NGO Monitors 2007 Report on HRW: Bias and Double Standards Continue
“The point is not to equate, excuse or balance abuses but to demonstrate that reporting is based on human rights principles rather than partisan considerations.” HRW’s Position Paper on the Durban Preparatory Committee (April 21, 2008):
The following analysis demonstrates that HRW’s own activities related to Israel continue to fall short of this basic standard of universality
- Analysis of Human Rights Watch‘s use of the rhetoric of international law and other terminology shows continued double standards and misleading or false claims.
- HRW accuses Israel of “collective punishment” of Palestinians in a way that is inconsistent with both international law and past and present usage of the term by HRW itself.
- HRW’s focus on Israel in 2007 dropped to 2005 levels after 2006 marked a return to the extreme bias of the 2000-2004 period. The change in 2007 allowed more resources to be focused on countries committing major human rights violations.
- However, disproportionate emphasis on Israel continued, with major reports covering 400 pages in 2007, using the same methodologies as in 2006 that lack credibility.
- This contrasts with the limited attention on human rights violations in Libya, Syria, and other countries in the region.
- Israel was the focus of more multimedia items (audio, video, graphics) than any other country in the region.
- Reports on Israel continue to be based on unverifiable evidence provided by “eyewitnesses,” selected journalists, and other inappropriate sources. In some cases hard evidence has shown this testimony to be blatantly untrue.
- HRW mentioned one or more of the kidnapped Israeli soldiers in a total of only 6 publications. Of these, only two refer to them by name; the other references are in passing.
This report includes quantitative analyses of publications from HRW’s Middle East and North Africa section, using a weighted scale methodology consistent with NGO Monitor’s previous analyses, and an assessment of the use of language in the HRW publications.
Agenda Analysis: Focus on Israel returns to 2005 levels
NGO Monitor’s analysis of Human Rights Watch’s (HRW) output devoted to each country in the Middle East and North Africa region shows that the focus on and condemnations of Israel in 2007 returned to 2005 levels (see 2005 report). This follows a sharp increase in 2006 (see 2006 report). However, Israel is still the focus of nine percent of HRW’s output on the region– significantly greater than Libya, Syria, Hamas, Hezbollah and other chronic human rights violators. The evaluation is based on a point system which accounts for both the volume of publications and their significance (a report clearly being much more significant and requiring more resources than a press release –see the methodological section of the appendix).
Figure 1 provides a comparison of the relative output (weighted as described in the methodology section in appendix) of countries in the Middle East over the past three years. Figure 1 demonstrates the degree that an increased focus on Israel correlates with a decreased focus on other countries in the region, such as Iran and Egypt.1 In absolute terms, HRW’s total output on the Middle East and North Africa has increased steadily over the period 2005-2007, the results presented here are in percentage terms to enable comparison.
Figure 1 Percentage of weighted sore 2005-2007
Figure 2 displays the results of the detailed comparative analysis of the activities of HRW’s Middle East and North Africa section in 2007. In 2007, 10 percent (121 points) of HRW publication effort in the Middle East and North Africa section focused on Israel, which is significantly lower than in 2006, and slightly higher in percentage terms than the 8 percent in 2005.2
Figure 2 Figure 2 Weighted score by country in 2007 in the Middle East / North Africa region. [Note that Iraq and Lebanon are included under the same criteria as pervious years- see appendix for further explanation]
Use of Multimedia items
In 2007 there were more multimedia items on Israel (6 audio, video and graphics items) than any other country (4 for Hamas, 4 for Iran and 3 for Egypt – other countries received 2 or fewer). Most countries, including Israel, featured in one audio item each as part of the 2007 World Report. Israel’s multimedia items also included audio commentary by Marc Garlasco on cluster munitions, pictures and audio on medical evacuations3, pictures and audio on students trapped in Gaza, and a video on cluster munitions. Multi-media material requires a higher level of investment by Human Rights Watch and can in many cases result in a larger impact than is the case in other types of HRW activities. When this increased impact is disproportionately focused on Israel, the issue of double standards becomes relevant.
Use of Reports
HRW puts a great deal of effort and resources into its major reports, often including a press conference and other public relations activities to accompany the publication. In many cases, the focus of a report is chosen weeks or months in advance of publication, and reflects a strategic decision regarding HRW’s priorities.
