The Building State Capability (BSC) program at the Center for International Development (CID) at Harvard University researches new strategies and tactics to build the capability of public organizations to execute and implement.

The BSC program is exploring the potential of a Problem Driven Iterative Adaptation (PDIA) approach, which rests on four core principles:

local solutions for local problems
pushing problem driven positive deviance
try, learn iterate, adapt
scale through diffusion

Recent Publications

The Risks to Education Systems from Design Mismatch and Global Isomorphism

Pritchett, Lant. 2014. “

The Risks To Education Systems From Design Mismatch And Global Isomorphism

The incredibly low levels of learning and the generally dysfunctional public sector schooling systems in many (though not all) developing countries are the result of a capability trap (Pritchett et al. 2010). Two phenomena reinforce persistent failure of schooling systems to produce adequate learning outcomes. One is the mismatch between system design—the allocation of activities across organizations and mechanisms of accountability—and the insights of the 'new institutional economics' from principal agent models and contract theory. In particular, many education systems attempt to manage teaching and learning as a 'thin' or 'logistical' activity that can be managed with top-down control and an emphasis on compliance. The reality is that teaching is a 'thick' or 'implementation intensive' activity that performs better when teachers and operators of schools are given performance standards, have multiple in-depth accountability channels, and are given greater autonomy. The second phenomena that facilitates persistent failure is global isomorphism on enrollment and inputs (Meyer et al. 1977; Boli et al. 1985; Meyer et al. 1997). That is, the field (in the sense of Bourdieu 1993) of global education has produced a near exclusive emphasis on enrollments and duration in school, adequacy of physical inputs, and formal qualifications that allowed, perhaps encouraged, national systems to ignore completely performance on child-learning (of any type, measured in any way). I conclude with a comparison in India of the national governments recent efforts in basic education which have been almost exclusively isomorphic.

Does successful governance require heroes? The case of Sergio Fajardo and the city of Medellín: A reform case for instruction

Andrews, Matt, and Alejandro Fajardo. 2014. “

Does Successful Governance Require Heroes? The Case Of Sergio Fajardo And The City Of Medellín: A Reform Case For Instruction

”. Publisher's VersionAbstract
The city of Medellín, Colombia was a cauldron of violence with 185 homicides per 100,000 people in 2002. By 2006, this rate had declined to 32.5. Such successful transformation was termed the ‘Medellín miracle’ and credited to policies of the city’s mayor, Sergio Fajardo. Fajardo came to office in 2004 and led a series of reforms that observers call visionary. The story of Medellín’s revival starts before Fajardo took office, however, and involved many more people than the mayor. This abridged version of the story offers instructors a classroom case to discuss leaders and leadership in governance reform.

Why Distributed End Users Often Limit Public Financial Management Reform Success

Andrews, Matt. 2014. “

Why Distributed End Users Often Limit Public Financial Management Reform Success

Externally supported Public Financial Management (PFM) reforms often have limited success in developing countries. The reforms commonly introduce new laws and systems that are not fully implemented or used, especially by distributed agents—budgeters, accountants, and such in sector ministries, provinces, and districts. This article asks why this should be expected and what could be done about it. It builds a theory of institutional change and tests such using data from a survey of public sector accountants in Eastern and Southern African countries—one sub-set of which was distributed. The evidence supports a simple explanation of why distributed end users often limit PFM reform success: they are likely to support incumbent institutions and question reform alternatives and are less engaged in reforms than more concentrated agents who champion reforms. The article suggests that research and practice needs to better account for the influence of distributed agents on externally supported reform success.


Doing Development Differently: The Manifesto

We pledge to apply these principles in our own efforts to pursue, promote and facilitate development progress, to document new approaches, to spell out their practical implications and to foster their refinement and wider adoption.