Low incomes are often unsteady within the year. By design, however, poverty headcounts in national statistics downplay the challenges of instability. We introduce the timecount, a measure that explicitly captures the shifting durations and intensities of poverty within the year. We show that, due to survey methodologies, timecounts have unintentionally become the de facto poverty rates in India and other countries, effectively replacing headcounts. In monthly longitudinal data from rural India, the timecount is 28% larger than the conventional headcount, captures deprivations of a wider population, and better predicts child health. We describe consequences for analysis, global policy, and ethics.

How we think about development hinges in large part on how we think about poverty. The world community has embraced the goal to end global extreme poverty as a cornerstone of development policy, but we show how success will hinge on how “poverty” is understood. We argue that global poverty will not be eliminated even if the global headcount of poverty is brought to zero. This is because much poverty is experienced within the year by people who are not typically designated as “poor”. Their deprivations may be substantial, but they systematically go uncounted through the process of annualizing data. We consider poverty as experienced during the year and describe the steps needed to truly achieve the goal of ending poverty.

Upward mobility for poor, rural workers often involves migrating to cities, but sending money home can be costly and unreliable. Through an RCT, we introduced mobile banking to migrant-household pairs from rural Bangladesh. One year later, use of the digital financial technology had increased urban-to-rural remittances, decreased rural poverty, and improved conditions in the lean season. Eight years later, treated households had acquired 33 percent more productive assets by value. The treatment and control groups were using mobile money to comparable extents by then, and treatment effects on rural household consumption, income, poverty, and financial outcomes were no longer detectable.