Remarks Prepared for Delivery by Special Inspector General John F. Sopko at Weill Cornell Medical College, New York, NY

“Afghanistan Reconstruction: Fact vs. Fantasy”

Good evening. I am John Sopko, and I am delighted to be here with you. I thank Weill Cornell Medical College for the invitation, and especially the kind introduction. Obviously, the thought was that with your expertise in treating intractable diseases, you would also be interested in Afghanistan reconstruction.

As Special Inspector General for Afghanistan Reconstruction-SIGAR for short-my business is checking alleged facts, digging out actual facts, and identifying weaknesses, failures, uncertainties, corrupt acts, and delusions. I hope that makes me the bureaucratic counterpart of a careful diagnostician, and not a low-budget version of Doctor Oz.

Sorting fact from fantasy in Afghanistan can be a challenge. But just as doctors must be willing to face the truth about whether a treatment is working, we in the United States must be willing to face the truth, and acknowledge the uncertainties, about our programs and policies if we want reconstruction to succeed. And it's important that we do succeed. The United States has committed more than $100 billion to rebuilding and developing Afghanistan. We have spent even larger sums on U.S. military operations there, and more than 2,000 U.S. service personnel have died there. Administrations of both major parties have determined that we have vital national interests in promoting stability in Afghanistan and preventing it from again serving as a platform for terror attacks on the United States.

No one in New York City, of all places, needs to be reminded that the United States intervened in Afghanistan in late 2001, rapidly overthrowing the Taliban regime that had sheltered, and refused to hand over, the mastermind of the 9/11 attacks, the late Osama bin Laden. By 2002, the U.S. program of regime change had morphed into one of reconstruction and nation-building.

Reconstruction activities have touched nearly every aspect of Afghan life. The United States and other international donors have funded activities that include standing up the Afghan army and police, digging wells and building roads, constructing schools and clinics, improving health care, creating a civil-aviation system, and developing alternatives to poppy cultivation as a way for farmers to earn a living. As you can see, it is a huge and far-reaching undertaking.

It is also very expensive. Since fiscal year 2002, Congress has appropriated almost $110 billion for Afghanistan reconstruction. And that's not counting the much greater costs of U.S. military operations in Afghanistan which have been estimated as nearly $700 billion through FY 2015. [1] Even now, 13 years on and with the U.S. military presence much diminished, nearly $15 billion more in reconstruction funds wait in the financial pipeline, appropriated but not yet spent. The United States and our allies have pledged billions more for years to come. To put things into some perspective, the $110 billion already appropriated for reconstructing Afghanistan, after adjusting for inflation, exceeds the value of the entire Marshall Plan effort to rebuild Western Europe after World War II.[2]

Unfortunately, from the outset to this very day, large amounts of taxpayer dollars have been lost to waste, fraud, and abuse. SIGAR professionals have documented details of U.S.-funded clinics that lack staff or medicines, schools that can collapse on their occupants because of shoddy construction, contracts that weren't performed properly or at all, aircraft that the Afghans can't fly or even maintain, troop rosters that can't be verified, cash assistance that can't be traced, and many other outrages. These disasters often occur when the U.S. officials who implement and oversee programs fail to distinguish fact from fantasy.

Apart from the ad hoc improvisations of early initiatives, problems include failure to define requirements and assess suitability, poor planning and execution, no arrangements for sustaining completed projects, lack of coordination within our own government and with other governments and organizations, theft and corruption by both American and Afghan military and civilians, collusion and bid-rigging on contracts, money laundering, and poor administration of U.S. funds provided directly to Afghan ministries. Note that I said, "problems include." The full problem list is much longer.

An overview of SIGAR

Congress eventually responded to these problems in 2008-six long years into the reconstruction effort-by creating SIGAR. Our authorizing statute [3] gave SIGAR a unique set of authorities. We are the only U.S. oversight agency focused solely on Afghanistan, and with the authority to audit, inspect, investigate, and otherwise examine any and all aspects of reconstruction, regardless of departmental ownership. We are not a component of any other federal entity, and unlike our oversight brethren, we are mandated to take a "whole of government" approach to reconstruction in Afghanistan, including U.S. coordination with foreign allies.

