Electoral Accountability and Corruption: Evidence from the Audits of Local Governments
American Economic Review, 2011 · paper · Political science
Uses Brazilian anti-corruption audit reports to test whether electoral accountability limits local corruption. Mayors eligible for reelection misappropriate 27 percent fewer resources than term-limited mayors, with larger effects where information access and judicial punishment are weaker.
Checked: PDF from nber.org, fetched 5 Oct 2026. Page numbers are those of that PDF.
12 findings
Substance Conclusions (p.24): "The estimated differential effect between first and second-term mayors is more pronounced among municipalities where the cost of corruption is higher - as measured by the presence of a judge". Table 8, the abstract and Section 6 show the differential is smaller where a judge is present, so 'higher' should read 'lower'.
- p. 24 Conclusions, para 2
more pronounced among municipalities where the cost of corruption is higher - as measured by the presence of a judge
- p. 2 Abstract
where the likelihood of judicial punishment is lower
- p. 38 Table 8 col 1
First-term × Judiciary district 0.037
- p. 21 Section 6, local context
in municipalities where the costs of engaging in corruption are higher, the difference in corruption between first and second-term mayors will also be lower
Judge's note: Table 8 col 1 (p.38 image): Mayor in first-term -0.049 [0.017]***, First-term × Judiciary district +0.037 [0.020]*. The implied gap with a judge is -0.012, against -0.049 without one. The abstract (p.2) says effects are "more pronounced ... where the likelihood of judicial punishment is lower". Section 6 (p.21) says the presence of a prosecutor raises the cost of corruption and that the differential is lower there. The conclusion states the opposite direction.
Second check: another model agreed
Report a problem with this finding
Substance The same R$148,000 is converted as "R$148,000 (US$54,688)" in the Introduction (p.2) and as "R$148,000 (US $60,000)" in Basic Results (p.14).
- p. 4 Introduction
R$148,000 (US$54,688)
- p. 16 Basic Results
R$148,000 (US $60,000)
- p. 4 Introduction
approximately R$148,000 (US$54,688)
- p. 16 Section 6
approximately R$148,000 (US $60,000)
Only the authors can say what is right here.
Judge's note: Read both passages on the page images. The introduction's own pair (5,459,209 / 2,017,259) implies about 2.706 R$/US$, which gives US$54,688. US$60,000 implies about 2.47. Rounding cannot account for a gap of roughly 10%. A reader of p.14 would take away a different dollar value.
Second check: another model agreed
Report a problem with this finding
Substance The text says "496 municipalities randomly selected across the first 11 lotteries", but footnote 18 gives 26 in lottery 1, 50 in each of lotteries 2-9 and 60 from lottery 10 ("May of 2004"), which totals 546. The May 2004 date also sits oddly with reports being "available in the beginning of 2004".
- p. 10 Section 4.1 and footnote 18
496 municipalities randomly selected across the first 11 lotteries
Only the authors can say what is right here.
Judge's note: Searched the paper for any exclusion or unavailability that would explain the gap; none is stated. Only the authors can say which number or date is right.
Second check: another model agreed
Report a problem with this finding
Substance Text: "By this measure [share of service items], first-term mayors are 23 percent less corrupt than second-term mayors". Table 5 cols 5-8 give -0.008 to -0.010, and Table 2 gives a second-term mean share of items of 0.059.
- p. 17 Basic results, para 1
first-term mayors are 23 percent less corrupt
- p. 35 Table 5 col 6
-0.010
- p. 32 Table 2 share of items, Corruption second-term
0.059
- p. 17 Section 6, Table 5 discussion
first-term mayors are 23 percent less corrupt than second-term mayors
Only the authors can say what is right here.
Judge's note: Checked p17, Table 5 (p35) and Table 2 (p32). The parallel 22% figure for counts does reproduce (0.467/2.127). The same method on the share of items gives 14-17%, and no printed baseline yields 23%. Only the authors can say what produced 23%.
