Millennium Development Goals Raised Primary Enrollment but Crowded Out Secondary

🔎 What Was Examined

Precise international metrics can push governments to reallocate resources to improve scores on those metrics. This study examines a secondary consequence of global performance indicators (GPIs): when a GPI targets a specific public good, governments facing finite resources may boost that targeted good at the expense of related goods.

🧭 Theory and Expectations

A formal argument predicts two linked effects of a targeted GPI:

• A main effect: increased provision of the measured public good.
• A substitution effect: reduced provision of related, untargeted public goods due to trade-offs in public spending.

Both effects are expected to vary with domestic institutions and information environments. The main and substitution effects should be largest in states that are least accountable (politically opaque or nondemocratic) and smallest in the most accountable states.

🧠 How This Was Tested

• A formal model formalizes the incentives created by GPIs and the conditions under which substitution arises.
• Empirical tests use cross-national data on primary and secondary school enrollment across 114 countries, leveraging the Millennium Development Goals (MDGs) as a case where primary enrollment was explicitly targeted while secondary was not.

📊 Key Findings

• Countries increased primary enrollment rates, the outcome directly targeted by the MDGs.
• Simultaneously, countries reduced (or underinvested in) secondary enrollment rates, consistent with a substitution away from the untargeted related public good.
• Both the boost to primary and the crowding-out of secondary are moderated by accountability: the effects are strongest in less accountable (more opaque, nondemocratic) governments and are mitigated as accountability rises.

💡 Why It Matters

Global performance indicators can produce perverse allocation incentives across public goods, not just the intended improvements on targeted indicators. Policymakers and international organizations should account for potential substitution effects and consider domestic institutional contexts when designing and promoting global targets.