Marceline Bauma, 8, center, copies her homework assignment at the Notre Dame du Congo school in Goma, Congo, in 2012. (Kem Knapp Sawyer/The Pulitzer Center)

Back in 2000, 164 governments met in Dakar, Senegal, to launch the ambitious United Nations goal of “Education for All” (EFA) — a commitment to provide quality basic education for all children and adults by 2015. “For all” and “by 2015” leaves little space for nuance: In 2016, no one should be left behind.

Agreeing on such a goal is one thing, realizing it quite another. So, 15 years after Dakar, one in six children has not completed primary school in low- and middle-income countries. Moreover, the quality of the education provided continues to be a challenge.

The education gap is particularly wide in fragile states — the low-income countries where the state is weak or unstable. We recently investigated progress in education in the Democratic Republic of Congo (DRC). Here’s what we found.

True, the DRC is an extreme case, where one-third of all children haven’t completed primary school. Yet the situation in the DRC illustrates the “last-mile” problems that haunt projects like EFA in fragile states. For now, EFA’s promises remain unmet in DRC, where the state cannot substitute for parents unable to pay for school.

The challenge of education in the DRC

Low-income countries are doubly challenged on the education front relative to high-income countries. On the one hand, their populations are relatively young and growing at 2 to 3 percent per year, which means they need to make a greater investment to achieve education for all. On the other hand, low-income countries have a lower capacity to tax. In the case of the DRC, total government expenses are below 15 percent of the country’s total income (GDP). While the education budget weighs quite high in government expenditures, as a percentage of GDP the DRC scores a world-record low of 2 percent of GDP in public education expenditures.

Despite these structural impediments, the DRC declared an end to all school fees effective as of the beginning of the new school year in September 2010. But there was no real increase in the government education budget. After an initial period of chaos and ambiguity, schools quickly resumed asking parents to pay school fees. What else could they do, other than simply close their doors? Recent estimates calculate that parents still finance 75 percent of primary education expenses in the DRC.

Fragile states like the DRC are not just operating in a low-income context; above all they lack the basic capacity to reach out to the whole population. States do not simply disappear, even when they are fragile or failing: Their basic infrastructure persists, and sometimes local and/or non-state actors take over. However, what remains of the state is spread quite unevenly over the state territory and affects different social groups in unequal ways.

A detailed look at the flow of salary expenses to teachers in the DRC indicates that large swaths of territory have to do with little or no education infrastructure. The state remains heavily involved in education in the capital city of Kinshasa, where 4 percent of the country’s primary school pupils live. Yet Kinshasa accounts for a disproportionate 22 percent of the budget for teachers.

In contrast, our research indicates that the Kivu provinces in the east have 15 percent of the country’s primary-level pupils, but receive only 7 percent of the teachers’ budget. In fact, less than half of all teachers are on the official payroll in the Kivu provinces. The rest are paid through parents’ contributions and do not “exist” officially, according to the Ministry of Education. And 20 percent of the schools in the Kivus are not actually registered, but exist outside of the state administration radar.

If the state had the means to pay for education, a fee-free policy would be a wonderful policy in Kinshasa, where 90 percent of school teachers are already on the state’s payroll — and parents’ contributions to pay for teachers’ top-ups could be substituted by increasing the teachers’ state salary at the stroke of a (Minister’s) pen.

It’s a different story in the Kivus, where abolishing school fees theoretically means that half the teaching staff, those teachers officially unregistered, and/or working in unregistered schools, would have to be laid off. An added problem is that in the absence of school fees, more parents are likely to send children to school, so the demand for teachers and supplies would be higher.

Here’s what actually happened

What happened in reality, in the Kivus as elsewhere, was that schools arranged with local authorities to circumvent the instructions to abolish school fees. The state proved largely incapable of enforcing its fee-free school policy against all other local interests, as schools and teachers largely survived on parent contributions.

It is an unfortunate coincidence that the provinces with less actual state presence (and state funds) also tend to be the provinces with the lowest school enrollment rates. The areas where the state education infrastructure is loosest are also the areas with the highest illiteracy rates. So even if the government could properly finance its fee-free policy, the program would likely have little impact in precisely those areas where the need for such a policy would be highest.

A final point is that the Education for All mission is about ensuring a quality basic education for all. It is one thing to get children to school, and quite another to ensure quality education. The quality of education may well be the last foot of the last mile.

In the DRC, school enrollment increased quite substantially in the past 16 years, with an 8 percent yearly bump in the number of pupils, far beyond the 3 percent rate of demographic growth. An estimated 57 percent of 15- to 24-year-olds have completed primary school in the DRC, which is better than the average for sub-Saharan countries (49 percent). At the same time, however, we recently reported that only 47 percent of this age group is able to read one complete sentence in the DRC, though the average is 63 percent for sub-Saharan Africa as a whole. The DRC may well have traveled too fast rather than too slow on the path toward education for all — and the quality of education may have suffered.

Does this mean that Education for All is not worth the effort? Of course not, after all, the international community has both a moral duty and a real interest to commit to such a goal. But given the structural hurdles in reaching such a goal, EFA should serve as a useful guide for the direction we want education to head, rather than as a fixed deadline or development strategy.

Tom De Herdt is the chairman of the Institute of Development Policy and Management at the University of Antwerp. You can follow him @tomdeherdt.

Kristof Titeca is a lecturer at the Institute of Development Policy and Management at the University of Antwerp. You can follow him @KristofTiteca.