One method is a simple time-series extrapolation or regression against time. These extrapolations may be by

(1) constant increments (collections increased by $5,000 in each of the last five years, so they are estimated to increase by $5,000 this year);

(2) constant percentage change (collections increased by 5 percent in each of the last five years, so they are estimated to increase by 5 percent this year);

(3) simple growth models using the average annual compounding formula developed in Chapter 4; and

(4) linear or nonlinear time trends in which revenue for the budget year is estimated as an arithmetic function of time (R a bt) or as a logarithmic function of time (lnR a bt), where R equals collections from the revenue source and t equals a time index, choosing between the trends according to which is judged most likely to produce a reasonable estimate.