Testing persistence of ammonia emissions with historical data from 1770 to 2019 in 37 OECD countries

Abstract

We examine the historical time series data of ammonia emissions from 1770 to 2019 in 37 OECD countries by looking at its statistical properties in order to determine if the series display time trends and persistence. These two properties are very common in environmental data, and our results indicate that reversion to the mean only occurs in the case of Finland, while the null hypothesis of a unit root cannot be rejected in the case of Norway or Iceland. In all the other cases, the estimated value of the differencing parameter is much higher than 1, and this is consistent for the two assumptions made regarding the error term. Thus, shocks are expected to be permanent in all cases except Finland.