Skip to main content

Table 2 Sample selection model estimates of seeing a GE label and use of label information

From: How consumers use mandatory genetic engineering (GE) labels: evidence from Vermont

Variable Binomial probit estimate: saw label Marginal effects of estimates
Did not use label Information cue and used label Preference signal and used label
Income $50,000 or greater 0.070** (0.031) 0.008 (0.145) − 0.025 (0.211) 0.016 (0.168)
“GMO” is transfer of genes that would not occur in nature 0.033 (0.031) − 0.078 (1.904) 0.100 (0.663) − 0.021 (0.267)
Spring data collection − 0.025 (0.030) − 0.090 (0.168) 0.060 (0.138) 0.029 (0.286)
Female − 0.029 (0.030) − 0.168** (0.071) 0.179** (0.066) − 0.010 (0.409)
Bachelor’s/professional degree 0.049 (0.031) 0.005 (2.758) − 0.059 (0.241) 0.053 (0.110)
Family with children − 0.012 (0.034) − 0.059 (0.277) 0.086 (0.190) − 0.026 (0.170)
Age − 0.004*** (0.000) − 0.003 (0.003) 0.005* (0.002) − 0.001 (0.002)
Oppose or strongly oppose GE − 0.019 (0.042) − 0.332*** (0.065) 0.441*** (0.052) − 0.108** (0.050)
Support or strongly support GE 0.027 (0.047) 0.365*** (0.096) − 0.269* (0.159) − 0.095 (0.105)
Active search dummy 0.245*** (0.050) − 0.470*** (0.072) 0.456*** (0.086) 0.013 (1.396)
Passive search dummy 0.131*** (0.039) − 0.437*** (0.064) 0.351*** (0.096) 0.085 (0.158)
Consults product information on front of packaging (low fat, reduced calorie, etc.) − 0.094*** (0.032)    
Consults ingredient list 0.026 (0.045)    
Consults nutrition information 0.004 (0.044)    
Looks for labeling indicating food is organic 0.023 (0.037)    
Looks for labeling indicating food is “natural” 0.073** (0.034)    
Estimated probability at data means   .36 .49 .15
  1. N = 942 respondents. Standard errors in parentheses. Multinomial logit with selection computed marginal effects at the mean
  2. ***, **, *Indicate significance at the 1%, 5%, and 10% level