Second check: another model agreed
Report a problem with this finding
Substance "When only 9 percent of the legislative council is from the same political party as the mayor ... there an 8.8 percentage point difference". Table 8 col 3 (-0.066 main effect, 0.120 interaction) gives about 5.5 pp at 9%.
- p. 22 Local context, para 2
there an 8.8 percentage point
- p. 38 Table 8 col 3
-0.066
- p. 22 Local context, para 2
there an 8.8 percentage point difference between first and second-term mayors
- p. 38 Table 8 col 3
First-term × Legislative support 0.120
Only the authors can say what is right here.
Judge's note: Table 8 col 3 read on p38 upright: first-term -0.066 [0.025], interaction 0.120 [0.060]. The same coefficients reproduce the text's zero point at 55%, and col 4 reproduces the text's 3.9 pp and -0.007. At 9% they give -0.055, not -0.088. No support share between 0 and 1 gives 8.8. The sentence is also missing the verb "is" ("there an").
Second check: another model agreed · checked by hand
Report a problem with this finding
Substance "which represents an average R$523,000 (US$237,592)" implies about 2.20 R$/US$. The paper's other conversions imply about 2.7 (R$5,459,209 = US$2,017,259; R$433 million = US$160 million).
- p. 22 Local context para 1
R$523,000 (US$237,592)
Only the authors can say what is right here.
Judge's note: Checked p22. At the rate used elsewhere in the paper, R$523,000 would be about US$193,000. No other exchange rate is stated anywhere in the paper.
Second check: another model agreed
Report a problem with this finding
Reference or label Eq.1 defines I_i as the first-term indicator, but the RDD model sets "I_i = 1[W_i ≥ 0]". Here W_i is the incumbent's vote-share margin, so I_i = 1 marks incumbent wins, which are second-term mayors.
- p. 14 Section 5.1
Ii = 1[Wi ≥ 0]
- p. 13 Equation 1
Ii indicates whether the mayor is in his first term
Judge's note: On p.11 I_i "indicates whether the mayor is in his first term". On p.12, W_i is the incumbent minus runner-up vote share, so W_i ≥ 0 means the incumbent was re-elected. The indicator as written equals the second-term dummy. Tables report the coefficient as 'Mayor in first-term'. The likely fix is 1[W_i < 0].
Second check: another model agreed
Report a problem with this finding
Reference or label Table 4 and Table 5 notes cite "Abadie and Imbens (2004) matching estimator", but the text (p.16) cites "Abadie and Imbens 2006" and the reference list has only the 2006 Econometrica entry.
- p. 34 Table 4 notes
Abadie and Imbens (2004) matching
- p. 16 Basic results
(Abadie and Imbens 2006)
- p. 28 References
Abadie, Alberto, and Guido Imbens. 2006.
- p. 35 Table 5 notes
Abadie and Imbens (2004) matching estimator
Judge's note: Read the table notes on the p.34 and p.35 images (2004), the text on p.16 (2006) and the reference list on p.28 (only 'Abadie, Alberto, and Guido Imbens. 2006 ... Econometrica 74 (1)'). No 2004 entry exists.
Second check: another model agreed
Report a problem with this finding
Reference or label Text: "In column 2 we control linearly for the difference in vote shares while in columns 3 and 4 we include a quadratic and a cubic term respectively". Table 6 has RDD linear, quadratic and cubic in columns 1, 2 and 3; column 4 is the re-elected comparison.
- p. 18 Ability section, para 1
In column 2 we control linearly for the difference in vote shares while in columns 3 and 4 we include a quadratic and a cubic term
- p. 36 Table 6 header
RDD linear RDD quadratic RDD cubic
- p. 18 Section 6, RDD paragraph
In column 2 we control linearly for the difference in vote shares while in columns 3 and 4 we include a quadratic and a cubic term respectively.
- p. 18 Section 6, RDD description
while in columns 3 and 4 we include a quadratic and a cubic term respectively
Judge's note: Table 6 (p36 upright) headers: (1) RDD linear, (2) RDD quadratic, (3) RDD cubic, (4)-(5) second-term vs first-term later re-elected. The same paragraph says "columns 1-3" are RDD, and the next paragraph correctly cites columns 4 and 5. The pointers should read column 1 and columns 2 and 3.
