Alexander Pfaff
The Forest in the South and North in the Context of Global Warming (book by WIDER).

PDF link iconHistorical data from the late 19th to the early 20th century are examined for New England. From the attempt to explain the reforestation that occurred, three main land-use claims arise: 1) population clearly does not fully dictate land use (e.g., de- or re-forestation); while population may well have an independent effect on land use, that effect clearly does not dominate all others; 2) factors that affect relative land-use returns, whether “external” to a region or not, clearly do affect land use; two examples are transport costs and productivity of other regions, which affect trade; and finally, 3) long-run analysis must consider shifts even in overall framework, such as from agriculture to migration and industrialization processes involving different economic dynamics. Support for these claims comes from limited historical data alongside relevant theory concerning optimal allocation of land between the four most relevant land uses: agriculture, manufacturing, forest (for timber or as a result of abandonment), and shelter (or, more generally, land uses other than for production). Supporting the “population” claim, previous New England farm expansion flattened out post-1850 and eventual reversed itself, even as population was increasing. Regarding the “returns” claim, the breakpoint in the 1790-1930 series of within-region measures (based on county-level data) of concentration of population is very clearly at about 1830, precisely the era in which the transportation revolution involving railroads, steamships and canals started to have its effect. Concerning the “long-run” claim, given an interest in land use there are grounds for attention to shifts in regional output, such as towards manufacturing from agriculture, as there is evidence that such shifts involved significant changes, in particular concentration of population within particular counties, along rivers and in particular locations along rivers.