1. INTRODUCTION
Protecting the health of the population by preventing and reducing the harmful use of alcohol
is one of the primary goals of governments. More alcohol use than alcohol consumption limits
recommended by the National Institute of Health and Care Excellence (NICE) has been
defined as the use of bad alcohol. It is stated that alcohol abuse can cause physical,
psychological, economic and social negative effects of individuals in this situation.
Due to the slower absorption of alcohol from the body than the absorption of the body's
normal functioning is causing a structure that causes drunkenness. Excessive alcohol
consumption along with intoxication, accidents, violence, and reduced productivity can lead
to undesirable consequences. It can also contribute to the development of cancer, brain, liver,
heart and bone pathologies and mental disorders. In 2012, nearly 3.3 million deaths or 5.9%
of global deaths were associated with alcohol consumption. There are significant gender
differences in the proportion of alcohol-related global deaths. For example, 7.6% of deaths
among males in 2012 and 4% of deaths among females were alcohol-dependent. In 2012 there
were 139 million lives that resulted in disability due to alcohol consumption. There is also
wide geographical variability in the rates of alcohol-related deaths and injuries.
In order to avoid the negative consequences of excessive alcohol use, governments and public
health organizations have interjected the rules of “sensitive drinking” globally. For example,
NICE recommends that those who choose to consume alcohol on a regular basis should not
exceed three to four units per day. The World Health Organization (WHO) states that alcohol
use is the third most important risk factor that increases the burden of disease in developed
countries. Besides, due to the alcohol policies they developed, governments aim to reduce the
harmful use of alcohol and the health and social burden that can be linked to alcohol in
society. Such policies can be formulated at a global, regional, multinational, national and subnational level. Most of the countries that are members of the World Health Organization have
been working to reduce the use of alcohol in recent years. The majority of reporting countries
have executed national alcohol policies. High-rate tax or price implementation for the solution
of the problem of excessive alcohol consumption is of great concern among public health
professionals and policy makers (Nelson, 2014).
Young adulthood, defined between the ages of 18 and 29, refers to a critical period of
development in one's life. It is evident that the health behaviors adopted by young adults in
these ages will affect long-term body health, which will continue in the following years.
Therefore, the positive habits they will gain during this period will have a positive effect on
their overall health status. The health behaviors adopted by young adults not only have
profound and long-term consequences for their health in their later lives but also will play a
crucial role in future generations when they become parents.
Average alcohol consumption per capita of adults remained almost constant in the last twenty
years between 1992 and 2012, but the increase in the last five years has been observed (WHO,
2014). In young people, alcohol consumption is increasing, especially in low-income and
middle-income countries. (WHO, 2011). Some of the reasons for young people consuming
more alcohol, low cost of alcoholic beverages, the presence of more extensive alcohol use
areas, alcohol promotion especially for young people and in many societies can be considered
as the acceptable form of alcohol consumption.
According to the OECD Health Statistics for 2016, Turkey ranks last among OECD countries
with 1.3 liters per person for alcohol consumption among individuals over 15 years of age. In
2012, an average adult in the OECD country consumed 9.1 liters of pure alcohol per year. The
highest alcohol consumption was expressed in Austria, Estonia, and France, where 12 liters
per person per year and above. The lowest alcohol consumption was reported in Turkey and
Israel. In countries such as Turkey, Israel, Indonesia and India, low alcohol consumption can
be said to be effective due to religion and cultural structure (OECD, 2015). Again, in OECD
reports, between 2008 and 2010, alcohol consumption of men and women in Turkey and
European countries were compared. According to this, the ratio of alcohol consumption per
capita for men and women aged 15 years and over in Turkey is 4,4 liters and 0,5 liters, which
is very low than the average consumption value of countries in the European continent, 10,6
liters. If we look at in 2010 in Turkey 15 years and older person of alcoholic drinks they
consume, and 63% portion of beer, 9% of the wine and if the luxury of 23% whiskey, vodka,
constitutes spirits such as raki. In Turkey, the ratio of the population of the heavy drinkers
aged 15 years and over to the total population was 0,2%, and the ratio between alcohol users
was 1,3%. In 2010, in Turkey, 65.9% of men 15 years and older, 92.4% of women and 79.6%
of both sexes did not use alcohol at all. 11.6% of men 15 years and older, 2% of women and
6.7% of both sexes in total are used to be drinkers.
