Essays in economic fluctuations: The effects of time-to-build and bottlenecks

This dissertation consists of three chapters (essays) in economic fluctuations. The first essay presents a time to-build model which is an alternative to the standard technology shock real business cycle model. The time-to-build model is appealing because it incorporates the time required to produce, deliver and install capital goods. Using both Solow and Cass-Koopmans versions of the time-to-build model, I prove that they produce endogenous cycles. In addition, I show that for reasonable parameter values, these models generate empirically relevant cyclic variations in output. Furthermore, time-series analysis shows that delivery lags on capital good—a proxy for production lags—have statistically significant explanatory power in a regression on industrial production.; The second essay incorporates the time to build into a model with adjustment costs. The bottleneck causes the price of capital rises when production lags increase. I show that this produces a concave growth curve as in a basic neoclassical growth model. Importantly, countries that have a higher bottleneck have a lower level of steady state income than countries with a lower bottleneck. As long as there are different bottlenecks across countries, per capita income convergence will not obtain. Therefore, this model can provide an answer for the lack of convergence observed in the data.; The third essay evaluates role of the bottleneck in causing economic fluctuations. I endogenize the bottleneck in this chapter and show that bottlenecks have a substantial impact on growth especially in countries with large capital stocks. Based on this model, a bottleneck series is generated from the data. This is an important step since the bottleneck data series was not, previously available. Using this data, a time series model for output variations is built using the bottleneck as an independent variable. This essay also examines time series models that built based on technology shocks and past output. All three models are used to forecast output and are compared to each other. I find that of the three models, output is most accurately forecasted using the bottlenecks. I conclude that bottlenecks are an important source of business cycle fluctuations.