Second check: another model agreed
Report a problem with this finding
Reference or label Table 9 numbers its five columns "(1) (2) (3) (3) (4)" and its notes say "Columns 1-4 present the results of an OLS regression", while the text (p.23) refers to the log(value) columns as "column 4" and "(column 5)".
- p. 39 Table 9 header
(1) (2) (3) (3) (4)
- p. 23 Discussion
As reported in column 4
- p. 39 Table 9 notes
Columns 1-4 present the results of an OLS
- p. 23 Section 7 text
As reported in column 4, we again do not any find
Judge's note: The p.39 image shows five data columns (five R-squared values: 0.18, 0.21, 0.21, 0.59, 0.59) under headers (1),(2),(3),(3),(4). The text on p.23 (PDF) cites columns 1-3 for the share outcome and columns 4 and 5 for log value audited. The header should read (1)-(5), and the notes' 'Columns 1-4' is also inconsistent.
Second check: another model agreed
Report a problem with this finding
Reference or label Table 3 labels the row "GDP per capita (R$ 1000)" but prints values of 5309.95 and 6189.44. These read as reais, not thousands of reais.
- p. 33 Table 3, GDP per capita row
GDP per capita (R$ 1000)
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: Brazilian municipal GDP per capita in 2002 was a few thousand reais, not millions. Read in thousands, the values would imply about R$5.3 million per capita.
Second check: another model agreed
Report a problem with this finding
Wording The same R$433 million (US$160 million) is called "almost half of what the federal government spent on the Bolsa Escola" program in 2002 (Introduction, p.2) and "more than half of what the federal government spent in the Bolsa Escola" (Conclusions, p.22).
- p. 4 Introduction
This is almost half of what the federal government spent
- p. 24 Conclusions
is more than half of what the federal government spent
- p. 4 Introduction, para 2
This is almost half of what the federal government spent on the Bolsa Escola
- p. 24 Conclusions, para 1
is more than half of what the federal government spent in the Bolsa Escola
Only the authors can say what is right here.
Judge's note: Both passages checked on the images. Same figure, same program, same year (2002), and the two descriptions cannot both be true. The paper does not give the Bolsa Escola total, so only the authors can say which wording is right.
Second check: another model agreed
Report a problem with this finding
37 minor (typos and the like; most checked by the judge only)
Wording Text: "From the bivariate relationship in column 1, we see that first-term mayors are associated with a 2.0 percentage point decrease in corruption... this estimate represents a 27 percent decline". Table 4 col 1 prints -0.019, and the Table 2 difference is 0.019.
- p. 16 Basic Results, para 2
with a 2.0 percentage point decrease in corruption
- p. 34 Table 4, column 1
-0.019
- p. 16 Basic results
with a 2.0 percentage point decrease in corruption. At an average corruption level of 0.074
- p. 2 Abstract
27 percent fewer resources
Only the authors can say what is right here.
Judge's note: Verified on Table 4 (p34 upright) and Table 2 (p32 upright). Col 1 is -0.019, and -0.020 appears only in cols 2-3. Because the table prints -0.019, the underlying value rounds to 1.9 pp, and 0.019/0.074 is 25.7%, not 27%. The 27% repeated in the abstract and discussion only follows from 0.020. The gap is small.
Second check: another model agreed · checked by hand
Report a problem with this finding
Wording Average federal transfer printed as "R$5,459,209 (US$2,017,259)" in the Introduction (p.2) but "on average R$5,459,054 was transferred" in Basic Results (p.14). Both sentences lead to the same R$148,000 figure.
- p. 4 Introduction, para 2
R$5,459,209 (US$2,017,259)
- p. 16 Basic Results, para 2
on average R$5,459,054 was transferred
- p. 4 Introduction, para 2
R$5,459,209 (US$2,017,259) of federal transfers, lame-duck mayors misappropriate approximately R$148,000 (US$54,688)
- p. 16 Basic results
R$5,459,054 was transferred to these municipalities, lame-duck mayors misappropriate approximately R$148,000 (US $60,000)
Only the authors can say what is right here.