This study, which aims to examine socio-economic determinants of alcohol consumption of
adults in Turkey, consists of five parts. In the second part, the literature summary of the
studies on determinants of alcohol consumption is shown. In the third part, the materialmethod of the study was examined and also the data, variables and statistical analysis
methods used in the study were discussed. In the fourth part, the results of the estimation
obtained with the multinomial probit model were analyzed, and interpretations were made
over marginal effects. In the fifth chapter of the final chapter, policy recommendations were
developed by taking into account the studies that coincide with the results obtained.
1. ABSTRACT OF LITERATURE
Given the negative effects of alcohol abuse on societies, there are many studies conducted by
researchers on this subject. Table 1 abstracts the studies carried out in several countries.
Table 1. Abstract of Literature
Contributor
(Year)
Country/Sample
size
England/9665
U.S.A./1292
Sweden/49321
Botswana/372
India/500
Portugal/2414
Ghana/3109
Thailand/41229
China/7853
Austria/267153
Belgium/1688
Australia/514
France/29566
U.S.A./2220
Turkey/50
Key Findings
Individuals with higher educational levels tend to consume more alcohol and become
alcohol dependent.
Body mass index, age, low-income variables reveal a negative impact on women's
alcohol use. Women who are not diagnosed with diabetes are more prone to drink
alcohol.
Individuals with low IQ scores tend to consume excessive alcohol.
Women who pay attention to religious life and have a high status in society are less
likely to consume alcohol. Men tend to consume more alcohol than women. Alcohol
consumption is lower in households with children.
Low educational level, the presence of alcohol consumption in family members and
cigarette consumption is related to the same way alcohol consumption. 41-50 age
group is more likely to consume alcohol. Low-income people consume more alcohol.
Excessive alcohol use is inversely related to the consumption of fruits and vegetables.
In individuals with low education levels, wine consumption is likely to be high. In
women who exercise regularly, they are likely to use mild alcohol. People who
consume excessive alcohol are men, advanced Age, low education level, smoking and
consuming low fruit vegetables.
Alcohol consumption varies between regions. Alcohol abuse around family and
friends has a significant effect on the individual's alcohol consumption.
A study of the household size and the staff is effective both on alcohol consumption
and on alcohol expenditure of individuals.
The rate of alcohol use in men between the ages of 35 and 65 is high in divorced or
separated and working at a low educational level and in an official institution. Women
living in the city have high rates of alcohol use compared to women living in rural
areas. Men in rural areas consume more alcohol.
In men, variables such as advanced age, high socio-economic level, low psychological
stress, and lack of treatment for depression are effective on alcohol consumption.
Being an only child at home and having a stepfather are factors that increase the
alcohol consumption of girls. Problems in the family increase the chances of girls
drinking alcohol.
Age, smoking, being born in Australia and being depressed increases the likelihood of
alcohol use.
Cigarette smoking in men, being high-income, healthy food consumption and alcohol
consumption is related. Men who have a history of cardiovascular disease or who are
depressed and women who have a history of cirrhosis are less likely to drink alcohol.
In women, age, income, working condition, alcohol consumption is inversely related.
Education and income are related to alcohol consumption in the opposite direction.
Alcohol consumption in divorced individuals is more likely. People with diabetes are
less likely to consume alcohol. Women who are working consume more alcohol.
Ethnic origin is valid for alcohol use.
The increase in the prices of alcoholic beverages during Ramadan has led to a
Spain/2254
Estonia/6500
Sweden/28800
Malaysia/2154
Albania/700
India/241
Turkey/9822
France/1103
Thailand/442
Belarus/1465
Malaysia/34539
Persia/1136
decrease in alcohol consumption. Raises in income, increases alcohol consumption.
In men, low age and daily cigarette consumption are positively associated with being
excessive alcohol consumption — the possibility of being a mild alcohol drinker
increases in married and working people. The increase in quality of life and regular
exercise increases the likelihood of people going to extremes from low alcohol
consumption to excessive consumption.
Excessive alcohol consumption increases the likelihood of calling an ambulance. The
use of mild alcohol in men is positively associated with receiving hospital care at the
hospital.There is a negative correlation between wine consumption and
hospitalization.
Age and gender are determinants of alcohol dependence. Young people are more
addicted to alcohol than old people. Being young, becoming a man, and
unemployment increases the risk of alcohol addiction.