While independent, we are subject to Congressional scrutiny. Congressional dissatisfaction with SIGAR's early results led to my predecessor's resignation. President Obama appointed me to serve as Special IG, and I started work in July 2012. I've spent decades as a federal prosecutor, as a Capitol Hill staffer, and an attorney in a private firm. But for the breadth, challenge, and importance of the work, none of those billets quite compares to SIGAR.

To carry out its work, SIGAR employs about 200 auditors, investigators, inspectors, attorneys, technicians, writers, special-projects and lessons-learned researchers, and other staff. Most of these professionals work in our Arlington, Virginia, headquarters, but we also maintain 42 billets in Afghanistan. So far, we have published 27 quarterly reports to Congress plus hundreds of financial and performance audits, inspections, testimonies, and other reports.

As of our most recent tally[4], SIGAR has:

• Completed 198 audit and inspection reports.
• Made 540 audit and inspection recommendations, with over 70 percent adopted.
• Issued 102 special-projects products ranging from inquiry and alert letters to fact sheets.
• Identified hundreds of millions of dollars that could be or have been put to better use, such as
  • More than $240 million in questioned costs from financial audits,
  • Over $570 million recouped for the U.S. government through fines, restitutions, forfeitures, civil settlements, and cost savings.
• Finally, SIGAR's work has led to convicting 86 people for offenses; 64 have been sentenced, some for long prison terms.

We can point to other results. Just this year, the Defense Department started acting on one of my longstanding goals, broadly publicizing referrals for suspension and debarment to help prevent people and businesses with links to insurgents or terrorists from getting federal contracts in Afghanistan. Acting on results of a SIGAR investigation, Dr. Ashraf Ghani, the new president of Afghanistan, cancelled a nearly $1 billion fuel contract and fired people involved in a bid-rigging scheme that appears to have criminally inflated costs to U.S. taxpayers by more than $200 million. That's a U.S. concern because, apart from paying to recruit, arm, equip, train, house, and pay Afghan security forces, we also buy fuel for them.

A legacy of mixed results

Please don't conclude that the Afghan picture is uniformly bleak, a landscape of nothing but wreckage, ruin, and waste. The assistance from the United States and other international donors does more than cover some 60 percent of the Afghan national budget; it has brought about some genuine gains. But if we are to achieve more with what is already a diminishing amount of aid, we have to be honest with ourselves about what we have and haven't accomplished.

There are more schools and more children in them. Deaths from childbirth and infant mortality are down. People are living longer. Roads, clinics, power stations, and irrigation facilities have been built.

My recent meetings with Afghan President Ashraf Ghani and Chief Executive Officer Abdullah Abdullah confirmed for me that these intelligent leaders are aware of their country's weakness, and of the urgent need for reform to continue receiving international aid and to better the lives of their people. I am cautiously optimistic for their prospects of bringing about positive change.

Still, we must face the facts before us. Afghanistan remains one of the world's poorest, least developed, and most corrupt countries, with weak ministries in a central government whose presence and support in rural areas is often minimal or tenuous. Though total U.S. funding for the Afghan National Defense and Security Forces has topped $65 billion since FY 2002, the Taliban and other insurgents are still fighting and capable of attacks across Afghanistan. The question then, is what have we and the Afghans really gotten for our 14-year-long, $110 billion investment?

Billions in Counternarcotics Funding Has Had Little Result

Which brings me to one of our top Afghan-reconstruction fantasies or facts - depending who you work for. Although Afghanistan ranks very low in international comparisons of income, education, health, corruption, and other measures, it leads the world in at least one area: opium production.

