There is a substantial amount of new, good research looking at not just the descriptive behavioral economics of investing and financial risk taking behavior but the actual brain mechanisms driving harmful economic and investing behavior.
These are focused mainly on inherited, congenital neurotransmitter deficits and impairments. These same conditions get worse with age and under stress or other physical ailments.
This study of brain processes is exciting because it offers the promise of understanding the conditions causing the harmful behavior, instead of just concentrating on the symptoms.
Moving from a focus on symptoms to the medical conditions is the mark of scientific progress and treatment potential.
This is a long paper and we have just excerpted some of the text. The full article pdf is attached. Please note this is not a final published version but a working paper. You will find full citations on the pdf.
“Risk preferences are of great practical importance given their relationship with economically significant behaviors such as competitiveness, career choice, savings behavior, and pension choice, among many others. We are only beginning to understand the potential role of variation in specific genes, such as the dopamine gene DRD4, in contexts involving risk preferences. This implies that more studies on DRD4 are merited, as well as the identification of other genes which may influence risk preferences.
Individuals differ significantly in their willingness to take risks. Such differences may stem, at least in part, from individual biological (genetic) differences. We explore how risk-taking behavior varies with different versions of the dopamine receptor D4 gene (DRD4), which has been implicated in previous studies of risk taking.
Discussion And Conclusion
Literature suggests that genetic contributions to individual (biological) differences have substantial implications for economic and behavioral studies. Further, genetic inheritance is a potentially important mechanism to consider when interpreting correlations in preferences between parents and offspring, and when considering determinants of preferences more generally.
This is not to say that heritable genetic factors fully determine behavior; experience and environment clearly matters. But the addition of genetic factors to economic models is highly likely to improve our understanding of behavior, thereby improving our models and increasing their predictive power.
There is a vast literature on risk taking.
- It reports significant heterogeneity in levels of aversion within and across populations
- Risk preferences appear to be a complex and multi-dimensional trait, perhaps explaining why some studies report correlations across risk domains.
- Though a variety of environmental forces, e.g., culture, no doubt contribute to such results, genetic variation may be a strong contributor as well.
This study focuses on variation in the dopamine receptor gene DRD4:
- This gene has previously been related to risk preferences in the economic domain, though with some inconsistent results
- This analysis seeks to deepen our understanding of the 7R+ genotype’s relationship to risk taking by looking at a variety of risk-related activities, with a focus on economic risk taking.
- Significant positive correlation between economic risk taking and general risk taking, and the positive effect of 7R+ among men on economic risk taking
- We find no relationship between 7R+ and self-reported general risk taking
- This latter finding is in line with the results of different genes correlating with economic and psychological risk measures.
This does not imply that risk preferences are unstable, nor that they are context dependent. In fact, Dohmen et al. (forthcoming) find evidence for a single trait operating in the different risk contexts they explore, but with some variation across contexts, perhaps due to differences in risk perception
We find evidence that:
- Individuals with a 7-repeat allele (7R+) of the DRD4 genetic polymorphism take significantly more economic risk in an investment game than individuals without this allele (7R-)
- This positive relationship is driven by the men in our study, while the women show a negative but non-significant result.
- Considering other risk measures, we find no difference between 7R+ and 7R- individuals in general risk taking or any of the risk-related activities.
Overall, our results indicate that:
- The dopamine system plays an important role in explaining individual differences in economic risk taking in men
- But not necessarily in other activities involving risk.
Risk preferences vary substantially across individuals, with women and older individuals. Some of this observed variation has been associated with biological factors. For example, twin studies on Swedish and Chinese twins suggest that genetic differences account for 20% and 57%, respectively, of individual differences in risk preferences in these two nations.
Relatively little is known about the specific genetic determinants of individual variation in risk preferences, although a number of recent studies explore possible associations between specific genetic loci involved in chemical signaling in the brain (neurotransmission) and economic risk preferences.
One neurotransmitter that has received particular attention is dopamine, due to its relation with reward processing in the brain. Activation of the dopaminergic reward pathways, and thus the release of dopamine neurotransmitters:
- Can generate feelings of pleasure and well-being that become associated with the behaviors that triggered the activation
- This makes dopamine a major player in reinforcement of behaviors that are associated with the anticipation of rewards.
Of the genetic markers for dopaminergic function, the dopamine receptor D4 gene (DRD4) has been identified as a candidate for explaining variation in economic behavior, and has received most of the attention in the literature thus far. As with many other genes, DRD4 comes in various versions (“alleles”), which differ among individuals.
There is a specific region of the gene which contains a repeated sequence of DNA base pairs. In different individuals, this sequence is repeated a different number of times (typically 2-11 times) on each of the two relevant chromosomes. The multiple versions of the gene are frequently divided into two dichotomous classes, those with fewer than 7 repeats on both chromosomes (7R- ) and those with 7 or more repeats on at least one chromosome (7R+).
