Investment in agriculture is relatively complex in IFs, because changes in investment are the key factor that allows us to clear the agricultural market in the long term. It is very similar to investment in energy, except that we do not need to compute type-specific investments—capital in agriculture is only used for the production function of crops.

We calculate a total agricultural investment need (INAG) to take to the economic model and place into the computation for investment among sectors. This investment involves multiple factors. These begin with the rate of investment within GDP of the previous year applied to the GDP of the current year, adjustment factors related to domestic and global crop stocks, and changes in the ratio of global crop demand to global GDP. This is expressed as

where

mulwst and mulst are adjustment factors related to global and domestic crop stocks, respectively. Both use the ADJSTR function described earlier. For mulwst, the values for the effects of the gaps between existing and desired stocks, and for the change in stocks, are hard coded with values of -0.3 and -0.9, respectively. For mulst, these parameters are hard coded with values of -0.2 and -0.4, respectively.

WAGDEM is the total global demand for crops

As an initial check against too rapid of a shift in demand for agricultural investment, INAG is not allowed to increase by more than 30 percent or decrease by more than 25 percent from the actual investment in the current year. A second check ensures that the demand is no less than 0.5 percent and no greater than 40 percent of current agricultural capital (KAG).

At this point the country-specific multiplier aginvm can boost or reduce INAG. One final check ensures that as long as GDP in the country is larger than it was in the first year, the demand for agricultural investment is not allowed to decline at an annual rate of more than 1 percent per year from the first year.

Investment need (INAG) then enters the economic model, which returns an adjusted value that feeds into further calculations in the agriculture model.