Excessive alcohol consumption is more common in rural areas than in urban areas.
Ethnic origin is significantly effective on the tendency to consume excessive alcohol.
Men with primary and pre-primary education are likely to become excessive alcohol
drinkers. There is a positive correlation between family financial losses, smoking, and
excessive alcohol consumption. Excessive religiousness, waist-hip ratio, and
retirement variables are related to excessive alcohol consumption in the opposite
direction.
Alcohol consumption in the immediate environment enhances the individual's alcohol
consumption.
Factors such as being in rural areas and education are positively related to alcohol use
and the level of expenditure on alcohol. The possibility of alcohol use of individuals
living in cities and having wealth is negatively related to alcohol expenditure. In rural
areas, individuals have an impact on the possibility of alcohol consumption, and the
income of the individual affects the amount of expenditure on alcohol. Having
compulsory health insurance is negatively related to alcohol consumption. Having
children under five years of age in rural areas reduces the possibility of alcohol use
and alcohol consumption.
In men, not to consume alcohol is negative, while excessive alcohol consumption is
positively correlated. Individuals suffering from financial distress have a high
tendency to consume excessive alcohol. Smokers and drug users are more likely to
have excessive alcohol consumption. The presence of individuals with psychological
disorders and alcohol use in the family is positively related to excessive drinking. As
the social status of the individual increases, the tendency to become excessive alcohol
consumers decreases.
Alcohol consumption is negatively related to individuals living in rural areas and
individuals over 45 years of age. On the other hand, being in the 25-45 age range is
positively correlated with alcohol consumption.
Alcohol use increases with the Arabs starting their education in Belarus; Excessive
alcohol use is quite high. Health problems associated with alcohol in men occur.
Students of Slavic origin have more alcohol-related health problems than students of
Arabic. Use of alcohol is influenced by factors such as religion, culture and social
norms.
Smoking and increase in income also enhance the plausibility of alcohol consumption.
Alcohol consumption varies according to ethnic origin and location.
The presence of alcohol-consuming individuals in the family, Smoking, occupational
status, marital status, gender, and substance abuse or alcohol consumption are
significant correlations. Although alcohol production and consumption are prohibited
in Iran, alcohol is the second most widely used substance.
1. MATERIAL and METHOD
1.1.Data
In this study, the horizontal-cross-section data obtained from the Turkish Health Survey, the
survey conducted by Turkish Statistical Institute in 2010 and 2012, were used. In the data
acquisition, the two-layered cluster sampling method was used. As an external tanning
criterion, the rural-urban separation was used. (The settlements with a population of 20000
and below are considered rural, and the settlements with a population of 20001 and above are
considered urban). The first phase sampling unit is the blocks selected from the clusters
containing an average of 100 households; the second phase sampling unit is the systematically
selected households from each cluster. In this study, horizontal-sectional data of individuals
over 15 years of age were used (TurkStat).
1.2. Variables
The dependent variable of the study, “have you ever used alcohol?” (Yes; No) and “do you
still drink alcohol?” (Yes, every day; yes, occasionally; I am not currently using) Is that the
state of alcohol abuse measured with questions. These two questions were combined into a
four-choice variable, “yes, every day; yes, occasionally; he does not use it at the moment; he
did not use it at all.”
The independent variables in the study were determined by conducting a literature study.
Independent variables are, place of residence (urban/rural), age (15-24, 25-34, 35-44, 45-54,
55-64 and 65+), gender (male/female), educational status (literacy/read and write but not
finished any school/ primary, elementary/secondary/ vocational secondary, high school and
equivalent schools, college/faculty/graduate/doctorate), marital status (never married, married,
wife died/divorced), monthly average income (less than $ 500,-,-,-, 2301+), general health status (very good/good, medium/bad/very bad), primary health
care provider psycho-social support health care status (yes/no), the frequency of brushing
your teeth (1 time per day/2 or more, 1 time per week/more than 1, never), body mass index
(underweight, average weight, overweight, obese), the frequency of eating fruit (at least once
a day, at least once a week at least once a month, not at all), exposure to violence, (the
situation was subjected to severe, a bit exposed and not exposed), exposure to tobacco smoke
at home status (never less than 1 hour per day, between 1-5 hours per day more than 5 hours
per day), the survey year (2010, 2012) and tobacco use status (Yes, No).