Judge's note: Checked PDF p4 and p16 on the images. Both sentences apply the column-6 estimate to the average transfer and reach the same R$148,000. The two averages differ by R$155, and neither passage says which sample each comes from (492 vs 476). Both averages give about R$147,000, so the conclusion is the same either way. The US$ conversion mismatch is handled under C002.
Second check: another model agreed
Report a problem with this finding
Wording The Introduction says that without local media "re-election incentives reduce political corruption by 9 percentage points". Section 6 says "almost 10 percentage points", and Table 8 col 2 prints -0.098.
- p. 4 Introduction
re-election incentives reduce political corruption by 9 percentage
- p. 22 Local context results
almost 10 percentage points (standard error=0.03)
- p. 38 Table 8 col 2
-0.098
- p. 4 Introduction
re-election incentives reduce political corruption by 9 percentage points
Judge's note: Verified on p4, p22 and Table 8 (p38 upright): col 2 is -0.098 [0.029]. 9.8 pp truncated to 9 conflicts with the paper's own "almost 10". The difference is minor. The separate exchange-rate problem for R$523,000 (US$237,592) is judged under C034.
Second check: another model agreed
Report a problem with this finding
Wording Section 4.1 says reports "were available in the beginning of 2004 for the 496 municipalities randomly selected across the first 11 lotteries". Fn 18 gives 26 + 8×50 for lotteries 1-9 and 60 per lottery "Starting on the tenth lottery in May of 2004", and Tables 1-2 use 492.
- p. 10 Section 4.1
for the 496 municipalities randomly selected
- p. 31 Table 1 notes
only computed for the 492 municipalities
- p. 20 Career concerns
Out of the 485 mayors in power during the 2001-2004 term
- p. 10 Section 4.1
the 496 municipalities randomly selected
Only the authors can say what is right here.
Judge's note: Checked p10 (text and fn 18), the Table 1-2 notes ("only computed for the 492 municipalities") and p20 (485 mayors). Fn 18's lottery sizes give 546 for 11 lotteries, not 496. A tenth lottery held in May 2004 cannot have reports available at the beginning of 2004. The 476 used in the regressions is explained (municipalities with a full set of covariates), but the paper never reconciles 496 with 492. The 485 mayors on p20 may be a different population (mayors with candidate data), so that figure is not treated as a defect.
Second check: another model agreed
Report a problem with this finding
Wording Discussion (p.22): "First-term mayors misappropriate 27 percent less resources than second-term mayors even when accounting for potential biases in political ability and experience". The 27% is the bivariate Table 4 estimate (0.020/0.074, p.16), not the ability/experience specifications in Tables 6–7.
- p. 22 Discussion, para 1
First-term mayors misappropriate 27 percent less resources than second-term mayors even when accounting for potential biases in political ability and experience
- p. 22 Section 7
misappropriate 27 percent less resources than second-term mayors even when accounting for potential biases in political ability and experience
- p. 16 Section 6
this estimate represents a 27 percent decline
Judge's note: Checked the p.16 crop: 27% comes from the column-1 bivariate estimate of 2.0 pp at a mean of 0.074. Table 6 (p.36) gives -0.029 to -0.040, and Table 7 (p.37) gives -0.029 to -0.056 for the first-term coefficient. Those are all larger in magnitude, so 27% works as a conservative lower bound and the wording is defensible. Still, the sentence ties the number to the adjusted analyses, which do not produce it. This is a wording/consistency point, not a wrong number.
Second check: another model agreed
Report a problem with this finding
Wording Table 5, column 8 (Tobit) prints "-0.44" in the row labelled "R-squared". An R-squared cannot be negative, so the row label (or the statistic reported) needs clarifying.
- p. 35 Table 5, R-squared row
-0.44
Only the authors can say what is right here.
