Ethnicity, Insurgency, and Civil War
American Political Science Review, 2003 · paper · Political science
Examines civil wars since 1945, finding their prevalence reflects accumulated long conflicts rather than the Cold War's end, and that ethnic or religious diversity does not raise risk. Poverty, weak states, rough terrain and large populations favoring insurgency explain risk.
Checked: PDF from web.archive.org, fetched 5 Oct 2026. Page numbers are those of that PDF.
9 findings
Substance The abstract says "after controlling for per capita incomes and growth rates" and lists "poverty and slow growth" among the conditions favouring insurgency, but no growth-rate variable appears in any reported model.
- p. 1 Abstract
after controlling for per capita incomes and growth rates
- p. 1 Abstract
These include poverty and slow growth
- p. 37 Table 5
gdpt−1
Only the authors can say what is right here.
Judge's note: Table 5 (p.37 image) has war, gdp, pop, mountains, noncontiguous, instab, democ, the diversity measures and civlib, and no growth term. A case-insensitive search of the full text finds 'growth' only in the abstract, in the literature review (p.5) and in footnote 19, where WDI growth rates are used to extend the income series. No table or text result reports a growth control or a growth effect. Only the authors can say whether growth models were run but not reported.
Second check: another model agreed
Report a problem with this finding
Substance The introduction says "each additional $1,000 of per capita income is associated with 35% lower odds of civil war onset". With the Table 5 eq. 1 coefficient of -0.300, $1,000 more income gives about 26% lower odds; the 35% figure is the increase in odds for $1,000 less, as Section 5.2 correctly states.
- p. 4 Introduction, paragraph on empirical results
additional $1,000 of per capita income is associated with 35% lower odds of civil war onset
- p. 19 Section 5.2, first results paragraph
$1,000 less in per capita income is associated with 35% greater odds of civil war onset
- p. 37 Table 5, eq. 1, gdp_t-1
-0.300
- p. 4 Introduction, para beginning 'In the empirical analyses'
each additional $1,000 of per capita income is associated with 35% lower odds of civil war onset
Judge's note: Table 5 eq. 1 prints gdp_t-1 = -0.300. Section 5.2 (p.19) says "$1,000 less ... 35% greater odds": exp(0.3) is about 1.35, which matches. The reverse direction is 1/1.35, about 0.74, so roughly 26% lower odds, not 35%. The introduction keeps the 35% figure but reverses the direction.
Second check: another model agreed
Report a problem with this finding
Substance 'States with noncontiguous territory are here estimated to have more than three times the annual odds of civil war onset', but the Table 6 eq. 2 noncontig coefficient of 0.976 gives an odds ratio of about 2.65.
- p. 22 Section 5.3 item 1
to have more than three times the annual odds of civil war onset
- p. 38 Table 6, eq. (2), noncontig
0.976
- p. 22 Section 5.3 item 1
estimated to have more than three times the annual odds of civil war onset
- p. 22 Section 5.3.1
States with noncontiguous territory are here estimated to have more than three times the annual odds
Only the authors can say what is right here.
Judge's note: exp(0.976) = 2.65, and eq. 3 gives exp(0.985) = 2.68. Both are below 3. In a logit the odds ratio follows directly from the coefficient, so simulation cannot raise it above 3: even the lognormal mean, 2.65*exp(0.241^2/2), is about 2.73.
Second check: another model agreed
Report a problem with this finding
Reference or label 'As discussed above regarding hypothesis H7c, another possible proxy is foreign support to governments' and 'the mechanism behind H7c' cite H7c, which is the separated-territory hypothesis. Foreign support is H7d, cited with H5b on p.13.
- p. 23 Section 5.3 item 2, second paragraph
As discussed above regarding hypothesis H7c, another possible proxy is foreign support
- p. 13 Section 3, measurement discussion
from foreign countries (H5b, H7d)
- p. 23 Section 5.3 item 2
the mechanism behind H7c rather than the effects of French colonial administration
- p. 11 H7 list
(c) A territorial base that is separated from the territorial base of the center by water or distance
Judge's note: On p.11, H7(c) is the territorial base separated by water or distance and H7(d) is foreign governments or diasporas. Page 13 links foreign support, including the support to governments by the USSR, the U.S. and France, to (H5b, H7d), and uses H7c for the noncontiguity dummy. Both H7c citations on p.23 point to the wrong hypothesis.