The disproportionate emphasis on Israel continued in 2007. The number of reports featuring Israel in 2007 (2)4 was equal to the number on Hezbollah, Egypt, and Iraq. Only Lebanon featured in more reports (3).5 Significantly, HRW officials chose not to issue any reports on Iran, Libya and Morocco in 2007. (In contrast, HRW’s press releases focusing on human rights violations in Iran constituted 17 percent of total press releases in the region, with Egypt targeted in a further 17 percent. Israel received 9% of the press releases, a proportionate amount given it received 10% of the over all focus).
Priorities within the Arab Israeli conflict
HRW’s priorities are also reflected in an analysis of the relative coverage of Israel vis-à-vis the Palestinian Authority, Hamas, and Hezbollah. The analysis shows minimal attention to internal Palestinian fighting and an over emphasis on Israeli actions in both the conflict with Hezbollah and the Israeli-Palestinian conflict. Figure 3 demonstrates almost the same distribution between publications focusing on Israel and the total focused on all groups in conflict with Israel (i.e. Palestinian groups and Hezbollah). This was despite the fact that nearly half of all Palestinian fatalities were caused by intra-Palestinian violence during 2007. 6
When Hamas seized control of Gaza by violent coup in June 2007, there were many clear violations of human rights.7 Human Rights Watch itself documented a few briefly. One example was the execution by Hamas of a cook for Mahmoud Abbas’s Presidential Guard (press release “Gaza: Armed Palestinian Groups Commit Grave Crimes” – June 13). The cook was executed by “throwing him to his death, with his hands and legs tied, from a 15-story apartment building in Gaza City”. Over the year, only 2 published items from HRW focused primarily on internal Palestinian violence, a third published item covered both internal Palestinian violence and attacks on Israelis. This is in comparison with seven documents focused on Palestinian attacks on Israelis.
Figure 3 Israel, Palestinians and Hezbollah
Further comparisons between Israel and those in conflict with her are available in the appendix, including a breakdown of the collective term Palestinians.8
As noted, major reports require far greater resources, and are designed to have a greater impact, including the use of press conferences and public relations campaigns. In 2007, HRW published two reports which ostensibly analysed alleged human rights violations by Israel and by its opponent (Palestinian terrorists or Hezbollah). However, in “Why They Died: Civilian Casualties in Lebanon during the 2006 War,” 122 pages describe alleged Israeli abuses while 23 pages describe alleged abuses by Hezbollah. In “Indiscriminate Fire: Palestinian Rocket Attacks on Israel and Israeli Artillery Shelling in the Gaza Strip,” 72 pages describe alleged Israeli abuses, while 19 pages describe alleged Palestinian abuses. (The page counts exclude pages on general background information, methodologies and non substantive appendices)
Language Analysis – Double Standards Continue
As explained in previous NGO Monitor reports, the language and terminology used by NGOs in their publications on human rights issues is an additional indicator of the group’s relative emphasis and agenda. Terms such as “war crimes”, “collective punishment”, “torture”, “violations of international law”, etc. are condemnatory, and their inconsistent or unjustified use is at odds with the core universality of human rights norms.
This analysis (results in Figure 4) was conducted using the same methodology as in NGO Monitor’s 2006 report.9 In 2006, Israel, Libya, Iraq and UAE all averaged around thirty occurrences for various terms or phrases. The analysis of HRW’s rhetoric and vocabulary in 2007, compared to previous years, shows that the use of condemnatory language has been reduced across the board in the Middle East and North Africa region. The words provided in the key in Figure 4 (and very close approximations) were counted when they referred directly to the country indicated, (explanations, including the many cases where occurrences were discounted, are provided in the appendix).
Figure 4: Language usage10
Misleading “balance” in HRW reporting
The relative increase and apparent “balance” in condemnation of Israel and its opponents is an improvement in reporting on the conflict, but still falls short of a even handed approach. This is because HRW’s reporting continues to distort the context and suggests a false moral equivalence between aggressor and defender. This is illustrated in an article by HRW’s Middle East and North Africa division director, Sarah Leah Whitson, published in Al-Akhbar on September 1, 2007. Whitson states:
“That Israel violated the laws of war does not justify the Islamic Resistance’s failure to abide by such laws. … that Israel carried out indiscriminate strikes on populated areas in southern Lebanon in no way justifies the Islamic Resistance’s retaliatory strikes on civilian areas in Israel.”
