Risk aversion is a concept in economics of how people behave when they are exposed to uncertain outcomes. The concept of being risk averse is defined as the propensity to prefer an offer with an expected lower but more certain outcome, compared to an offer with a higher expected outcome with more risk. A more risk averse person is thus less inclined to make an investment with a fairly rewarding outcome if the outcome is uncertain, than a person who is less risk averse. For a poor peasant, this might imply little or no investment in modern production technologies that could lead the peasant out of poverty. Knowledge of risk behavior is thus of importance for public policy making in a pro-growth policy agenda.
Are poor farmers more risk averse than western students? And is the degree of risk aversion in part determined by gender? This thesis examines risk attitudes amongst low-income peasants in the rural highlands of Peru. I will investigate gender specific differences in risk behavior amongst peasants in two regions in the rural southern Peruvian highlands. I will examine whether or not the Peruvian women in our sample are more risk averse than men, with my hypothesis being Peasant women in the Peruvian highlands are more risk averse than men.
The empirical analysis is based on both quantitative and qualitative data. The main method used in this thesis is experimental. By using a risk game with real monetary rewards, conducted in the two regions in the southern Peruvian highlands, I was able to deduce the participants’ degree of risk aversion. The hypothesis will also be addressed by the data generated from both a questionnaire connected with the experiment and a survey of the larger research project Land and Gender in Peru. In addition, I collected qualitative data in the same period and the same area as the experiments of the one region was conducted. The qualitative data collection consisted of observing the participants and their decision-making process during the experiments, conversation with the participants after the game and conversations and interviews with other locals.
In section 2 I give a textbook presentation of the theory of behavior under uncertainty relevant for my thesis. Most of the existing experimental literature on risk aversion and gender find women to be more risk averse than men. This is presented in section 3. However, I also present some papers suggesting no apparent link between gender and risk aversion. There also seems to be a distinction between the findings in the western world and the underdeveloped word. Most of the experimental research on risk aversion in the western world is conducted with a university student sample. As opposed to much of the literature from developing countries, the majority finds a significant gender-specific difference in risk aversion. There also appear to be a relationship between poverty and risk behavior. In section 3.3 I give a short presentation of the theoretical implications of risk aversion and poverty in the theory of poverty traps, postulating that poor peasants are highly risk averse because they are more vulnerable to negative shocks. And since they are more risk averse, they are more reluctant to undertake risky investments, which might push them out of poverty. In section 4, I present the method used to elicit risk aversion, the experimental design. I also discuss briefly some of the shortcomings with the method and the experimental design. The results and the discussion are shown in section 5. The sampling characteristics display the characteristics of the participants, who in short can be summarized as low-educated, low-income peasants with a highly gender segregated labor sphere. Nonetheless, in general they report to be satisfied with both their economic situation and health. The participants risk behavior is consistent with the expected behavior of poor peasants; the majority is either highly risk averse or risk loving. According to my regression analysis in section 5.3, women are significantly more risk averse than men. However, after controlling for principal work activity, there is no statistical evidence of gender-specific differences in risk behavior. Still, the only work activity with a significant effect on risk aversion, is housework. Only a minority of women has housework as their main activity, suggesting that there is still a gender difference with women being more risk averse than men. Furthermore my findings also suggest that men are significantly more extremely risk loving than women, and that this difference might be the driving force of gender differences in risk behavior. All descriptive and econometric analysis has been executed by the use of the statistical software Stata11.