“Folks, We Are Going to Have a Bad August”
In May 2021, I was on a coordination call about Afghanistan. The United States was preparing to withdraw, the Taliban was advancing, and the backlog of Afghans who had supported the American mission and were still waiting for Special Immigrant Visas had become impossible to ignore. I had spent years working on behalf of America’s wartime allies, and I understood what those numbers meant. These were not just applications in a system. They were people who had worked alongside Americans and had been told we would not abandon them because of the risks they took. As we discussed the withdrawal and the months ahead, I remember seeing everything converge and saying, almost matter-of-factly, “Folks, we are going to have a bad August.”
I hadn’t started out working on Afghanistan. My involvement grew from a much smaller effort that became something I never expected.
I first became involved with America’s wartime allies through the Special Immigrant Visa program. I learned about Afghans who had worked alongside American forces and organizations, often at great personal risk. They translated for troops, supported American missions, and put themselves and their families in danger because of their association with the United States. What stood out to me was the gap between our promises and their reality. The United States had created a pathway to safety, but the system was painfully slow. Applications could sit for months or years while the security situation around them worsened. Over time, it became impossible to treat this as just a bureaucratic problem.
That work eventually brought me into the Association of Wartime Allies. By 2021, I was working closely with Matt and Kim, along with everyone at AWA and the Evacuate Our Allies coalition. We were trying to understand what was happening and, more importantly, what was about to happen. We were looking at the SIV backlog, tracking the pace of the withdrawal, analyzing the numbers, and trying to turn an enormous amount of information into something policymakers, the media, and the public could understand.
AWA and the Evacuate Our Allies coalition had been warning for months that the United States needed to move much faster. By the spring of 2021, it was becoming increasingly difficult to see how the existing system could possibly keep up with the danger developing on the ground. The problem wasn’t simply that there were too many applications. The problem was that the clock was now running against the people who were waiting.
As the withdrawal approached, our work became more concrete. We were publishing reports and data analysis, putting out almost daily updates, and tracking the number of aircraft that would be required to evacuate the people who were eligible to leave. We started counting down the planes we would need, because we wanted to make the scale of the problem impossible to ignore. There was a finite number of days before the withdrawal and a finite capacity to move people. The math was not complicated. What was difficult was getting people to understand what the math meant.
I remember spending enormous amounts of time with Matt and Kim working through the latest information, checking numbers, updating our analysis, and figuring out what we could say with confidence. The situation changed so quickly that something we had written in the morning could be overtaken by events by the afternoon. We were constantly revising our understanding of what was happening and trying to stay ahead of it. It was exhausting, but I also felt a growing urgency that we couldn’t simply say that the situation was bad. We needed to document it, quantify it, and keep putting it in front of people who had the ability to act.
By May, I had been watching this problem for years. I knew the backlog. I knew how many people were still waiting. I knew how slow the system was even in normal conditions. And I knew the United States was preparing to leave.
That is what I was thinking about on that call.
The withdrawal was still months away, but the warning signs were clear. The Taliban was gaining ground, and the Afghan government was struggling to hold territory. There was still time to prepare, but the scale of what needed to be done was enormous. The work we were doing through AWA and the Evacuate Our Allies coalition was increasingly focused on one question: how do you move tens of thousands of people out of a country before the American military leaves?
At the time, August still felt distant. That was part of the frustration. We knew the withdrawal was coming. We knew the Taliban was advancing. We knew tens of thousands of people were stuck in a system already under strain. There was still time, but only if people were willing to accept how fast things could fall apart.
Then the summer accelerated.
One city fell, then another, then another. The map of Afghanistan changed every few days. What had looked like a difficult transition began to look like a collapse. From the United States, I watched it unfold with growing unease. The warning I had made in May no longer felt like a prediction. It felt like something we were watching happen in real time.
The work at AWA and across the Evacuate Our Allies coalition intensified with it. Matt, Kim, and everyone involved were no longer simply trying to prepare for a crisis. We were trying to keep up with one.
The reports and analysis that had begun as advocacy became a running record of a crisis unfolding in real time. We were constantly updating the numbers, tracking the evacuation effort, following the situation on the ground, and trying to communicate the magnitude of what was happening to people outside the relatively small community that had been paying attention to the SIV crisis.
