Loading field map
How should developing countries structure their financial systems to promote growth and reduce poverty? This question has generated five successive and overlapping frameworks, each offering a different diagnosis of what goes wrong in financial markets and what policy should do about it. The frameworks range from heavy state direction of credit to market liberalization, from microcredit to institutional reform, and from cross-country regressions to randomized experiments. Understanding the sequence of these frameworks—and the tensions that remain between them—is essential for grasping how development finance has evolved and where it stands today.
In the decades after World War II, newly independent countries in Asia, Africa, and Latin America faced a pressing problem: how to finance industrialization when private capital markets were thin or nonexistent. The dominant answer was state-led development banking. Governments created specialized development banks to channel subsidized credit to priority sectors such as heavy industry, agriculture, and infrastructure. Interest rates were kept artificially low—a policy known as financial repression—to reduce the cost of borrowing for favored projects. The rationale was straightforward: private banks, left to themselves, would avoid long-term, high-risk industrial loans, so the state had to step in.
This framework treated financial markets as inherently prone to failure in developing economies. It assumed that the state could identify strategic sectors and allocate credit more effectively than markets. By the 1970s, however, the costs of this approach had become visible. Negative real interest rates discouraged savings, while directed credit led to widespread non-performing loans and political interference in lending decisions. Development banks accumulated large losses, and the financial systems they dominated remained shallow and fragile. The stage was set for a radical alternative.
The financial liberalization framework emerged as a direct challenge to state-led development banking. Drawing on the work of Ronald McKinnon and Edward Shaw in the early 1970s, liberalization theorists argued that financial repression was itself the cause of underdevelopment. By keeping interest rates below market-clearing levels, governments discouraged savings and misallocated the limited funds that were available. The solution was to free interest rates, privatize banks, and open financial markets to foreign competition. Higher real interest rates would mobilize savings, and market forces would allocate capital to its most productive uses.
This framework dominated policy advice from the 1980s onward, especially through the structural adjustment programs of the World Bank and IMF. Yet the results were mixed. In Latin America, rapid liberalization in the Southern Cone led to banking crises and high real interest rates that choked off investment. The East Asian financial crisis of 1997–98 dealt a further blow: countries that had followed liberalization prescriptions suffered devastating collapses, while those with more regulated systems (like China and India) weathered the storm better. The liberalization framework had to be tempered: the new consensus called for sequencing reforms carefully, strengthening bank supervision before opening markets, and recognizing that liberalization could fail in the absence of sound institutions.
Around 1990, two new frameworks emerged that coexisted and addressed different aspects of the financial development puzzle. They did not directly replace liberalization but rather shifted the conversation in distinct directions.
Microfinance burst onto the scene with the Grameen Bank in Bangladesh, founded by Muhammad Yunus. The core insight was that poor households lacked access to formal credit not because they were unreeditworthy, but because banks lacked the mechanisms to lend small amounts without collateral. Group lending—where borrowers form small groups that jointly guarantee each other's loans—used social collateral to overcome information and enforcement problems. The initial narrative was transformative: microcredit would lift millions out of poverty by enabling self-employment and small enterprise.
This framework differed sharply from both state-led banking and liberalization. Unlike state-led development banks, microfinance institutions were often non-governmental and focused on the very poor rather than on large industrial projects. Unlike liberalization, which assumed that freeing markets would automatically expand access, microfinance argued that specific institutional innovations were needed to reach excluded populations. By the 2000s, microfinance had expanded globally, but experimental evidence—discussed below—tempered the early claims. The framework broadened into financial inclusion, encompassing savings, insurance, and digital payments, and it remains a vibrant area of policy and research.
At the same time, a separate line of research asked a different question: why do some countries have deep, well-functioning financial systems while others remain financially shallow? The answer, developed by economists such as Robert King, Ross Levine, and later Rafael La Porta, Florencio Lopez-de-Silanes, Andrei Shleifer, and Robert Vishny, pointed to legal origins and property rights. Countries with legal systems that protect outside investors—especially those rooted in common law—tend to have larger stock markets and more credit to the private sector. Weak contract enforcement and insecure property rights, by contrast, stifle financial intermediation.
This institutional framework offered a diagnosis for why financial liberalization had failed in many settings: freeing interest rates and privatizing banks did little good if the underlying legal and regulatory infrastructure was weak. It also provided a rationale for a different kind of state action: not directing credit, but building the institutions that support arm's-length finance. The methodology was cross-country regression, using indicators of legal origin, creditor rights, and accounting standards to explain variation in financial depth. This framework coexisted with microfinance but addressed a different scale: it focused on the overall financial system rather than on specific lending technologies for the poor.
The most recent framework is not a new theory of what drives financial development but a methodological revolution in how to test claims about financial interventions. Experimental development finance applies randomized controlled trials (RCTs) and quasi-experimental designs to evaluate the impact of financial products and policies. The central problem it addresses is attribution: how can we know that a microcredit program, a savings account, or a financial literacy course actually caused the observed outcomes?
This framework emerged in the early 2000s, led by researchers such as Abhijit Banerjee, Esther Duflo, and Dean Karlan. Its first major target was microfinance. A wave of RCTs in countries from India to Morocco found that access to microcredit had modest effects on business investment and income but did not produce the dramatic poverty reduction that advocates had claimed. These findings did not discredit microfinance but narrowed its promise: microcredit was a useful tool for some households, not a silver bullet. The experimental framework also expanded the agenda beyond credit to savings, insurance, and commitment devices, showing that poor households valued safe savings accounts and weather insurance as much as or more than loans.
Experimental development finance differs from the institutional economics framework in both method and scope. Institutional economists use cross-country regressions to identify broad correlates of financial development; experimentalists use within-country randomization to identify causal effects of specific interventions. The two approaches often disagree on what constitutes credible evidence. Institutionalists argue that experimental results are too narrow to inform system-wide reform; experimentalists counter that cross-country regressions are plagued by endogeneity and omitted variable bias. This methodological tension remains unresolved.
Today, three frameworks remain active: microfinance and financial inclusion, new institutional economics of financial development, and experimental development finance. Each has carved out a distinct domain. Financial inclusion research and practice focus on product design, delivery channels (including mobile money), and the welfare effects of access. Institutional economics continues to inform debates about legal reform, creditor rights, and the role of financial regulation. Experimental methods have become the gold standard for evaluating specific interventions, and their influence has spread to other areas of development policy.
What do these frameworks agree on? All three recognize that financial markets in developing countries are not self-correcting and that institutions matter—whether those institutions are group lending mechanisms, legal systems, or regulatory frameworks. All three also accept that the state has a role, though they differ sharply on what that role should be. The main disagreements are about method and about the relative importance of system-wide versus targeted interventions. Institutional economists favor cross-country evidence and argue for broad legal and regulatory reforms; experimentalists favor randomized evidence and argue for testing specific products and policies before scaling them. Meanwhile, the older frameworks have not disappeared entirely. Development banks have seen a modest revival in the context of green finance and pandemic response, and some economists argue for a return to directed credit under certain conditions. The liberalization framework, though chastened, still underpins much of the advice on financial openness and capital account liberalization.
The central tension of development finance—how much to rely on markets, states, institutions, and experiments—remains unresolved. Students entering the field will find a lively debate, not a settled consensus, and a set of tools that continue to evolve.