World Development Report 2018: Learning to Realize Education’s Promise

Education systems are complex. Aligning an education system’s goals, financing, and incentives with student learning is difficult for technical reasons. But there are also political reasons systems do not prioritize student learning. Political impetus to fix misalignments can help achieve important educational objectives—as it has in Chile, England, and India (see chapter 11)—but unhealthy politics can make things worse. Too often, education interventions, whether big reforms or day-to-day implementation steps, are compromised because powerful individuals or groups can make others act in ways that serve private interests rather than the collective good. Powerful actors frequently benefit from the status quo and devise mechanisms to preserve it, regardless of the impact on system performance. These mechanisms result in actors being trapped in low-learning equilibriums.

Unhealthy politics can intensify misalignments in education systems

Many education systems encounter political impediments and rent-seeking, making alignment much harder to achieve. Consider these examples:

• Using computers to educate students requires difficult technical decisions on program design. But even when there is consensus on technical design, students may not benefit. For example, in 1996–97 the superintendent of New York City’s District 29 rigged a $6 million contract, awarding it to a computer company affiliated with a politically connected property developer. In return, the company gave the superintendent expensive gifts, while delivering archaic or nonfunctioning computers to students. Teachers had been counting on decent computers to help their students in math; without the computers, the students lost out.

• In 2009 Mexico’s federal government introduced a plan for competitive recruitment of teachers, whereby all candidates were required to take a test covering content knowledge, pedagogical mastery, and ethics. Designing the tests was technically difficult. But the technical challenges paled next to the political impediments created by local affiliates of Mexico’s teachers’ union, the Sindicato Nacional de Trabajadores de la Educación (SNTE), which has 1.4 million members. The policy change meant that the opportunity for patronage-driven hiring would vanish. Because of strong opposition from the SNTE, the reform was diluted, making it applicable only to a small pool of vacancies. Estimates suggest that up to 85 percent of hiring in 2010 was discretionary rather than competitive. Recent evidence indicates that the teachers hired through discretionary methods were much less effective at improving student learning than those hired competitively. • Vyapam (http://www.vyapam.nic.in) is the government-run professional examination board in Madhya Pradesh, India. It conducts large-scale entrance tests for admission into courses such as medicine and for recruitment into state government jobs such as the police. Designing entrance tests and ranking candidates are technically challenging when there can be more than 100,000 candidates. But political economy factors intrude as well: recently, rent-seeking is alleged to have undermined the goal of fair, transparent admissions. In 2013 an independent probe exposed a potential multibillion-dollar scheme in which senior politicians and government officials had allegedly set up a system allowing unqualified candidates to pay bribes, often to middlemen, to receive high rankings in entrance tests. In 2015, the Supreme Court of India transferred the case from the state government to the country’s premier investigative agency, the Central Bureau of Investigation, which is currently pursuing the investigation.

Education systems involve many stakeholders with multiple, often contradictory, interests. These systems are not just about students, teachers, or principals. They also involve politicians, bureaucrats, the judiciary, private players, and more. Participants linked to these institutions have a vested interest in how the system works, including its structure and funding. A textbook supplier may want to provide a quality product, but it also cares about profits. A politician may want to make teachers accountable for student learning, but also realizes the electoral risks of teacher opposition. A bureaucrat may support meritocratic admissions, but also accepts a “token of appreciation” for ensuring the admission of an acquaintance’s child to a desirable school. A parent may want to complain about a teacher, but worries that her child could suffer retaliation.

Vested interests are not confined to private or rent-seeking interests. Actors in education systems are often driven by their values or ideology, especially when the consequences of education policies are not readily apparent. Examples include a commitment to public schools versus public-private choice, secular education versus religious, and accountability for test scores versus a focus on teacher qualifications. In addition, education systems can be used by dominant ethnic groups—especially in multilingual or multi-religious societies—to promote their positions while suppressing minorities.

Multiple interests jeopardize learning goals. Balancing multiple interests is difficult. When education interventions threaten interests, whether they be a person’s financial, ideological, or status-related interests, resistance from different parts of the system can be expected. The net effect: the system is pulled away from a focus on learning (figure 10.1).

