Measuring technological change using a latent variable approach
提出一种潜变量结构方程方法，将参数生产函数与离散全要素生产率指数结合，用于测量美国农业技术变化，发现估计的技术指数年增长率高于传统方法。
The state of technology is an unobservable variable in transformation functions. The study introduces a latent variable structural equation approach to modelling technological change in US agriculture. The proposed approach combines the parametric production (cost) function and the discrete total factor productivity index approaches together and minimises measurement errors. The technological change for the US agriculture is assumed to be caused by agricultural research, extension, and the farmers' educational backgrounds. The results are compared with the conventional total factor productivity measurement and show that the estimated technology index has a higher annual growth rate than the discrete total-factor-productivity index estimates