Industrial Time Series of Nigeria, 1970-2009: Evolution and Unit Root Testing in the Presence of Multiple Endogenous Structural Breaks

This paper examines the time series properties of eighteen (18) Nigeria industrial variables using annual data from 1970 to 2009. Along with traditional Augmented Dickey-Fuller (ADF) and Phillip –Perron (PP) unit root tests, we use the Lee and Strazicich (2003) minimum Lagrange multiplier (LM) approach to test the null hypothesis of unit root against the break-stationary alternative. ADF and PP tests provide evidence against the unit root null in only 6 (ADF) and 5 (PP) of the variables that challenge these findings. Lee and Strazicich results reveal that contrary to ADF and PP conclusions, all the industrial time series data except one are stationary at the level at 10 percent level of significance. At the same time the Lee and Strazicich LM test identifies endogenously the two most significant structural breaks in the data impacting on both intercept and trend in each and every time series examined. Specifically Lee and Strazicich unit root tests provide evidence against the unit root null in fifteen (15) out of the eighteen (18) industrial time series data at the 5 percent level of significance and seventeen (17) at the 10 percent level during the period under investigation. Thus, failure to account for structural breaks in unit root tests may lead to spurious results and conclusions. We found that the dates of the structural breaks in most cases correspond to (i) the fall in the price of oil in 1981 and the revision of the Fourth National Development Plan 1981-85 as well as the promulgation of the economic stabilization Act or Austerity measures in 1982, (ii) the adoption of structural adjustment programme 1986-88, (iii) the 1988-1994 tariff policy reforms, and (iv) the 1995-2001 policy shift towards measures to promote capacity utilization, increase manufacturing output and grant tax concessions to exporters The policy implications of the results for business cycle theories, econometric modelling, and accurate evaluation of programmes to bring about structural changes are highlighted.