Education systems are vulnerable to political interference because they are opaque and because teachers constitute a large base of government employees. The opacity of education systems, coupled with uncertainty about how a specific education policy will affect learning, is fertile ground for contestation of reforms. Teachers—the most important factor in learning— have traditionally been important grassroots political actors, because of their geographic spread and regular interaction with parents. Two characteristics make teachers especially attractive as patronage appointees. First, entry costs to the profession are often low. Second, the impact of incompetent patronage appointees on learning is not immediately visible, so it has few reputational consequences for politicians, especially if they are already operating on a short time horizon.

Because of the size of the teaching force, teachers’ unions can be politically important. The political power of a union depends on how effectively its leadership can mobilize teachers, which varies widely within and across countries. In many countries, not all teachers are union members or engaged in union activity. Whether union activity helps or hinders education reform ultimately depends on several factors (box 10.1).

Multiple actors and interests: Pulling the system out of alignment at each step of the policy cycle

Personal interests influence reform at every step. Vested interests—of teachers, principals, bureaucrats, politicians, parents, students, the judiciary, civil society organizations, the private sector—are influential at every step of the education policy cycle. Broadly, these steps are setting policy goals, designing policies, implementing policies, evaluating policies, and sustaining policy reforms. The forces that detract from alignment tend to be magnified in conflict settings (box 10.2).

In many cases, policies are not chosen for their effectiveness in improving learning. Often, they are guided instead by the vested interests of powerful actors. Policies to hire teachers tend to be popular with politicians, teachers, and parents because they bring visible, immediate benefits. Likewise, large-performance appraisal, accompanied by tools to help scale school construction programs tend to attract teachers improve. Yet, because such reforms could considerable support. In a diverse range of countries expose poorly performing teachers, the reforms rarely (Cambodia, Colombia, Mozambique), policy makers reach the policy arena. Other policies that threaten have invested in building preschools instead of in teaching jobs include school consolidation or closing.

less visible but more effective process-oriented early

Such policies have been difficult to implement, because childhood initiatives, such as programs to improve parental support for local schools makes it politically parent-child interactions. In Bangladesh, until recently infeasible to close small, high-cost rural schools. In it was much easier to unite elites around the need for

Bulgaria, school principals have been reluctant to let mass education than around raising educational stanteachers go, despite declining school-age populations.

dards in schools.

In several countries, strong teachers’ unions have pre-It is also difficult to adopt a policy goal that threatvented large-scale teacher redundancies.

ens or reconfigures jobs, as is true for most quality-enhancing education policies. For example, an alternative to the politically popular policy of reducing

Even when the goal of a policy is to improve stuclass size would be to introduce serious teacher dent learning, its final design often reflects what powerful interests want, which can undermine the goal. Decentralization policies aim to increase policy responsiveness and accountability, but many times they delegate accountability for results without the authority or resources to achieve them. In Indonesia, Pakistan, and some Latin American countries, major decentralization efforts have struggled (at least initially) to find the right balance between central and local funding, or between central and local authority. Central authorities often attempt to limit the power of lower units of government because local governments—being closer to the people—can threaten the political power of more distant governments. At the same time, local governments may be unwilling to assume greater responsibility or adopt national norms—for example, on the inclusion of marginalized groups.

Policy makers may face little resistance when signing offon a policy, but implementation can be compromised if the policy threatens powerful interests. Policies designed to measure teacher performance have been particularly difficult to implement. In 2000 South Africa’s (then) Department of Education introduced the National Policy on Whole-School

Evaluation to establish standardized procedures for monitoring school performance and establishing the support needs of schools. Though the policy was meant to be supportive, premised on building collaborations and mentorship, the South African Democratic Teachers’ Union—the country’s largest— remained opposed to it. Many of the union’s chapters blocked the adoption of the policy in schools in their areas. A similar situation occurred in Mexico in 2012.

Well-intentioned reforms may threaten the legal entitlements of individuals—and when, understandably, they turn to the courts for redress, reforms risk being stalled. In Peru, unions resisted a new law on teacher evaluations by challenging its constitutionality. The ensuing court process then delayed implementation of the first round of evaluations. Though the court eventually upheld the law, for political reasons the union was given a major concession: the law applied only to newly hired teachers. Similarly, in 2002 teachers in Andhra Pradesh, India, stalled implementation of a policy on teacher transfers by filing a court case.

