Monday, May 30, 2016

People who trust NGOs are more active


NGOs play an important role in Georgian society, weighing in on issues as diverse as energy policy and the budget. An important question however is, how much do people trust NGOs and are those who trust NGOs different from those who don’t? This blog post looks at how people who report trusting NGOs in Georgia differ from those who report distrusting them in terms of their social and political engagement.

CRRC’s Caucasus Barometer surveys (CB) regularly ask about trust in NGOs. In 2015, 23% of the population reported fully or rather trusting NGOs, while about the same share reported fully or rather distrusting them. In addition, a large share of the population (42%) reported neither trusting nor distrusting NGOs. The share who answered that they “don’t know” if they trust or distrust NGOs has declined from 31% in 2012 to 14% in 2015, which may indicate that awareness of NGOs has increased in recent years. The share of the Georgian public which trusts NGOs (sum of “fully trust” and “rather trust”) has more or less stayed the same since 2011, while the share which distrusts them (sum of “fully distrust” and “rather distrust”) has increased from 9% in 2012 to 20% in 2015.




People who trust NGOs are slightly more socially and politically active (i.e. attend public meetings, vote, etc.), than those who distrust NGOs. In 2015, 25% of those who trust NGOs reported attending a public meeting during the six months prior to the survey, while only 15% of those who distrust NGOs reported doing the same. This finding has been consistent over the past few years.



Note: Answer options “fully trust” and “rather trust” are combined in the columns “Trust”, and answer options “fully distrust” and “rather distrust” are combined in the columns “Distrust”. 


Those who report trusting NGOs also say they would participate in presidential elections if they were held next Sunday more often than those who distrust NGOs. In 2015, 78% of those who trust NGOs said they certainly or most probably would participate in elections if held next Sunday, while a slightly lower share (69%) of those who distrust NGOs said the same.



People who report trusting NGOs are more socially and politically active than those who distrust them, as demonstrated by CB data on public meeting attendance and intention to participate in elections.

For more on trust in NGOs in Georgia, see this Caucasus Analytical Digest article and for more data from CB 2015 take a look at our Online Data Analysis tool here.

Monday, May 23, 2016

The fury before the storm

The Georgian Parliamentary by-elections held on October 31, 2015 are regarded by some Georgia watchers as a ‘final rehearsal’ for the 2016 general elections,  and the results have been hotly debated. Tamar Khidasheli, who represented the Republican Party and the Georgian Dream Coalition (GDC), defeated her opponent from the Alliance of Patriots of Georgia, Irma Inashvili, by a margin of less than one percentage point.  The results were met with intensive questioning not only from representatives of the Alliance,  but some members of the ruling coalition as well.  Opponents were especially critical of the ‘special electoral precinct’ of the Ministry of Defense where Mukhrovani Military Base personnel voted.  Although, the special precinct existed during the previous elections, Khidasheli’s opponents argued that votes from the military made an outsized contribution to the victory of the Republican candidate.  They also accused the former majoritarian of Sagarjo and current Defense Minister of the Republican Party, Tinatin Khidasheli, of unlawful interference in electoral matters.  The situation reached a nadir when Transparency International Georgia approached the prosecutor’s office to start an investigation of the Sagarejo elections, although the investigation was later stopped.  This blog post does not look at any would-be procedural violations in the Sagarejo elections, but does describe the geographic, demographic and ethnic peculiarities of voters which could have contributed to Tamar Khidasheli’s victory in the elections and will likely be of consequence to the general elections.

The absolute difference in votes between Khidasheli and Inashvili constituted only 559 votes. Votes cast at special precincts are counted at “mother” precincts, and vote counts at special precincts are published together with the “mother” precinct count. Hence, it is impossible to distinguish between the vote counts of the two.  As the graph below shows, even if the results at the special precinct and its “mother” precinct were to be annulled, Khidasheli would be the likely winner. However, the margin of victory would have been only 87 votes.  Clearly, every vote was significant for Khidasheli’s victory.


Note: The height of the bars corresponds to the number of votes for each candidate, the labels denote vote share. Percentages may not add up to 100% due to rounding error.