The United Nations reports that Afghanistan produces 90 percent of the world's illicit opiates.[5] U.S. and United Nations estimates agree that Afghanistan has roughly 500,000 acres, or about 780 square miles, devoted to growing opium poppy. [6] That's equivalent to more than 400,000 U.S. football fields, including the end zones. This enormous acreage devoted to opium feeds a huge tragedy-fostering heroin addiction and crime around the world, including here-as well as a strategic threat. Taxes on opium are a major revenue source for Afghan insurgents, and a powerful prod to corruption among Afghan officials.

Here is where facts and fantasies as well as questions about metrics and indicators come into play. As of this March, the United States has provided $8.4 billion-I repeat, billion-for counternarcotics programs in Afghanistan.[7] DOD and State give information to SIGAR every quarter on their successes such as drug-treatment centers built, rehab workers trained, tractors donated (31 last year), alternative-livelihood programs executed, drug seizures, and so on. Yet, despite all this, we see record and rising levels of opium production.

The United States also highlights its poppy-eradication programs. Unfortunately, last year's successful eradication effort reduced that record-high opium-cultivated acreage by a bit over 1 percent.[8] Lists of counternarcotics programs, people trained, equipment donations and such, even if accurate, represent activity and outputs, not outcomes.

Like New Year's resolutions, such inputs and outputs means very little if they don't lead to actual positive outcomes. The bottom line-record opium cultivation and production-clearly shows we are not winning the war on drugs in Afghanistan. Of course, the U.S. government announced its own war on drugs in 1971, almost 45 years ago, and we haven't won that, either. Many of you here today are no doubt painfully familiar with the human consequences of those dual failures.

Even when data sounds cheering, verify before celebrating

I mention the counternarcotics effort in Afghanistan to emphasize that it is all too easy to focus on metrics and measurements without asking how salient they are, what they really signify, and what they suggest needs to be done, not done, or stopped. After all, if you're at a family reunion and someone runs in to report Uncle John's body temperature is 91 degrees Fahrenheit, what you should do next depends on context. If he's just been hauled out of a frigid pond, you wrap him in blankets. If he's dead and approaching room temperature, you send a cousin out for a bag of ice. Some measurements can be precise, accurate, and ambiguous, all at the same time.

SIGAR has seen too many reconstruction efforts in Afghanistan that entail dubious, unverifiable, misleadingly precise, or irrelevant data and metrics. Placing more credence and confidence in them than they deserve, or misreading their implications, creates an enormous risk of wasted work and money, and of failure to achieve our reconstruction objectives.

The late guru of total-quality management, Doctor W. Edwards Deming, warned about this tendency. He was writing about "deadly diseases [that] afflict most companies in the Western world," but the applicability is broader. One of the top five deadly diseases, Deming said, is "running a company on visible figures alone (counting the money). . . . the most important figures one needs for management are unknown or unknowable." [9] His point was that some empirical measures may mislead, and some vital elements of success, like leadership, morale, and pride, may be difficult or impossible to measure.

Like medical diagnosticians, forensic experts, and epidemiologists, inspectors general need to apply professional skepticism and critical thinking to the data before us, lest we jump to conclusions and abandon ourselves to fantasy. We must keep in mind the wise caution in Gilbert and Sullivan's jolly 1878 operetta, HMS Pinafore:

Things are seldom what they seem,
Skim milk masquerades as cream.

In light of that, let us start with a very basic question of health and well-being that should be of particular interest to you here at Weill Cornell: Do Afghans live as long as recent reports claim they do?

How much has Afghan life expectancy changed?

In 1980, just after the Soviet invasion, Afghanistan had one of the world's lowest life expectancies at birth, 41 years, according to the World Bank. [10] (Cambodia, then in a period of brutal repression, was lowest at 30 years.)

Even today, Afghanistan is mostly rural, largely illiterate, very poor, and scantily supplied with public goods, so it is no surprise that both basic health care and basic health information are in short supply for millions of its people. The U.S. has tried to improve the public health situation, training thousands of midwives, building clinics and hospitals, and a modern computer-savvy Ministry of Public Health. USAID support for these programs totaled about $1 billion as of last December, with nearly $400 million more to come through FY 2018.[11] So there has been progress in reducing birthing deaths and childhood mortality, as well as in treating causes of adult mortality. But how much?

