RAIN FALLS, POVERTY RISES: Study quantifies how rainfall shocks push millions deeper into destitution

By Francis Allan L. Angelo

In a country where rain is both life and ruin, researchers have now put hard numbers on what Filipino farmers have long known in their bones: when the rains fail, so does everything else.

A new science-policy brief (SPB) released by the Oscar M. Lopez (OML) Center, in collaboration with De La Salle University, delivers what may be the most rigorous quantitative evidence to date on the relationship between rainfall variability and poverty in the Philippines.

The data reveals a troubling trend: below-average rainfall can push poverty incidence up by as much as 6 percentage points nationally — and nearly 10 percentage points among agricultural households alone.

At a national poverty incidence of 15.5 percent in 2023, representing some 17.54 million Filipinos, a swing of that magnitude translates into millions of additional people falling below the poverty line when the rains fail to arrive.

“Rainfall variability is a critical but often underexamined pathway through which climate change affects poverty, especially in countries where livelihoods remain highly climate-sensitive,” the brief states in its opening lines — a framing that sets up what becomes a detailed, province-level diagnosis of a crisis unfolding in slow motion across the Philippine archipelago.

Behind the numbers

The brief, titled “Do Rainfall Shocks Affect Poverty? Evidence from Philippine Household and Climate Data,” is the third in the OML Center’s Science-Policy Brief series. It was authored by Alfi Lorenz B. Cura, Jean Mariel M. Rañises, Ayn G. Torres, and Krista Danielle S. Yu, with Dr. Rodel D. Lasco serving as executive editor.

The study draws from two primary datasets: the Philippine Statistics Authority’s Family Income and Expenditure Survey (FIES) for 2018 and 2023, and rainfall records from the Department of Science and Technology’s Philippine Atmospheric, Geophysical and Astronomical Services Administration (DOST-PAGASA).

The 2021 FIES round was deliberately excluded to avoid the confounding effects of the COVID-19 pandemic on household income patterns.

Table 1 shows that poverty rates differ across rainfall conditions. In general, poverty tends to be higher during periods of below-average rainfall and lower during above-average rainfall. The differences are more pronounced among agricultural households, highlighting their greater sensitivity to rainfall variability.

In total, the analysis encompasses more than 82,000 household observations — 47,837 from 2018 and 34,610 from 2023 — across 20 provinces with synoptic weather stations. This province-level matching of climate data to household survey data is methodologically significant: it allows researchers to link specific rainfall conditions in a given location and year directly to the poverty status of households living there.

To estimate causal effects, the research team employed a Difference-in-Differences (DID) approach, comparing poverty outcomes between 2018 and 2023 across areas that experienced different rainfall conditions.

For the agricultural household analysis, they used a more advanced Triple Difference (DDD) specification, which isolates the additional effect of rainfall shocks on farming households relative to non-farming ones while controlling for time-invariant provincial differences and national year-to-year trends.

Rainfall shocks are defined as deviations from historical norms: a positive shock means above-average rainfall, a negative shock means below-average.

Both types, the study finds, have measurable and statistically significant effects on poverty — but in opposite directions, and with notably asymmetric magnitudes depending on which sector of the population is being observed.

Dry spells are the bigger threat

The headline finding is that drought-like conditions — periods of below-average rainfall — are strongly associated with higher poverty incidence. The estimates suggest that below-average rainfall periods increase poverty incidence by approximately 3.9 to 6.0 percentage points across the general household population.

For agricultural households, the effect is dramatically larger. Farming families, who depend directly on precipitation for crop production and therefore for income, see poverty incidence rise by nearly 10 percentage points (pp) — specifically, 9.96 pp — during periods of below-average rainfall. This figure captures the compound vulnerability of households that have few alternative income sources, limited access to irrigation, and almost no financial cushion to weather a bad harvest.

Food poverty — defined as the condition where household income is insufficient even to meet basic nutritional needs — tells a similarly troubling story. Below-average rainfall periods are associated with a rise in food poverty of around 2.2 percentage points overall, with agricultural households bearing the brunt of an additional 5.7 pp increase over and above that baseline effect.

The descriptive data in the brief add texture to these estimates. Table 1 shows that poverty rates consistently differ across rainfall conditions, with the gaps most pronounced among agricultural households — a pattern consistent across both survey years.

The silver lining

The brief’s findings are not uniformly dire. Above-average rainfall, perhaps counterintuitively in a country prone to flooding, is associated with lower poverty and food poverty rates — at least in the moderate range of surplus rainfall examined in this study.

The estimates suggest that above-average rainfall conditions are associated with a decline in poverty incidence of 3.6 to 4.4 percentage points, with food poverty declining by 1.6 to 2.3 percentage points. Agricultural households again see larger gains: food poverty among farming families falls by an additional 3.6 percentage points under above-average rainfall compared to non-farming households.

