POLA PENGGUNAAN HASIL, MANFAAT, DAN FAKTOR PENYEBAB PENJUALAN KOPI PETANI KEPADA TENGKULAK DAN EKSPORTIR DI KECAMATAN PULAU PANGGUNG KABUPATEN TANGGAMUS

This study aims to assess: business performance of coffee middlemen and exporters, coffee allocation and usage patterns total sales proceeds of coffee farmers, the economic benefit when farmers selling coffee to middlemen and exporters, and factors cause farmers decisions determine coffee sales. This research was conducted in two villages, Gunung Megang Village and Tekad Village in Pulau Panggung Subdistrict of Tanggamus Regency. The method of analysis used in this study is an instrument test (validity and reliability), perception (qualitative description), Willingnes to Pay method (WTP), and multinomial logit regression which two models. The research samples were chosen by propotional simple random sampling from four farmer groups of two villages. The results showed that from each of the 65 respondents coffee farmers assessed both of the exporter and middlemen with score 87.7 percent and 75.4 percent. It meant that farmers assessed both exporters and middlemen as the sale of coffee, but more respondents rate the better exporters as evidenced by the value of perceptions. Allocation of coffee farmers production obtained from 65 samples showed that 31 farmers (47.69%) sold most of his coffee to middlemen, 12 farmers (18.46%) equally, and 22 farmers (33.85%) sold most of his coffee to the exporter with an average amount of coffee to middlemen at 565.98 kg and 509.83 kg to exporters from total average production of 1108 kg per farmer and the use pattern result was used for subsistence in daily. Economic benefits obtained when farmers sell to middlemen was Rp212,779.65 per farmer and the exporter was Rp158,367.49 per farmer. Factors affecting farmers decision for determining the portion of sales seen from Likelihood Ratio (LR) indicated that all variables were significant simultaneously, while the individual testing for both logit model logit 1 showed the significant variables were education (X2), and the logit 2 significant variables were education (X2) and experience (X3). Key words: benefit, coffee, multinomial logit, perceptio