1.3.Research Methodology
Multi-state probit model; zero mean, arbitrary variance-common variable matrix and common
multivariate normal distribution and are expressed as a coincidental utility function model
with error terms. In this case, the variation of the error term in the multistate probit model
may be different and related (Daganzo, 1979: 17). However, it is known that there is a close
similarity between the multi-state probit model and the multi-term probit models in which
two-state preference decisions are observed during multiple periods (McCulloch and Rossi,
1994: 207-208). Although the multi-state probit model is one of the intriguing models of
alternatives among the hidden benefits of the alternatives, and among the attractive models
among the preference behavior models, it is not widely used in the literature since the
problem of calculating the highest likelihood estimator in the application is encountered.
Nevertheless, the fact that the multi-state probit model alleviates the independence
assumption of unbound alternatives is considered as superiority to the multi-state Logic
Model (Geweke, Keane and Runkle, 1994: 694).
2. CONCLUSIONS
2.1. Definitive Statistics
Socio-economic and demographic variables are shown in Table 2. In this study, the total
values of some variables are different due to the loss observation values. 73% of the
individuals in the study reside in cities. The highest participation is in the 25-54 age group.
While the percentage of men is 45%, the percentage of women is 55%. SSI covers the
treatment cost of the vast majority of individuals (81%). While 36% of individuals are
primary school graduates, 12% are university graduates and higher education graduates. The
proportion of married people is about 70%. Approximately 40% of the individuals in the
study had monthly net income of TL-. It was determined that 54% of the individuals
have good health, 62% brushed their teeth once or twice a day, and 43% had an average
weight. Also, 57% of individuals consume fruits at least once a day, while 67% consume
vegetables at least once a day. It was found that 23% of individuals were exposed to tobacco
smoke for at least one hour or more a day.
2.2.Model Prediction
In the present study, the multinomial probit regression model was used to determine the
factors that are effective in alcohol use in individuals over 15 years of age. The results of the
estimated multinomial probit regression model are given in Table 3.
In the predicted multinomial probit model, the state of “no alcohol use” was determined as the
main group. The coefficient interpretation cannot be made from the predicted model, but only
the signs of variables can be determined through this model. According to this study, it has
been determined that alcohol consumption has a positive correlation with sex, age, education,
marital status, high income, dental brushing frequency, fruit consumption, violence, tobacco
use, and home smoke exposure. The inverse relationship was found between alcohol
consumption and general health status variables. Table 4 shows the marginal effects predicted
from the model.
According to the results of the analysis, gender had a significant effect on alcohol use. It has
been determined that men are 0.3% more likely to drink alcohol every day than women and
9.5% more likely to drink alcohol every day. When age was examined, it was found that
individuals in 25-34, 35-44, 45-54 and 55-64 age groups had a higher incidence of occasional
alcohol use than individuals in 15-24 age group (4.3%, 5.7%, 5.8%, and 5.5%, respectively).
It has been found that individuals 65 years and older are 1.2% more likely to use alcohol than
individuals 15-24 years of age, and 11.4% more likely to have used alcohol before. Primary
School, elementary, High School, and university graduates have been found to be,
respectively, 8.6%, 12.2%, 16.4% and 24.3% more likely to use alcohol from time to time
than those who do not know how to read and write. It has been determined that singles are
2% more likely to use alcohol from time to time than married ones. It has been found that the
probability of alcohol use is 3.1% more than that of married people who have died or
divorced spouse. When monthly income was examined, it was found that those whose income
was between- TL were 2.7% more likely to drink alcohol than those whose income
was less than 500 TL. It has been determined that those whose income is between TL- are 1.9% more likely to use alcohol from time to time than the reference group. It has
been determined that those with income of TL 2301 and above are 8.2% more likely to use
alcohol from time to time than the reference group.
According to the general health status, it was found that those with moderate health status
were 2.04% more likely to use alcohol from time to time than those with poor health status
and those with good health status were 1.93% more likely to use alcohol from time to time
than those with poor health status. It has been determined that the individuals who brush daily
are 1.7% more likely to use alcohol than those who do not brush at all.
