Agriculture and modern industry are different from each other in many aspects. Here I use the supply-demand analysis as below.
On one hand, the supply of agricultural productions, such as rice and corn, is heavily relied on environmental conditions like temperature, rain and wind etc. On the contrary, the supply of industrial productions is always steady, or it is only determined on the input of raw materials. Although technology is vital, we can ignore it here since it plays an important role on both (or it will be considered later because it will also influence the degree of uncertainty).
On the other hand, the demand of food (agricultural productions) is always stable, or we can say that the price-elasticity is small, while the demand-elasticity of industrial productions is much bigger. Thus, the most interesting thing is that there is significant difference between them both in the supply and demand side, and if we also take the service-industry into account, this question will become more and more complex and interesting.
Under the traditional assumptions, we usually consider the industry as the typical model, then establish a general equilibrium model. It is reasonable because modern industry make the largest part of GDP. But at the same time, since the uncertainty of output is very limited, there is no need to waste too much on supply analysis. Therefore, we usually ignore the role of uncertainty which do exist in the real world.
Here, I wonder that if we apply this method to the economic analysis of developing or undeveloped countries, where agriculture is still very important in national production, there may be some unrealistic expectations derived from this simplified model.
So if possible, can we make a consideration of uncertainty as well? Maybe a simple way is to assume that var(output) doesn't equal 0. I don't know how to go further, but I think the results will be certainly interesting, especially in the price-determination and equilibrium, and may also be meaningful for policy-makers.