 4 Panel B (p.74), industry+occupation is 9.6+10.7=20.3% (1980) and 17.6+32.9=50.5% (2010). Adding unionization gives 26.5% and 49.2%, which are the printed 27% and 49%. On p.9 the paper itself treats union coverage separately ('20% ... with ... union coverage contributing an additional ... 6 percent'). 'Factors like' may be meant to include unionization, but the text does not say so, so the reader sees two sets of numbers for 'occupation and industry'.

Second check: another model agreed

• p. 12 Section 2.2, discussion of Table 6

  resulting in comparable declines of 0.09-0.10 log points in the gender wage gap across the distribution

• p. 76 Table 6 Panel A

  0.2381

• p. 12 Section 2.2

  comparable declines of 0.09-0.10 log points in the gender wage gap across the distribution

• p. 76 Table 6 Panel A

  0.1767 0.2729 0.0721 0.1648

Judge's note: Read Table 6 on the p76 page image. The 1980-minus-2010 covariate effects are 0.105, 0.098 and 0.087 (human capital) and 0.108, 0.111 and 0.104 (full). Several of these fall outside the stated 0.09-0.10 range. By contrast, the coefficient-effect ranges stated in the same paragraph (0.18-0.20 and 0.16-0.18) do match the table. The stated range is only an approximate summary.

Second check: another model agreed

• p. 76 Table 6 Panel C, 1980 and 2010

  0.4725 0.4615 0.1855 0.1967

• p. 71 Table 1 PSID

  1980 2282 1491 62.1% 64.8% 60.1% 62.4%

• p. 12 footnote 16

  the results closely match the actual percentiles

• p. 76 Table 6, Panel C, 1980 and 2010 (PSID)

  0.4725 0.4615

Only the authors can say what is right here.

Judge's note: I checked Table 1 (p.71) and Table 6 (p.76) on the page images. Both cover nonfarm wage and salary workers aged 25-64 with at least 26 weeks of employment, and Table 1 is titled full-time workers. Table 1 entries are exp(D), the female-minus-male percentile log wage, so the implied gap is -ln(ratio). The 1980 median matches: -ln(0.601)≈0.509 against 0.5091. The 2010 90th percentile is close: -ln(0.739)≈0.302 against 0.3014. The others differ: the 1980 10th is ≈0.434 against 0.4725, the 1980 90th ≈0.472 against 0.4455, the 2010 10th ≈0.205 against 0.1855, and the 2010 50th ≈0.194 against 0.2073. Table 1 thus puts the 1980 90th-percentile gap above the 10th-percentile gap, while Table 6 does the reverse. Panel C sums differ between the Human Capital and Full columns (0.4725 vs 0.4615), so they are model-based reconstructions and need not equal the raw gaps exactly. Still, footnote 16's 'closely match' is hard to square with differences of about 0.04 and a reversed ordering. Only the authors can say whether the sample, the weights or the approximation explain this. The Panel C sums of Panels A and B check out.

Second check: another model agreed

• p. 28 Section 3.7

  much of this effect was offset by adverse (to women) movements in returns to occupations

• p. 75 Table 5, Occupation Variables

  0.0506

Judge's note: Read from the Table 5 image (p.75). The net occupation effect is +0.0253 under both bases. This fits the p.10 statement that occupation accounted for a larger gap in 2010 than in 1980. 'Much of' understates a more-than-complete offset, so this is a wording point only.

Second check: another model agreed

• p. 2 Abstract

  more slowly at the top of the wage distribution that at the middle or the bottom

• p. 51 Conclusion, first paragraph

  at the top of the wage distribution that at the middle or the bottom

• p. 4 Introduction

  in Section 6, we highlights that the overall structure

• p. 2 Abstract

  much more slowly at the top of the wage distribution that at the middle or the bottom

Judge's note: Checked the page images. The abstract crop and the conclusion crop both show 'that at the middle'. The p. 4 crop shows 'we highlights'. These are plain typos.

Second check: none (judge only)

• p. 2 Abstract, first sentence

  Using PSID microdata over the 1980-2010, we provide

Judge's note: The page image of the p2 abstract shows the text as quoted. On p5 the paper itself writes 'the 1980-2010 period'.

Second check: none (judge only)

• p. 4 Introduction, Section 5/6 overview

  Section 6, we highlights that the overall structure

• p. 4 Introduction, plan of the paper

  Section 6, we highlights that the overall structure of wages

• p. 4

  in Section 6, we highlights that the overall structure

Judge's note: Confirmed on the p4 page image.

Second check: none (judge only)

• p. 4 Introduction, Section 4 preview

  not accounted for gender differences in measured qualifications

Judge's note: Confirmed on the p4 page image.

