Persistence of Shock in a Time Series

Abstract

Services provided by Bangladesh Railway (BR), a popular mode of transportation, have been vital to the growth and development of the economy since their inception. However, the declining share of its revenue to the gross domestic product (GDP) remains a major policy concern which needs to be addressed. Public policy, particularly short-term stabilization policy, aimed at dealing with this concern would be effective if the series (share of revenue to GDP) is trend stationary. However, if the series is found to be stochastically non-stationary, such stabilization policy would be ineffective and would signal the need for long-term restructuring and reforms. Given this perspective, the article examines if the relevant time series (in this case, railway revenue as per cent of GDP) contains a deterministic or stochastic trend to assess whether short-term public stabilization policy would be effective or long-term structural adjustments and reforms would be necessary. In addition, the article also provides a measure of persistence in the above mentioned time series following Cochrane (1988). Loss of railway’s share appears to be due to long-term shifts in demand due to inter-modal competition as well as supply and capacity constraints.