Boundaries, Borders, and Front Lines: Coding Territorially Ambiguous Units

Arguably one of the most common misconceptions about the PTS is that the our scores constitute an assessment or judgment about true state of human rights conditions in countries around the globe. In this view, a PTS score of say 5 is understood to indicate that human rights conditions in a particular country are horrendous, whereas a score of 1 provides a clean bill of health – human rights are well protected and no abuses of basic physical integrity rights have taken place.

In reality, the scope of the PTS is much more limited. What we provide are simply assessments of human rights conditions as reported by human rights monitoring organizations (i.e., Amnesty International, Human Rights Watch, the U.S. State Department). PTS coders read human rights reports and assign scores based on the conditions described in the annual human rights reports. In other words, we assign scores to reports and not to countries.

More accurately then, a score of 5 indicates that a monitoring organization (e.g., Amnesty International) produced a damning report describing human rights conditions as especially dire. A score of 1 would indicate that no abuses were reported by the organization for a particular country. While this distinction might appear arcane and technical, it is critical to understand and interpret the scores we produce. As such, a score of 5 only indicates that a monitoring organization wrote a devastating report. The reality on the ground may very well be different. Similarly, a score of 1 does not necessarily mean no abuses have taken place but merely that none were reported.¹

Importantly, the PTS’s focus on reports as the unit of analysis (see also: PTS Codebook) has important implications for our work and creates a vexing challenge for PTS coders every year. Because we score or classify human rights reports based on the reported abuses that were perpetrated by state actors rather than by private or civil society actors (see: State Actor Requirement in the PTS Codebook) inside the territorial boundaries of the entity in question (see: Territoriality Requirement in the PTS Codebook), coders have to determine whether any reported abuse was perpetrated by a state actor and whether it took place inside of the boundaries of the country or territory in question.

It should be noted here that our emphasis on state actors inside of the territorial boundaries of the state or territory in question, implies that PTS scores often do not accurately represent the extent to which a country protects or respects human rights. All too frequently, a country’s human rights record outside of its territorial borders is far worse than its human rights practices domestically. Similarly, the PTS is likely an incomplete measure of the human rights conditions citizens actually experience, as only human rights violations perpetrated by state actors are considered by coders. State perpetrated abuses are likely dwarfed by non-state sources of human insecurity.

Applying the State Actor Requirement and especially the Territoriality Requirement to reports is complicated by the fact that all three monitoring organizations whose reports we rely on (Amnesty International, Human Rights Watch, and the U.S. State Department) regularly produce reports covering entities that are not conventional nation-states or whose boundaries are contested. As a result, the coding rules require coders to determine who qualifies as the relevant state actor in territories that are not conventional states and whose borders are contested and often in flux.

For example, all three have over the years issued reports for Western Sahara (a non-self-governing territory under partial Moroccan military occupation). Amnesty International and the U.S. State Department publish annual reports on Taiwan (a territory claimed by the People Republic of China). The State Department provides sub-reports (embedded under the China report) for Hong Kong, Macau, and Tibet, Amnesty International a report for Puerto Rico (a U.S. territory), and Human Rights Watch reports for the European Union and the African Union (both supra-national organizations). The U.S. State Department also issued reports for Crimea (which Russia annexed from the Ukraine in 2014) and more recently all three monitors published reports (or sub-reports) for Russia-occupied areas in the Ukraine. A final set of territorially ambiguous or contested reports (or sub-reports) concern Israel, the Israeli Occupied Territories, Gaza, the West Bank, the Palestinian Authority, and Palestine.