Second check: none (judge only)

• p. 5 Section 2, paragraph on Table 1

  Because i earnings refer to the previous year

• p. 4 Intro

  in Section 6, we highlights that

Judge's note: The p. 5 crop shows a plain 'i' after footnote marker 3; it is not an extraction artefact. The 'we highlights' item belongs to C001.

Second check: none (judge only)

• p. 25 footnote 43

  There are also a number studies

• p. 4 Introduction

  in Section 6, we highlights

• p. 17 footnote 23

  for the group with low-education, low-experience group

Judge's note: Both crops match the quotes. The 'we highlights' item is covered in C001.

Second check: none (judge only)

• p. 8 footnote 13

  taking a women and valuing her characteristics

• p. 8

  the experiment of taking a women and valuing her characteristics

Judge's note: The footnote 13 text in the crop matches the quote.

Second check: none (judge only)

• p. 10 Section 2.1, Table 5 introduction

  Juhn, Muphy and Pierce (1991)

Judge's note: The p. 10 crop shows 'Muphy'. The reference entry is 'Juhn, Chinhui, Kevin M. Murphy, and Brooks Pierce. 1991', and the later text spells it 'Murphy'.

Second check: none (judge only)

• p. 10 Section 2.1

  unexplained factors continue account for a substantial share

Judge's note: Confirmed on the p10 page image.

Second check: none (judge only)

• p. 13 Section 2.2

  Kassenboehmer and Simming (2014)

• p. 62 Reference list

  Kassenboehmer, Sonja C. and Mathias G. Sinning.

• p. 62 References

  Kassenboehmer, Sonja C. and Mathias G. Sinning. 2014.

• p. 62 Reference list

  Kassenboehmer, Sonja C. and Mathias G. Sinning

Judge's note: Checked the p. 13 text and the p. 62 reference image. I also checked p. 36 against the reference entry (Heyman, Fredrik, Helena Svaleryd...). The Muphy and Sapenza sub-items are covered by C013 and C026.

Second check: none (judge only)

• p. 36 Section 3.9

  Even to the extent this the case

• p. 36 Section 3.9

  quantitatively important in explaining of the gender pay gap

• p. 20 Section 3.3

  raw gender pay at a point in time

• p. 27 Section 3.6

  expecting that the relationship to be causal

Judge's note: Each phrase appears as quoted on the page images.

Second check: none (judge only)

• p. 24 Section 3.4

  our estimate of the overall the gender pay gap

Judge's note: The image matches the quote.

Second check: none (judge only)

• p. 27 Section 3.6

  Indeed, it is has been found

Judge's note: The image matches the quote.

Second check: none (judge only)

• p. 35 Section 3.9, fifth study

  Correll, Bernard, and Paik (2007)

• p. 58 Reference list

  Correll, Shelley J., Stephen Benard, and In Paik. 2007.

• p. 35

  Fifth, we point to the results of the study by Correll, Bernard, and Paik (2007)

Judge's note: The p. 35 crop shows 'Bernard'. The reference entry on p. 58 reads 'Stephen Benard', and the earlier text also uses 'Benard'.

Second check: none (judge only)

• p. 35 Section 3.9, fourth study

  Reuben, Sapenza, and Zingales (2014)

• p. 65 Reference list

  Reuben, Ernesto, Paola Sapienza, and Luigi Zingales. 2014.

• p. 35

  A fourth study, by Reuben, Sapenza, and Zingales (2014)

Judge's note: The misspelling is confirmed. One reviewer also said 2014 conflicts with 2015, but that part is wrong: the reference list has two separate Reuben, Sapienza and Zingales entries, 2014 (PNAS) and 2015 (NBER WP).

Second check: none (judge only)

• p. 35 Section 3.9

  They found that that IAT scores

Judge's note: The image matches the quote.

Second check: none (judge only)

• p. 36 Section 3.9

  Heyman, Svalerty, and Vlachos’ (2013)

• p. 61 Reference list

  Heyman, Fredrik, Helena Svaleryd, and Jonas Vlachos. 2013.

Judge's note: Checked the p.36 crop and the p.61 reference entry.

Second check: none (judge only)

• p. 39 Section 4.1, paragraph on Table 7

  industry effects may have due to psychological factors

• p. 41 Section 4.1

  Mueller and Plug(2006) found

• p. 39 Section 4.1

  controls for these measurable may

• p. 39 Section 4.1

  may have due to psychological factors

Judge's note: All three slips are visible on the page images.

Second check: none (judge only)

• p. 77 Table 7 and note

  Manning and Swafford (2008)

• p. 41 Section 4.1

  Manning and Swafford (2008) used separate

• p. 63 Reference list

  Manning, Alan and Joanna Swaffield. 2008.

