Sequence and cluster analysis used to identify key spatial trajectories in fuel poverty based on definitions and indicators used in national policy in England. The full code to replicate the analysis and outputs in R is available here.
We analyse sub-regional fuel poverty indicators from 2010 - 2019. In order to analyse sequences within fuel poverty data over time, it is necessary to transform the data into a categorical, rather than continuous dataset. This allows for analysis of how Local Authorities move through different states over time. Relative deciles are selected as an appropriate categorical classification for the dataset, classifying Local Authorities into deciles (i.e. the top 10%, top 10% etc.) for each year between 2010-2019.
LA_variables.csv includes the LA scale contextual variables (e.g. unpaid care, deprivation etc.) and data_deciles.csv includes the raw fuel poverty data (as a percentage) followed by a decile classification for each year. LA_finaldataset.csv includes all variables, fuel poverty data and the cluster assigned.