Second check: another model agreed
Report a problem with this finding
Reference or label 'Equation 3 in Table 7 includes two additional geographic variables' should say Table 6. Table 7 is the French vs non-French colonies cross-tab and has no equations.
- p. 22 Section 5.3 item 1, anticolonial wars
Equation 3 in Table 7 includes two additional geographic variables
- p. 38 Table 7
Civil war onsets in French vs. non-French former colonies in sub-Saharan Africa
- p. 22 Section 5.3 item 1
Equation 3 in Table 7 includes two additional geographic
- p. 38 Table 6 eq. 3
log(totland)
Judge's note: On p.38, Table 6 eq. 3 has the mountains row and log(totland). Table 7 is a descriptive onsets table. The surrounding text on p.22 discusses Table 6 equations 1, 2 and 4.
Second check: another model agreed
Report a problem with this finding
Reference or label Table 6, eq. (3): the coefficient row "0.225 (0.083)**" between noncontig and log(totland) has no variable label.
- p. 38 Table 6, eq. (3), row between noncontig and log(totland)
0.225
- p. 22 Section 5.3.1
the mountains measure and total land area, both logged
- p. 38 Table 6, between noncontig and log(totland) rows
0.225 (0.083)**
Judge's note: The Table 6 image (p.38) shows a blank label cell beside 0.225 / (0.083)** in the eq. 3 column. The text on p.22 says equation 3 adds 'the mountains measure and total land area, both logged', so the missing label is presumably the logged mountains variable. (That same p.22 passage calls this 'Table 7', but that is outside this cluster.)
Second check: another model agreed
Report a problem with this finding
Reference or label p.22 (PDF 23): French sub-Saharan effect attributed to "the mechanism behind H7c", but H7c is the noncontiguous-territory hypothesis; foreign military support is discussed under H7d.
- p. 23 Section 5.3 item 2, end of French paragraph
the mechanism behind H7c rather than the effects of French colonial administration or law
Judge's note: Page image p.23 checked. Hypothesis list (PDF p.11): H7(c) is a territorial base separated from the centre by water or distance; H7(d) is foreign governments or diasporas supplying weapons, money or training. On PDF p.13 the authors measure noncontiguity for H7c and discuss foreign support to governments (Soviet, U.S., French policy) under 'H5b, H7d'. The French paragraph argues about French military support to governments, so H7d fits. The same mislabel occurs earlier on p.23 ('As discussed above regarding hypothesis H7c, another possible proxy is foreign support to governments').
Second check: another model agreed
Report a problem with this finding
Reference or label Young-males check is added "in equation 1, Table 1", but Table 1 is 'Civil war onsets by decade' and has no equations; the main specification is equation 1 of Table 5.
- p. 25 Section 5.3(d) Young Males
in equation 1, Table 1, it has the expected positive sign
- p. 34 Table 1
Table 1: Civil war onsets by decade
- p. 25 5.3(d)
in equation 1, Table 1, it has
- p. 25 Sec 5.3 (d) Young Males
in equation 1, Table 1
Judge's note: Page 25 image shows the text 'in equation 1, Table 1'. Page 34 image shows Table 1 as a descriptive table by decade. Equation 1 is in Table 5 (p.37), and other robustness checks refer to it as 'equation 1, Table 5' (e.g. p.23).
Second check: another model agreed
Report a problem with this finding
Reference or label "Equation 3 in Table 8 shows that at least for the period 1976 to 1994...": hrabuse_t-1 (0.680, SE 0.207) appears only in column (2) of Table 8 (N=2324); column (3) is the <5000-death recode, which p.28 also calls 'Equation 3 in Table 8'.
- p. 27 Section 5.3(g) Human Rights Abuses
Equation 3 in Table 8 shows that at least for the period 1976 to 1994
- p. 39 Table 8, column (2), hrabuse_t-1
0.680
- p. 39 Table 8, eq (2)
hrabuset−1 0.680
- p. 28 Section 5.3 item 4
Equation 3 in Table 8 shows the results for our main specification when the 35 conflicts
Judge's note: Table 8 image (p.39): hrabuse_t-1 has an entry only in column (2), with N=2324, which fits the 1976-94 window. Column (3) has N=6321 and no hrabuse. On p.28 the text cites 'Equation 3 in Table 8' for the recode of the 35 conflicts under 5000 deaths. The human-rights pointer should be equation 2.