Although statistically such references count once against each party and are thus “balanced”, the context is clearly portraying Israel as the aggressor.
Similarly, a HRW press release (July 12, 2007) headlined “Israel/Lebanon: A Year Later, No Justice for War Violations,” states:
“During the conflict, Human Rights Watch documented serious violations of humanitarian law by both the IDF and Hezbollah… the Israeli army… thus systematically failed to distinguish between civilians and combatants, in violation of humanitarian law… [commanders] may be guilty of war crimes.”
After three paragraphs on Israel, the authors of the report turn to Hezbollah, providing some facts without the emotional comment. This section ends by stating,
“while Hezbollah appeared to target some of its rockets at military objectives…such attacks were at best indiscriminate attacks on civilian areas, at worst direct attacks against civilians…[Commanders] may also have committed war crimes and should be investigated.” (Emphasis added by NGO Monitor).
In the case of Israel the phrasing conveys culpability; in the case of Hezbollah, it only suggests a need to investigate.
Similarly, this text uses the phrase “violation of humanitarian law” to refer to Israel’s alleged failure to distinguish between civilians and combatants. In the case of Hezbollah, the same is suggested to be a “best case” analysis, with the worse case being the targeting of civilians. While Hezbollah’s actions are clearly a “serious violation” of international law, HRW officials refer to this as a hypothetical “worst case”. The use of such language shows a political bias against Israel in style and commentary.
This politically based distinction is repeated in HRW’s 249 page report that returns to the question of Israeli actions in the 2006 war with Hezbollah, headlined “Why They Died: Civilian Casualties in Lebanon during the 2006 War”, (September 2007). The authors describe events and simultaneously claim numerous violations of international law by Israel, eg “Israel often attacked targets that, under the laws of war, could not be considered military objectives subject to attack.”
HRW lacks both the expertise and the intelligence information to make this judgement, and it places the burden of proof on Israel, demanding additional information not known to HRW. In contrast, the section on Hezbollah’s use of human shields uses the language of uncertainty: “…we found a handful of instances but nothing to suggest a widespread practice” (page 6).
Such bias and double standards is also evident in a press release “Israel: Threatened Sanctions on Gaza Violate Laws of War” (September 20, 2007) written by Sarah Leah Whitson. According to Whitson, “Palestinian armed groups are clearly violating the laws of war by firing rockets deliberately or indiscriminately at Israeli civilians… Israel has the responsibility to protect its citizens, but not by collectively punishing the people of Gaza, which seriously violates the laws of war.” The term “collective punishment” (which is not applicable to this situation under international law – see detailed discussion below) is used to condemn Israeli responses to aggression, but not with respect to the deadly rocket barrages from Gaza. And as in many other instances, Whitson avoids holding Hamas responsible by name.
Whitson’s press release goes on to argue that “Israel remains an occupying power in the Gaza Strip… because it continues to exercise effective day-to-day control over most aspects of life in Gaza.” The dimensions listed are airspace, sea space and borders. In contrast, the accepted definition under international law of “day-to-day control over most aspects of life” as a condition of occupation, refers to the running of local government and facilities.11 Whitson also ignores Egypt’s role in Gaza.
Promoting the false claim of “continuing occupation”
Building on the false claim of “continuing occupation”, HRW holds Israel to a unique standard. Listing the obligations of “an occupying power” (i.e. supply of food and medical supplies to the civilian population), HRW then extends these obligations to “power and fuel” without legal basis. The reduction of fuel supplies between countries is not unique to Israel (e.g. Russia / Ukraine in 2006),12 and is not prohibited under current international conventions.13 HRW did not condemn Russia for this action nor refer to its control over fuel as creating a situation of “occupation”.
Areas of improvement
Finally, from multi-region documents examined in this analysis, there is evidence that these problems notwithstanding, HRW has heeded some of NGO Monitor’s previous analysis on language use. In one document “More Business Than Usual: The Work Which Awaits the Human Rights Council” (12 March) HRW officials outline their suggested agenda and priorities for the UN Human Rights Council. The document refers to “serious human rights abuses throughout Afghanistan”, “serious abuses, including extrajudicial killings, torture, beatings, arbitrary arrests, and rapes” in Chad, “serious human rights abuses, particularly those abuses committed by security forces” in Guinea, “militias were responsible for serious abuses of civilians, including killings, torture, and rape and sexual assault” in Sudan, and says the “police and army continue to commit serious and persistent human rights abuses” in Zimbabwe. Israel, Egypt, Iran, Iraq, and Saudi Arabia are also included in this report, yet with far more caution in the use of language.
