This study is an exploration of the relationships between income, demographic pressure, technological change in agriculture, and the structure of political economies in light of cross-country differences in deforestation. The study focuses on small farmers and shifting cultivation. The analysis is based on a model developed by Larson (1994) that accounts for rural poverty, rootlessness, and distribution of landholdings. Regression equations model the average annual rate of deforestation, the relative area under forests, and a recursive model that includes both the deforestation rate and the forested area. Deforestation was reasonably well explained by a dummy variable for Asia, a rank order variable of the amount of forested area in 1980, the gross domestic product per capita in 1990, the average annual population growth rate during 1981-90, and the percentage increase in value added to agriculture during 1981-90 in 1990 dollars. Findings indicate that a 10% increase in the population growth rate increased the rate of deforestation by 10.6%. A 10% increase in income per capita increased deforestation by 49.5%. The influence of income on deforestation followed Kuznet's U-shaped curve. The turning point for reduced deforestation was income of $3500 per capita. Only Central and South America are near this income level. An increase in 1 agricultural worker per household increased deforestation by 50%. A 10% increase in smallholders' share of agricultural land reduced deforestation by 3.4%. Countries with high rural rootlessness had 23.6% less relative area under forests, suggesting that rural rootlessness rather than poverty per se leads to deforestation. The recursive model shows that demographic pressures led to deforestation and were mediated by technological change. Political economy theories of deforestation received strong empirical support.

中文翻译：

---

贫穷国家的鹳、犁、农村社会结构和热带森林砍伐？

本研究探讨了收入、人口压力、农业技术变革和政治经济结构之间的关系，鉴于跨国森林砍伐的差异。该研究侧重于小农和轮垦。该分析基于 Larson (1994) 开发的模型，该模型解释了农村贫困、无根和土地所有权分布。回归方程对平均年毁林率、森林下的相对面积以及包括毁林率和森林面积的递归模型进行建模。亚洲的一个虚拟变量、1980 年森林面积的排序变量、1990 年人均国内生产总值、1981-90 年平均年人口增长率、以及 1981-90 年农业增加值的百分比增长（以 1990 美元计）。调查结果表明，人口增长率每增加 10%，森林砍伐率就会增加 10.6%。人均收入每增加 10%，森林砍伐就会增加 49.5%。收入对森林砍伐的影响遵循库兹涅特的 U 形曲线。减少森林砍伐的转折点是人均收入 3500 美元。只有中美洲和南美洲接近这个收入水平。每户增加 1 名农业工人，森林砍伐就会增加 50%。小农在农业用地中所占份额增加 10%，森林砍伐减少了 3.4%。农村无根率高的国家森林下的相对面积减少了 23.6%，这表明农村无根率而不是贫困本身导致了森林砍伐。递归模型表明，人口压力导致森林砍伐，并受到技术变革的调节。森林砍伐的政治经济学理论得到了强有力的实证支持。