The research team, led by Thandar Win Maung and Sayamol Charoenratana of Chulalongkorn University together with Nguyen Diem Tien Ho and Peter Michael Rosset, focused on two deliberately contrasting sites: Lat Lum Kaeo District in Thailand’s irrigated central plains and Meiktila Township in Myanmar’s Central Dry Zone. The choice was strategic. Thailand is a mature, leading global rice exporter, while Myanmar, despite producing rice more cost-effectively than most competitors, struggles with unstable prices, fierce competition and disrupted supply chains. Agriculture contributes roughly 30 percent of Myanmar’s GDP and supports more than 30 percent of Thailand’s population, making both countries acutely sensitive to climate-driven yield instability. The World Bank has projected that climate-induced threats could cost Thailand 7 to 14 percent of GDP by 2050 under current scenarios, while the World Food Programme estimates that over 12 million people in Myanmar are acutely food insecure, a crisis worsened by climate shocks and socioeconomic instability.

The study zeroed in on four adaptation practices that farmers themselves reported most often: crop rotation, the use of high-yielding varieties, adjustment of cropping calendars, and modification of planting methods. These are not exotic technologies. They are largely knowledge-based, low-cost adjustments rooted in indigenous farming experience, sometimes combined with government-initiated techniques. Farmers who implemented at least one of these practices during the 2023 rainy season were classified as adapters, while those who reported none were treated as non-adapters. In Lat Lum Kaeo, 62.2 percent of farmers had adopted at least one measure, compared with 51.9 percent in Meiktila. Thai farmers leaned toward more proactive, diversified responses, with 51.1 percent practicing crop rotation and 47.8 percent changing planting methods, while Meiktila farmers relied most heavily on rotation at 39.5 percent and on high-yielding varieties at 30.8 percent.

Before estimating any impacts, the researchers checked whether farmers’ perceptions of climate change matched the meteorological record, and they largely did. Weather station data from 2006 to 2023 show rising temperatures in both districts, with Lat Lum Kaeo warming from an average of 29.25°C to 30.28°C and Meiktila from 27.22°C to 28.28°C. Rainfall trends diverged: Meiktila’s total annual rainfall declined while Lat Lum Kaeo’s increased. When asked about the past decade, 84.3 percent of Meiktila farmers and 63.2 percent of Lat Lum Kaeo farmers perceived rising temperatures, and 77.8 percent of Meiktila farmers reported declining rainfall against 73.5 percent of Thai farmers who saw increases. The alignment between perception and measurement matters, because farmers only adapt when they believe the climate is genuinely shifting.

The methodological core of the study is its handling of self-selection bias, a persistent problem in observational economics. Farmers who adapt may simply be more skilled, better informed or better resourced to begin with, so naively comparing adapters with non-adapters can produce misleading conclusions. The team used a random utility framework to model the adaptation decision, in which a farmer adapts only when the expected net benefit is positive, and then applied propensity score matching to construct a fair comparison. Each of the 209 adapters was matched with three similar non-adopters from the 156-household control group using nearest-neighbour matching with replacement. Diagnostic tests confirmed the approach worked: the pseudo-R-squared fell from 0.122 to 0.026 after matching, the mean standardized difference dropped from 21.3 percent to 10.2 percent, and total bias reduction reached 52.1 percent, indicating that systematic differences between the groups were substantially eliminated.

The headline results are striking. Adaptation raised the probability of being food secure by 19.9 percentage points among adopters, and increased net farm income by 242.22 US dollars per hectare, both significant at the 1 percent level. Food security was measured with the World Food Programme’s Food Consumption Score, which tracks how often households consume different food groups over a week, with households classified as poor, borderline or acceptable. Before matching, the raw gap was even wider: 78 percent of adapters were food secure compared with only 51 percent of non-adapters, and adapters earned an average of 1,097.75 dollars per hectare against 913.92 dollars for non-adapters. Non-parametric Wilcoxon rank-sum tests and 95 percent confidence intervals that excluded zero reinforced the reliability of the positive effects.

Perhaps the most intriguing finding concerns intensity. The distribution of adoption was starkly polarized: 42.7 percent of farmers took no adaptation action at all, while a combined 34.3 percent adopted either three or all four strategies. Farmers who stacked multiple practices consistently outperformed those using fewer or none, with both food security and income rising in a broadly linear fashion as the number of strategies increased. This suggests the practices are complementary rather than substitutes, each additional measure delivering incremental welfare gains. It also carries a warning: the widening gap between non-adopters and multi-strategy users hints at growing inequality if support remains uneven across regions and income groups.

