The Process of an Innovation Cycle 



137 



Notes: 

 Data are annua! observations (area planted, in dunam [=0.1 hectare!) 

 1961/62-1966/67, except for cucumbers and peppers, for which data were avail- 

 able starting with 1958/59 season. 

 y^ area in year / 



*R^ correlation coefficient between the calculated and observed y values. 

 Estimated by the method suggested by Tintner (1952), pp. 208-11. 

 The equation for the established moshavim did not have a real value solution; 

 the estimates for the pooled data of the total area in the moshav sector are re- 

 ported. 



low for the young moshavim. The table also reveals a higher rate 

 of adoption (the parameters a) for the latecomers— the Arabs, the 

 private growers, and the young moshavim — which suggests a 

 faster rearrangement in positions of static comparative advan- 

 tages than rates of early (experimental) adoptions. A possible ex- 

 planation for this faster rearrangement, suggested by Finis 

 Welch, is that the variance of skill distribution is lower in these 

 sectors, making for a speed-up in the diffusion process in the 

 lower-skilled groups. (This explanation is consistent with the 

 form of n(w) in figure 7.2) 



The diffusion process is completed when the market is in the 

 long-run equilibrium. In addition to the estimates of the sigmoi- 

 dal equation reported in table 7.2, the same function was esti- 

 mated from the same data set imposing projected equilibrium A- 

 values, explicitly recognizing in this way the equilibrating proper- 

 ties of the process (table 7.3). 