The researchers are careful to qualify this finding. The study captures average effects and is not designed to model extreme rainfall events such as severe flooding, which can devastate crops and infrastructure even as aggregate rainfall statistics appear favorable. The relationship between rainfall and agricultural productivity is likely nonlinear: moderate surpluses benefit rainfed farming systems, while extreme precipitation events carry their own destructive consequences.

Still, the directional finding is meaningful for policy: it confirms that in the Philippines’ largely rainfed agricultural landscape, adequate rainfall is productivity-enabling — and its absence is poverty-inducing.

A climate system becoming more erratic

The researchers situate their findings within a broader and accelerating climate context. The Philippines is not merely experiencing normal year-to-year rainfall variability; it is experiencing a systematic shift in precipitation patterns driven by global warming.

Drawing on the OML Center’s own State of the 2023 Philippine Climate report and the 2024 Philippine Climate Change Assessment (PhilCCA), the brief documents that tropical cyclone-induced rainfall in the Philippines has increased significantly over recent decades. Since 2000, the contribution of typhoon rainfall to total annual rainfall has risen by 16.9 to 19.3 percent per decade — a staggering rate of change. From 1958 to 2017, high-precipitation-event (HPE) days and total HPE rainfall also increased by 6.0 and 12.7 percent per decade, respectively, driven largely by typhoon activity.

These trends are unfolding against a backdrop of rising temperatures. The year 2023 saw minimum temperatures in the Philippines rise approximately 0.6°C above normal, compounding moisture dynamics and intensifying rainfall-related risks. Globally, 2023 was confirmed as the warmest year on record, exceeding the previous peak in 2016 by 0.13°C to 0.17°C across multiple global datasets — a finding the brief cites from the IPCC’s Sixth Assessment Report Synthesis as consistent with national observations.

The implication is significant: the rainfall shocks documented in this study are not a static risk. They are intensifying. A climate system delivering more extreme precipitation events — more severe droughts and more powerful typhoon rains — will generate more frequent and more severe poverty shocks to agricultural households unless adaptation measures are substantially scaled up.

Frameworks without delivery

The Philippines is not without laws. The brief surveys a comprehensive national policy architecture: the Climate Change Act (Republic Act 9729), the Philippine Disaster Risk Reduction and Management Act (RA 10121), the National Climate Change Adaptation Plan (NCCAP), and the National Disaster Risk Reduction and Management Plan (NDRRMP). The Department of Agriculture operates an Adaptation and Mitigation Initiative in Agriculture (AMIA Program). These are serious instruments with serious mandates.

Yet the brief’s assessment of their effectiveness is pointed: implementation gaps continue to limit their effectiveness in addressing the impacts of rainfall variability. Several structural failures are identified.

First, policy siloing. Climate risks, agricultural production, and poverty reduction are frequently treated as separate concerns managed by different agencies, leaving integrated responses to be resolved at the local level — where capacity is often least available to do so.

Second, the reactive posture of disaster response. When Typhoons Kristine, Leon, Marce, Nika, Ofel, and Pepito struck the Philippines in 2024 in rapid succession, the government reported providing emergency aid worth PHP 1.1 billion. The brief does not dispute the necessity of that aid. But it argues that counterpart investment in long-term resilience-building measures must be proportionate — and currently is not.

Third, uneven local capacity. National policies do not automatically translate into consistent local delivery. Gaps in technical expertise, resources, and access to climate information mean that programs reach communities unevenly, with the most remote and most agriculture-dependent areas often the worst served.

Fourth, limited anticipatory action. Across the country, the practice of using climate forecasts to trigger pre-emptive interventions before impacts materialize — what the brief calls anticipatory action — remains poorly institutionalized. This represents a missed opportunity: modern climate forecasting is increasingly capable of providing actionable advance warning, but the policy machinery to act on those warnings before households are harmed is still largely absent.

There is also a notable policy gap in agricultural strategy. The National Framework Strategy on Climate Change 2010-2022 outlined strategies for climate-resilient agriculture, but an updated framework beyond 2022 has not yet been released. The DA’s AMIA Program, meanwhile, continues operating without a clear updated strategic framework linking it to the country’s current climate risk assessments.

Five recommendations

Against this backdrop of documented vulnerability and policy underperformance, the brief advances five actionable recommendations directed at national government agencies, local government units, and legislative bodies.