When the situation of violence is examined, it was found that individuals who are severely
exposed to violence are 4.4% more likely to use alcohol than those who do not see any
violence, and those who are exposed to some violence are 1.8% more likely to use alcohol
from time to time compared to the reference group. It has been determined that those who are
exposed to 5 hours of tobacco at home are 1.3% more likely to use alcohol from time to time
than those who are not exposed to it and 0.1% more likely to use alcohol every day. It was
found that individuals who were exposed to smoke for more than 5 hours at home were 0.3%
more likely to drink alcohol each day than those who did not. It was found that the probability
of alcohol use is 0.1% per day compared to those who do not use tobacco and the probability
of alcohol use is 11.9% per day.
3. DISCUSSION
In this study, it was aimed to examine socio-economic determinants of alcohol consumption
of adults in Turkey. For this purpose, the horizontal-cross-section data obtained from the
Turkish Health Survey, conducted in 2010 and 2012 by the Turkish Statistical Institute, were
analyzed with the multinomial Probit model. According to the analysis results, gender
differences are one of the primary factors in alcohol use. All age groups are active in drinking
alcohol. Income was found to have a positive effect on alcohol use. One of the indicators of
healthy life is fruit consumption frequency. It has been determined that the state of no fruit
consumption is a decisive factor in occasional alcohol use. The same-way relationship
between alcohol consumption and psycho-social support of individuals who consume
excessive alcohol was found and the opposite-way relationship between alcohol consumption
of people who have good health status. Education has been shown to be effective in alcohol
use in all groups. It has been determined that individuals with income of 2500tl and above
tend to drink alcohol. It has been found that individuals who use tobacco tend to use alcohol.
In parallel with the results obtained by, Burazeri et al. (2010), Andréasson et al. (2013), Pärna
vd. (2010), Grittner et al. (2012), Zho et al. (2006), Chavez et al. (2013), the study found that
gender had significant effects on alcohol use. Men are more prone to alcohol consumption
and dependence than women. Easwaran et al. (2014), Dias et al. (2011), Coulson et al. (2014),
Patrice and Raul (2014), Soler-Vila et al. (2014) reported that age had a significant effect on
alcohol consumption. However, it is observed that alcohol consumption in the age group is
higher and varies according to the regions where the study was conducted. In this study,
especially adult individuals are more likely to consume alcohol from time to time compared to
young people. Easwaran et al. (2014), Dias et al. (2011) while their findings were similar,
Patrice and Raul (2014) reported that young people tend to use more alcohol than older
people. In most of the studies, it was perceived that the education variable was used.
According to the study, the increase in educational levels also increases the possibility of an
individual using alcohol from time to time. The findings obtained by Huerta and Borgonovi
(2010) coincide with the results of the study, Easwaran et al. (2014), Bureri et al. (2010),
Patrice and Raul (2014) reported that people with low education tend to consume more
alcohol in their studies. The increase in the income of the individual has been found to have
led to an increase in the probability of alcohol consumption. Similar results were obtained
from Bryden et al. (2013), Bonevski et al. (2014), Çeken (2015) 's studies. The probability of
alcohol consumption of tobacco users was also high. Alavi et al. (2016), Cheah et al. (2017),
Yaogo et al. (2013), Huerta and Borgonovi (2010) and Dias et al. (2011) found the same
results in their studies and stated that the use of tobacco products in the individual increased
alcohol use.
In the case of marital status, the probability of alcohol consumption of married and divorced
or deceased spouses was found to be higher than singles. Soler-Vila et al. (2014), Martinez et
al. (2011) in their study, they stated that married individuals affected alcohol consumption.
Zho et al. (2006) reported in their studies that divorced or separated people tend to consume
alcohol.
Alcohol use is an individual choice, but it can have many social effects. In particular,
excessive alcohol use can damage the individual's health, career, and personal relationships,
as well as cause anti-social, violent behavior or traffic accidents, which can damage others.
It is apparent that alcoholics need help to get rid of an addiction. Therefore, it would be
inevitable for governments to intervene by national and international public health authorities.
In particular, the ability of governments to design and implement comprehensive prevention
strategies that combine the strengths of different policy approaches is critical to success.
Alcohol advertisements become uncontrollable using the internet and social media. In
particular, it is essential to reduce the probability of individuals in the youth age to encounter
these advertisements. It is seen that the sanctions used in the Prevention of alcohol use of
individuals vary among countries. This situation needs to be put into a certain standard and
implemented worldwide.