Climate change, sea level rise, and the agricultural revolution

Ridhika Mishra; Richard S. J. Tol

AI summary · arXiv: 2610.37306v1 · econ.GN

Climate change, sea level rise, and the agricultural revolution

Ridhika Mishra, Richard S. J. Tol

Original paper Submitted to arXiv on 27 Oct. 2026 License: NONEXCLUSIVE-DISTRIB 1.0 Also in: fr

How to read this summary? Illustrated guide

The essentials

Seasonality, population pressure, and diffusion help explain early agriculture, while technological expansion of its climatic niche outweighed sea-level rise and climate change in expanding agricultural potential.

Question

Which environmental and demographic factors explain the invention and spread of agriculture? Could now-submerged coastal lands have supported early farming, and how much did environmental change versus technological progress expand agricultural potential?

Method

The authors combine a global panel of 1,090 grid cells in 500-year periods with HYDE population and land-use data, two TraCE-21ka climate reconstructions, and sea-level, elevation, ice-cover, and vegetation data. They replicate and extend cross-sectional regressions and panel logit models, initially covering seven independent inventions, add New Guinea as an eighth event, and estimate Cox duration models. They reconstruct counterfactual population distributions and submerged-land invention probabilities, then use climatic convex hulls and a Shapley decomposition across 24 factor orderings to attribute changes in potentially agricultural area.

Findings

Authors’ interpretation The authors interpret their replication and extensions as supporting increased temperature seasonality as an explanation for agricultural invention and spread.
See the passage in the paper
His main hypothesis—increased temperature seasonality stimulated the invention and spread of agriculture—holds up with alternative data and estimators and when tested against competing explanations.
Result Population density predicts invention in the extended logit models but not adoption, whereas neighboring adoption strongly predicts agricultural spread.
See the passage in the paper
Diffusion has a large and highly significant effect. This is still the case if the sample is restricted to 10,000 BCE and later (column 4).
Result Reconstructed environmental changes increased population density in early-adopting areas by 0.0049 people per square kilometre, less than 2% of their population density.
See the passage in the paper
Population density increases by 0.0049 people per square kilometre. This increase is statistically significant, but not substantial: Population density was 0.3207 (0.0022) people per square kilometre, so the increase was not even 2%.
Authors’ interpretation Logit extrapolations identify some submerged areas as suitable for early agriculture, without demonstrating that farming actually occurred there.
See the passage in the paper
Some now-submerged areas in the Persian Gulf, Mediterranean, and Yellow Sea were suitable for early agriculture based on the estimated logit model for the invention of agriculture.
Authors’ interpretation The authors attribute most long-run expansion of agricultural potential to technological improvements that widened agriculture’s climatic niche.
See the passage in the paper
In sum, the net effect of sea level rise on ancient agriculture is small. Climate change had a larger effect, but fluctuated between positive and negative. Both effects are dominated by technological progress that expanded the extensive margin.

Limitations

The archaeological record is selective, and underwater archaeology makes the proposed existence of farming on submerged shelves difficult to test.
See the passage in the paper
The archaeological record is always changing, and skewed towards promising and accessible sites. Underwater archaeology is difficult, which makes it hard to prove, let alone disprove our hypothesis of farming on the continental shelf.
Limited vegetation and climate reconstructions leave uncertainty about the counterfactual pasts poorly characterized.
See the passage in the paper
We use a single reconstruction of past vegetation, and two reconstructions of past climate from a single model; we therefore have little idea of the uncertainties about our counterfactual pasts.
The invention duration model has limited statistical power, with only seven events and no statistically significant parameters under cell-clustered standard errors.
See the passage in the paper
No parameter is statistically significant, as there are only 7 events and standard errors are clustered at the cell level.

Why it matters

The authors argue that seasonality currently has the strongest empirical support among the tested explanations for early agriculture, supplemented by population pressure and technological diffusion. Their reconstruction shifts emphasis from sea-level displacement toward farmers’ capacity to extend cultivation into previously unsuitable climates.

Written by AI from the paper. Passages are checked word for word against the text we analysed; the rest may contain errors.

AI-generated keywords: Agricultural origins; Climate seasonality; Sea-level rise; Population pressure; Technological adaptation

