136 Agricuhural Research and Product ivity 



with some farmers very progressive and others far behind. The 

 traditional Arab sector is generally the last to follow. 



The early stages of the adoption of plastic covers were also in 

 line with these general patterns. The innovation was first picked 

 up by several entrepreneurial private farmers who were close to 

 the experiment station, and by a few kibbutzim. Early stages were 

 characterized by experimentation, learning (the right kind of 

 plastic material, optimal timing of treatments, appropriate soils, 

 etc.), and by more than a few disappointments. The following two 

 examples are typical of the kind of activities that form the indus- 

 try's learning process at the different skill levels: (1) A kibbutz 

 that operated a factory for plastic products developed and tested 

 sheeting for covers and a tractor-mounted implement to con- 

 struct the protective tunnels mechanically. (2) A great majority of 

 hired laborers in the private sector were Arab villagers, and on 

 their jobs they had the opportunity to learn by doing and thus 

 could bring the innovation back to their own farms. 



if sectors are homogeneous in skills and other initial endow- 

 ments, then, according to the model, all farms in a sector will 

 adopt the innovation simultaneously. But the sectors, though dis- 

 tinct, are not homogeneous. For example, some kibbutzim have 

 comparatively low labor-opportunity costs. Also, with experience, 

 producers specialized and increased the volume of production, a 

 possibility not considered in the model. Therefore, in terms of our 

 data, the innovation cycle is manifested as shifts in relative sec- 

 torial shares of production. 



Moshavim, both established and young, planted by 1961/62 

 half the covered winter vegetables (table 7.1 , line 8). The share of 

 the kibbutzim had substantially declined by 1966/67, while the 

 shares of the young moshavim and the Arab sectors increased to 

 a level larger than the shares of these sectors in total farm prod- 

 uct.'^ Similar findings are reported in table 7.2, which gives esti- 

 mates of the logistic function parameters (a graphical examination 

 confirmed the logistic time pattern of the diffusion process). The 

 intercept [the parameter kl(\-\-b)\ is zero for the Arab sector and 



12. By 1970/71 ihc young moshavim and Arab sectors shares were 46.3% and 

 25.6%, respectively. 



