Synthetic Control Methods for Comparative Case Studies: Estimating the Effect of California's Tobacco Control Program

• p. 35

  Abadie, A. and Gardeazabal, J. (2003), “The Economic Costs of Conflict: A Case Study of the

Only the authors can say what is right here.

Judge's note: Rests on outside knowledge: the published article (AER 93(1), doi:10.1257/000282803321455188) begins on page 113, not 112. The p.35 image confirms the printed '112-132'. The files cannot show the true first page.

Second check: another model agreed, code check

• p. 35

  Card, D. (1990), “The Impact of the Mariel Boatlift on the Miami Labor Market”, Industrial

Only the authors can say what is right here.

Judge's note: Rests on outside knowledge: the article appeared in ILR Review vol. 43, no. 2 (1990), not vol. 44. The p.35 image confirms the printed 'vol. 44'. The files cannot show the true volume.

Second check: another model agreed, code check

• p. 13 II.D

  to estimate native employment rates in Miami

• p. 3 footnote 1

  estimate the un- employment rates of native workers

Judge's note: Both passages were checked on the page images. They describe the same study's outcome in different terms. The two measures are related labour-market outcomes, so this is a minor consistency point.

Second check: another model agreed

• p. 7 Section II.B

  all i ∈ {1, . . . , N}

• p. 10 Section II.B, below eq. (4)

  {w∗j}2≤j≤N

• p. 34 Appendix C

  {w∗j}2≤j≤N is a deterministic

• p. 34 Appendix C

  {w∗ j}2≤j≤N

Judge's note: Page 7 defines units i = 1,…,J+1 and in the same sentence uses 'all i ∈ {1, . . . , N}'. Page 10 has F_t indexed 1≤j≤N and {w*_j}_{2≤j≤N}, followed directly by eq. (5), which sums j=2 to J+1. Page 34 (Appendix C) repeats {w*_j}_{2≤j≤N}. A grep finds no definition of N as a count. N evidently stands for J+1, but it is never introduced and clashes with the Y^N superscript.

Second check: another model agreed

• p. 19 Section III.B

  These variables are averaged over the 1980-1988 period

• p. 48 Table 1 note

  (beer consumption is averaged 1984-1988)

• p. 19 III.B

  averaged over the 1980-1988 period

• p. 48 Table 1 note

  beer consumption is averaged 1984-1988

Only the authors can say what is right here.

Judge's note: Page 19 lists beer consumption among the predictors and gives a single averaging window of 1980-1988. The Table 1 note on page 48 gives an exception of 1984-1988 for beer. Appendix A (the beer data source) says nothing about availability before 1984. The two descriptions disagree, and only the authors can say which window was used and why.

Second check: another model agreed

• p. 21 Section III.C

  Combined with the high balance on all smoking predictors (Table 1)

• p. 21 Section III.C

  the synthetic California accurately reproduces the values that smoking prevalence and smoking prevalence predictor variables had in California

• p. 48 Table 1, Ln(GDP per capita) row

  10.08

Only the authors can say what is right here.

Judge's note: Table 1 (page 48) shows close matches on the other predictors. The income row shows no improvement over the donor average. The 'all predictors' wording on page 21 therefore overstates the balance. The V-weight on income may be small, but the text does not say so.

Second check: another model agreed

• p. 7 Footnote 2

  cases studies are used to test hypotheses

Judge's note: Seen in footnote 2 on the page 7 image.

Second check: none (judge only)

• p. 7 II.B

  Let T0 be number of

Judge's note: Seen on the page 7 image, Section II.B.

Second check: none (judge only)

• p. 11 Footnote 7

  similar to the exposed units in term of the values of X1

Judge's note: Seen in the footnote 7 crop from page 11.

Second check: none (judge only)

• p. 12 Footnote 8

  the variables in in X0 and X1

• p. 12 footnote 8

  variables in in X0 and X1

• p. 12

  le variances of the variables in in X0 and X1 (normaliz

Judge's note: Seen in the footnote 8 crop from page 12; the duplication is in the typeset text, not an extraction artefact.

Second check: code check

• p. 13 Section II.D

  our ignorance about the the ability of the control group

• p. 13 II.D

  ignorance about the the ability

• p. 13

  ives from our ignorance about the the ability of the cont

Judge's note: Seen on the page 13 image.

Second check: code check

• p. 15 footnote 12

  Health Educa- tion(20 percent)

• p. 17 III.A

  Evans, Farrelly, and Montgomery, (1999)

Judge's note: Both seen on the page images (crops C009-a and C009-b).

Second check: none (judge only)

• p. 18 Section III.A, Fichtenberg and Glantz paragraph

  first temporally eliminated and then reestablished

Judge's note: Seen on the page 18 image.

Second check: none (judge only)

• p. 21 Section III.C, Table 2 paragraph

  smoking trends in California prior to the passage of Proposition 99 is best reproduced

Judge's note: Seen on the page 21 image.

Second check: none (judge only)

• p. 31 Appendix B data sources

  Export plus Imports as percentage of GDP

Judge's note: The text layer shows the wording as quoted. This is a trivial grammatical inconsistency: 'Exports plus Imports' would be parallel. The meaning is unaffected.

Second check: none (judge only)

• p. 32 Appendix C

  restrict ourselves to the case that F = 1, so there is one unobserved factor

• p. 32 Appendix C, first paragraph

  we will restrict ourselves to the case that F = 1, so there is one unobserved factor

Judge's note: A text search finds 'F' as a symbol only at this location; elsewhere it occurs only in names. The model in Section II.B and at the start of Appendix C has a single factor λtμi, and no symbol is introduced for the number of factors. The following clause makes the meaning clear, so this is a minor notation gap.

Second check: none (judge only)

• p. 32 Footnote 19

  correlated across units and in time because the presence of the term λtµi

Judge's note: The p.32 image confirms the wording 'because the presence of the term'. It should read 'because of the presence of'.

Second check: none (judge only)

• p. 34 Appendix C, last sentence

  the discrepancies in equation in equation (5) go to zero

• p. 34 Appendix C, last paragraph

  discrepancies in equation in equation (5)

• p. 34

  equation in equation (5) go to zero

Judge's note: The p.34 image, last sentence, shows the duplicated words.

Second check: none (judge only)

• p. 34 Appendix C

  there exist two finite constant λmin > 0

Judge's note: The p.34 image confirms 'two finite constant'.

Second check: none (judge only)

• p. 36 References, Meinhardt et al. (1995)

  Wirtschaftfroschung

• p. 36 References, Meinhardt et al. (1995)

  Wirtschaftfroschung, Sonderheft 154

Judge's note: The p.36 image confirms the spelling 'Wirtschaftfroschung'. This is a transposition of 'or' plus a missing 's'.

Second check: none (judge only)

• p. 36 References, Longo et al.

  Krusea, R.

Only the authors can say what is right here.

Judge's note: Rests on outside knowledge: the third author of Longo et al. (2001), Tobacco Control 10, 267-272, is R. Kruse. I checked the page 36 crop and it reads "KRUSEA, R." as quoted. The trailing 'a' looks like a typo.

Second check: none (judge only)

• p. 48 Table 1

  Cigarette sales Per capita 1988

Judge's note: I checked the Table 1 image on p.48. The 1988 row reads "Per capita" with a capital P, while the 1980 and 1975 rows read "per capita". The inconsistency is cosmetic only.

Second check: none (judge only)