Monday, March 26, 2018

Women Significantly Less Likely to Go Out to Eat in Georgia

[Note: This post was published with OC-Media. The post was written by CRRC-Georgia's President, Koba Turmanidze.]

Busy restaurants and cafes are a common sight in Georgia, and CRRC’s Caucasus Barometer data suggest that restaurants and cafes have become busier over the last five years. While 27% of Georgia’s population reported going to a restaurant in 2012, five years later 50% did. There is an upward trend for both men and women, yet the data also suggests there is a significant gender gap. Taking into account other social and demographic characteristics, women are significantly less likely to go to restaurants than men.


Note: According to the instructions to this question, restaurants included pizzerias, khinkhali houses, McDonald’s, etc.

A number of factors including settlement type, age, social status, economic condition, and gender influence whether an individual goes out to eat. The findings are hardly surprising in many respects: residents of Tbilisi are more likely to go to restaurants compared to the residents of villages. Irrespective of whether a person reports being employed or unemployed, he or she is more likely to go to a restaurant than individuals who are outside the labor force, i.e. those who do not work and are not looking for a job either. Likewise, people living in households with low reported expenditures per month (250 USD or less) are less likely to go to a restaurant. Also unsurprisingly, age is negatively related with eating out: the older a person the less likely is he or she to go out to eat. Actual and perceived social status show the opposite effects of age: the more years a person spent studying in formal educational institutions, the higher are his or her chances to have gone to a restaurant. In the same manner, the higher along a hypothetical ten step ladder representing the society a person places him/herself, the more likely they are to visit restaurants.



Note: The chart displays the effect of each factor on an individual’s probability of reporting they went to a restaurant during the past six months. ‘Diamonds’ are point estimates, whereas lines show 95% confidence intervals. The further the ‘diamond’ is from the red dotted line, the larger the effect. The few ‘diamonds’ right on the red dotted line are reference categories for a variable. Rural settlements, males, individuals who do not belong to the active labor force, and individuals who did not report household spending are reference categories. Every other category should be interpreted in relation to corresponding reference category (e.g. capital residents in relation to rural residents, females in relation to males, etc.)

While all the above factors influence whether a person goes to a restaurant, gender has the largest effect of all: all else equal, women are about 10 percentage points less likely to go to a restaurant than men. Further analysis shows that more educated women are no more or less likely than less educated women to go to restaurants, women from relatively wealthy households are not different from women from poorer households, and so on.

When looking at the impact of other socio-demographic factors across the two gender groups, women are worse off in terms of going to restaurants simply because they are women. The chart below demonstrates that if we pick a male and a female of the same age between the ages of 20 and 78, the male will always have a higher chance to have reported going to a restaurant.

Similarly, if we take two people of a different gender, but identical years of education, the man will still be more likely to have eaten at a restaurant in the last six months than a women. Notably, the significant difference in terms of years of education is maintained in the group who studied for 10 to 16 years, which constitutes 84% of the population according to the 2017 Caucasus Barometer survey.



Employment status and a household’s expenditure do not entirely diminish the impact of gender either: while both males and females are equally likely to go to a restaurant if they do not belong to the active labor force, in the unemployed and employed groups, females are disadvantaged. Moreover, females from households that spent up to USD 400 in the month prior to the survey are also less likely to have eaten in a restaurant in the past six months. Interestingly, there is no gender difference in the group of relatively high spending (more than USD 400) as well as in the group which did not report their household expenditure.


The findings of this analysis suggest that gender is the single most important factor that predicts whether an individual will go to a restaurant in Georgia. Regrettably, females are disadvantaged in this regard compared to males of the same age, education, social-economic standing and settlement type, demonstrating yet another form of gender inequality in Georgia.


To explore the data used in this blog post, visit CRRC’s Online Data Analysis platform. The code used for data analysis is available here.


Koba Turmanidze is CRRC-Georgia’s President.

Monday, March 19, 2018

Temporary emigration intentions from Georgia: Do migration networks count?

