With continuing U.S. causalities—including the loss of the highest ranking officer killed in a combat zone since Vietnam —and the perceived inability to peacefully transition to a new president, many Americans are questioning what we have to show for 3,469 fatalities and more than USD 640 billion spent on Afghanistan since 2001.

The Department of State (DOS) and Department of Defense (DOD) can cite statistics to show that our efforts have not been in vain. In my own experience, though, the Justice Center in Parwan (JCIP) is a good example of how the U.S. should define “success” in Afghanistan.

The JCIP is a joint U.S.-Afghan project to establish Afghanistan’s first national security court. It was originally envisioned to be the largest court complex in Afghanistan, bringing together judges, prosecutors and defense counsel to work through a backlog of national security cases. And, from 2011-2012, I was a member of the U.S. Justice Advisor team tasked with training, assisting, and advising our Afghan partners.

When I joined the team, the JCIP was limited to a single building with a single panel of judges, hearing roughly 5-10 cases a month. During my 11 months, I saw the JCIP grow from one judicial panel to five, try over 300 cases, and eventually average 45-50 hearings a month, largely through the hard work and dedication of our Afghan partners.

However, even during this successful growth, there were problems that made the JCIP future unclear. We learned that certain Afghan colleagues were seeking transfers from the JCIP because U.S. involvement did not afford them an opportunity to accept bribes. Additionally, the defendants, accused of terrorist activities or membership, frequently threatened the court members and their families. These factors led to occasional low morale among our Afghan colleagues and tensions with the U.S. Justice Advisors. As I left in spring 2012, I was asked by a Prosecutor “What do you think you’ve accomplished here?” At that moment, I had no answer.

I have now returned to Afghanistan and I recently had the opportunity to visit the JCIP. During the visit, I found that the DOS had pulled their Justice Advisors a week before my arrival and the DOD was anticipating more limited support in 2015. I also learned that court hearings were taking place in a former police academy because the court house had never been completed.

I also found some of my former colleagues. I spoke with them and discovered that while a number of staff I once knew had left, there was a core group of judges and prosecutors who remained. While they said they will miss working closely with their U.S. partners, they were looking forward to continuing the JCIP mission. They explained that the court is now run entirely by Afghans and was as efficient as it had ever been. I know the current JCIP does not look like the end-state originally envisioned by the U.S., but I would say it is a success nonetheless: the JCIP will continue into 2015 led and run by Afghans.

This is how the U.S. should define “success” in Afghanistan. It is unlikely that all of our projects will endure or that all of our ambitions for democracy will be met. However, success in Afghanistan is not a binary outcome (success or failure) and it cannot be measured by U.S.-defined metrics.

The fact that eight million children now attend school, up from just one million a decade ago, and that one out of three is female, is a success even if there is still room for improvement. The fact that over seven million Afghans participated in the recent election is a success even if the outcome is still undecided because fourteen years ago elections were not even an option. And, the fact that JCIP prosecutors and judges are still willing to pursue the rule of law against accused Taliban and Al Qaeda members—even given the dangers—is success.

Ultimately, all of these improvements are cause for celebration. There will undoubtedly be setbacks but, we should not judge our time in Afghanistan in absolutes. Success is best defined as finding an Afghan solution to an Afghan problem.