[R-sig-Geo] Fitting hierarchical spatial panel data models

[Denys Dukhovnov]

Hello,

I am aiming to run an SLX or and SDEM local spillover type spatial model and I am looking for advice on existing methods and their implementations. 

I have a panel data with about 30 years of observation and nested spatial structure (such as US counties within US states). I would ideally like to account for both the hierarchical structure of the data as well as the time component. 

I understand that hierarchical spatial econometric models are well established (for example, as per Lacombe & McIntyre (2016)), but I struggle to find any prior reference that combines the nested/hierarchical structure with spatial panel data with a longitudinal component.

I am curious if someone could point me to where I might want to look for a solution and any existing software implementation in R. Alternatively, ignoring the temporal component, is it possible to specify a multi-level model within the splm R package for SDEM type process? 


Many thanks in advance.

Best,
Denys