Judge's note: Viewed page 35. The value -0.44 is printed under the Tobit column in the R-squared row. A pseudo-R² from a Tobit model can be negative, so this is most likely a labelling issue: the row does not say which statistic it reports. The Tobit column in Table 4 reports 0.20 in the same row. Only the authors can say which statistic this is.
Second check: another model agreed
Report a problem with this finding
Wording Table 7 column 1, labelled "Full sample", reports Observations "467", while the full sample is 476 municipalities elsewhere (Table 3 notes, Table 8 and Table 9 notes: "all 476 municipalities").
- p. 37 Table 7 col 1
467
- p. 38 Table 8 notes
all 476 municipalities
- p. 37
Observations 467
Only the authors can say what is right here.
Judge's note: The p.37 image shows 467 under 'Full sample'. The p.38/p.39 notes say 476, and the Table 3 notes say 476 municipalities have non-missing characteristics. I found nothing in the text explaining the loss of 9 observations, for example missing political-experience data. It could be a transposition or a real sample difference; only the authors can say.
Second check: another model agreed
Report a problem with this finding
Wording Table 8: in column 4, "-0.045 [0.022]***" and, in column 2, First-term × Media "0.084 [0.029]**". The significance markers and the coefficient/standard-error ratios do not correspond under the declared 99/95/90 percent levels.
- p. 38 Table 8 col 4 and col 2 interaction
[0.022]***
Only the authors can say what is right here.
Judge's note: Viewed page 38. In column 4, the ratio is about 2.05, which reaches the 95 percent level but not 99 percent, yet the entry carries ***. For First-term × Media, the ratio is about 2.9, above the 99 percent cutoff, yet the entry carries only **. The other entries in the table match their markers. The authors should check how the stars were assigned.
Second check: another model agreed
Report a problem with this finding
Typo "This allow us to control" should read "This allows us to control".
- p. 4 Introduction, RDD paragraph
This allow us to control for many unobserved
Judge's note: Verified in the p.4 crop.
Second check: none (judge only)
Report a problem with this finding
Typo "electoral accountability act as a powerful mechanism" should read "acts".
- p. 5 Introduction, para 3
electoral accountability act as a powerful mechanism
Judge's note: Verified on p.5. 'Suggest' here means 'indicate', not 'propose', so the mandative subjunctive does not apply and 'acts' is required.
Second check: none (judge only)
Report a problem with this finding
Typo "a growing literature that test" should read "tests".
- p. 5 Introduction, para 3
a growing literature that test the traditional political economy
Judge's note: Verified on p.5.
Second check: none (judge only)
Report a problem with this finding
Typo "by using objective measure of corruption" is missing an article; it should read "an objective measure" or "objective measures".
- p. 5 Introduction, para 3
by using objective measure of corruption
Judge's note: Verified on p.5.
Second check: none (judge only)
Report a problem with this finding
Typo Footnote 7: "is an adaptation the political agency model" is missing "of".
- p. 5 Footnote 7
is an adaptation the political agency model
Judge's note: Verified on p.5.
Second check: none (judge only)
Report a problem with this finding
Typo The number of Brazilian municipalities is given as "5,560" (p.6) and as "5500" in footnote 16 (p.7).
- p. 8 Section 3
the 5,560 Brazilian municipalities
- p. 9 Footnote 16
Brazil’s 5500 municipalities
Judge's note: Footnote 16 is phrased approximately ('approximately 8 percent'), so 5500 is plausibly a rounded figure. It is a minor inconsistency only.
Second check: none (judge only)
Report a problem with this finding
Typo "are all invited witness the lottery" is missing "to".
- p. 9 Section 3.2
are all invited witness the lottery
Judge's note: Verified on p.7.
Second check: none (judge only)
Report a problem with this finding
Typo Footnote 22 has a stray word: "were decentralized during of the late 1980s".
- p. 11 Footnote 22
were decentralized during of the late 1980s
Judge's note: The crop of footnote 22 on p.9 shows the extra 'of' as quoted.
Second check: none (judge only)
Report a problem with this finding
Typo Copy-editing errors: "the Hefindahl index was 0.08" (p.20; spelled Herfindahl elsewhere), "we again do not any find" (p.21), "decentralized during of the late 1980s" (fn 22), "Quaterly Journal of Economics" (Reinikka & Svensson reference).