Voting patterns in Georgia differ across location, ethnicity and religious denomination.  Eighty percent of Sagarejo’s population lives in rural areas, and about forty percent are ethnic Azerbaijanis.  Historically, ethnic minority votes in Georgia constitute an important source of electoral support for ruling political parties.  Looking at the demographic peculiarities of the Sagarejo by-elections gives interesting insights into the voting behavior of municipality residents.

The graph below shows election results disaggregated by settlement type. In urban areas, Inashvili won decisively. The race was relatively close in Georgian villages, but again the opposition candidate came out on top. In Azerbaijani precincts, the Republican candidate overwhelmed her opponent.


Note: The height of the bars corresponds to the count number of votes for each candidate, the labels denote vote share received in each settlement. Percentages may not add up to 100% due to rounding error.

While the chart above shows that votes from Azerbaijani villages were decisive for the victory of Tamar Khidasheli, when comparing the geographic and demographic peculiarities of each candidate’s supporters, it is clear that the socio-demographic make-up of Khidasheli’s voters was similar to that of participating voters overall. On the other hand, urban voters and voters from Georgian villages disproportionately supported Inashvili.



It can be argued that the victory of the ruling coalition candidate in the Sagarejo elections was largely a result of the support of ethnic Azerbaijani voters, whilst Inashvili’s supporters were mainly ethnic Georgians. This finding isn’t all that surprising, and follows a general pattern from Georgian elections past: Azerbaijanis in Sagarejo municipality almost always support the government. In 2012 parliamentary elections, the United National Movement gained 83% of votes in Azerbaijani villages, whilst in Georgian settlements the party barely won one third of votes. A year later, the Azerbaijani population of Sagarejo voted overwhelmingly (58%) for the presidential candidate of the new government.

The victory of the government-endorsed candidate in Sagarejo by-elections was mostly influenced not by the special precinct, but by the support of the municipality’s ethnic Azerbaijani population. As in the past, in the 2016 parliamentary elections, ethnic minority support for the ruling party is likely to be significant.

Monday, May 16, 2016

Fearing for the children – how living with children affects homophobic attitudes in Tbilisi

Following the controversial events of May 17, 2013, CRRC-Georgia conducted a survey in order to gauge the opinions and attitudes of the adult residents of Tbilisi towards homosexuals and their rights. Among the various outputs following the survey was  a series of blog posts exploring statistical predictors of homophobia. The findings indicated that a low level of education was one of the strongest predictors of homophobia among Tbilisi residents and that men had a higher probability of being homophobic than women, particularly when the men believed that homosexuality was an inborn rather than an acquired trait. This blog post looks deeper into the predictors of homophobia in Tbilisi by testing for a statistical relationship between homophobia and living in a household with one or more children under the age of 18. Our findings suggest that there is a significant relationship between these two variables, although it is different for men and women.


Unlike the previous series of blog posts on homophobia, weighted data is used for the descriptive analysis in order to make more accurate projections about the attitudes of Tbilisi residents. As was the case with the previous blog posts, we measure homophobia using the question, “[Whom] would you not wish to be your neighbor most?” Respondents were asked to choose one of the six groups presented on a show card: drug addicts, black people, adherents of a different religion, people having different political views, homosexuals, and criminals. Those who chose homosexuals (31% of those who answered this question) were deemed to be homophobic.

As the chart below shows, overall, there is no difference in the share of people reporting homophobic attitudes between those living in households with children and those living in households without children (32% and 31%, respectively). The findings, however, are very different separately for men and women. Specifically, we find that the share of Tbilisi females who are homophobic is 7% higher when one or more children live in their household, while it is 13% lower among men.



Logistic regression confirms the importance of having one or more children living in the household as a predictor of homophobic attitudes. According to the model, women with child(ren) in the household are 207% more likely to be homophobic when age, gender and education are controlled for. For men, though, the likelihood decreases.