Parents can also make it difficult to implement learning-focused policies. A common example is parents helping children to cheat on examinations, which makes it hard to measure student learning. In 2015 the global media broadcast images of family members in Bihar, India, handing cheat sheets to children inside a building taking exams. Perhaps parents are aware that their children have not learned much in school, leaving them uncompetitive against better-prepared or more affluent children.

Indicators of the effectiveness of policies are often chosen in a way that lets powerful groups offthe hook. When a policy fails, frontline bureaucrats or principals may face repercussions regardless of whether failure was in their control. As a result, decisions on what to measure and track are less a reflection of what the education system values than of who is willing to be held accountable for what. For example, India’s landmark Right to Education Act (Act No. 35, 2009) did not originally contain any measure of teacher effectiveness or of student learning (although subsequent rules and amendments have sought to introduce the quality dimension). Similarly, accreditation systems in higher education tend to focus on inputs—such as number of classrooms, amount of equipment, or faculty-student ratios—instead of what students have learned or whether they become employed.Such an approach limits liability, but jeopardizes learning goals.

Data can be manipulated. Even when indicators track meaningful variables, data quality may be compromised. Data on outcomes can be gamed; decisions on who collects data and how often are made using subjective criteria. Gaming might take the form of candidates hiring test takers, parents facilitating cheating, teachers misreporting student test scores, or government officials encouraging teachers to modify test scores. In several countries, comparisons of national enrollment data with household survey data find systematic discrepancies, with official statistics sometimes exaggerating progress.

A subtler barrier to effective monitoring and evaluation is when governments collect mountains of data but not in a format that facilitates decision making. In some countries, the many efforts to collect data on indicators create the illusion that policy makers are actively engaged in data-driven decision making to improve school quality. But by the time data entry is completed, it is time for the next round of data collection. No serious analysis is conducted, feedback is not provided to schools, or the data are too broad to be useful. Such instances devalue data in decision making.

Even when difficult reforms are implemented, they can be undone. Reversal can be incremental, with policy makers softening elements to appease specific groups. In the late 1990s, the government of Madhya Pradesh, India, began hiring teachers from the newly created shiksha karmi cadre, under which all new teachers were to be locally recruited and put on 10-month contracts. In response, teacher applicants filed court cases arguing that the policy violated their constitutional rights, which emphasize that no citizen can be ineligible for office based on criteria such as place of birth. Burdened with litigation and pressure, the government redesigned the policy, making concessions on local recruitment and qualifications. Similarly, in São Paulo, Brazil, reforms of teacher career tracks introduced in 2009 were gradually undone by 2011 under a new education minister.

Reversal can be sudden. In Ghana, an early childhood care and development body was set up under the office of the president, with high-level support. But a change in administration put the office under the Ministry of Gender, Children, and Social Protection, lowering the priority and visibility given to early childhood issues. In República Bolivariana de Venezuela, decades of reforms that had created a strong higher education system were reversed when a new populist government set the goal of universalizing higher education. That effort, without prepared students, adequate faculty, or the appropriate infrastructure, has weakened the country’s education system.

These cases raise two important issues. First, why do parents and students have such a limited voice in influencing the vested interests that jeopardize quality-enhancing reforms? It could be that those most likely to benefit from reforms—especially parents and students—are often poorly organized. Moreover, the immediate gains of any proposed policy tend to be uncertain, making it harder to mobilize support for the reform. Parents may also find that the potential ramifications of opposing a teacher or politician could be formidable for their children. By contrast, those who stand to lose from reforms tend to be better aware of their losses and, in many cases, are better organized for collective action.

Second, more generally, why do these low-learning equlibriums persist? For every teacher, bureaucrat, politician, judge, or businessperson who jeopardizes learning, there are several who feel deeply accountable for student learning and act to strengthen education systems. Yet individual actors find it hard to escape these traps. Why?