- p. 22 Local context para 2
Hefindahl index was 0.08
- p. 22 Local Context, para 2
the Hefindahl index was 0.08
- p. 23 Section 7
we again do not any find
- p. 21 footnote 40
whether the unicipality is a judiciary district
Judge's note: Each one seen on the page image or crop. The "unicipality" typo also listed here is judged under C031.
Second check: none (judge only)
Report a problem with this finding
Typo Section 4.3 reads "loses statistical significant once we account"; the noun 'significance' is intended.
- p. 12 Section 4.3
loses statistical significant once we account
Judge's note: The crop of p.10 (PDF p.12) shows the wording as quoted.
Second check: none (judge only)
Report a problem with this finding
Typo "share of services items audited" is inconsistent with the paper's usual 'service items'.
- p. 12 Section 4.3, first paragraph
share of services items audited
Judge's note: The crop shows the phrase as quoted. The meaning is unchanged.
Second check: none (judge only)
Report a problem with this finding
Typo "the potential bias associate with the effects of political experience" should read 'associated with'.
- p. 14 Section 5, after Eq. 3
potential bias associate with the effects of political experience
Judge's note: The crop of PDF p.14 shows the wording as quoted.
Second check: none (judge only)
Report a problem with this finding
Typo Subject–verb disagreement: "the cost of rent extraction are lower".
- p. 15 Section 6 intro
the cost of rent extraction are lower
Judge's note: The crop of the Section 6 intro shows the wording as quoted.
Second check: none (judge only)
Report a problem with this finding
Typo Footnote 25 reads "increases one chances of getting re-elected"; 'one's chances' is intended.
- p. 15 Footnote 25
increases one chances of getting re-elected
Judge's note: The crop of footnote 25 shows the wording as quoted.
Second check: none (judge only)
Report a problem with this finding
Typo Terms of office are dated inconsistently: "1989-1992, 1993-1996, and 1997-2000 administrations" (p.13 and p.17), but "(either from 1988-1992 or 1993-1996)", "during 1992-1996" and "previous administration (1996-2000)" on p.17 (fn 34 also says 1996-2000).
- p. 19 Experience section
either from 1988-1992 or 1993-1996
- p. 19 Experience section
during the previous administration (1996-2000)
- p. 15 Section 5.2
1989-1992, 1993-1996, and 1997-2000
- p. 19 Controlling for Experience, para 2
(either from 1988-1992 or 1993-1996)
Judge's note: Checked p15 (C028-c crop) and p19. The same administrations are written with election years in some places and with term years in others. Evident fix: use 1989-1992 and 1997-2000 throughout.
Second check: none (judge only)
Report a problem with this finding
Typo Footnote 26 reads "the estimate remain almost identical"; the singular subject needs 'remains'.
- p. 17 Footnote 26
the estimate remain almost identical
Judge's note: The crop of footnote 26 shows the wording as quoted.
Second check: none (judge only)
Report a problem with this finding
Typo Footnote 37 has a stray word: "likelihood of reelection on among mayors".
- p. 20 Footnote 37
likelihood of reelection on among mayors
Judge's note: The crop of footnote 37 shows the wording as quoted.
Second check: none (judge only)
Report a problem with this finding
Typo Footnote 40: "whether the unicipality is a judiciary district" should read "municipality".
- p. 21 Footnote 40
whether the unicipality is a judiciary district
- p. 21
unicipality is a judiciary district
Judge's note: Seen in the crop of p21 footnote 40.
Second check: none (judge only)
Report a problem with this finding
Typo "less corrupt that second-term mayors" should read "than".
- p. 22 Local Context, para 2
less corrupt that second-term mayors
Judge's note: Visible in the crop of p.22.
Second check: none (judge only)
Report a problem with this finding
Typo "we again do not any find" should read "do not find any", and "was audit during the 2004" should read "was audited".
