30 Dec

Trust and Gender in the Caucasus

I wrote a blog post a few weeks ago about interpersonal trust (using EBRD data). One of my arguments is that Azerbaijanis have very low interpersonal trust. After a conversation with a girlfriend, I had the idea to look at gender differences in trust as well. This is with CRRC data.

First, without gender, you can see that generally Azerbaijanis have much lower trust in others than Armenians or Georgians do, but this varies by question. (These are on a 1-10 scale with 1 not trust, 10 trust completely or 1 total disagree and 10 totally agree).






In this next chart with gender, you can see gender differences in a variety of questions related to interpersonal trust. The only time that there was no significant difference in trust between men and women is for Armenians with the question “there are many people I can trust completely” – otherwise all these differences are significant.


But looking at the two extremes – completely agree and completely disagree, things are also interesting…


In this question of people that one can trust completely, Georgians seem to have it the worst, with only 3 or 4% having someone like that. Overall gender differences aren’t that stark in this question.


Here for money lending, big country differences. Georgians can’t get a break, huh? Gender differences here exist, but not too harshly.


For reliable people, Georgians again are in a bad spot.


Help repair the apartment – well, this is a lot to ask of people, right? Georgians again are not doing well here.


I don’t want to get sick in the Caucasus. Looks like people have trouble finding people to  help care for them. Again, no major gender differences here.

So overall, despite there being a statistically significant difference in trust between genders, it doesn’t really manifest in these practical questions.


30 Dec

Material Deprivation in 2012

A few years ago I wrote a piece about measuring poverty in the South Caucasus.

That paper ended with 2010, but I was talking about it with someone yesterday, so I did some quick 2012 updates.

As far as monthly household income in 2012, the distribution hasn’t changed much since that 2010 paper, but you can see that Armenians and Azerbaijanis are doing much better economically.




In the paper I argue that material deprivation is the best measure of economic wellbeing because it reflects household ability to buy consumer goods.

This is the 2009-2010 distribution of material deprivation:


And this is the 2012 update.


Armenia didn’t change much. Azerbaijan has less people at the poverty end of the scale, so that’s a positive change. Georgia had a bit of a change in more people in the enough for food category, but about the same in poorest category.


It is hard to see economic changes in such a short period of time – it would certainly be worthwhile to do some trend analysis here. Maybe next year!

03 Dec

#putinout hashtag analysis

civilnet (Via Civilnet.am)

There is a series of protests in Armenia right now, focused on Vladimir Putin’s visit, but overall critiquing Armenia’s possibly joining the Eurasian Union.

Unzipped does a nice summary here.

Here’s my Dec 3 hashtag analysis that is the following relationships. It is basically Armenians (far left), foreigners (middle), and Russians (right).


You can see that pretty much everyone in Armenia that is on Twitter follows each other. 🙂

This is the Dec 3 reply analysis of the hashtag – much clearer about small communication networks.



Here is Dec 4 reply analysis.


21 Oct

What does the Internet think of the South Caucasus?

Well, really just Armenia and Azerbaijan… Googling for Georgia is too difficult.

Inspired by this campaign that looked what Google automatically fills in for searches about men and women, I tried to do this with Armenia and Azerbaijan. I was logged out of Google, so these are the “natural” search results, not influenced by my own.


Yes, Azerbaijan is rich! And absolutely, the question of its position in Europe or Asia is an important one. Which country? Well, I guess with Southern Azerbaijan in Northern Iran, this could  be a little confusing.


Again, the continent thing is really confusing. I don’t blame you for searching for this, users of Google.


Here is an interesting selection. No history? Impossible.


Armenia Has Talent! Awesome. And no nukes, sorry.


Excellent questions, Google users. Russia versus USSR is a tough one for you young people, I guess.


Yes-  first Christian nation!


I ask two of these four questions often. Guess which ones?


I ask three of these four questions often. Guess which ones?


Well, some of these are not nice and some are interesting empirical questions.


Interesting questions here too.


I love William Saroyan!


Area code and ethnicity – excellent questions.


Oh wow – this sums up a lot of the questions swirling in someone’s head.





Weird that nothing came up for these.


So interesting stuff people of Google!

19 Oct

Demand for Democracy in Europe and Eurasia

Demand for democracy is one of my favorite outcome (dependent) variables. It isn’t perfect, but it is used frequently in public opinion surveys and has been shown to be correlated with all sorts of interesting things.

In a transitional society, popular demand for democracy (or legitimation) takes the form of a choice between competing regime types with which people have some degree of familiarity. Thus, survey questions should preferably not ask people how much they like democracy in the abstract (for example, through agreement or disagreement with one-sided Likert scale statements). Instead, they should offer respondents realistic choices between democracy and its alternatives.” (from here)

“Which of these three statements is closest to your own opinion? (A) Democracy is preferable to any other form of government; (B) In certain situations, a nondemocratic government can be preferable; or (C) To people like me, it doesn’t matter what form of government we have.”

A study that I did with Erik Nisbet and Elizabeth Stoycheff found that: Internet use, but not national Internet penetration, is associated with greater citizen commitment to democratic governance. Furthermore, greater democratization and Internet penetration moderates the relationship between Internet use and demand for democracy.

So, I was playing around with the EBRD Life in Transition II data today and did some graphics.


Wow, Swedes love democracy.


The variance in the former Soviet states is really interesting.


And then we have the Caucasus. Looks like Armenians LOVE democracy.
Only half of Azerbaijanis prefer democracy.
Pretty high don’t knows in Azerbaijan and Georgia.

