Farmers in the developed countries do not hire their workforce or rent machinery and land afresh each day or week because it is more profitable to have longer term arrangements/contracts. Hiring/training and firing/terminating workers, searching/learning to use and refurbishing/returning machinery, and searching/learning to use and returning land to its original condition are all costs over and above a per-unit time rental rate. These costs insure that farmers* demand for most inputs depend not only on current exogenous factors but also on past use and expectations about future use. These are arguments that agricultural input demand functions, at least for the developed countries, are dynamic requiring some time for full adjustment to exogenous economic shocks to occur.

Friesen, Capalbo, and Denny (1992) identify two different approaches to dynamic input demand. First, there are th^ry-based models where dynamics arise from optimal agent behavior. Th^e models have generally taken an adjustment-cost route, e.g., see Lucas 1967a, Nichell 1986, Chambers and Lopez 1984, Vasavada and Chambers 1986, Vasavada and Ball 1988, or resources deterioration with use, e.g., Tegene, Hufliman and Miranowski 1988. Second, data-based dynamic models have been used where dynamics are used to describe the pattern of input use but do not arise from explicitly optimal agent behavior, e.g., see Friesen, Capalbo, and Denny (1992). Both of these approaches have claimed advantages and disadvantages.