4.2 Estimation Results
The statistical results for the preferred models for domestic and export demand are shown in
Table 1. Domestic demand was estimated using 2SLS because of the endogeneity of the price of cotton.
Because first-order autocorrelation in the residuals was found, 2SLS was applied with correction for first-
order autocorrelation in the residuals using the two-step procedure developed by Hatanaka (1976).
Overall, the results seem quite reasonable and suggest a strong and significant impact of promotion and
research on mill consumption of cotton. There is significant seasonality in mill consumption as indicated
by highly significant monthly dummy variables (not included in table). The own-price elasticity of
Table 1. Regression Results for Monthly Demand for U.S. Cotton, 1986-2000
Domestic Demand |
Export Demand | ||||||
Independent |
2SLS and First-Order |
Independent |
2SLS and First-Order | ||||
Para-meters |
t-values |
Elasticity |
Parameters t |
-values |
Elasticity | ||
CONSTANTt |
1.75181 |
2.07 |
CONSTANTt |
338.440 |
0.77 | ||
PCOTTONt |
-0.01089 |
-3.21 |
-0.413 |
EXPORTSt-I |
0.601 |
9.27 |
0.601 |
PPOLYt |
-0.00361 |
-1.65 |
-0.129 |
PCOTTONt |
-6.757 |
-1.98 |
-0.692 |
PRAYONt |
0.00261 |
1.50 |
0.137 |
PPOLYt |
0.748 |
0.31 |
0.072 |
DTEXWAGEt |
-0.13169 |
-0.87 |
-0.453 |
WPCOTTONt |
7.266 |
2.23 |
0.732 |
WPCOTTONt |
0.01264 |
4.08 |
0.427 |
DECIt |
-0.389 |
-0.13 |
-0.037 |
DECIt |
-0.00723 |
-2.14 |
-0.256 |
FGDPt |
0.016 |
0.52 |
0.211 |
DPIt |
-67879.7 |
-0.87 |
-0.616 |
ROWSTKt |
-3.790E-09 |
-0.57 |
-0.103 |
FGDPt |
0.000061 |
0.79 |
0.309 | ||||
SAGPROMt |
2.12E-08 |
2.00 |
0.023 | ||||
SAGNARESt |
5.12E-07 |
4.72 |
0.152 | ||||
SAGNARESt-1 |
7.30E-08 |
0.68 |
0.022 | ||||
SAGNARESt-2 |
2.79E-07 |
2.75 |
0.083 | ||||
SAGNARESt-3 |
3.16E-07 |
3.06 |
0.094 | ||||
rho |
0.19303 |
2.62 |
rho |
0.07600 |
-0.53 | ||
N |
176 |
N |
178 | ||||
R2 |
0.7990 |
R2 |
0.7540 | ||||
R |
0.7671 |
R |
0.7290 | ||||
DW |
2.0318 |
DW |
1.8939 |
20