04 Oct

Social Networking Sites in Armenia – Facebook for Elites, Odnoklassniki for Everyone Else

(This is a continuation of a post I made in 2010 about this issue. This analysis based on the 2013 Alternative Resources in Media dataset.)

So, the social networking site that one spends time on isn’t arbitrary. Research tells us that generally people go where their friends are, but also that there are demographic differences in site choice. A study of MySpace versus Facebook in 2007 is the classic case of this. For a variety of reasons, wealthier and more educated people were on Facebook and poorer and less educated people were on MySpace. Nowadays in the U.S., Facebook has sort of taken over the social networking space and a lot of those demographic differences have gone away.

In the 2010 analysis posted above, Facebook in Armenia was quite elite while Odnoklassniki was not. Some reasons include that Odnoklassniki was more accessible on for free or cheap on mobile devices, the Russian language interface was more accessible to those without English skills (at the time Facebook was mainly an English language platform), Odnoklassniki was more about fun (and porn), and Facebook wasn’t great on a mobile device.

Fast forward to 2013, and things have changed. Facebook has grown a lot globally. The Facebook mobile platform is very user-friendly. Russian and Armenian versions of Facebook work quite well. More Armenians are online as well.

So I looked once again at use. Let’s first look at the overall picture before going into the demographic differences.

(Note that this is PRIMARY SNS, they could have accounts on the other site.)
First you can see that a third of all Armenian adults are on a social networking site and 70% of all adult Armenian Internet users are on a social networking site. The Odnoklassniki versus Facebook breakdown is that 10% of all Armenian adults are on Facebook, but 20% of all Armenian adults are on Odnoklassniki. Of Internet users, less than a quarter are on Facebook while nearly half (45%) are on Odnoklassniki. The comparative breakdown is a third of SNS users on Facebook and two-thirds on Odnoklassniki.


In terms of the time that is spent on the social networking site, there weren’t tremendous differences. About a third are on the SNS several times a day and most of the rest of the users are on the SNS at least once a day.


So who is on each site?

(This is based on an ANOVA):

There is no difference in age between Facebook and Odnoklassniki and other SNS users in AGE. However, non SNS users are much older. It is the same story with ECONOMIC WELLBEING – non SNS users are poorer than SNS users. The same is for RUSSIAN skills – the only difference is with non-SNS users and users.
ENGLISH skills – Facebook users (and other SNS users, but they’re weird, so let’s ignore them) have statistically significantly higher English skills than others.
Facebook users are also statistically significantly higher in EDUCATION than everyone else.
No differences in SEX.

So, we can summarize that Facebook users are more elite – English speaking and better educated. Thus, not a big chance from 2010.

(This is based on a multinominal logistic regression):
Looking at this 2013 data, the determinants of being an Odnoklassniki user are the following. In this analysis, all the different factors consider each other, so for example, the influence of higher education on English language skill is cancelled out, so each variable is really telling its own story.

First, Sex – men are more likely to be on Odnoklassniki than women. Next, Russian language skills – those with better Russian are more likely to be on Odnoklassniki than those with poor Russian. Then English skills matter. Then education.
Determinants of being on Facebook are first English skill. Better English means more likely to be on Facebook. Next, higher education. Russian language skill matters next.

For both Odnoklassniki and Facebook, age and economic status didn’t matter much. However, younger and wealthier people are more likely to be online. So it appears that once you’re online, your choice of social networking site is more about language skills and education.

So, with that, what are people doing on social networking sites?

Remember that Facebook users are more “elite” – so the activities they do will likely be more elite as well.

In terms of communicating with friends, posting photos, and entertainment no big differences between the sites. However, Facebook users are more interested in getting and sharing information. Odnoklassniki users are also playing more games than Facebook users.


When asked what the most important activity on a social networking site, Facebook users were much more likely to say getting information.


When asked about sharing political and social information, Facebook users are much more likely to share than Odnoklassniki users. But the majority of all social networking site users aren’t sharing (they say!).


17 Sep

Transnational Families in Armenia and Information Communication Technology Use

New study out today.

Considerable evidence has shown that migrants use ICT to maintain connections to their families and home. Unlike most previous researchers of migrant ICT use, we study those left behind. Using a nationally representative sample in which two thirds of respondent households have a migrant, we determine the effect of this on ICT use. Multivariate analyses including relevant demographic factors that influence both migration and ICT use reveal that transnational (migrant) family status influences frequency of Internet use, Internet ownership, and Skype use, but not other activities. Given the positive social effects of maintaining family connections, ICT use may lessen the negative effects of migration on families and society.

This study uses the 2011 Alternative Resources in Media dataset done by CRRC.

16 Sep

The Internet Cafe is dead – most Armenians get online at home

More from the new Alternative Media USAID-CRRC dataset.

Most Armenians are getting online (primarily) from home. (I wish that it had had some sort of ranking or estimate of hours, but it is what it is.)



And how do people get online at home? Cable, 3G flash card, and still dial up (I can do a breakdown by urban/rural if people are interested.)



I’ll do another post about mobile Internet in the future.


15 Sep

What are Armenians doing online in 2013?

Here’s an update to this 2012 post.


Green is of Internet users and blue is for all Armenian adult citizens.

In the near future I’ll do a break down of these activities by region, sex, etc.

In this recent publication of mine, we show that rural, less educated, and poorer Armenians weren’t engaging in capital-enhancing activities like news reading or blogging. Will this still be true in 2013?