The UN estimates the number of international migrants worldwide to be on the rise. Academics and policy makers continue to pay considerable attention to drivers of international migration, i.e. the factors that cause people to move from their home country, either temporarily or permanently.  While a significant body of scholarship exists on the structural ‘push’ factors of international migration, such as limited economic opportunities, poverty, poor governance, or war in migrants’ home countries, interpersonal factors are no less important in shaping migration.  This blog post investigates the latter, seeking to examine how individuals in Georgia with and without close friends and family living abroad differ in their willingness to emigrate from the country temporarily. 

Studies have been conducted that demonstrate the impact of personal networks on migration behavior.  One central theory guiding these studies is the ‘migration network theory,’ which posits that the reduced social, economic, and emotional costs of migration stemming from existing contacts who are able and willing to help new migrants ultimately ease migration, and, to a certain degree, promote it. Understanding migration networks permits a more comprehensive view of migration as a dynamic process, rather than a mechanical outcome of economic or political deprivation.  Migration networks include family, friends, neighbors, and former colleagues — essentially anyone an individual can rely on and share information about opportunities abroad, including settlement assistance.

Emigration has been an important coping strategy for the population of Georgia since the dissolution of the Soviet Union. CRRC’s Caucasus Barometer survey data from 2010 through 2017 indicates that the share of people in Georgia willing to temporarily emigrate has increased slightly.  In CB 2017, 55% of the adult population of Georgia responded ‘yes’ to the question: “If you had a chance, would you leave Georgia for a certain period of time to live somewhere else?”  In 2010, this share was 47%.


CB also asked two questions that can help see individuals’ temporary migration intentions in light of the migration networks they might have.  Of those who had a close relative living abroad at the time of the survey, 59% responded that they would leave Georgia temporarily to live somewhere else.  In contrast, only 40% of those without close relatives living abroad responded that they would emigrate temporarily. Similarly, individuals who reportedly had a close friend abroad were more likely to report a willingness to temporarily emigrate than those who did not. It is still important to mention, though, that about 40% of those not having a close friend or relative abroad still report willingness to temporarily emigrate from the country.


The findings presented in this blog post suggest, in accordance with the migration network theory, that social networks may play a role in people’s willingness to temporary emigrate from Georgia. Individuals with a close contact who was living abroad at the time of the survey were more likely to respond that they would leave Georgia for a certain period of time to live somewhere else.  It should be emphasized, however, that CB does not present data on actual emigration, but rather reported intentions that may or may not result in individual actions.

To explore the data used in this blog post further, visit our Online Data Analysis platform.


Monday, March 12, 2018

Dissecting Attitudes towards Pre-Marital Sex in Georgia

Many in Georgia embrace conservative attitudes about premarital sex, as a previous CRRC blog post highlighted. Attitudes are different, however, depending whether it’s a male or a female having the premarital relationship. This blog post uses data from CRRC’s 2017 Knowledge of and attitudes toward the EU in Georgia survey (EU survey) conducted for Europe Foundation to describe how justified or unjustified people of varying ages, genders, and those living in different types of settlements believe pre-marital sex to be for men and women.

In 2017, when asked, “In your opinion, how justified or unjustified is it for a woman to have a sexual relationship before marriage?” 71% of people in Georgia reported that it is ‘never justified.’  In contrast, only 38% responded that it is ‘never justified’ for a man to have a sexual relationship before marriage. Both men and women are more conservative towards women engaging in pre-marital sexual relationships than men.  However, women report that it is ‘never justified’ for a man to have pre-marital sex slightly more often than men.


Variations in the level of justification of male and female pre-marital sex can also be observed by age group and settlement type. Unsurprisingly, older people (56+) hold more conservative attitudes toward pre-marital sex than younger individuals, responding more frequently that it is ‘never justified’ for both men and women to have a sexual relationship before marriage. Nonetheless, people above the age of 55 exhibit much greater acceptance of a man having a sexual relationship before marriage than of a woman.