Trapped in low-accountability, low-learning equilibriums

The formal rules of the game—that is, the laws and policies governing education systems—already reflect power asymmetries. When specific policy goals are chosen, when finance is allocated to certain tasks, when teachers’ unions bargain for concessions, preexisting power asymmetries and struggles are expressed through policy.

But such decisions also reveal the informal contracts that determine which formal rules are chosen or followed. Unwritten codes of conduct derive from the values, expectations, and cultural norms in a social setting, and they are important in determining the extent, nature, and strength of politics in that setting. In Indonesia, where older colleagues are treated with considerable courtesy, school mergers have often been delayed informally until principals who stood to lose their jobs retired. In rural Rajasthan, India, field research finds that teachers often have to pay bribes to get needed services, such as a transfer. Interestingly, the norm differs by gender: male teachers make the payments directly, while female teachers typically go through a male relative.

The widespread operation of informal networks reveals a lack of generalized trust within systems. Unwritten codes of conduct between individuals can thrive only if there is sufficient trust between them. Each must trust that the other will behave as expected. Yet as individuals cultivate personalized trust-based relationships—often undermining learning or equity goals in the process—overall trust in the system suffers.

As systems grow more complex and the number of actors and interactions increases, uncertainty multiplies. Trusting others becomes increasingly difficult. Creating reciprocal obligations helps manage the uncertainty. These obligations do not need to be spelled out; the social setting ensures they are understood. During the Suharto era in Indonesia, teachers were required to display “mono-loyalty” to the state and teach compulsory courses in the state ideology, Pancasila. If they did not, they knew they risked demotion or transfer to schools in undesirable areas. In SNTE-dominant parts of Mexico, teachers knew that if they did not support the SNTE, they risked unfavorable transfers or being sidelined.

Reciprocal obligations complicate accountability. Power relations between entities and groups depend on context. One group may be more dependent on another—and therefore less powerful—in one context.

But in another context, dependence patterns could reverse. In 2007, when teachers in Rajasthan threatened the ruling party with electoral sabotage, they were the ones who wielded power, with the ruling party dependent on them for victory. But in another context at the same time, the ruling party controlled individual teachers through patronage-based appointments and transfers (figure 10.2). Because these opposing relations occurred simultaneously, the distinctions between who was more dependent and who was accountable to whom became blurred. Such interdependencies govern relationships between various participants in education systems, such as parent-teacher or bureaucrat-middleman interactions.

Interdependencies can become coercive and entrenched. This happens when actors are unable to break out of informal contracts. In the Vyapam case in Madhya Pradesh, India, several bureaucrats, fearing adverse career repercussions, allegedly joined the scam, making it much worse than otherwise possible. Then others joined—with middlemen purportedly profiting offthe connections made between the various players. What started out as a small-time operation allegedly became institutionalized (albeit informally) as people began to believe they would lose out if they questioned the status quo. Likewise, in New York City the unwritten power of school board members forced superintendents and principals to routinely allow wasteful practices. This pattern repeats itself across cases, countries, and time.

As participants get trapped in unhealthy interdependencies, they devise mechanisms to protect themselves from undue blame and punishmen—and avoid taking risks. Fearing repercussions for uncooperative behavior, actors make choices that provide the appearance of change—for example, when a politician presides over school openings but does not address teacher absenteeism, a judge delays case hearings endlessly, or a parent sits on a dormant school committee. These actors become averse to taking risks or innovating. Such behavior coexists with a perverse form of information management. For fear of being wrongly implicated in illegal behavior, officials sometimes generate mountains of paper, files, and data, paralyzing the system instead of providing relevant information. The opacity, stickiness, and low capacity of education systems make it easier to exaggerate accomplishments and cover up performance problems.

Abdicating responsibility and avoiding blame erode an education system’s ability to function, thereby perpetuating a low-accountability, low-learning equilibrium. Teachers, bureaucrats, judges, or politicians who fail to cooperate with the status quo are likely to put themselves at considerable professional risk. The system leaves them little choice but to conform. The problem is not limited to specific individuals, but arises from the multiple interests of actors and the underlying incentives in education systems. The accountability needed to ensure student learning becomes secondary.

This is the story of unhealthy politics. Healthy politics can generate the momentum for reform and deliver results for education outcomes, as chapter 11 shows.