1. Invest in water access infrastructure. The brief calls for expanding irrigation infrastructure, small reservoirs, and rainwater harvesting systems to buffer the effects of dry spells and stabilize agricultural production. It invokes a largely dormant legal instrument: Republic Act 6716, which mandates the Department of Public Works and Highways to construct rainwater collectors in every barangay in the country, to be managed by community-level Barangay Waterworks and Sanitation Associations. A pending Senate Bill (SB 243) seeks to institutionalize standards for these facilities. The brief recommends immediate enforcement of the existing law while the Senate measure works its way through the legislative process.
2. Support climate-smart agriculture. The brief advocates promoting drought-resilient crop varieties, diversifying toward low-water-demand commodities, and strengthening technical support for farmers. In livestock and fisheries, adaptation includes shifting to climate-resilient inputs — drought-tolerant feed and heat-resistant fish varieties. This recommendation envisions a broader transformation of Philippine agricultural systems away from climate-vulnerable monocultures and toward diversified, climate-adapted production.
3. Strengthen risk financing mechanisms. The brief recommends encouraging agricultural insurance uptake and developing index-based crop insurance, where payouts are automatically triggered by measurable climate thresholds — such as below-average rainfall or extreme temperature changes — rather than requiring costly loss assessments. The mechanism is not novel: the United Nations Development Programme piloted a weather-based insurance index with the Philippine Crop Insurance Corporation (PCIC) before the COVID-19 pandemic. House Bill 6519 currently proposes expanding this into a formal automatic payout system backed by a proposed Php 5.8 billion subsidy fund. The brief describes institutionalizing this mechanism as “a breakthrough in both short- and long-term approaches to mitigating agricultural shocks.”
4. Expand climate monitoring and link it to response mechanisms. The Philippines currently operates only 57 synoptic weather stations covering 45 of its 82 provinces — a coverage gap that leaves large swaths of agriculture-dependent territory without localized, accurate rainfall data. The brief calls for expanding the weather station network, accelerating adoption of satellite-based monitoring technologies, and systematically linking climate information to response mechanisms so that data translates into timely, location-specific interventions. It cites existing partnerships between the Philippine Space Agency (PhilSA) and the DA and Philippine Rice Research Institute as models to be expanded beyond rice, corn, and onion monitoring to a broader range of crops and geographic areas over the next five to ten years.
5. Strengthen institutional systems and inter-agency coordination. The brief calls for institutionalizing the use of a Results-Based Monitoring and Evaluation System (RBMES) within the NDRRMP to track preparedness spending and reorient budget rules toward anticipatory measures rather than post-disaster relief. It recommends strengthening coordination across water, agriculture, and disaster risk management agencies through the Water Resources Management Office, with shared data protocols and clarified operational roles. It also highlights the eCCET Helper — a decision-support tool that guides local government units through the government’s Climate Change Expenditure Tagging (CCET) system — as a practical instrument for translating climate information into coordinated local action.

What this means for the 7 million

Behind the regression tables and policy recommendations is a human reality that the brief grounds clearly in national statistics.

In the Philippines, approximately seven million households depend on agriculture for their livelihoods, according to the 2022 Census of Agriculture and Fisheries.

Among these farming households, the poverty incidence in 2023 stood at 27 percent — nearly 2.10 million individuals — compared to the national average of 15.5 percent. Farmers are already among the poorest Filipinos, and they are precisely the group the study identifies as most exposed to rainfall shocks.

The numbers tell a story of structural vulnerability compounded by climatic risk.

A farming household in a rainfed province has no irrigation backup when the rains fail. It has, in most cases, no crop insurance to replace lost income. It likely has no access to credit on terms that allow meaningful recovery before the next planting season.

And it is served by local government units that may lack the resources, the technical staff, or the climate information needed to intervene effectively.

The brief is not pessimistic about solutions. Its five recommendations are grounded in existing legislation, proven programs, and pending bills — none of them require inventing something from scratch.

What they require is political will, funding, and the institutional coherence to move from frameworks on paper to programs on the ground.

The OML Center’s Science-Policy Brief series aims to translate research findings into actionable policy guidance for Philippine decision-makers.

The third issue in the series — this rainfall-poverty study — builds on a growing body of international evidence linking climate variability to household welfare outcomes in developing countries, including comparable studies from Vietnam, Ecuador, and other climate-exposed economies.

The researchers acknowledge limitations. The analysis is constrained to 20 provinces with available synoptic station data, which may affect representativeness in underserved areas — precisely the regions where rainfall monitoring is most needed and least available.

The DID methodology assumes parallel trends between affected and unaffected groups, an assumption that cannot be directly verified. And the study captures average effects, not the tail risks of extreme flooding events.

These caveats are noted not to diminish the findings but to underscore the case for one of the brief’s own recommendations: expanding climate monitoring infrastructure so that future research — and future policy — can draw on more comprehensive, more granular data.

The science-policy brief is available for download at https://bit.ly/OMLC-SPB3. Previous briefs in the series can be accessed at https://bit.ly/OMLC-SPBs.

The Oscar M. Lopez Center for Climate Change Adaptation and Disaster Risk Management Foundation, Inc. is based in Pasig City and conducts research at the intersection of climate science, disaster risk, and development policy. The Center operates in partnership with De La Salle University and other academic and government institutions.