Second check: another model agreed
Report a problem with this finding
39 minor (typos and the like; most checked by the judge only)
Reference or label The text says "Figure 1 shows the number of countries with ongoing civil wars by year", but the figure's left axis and legend read "# ongoing civ wars".
- p. 2 Introduction, first sentence
Figure 1 shows the number of countries with ongoing civil wars by year
- p. 41 Figure 1, left axis and legend
# ongoing civ wars
Only the authors can say what is right here.
Judge's note: The p.2 crop and the Figure 1 page image (p.41) show the mismatch. The left series is labelled as a count of wars, not of countries. These differ when a country has several concurrent wars. The second series (% countries with an ongoing war) matches its description in the text.
Second check: another model agreed
Report a problem with this finding
Reference or label Footnote 50 cites "Wallenstein and Sollenberg", which matches no reference entry; the armed-conflict data are Gleditsch, Strand, Eriksson, Sollenberg and Wallensteen (2001), with Wallensteen spelled 'ee'.
- p. 28 Footnote 50
We coded only those cases from Wallenstein and Sollenberg
- p. 31 Reference list, Gleditsch et al.
Peter Wallensteen
- p. 31 References
Margareta Sollenberg and Peter Wallensteen. 2001.
- p. 28 fn 50
Wallenstein and Sollenberg
Judge's note: Page 28 footnote 50 shows 'Wallenstein and Sollenberg'. The reference list (p.31) has only the Gleditsch et al. 2001 entry, with 'Peter Wallensteen'. The main text cites the data as Gleditsch et al. (2001). The name is misspelled and the citation form is inconsistent.
Second check: another model agreed
Report a problem with this finding
Reference or label Fearon & Laitin 1996 reference reads "American Political Science Review 4:715–35"; in the list's volume:pages format, 4 is not the APSR volume (other entries give 89 for 1995 and 94 for 2000, which implies 90 for 1996).
- p. 30 References
American Political Science Review 4:715–35
Judge's note: Reference page 30 image checked. Other APSR entries in the list: Licklider 1995 is 89:681–90, Doyle & Sambanis 2000 is 94:778–801, Hegre et al. 2001 is 95. The '4' appears to be the issue number printed where the volume belongs.
Second check: another model agreed
Report a problem with this finding
Wording Footnote 3 says "conjoined with three other variables" and then names four: exogenous social change, mobilizing structures, framing processes and repertoires of contention.
- p. 3 Footnote 3
is conjoined with three other variables to account for contentious events
- p. 3 Footnote 3
is conjoined with three other variables to account for contentious events: exogenous social change, mobilizing structures
Judge's note: The image shows the list. The semicolons create three groups, but the first group contains two named items ('exogenous social change, mobilizing structures, that is, ...'). So the count of 'three' is ambiguous rather than plainly wrong: exogenous social change may be meant as background.
Second check: another model agreed
Report a problem with this finding
Wording The introduction states as an empirical finding that wars in former French sub-Saharan colonies "have killed far fewer when they occurred", but no death or magnitude analysis is reported.
- p. 4 Introduction, summary of findings
these have killed far fewer when they occurred
Only the authors can say what is right here.
Judge's note: The full-text search finds French-colony evidence only on onset rates (Table 7 and the related text). Footnote 49 says the death data are 'the subject of a paper in progress'. The text near line 499 only states, as a hypothesis, that such states should have fewer deaths.
Second check: another model agreed
Report a problem with this finding
Wording 'Indonesia is the most civil war-prone country in this sample with seven onsets, followed by India with five.' However, the case list gives six onsets to FRANCE, and the 126-conflict sample includes the colonial wars.
- p. 16 Section 4 summary statistics
Indonesia is the most civil war-prone country in this sample with seven onsets, followed by India with five.
- p. 22 Section 5.3 item 1
which had 6 of the 13 imperial wars by our codings
- p. 40 Case list, FRANCE entries
FRANCE
Only the authors can say what is right here.
