Billions of U.S. tax dollars are spent overseas every year by the U.S. Agency for International Development, but many projects never achieve their multimillion-dollar goals, despite the agency's very public emphasis on rigorous project evaluation.

Reports by the agency's office of the inspector general watchdog show a pattern of repeated problems that lessen projects' impact and waste taxpayer dollars.

More than half the projects the DIG has reviewed since March 2013 had such significant problems that they failed in at least one major objective. Six more had problems serious or numerous enough to limit their impact. Only two of the 19 projects achieved their goals without serious snags.

Delays, sustainability, dangerous security situations, inexperienced staff, difficulty hiring people and a lack of accurate data hindered most of the floundering projects, according to the DIG reports.

Some of these problems, like instability, are inherent to international development. Development work is particularly vulnerable in "frontline states" like Afghanistan, Pakistan and Iraq, said Mark Lotwis, vice president of Public Policy for InterAction, a network of nongovernmental international organizations.

"There are a lot of problems with delivery of programs, there's no doubt," Lotwis told the Washington Examiner.

Another aspect of the problem is the large amount of money being spent quickly in unstable conditions, he said. In frontline states, part of USAID's development funding comes from overseas contingency operations funds allocated by Congress for wartime spending.

Because the money has to be returned to the Treasury if it isn't spent within a certain time frame, it can create a "use it or lose it" mentality and lead to hasty awards, Lotwis said.

"When there's tremendous amount of money being processed not in normal ways ... there are going to be problems," he said.

USAID's partial success rate is to be expected in volatile areas, said Brett Shaefer, who follows international regulatory issues for the Heritage Foundation. "You do see an unreliable success rate," he said. "It's a coin flip."

No amount of money can overcome a poor development environment, he said.

But instability and funding are only part of the story. In 2012, only 22 percent of USAID's funding in frontline states came from overseas contingency funds, and problems with project effectiveness weren't limited to those countries.

Ssuccess was also limited by poor planning, lax oversight by USAID, unreliable data and poor coordination, according to IG reports.

A $75 million grant to improve education in Lebanon by renovating schools and training teachers, for example, suffered from poor oversight and planning. Two years into the five-year project, 45 percent of the funds allotted to renovating schools had been spent on 13 percent of the designated schools, and 44 percent of teacher training funds had been spent on three percent of the intended teachers.

Another example of poor oversight was a project in El Salvador intended to improve the local business environment. But USAID didn't conduct required site visits, didn't establish indicators to measure progress, and didn't use accurate data to measure jobs created. As a result, the IG said it was impossible to determine if the project was successful.

Another government watchdog, the Special Inspector General for Afghan Reconstruction, recently described some of USAID's measurements for success as "ambiguous" and "hazy."

In one instance cited by SIGAR, USAID claimed a program in Afghanistan improved access to health care facilities, but the data couldn't prove a relationship between the program and access because Afghanistan and the numbers the agency gave weren't even specific to the regions they worked in.

"The causal haziness around the edges of this indicator suggests that careful attention to selection, logic, and measurement protocol is warranted," Inspector General John Sopko wrote.

SIGAR cited several other examples where USAID claimed success without being able to prove a direct link between its work and the improvements.

In 2011 the agency trumpeted a new approach to evaluating projects more aggressively, and publishes third-party reports in a special section of its website.

Its renewed emphasis on evaluation doesn't seem to be improving USAID's success rate, though. In a 2008 report to Congress, the agency said 40 percent of the projects in its report met their goals; 21 percent improved conditions but didn't meet goals; and 37 percent of projects failed to do either.