consumption in the US, the spatial SUR models are run separately for the whole time period, the recession period from 2009 to 2012 and the compared time period from 2005 to 2008, and the results are shown in Tables 1e6, with t-statistics in parentheses. There are strong direct spatial dependences among biomass consumption for 2000e2012 (Table 1), implying that biomass
consumption in one state will interact directly with biomass consumption in other states. Except for the direct spatial dependence in 2001 which is negative, all others are positive, indicating that states with relatively high (low) biomass consumption are located close to other states with relatively high (low) biomass consumption more often than it would be observed, if their locations were purely random. There is an obvious division in the direct spatial dependences for 2000e2012. The direct spatial dependence before 2009 is smaller than that after 2009. This may imply that the interdependence among biomass consumption between states tends to be stronger for the recession period. The strong indirect spatial dependences among biomass consumption for 2000e2012 are also shown in Table 2, indicating that