Judge's note: The p.40 case list has Indonesia 7 (1950, 1953, 1958, 1965, 1975, 1991, 1997), India 5 (1947, 1948, 1952, 1982, 1989) and FRANCE 6 (1945, 1947, 1952, 1953, 1954, 1955). On p.16 'this sample' is the 126 conflicts, 13 of them anticolonial. P.22 says the French empire 'had 6 of the 13 imperial wars'. India is therefore second only if empires are not counted as 'countries', and the paper does not say this. The authors should clarify.
Second check: another model agreed
Report a problem with this finding
Wording The text says onsets were 'particularly common (relative to country years) in the late 1940s and in the 1990s', but in Table 1 the 1950s rate (2.30) is higher than the 1990s rate (2.22).
- p. 16 Section 4 summary statistics
particularly common (relative to country years) in the late 1940s and in the 1990s
- p. 34 Table 1, Onsets/100 country-years
2.30
- p. 16 Section 4
civil war onsets have been particularly common (relative to country years) in the late 1940s and in the 1990s
- p. 34 Table 1
Onsets/100 country-years 4.01 2.30 1.68 1.79 1.09 2.22
Judge's note: Table 1 (p.34) gives onsets per 100 country-years of 4.01 (1945-9), 2.30 (50s), 1.68 (60s), 1.79 (70s), 1.09 (80s), 2.22 (90s) and 1.91 overall. Both the 1950s and 1990s are above the overall rate, so the claim is loosely defensible, but naming the 1990s rather than the 1950s does not match the table's ordering.
Second check: another model agreed
Report a problem with this finding
Wording The text says 'The estimated coefficient changes from -.29 to -.24 when a dummy for "the West" is included', but Table 5 eq. 1, the specification discussed in this paragraph, gives gdp_t-1 = -0.300.
- p. 19 Section 5.2, discussion of the West dummy
The estimated coefficient changes from -.29 to -.24
- p. 37 Table 5, eq. (1), gdp_t-1
-0.300
- p. 19 Section 5.2
The estimated coefficient changes from -.29 to -.24 when a dummy for “the West” is included
- p. 19 Sec 5.2
changes from -.29 to -.24 when a dummy
Only the authors can say what is right here.
Judge's note: In Table 5 (p.37) gdp_t-1 is -0.300 in eq. 1, -0.288 in eq. 2 and -0.291 in eq. 3. The paragraph on p.19 discusses equation 1, and its '35% greater odds' figure matches -0.300. The -.29 baseline matches eq. 2 or 3, or perhaps a different sample. Only the authors can say which.
Second check: another model agreed
Report a problem with this finding
Wording 'a 43% decrease in risk going from a country with 30% versus a 75% ethnic plurality', but the Table 5 eq. 3 ethplural coefficient (-0.778) gives about a 30% decrease in odds over that range.
- p. 20 Section 5.2, ethplural paragraph
a 43% decrease in risk going from a country with 30% versus a 75% ethnic plurality
- p. 37 Table 5, eq. (3), ethplural
-0.778
Only the authors can say what is right here.
Judge's note: With ethplural as a proportion, the change in log-odds is -0.778*0.45 = -0.350. exp(-0.350) = 0.705, a decrease of about 30%. The reverse direction, exp(0.350) = 1.42, is a 42% increase, so 43% looks like the change measured from the 75% baseline. Clarify simulation at low base rates cannot account for the gap. The authors should check the figure.
Second check: another model agreed
Report a problem with this finding
Wording 'doubling country population is associated with only a 20% increase in the annual odds', but the Table 5 eq. 1 log(pop) coefficient of 0.218 gives about 16%.
- p. 21 Section 5.2, population paragraph
doubling country population is associated with only a 20% increase
- p. 37 Table 5, eq. (1), log(pop_t-1)
0.218
- p. 21 Section 5.2, third set of points
doubling country population is associated with only a 20% increase in the annual odds
Only the authors can say what is right here.
Judge's note: For a doubling, 2^0.218 = exp(0.218*0.693) = exp(0.151) = 1.163, a 16% increase in odds. Eq. 2 (0.271) would give about 21%. The text refers to the main results (eq. 1). This could be loose rounding, so it is a query.
