This paper attempts to shed new empirical light on infrastructure's role in economic growth using a semiparametric smooth coefficient model to avoid specification problems in some existing studies and admit infrastructure-induced nonlinearity and parameter heterogeneity. Estimated by a three-step procedure that controls for endogeneity in both the regressors and the environmental variable (infrastructure), the model is applied to the empirical context of telecommunications infrastructure in a fast-growing economy, China. The results reveal that telecommunications contribute to output through various sources, namely its neutral and non-neutral impacts. The total/net effect is positive but largely decreases with telecommunications stocks. (JEL L96, O4, O1).
|Number of pages||17|
|Early online date||25 Feb 2019|
|Publication status||Published - 1 Jul 2019|
Research Beacons, Institutes and Platforms
- Global Development Institute