- Functionally, individuals with the 7R+ genotype are putatively less sensitive to dopamine uptake
- Therefore 7R+ individuals require higher levels of dopamine to produce a response of similar magnitude to that of 7R- individuals
- In order for 7R+ individuals to achieve a comparably satiating response in the brain’s corticomesolimbic dopamine reward pathway, they may engage in more stimulating behaviors than do 7R- individuals.
Such genetic variation in response to dopamine may thus contribute to individual differences in those personality and behavioral traits that are associated with the dopamine system. Such traits include:
- novelty seeking
- pathological gambling
- attention deficit/hyperactivity disorder
- behavioral disinhibition
- sexual promiscuity
- and many other behaviors.
Economic risk taking may be another important behavioral trait related to the dopamine system. Four recent studies explore this possibility.
Two of them find a positive association:
- 7R+ men invest significantly more money into a risky investment than do 7R- men. They also examine a second dopamine receptor gene DRD2, and find no relationship between genetic variation in DRD2 and risk taking.
- Kuhnen and Chiao (2009) similarly find a positive relationship between the 7R+ genotype and risk preferences in a laboratory measure. They also find that the serotonin transporter gene 5-HTTLPR helps to predict risk preferences.
However, when the probabilities are ambiguous or when losses are possible, they find that 7R+ individuals do make riskier choices than 7R- individuals, in accord with the other studies.
7R+ men who have been given L- DOPA become more risk taking than 7R- men given the drug.
We also examine the connection between the 7R+ genotype and self-reported general risk taking, as well as behavior in self-reported risk-related activities. We hypothesize that 7R+ individuals will be more risk taking than 7R- individuals on all risk measures.
Our dependent variable for economic risk taking is the amount of money participants put at risk:
- 7R+ individuals are significantly more risk taking than 7R- individual
- However, there is a significantly different effect of 7R+ on men and women, seen by interacting the variable for being a woman and 7R+
- This indicates that 7R+ may have different effects on economic risk taking in men and women in our sample, and we thus pursue the analysis of men and women separately.
Looking Only At Males:
- 7R+ men take significantly more risk than their 7R- counterparts when controlling for age and height.
- The effect is sizeable: 7R+ men take 22% more economic risk than 7R- men
Age is also significantly related ; older individuals take less risk. The effect of 7R+ on risk taking persists when neither age nor height is controlled.
Looking At Women Only:
- The effect of 7R+ is non-significant when controlling for age and height and when neither covariate is included
- The sign of the effect of the 7R+ genotype on economic risk taking is negative in the female sample, the opposite of what is observed in the male sample.
- The lack of statistical significance of among women is not surprising given the very low number of 7R+ women.
Economic vs. General Risk Taking
We now consider general risk taking. There is no significant relationship between general risk taking and our variable stocks/(stocks+bonds) in the male or female subsamples. General risk taking proved to be not related to having started a company, smoking or drinking alcohol for women.
For men, however,the relationship between general risk taking and entrepreneurship is significant and in fact negative.
Given the profound effect of 7R+ on economic risk taking in males, we might expect it to affect general risk taking, as indicated by the response on a 10-point scale to the question:“Are you a person who is generally prepared to take risks or do you try to avoid taking risks?”
Indeed, in our sample economic risk taking was significantly positively correlated with general risk taking. Nevertheless, looking at the whole sample, perhaps surprisingly, 7R+ has no significant effect when controlling for gender, age and height, nor when no controls are included.
We also find no significant effect of 7R+ on general risk taking for either gender, or that the effect differs between the genders.
The only significant predictor of general risk taking is gender, where women take fewer risks.
First we consider economic risk taking. Correlating economic risk taking with stocks/(stocks+bonds),and including both genotyped and non-genotyped participants, the relationship is not significant for either men or women.
We also find no relationships between economic risk taking and entrepreneurship (having started at least 1 company). Note however, that about half of both men and women in our sample have started a company. This is many times higher than the national (US) average.
There are no significant associations between economic risk taking and being a smoker. Looking at the relationship between economic risk taking and drinking alcohol, the relationship is not significant for either men or women.
There is also a significant positive relationship between general risk taking and drinking alcohol in the male subsample.
7R+ And Risky Activities
To explore potential effects of the 7R+ genotype on investment allocations, we look at our variable stocks/(stocks+bonds). In a regression analysis, we find no significant relationship between the 7R+ genotype and investment in stocks and bonds either with or with covariates.
Being female is however negatively associated with more risky, whereas height is positively effects in the expected direction. The interaction effect between 7R+ and gender is not significant, thus we do not look at the male and female samples separately.