A USAID fact sheet from last year says of Afghanistan:

Life expectancy has increased from 42 years in 2002 to over 62; maternal mortality rates have declined by 80 percent and child mortality rates by almost 50 percent.[12]

If this is true, this may be one of the biggest development successes in history. But opinions differ on the life-expectancy claim. The current Central Intelligence Agency World Factbook gives a 2014 estimate that average Afghan life expectancy at birth is 50.49 years, not 62. [13] That number, if accurate, still represents a respectable improvement from the ghastly level of the past, but is also significantly different from the number cited by USAID-and from the World Bank's 2013 estimate of 61 years. [14] The UN Population Division, the U.S. Census Bureau, and the World Health Organization also offer lower estimates of Afghan life expectancy.[15]

What to believe? The more glowing estimates are drawn from the USAID-funded Afghanistan Mortality Survey, conducted in 2010 by the Afghan Ministry of Public Health. I note in passing that a 2013 SIGAR audit of USAID's support for the ministry found, among other things, that USAID had not verified what, if anything, the ministry had done to address deficiencies in its internal audit, budget, accounting, and procurement functions. [16] These concerns raise questions about the ministry's ability to provide solid, verifiable data on its operations.

Nonetheless, the mortality survey was a major effort, and I am not suggesting that its organizers didn't do their best. But conducting a comprehensive public-health survey in a war zone is bound to entail some uncertainty. A National Public Radio report in January 2012 put the question plainly: "Gains in Afghan Health: Too Good to be True?"

The NPR story quoted Dr. Kenneth Hill, a demographer at Harvard University and a technical advisor for the survey, saying "There are still huge question marks hanging over the estimates, [so] I'm not sure there is enormous value in the data." The story noted that some parts of the country were off-limits because of the security situation, while cultural taboos about discussing family health could lead to underreporting maternal and infant deaths. In addition, the reported gain in life expectancy was so large, NPR said, that "It's unlikely that both the old life expectancy of 42 years and the new estimate of 62 years are correct."[17]

Another view on Afghan health comes from that fine organization, Medicins Sans Frontieres, or MSF. They have been working in Afghanistan more than 30 years, and have more than 1,600 staff there, mostly Afghans. In February 2014, MSF released a study drawing on six months of research among hundreds of ordinary Afghans using health-care facilities.

The MSF research was motivated, they explain, by "alluring narratives of success-crafted to suit political and military agenda" and "overly optimistic rhetoric" that benefits from a dearth of reliable statistics.^[18] The MSF report notes that health statistics in Afghanistan are "notoriously unreliable" for reasons including lack of monitoring and lack of data gathering in some remote and high-risk areas. MSF researchers found that cost, distance, and danger prevent many Afghans from seeking health care, and that 79 percent of their sample had bypassed local clinics because of perceived problems with staff availability or skill, or the quality of services. ^[19]

The MSF findings suggest that much of the official happy talk on health should be taken with a grain of salt-iodized, of course, to prevent informational goiter. Accordingly, SIGAR remains skeptical about what, if anything, our programs contributed to whatever improvements have occurred there, since USAID, the Embassy, and other development contractors appear unable to explain the important logical connection between inputs, outputs and outcomes. Their inability is best illustrated in my request two years ago to USAID, State and DoD to provide me with their top ten successes in Afghanistan and their explanation on how they came to that conclusion. I am still waiting for their list. One agency official even said it was like asking him to rank his children. But children are not-unless you count college tuition-multi-billion-dollar projects. You can't rationally decide how to start, modify, or stop projects, or how to reward or hold accountable their managers, unless you compare them.

I am happy to note that, with support from USAID, the Ministry of Public Health is launching a new, $5.6 million "gold-standard" Demographic and Health Survey across all 34 Afghan provinces, to be completed in September 2018. Until those results are compiled, vetted, and statistically adjusted, SIGAR remains politely skeptical that there is a truly solid basis for trumpeting gains and making plans that presume their accuracy.