What drives farmers to adapt in the first place? The logistic regression underlying the propensity scores identified five significant determinants. Education was the strongest, positive and significant at the 1 percent level, followed closely by access to climate information. Access to credit was significant at the 5 percent level, while farming experience and road access mattered at the 10 percent level. Notably, access to agricultural extension services and attendance at training, the conventional channels for promoting climate-smart agriculture, were positive but not statistically significant. The authors suggest this may reflect poor extension support, a reliance on traditional knowledge, or training that failed to address appropriate topics or methods, a finding with uncomfortable implications for how adaptation programs are currently designed and delivered.

The policy implications flow directly from these results. Because credit and road infrastructure significantly boost adaptation, the authors argue for policies that ease financial constraints and expand investment in rural transport, which reduces costs, improves market connectivity and cuts postharvest losses. Because extension and training underperformed, they call for a comprehensive review of their quality and curricula, shifting away from generic advice toward specialized, context-specific programs that align with farmers’ practical needs and existing indigenous knowledge. They also recommend strengthening formal and informal education, building farmer networks and communication platforms for peer learning, and involving rural communities directly in adaptation planning to improve local ownership. Governments, the private sector and NGOs, the study concludes, must scale region-specific strategies that integrate formal technologies with farmer-led practices.

The authors are candid about the study’s limits. The cross-sectional design cannot strictly establish temporal ordering between adoption and income, the survey did not record when farmers first adopted practices that are often traditional and long-established, and propensity score matching controls only for observable characteristics, leaving possible bias from unobserved factors such as managerial ability and soil quality. Adoption was self-reported, the data cover selected areas only, and Myanmar’s socio-political instability was not explicitly controlled for even though it may influence both adaptation decisions and food security. Future work, they note, should incorporate longitudinal data and region-specific indicators of political stability. Even with these caveats, the message is clear and consequential for a region where rice is life: adaptation works, it pays, and the farmers who combine multiple strategies stand to gain the most, provided the information, credit and infrastructure reach those currently left behind.

Subject of Research: Impacts of climate adaptation strategies on food security and farm income among rice-farming households in Thailand and Myanmar

Article Title: Rice farming resilience to climate variability: Impacts of adaptation strategies on food security and farm income in Thailand and Myanmar

Article References: Maung, T. W., Charoenratana, S., Ho, N. D. T., & Rosset, P. M. (2026). Rice farming resilience to climate variability: Impacts of adaptation strategies on food security and farm income in Thailand and Myanmar. Environmental and Sustainability Indicators, 32, Article 101560. https://doi.org/10.1016/j.indic.2026.101560

Image Credits: AI Generated

DOI: 10.1016/j.indic.2026.101560

Keywords: rice farming, climate change adaptation, food security, farm income, Thailand, Myanmar, propensity score matching, crop rotation, high-yielding varieties, climate variability, Southeast Asia, smallholder farmers

The research team surveyed 400 small and marginal farmers across selected districts of Rayalaseema, deliberately sampling both recipients and non-recipients of microcredit. This two-group design allowed the researchers to compare outcomes between farmers who had access to small loans and those who did not, holding the shared environment of drought risk, soil constraints, and market conditions broadly constant. The region is a natural laboratory for this question: agriculture there is dominated by rainfed farming, water scarcity is chronic, and farmers have historically faced limited access to formal credit, leaving many dependent on informal lenders whose high interest rates discourage investment in anything beyond immediate survival.

The headline finding is striking. Sixty percent of microcredit recipients adopted agricultural innovations, ranging from improved inputs to modern cultivation techniques, compared with only 35 percent of non-recipients. That near-doubling of the adoption rate matters because technology adoption is widely regarded as the central pathway through which smallholders can escape low-productivity traps. In the framework of innovation diffusion, farmers without discretionary capital cannot afford the upfront cost or the risk of trying a new practice; credit relaxes that constraint, giving them the liquidity to experiment and the buffer to absorb a failed season.

But the authors did not stop at simple comparisons, which can be misleading. Farmers who take microcredit may differ systematically from those who do not, in education, landholding, motivation, or social connections, and those differences, rather than the credit itself, could drive the observed gains. To address this problem of selection bias, the team deployed two complementary econometric techniques. Structural Equation Modelling, or SEM, was used to map the network of factors that influence both microcredit access and farmer empowerment, testing whether the hypothesized relationships between socio-economic variables, credit use, and outcomes fit the observed data. Endogenous Switching Regression, or ESR, was then applied to estimate the income effects of credit while explicitly correcting for the fact that credit receipt is a choice rather than a random assignment.