While we have by no means been consistent over the years when coding these reports (see the section: The Unique Case of Israel, the Occupied Territories, and Palestine below), our approach to coding reports for these units, entities, or territories is as follows:

Example 1: Morocco and Western Sahara

1. If a monitoring organization releases a separate report (or a nested sub-report under the Morocco report) for Western Sahara, we treat both Morocco and Western Sahara as distinct units and assign a score for each. Coders are asked to only take into account abuses perpetrated by Moroccan state agents (e.g., Moroccan police or military) inside Morocco – that is not in Western Sahara. Any human rights violation by Moroccan authorities inside Western Sahara is to be treated as extra-territorial and should not affect Morocco’s human rights score at all.²: Similarly, to assign a score to the Western Sahara report, coders are instructed to only consider actions by Western Saharan authorities inside of Western Sahara.

2. If a monitoring organization issues only a report for Morocco (and not for Western Sahara) which treats Western Sahara as a part of Morocco, coders are instructed to consider any abuses by Moroccan authorities inside Morocco and Western Sahara when assigning a score. Abuses perpetrated by non-Moroccan authorities should be disregarded like those of other non-state actors (e.g., rebel or insurgent groups, private citizens, corporate actors).

Example 2: The European and African Union

When coding Human Rights Watch reports for the African and European Union, coders are to consider any human rights violations perpetrated by African or European authorities as well as those perpetrated by authorities of individual member states that occur inside the territorial boundaries of member states.

Example 3: Russia, Ukraine, Crimea, and the Russia-occupied areas of the Ukraine

In 2019, 2021, and 2022, the U.S. State Department issued a sub-report (under the Ukraine report) covering human rights conditions in Crimea, which Russia had annexed in 2014. Coders are instructed to consider physical integrity rights violations perpetrated by the de-facto authorities in Crimea (i.e., Russian police, military, et cetera) when coding this sub-report. Importantly any abuse perpetrated by Russian or Ukrainian authorities in Crimea must be disregarded as extra-territorial for purposes of coding the reports for Russia and the Ukraine.

Starting in 2022, the State Department also produced a sub-report for Russia-Occupied Areas as part of its Ukraine report.³ Just as the report for Crimea, coders are tasked with treating the Russian occupiers as the de facto state actor and to consider only violations perpetrated by Russian occupation authorities (e.g., military) in the occupied areas when assigning a score. Again, any abuses perpetrated by Russian or Ukrainian forces within the boundaries of the occupied areas must be disregarded as extra-territorial when coding the Russia and Ukraine reports.

The Unique Case of Israel, the Occupied Territories, and Palestine

The most prominent, and likely most controversial, example of territorial ambiguity concerns reports and sub-reports for Israel, the Occupied Territories, the West Bank, Gaza, the Golan Heights, East Jerusalem, and Palestine. All three monitoring organizations, on whose reports we base our scores, have reported on these entities differently. Since 1977, the U.S. State Department has issued a report for Israel containing a detailed sub-report for either the Occupied Territories, or for Gaza and the West Bank. Amnesty International has published reports on Israel including the Occupied Territories since 1977, but in addition has produced a separate report for the Palestine Authority/Palestine starting in 1997. Human Rights Watch reported only on the Occupied Territories from 1989 to 1997 and began publishing a single report for Israel and the Occupied Territories/Palestine since 1998. However, these combined reports contain explicit – albeit brief – subsections on the Palestinian Authority (in the West Bank) and Hamas (in Gaza).

History of the Reports for Israel, the Occupied Territories, and Palestine

Below, I provide a summary of the changes to report and sub-report titles as well as section headers for the annual reports published by Amnesty International, Human Rights Watch, and State Department since 1977.

U.S. State Department Israel Reports

• 2020–2025: Israel, West Bank and Gaza
  • sub-report: West Bank and Gaza
• 2018–2019: Israel, Golan Heights, West Bank, and Gaza
  • sub-report: West Bank and Gaza
• 1977–2017: Israel and the Occupied Territories
  • sub-report: The Occupied Territories

Amnesty International Israel Reports

• 2008–2025: Israel and the Occupied Palestinian Territories
• 1998–2007: Israel and the Occupied Territories
• 1995–1997: Israel (the State of) and the Occupied Territories, including areas under the Palestinian Authority’s jurisdiction
• 1977–1994: :Israel (the State of) and the Occupied Territories