Second check: another model agreed
Report a problem with this finding
Wording Text says Poe, Tate and Keith (1999) developed a scale using reports on "142 countries between 1976 and 1994", but the reference title says "Covering the Years 1976-1993".
- p. 27 Section 5.3(g)
reports on 142 countries between 1976 and 1994
- p. 32 Reference list, Poe et al. 1999
Covering the Years 1976-1993
- p. 27 Sec 5.3 (g)
142 countries between 1976 and 1994
Only the authors can say what is right here.
Judge's note: Page 27 image shows the text '1976 and 1994' and later 'for the period 1976 to 1994'. Reference list p.32 gives the title '...Covering the Years 1976-1993'. The authors may have used an extended release of the data, but the sentence attributes 1976-1994 coverage to the 1999 study. Only the authors can say which is right.
Second check: another model agreed
Report a problem with this finding
Wording "With one exception, the mountains variable, the results are remarkably consistent", yet in Table 8 eq. 4 noncontig is 0.440 (0.293) without a star, versus 0.670 (0.271)* in Table 5 eq. 1.
- p. 28 Section 5.3 item 4, last paragraph
With one exception, the mountains variable, the results are remarkably consistent.
- p. 39 Table 8, eq. (4), noncontig
0.440
- p. 37 Table 5, eq (1), noncontiguous
0.670
Only the authors can say what is right here.
Judge's note: Table 8 (p.39), column (4): log(%mountain) 0.156 (0.088) and noncontig 0.440 (0.293), both unstarred. Noncontig is starred in Table 5 eq. 1 and in Table 8 eqs. 1-3. So noncontiguity also loses significance, and the text naming a single exception understates the change.
Second check: another model agreed
Report a problem with this finding
Typo Footnote 1 reads "Cf. Gurr (2000a), notes the decline in ethnic war", which is ungrammatical (a stray 'Cf.' or a stray comma).
- p. 2 fn 1
Cf. Gurr (2000a), notes the decline
Judge's note: The page image confirms the wording. The fix is, for example, 'Gurr (2000a) notes...'.
Second check: none (judge only)
Report a problem with this finding
Typo The introduction reads "(e.g., the Soviet Union in Eastern Europe the former Soviet republics, ...)": a conjunction or comma is missing.
- p. 4 Introduction
the Soviet Union in Eastern Europe the former Soviet republics
Judge's note: The page image confirms the text as quoted.
Second check: none (judge only)
Report a problem with this finding
Typo Two missing words: "the sources nationalist mobilization" (p.5) should have 'of', and "for the countries that category" (p.17) should have 'in'.
- p. 5 Section 2, first paragraph
the sources nationalist mobilization
- p. 17 Section 5.1, Table 2 description
for the countries that category
Judge's note: Both page images confirm the wording.
Second check: none (judge only)
Report a problem with this finding
Typo Footnote 14 has "rebels leaders" (should be 'rebel leaders'), and p.5 has "national distri-butions systems" (should be 'distribution systems').
- p. 10 Footnote 14
rebels leaders
- p. 5 Introduction, Collier-Hoeffler paragraph
national distri-
- p. 10 fn 14
rebels leaders have either spent time
Judge's note: Both page images confirm the wording.
Second check: none (judge only)
Report a problem with this finding
Typo Footnote 7 has "Encyclopedia Brittanica", a misspelling of Britannica.
- p. 7 Footnote 7
Encyclopedia Brittanica
Judge's note: The page image confirms the spelling.
Second check: none (judge only)
Report a problem with this finding
Typo The measures list runs "(1) ...; (2) ...; and (3) ... based on Grimes and Grimes (1996); a measure of religious fractionalization ...": the fourth item has no number and follows the 'and' of item (3).
- p. 7 Sec 2
based on Grimes and Grimes (1996); a measure of religious frac-
Judge's note: The page image confirms the punctuation. This is a minor style and punctuation point.
Second check: none (judge only)
Report a problem with this finding
Typo "mechanisms associating lower per capita with civil war onset": the word 'income' is missing.
- p. 11 Section 3, after H6
mechanisms associating lower per capita with civil war onset
Judge's note: The page image confirms the wording.
