Projecting county pulpwood production with historical production and macro-economic variables
We explored forecasting of county roundwood pulpwood produc-tion with county-vector autoregressive (CVAR) and spatial panelvector autoregressive (SPVAR) methods. The analysis used timberproducts output data for the state of Florida, together with a set ofmacro-economic variables. Overall, we found the SPVAR specifica-tion produced forecasts with lower error rates compared to CVARspecifications. Nonetheless, high forecast errors across countiesrevealed the uncertainty associated with projecting volumes ofcounty pulpwood production.Published by Elsevier GmbH. on behalf of Department of ForestEconomics, Swedish University of Agricultural Sciences, Umeå.