Can public policies change risk aversion?

Macroeconomics - Economic growth - Monetary Policy
Fiscal Policy - Public and Welfare Economics

Titling program lead to a significant reduction in risk aversion

Attitudes towards risk can influence human behavior and determine economic outcomes such as investment decisions and labor supply.  There is growing evidence, however, that risk attitudes are not fixed but can be shaped by several events such as natural disasters, financial crises, and violent conflict. [1]

 In a recent paper (Aragón et al., 2020), we use a large urban titling program in Peru to test whether public policies can affect risk aversion. Our findings show that this program significantly reduced risk aversion. The effects are sizeable and seem to be driven by the security of tenure and changes in perceived background risk (i.e. non-insurable, unavoidable risk).

The Peruvian urban titling program issued over 1.2 million property titles to urban squatters and is considered one of the largest programs of this kind in the developing world. Before its implementation, obtaining a title was a slow and costly process, and informality rates were high. Previous studies have found significant effects of this program on labor supply and residential investment, among other outcomes (Field, 2005; Field, 2007). According to their findings, these effects seem to be driven by increased security of tenure rather than better access to credit.

Our empirical analysis uses a novel household dataset with questions designed to elicit risk preferences based on the methodology developed by Barsky et al. (1997). Similar to Field (2007), our identification strategy exploited the timing in the implementation of the program through a difference-in-difference approach. Specifically, we compared the difference in risk aversion between eligible and non-eligible households in areas reached by the program relative to areas yet to be reached.

Our main results provide convincing evidence that the titling program reduced risk aversion. The effect is economically significant: the proportion of individuals with high-risk aversion (the most common category, with around 73% of the sample) decreased by almost 10 percent. This represents a rise in risk tolerance of 13% relative to the average value.

The magnitude of this result is comparable with the estimates documented for the effect of natural disasters on risk preferences. For instance, Hanaoka et al. (2017) find a reduction of 8.1 percent in their measure of risk aversion after the Great Japan earthquake, while Samphantharak and Chantarat (2015) document an increase of almost 10 percent in risk aversion after the Thai 2011 mega flood.

The impact of titling was driven by a change in the distribution of risk attitudes. Specifically, we classified respondents into four categories based on their risk aversion ranked from low to high. Figure 1 shows that the program reduced the probability of having low-risk tolerance (the most common category) by almost 9% and increased the probability of being in the highest category by around 7%.

These effects also seem to be persistent. Two findings support this interpretation. First, the individuals in our sample were surveyed 3-7 years after receiving a property title. Thus, our main results are already purged from short-term effects. Second, we cannot rule out that the impact diminished over time. The reason is that the magnitude of the program’s impact between beneficiaries that received the program early on and those who did closer to the time of the survey was similar.

In general, our results are robust to the use of different measures of risk aversion. Our findings also hold under alternative specifications, such as adding a rich set of control variables and district fixed effects, controlling for individuals’ expectations to receive a title, or using an alternative identification strategy comparing only individuals with and without a title. Similar to Hanaoka et al. (2017), we also find that most of the change occurs among men.

Changes in risk preferences appear to be driven by reductions in background risk

We examined potential mechanisms that could explain why the titling program reduced risk aversion. Due to data limitations, our approach consisted of ruling out explanations. Hence, these findings should be interpreted as suggestive evidence. One possible explanation is that the titling program affected levels of income or wealth. This could happen, for example, if having a formal title increases market employment, property values, facilitates access to credit, or fosters productive investment. Under certain risk preferences, an increase in income would lead to a reduction in risk aversion. To test this, we included additional measures of self-reported monthly expenditure, hours worked, and perceived change in house prices after obtaining a formal title, as well as proxies for the expected increase in income after the program to our baseline model. Interestingly, our findings remained roughly similar, which suggests that the impact was driven by another mechanism.

An alternative explanation is that a reduction in background risk could lead to a decrease in risk aversion (Gollier & Pratt, 1996; Eeckhoudt et al., 1996). Intuitively, when individuals assess the risk of a particular decision, they take into account the overall level of risk to which they are exposed. While not conclusive, several observational and experimental studies document a positive relation between background risk and risk aversion (Heaton and Lucas, 2000; Guiso and Paiella, 2008; Lin, 2009; Cameron and Shah, 2015; He and Hong, 2018).