HRW and the Kidnapped Israeli Soldiers
Two Israeli soldiers, Eldad Regev and Ehud Goldwasser, were kidnapped in July 2006 during a cross border raid by Hezbollah, in which 8 other Israelis were killed. Gilad Shalit was kidnapped from an Israeli position near Gaza in June 2006. All three remain in captivity and Hezbollah and Hamas have refused to allow visitation by the Red Cross.
In 2007, HRW mentioned one or more of the soldiers in a total of only 6 publications (There were 29 documents on Israel, 18 on the Palestinians and 12 on Hezbollah, many of which could have made reference to the soldiers). Of these, only two refer to them by name; the other references are in passing. In one report, HRW “criticized Hezbollah” for “illegally refusing” to confirm the condition of the soldiers and for not allowing the Red Cross access to them (“Why they died”, September 2007).
Charges of “Collective punishment”
In 2007, Human Rights Watch began using the phrase “collective punishment” to refer to Israeli self-defensive measures against rocket fire emanating from Gaza. This sweeping interpretation of the term is not consistent with international law,14 nor with the historic or current usage of the term by Human Rights Watch in other conflicts.
Using a Google search on HRW’s site as a heuristic, 55 percent of HRW’s content referring to blockades and collective punishment is related to Israel.15 (Note that this is 55% of all HRW material and not limited to the Middle East and North Africa section). However, Israel is the only case where HRW uses “collective punishment” to refer to a blockade and the potential impact on civilian life. In other cases, this term is used to describe beatings, murder and destruction of property as indiscriminant retaliation against a group of people for the acts of members of that group.
Other cases of blockades that are not termed “collective punishment,” include Azerbaijan’s blockade of Nagorno Karabakh and Armenia described in Human Rights Watch 1994 World Report:
“Electricity, gas, oil and grain-necessary for the basic human needs of civilians in Armenia-were in extremely short supply… … The lack of gas and electricity deprived Armenians of heat in the freezing winter… a rise in deaths among the newborn and the elderly was accompanied by a higher suicide rate and growing incidence of mental illness. The blockade had ruined Armenia’s industry…”
The report does not refer to this “blockade” as “collective punishment,”, and indeed recommends that “all but humanitarian aid should be withheld from Armenia because of Armenia’s financing of the war”. It is not clear why HRW promotes a policy of economic isolation for Armenia, but when Israel must respond to daily rocket attacks on civilian population centers, HRW condemns a similar policy as “collective punishment.”
Similarly, in a 1999 press release on Chechnya, “Russian Ultimatum to Grozny Condemned” (8 Dec, 1999) HRW described the humanitarian situation as
“rapidly deteriorating, with no functioning hospitals, electricity, running water, gas, or heating since the beginning of November, and dwindling food supplies”.
This is clearly a more urgent humanitarian situation than Gaza in 2007 (where humanitarian aid enters daily16), but HRW did not classify it as “collective punishment.”
In 2007, the term “collective punishment” was used by HRW in 13 items not referring to Israel (see Table 1). These cases generally provide evidence of punitive intent against third parties either at the family or community level.
Table 1 Collective punishment in 2007 HRW publications
The items in Table 1 show HRW’s use of “collective punishment” in highlighting reprisal actions against third parties. For example, in 2007, testimony to a US House Committee on Foreign Affairs, Subcommittee on Africa and Global Health, Sam Zarifi (HRW’s “Washington Advocate”)17 stated,
“in the Ogaden, we have documented massive crimes by the Ethiopian army, including… villages burned to the ground as part of a campaign of collective punishment”.
Another example is found in an August 2007 Guardian article, “Ethiopia’s dirty war” authored by HRW’s London Director, Tom Porteous18. He asserts that
“dozens of civilians have been killed in what appears to be a deliberate effort to mete out collective punishment against a civilian population suspected of sympathising with the rebels.”
These results show that HRW’s application of “collective punishment” is inconsistent and arbitrary. No other “blockade” is described in these terms, and other cases of “collective punishment” involve beatings, murder and destruction of property as indiscriminant retaliation against a group of people for the acts of members of that group.