What goes on at Afghanistan's schools?

Doubtful data also complicates assessments of education in Afghanistan. In a mostly illiterate country where girls have often been completely excluded from schools, education is critical to promote quality of life, equality of opportunity, and economic growth. As of this March, USAID has disbursed more than $768 million for education programs in Afghanistan.

The Afghan Ministry of Education reported in 2014 that the country had 8.35 million students enrolled at 14,226 general-education schools. As of the prior year, 6.6 million students were listed as "present," while 1.55 million were considered absent.[20]

The ministry maintains statistics in its electronic education-information system. Unfortunately, the ministry's system doesn't track whether schools are open or closed at any particular time, nor whether they've been subject to security threats. The student numbers are also less than they might appear to be. For one thing, they are not independently verified. For another, as SIGAR reported last year, the ministry counts absent students as "enrolled" for up to three years before dropping them from the rolls. That's right: a student who has not attended school in nearly three years is still considered as "enrolled." That's like saying a spouse who packed up and left three years ago is still committed to you.

The official rationale was that absent students might decide to return. Maybe, but as SIGAR has noted, the fact that some program funding levels are keyed to enrollments also creates an incentive to maximize reported enrollments.[21] To cite just one example of the gulf between official indicators and grubby reality, the Afghan Analysts Network sent observers to schools in Ghor Province, where 13 teachers and 767 students were supposedly engaged in education. They found 5 teachers and some 20 students. The AAN researchers also were told of fake schools for girls, teachers' salaries being stolen by warlords, security threats causing interruptions in schooling, and parents sending children to religion-centered madrasa schools rather than public facilities.[22] To take our concern to a macro level, I can report that a ranking USAID official in Afghanistan has told us that the number of students actually attending Afghan schools may be on the order of four million, not the eight million widely reported as enrolled. This lower number has been confirmed by a number of Afghan civil society organizations also.

More important from the viewpoint of Afghan children, the country's future, and U.S. taxpayers, there is very little visibility into what is actually being taught at Afghan schools. Most Afghan students, even if accurately counted and actually at school, are in rural areas. Security is often a problem, and the numbers and mobility of oversight personnel continue to shrink. We have, therefore, very little insight into whether teachers are even literate themselves, present and are using the official curriculum and texts, or whether they or insurgent influence are injecting extremist precepts into the minds of Afghanistan's future adult citizens.

For all we know, millions of kids could still be absorbing lessons on extremist violence, intolerance, and gender inequality. Once again, unverified data about inputs and activities cannot give us a reliable read on an important and much-desired outcome. As sociologist William Campbell wrote many years ago, "Not everything that can be counted counts, and not everything that counts can be counted."[23]

How many Afghan security personnel are there?

For my last example of the difficulties and importance of extracting accurate, verifiable, and pertinent information from our commitment to Afghan reconstruction, I turn to what sounds like a simple question: How many people serve in the Afghan National Defense and Security Forces, the ANDSF?

As of February 2015, the U.S.-military-led Combined Security Transition Command-Afghanistan (mercifully known as CSTC-A) reports ANDSF personnel actually assigned to duty numbered 328,805, roughly split between the army and the national police.[24]

That is a very exact number. But even if it were accurate at some point in February, the half-life of its accuracy was probably a matter of hours. People get killed. Recruits sign up. Soldiers and police desert. Enlistments end. It's the same reason the U.S. Census will never get a precise count of the U.S. population. But that's a trivially true observation.

Unfortunately, there are more basic problems with the reliability of ANDSF numbers. Let me give you the highlights.