In our context, a reduction in background risk could occur for at least two reasons. First, a formal property title could reduce the perceived risk of expropriation. For instance, Field et al. (2007) documented that granting property titles in Peru led to a sizable reduction of a (perceived) risk of eviction. Second, having a title could increase the perceived likelihood of accessing credit. In turn, this increased access could allow new coping strategies against unexpected income shocks (like job loss).

Due to the unavailability of data on perceived background risk, we did not examine these hypotheses directly. However, we examined whether the program had heterogeneous effects based on beneficiaries' beliefs regarding the security of tenure and access to credit. Our findings show that the impact of the program on risk aversion was only statistically significant for people who believed their security of tenure would improve. While far from conclusive, these results suggest that changes in background risk could, in part, explain why titling leads to a reduction in risk aversion.

A new direction for policy design and evaluation

Our findings raise provocative questions regarding the potential for policy interventions to shape economic outcomes not only by changing the economic environment but also attitudes and preferences. This result adds to previous evidence that property reforms may affect market beliefs, as documented in Di Tella et al. (2007). However, it is possible that many different programs also affect preferences in ways that previous research has not accounted for. We believe that a better understanding of these mechanisms is relevant for more effective policy design and evaluation.

1. See for instance Paet et al. (2014), Cameron and Shah (2015), Hanaoka et al. (2017), and Jakiela and Ozier (2016).


Aragon, F. M., Molina, O., & Outes-Leon, I. W. (2020). Property rights and risk aversion: Evidence from a titling program. World Development, 134(105020).

Barsky, R. B., Juster, F. T., Kimball, M. S., & Shapiro, M. D. (1997). Preference Parameters and Behavioral Heterogeneity: An Experimental Approach in the Health and Retirement Study. The Quarterly Journal of Economics, 112(2), 537-579.

Cameron, L., & Shah, M. (2015). Risk-taking behavior in the wake of natural disasters. Journal of Human Resources, 50(2), 484-515.

Chuang, Y., & Schechter, L. (2015). Stability of experimental and survey measures of risk, time, and social preferences: A review and some new results. Journal of Development Economics, 117, 151-170.

Di Tella, R., Galiani, S. F., & Schargrodsky, E. S. (2007). The Formation of Beliefs: Evidence from the Allocation of Land Titles to Squatters. Quarterly Journal of Economics, 122, 209-241.

Eckel, C. C., El-Gamal, M. A., & Wilson, R. K. (2009). Risk loving after the storm: A Bayesian-Network study of Hurricane Katrina evacuees. Journal of Economic Behavior & Organization, 69(2), 110-124.

Eeckhoudt, L., Gollier, C., & Schlesinger, H. (1996). Changes in Background Risk and Risk-Taking Behavior. Econometrica, 64(3), 683-689.

Field, E. (2005). Property rights and investment in urban slums. Journal of the European Economic Association, 3(2-3), 279-290.

Field, E. (2007). Entitled to work: Urban property rights and labor supply in Peru. The Quarterly Journal of Economics, 122(4), 1561-1602.

Gollier, C., & Pratt, J. W. (1996). Risk vulnerability and the tempering effect of background risk. Econometrica, 64(5), 1109-1123.

Hanaoka, C., Shigeoka, H., & Watanabe, Y. (2017 (forthcoming)). Do risk preferences change? Evidence from panel data before and after the great east Japan earthquake. American Economic Journal: Applied Economics.

Jakiela, P., & Ozier, O. (2016). The Impact of Violence on Individual Risk Preferences: Evidence from a Natural Experiment. IZA DP No. 9870.

Page, L., Savage, D. A., & Torgler, B. (2014). Variation in risk-seeking behaviour following large losses: A natural experiment. European Economic Review, 71(C), 121-131.

Samphantharak, K., & Chantarat, S. (2015). The effects of natural disasters on household's preferences and behaviours: Evidence from Thai farmers during and after the 2011 mega flood. Disaster Risks, Social Preferences, and Policy Effects: Field Experiments in Selected ASEAN and East Asian Countries, 57-84.

Voors, M. J., Nillesen, E. E., Verwimp, P., Bulte, E. H., Lensink, R., & Van Soest, D. P. (2012). Violent conflict and behavior: A field experiment in Burundi. American Economic Review, 102(2), 941-964.

Share this