Second check: none (judge only)
Report a problem with this finding
Typo Footnote 19 has "evidence on energy per consumption" where 'energy consumption per capita' is meant.
- p. 12 Footnote 19
evidence on energy per consumption
Judge's note: The page image confirms the wording. The same footnote uses 'energy consumption per capita' correctly a line earlier.
Second check: none (judge only)
Report a problem with this finding
Typo Footnote 21 has "the land area containing the capitol city"; the word should be 'capital'.
- p. 13 Footnote 21
capitol city
- p. 13 fn 21
containing the capitol city
Judge's note: The page image confirms the spelling.
Second check: none (judge only)
Report a problem with this finding
Typo Section 4 names "the Institute for International and Strategic Studies (IISS 2000)"; the organisation is the International Institute for Strategic Studies.
- p. 14 Sec 4
Institute for International and Strategic Studies (IISS 2000)
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: IISS stands for the International Institute for Strategic Studies. The page image shows the quoted text, and the reference-list entry (IISS 2000) gives the same form as publisher: 'London: Institute for International and Strategic...'.
Second check: none (judge only)
Report a problem with this finding
Typo Misspellings 'occuring' (p.15: 'occuring within the colonial empire', '5 colonial wars occuring'; p.23: 'occuring in almost 45%') and 'occured' (fn 27, 'if any occured'); should be 'occurring'/'occurred'.
- p. 15 Section 4, colonial wars paragraph
We count these as occuring within the colonial empire
- p. 16 Footnote 27
occured
- p. 23 Section 5.3 item 2
occuring in almost 45% of the country-years
- p. 15 Sec. 4
occuring within the colonial empire
Judge's note: Saw all the misspellings on the page images: p.15 (twice), p.23 and footnote 27 on p.16.
Second check: none (judge only)
Report a problem with this finding
Typo Stray article in 'use the country-level data to a produce such estimates'.
- p. 15 Section 4, practical considerations paragraph
to a produce such estimates
Judge's note: The page 15 image shows 'to a produce such estimates'. The 'a' should not be there.
Second check: none (judge only)
Report a problem with this finding
Typo 'treciles' (p.16 'equal size groups (treciles)'; fn 47 'broken into treciles or quartiles') should be 'terciles'.
- p. 16 Section 5.1
equal size groups (treciles)
- p. 26 Footnote 47
broken into treciles or quartiles
Judge's note: Seen on the p.16 and p.26 images.
Second check: none (judge only)
Report a problem with this finding
Typo Fn 28 'slightly lower mean number of civil war onsets then countries that aren't' ('then' for 'than'), and p.17 'for the countries that category' (missing 'in').
- p. 17 Footnote 28
mean number of civil war onsets then countries that aren’t
- p. 17 Sec 5.1 para 2
for the countries that category
- p. 17 fn 28
slightly lower mean number of civil war onsets then countries
Judge's note: Both are visible on the page 17 images.
Second check: none (judge only)
Report a problem with this finding
Typo 'A country with noncontinguous territory' should be 'noncontiguous'.
- p. 18 Section 5.1, Table 3 eq. 3 discussion
noncontinguous territory
Judge's note: Visible on the p.18 image. The same paragraph spells the word correctly earlier.
Second check: none (judge only)
Report a problem with this finding
Typo Fn 34: 'spline or dummy coefficents jointly insignificant' should be 'coefficients'.
- p. 19 Footnote 34
spline or dummy coefficents jointly insignificant
Judge's note: Visible on the p.19 image.
Second check: none (judge only)
Report a problem with this finding
Typo Religious fractionalization is labelled "relfrac" in Table 5 but "religfrac" in the text (Section 5.2, "(ethfrac and religfrac)") and in the other tables.
- p. 37 Table 5 row label
relfrac
- p. 19 Section 5.2
(ethfrac and religfrac)
- p. 19 Section 5.2
ethnic and religious fractionalization (ethfrac and religfrac)
Judge's note: The Table 5 image (p.37) shows the row label 'relfrac'. A text search finds 'religfrac' at text-layer lines 750 (p.19), 1451 and 1867 (other tables), and 'relfrac' only in Table 5. The variable naming is inconsistent; the meaning is clear.