- p. 23 Section 7, para 2
we again do not any find
- p. 23 Section 7, para 2
the municipality was audit during the 2004
Judge's note: Both are visible on p.23.
Second check: none (judge only)
Report a problem with this finding
Typo "is not only statistical significant" should read "statistically significant".
- p. 24 Conclusions
is not only statistical significant
- p. 24 Section 8, para 1
not only statistical significant, but economically
Judge's note: Visible on p.24.
Second check: none (judge only)
Report a problem with this finding
Typo "Programa de Atenção Bsica" is missing the letter 'á' (Básica).
- p. 26 Appendix A, first example
Programa de Aten¸c˜ao Bsica
Judge's note: The rendered page shows "Bsica", so this is not only a text-layer loss.
Second check: none (judge only)
Report a problem with this finding
Typo Appendix A (p.26): "approximatly nine kilometers" should read "approximately", and "it was late found" should read "it was later found".
- p. 26 Appendix A, third example
approximatly nine kilometers
- p. 26 Appendix A, third example
it was late found
- p. 26
to build approximatly nine kilometers of a road
Judge's note: Both misspellings are visible in the p.26 crops.
Second check: none (judge only)
Report a problem with this finding
Typo Two reference page ranges are garbled: "92 (5): 1594 605" (Duggan and Levitt, which also has "Isnt") and "91 (4): 10951102" (Fisman).
- p. 29 References, Duggan and Levitt
92 (5): 1594 605
- p. 29 References, Fisman 2002
91 (4): 10951102
- p. 29 References
Fisman, Raymond. 2002. “Estimating the Value of Political Connections.” American Economic Review 91 (4): 10951102.
Judge's note: The rendered page shows the missing dashes and digits. The claimed Fisman volume/year mismatch rests on outside knowledge and is not relied on here.
Second check: none (judge only)
Report a problem with this finding
Typo References (p.30): "Quaterly Journal of Economics" (Reinikka and Svensson 2004) should read "Quarterly", and "Cambridge, Massachussets" (Persson and Tabellini 2000) should read "Massachusetts".
- p. 30 References, Reinikka and Svensson
Quaterly Journal of Economics
- p. 30 References, Persson and Tabellini 2000
Cambridge, Massachussets
- p. 30
Quaterly Journal of Economics 119 (2): 679–705
Judge's note: Both misspellings are visible in the p.30 crops. The same reference list spells 'The Quarterly Journal of Economics' correctly in the Persson, Roland and Tabellini entry, and the title page has 'Massachusetts Avenue'.
Second check: none (judge only)
Report a problem with this finding
Typo The table notes read "share of resources found to corrupt" and "log GDP per capita per in 2002". These recur in Tables 4-9 and Appendix Table 1.
- p. 34 Table 4 notes
share of resources found to corrupt
- p. 34 Table 4 notes
log GDP per capita per in 2002
Judge's note: Visible in the notes on p.34 and repeated in the later tables.
Second check: none (judge only)
Report a problem with this finding
Typo Table 6, column 5 prints its standard error as "[0.0177]*" (four decimals), while every other entry in the table uses three; the text gives it as 0.017.
- p. 36 Table 6 column 5
[0.0177]*
Judge's note: Viewed page 36; four decimals confirmed. This is a cosmetic precision inconsistency only.
Second check: none (judge only)
Report a problem with this finding
Typo The notes to Tables 6 and 7 read "first-mayors that get re-elected"; the word "term" is missing (should be "first-term mayors").
- p. 36 Table 6 notes
first-mayors that get re-elected
- p. 37 Table 7 notes
first-mayors that get re-elected
Judge's note: Viewed the notes on pages 36 and 37; the phrase is printed as quoted in both.
Second check: none (judge only)
Report a problem with this finding
Typo Appendix Table 1 column headers read "(1) (3) (4) (5)" for its four columns; column (2) is skipped.
- p. 40 Appendix Table 1 header
(1) (3) (4) (5)
Judge's note: Verified on the p.40 image. The text cites only column 1 of Appendix Table 1, so no pointer is broken; this is a cosmetic numbering slip.
Second check: none (judge only)