Supply response to countercyclical payments and base acre updating under uncertainty: an experimental study.

(25) The Lagrange Multiplier test (p-value less than 0.005) indicated that random-effects model is preferred to the classic regression model (Greene 2000). The fixed-effects model cannot include the treatment or the risk preference variables due to perfect colinearity with the individual intercepts. The Hausman test was conducted on the above model excluding T2, RAVER, and RLOV. The Hausman (1978) test (p-value of 1.00) indicated no significant difference between the fixed and random-effects models, which suggests that the random-effects model may be preferred since it is efficient (Greene 2000). The fixed-effects model yielded similar results with slightly lower case coefficients (0.0472 for CCP case and 0.0647 for Policy risk case) and lower significance levels (p-values 0.018 and 0.060). The two-way fixed effects model was not an option since there is perfect colinearity between the time periods and cases.

(26) This linear wealth term may be simple; we ran other specifications of wealth effects such as a squared term and interaction terms with the risk preference variables without improvement to the model.

(27) The risk-averse classification in the X-test was used for any player taking the $2.50 sure bet in all games, chose the sure bet in games 1-7, 1-8, or 1-9, or chose the sure bet in at least seven of ten games. A player was risk loving if they never took the sure bet, played the lottery in games 1-9, 2-9, or 3-9, or chose to play the lottery in at least seven of ten games. All players not in either of these two categories were considered risk neutral; this category contained a majority of the players.

(28) The United States has not yet notified the WTO of which category of domestic support CCPs would be placed.

Christopher R. McIntosh is assistant professor, Department of Economics, University of Minnesota-Duluth, Jason F. Shogren is Stroock Professor of Natural Resource Conservation and Management, Department of Economics and Finance, University of Wyoming and the King Carl XVI Gustaf's Professor of Environmental Sciences, Umea University, and Eric Dohlman is agricultural economist with the Field Crops Branch, Market and Trade Economics Division, Economic Research Service, USDA.

We thank the ERS/USDA for financial support on cooperative agreement #43-3AEK-2-80080. Helpful comments were received from reviewers and presentations at University of Minnesota-Duluth, Penn State, and ERS. Shogren thanks the Norwegian University of Life Sciences for the hospitality needed to finish this paper. All errors remain our own. The views expressed here are those of the authors, and may not be attributed to the Economic Research Service or the U.S. Department of Agriculture. Table 1. Lottery Probabilities, Prices, Expected Values per Token, and Variances per Toke Used in the Experiment Lottery Prob ZP Prob LP Prob HP ZP

Program Crop (Blue) 1 0.1 0.6 0.3 0 2 0.1 0.4 0.5 0 3 0.1 0.4 0.5 0 4 0.1 0.5 0.4 0 5 0.1 0.6 0.3 0 6 0.1 0.4 0.5 0 7 0.1 0.4 0.5 0 8 0.1 0.4 0.5 0 9 0.1 0.3 0.6 0 10 0.1 0.5 0.4 0

Nonbase Crop (Red) 1 0.1 0.55 0.35 0 2 0.1 0.4 0.5 0 3 0.1 0.5 0.4 0 4 0.1 0.4 0.5 0 5 0.1 0.4 0.5 0 6 0.1 0.4 0.5 0 7 0.1 0.5 0.4 0 8 0.1 0.5 0.4 0 9 0.1 0.5 0.4 0 10 0.1 0.3 0.6 0

EValue/ Var/ Lottery LP HP Token Toke

Program Crop (Blue) 1 13 17 12.9 22 2 11 19 13.9 36 3 13 19 14.7 32 4 12 18 13.2 27 5 13 18 13.2 24 6 12 17 13.3 25 7 8 17 11.7 33 8 12 21 15.3 44 9 11 32 22.5 144 10 12 18 13.2 27

Nonbase Crop (Red) 1 12 18 12.9 26 2 12 21 15.3 44 3 10 28 16.2 101 4 11 23 15.9 60 5 10 24 16 72 6 13 14 12.2 17 7 10 14 10.6 16 8 13 15 12.5 18 9 18 23 18.2 42 10 14 15 13.2 20 Note: ZP = Zero price; LP = Low Price; HP = High Price; EValue = Expected Value; Var = Variance. Table 2. Lottery Expected Value/Variance Test and Bonus Schedule for Each Lottery

Expected Lottery Value Variance 1 Same Blue < Red 2 Blue < Red Blue < Red 3 Blue < Red Blue << Red 4 Blue << Red Blue < Red 5 Blue < < Red Blue < < Red 6 Blue > Red Blue > Red 7 Blue > Red Blue >> Red 8 Blue >> Red Blue > Red 9 Blue >> Red Blue >> Red 10 Same Blue > Red

Bonus Schedule Lottery BONUS1 BONUS2 BONUS3 1 150 1,350 50 2 150 1,350 250 3 150 1,350 50 4 150 1,350 150 5 150 1,350 50 6 150 1,350 150 7 150 1,350 550 8 150 1,350 150 9 150 1,350 250 10 150 1,350 150 Table 3. Descriptive Statistics of the Dependent Variable for Each Case Case Mean SD Min Max Baseline

Blue proportion 0.4953 0.2916 0 1.00 CCP

Blue proportion 0.5245 0.31 0 1.00 Policy Risk

Blue proportion 0.5236 0.3071 0 1.00 Case Cases Skewness Kurtosis Baseline

Blue proportion 880 0.009 2.0711 CCP

Blue proportion 880 -0.1059 1.8659 Policy Risk

Blue proportion 880 -0.0577 1.966 Table 4. Random-Effects Regression Results

Predicted Std. Variable Coefficient Coefficient Error T2 Zero 0.0130 0.0152 CCPcase + 0.0543 *** 0.0186 Policyriskcase + 0.0792 ** 0.0311 LOTTO2 - -0.235 *** 0.0224 LOTTO3 ? -0.130 *** 0.0224 LOTTO4 - -0.234 *** 0.0224 LOTTO5 - -0.185 *** 0.0224 LOTTO6 + 0.146 *** 0.0224 LOTTO7 ? 0.0780 *** 0.0224 LOTTO8 + 0.170 *** 0.0224 LOTTO9 + 0.129 *** 0.0224 LOTTO10 - -0.0798 *** 0.0224 LDCE - -0.00292 * 0.00166 LHITIND - -0.0149 0.0317 RAVER + -0.0494 * 0.0266 RLOV - 0.0118 0.0249 Constant + 0.537 *** 0.0208 Note: N = 2,640, [R.sup.2] = 0.255. A single asterisk (*) indicates significance at the 10% level, a double asterisk (**) indicates significance at the 5% level, and a triple asterisk (***) indicates significance at the 1% level. There was no statistical difference between one-way and two-way random-effects. Adding heteroscedasticity in either the general or group form did not lead to a significant difference in results. The autocorrelation coefficient was insignificant (estimated RHO = -0.056722) so no correction was made.