Use of eye witnesses and unverifiable evidence
In many reports on Israel, Human Rights Watch relies extensively on unverifiable evidence, including accounts from “eyewitnesses” and selected journalists.
In the 145 page (English) report “Indiscriminate Fire”19 Human Rights Watch returns to the disputed Gaza beach incident from 2006.20 This account presents interviews with family members of the victims who claim Israeli shelling was responsible for the causalities. Their stories, however, were contradicted by evidence available to HRW prior to the publication of this report. This evidence includes IDF records of all artillery fired prior to the explosion, the fact that no missiles were actually fired within 10 minutes of the explosion, and forensic evidence of shrapnel removed from victims which was not consistent with an Israeli shell21. When presented with the evidence by the IDF at the time, Human Rights Watch noted (Jerusalem Post, June 19, 2006) that the IDF investigation made “a good assessment” in ruling out an IDF shell as the cause of the blast.22 The use of emotionally involved witnesses, in this case a year after their testimony is invalidated by evidence, is a systematic fault in HRWs research methodology. Their approach favours narrative over verifiable evidence-based research.23
Another section of the “Indiscriminate Fire” report discusses an exchange of fire in the vicinity of the Nada Apartments Complex, situated on the western edge of Bet Hanoun in Gaza. A footnote states that, “the testimony of one resident that a rocket may have been fired from the roof of one of the buildings on the evening of July 24 was not supported by Human Rights Watch’s on-site investigation”. This investigation was a physical inspection which found “no burn marks or other indications that a rocket had been fired from there”. Such a lack of evidence cannot prove anything — it only fails to support the eyewitness claim. The report goes on to say that “Many Nada residents freely acknowledged Palestinians firing rockets from the nearby open areas, sometimes close to the apartments, but all except for this person insisted that no rockets had ever been fired from the grounds of the complex or the roofs of the buildings.” In this case, HRW tries to bolster the credibility of eye-witnesses, who have every reason to lie (they surely know that rocket-fire from their homes would render them legitimate targets for IDF reprisal). In order to support their conclusions, HRW does its own investigation which finds nothing, and rejects the evidence of one eyewitness whose counter-evidence is the only non-interested testimony HRW has.
In a third example, HRW’s report “Why they died”24 acknowledges the errors resulting from reliance on eyewitnesses. It states, “In our earlier report, Fatal Strikes, Human Rights Watch did not have information about Hezbollah firing from the area.” The authors further note that the eyewitness stated “To my knowledge, Hezbollah was not operating in the area, but I can’t be 100 percent sure because we were sleeping.’” This quote omits the first sentence of the original quote from Fatal Strikes which stated, “I am positive the family had nothing to do with Hezbollah” (pg 25).
Nevertheless, “Why they died” also uses eyewitness accounts to assert that “the handful of cases of probable shielding that we did find does not begin to account for the civilian death toll in Lebanon.” The cases HRW found may well be a fraction of the total cases. In an area still largely controlled by Hezbollah, there is a strong incentive against whistle blowing on the use of Human shields, and many uses of civilian space will leave no physical evidence. A lack of evidence cannot be used to make categorical and empirical claims.
Although HRW’s relative focus on Israel in 2007 is less than in 2006, the disproportionate emphasis, and the examples of bias and double standards continue. Despite a major increase in internal Palestinian violence during 2007, including the Hamas takeover of Gaza, HRW’s focus on Israel has not changed significantly. Similarly, in the selective use of language, especially the arbitrary accusation of “collective punishment,” HRW’s strong political agenda with respect to Israel is clear.
See the appendix to this report
1 Lebanon and Iraq have been excluded from this longitudinal analysis due to the impact and separate reporting on external actors in these countries. These factors make data analysis on these countries between years inconsistent. This does not impact on the accuracy of reporting within a given year.
2 121 of 1260 points were devoted to Israel in 2007, compared to 59 of 702 points in 2005. In 2006 223 of 1088 points (over 20%) were focused on Israel, this is partly explained by the Second Lebanon War. The total publications (in terms of both weighted score and in terms of actual number of documents) has steadily increased across the region over the last three year. The percentage of these documents dedicated to Israel has been erratic over the three year period.