So why do women tend to be more homophobic when they live with children, and why do men tend to be less homophobic? While further research focused specifically on this issue would be needed to determine the exact cause behind this finding, one possible explanation to the first part of the question is that women may be confusing homosexuality for pedophilia; another possibility is that women may be afraid that homosexuals potentially living next door might influence children into homosexuality (especially provided that, according to the findings presented in a previous blog post, women, more so than men, believe homosexuality to be an acquired trait). In regards to the second part of the question, men may feel that their masculinity has been confirmed once they have children (see here for more information), and thus they feel less threatened. It should be noted, however, that this latter explanation would only apply to those men who have their own children, while the survey data only tells us whether they live in a household with children, but not whether those children are or are not their offspring. For this reason, it may be useful in future studies to investigate the effect of having one’s own child(ren) on homophobic attitudes, in addition to more generally living with children.

The findings presented in this blog post show that women are more likely to be homophobic when there are child(ren) living in their household whereas the opposite is true for men. This finding suggests possible approaches to address homophobia, such as information programs speaking directly to mothers, and women in general, regarding the “nature versus nurture” debate, pedophilia, the nature of homosexual companionship, and the harms of viewing homosexuality as an abnormality. For men this may be more of a struggle with personal sexual identity and insecurities. In any case, a dialogue on the root-causes of homophobia is an important part to combating discrimination against sexual minorities.

The dataset of this survey, as well as respective documentation are available at CRRC’s Online Data Analysis portal.

Monday, May 09, 2016

Trends in the Data: Public support for democracy is slowly waning in Georgia


Following the first ever peaceful transition of power in Georgia’s 2012 parliamentary elections, the country improved its position in the Freedom House and Polity IV democracy rankings. Results from the latest polls, however, show that public support for democracy in Georgia has declined over the past few years. Weakening support for democracy can pose a serious problem for the process of democratic consolidation in Georgia, that is, the institutionalization and maturation of Georgia’s democracy. Reversion to a non-democratic system is unlikely in consolidated democracies, where a democratic system is accepted as “the only game in town”. Public opinion polls make it possible to measure public’s support for democracy. This blog post looks at how attitudes towards democracy have changed in Georgia in recent years.

According to CRRC’s 2015 Caucasus Barometer (CB) survey, nearly half of the population of Georgia (47%) agrees that “democracy is preferable to any other kind of government”, and only 16% thinks that “in some circumstances, a non-democratic government can be preferable”. However, if we compare these findings with previous waves of CB, it is clear that during the past four years, support for democracy has declined in Georgia. It was rather stable from 2011 through 2013, with approximately two thirds of the population reporting that democracy is the best form of governance. While in 2011, only 8% thought that “in some circumstances, a non-democratic government can be preferable”, this share doubled in 2015. The share of those who answered ”for someone like me, it doesn’t matter what kind of government we have”, also increased during the same period.



Note: Options ‘Don’t Know’ and ‘Refuse to answer’ are not shown on the above chart. CB was not carried out in 2014. 

Notably, the share of the population who consider Georgia a democracy has declined. According to CB 2011, half the population characterized Georgia as either ”a full democracy” (8%) or ”a democracy but with minor problems” (42%). In 2015, only 3% and 20% reported the same. The share of those who characterized Georgia as ”a democracy but with major problems” or ”not a democracy” increased.


It might be expected that one of the reasons why support for democracy declined would be the weakening of democratic values. Although a number of indicators can be used to test this,  CRRC data  do not suggest that this is the case. Support for democratic values has, in fact, visibly increased. For instance, the share of the population who agrees with the statement that “people should participate in protest actions against the government, as this shows the government that the people are in charge” has increased.  Since 2012 the share of the public agreeing with this statement has been at least twice as large as the share of those who agreed with the opposite statement.


Although, over time, the share of the population who regard Georgia as a democracy declined, the data shows that freedom of speech has likely been strengthened. For instance, since 2009 the share of those who agree that in Georgia people have the right to openly say what they think increased from 55% to 72% in 2015, while the share of those disagree with the opinion halved.



There is, thus, no empirical evidence confirming that the decline in public support for democracy in Georgia is caused by the weakening of democratic values. Hence, the reasons for the decline described in this blog post likely lay elsewhere.

When Georgia’s democratic development is discussed, public opinion is often forgotten. As this blog post has tried to demonstrate, the political elite should not take public support for democracy for granted. Attitudes towards democracy, like other attitudes, may often be changing. Further research is needed to understand the causes of this, since democratic consolidation is less likely without public support.

To find out more about public attitudes in Georgia, visit CRRC’s online data analysis tool.