Both men and women in the capital and other urban settlements are more liberal than those residing in rural and ethnic minority settlements.  However, men and women in Tbilisi generally demonstrate greater acceptance of premarital sex than those in other urban settlements of Georgia. While people living in rural and ethnic minority settlements hold the most conservative attitudes in general, they are more strongly opposed to women having pre-marital sexual relationships than men, further highlighting how standards of ‘justifiable’ sexual behavior are applied to men and women differently.



The data presented in this blog post highlights a number of findings.  First, a majority of individuals in Georgia believe that women should adhere to conservative standards of sexual ‘purity,’ while men are granted greater liberty in this regard.  Secondly, even within populations that are more liberal toward pre-marital sex — men and women aged 18-35 and those residing in the capital — most people still report it is never justified for a woman have a pre-marital sexual relationship, while they are more liberal with men.  The fact that women tend to respond more frequently that it is ‘never justified’ for a woman to have a pre-marital sexual relationship than responding the same about a man demonstrates the extent to which women have internalized gendered norms regarding sexual behavior.

To explore the data used in this blog post further, visit our Online Data Analysis platform.

Monday, March 05, 2018

Partisanship and Trust in TV in Georgia

[Note: This post was first published on OC-Media. The post was written by David Sichinava, a Senior Policy Analyst at CRRC-Georgia. The views presented in this blog do not represent the views of CRRC-Georgia, the National Democratic Institute, or any affiliated entity.]


One of the outcomes of the stark polarization of news media sources globally is that people tend to align to the media outlets which resonate most with their ideological beliefs. In most cases, consumption of a particular ideological media source can only reinforce one’s beliefs, which might lead to an even further polarization of the audience. These patterns can be characteristic of mass media in contexts as different as, for instance, the United States and Lebanon. As the data from the December 2017 wave of CRRC/NDI survey shows, people in Georgia also appear to be selective in trusting media that aligns with their political beliefs as well.

The two largest TV networks in Georgia, Imedi TV and Rustavi 2, tend to support different political parties in their coverage of current events. A long-term media monitoring project which was funded by the European Union (EU) and the United Nations Development Program (UNDP) unveiled that throughout the 2016 parliamentary and the 2017 municipal election campaigns, Imedi TV allocated more airtime to and provided positive and neutral coverage of the governing Georgian Dream (GD) party. The network dedicated less airtime to and had more negative coverage of the opposition United National Movement (UNM). In contrast, Rustavi 2 covered the ruling party negatively, while covering the UNM with neutral or positive tones. The UNM also received more airtime on Rustavi 2.

Unsurprisingly, those who name the Georgian Dream as a party closest to their views were more likely to trust Imedi TV for accurate information on politics and current affairs in Georgia than were those who named the United National Movement on the December 2017 CRRC/NDI survey. At the same time, those who named the UNM as a party closest to their views were more likely to trust Rustavi 2. No specific preference could be seen in the case of those who answered “No party”, “Don’t know”, or refused to answer the question (i.e., non-partisans).

This tendency persists across the population of different settlement types and endures even when controlling for major demographic characteristics, such as gender, age, and ethnicity. Tbilisi residents who reported either of the two political parties as being closest to their views were the most polarized: those identifying themselves with the UNM in the capital had a relatively small probability (27%) of trusting Imedi TV, while those who identify themselves with GD had a comparably low probability of trusting Rustavi 2 (32%).


Note: Points on the chart display predicted probabilities of trusting Imedi or Rustavi 2 by settlement type and party preference, while bars correspond to 95% confidence intervals. For example, the probability for a person who identifies with Georgian Dream and resides in Tbilisi to trust Imedi is about 63%, while the probability for a UNM supporter in Tbilisi to trust this TV channel is as low as 27%. These probabilities are calculated using logistic regression models. Replication data and corresponding R code can be found here.

Unsurprisingly, those who report any of the two major Georgian political parties to be closest to their views tend to trust the TV network that favorably covers their party. In contrast, the non-partisan population does not systematically differ in trusting either TV network.

To have a closer look at CRRC/NDI survey results, visit our Online Data Analysis portal.