Second check: none (judge only)
Report a problem with this finding
Typo Duplicated word: "Mali (1989, a case that that really belongs with those of the 1990s)".
- p. 23 Section 5.3 item 2, French paragraph
a case that that really belongs
Judge's note: The page image p.23 shows 'that that'.
Second check: none (judge only)
Report a problem with this finding
Typo "Mitterand’s government" misspells the French president's name (Mitterrand).
- p. 23 Section 5.3 item 2
Mitterand’s government
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: the president's name is spelled Mitterrand. The page image p.23 shows 'Mitterand’s'.
Second check: none (judge only)
Report a problem with this finding
Typo "we turn Poe and Tate’s question around" names only two authors, while the cited study in the same paragraph is Poe, Tate and Keith (1999).
- p. 27 Section 5.3(g)
we turn Poe and Tate’s question around
- p. 27 5.3(g)
turn Poe and Tate’s question around
Judge's note: Page 27 image checked. This may be informal shorthand, but it is inconsistent with the citation and the reference list, which have no Poe and Tate entry.
Second check: none (judge only)
Report a problem with this finding
Typo Footnote 50 ends "(1000 killed, total." and the parenthesis is never closed.
- p. 28 Footnote 50
(1000 killed, total.
- p. 28 footnote 50
Wallenstein and Sollenberg that they judged to have reached the “intermediate” level of violence (1000 killed, total.
Judge's note: The page 28 image shows the footnote ends without a closing parenthesis.
Second check: none (judge only)
Report a problem with this finding
Typo Missing article: "to provide check on how dependent our results are" (should read 'provide a check').
- p. 28 Sec 5.3 item 4
to provide check on how dependent
Judge's note: Page 28 image checked.
Second check: none (judge only)
Report a problem with this finding
Typo Huntington 1996 reference gives the publisher as "New York: Simon and Shuster."
- p. 31
New York: Simon and Shuster.
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: the publisher is spelled Simon and Schuster. The page 31 image shows 'Shuster'.
Second check: none (judge only)
Report a problem with this finding
Typo Reference list misspellings: Weede (1981) title printed as "Income inquality, Average Income, and Domestic Violence" and Huntington (1996) publisher printed as "Simon and Shuster".
- p. 33 Reference list, Weede 1981
Income inquality
- p. 31 Reference list, Huntington 1996
Simon and Shuster
- p. 33
“Income inquality, Average Income, and Domestic Violence.”
Judge's note: I checked the page images. On p.33 the Weede entry reads "Income inquality", and on p.31 the Huntington 1996 entry reads "New York: Simon and Shuster." 'inquality' is plainly a typo for 'inequality'. The publisher's spelling 'Schuster' rests on outside knowledge: the publisher is Simon & Schuster.
Second check: none (judge only)
Report a problem with this finding
Typo Table 2, Medium fractionalization/High income cell is printed "0.18" (two decimals); every other cell has three decimals (Table 4 also prints ".818" with no leading zero).
- p. 34 Table 2, Medium ethnic fractionalization, High income
0.18
Judge's note: In the Table 2 image the Medium/High cell reads 0.18 with n=22, while the other cells read e.g. 0.765 and 0.609. The value is consistent with 4/22, so the issue is formatting only. The Table 4 image shows ".818" next to 0.636 and 0.286.
Second check: none (judge only)
Report a problem with this finding
Typo Case list: the entry "RUSSIA 1999-" has no description, whereas every other entry names the conflict or rebels.
- p. 40 Case list, second column, RUSSIA 1999-
1999-
Only the authors can say what is right here.
Judge's note: The p.40 image shows RUSSIA 1999- with an empty description column; the entry above reads RUSSIA 1994-96 Chechnya. The missing text is presumably 'Chechnya', but only the authors can confirm.
Second check: none (judge only)
Report a problem with this finding
Typo Figure 1 legend is truncated: "% countries with ongoing civ wa" (the right axis title reads "civ war").
- p. 41 Figure 1 legend
% countries with ongoing civ wa
Judge's note: The p.41 figure image shows the legend text cut off at 'civ wa', while the right axis label reads '% countries with ongoing civ war'. This is cosmetic.
Second check: none (judge only)