Boyd 
Boyd's investment model is an extension of the 
utility model he developed and estimated for the 
harvesting decision. Using microlevel data from a 
survey of landowners in seven North Carolina coun- 
ties, Boyd reports that a landowner's decision to 
undertake any one of several forestry investment 
activities, including reforestation, is affected by 
stumpage prices, size of holdings, and programs of 
financial and technical assistance. Farmers and ab- 
sentee owners are less likely to invest. By compar- 
ing the higher price elasticity of the landowners' 
harvesting decisions to the price elasticity of their 
investment decisions, Boyd notes that landowners 
are more likely to harvest than to invest in response 
to improving prices. This suggests a diminished im- 
portance of market signals as determinants of the 
investment decision. Boyd's estimated equations 
show further that both technical assistance and 
knowledge of cost-sharing increase the likelihood of 
investment. Boyd compares the greater derivative 
(elasticity) for technical assistance with that for fin- 
ancial assistance and concludes that "a govern- 
ment policy which relies more on dissemination of 
technological and market information is probably a 
better means of increasing timber supply than a 
policy involving subsidies." 
Hyberg 
Hyberg (1986), like Boyd, examines both the 
harvesting and reforestation choices of landowners 
using the same survey data from North Carolina and 
additional Statewide survey data from Georgia. Hy- 
berg's objective was to test the appropriateness of 
the utility- and profit-maximization models for har- 
vesting and reforestation decisions. His estimations 
generally support the utility maximization model for 
reforestation, although he concludes that the data 
do not allow a strong rejection of the profit- 
maximization model. In the reforestation model, ex- 
ogenous income (income generated from sources 
other than timber) has a positive but not significant 
effect on reforestation. Stumpage prices have a 
positive but modestly significant effect. Reforesta- 
tion costs have a negative and highly significant 
effect. 
de Steiguer 
The model developed by de Steiguer (1985) 
uses aggregated data rather than survey data to 
examine the effects of stumpage prices, income, 
interest rates, and cost-sharing. De Steiguer's de- 
pendent variable is the number of autonomous (un- 
subsidized) acres of trees planted across the South. 
The estimated model reveals strong positive effects 
of income and interest rates on autonomous invest- 
ment, while the effects of stumpage prices and cost- 
sharing are not statistically different from zero. De 
Steiguer's results, unlike Boyd's, demonstrated an 
income constraint. His finding that interest rates af- 
fect investment behavior suggests a sensitivity of 
landowners to capital markets. The absence of a 
price response reflects a limited market effect, while 
de Steiguer views the absence of a cost-share influ- 
ence as evidence that the substitution of public cap- 
ital for private capital is not a valid contention. If 
substitution had been widespread, the coefficient 
for cost-sharing (when the dependent variable is 
autonomous investment) would have been negative 
and statistically different from zero. The conclusion, 
therefore, is that autonomous reforestation does not 
decline as the result of increased cost-share dollars 
being available. 
Brooks 
Brooks (1985) uses aggregated data similar to 
de Steiguer's but specifies reforestation costs as a 
primary determinant of investment while dropping 
interest rates and income. He also uses expected 
returns rather than current prices as the primary 
market determinant. Like de Steiguer, Brooks found 
minimal effects of market signals on forestry invest- 
ments. He did, however, demonstrate a strong neg- 
ative effect of costs in the South Central States, but 
not in the Southeast. Cost-sharing was found to 
have a strong positive influence on investment lev- 
els throughout the South. 
Romm 
The study by Romm and others (1985) uses 
microlevel survey data from telephone interviews 
with nonindustrial private forest landowners in Cali- 
fornia to examine the effects of the California Forest 
Improvement Act, a 90-percent cost-share program 
for landowners holding less than 5,000 acres of 
83 