• p. 39 Section 4.1

  Manning and Swafford (2008)

Judge's note: The Table 7 crop (p.77) shows 'Manning and Swafford (2008)'. The reference on p.63 spells the name 'Swaffield'. A grep found 'Swafford' at every in-text occurrence (pp.39 and 41, footnote 66, Table 7 and its note). This is a consistent misspelling of the co-author's name.

Second check: none (judge only)

• p. 41 Footnote 62

  Niederle and Westerlund (2007)

Judge's note: Checked the footnote crop. The reference list and the other in-text citations spell it Vesterlund.

Second check: none (judge only)

• p. 41 Section 4.1, Manning and Swafford paragraph

  then then the male and female

• p. 41

  then then the male and female wage functions

• p. 41

  were to change their traits, then then the male and female

Judge's note: The image matches the quote.

Second check: code check

• p. 41

  Westerlund (2007). Men were found to be more competitive

Judge's note: In the crop, footnote 62 reads 'Niederle and Westerlund (2007)'. The reference list entry is 'Niederle, Muriel and Lise Vesterlund. 2007', and every other in-text citation says Vesterlund.

Second check: none (judge only)

• p. 41

  competitiveness did not appear explain the

Judge's note: The crop of footnote 62 shows the text as quoted. 'appear explain' needs 'to'.

Second check: none (judge only)

• p. 44 Section 4.3

  nonmathmatically-oriented classes

Judge's note: I checked the crop of p.44. The image reads 'nonmathmatically-oriented classes', so the misspelling is in the PDF and is not an extraction artefact.

Second check: none (judge only)

• p. 45 Section 4.5, first sentence

  Recent work by Bertand, Kamenica, and Pan (2015)

• p. 45 Section 4.5

  Bertand, Kamenica, and Pan (2015)

Judge's note: On the p.45 crop, the first sentence of Section 4.5 reads 'Bertand'. The next paragraph spells the name 'Bertrand, Kamenica, and Pan', and so does the reference list. The slip is internal to the document.

Second check: none (judge only)

• p. 51 Conclusion, second paragraph

  explain relatively little of the gap wage gap in the aggregate

• p. 51 Conclusion

  relatively little of the gap wage gap in the aggregate

Judge's note: The p.51 crop shows the same garbled phrase as the text layer.

Second check: none (judge only)

• p. 51 Conclusion, first paragraph

  declined much more slowly at the top of the wage distribution that at the

• p. 51

  slowly at the top of the wage distribution that at the

Judge's note: The crop shows 'that at the middle'. The text extract shows the same wording in the abstract (line 35).

Second check: none (judge only)

• p. 52 Conclusion, psychological attributes paragraph

  as well as real-world settings, will likely continue to provide insights

Judge's note: The crop shows the sentence as quoted. After the relative clause, 'will likely continue' has no subject. The meaning is still recoverable.

Second check: none (judge only)

• p. 55 References, Bell 2005

  Institute for the Study of Labor Working (IZA) Working Paper No. 1689

Judge's note: The p.55 crop shows the duplicated 'Working' in the series name. It is cosmetic.

Second check: none (judge only)

• p. 60 References

  Quarterly Review of Economics, 118 (3): 1049-74

• p. 58 References

  Psychological Science 15 (3):75-141

• p. 60 References

  “Engines of Liberation.” 2005. Review of Economic Studies

• p. 57 References

  Blundell, Richard, Amanda Gosling, Hidehiko Ichimura, and Costas Meghir. .2007

Judge's note: I checked the crops on pp.57, 58 and 60. The duplicated year and the misplaced period are visible as printed. The Gneezy et al. journal-name point is covered under C044. Rests on outside knowledge: that the Ceci et al. journal is Psychological Science in the Public Interest. That item should go to the authors as a query.

Second check: none (judge only)

• p. 71 Table 1 panel headers

  Panel Study of Income Dynamic (PSID)

• p. 71 Table 1 panel headers

  March Current Populations Survey (CPS)

Judge's note: The p.71 crops show both misspelled headers as printed.

Second check: none (judge only)

• p. 71

  Panel Study of Income Dynamic (PSID)

Judge's note: The page image of Table 1 (p.71) shows "Panel Study of Income Dynamic (PSID)", with no final 's'. This is a cosmetic misspelling of the survey name.

Second check: none (judge only)

• p. 75 Table 5 notes

  is the the change over the 1980-2010

• p. 76 Table 6 notes

  Standard error are in parentheses

• p. 77 Table 7 title

  Gender Pay gap

Judge's note: All three are visible in the crops of pp.75, 76 and 77.

Second check: none (judge only)

• p. 76 Table 6 notes

  Standard error are in parentheses

Judge's note: The Table 6 notes on the p.76 image show "Standard error are in parentheses and are computed by bootstrapping with 100 repetitions." The subject and verb do not agree in number.

Second check: none (judge only)