The statistical diagnostics suggest the models are trustworthy. The comparative fit index, which measures how well the structural model reproduces the observed covariance patterns, reached 0.977, and the Tucker-Lewis index came in at 0.933, both above the commonly recommended thresholds of roughly 0.90 or 0.95. The root mean square error of approximation, which penalizes model complexity and measures the discrepancy per degree of freedom, was 0.094, indicating a reasonable approximation of the data despite falling slightly above the strictest conventional cutoff. In practical terms, the fitted model captures the relationships among credit access, empowerment, and farm outcomes with enough fidelity to support the study’s substantive conclusions.

The SEM analysis also identified which farmers are best positioned to convert credit into agricultural improvement. Education emerged as a significant socio-economic factor, consistent with the idea that literate farmers can better evaluate new technologies, navigate loan procedures, and interpret extension advice. Membership in cooperative societies was another strong predictor of productive credit utilization. Cooperatives serve as conduits for information, collective bargaining, and peer demonstration effects, so a farmer embedded in such a network is more likely to know what a loan could profitably finance and more likely to see neighbors succeeding with the same investments. These findings echo a broader literature suggesting that financial inclusion works best when it is bundled with social and informational infrastructure rather than delivered as cash alone.

The income results are where the study becomes most consequential for policy. Using the switching regression framework, the researchers found that farmers with credit access earned substantially higher returns than their counterparts without it. The effect was particularly pronounced among larger landholders: farmers owning more than 10 hectares saw their income rise from 40,833 rupees without credit to 66,733 rupees with credit, an increase of 63 percent. That figure illustrates a sobering dynamic in rural finance. Larger farmers, who can offer collateral, absorb risk, and scale the returns on any technology they adopt, are best able to translate a loan into profit. Small and marginal farmers, who make up the majority of India’s agricultural workforce, benefit as well, but the study’s findings imply that credit alone may not fully equalize opportunity across the landholding spectrum.

Indeed, the authors are careful about the limits of microcredit. Its impact, they conclude, is constrained by structural issues that no loan officer can fix: inadequate irrigation infrastructure, poor market access, and the broader climate risks of a semi-arid region where a single failed monsoon can wipe out the returns on any investment. A farmer may borrow to buy drip irrigation equipment or improved seed, but if canal water never arrives or the nearest regulated market is hours away over rough roads, the technology cannot deliver its full yield and income potential. This is a recurring theme in development economics: financial instruments are necessary but not sufficient, and their effectiveness depends on the physical and institutional environment in which they operate.

The policy implications follow directly. The study argues for interventions that integrate financial inclusion with infrastructure development and agricultural extension services. In practice, that means pairing microcredit programs with investments in water harvesting and irrigation, rural roads and market linkages, and front-line advisory services that help farmers choose and correctly deploy new technologies. It also means designing credit products that fit the cash-flow realities of rainfed agriculture, where income arrives in one or two lumpy harvests and drought years can interrupt repayment. Financial literacy programs, which the authors’ framework positions as a driver of farmer empowerment, can further raise the return on every rupee borrowed by helping households plan investment and manage debt.

For the wider world of climate adaptation research, the Rayalaseema study adds a valuable data point from a region that sits squarely on the front line of climate stress. Semi-arid farming systems across the developing world, from the Sahel to southern Africa to South Asia, face the same triad of constraints: scarce credit, slow technology diffusion, and mounting climate risk. Evidence that microcredit can nearly double technology adoption rates, and that adoption translates into measurable income gains, strengthens the case for microfinance as a component of resilient farming strategies. At the same time, the 63 percent income boost concentrated among larger landholders is a caution that well-intentioned financial inclusion can widen gaps unless it is deliberately paired with the infrastructure, cooperatives, and extension support that allow the smallest farmers to benefit too. In Rayalaseema, the loan is only the beginning; what farmers can do with it depends on everything else around them.

Subject of Research: The effect of microcredit on sustainable farming technology adoption and farm income among smallholders in Rayalaseema, India

Article Title: Microcredit and sustainable farming practices in Rayalaseema, India with emphasis on technology adoption and income enhancement

Article References: Vedavathi, V., Kesavaraj, G., & Lodi, F. K. (2026). Microcredit and sustainable farming practices in Rayalaseema, India with emphasis on technology adoption and income enhancement. SN Social Sciences, 6(10), Article 450. https://doi.org/10.1007/s43545-026-01723-y

Image Credits: AI Generated

DOI: 10.1007/s43545-026-01723-y

Keywords: microcredit, sustainable agriculture, technology adoption, smallholder farmers, Rayalaseema, India, structural equation modelling, endogenous switching regression, farm income, climate resilience, financial inclusion, rural development