Evicted and Abandoned

About this project

About this project

"Evicted and Abandoned" is an 11-month investigation by the International Consortium of Investigative Journalists (ICIJ), Huffington Post, the Investigative Fund, the GroundTruth Project and more than 20 other media partners around the world. It looks at how projects funded by the World Bank have forced people from their homes, taken their land or damaged their livelihoods.

ICIJ found that nearly 1,000 projects approved between 2004 and 2013 have physically or economically displaced an estimated 3.4 million people. The World Bank’s “social safeguards” policy says resettlement should be avoided when possible. If not, “displaced persons should be assisted in their efforts to improve their livelihoods and standards of living or at least to restore them ... to pre-displacement levels.”

ICIJ discovered that basic information – such as how many people would be negatively affected by projects – was difficult to find. To track down estimates for the number of people affected by the bank-funded projects, our data and research unit reviewed tens of thousands of pages of documents in four languages. We also had several meetings and many email conversations with World Bank officials to clarify questions about our methodology and make sure we were accurate.

By providing detailed information on a per-project basis, as well as aggregates by countries and sectors, ICIJ seeks to make information about the World Bank-funded projects causing displacement more transparent. This data is not searchable on the bank’s Internet sites.

ICIJ doesn’t assert that all affected people mentioned in this interactive have not been fairly compensated for their losses. The World Bank says that in many cases it has restored or improved people’s living conditions and livelihoods, but it acknowledges that shortcomings in its oversight of displacement cases often makes it impossible to say precisely how many people were given enough help and how many weren’t.

Where does the data in this interactive come from?

It comes entirely from public information made available online by the World Bank, except for a few documents the bank provided directly to ICIJ. To create this database we had to manually review more than 4,000 files and 1,500 projects. You can download the data used in this interactive here (updated May 26, 2015).

All the documents about the number of people affected and the following project details were gathered from the bank’s website: approval and closing dates, status, amounts, sectors of investment and environmental categories. Information related to complaints comes from the website of the Inspection Panel, the bank’s independent accountability arm.

This interactive does not include any projects funded by the International Finance Corporation (IFC), which reports information about resettlement differently and in less detail.

How up-to-date is the data?

All the projects included in this database were approved between 2004 and 2013. But a project can last years and its information may be updated frequently, so ICIJ did not take into account any document made available online by the bank after Dec. 31, 2014.

Data from the World Bank website, including dates, amounts or project status, was last updated April 2, 2015. Information about complaints was last updated Feb. 3, 2015.

This database was released on April 16, 2015, the day the stories based on this investigation began to be published.

On May 26, 2015 ICIJ removed three projects previously listed in the database as involving “possible displacement” because the end-of-project documentation indicated that physical or economic displacement did not ultimately happen. The removal of these projects did not affect the estimated number of people displaced.

On July 15, 2015 ICIJ added one complaint that was not previously listed in the database after further review of the cases received by the World Bank's Inspection Panel.

How did ICIJ estimate the number of projects causing displacement?

ICIJ looked at all projects approved between 2004 and 2013 that had a resettlement document available on the World Bank website to determine if there was confirmed or possible displacement. We also included projects for which no resettlement document was available but for which the bank received a complaint about land acquisition or involuntary resettlement.

We excluded projects for which World Bank documents indicated no displacement had happened as well as those that were not finally approved, that did not have approval dates or that could not be could not be identified with any particular country. We also excluded additional financing projects connected to projects approved before 2004.

In order to avoid duplication, we determined which projects were additional financing or supplemental to main projects. If the main project was already in our data, we counted all those projects – the main project and the subprojects – as a single project.

ICIJ divided projects with displacement into two categories: possible and confirmed. Displacement was confirmed by the bank through a resettlement plan or through documents produced after the project was completed. If bank documents indicated resettlement was likely or the project could potentially restrict access to natural resources, we included it in the “possible” category.

How did ICIJ estimate the number of people displaced?

ICIJ extracted the number of people physically and economically displaced from estimates in documents called “Resettlement Action Plans,” published online by the World Bank. We used all the documents available up to Dec. 31, 2014, as well as 43 documents sent directly by the bank. When projects were closed and there were no resettlement documents, ICIJ looked for estimates of displacement in Implementation Completion Reports (ICR). Most ICRs, however, did not include a census of the number of people displaced.

Projects had one or multiple resettlement plans attached to them. To avoid counting affected people multiple times, resettlement plans referring to the same displaced people were only counted once. Some documents only available in certain languages (Indonesian, Chinese, Arabic) were not included in our analysis if the bank did not provide an English version. All documents in English, French, Spanish and Portuguese were included in the analysis.

Inside the documents, ICIJ looked for what the bank identifies as “Project-Affected People (PAP)” or “displaced people.” Inconsistent documentation makes it very difficult to have an exact number of displaced people by project or by country. The World Bank recently acknowledged “major problems” in its handling of resettlement, admitting that projects involving physical or economic displacement “often had poor or no documentation.”

Some projects only count households, while other count individuals. Some detail the type of impacts while other don’t, only identifying an overall number of affected individuals.

For these reasons, ICIJ focused on the broader category of project-affected people, which was the most commonly used term across projects.

What does a “minor impact” mean in terms of displacement?

The range of resettlement impacts stated in the World Bank documents can be varied and incoherent. Some individuals will have to move to another house, but the smallest impacts can be as modest as losing less than 1% of their land.

Some projects indicate the percentage of impacts that the borrower considers “minor.” Others list the precise percentage of land taken. Some provide little information, which makes it difficult to go into detail about the type of consequences people suffered across projects.

In a 2012 internal report, the World Bank acknowledged that resettlement plans often use the term “project-affected people” (PAPs) “without disaggregating the type of impacts those people faced.”

ICIJ includes a footnote in this interactive if a project affected more than 20,000 people but bank documents said most were expected to experience only minor impacts.


News Application Developer

Matthew Caruana Galizia

Data Journalist

Cécile Schilis-Gallego

Data Editor

Mar Cabra


Álvaro Ortiz

Assistant News Application Developer

Cristiam Da Silva


Friedrich Lindenberg, Shane Shifflett, Hamish Boland-Rudder and Allison Pohle


Will Fitzgibbon, Marcos García Rey, Delphine Reuter and Emilia Díaz-Struck

Photo by Sami Siva / International Consortium of Investigative Journalists