3 The pictures and audio alleging that Israel prevented residents of Gaza from leaving to seek external medical treatment are also linked at the top to a press release. This has been included as a multimedia item in this report.
4 “Indiscriminate Fire” and “Why they Died.”
5 One on the Lebanon War (Why they died), one on Iraqi refugees (Rot here or die there) and one on domestic workers (Exported and Exposed)
6 According to the data published by B’tselem, 351 Palestinians were killed by other Palestinians (http://www.btselem.org/english/Statistics/Casualties_Data.asp?Category=23) compared with 379 casualties from clashes with Israeli security forces (http://www.btselem.org/english/Statistics/Casualties_Data.asp?Category=1). Such data is not reliable, and if anything, underestimates the extent of internal Palestinian violence.)
7 Examples include “Hospitals offer no safety in Gaza strip,” ABC News, 13 June, 2007 and “Crowd loots Gaza home of Arafat,” Ali Waked and Reuters, 16 June, 2007.
8 To avoid double counting, publications with attention given to at least one of: the Palestinian Authority / Fatah, Hamas or “Palestinian armed groups” were included in the Palestinian count.
9 In the language analyses of previous years, NGO Monitor included “torture”. In 2007, HRW did not use the term torture in any reports on Israel, while Palestinian torture related to factional violence received a single mention. This data is presented in a separate graph in the appendix.
10 See methodology section for more details
11 “Is Israel Bound by International Law to Supply Utilities, Goods, and Services to Gaza?” Abraham Bell, JCPA Vol. 7, No. 33 28 February 2008
12 Peter Finn, Russia Cuts Off Gas to Ukraine In Controversy Over Pricing, Washington Post, Monday, January 2, 2006, Page A07
13 “Is Israel Bound by International Law to Supply Utilities, Goods, and Services to Gaza?” Abraham Bell, JCPA Vol. 7, No. 33 28 February 2008
14 See “International Law and Gaza: The Assault on Israel’s Right to Self-Defense,” Jerusalem Viewpoints (Jerusalem Center for Public Affairs) Vol. 7, No. 29, Abraham Bell, January 28, 2008; and Dr. Avi Bell, “Is Israel Bound by International Law to Supply Utilities, Goods, and Services to Gaza?” Jerusalem Viewpoints (Jerusalem Center for Public Affairs) Vol. 7, No. 33, 28 February 2008,
15 When searching site:hrw.org blockade and “collective punishment” +Israel 77 results are returned. When searching site:hrw.org blockade and “collective punishment” -Israel 63 results are returned. The 77 documents represent 55% of the total. (Some publications include separate sections in which both terms are used, and are counted twice.) These are heuristic measures only as HRW reports may be listed two or more times due to posting as pdf files as well as web pages.
16 See IDF Spokesperson reports of trucks entering Gaza with supplies; Aluf Benn, Dichter: Israel to allow aid supplies, food into Gaza, Haaretz, 16 June, 2007
17 “Ethiopia and the State of Democracy: Effects on Human Rights and Humanitarian Conditions in the Ogaden and Somalia,” HRW, October 2, 2007
18 See “Ethiopia’s dirty war,” Tom Porteous, London director HRW, Published in Guardian Unlimited August 05, 2007.
19 “Indiscriminate Fire: Palestinian Rocket Attacks on Israel and Israeli Artillery Shelling in the Gaza Strip,” HRW, July 1, 2007.
20 See “Gaza beach incident: Timeline of HRW involvement and activities June 9-21 2006,” NGO Monitor, June 21, 2006
21 Yaavoc Katz, IDF not responsible for Gaza blast, Jerusalem Post, Jun 13, 2006
22 Yaakov Katz and Judy Siegel-Itzkovich, Gaza beach blast victim wakes, Jerusalem Post, June 19 2006 (updated June 20th)
23 HRW has a history of publishing “facts” despite being aware of contradictory evidence. For example after the July 30, 2006 Qana incident in the Lebanon War, Lucy Mair of HRW disregarded the Red Cross estimate of 28 casualties in favor of the higher estimate of 54 provided by an alleged “survivor” [see Kim Murphy, “Warfare in the Middle East: Officials Say 28 Die in Qana not 54,” The Los Angeles Times, August 4, 2006.].
24 “Why They Died: Civilian Casualties in Lebanon during the 2006 War,” HRW, September 5, 2007