Does time preference affect smoking behavior? A dynamic panel analysis

I analyze the effect of time preference on smoking behavior while controlling for the effect of past smoking behavior, using a dynamic panel analysis. I use nationwide micro panel data from Japan and make three main findings. First, as in the previous literature, without controlling for the effect of past smoking behavior, I find positive associations between time preference and smoking behavior, with the association significant for male. Second, by using a dynamic panel analysis, I find that time preference affects smoking behavior significantly for female but not for male. Third, male smokers who have a low discount rate are more likely to want to quit or reduce smoking behavior, whereas female smokers with a low discount rate are no more likely to want to quit or reduce smoking behavior. This result implies that while male smokers are concerned about the future consequences of smoking, they cannot manage to quit.