There was no single dashboard that told us what was happening. We had to pull information together from different sources, reconcile it, analyze it, and then try to turn it into something useful. The work was relentless. We were publishing updates almost daily, continuing the plane countdown, producing analysis, and trying to make sure that the urgency of the situation did not get lost in the noise surrounding the withdrawal.
When Kabul fell on August 15, everything changed.
The images from the airport were hard to process. Thousands of people were trying to get inside while American aircraft came and went. Families were trying to understand where to go, what documents they needed, and whether they had any chance of getting through the gates. After twenty years, the American presence had collapsed into a few square miles around an airport.
I had spent years working on behalf of the people caught in that withdrawal, so I understood what I was seeing. Many in those crowds had worked with the United States. They had believed our promises, and now those promises depended on whether they could reach an airport in the middle of a rapidly deteriorating security situation.
The systems meant to process them were never built for this. Information was scattered across databases, spreadsheets, emails, and organizations that were never designed to function as one evacuation system. People who had waited for years suddenly had days, sometimes hours, to move through a process that was already overwhelmed.
Our work changed again. We were no longer trying to warn about what might happen. We were documenting what was actually happening.
I became consumed by it. The days blurred as we tried to reconcile data, identify eligible people, coordinate with those on the ground, update our analysis, publish information, and figure out what could still be done before the last aircraft left. The work that had begun months earlier now had a deadline measured in days.
I was exhausted, frustrated, and increasingly afraid of what would happen to those we could not reach. The scale of the need far exceeded our capacity to respond. There were more people than seats, more people than time, and more problems than anyone could solve in those final weeks.
What made it hardest was that there were no good options, only decisions made with incomplete information under extreme time pressure. Every delay mattered. Every missing document mattered. Every departing aircraft meant an opportunity that would not return.
Then, on August 26, the airport was attacked. American service members and Afghan civilians were killed. The evacuation’s end no longer felt theoretical. It felt imminent.
Five days later, on August 30, the last American military aircraft left Afghanistan.
What I remember most is what came after.
I didn’t feel relief or victory. I felt a profound sense of unfinished business. The planes were gone, the news cycle moved on, and the political debate began. But the situation wasn’t over for me, or for the people we had spent months trying to help. People who had supported the United States were still there. They had been promised we would not forget them, and now there was no airport left to reach.
I kept thinking about what I had said in May.
I had been right, but there was no satisfaction in it. I wished I had been wrong.
The work didn’t stop when the planes left. If anything, the nature of the work changed again. The evacuation had become history almost immediately, but for the people who remained in Afghanistan, it was the beginning of a very different reality.
Through AWA and the Evacuate Our Allies coalition, we continued documenting what happened to those who remained. We surveyed more than 20,000 Afghans and worked to understand what their lives looked like after the withdrawal. The findings were devastating. People lost jobs and income. Women lost freedoms they had spent twenty years building. Some were imprisoned, tortured, or killed.
But what stayed with me was not the data. It was the people behind it: people who had trusted the United States, taken risks because they believed in what we said, and been told they would not be forgotten.
Five years later, I still think about Afghanistan. I think about how fast everything changed, how much warning we had, how hard Matt, Kim, and everyone at AWA and the Evacuate Our Allies coalition worked during those final months and weeks, and how many we still could not reach. I don’t tell this story because I was a hero. I wasn’t. I was one person trying to understand a crisis as it unfolded and doing what I could with the time and tools available.
But I also don’t want the work of that summer to disappear into the history of the withdrawal. We spent months trying to quantify what was coming. We published the analysis. We counted the planes. We tracked the evacuation. We documented what was happening as the situation changed by the day. We tried to make the numbers impossible to ignore.
History matters because memory fades, and when memory fades, it becomes easier to forget what people knew, when they knew it, and what they tried to do about it.
What stays with me most is this: in May, there was still time. The collapse had not happened. The airport was still open. The planes were still there. There was still a chance to prepare.
And then August arrived.
Michael Trudeau
Data Science Principal
Association of Wartime Allies