Amnesty International Palestine Reports

• 2014–2025: Palestine (State of)
• 2000–2013: Palestinian Authority
• 1997–1999: Palestinian Authority (areas under the jurisdiction of the)

Human Rights Watch Israel/Palestine Reports

• 2013–2025: Israel and Palestine
  • subsection: The Palestinian Authority
  • subsection: Hamas and Palestinian Armed Groups
• 2012: Israel/Occupied Palestinian Territories
  • subsection: The Palestinian Authority
  • subsection: Hamas
• 2010–2011: Israel/Occupied Palestinian Territories
  • subsection: The Palestinian Authority
• 2009: Israel/Occupied Palestinian Territories
  • subsection: The Palestinian Authority (PA) and Hamas
• 2006–2008: Israel/Occupied Palestinian Territories
• 2005: no country reports published
• 1998–2004 Israel, the Occupied West Bank & Gaza Strip
  • subsection: Palestinian Authority Territories
• 1992–1997 Israeli-occupied West Bank and Gaza Strip
• 1989–1991 Israeli-occupied Territories

How Have We Coded These Reports?

The short answer is inconsistently. To cover the bewildering diversity in reports among the three monitoring organizations, we have over the years coded the following units in our data set:

• Israel (1976–2015) – Israel including Occupied Territories
• Israel (2016–2024) – Israel not including Occupied Territories
• Israel in the Occupied Territories (1994–2015, 2023, 2024)
• Israel in pre-1967 borders (1994–2015)
• Palestine, State of(1994–2024)
• Gaza (Hamas) (2011-2014)

Beginning with Amnesty International’s and the U.S. State Department’s 1977 reports, the PTS team has coded the unit Israel. Up until the 2016 reports, coders treated the West Bank, Gaza, the Golan Heights, and East Jerusalem as under Israeli control for coding purposes.⁴ Any physical integrity rights violations by Israeli authorities in Israel or the Occupied Territories were factored into Israel’s score.

In 1995, we began to code three additional units: Israel in the Occupied Territories which accounts only for abuses perpetrated by Israeli authorities in the Occupied Territories; Israel in pre-1967 borders for violations perpetrated by Israeli authorities inside of the boundaries as they existed prior to the 1967 Six-Day War (i.e., not including the Occupied Territories); and Palestine, State of which captures violations perpetrated by the Palestinian Authority in the Occupied Territories, and later those by Hamas in Gaza.⁵

It is clear that the coding of these units is entirely inconsistent with the PTS Codebook in so far as we claim to code reports, yet none of the reporting organizations provide a report for Israel in the Occupied Territories or Israel in pre-1967 borders, and only Amnesty International has published corresponding reports for the Palestinian Authority or Palestine, State of.

In 2012 we added the unit Gaza (Hamas) to distinguish abuses perpetrated in the parts of the Occupied Territories controlled by the Palestinian Authority, from those perpetrated by Hamas after its takeover of Gaza in 2007. As such, from 2012 to 2016 we were coding five distinct units (most of which did not correspond to a report from either Amnesty International or the State Department). We produced a score for Israel and the Occupied Territories, Israel in the Occupied Territories, Israel excluding the Occupied Territories, and separate scores for the Palestinian Authority and Hamas in the territories they controlled, respectively.

Producing five scores based on two Amnesty International reports and five scores based on one State Department report (and one sub-report) was incredibly challenging for our coders who were tasked annually to parse the often vague reporting of dynamic borders, boundaries, and front lines, as well as to identify the appropriate de-facto state actor for each unit. To improve matters and bring our coding activities into line with the PTS codebook, in 2017 we reduced the number of units to two: Israel and Palestine, State of.

Thus, the score for Israel would now exclude Israeli abuses perpetrated in the Occupied Territories and treat them as extra-territorial. It is therefore equivalent to the dropped unit Israel in pre-1967 borders. Importantly, this implies that PTS scores for Israel from 1976 to 2015 are not comparable to scores for Israel from 2016 to the present, as they are two distinct territorial units merely sharing a name. The score for Palestine, State of now includes abuses perpetrated by any de-facto state actor (i.e., the Palestinian Authority and Hamas) against individuals in territories under their control (i.e., the West Bank and Gaza).

