The timeline above reveals where humanitarian crises have been most concentrated in the last three decades— and where aid workers have been killed, wounded, or kidnapped as a result of deliberate violence. Unsurprisingly, regions affected by prolonged conflict show consistently higher numbers of affected humanitarian personnel over the years.
Humanitarian aid workers risk their lives to support those enduring inhumane and unjust conditions. They step into volatile, war-ridden, often dangerous environments out of a commitment to serve others. They should not be targets for standing up for humanity.
The AWSD dataset exists to document these acts of violence, preserving a record of what has been done to those who dedicate themselves to helping others. These records are essential—not only for accountability, but for informing stronger protection, prevention, and recovery strategies moving forward.
Data Source
The Aid Worker Security Database (AWSD) is a comprehensive, publicly accessible repository that documents major violent incidents against humanitarian workers globally. It is maintained by the nonprofit research organization Humanitarian Outcomes and funded by USAID. AWSD gathers its information from multiple sources including media reports, security incident reports from humanitarian organizations, UN security reports, and direct reporting from field security personnel. This data is then reviewed and managed by expert personnel.
In AWSD, each event represents a unit of analysis. The database documents instances of major violence against aid workers, which includes:
Killing
Wounding (serious injury requiring medical attention)
Kidnapping
Rape and sexual assault
It also attempts to provide reliable and transparent information on a variety of other variables, including the location of the attack, the means of violence, the humanitarian aid organization the victim(s) belong to, and the identity and affiliation of the perpetrator.
Aid workers are defined as employees and associated personnel (both paid and volunteer) of nonprofit aid agencies providing material and technical assistance in humanitarian contexts. This encompasses:
Intergovernmental Organizations (IGOs):
United Nations (UN) IASC Agencies: FAO, OCHA, UNDP, UNFPA, UNHCR, UNICEF, UN-Habitat, WFP, WHO
Local volunteer groups and civil society organizations
The AWSD’s definition of a humanitarian aid worker includes locally contracted staff like drivers and security guards. It also covers contracted workers and vendors of humanitarian organizations when they are affected by violence while supporting humanitarian missions.
Explore the visualization below to compare and follow how different types of humanitarian aid organizations have been impacted by violence over time.
Code
import pandas as pdimport plotly.graph_objects as go# Aggregate data by year for each organization typeyearly_data = df.groupby('year')[['un', 'icrc', 'nrcs_and_ifrc', 'ingo', 'nngo', 'other']].sum()# Reset index to make 'year' a column againyearly_data = yearly_data.reset_index()# Create a prettier name mapping for labelsorg_names = {'un': 'United Nations','icrc': 'Int. Committee of the Red Cross','nrcs_and_ifrc': 'National Red Cross & IFRC','ingo': 'International NGO','nngo': 'National NGO','other': 'Other Organizations'}# Create the interactive plotfig = go.Figure()# Red color palettecolors = ['#0294f5', '#870000', '#e30000', '#fca103', '#fcc203', '#fcde65']# Add each organization as a separate linefor i, org inenumerate(['un', 'icrc', 'nrcs_and_ifrc', 'ingo', 'nngo', 'other']): fig.add_trace(go.Scatter( x=yearly_data['year'], y=yearly_data[org], mode='lines', name=org_names[org], line=dict(color=colors[i], width=3), hovertemplate=f"{org_names[org]}<br>Year: %{{x}}<br>Victims: %{{y}}<extra></extra>" ))# Update layout with Merriweather fontfig.update_layout( title='Violence Victims by Aid Organization (1997-Present)', xaxis_title='Year', yaxis_title='Number of Victims', legend_title='Aid Organization', hovermode='closest', template='plotly_white', font=dict( family="Merriweather, serif", size=12 ), title_font=dict( family="Merriweather, serif", size=20 ))html_file ='aid_orgs.html'pio.write_html(fig, file=html_file, include_plotlyjs=True, full_html=True)# import font and add JavaScript fixhtml_template ="""<style>@import url('https://fonts.googleapis.com/css2?family=Merriweather:ital,wght@0,300;0,400;0,700;0,900;1,300;1,400;1,700;1,900&display=swap');</style>{js_code}{plot_html}"""# display in rendered htmlfrom IPython.display import IFrame, HTML, displayplot_html = pio.to_html(fig, include_plotlyjs=True, full_html=False)final_html = html_template.format(js_code=js_code, plot_html=plot_html)display(HTML(final_html))
The AWSD data shows that International NGO workers (excluding UN staff) have experienced the highest number of targeted violence incidents over time, followed by National NGO workers and then UN staff.
Interactive Timeline of Humanitarian Attack Data
This alternative interactive visualization below offers both a high-level overview and detailed insights into humanitarian attack events. This will allow you to gain insight on the scope and trends of violence over time in different parts of the world.
You can customize your exploration by selecting a variable of interest for the y-axis using the dropdown menu labeled “Y-axis” at the bottom. Each point is color-coded by the region where the incident occurred. You can zoom in and out across time, and hovering over a point reveals detailed information—including the country, date, and the number of people affected, based on your selected variable.
