Problem: Alcohol use is decreasing in Russia, alongside a gradual decline in alcohol-related mortality and crime. However, according to the WHO, Russia is still among the top twenty countries in the world with the highest per capita alcohol consumption.
Solution: The social environment is not the only thing that shapes individual attitudes around alcohol. Other factors include non-cognitive, or soft skills, which refer to personality traits and socio-emotional behaviours. However, their impact has not been sufficiently examined or adequately addressed in alcohol policies.
Most studies focusing on the impact of personality traits on alcohol consumption have been based on relatively small samples and disconnected from socio-economic indicators, limiting the validity of their findings. Instead, HSE researchers Ksenia Rozhkova, Sergey Roshchin and Yana Roshchina used data from a national household survey and studied various parameters to measure the impact of non-cognitive skills on the likelihood and volume of individual alcohol consumption in Russia. Their study confirms the role of personality traits in alcohol use and offers new insights into whether educational attainment is really as important in reducing individual drinking levels as previously thought. A preprint of the paper is available on the HSE website.
The study examined the impact of the so-called Big Five personality traits on individual drinking patterns. The Big Five model, introduced by American psychologist Lewis Goldberg, is widely used for personality assessment.
The Big Five model includes the following dimensions of the human personality:
conscientiousness. Refers to characteristics such as perseverance, meticulousness, diligence, commitment and striving for order;
extraversion. Refers to directing one’s energy and interest towards the outside world rather than one's subjective experience (sociability, enthusiasm, emotional warmth, propensity for adventure, and activity);
neuroticism. Refers to emotional (in)stability (anxiety, impulsivity, insecurity, irritability, vulnerability to stress, and hostility);
openness. Refers to creativity, curiosity and good imagination;
agreeableness. Refers to being friendly and willing to compromise (being accommodating, cooperative, altruistic, trusting and modest).
The research was based on data from the 2016–2018 RLMS-HSE, which included information on sociodemographic indicators, non-cognitive skills (based on answers to 24 questions of the Big Five Personality Test), and alcohol consumption (answers to questions about abstinence, preferred drinks, frequency of drinking and amount of alcohol consumed) of 5,300 Russians aged 20 to 60.
The data was analysed using various econometric models, in particular, random-effects probit and tobit models.
The former is a binary choice model used to predict the probability of an event as a function of several variables. The latter (tobit) model is used when there is censoring in the dependent variable, i. e. all values at or below/above some threshold are censored. In this case, the volume of alcohol consumption could not be below zero. A failure to account for this censoring would invalidate the estimates. Random effects models allowed the researchers to account for individual-level heterogeneity in alcohol consumption.
The data was controlled for 19 individual and regional characteristics, ranging from gender, age, ethnicity, marital status and bodyweight to per capita income, alcohol prices and average air temperatures in their region of residence.
Agreeableness was the only characteristic not statistically related to alcohol consumption. The impact of other characteristics varied across the sample and between men and women, but the overall conclusion is that individual personality traits can predict the likelihood and amount of alcohol consumption.
In particular, the more conscientious a person (persistent, hardworking and reliable), the less they tend to drink and the more likely they are to abstain. Since conscientiousness is linked to long-term strategic thinking, conscientious people often have better self-control and are more likely to invest in their health and education and to maintain a healthy lifestyle.
According to the researchers, a one standard deviation increase in conscientiousness is associated with an increase in the likelihood of abstinence of 2.5 percentage points for men and 1.7 percentage points for women and a decrease in the volume of consumption by 38 pp. and 28 pp. respectively.
Compared to people measuring high on conscientiousness, extroverts tend to drink more and people who are open to experience drink less in terms of volume. The effect of extraversion on drinking is opposite to that of conscientiousness. Extroverts enjoy human interaction and are more likely to engage in social activities, including those which involve drinking, and less willing to turn down a drink.
Thus, a one standard deviation increase in extraversion reduces the probability of abstinence by 2.1 pp. for men and 2.7 pp. for women and increases the log volume of consumed alcohol by 27 pp. for men and 24 pp. for women.
In contrast, a one standard deviation increase in openness to experience, which is usually related to novelty seeking, creativity and good imagination, is associated with a 7 pp. reduction in the log of alcohol consumption volume, but only for men. No effect of openness on drinking was found in women. However, neuroticism, associated with changeable moods, insecurity, anxiety and vulnerability to stress, can push women to drink more. A one standard deviation increase in neuroticism increases the log volume of consumed alcohol by 7 pp. for men and 13 pp. for women.
This could be due to observed gender differences in drinking motives. 'While drinking among men is frequently associated with conformity and enhancement motives, women are more prone to use alcohol as a coping mechanism to help them escape negative emotions', the researchers explain. ‘Higher neuroticism is also associated with depressive episodes, which are more likely to occur in women, who may then use alcohol to self-medicate'.
Education is generally believed to reduce alcohol consumption. The HSE study confirms this association: a university degree decreases the log quantity of monthly consumed alcohol by 50 pp. for men and 31 pp. for women, while vocational training lowers it by 37 pp. and 21 pp., respectively. This finding, however, does not account for non-cognitive skills.
Once the Big Five were included in the analysis, the effects of education dropped approximately by half to 26 pp. for men and 16 pp. for women with university education and to 24.8 pp. and 11.6 pp., respectively, for those with vocational training.
According to the researchers, 'By excluding personality traits, we tend to overestimate the effects of education on alcohol consumption. Although universities and colleges provide students with a new social environment, those personality traits which can impact alcohol consumption in either direction (i. e. conscientiousness, extraversion and neuroticism) have been formed long before a person enrols in university or vocational school. As demonstrated by twin studies, between 30% and 50% of self-rated personality traits are inherited, with extraversion and neuroticism being the most inheritable categories of the Big Five and conscientiousness mostly representing the result of primary socialisation'.
Russia has adopted a policy of reducing alcohol consumption. Based on the study findings, interventions which address the personality aspect of drinking patterns could make this policy more effective. At the same time, these new findings on the impact of educational attainment on drinking raise a critical question: could it be too late to begin alcohol abuse prevention in college? Given that the personality traits making one likely to drink excessively have already developed by that time, perhaps such preventive work should start much earlier, in secondary school, and focus more on developing soft skills with a protective effect, such as conscientiousness.