It is worth reiterating that the PTS aims to capture state perpetrated physical integrity rights violations against individuals within a state’s territorial jurisdiction. And although the decision to code only Israel and Palestine, State of certainly reduced the complexity of our coding effort, it also meant that Israeli abuses in the Occupied Territories did not affect Israel’s PTS scores, while violations by Palestinian authorities (such as Hamas) inside of Israel’s pre-1967 borders were excluded from Palestine’s PTS scores.

Although, disregarding extra-territorial abuse may align with our coding rules, applying our rules rigidly amid the contested, indeterminate, and legally ambiguous status of the Occupied Territories and Palestine in relation to Israel calls the substantive value of the resulting PTS scores into question. In large part because of these concerns, following Hamas’ October 7, 2023 attack on Israel and the ensuing Gaza War, we suspended coding the Israel and Palestine reports in 2024.

This year (2025) we resumed our coding of the Israel and Palestine reports and also coded the reports from last year. In response to our own concerns as to how to approach these reports, we brought back the unit Israel in the Occupied Territories to go along with our scores for Israel (which again is equivalent to Israel in pre-1967 borders) and Palestine, State of. Our Israel scores accounts for reported abuses by Israeli authorities inside Israel’s pre-1967 borders excluding abuses perpetrated in the Occupied Territories. Our scores for Israel in the Occupied Territories cover violations by Israeli authorities in the Occupied Territories (i.e., the West Bank, Gaza, the Golan Heights, and East Jerusalem). Scores for Palestine, State of account for abuses by the Palestinian Authority and Hamas in the areas they nominally control. As such, our coding of the reports covering the various Israel and Palestine reports mirror our approach to coding Russia, the Ukraine, and the Russian occupied areas of Ukraine (i.e., Crimea, and parts of Eastern Ukraine). The 2023 and 2024 PTS scores for these units are reported in Table 1, below.

Unit	PTS_A 2023	PTS_A 2024	PTS_H 2023	PTS_H 2024	PTS_S 2023	PTS_S 2024
Israel	3	4	4	4	4	3
Israel in the Occupied Territories	5	5	5	5	5	4
Palestine, State of	4	4	4	3	4	4
Russian Federation	4	5	4	4	5	5
Russia-Occupied Areas (Ukraine)	5	–	5	–	5	5
Ukraine	3	1	4	2	3	2

Note: Shown are the most recent PTS scores included in this years release.

Our decision to code three units for 2023 and 2024 represents a judgment call in favor of greater consistency with our coding rules. The drawback, however, is that the human rights conditions experienced by citizens on the ground – specifically Israelis within the pre-1967 borders – are not fully reflected.

Measuring Human Insecurity: Consistency at the Expense of Validity?

The case of Israel and Palestine is, unfortunately, only the most prominent – and arguably the most controversial – example of an inconsistent application of our coding rules. Other cases where the State Actor Requirement and the Territoriality Requirement pose challenges include many states experiencing civil conflict (e.g., Yemen, Libya), where it is often unclear who the state actors are and what territory they actually control. “Failed states” (e.g., Somalia) and unrecognized states (e.g., Somaliland) represent another category that is almost impossible to code.

Ultimately, the challenges and the trade-offs discussed above highlight both the necessity and the limitations of our approach. By coding reports rather than attempting the impossible task of reconstructing a complete record of human rights practices, we acknowledge that our data can never fully capture the complexities and realities on the ground. At times, the decisions we make may seem arbitrary, especially in contested, fragile, or ambiguous political contexts. Still, by striving for transparency and adhering as consistently as possible to our coding rules, we aim to provide researchers with a tool that, while limited, remains useful for systematic comparison and long-term analysis.

---