基于三阶段最小二乘法探究诱增交通量

在诱增交通量的预测中，中国现有统计数据在精度和广度上的局限性，容易造成计算结果存在遗漏变量偏误和测量误差偏误；同时变量之间的相关性容易导致计算结果存在联立性问题。为克服这些问题，采用工具变量法中的三阶段最小二乘法探讨道路供给与交通需求之间的关系。基于国家统计数据库数据，选择车辆出行距离、民用汽车拥有量和交通拥堵水平三个内生变量构建联立方程模型，采用三阶段最小二乘法对联立方程进行估计，并给出短期、长期弹性系数的求解方法。结果表明，道路建设会导致车辆出行距离增加，不能解决中国城市交通拥堵问题，而大力发展公共交通可以降低交通拥堵水平。

Induced Traffic Forecasting Based on a Three-stage Least-Square Approach

Due to the limitation in precision and span of the statistical data, there often exists omitted variable bias and measurement error bias in induced traffic forecasting in China. Meanwhile, the correlation between variables may result in multicollinearity issues in calculation results. To contend with these problems, this paper investigates the relationship between roadway supply and traffic demand using a threestage least-square （3SLS） estimation of instrument variables. Based on the data from China Statistical Database, the paper establishes the simultaneous equations with three endogenous variables including vehicle travel distance, civil car ownership, and level of traffic congestion. The simultaneous equations are estimated with the 3SLS and the methods for solving short-term and long-term elasticity coefficients are also discussed. Results show that roadway construction leads to the increase of vehicle travel distances, hence cannot solve the problem of urban traffic congestion in China; instead, the development of public transportation can alleviate traffic congestion.