Earlier this year, SIGAR auditors made unannounced visits to the headquarters of two corps of the Afghan National Army and to an Afghan Air Force wing. Our auditors asked for identification from the 134 personnel present for duty. Thirty-one of those people, or 23 percent, could not be verified from Afghan personnel data.[25] Some of the people had no ID, official or otherwise. And those are results from two corps headquarters and an air wing based in Kabul. Results from lower-echelon or more remote locations might be worse. But with oversight travel increasingly constrained, we may have no good way to find out. So it is no surprise that our auditors' key conclusion from their work on the Afghan army is that, "Despite 13 years and billions of dollars in salary assistance to the Afghan government for the ANA, there is still no assurance that personnel and payroll data are accurate,"[26]

Unfortunately, that finding in April echoed the findings of another SIGAR audit earlier this year of the other half of Afghanistan's national-security apparatus, the Afghan National Police. In that audit of ANP personnel and payroll information issued in January, SIGAR found that while the United States has provided more than $16 billion for the national police since 2002, "There is still no assurance that personnel and payroll data are accurate." [27]

Problems in obtaining reliable numbers from the ANDSF stem from a variety of weaknesses. Our auditors found that daily roster checks are often unsupervised, and some rosters have illiterate soldiers' X marks apparently entered by a single hand. Neither Afghan nor U.S. officials have written procedures for verifying data accuracy. The Afghan reporting system relies heavily on paper records, manual entry, and hand delivery of reports. The Ministry of Defense data system requires manual keyboarding, isn't linked to other systems, and can't generate reports on individual soldiers' pay.

Among other problems, the system creates opportunities for unit commanders to include dead or deserted soldiers on the rolls to collect more cash. On a larger scale, dubious and unverifiable numbers reporting makes it difficult for Afghan military leaders and U.S. advisors to obtain a clear picture of the strength and capabilities of the security forces. And because the United States will be supporting a nominally 352,000-strong ANDSF through at least 2017, dubious numbers entail the risk that U.S. taxpayers could be paying large sums for security that will never be real.

Meanwhile, the ghosts of Dr. Deming and Professor Campbell, remind us that even accurate data can't clinically capture things like leadership, motivation, morale, and unit cohesion that are essential elements for an army or a police force. And last year's evaporation of two entire Iraqi army divisions in the face of 800 to 1,000 ISIS insurgents at Mosul brings home the importance of not only knowing how many soldiers are present, but how capable and willing they are to fight.

Conclusion

For both humanitarian and national-security reasons, the U.S. mission to reconstruct Afghanistan is critical. And with $15 billion currently awaiting disbursement, with billions more to follow, there is both need to improve the effectiveness of the effort, and time to make a difference in the outcome.

We must not kid ourselves about Afghanistan. It will be a long struggle. Defeating a determined insurgency, improving health and education, altering attitudes toward women, reducing corruption, and building governmental competence are not casual, short-term undertakings. Impatience driven by temperament, election cycles, or fiscal-year budgeting can only impede progress.

We can also safely say that the struggle in Afghanistan won't be shortened, much less won, by official happy talk and cheerleader-style press releases. Improving the likelihood of mission success requires, as a start, accurate, verifiable, and pertinent data-accompanied by a recognition that some key indicators require subjective evaluation by experienced and independent observers in the field. Let me emphasize the independence issue. Poor data and disregard of nonquantitative assessments that is biased by self-interest and turf protection can only lead to unrealistic judgments, unjustified hopes, and outright fantasies with no link to reality.

The 19th-century American humorist Josh Billings said, in the manner of Socrates, "I honestly believe it is better to know nothing than to know what ain't so."[28] SIGAR's clinical examination of American reconstruction operations in Afghanistan has persuaded me that we know a lot more than nothing, but a lot less than we think. Budgeting, planning, oversight, course corrections, and decisions to adjust the targets, duration, and intensity of U.S. efforts there all require reliable information. At the moment, that is all too often a scarce commodity and, accordingly, our programs as we go forward may be based more upon fantasy than reality.

SIGAR will press on in the years ahead to carry out its assignment of pinning down facts; calling out fluff, felonies, and fantasies; and recommending improvements. We welcome your interest and support in that mission, as I welcome your comments and questions. Thank you.