Read further

Comprehensive Summary
Layman's Summary
Blog article
Original abstract

Ridhika Mishra and Richard S. J. Tol investigate why agriculture emerged and spread after the last Ice Age. They compare explanations involving climatic seasonality, warming, carbon dioxide fertilization, population pressure, sea-level rise, and technological progress. In particular, they examine whether coastal flooding displaced populations toward farming and whether agriculture could have originated on land now submerged. Their analysis also replicates and extends Matranga’s 2024 finding that stronger seasonal contrasts encouraged the transition from hunting and gathering to agriculture. The authors combine archaeological dates for agriculture’s invention and adoption with reconstructions of population, agricultural land use, climate, vegetation, elevation, ice cover, and sea level. Their principal agricultural dataset covers 1,090 geographical cells at 500-year intervals and includes seven independent invention events; one extension adds New Guinea as an eighth. They apply cross-sectional regressions, panel logistic models, and survival analysis, comparing two versions of a paleoclimate simulation. To assess coastal displacement, they reconstruct population distributions and model counterfactual scenarios that hold sea level, climate, or vegetation at their 10,000 BCE conditions. They also define agriculture’s climatic envelope using reconstructed cropland locations and distinguish the contributions of environmental change from those of an expanding range of usable climates. Temperature seasonality receives the most consistent support among the climatic explanations. In the invention models, odds ratios generally range from 1.15 to 1.19 per unit increase in temperature seasonality. Its association with adoption also persists across alternative climate reconstructions and estimation methods. However, the survival model for independent invention yields no statistically significant predictors when standard errors are clustered geographically, reflecting the small number of invention events. Population density predicts invention in several specifications but not subsequent adoption. By contrast, neighboring agriculture strongly predicts adoption: estimated odds ratios range from approximately 7 to 21, while survival models produce hazard ratios of about 3.7–3.9. Carbon dioxide offers no convincing independent explanation because its nearly uniform spatial distribution and rising time trend complicate interpretation. Although roughly 8 million square kilometres of dry, ice-free land were submerged between 10,000 and 5,000 BCE, the estimated effects on early agriculture are modest. Holding sea level constant would have increased the land area within the earliest agricultural climatic envelope by approximately 5.6% globally, 6.8% around the Mediterranean, 8.9% around the Black Sea, and 23.4% around the Caspian Sea. Combined environmental changes increased reconstructed population density in early-adopting areas by just 0.0049 people per square kilometre—less than 2% of their estimated density. Model extrapolations identify potentially favorable submerged locations in parts of the Persian Gulf, Mediterranean, and Yellow Sea. Predictions vary substantially: maximum probabilities reach about 9% in the Yellow Sea under the authors’ specifications and 31% in the Gulf of Suez using Matranga’s coefficients. These estimates are model predictions, not archaeological evidence. From 9,000 BCE to approximately the beginning of the Common Era, the land area within agriculture’s climatic envelope expanded from about 4.8 million to 124.7 million square kilometres. The decomposition attributes 103% of the net increase to an expansion of that envelope, interpreted as technological progress. This contribution was offset principally by sea-level rise, at −3.3%, while climate and ice changes accounted for only small net shares. These estimates measure potential climatic suitability, not cultivated acreage. The authors conclude that seasonality, population pressure, diffusion, and especially technological adaptation best explain the observed patterns. Important limitations include incomplete and geographically biased archaeological records, uncertain population reconstructions, simplified inundation modelling, missing historical data on soils and prey abundance, and limited climate-model diversity. Submerged agricultural origins therefore remain speculative, and the estimated relationships do not establish definitive causal mechanisms.

Why Did Farming Begin? Climate, Rising Seas, and the Agricultural Revolution

After the last Ice Age, people began a transformation that would reshape human life: they started farming. But why did agriculture emerge in some places and spread to others? Warmer conditions, growing populations, seasonal food shortages, and technological innovation have all been proposed as explanations. Rising seas add another possibility: perhaps coastal flooding pushed people toward cultivation—or concealed farming’s earliest beginnings beneath the water.

In “Climate change, sea level rise, and the agricultural revolution,” Ridhika Mishra and Richard S. J. Tol investigate these competing ideas. Their findings favor a combination of seasonal contrasts, population pressure, the spread of farming between neighbors, and technological adaptation. Sea-level rise, despite dramatically redrawing coastlines, appears to have played a smaller role.

Reconstructing a Prehistoric Transition

Studying agriculture’s beginnings requires connecting fragmentary archaeological evidence with reconstructions of a world that no longer exists. The authors combine dates for agriculture’s invention and adoption with estimates of population, agricultural land use, climate, vegetation, elevation, ice cover, and sea level.

Their principal agricultural dataset covers 1,090 geographical cells at 500-year intervals and includes seven independent invention events. One extension adds New Guinea as an eighth. They use several statistical approaches, including cross-sectional regressions, panel logistic models, and survival analysis, which examines the timing of transitions. Comparing two versions of a paleoclimate simulation helps test whether their findings depend on a particular reconstruction.

An important distinction runs throughout the study: inventing agriculture is not the same as adopting it. Conditions that encouraged communities to develop farming independently may differ from those that encouraged their neighbors to take it up.

Seasonality Matters More Consistently Than Warming

Among the climatic explanations, temperature seasonality receives the most consistent support. Stronger seasonal contrasts could make food supplies less reliable throughout the year, potentially increasing the appeal of cultivation and storage. That is a plausible interpretation, however, rather than a mechanism definitively established by the statistical results.

