The weather-information problem faced by settlers of semi-arid regions of the Great Plains hindered their attempts to adapt their crops, techniques, and farm sizes. Episodes of homestead settlement and collapse in western Kansas in 1893–1894 and in eastern Montana in 1917–1921 are examined. A Bayesian learning model indicates how new climate information was incrementally incorporated to revise views of agricultural prospects. Primary data show homesteaders' lagged response to new drought information and illustrate drought's differential impact on small farms. Dryfarming doctrine, despite its optimistic claims, was an imperfect response to drought. Indeed, some dryfarming practices hastened homestead failure.

“No one need be in doubt about the sharp change in climate that occurs somewhere between the 96th and 100th meridians. It can be felt on the lips and skin, observed in the characteristic plant and animal life, seen in the clarity and/or dustiness of the atmosphere, determined by measurements of rainfall and evaporation, tested by attempts at unaided agriculture. Practically every western traveler in the early years remarked the facts of aridity, though not all used the word ‘desert’. …”Stegner, Beyond the Hundredth Meridian, p. 399.