In the invention models, odds ratios generally range from 1.15 to 1.19 per unit increase in temperature seasonality. These figures describe changes in odds, not equivalent increases in probability. The association between seasonality and adoption also persists across alternative climate reconstructions and estimation methods.

The evidence is less decisive when the authors examine the timing of independent invention. With geographically clustered standard errors, the survival model produces no statistically significant predictors. The small number of invention events limits what can confidently be inferred. Carbon dioxide also provides no convincing independent explanation: its nearly uniform spatial distribution and rising time trend make its influence difficult to separate from other changes.

Population Pressure and the Power of Neighbors

Population density predicts agricultural invention in several specifications, consistent with the idea that pressure on wild resources encouraged experimentation. But density does not similarly predict later adoption.

Instead, neighboring agriculture stands out. Estimated adoption odds ratios range from approximately 7 to 21, while survival models produce hazard ratios of about 3.7–3.9. These strong associations point toward diffusion as an important part of the story. They do not, by themselves, establish whether farming spread through migration, knowledge exchange, imitation, or some combination.

Did Rising Seas Push People Into Farming?

Between 10,000 and 5,000 BCE, roughly 8 million square kilometres of dry, ice-free land disappeared beneath rising seas. To investigate the consequences, the authors reconstruct population distributions and model scenarios holding sea level, climate, or vegetation at their 10,000 BCE conditions.

The estimated displacement effect is modest. Combined environmental changes increased reconstructed population density in early-adopting areas by just 0.0049 people per square kilometre—less than 2% of their estimated density. Holding sea level constant would have increased land within the earliest agricultural climatic envelope by about 5.6% globally, although regional effects were larger, reaching 23.4% around the Caspian Sea.

Could Agriculture Have Begun on Submerged Land?

Model extrapolations identify potentially favorable submerged locations in parts of the Persian Gulf, Mediterranean, and Yellow Sea. Predictions vary considerably: maximum probabilities reach about 9% in the Yellow Sea under the authors’ specifications and 31% in the Gulf of Suez under an alternative coefficient set. These are model predictions, not archaeological discoveries. Submerged agricultural origins remain speculative.

Technology Expanded Farming’s Possibilities

The study’s broadest finding concerns agriculture’s expanding climatic envelope—the range of climates in which reconstructed cropland occurred. Between 9,000 BCE and approximately the beginning of the Common Era, land within that envelope grew from about 4.8 million to 124.7 million square kilometres.

The authors attribute 103% of the net increase to expansion of the envelope itself, interpreted as technological progress. That exceeds 100% because other changes offset part of the gain, principally sea-level rise at −3.3%. These estimates describe potential climatic suitability, not actual cultivated acreage, and the technological interpretation is indirect.

The overall picture is therefore not simply that a warmer world produced farming. Seasonal pressures, population, neighboring farmers, and adaptation all mattered. Yet incomplete archaeology, uncertain population estimates, simplified inundation modelling, missing soil and prey data, and limited climate-model diversity leave substantial uncertainty. The research sharpens the explanation of agriculture’s rise without claiming to settle its causes.

Created on 02 Nov. 2026

Was this synthesis useful?

Score: 0

Ask questions about this paper to our AI assistant

How to chat with this paper? Illustrated guide

Compare with the PDF

You can also chat with multiple papers at once here.

AI assistant instructions?

Tools for this paper

Explain it differently 3 free rewrites left today

Save to a collection Group papers in collections to compare them, chat with them and export them. Create a free account Log in

Export BibTeX · RIS (Zotero, Mendeley, EndNote) · Obsidian (Markdown) · Notion (Markdown)

Advanced tools Premium

For researchers who follow a field closely: citation alerts, journal-club slides, version comparison and Zotero export. Discover Premium

Zotero

Citation alerts Citation alerts guide

Who works on it? The people and institutions who publish on its topic

Weekly letter Presented in "Economics and finance this week" of 2 November 2026. Economics and finance this week: When the measuring tool changes the answer

Journal club Premium+ Slides to present it Propose for the journal club

Versions What changed between versions

Talk about it in public Premium+ Press kit: release, questions, hype check Press kit guide

Disclaimer: The AI-based summarization tool and virtual assistant provided on this website may not always provide accurate and complete summaries or responses. We encourage you to carefully review and evaluate the generated content to ensure its quality and relevance to your needs.

Climate change, sea level rise, and the agricultural revolution

The essentials

Question

Method

Findings

Limitations

Why it matters

Read further

Definitions

Why Did Farming Begin? Climate, Rising Seas, and the Agricultural Revolution

Reconstructing a Prehistoric Transition

Seasonality Matters More Consistently Than Warming

Population Pressure and the Power of Neighbors

Did Rising Seas Push People Into Farming?

Could Agriculture Have Begun on Submerged Land?

Technology Expanded Farming’s Possibilities

Ask questions about this paper to our AI assistant

Tools for this paper

Advanced tools Premium

Papers around this one

Closest papers

Prior work it builds on

Go further

Cite this paper (BibTeX)