A Hard Look in the Mirror: Central Asia Scores Poorly on Perception of Corruption

by Ryan Weber

Transparency International Released its 2010 Corruption Perception Index (CPI), and as the regional map shows (above), Central Asia is left looking rather red in the face. On a scale of 0 to 10, in which lower numbers were judged more corrupt by respondents, the regional regimes scored between 2.9 (Kazakhstan) and 1.6 (Uzbekistan & Turkmenistan). Georgia, China and Russia provide some context, with the first two - by no means poster children for anti-corruption - scoring higher than any Central Asian states at 3.8 and 3.5 respectively, while Russia, the dominant political force in the region, held steady with the pack at an embarrassingly low score of 2.1.

The Index, which is compiled from a series of surveys filled out by citizens of each country then averaged and adjusted according to the survey biases, is just one measure, and an admittedly imperfect one, to evaluate the level of corruption active in the everyday, as well as institutional levels of government. Corruption is a difficult concept to define, ranging from police traffic bribes to no-bid government contracting, and even more complicated to track accurately. The index provides a "rough barometer" of public perceptions, captured at the moment they take a survey, and not held against a baseline experience. As such, as the Transparency International website warns,

"The CPI is not a tool that is suitable for trend analysis or for monitoring changes in the perceived levels of corruption over time... Year-to-year changes in a country’s score can result from a change in the perceptions of a country’s performance, or changes in the methodology resulting from TI’s efforts to improve the index." (link)

Well and good, but trend analysis may in fact provide some interesting results when considered in light of the events affecting the various countries, or as a measure of the similar or dissimilar year-by-year corruption perception of citizens from different countries within a shared region. It must still be stressed, what such analysis will show is not the level of corruption in any country, but the perception of corruption among its citizens.

In a way, this can be misleading - institutional corruption can involve much larger sums of money, but be less apparent to those outside of governments. But it is also instructive in sociological ways that a more accurate "corruption score" wouldn't record. For example, if citizens felt more positive about the direction their government was moving in, or that it was making progress relative to earlier corruption, they might rank it higher. Likewise, it's worth saying that cross-country analysis is also difficult, as different citizens will be judging their own governments differently based on expectations of what is, or is not, an acceptable act or presence of corruption.

Before we start going against TI's well-established grain, let us first consider the 2010 Index itself; what it says about the countries of our focus, and how Central Asia and the Caucasus compared to their neighbors and distant (comparables).

What is immediately clear is that the region in question does not fare well in perception of corruption by its own citizens. The average score of the core countries in question (Armenia, Azerbaijan, Kazakhstan, Kyrgyzstan, Georgia, Tajikistan, Turkmenistan, Uzbekistan) is 2.4, a number which would lower with the addition of regional neighbors Iran, Afghanistan, Pakistan, or Russia. Bracketing this region are China (3.5), India (3.3) and Turkey (4.4), all of which enjoy dramatically higher scores, despite still being unimpressive considering the 10 pt scale involved. Turkey, the highest-scoring neighbor country, only ranks 56th out of the 178 countries studied. The top end of the scale is mostly found in Scandanavia, where Denmark, Finland & Sweden all score over 9. The US, UK, and Japan are all in the 7's.

Interestingly, while per capita GDP and the CPI seem to have some correspondence - with higher scores among higher-earners - the countries we are most interested in present a series of outliers. This is true both when ranked by CPI score (above right; red highlights for all GDPs over $9,000) or when ranked by per capita GDP (below left; blue highlights for scores at or below the 2.4 regional avg.).

Of the 8 core countries, Uzbekistan and Turkmenistan scored the lowest at 1.6 each, sharing the 172nd rank with eachother as well as Sudan. Only Iraq (1.5), Afghanistan (1.4), Burma (1.4) and Somalia (1.1) scored lower. Kyrgyzstan ranked just 7 spots ahead (#164) with a score of 2.0 - the same as the Democratic Republic of Congo - while Tajikistan and Russia tied with a bevy of other under-performing states at 2.1 .

Georgia was the highest-scoring core country at 3.8, followed - distantly - by Kazakhstan (2.9). Armenia (2.6) and Azerbaijan (2.4) were middle-of-the-pack in this unglamorous regional portrait.

Because the CPI is based on perception, its authors and critics are both quick to point out how events, situations, and even rumors, beyond the control of the governments under review can affect the results. This is mitigated to some degree by the size of survey sample, the use of different survey types, and the distribution of survey periods over a long period, from January 2009 to September 2010. This also provides an inherent disservice to a strictly one-year view because the resulting Index represents in effect an averaging of citizen opinion over that period. Should a dramatic event occur that would alter that perception - in either direction - the change, or we might think of its momentum - will not appear in the Index.

The recent turmoil, nationalizations, and seeming-democratic breakthrough in Kyrgyzstan is one possible example. The study period includes surveys from before former-President Bakiev's questionably re-election campaign, during the elevated levels of corruption perception following his appointing his unpopular son, Maksim, to oversee a major Russian aid package, and during the difficult political overthrow and inter-ethnic violence that marked the spring and summer of 2010. It is difficult to imagine a static public corruption perception throughout this time period, but it is equally dubious to predict or assume how each event affected citizen perceptions of current and future corruption. The possibility of a more stable government with greater transparency on the horizon bodes well for the Kyrgyz CPI 2011, and hopefully for the Kyrgyz people in general, but only time and further survey data will tell.

With this as an acknowledged short-coming of the data-as-meant-to-be-used, I would like to intentionally engage in misusing the CPI dataset in the interest of seeing what it turns up.

To do this, I first looked at a data table of the respective CPI scores for each of the 8 core countries as well as the dominant regional players, Russia and China, from the years 2005 to 2010. I could have easily extended the study back to 2000, but given the tenuous nature of the endeavor already, the incredible transformation of most regional countries over the past decade, and the dramatically improved survey methodology, I decided to limit my analysis to the past 5 years (6 reports).

In addition to the raw scores, I included a simple calculation of Change, comparing the 2005 and 2010 scores. This helps negate some of the year-to-year 0.1 and 0.2 score variance, and clarifies which countries experience sustained change in their corruption perceptions. The number that immediately jumps out is that of Georgia, which since 2005 has improved its CPI score a full 1.5 pts, from 2.3 to 3.8. Most of that change came dramatically in 2006-2008, peaking at 4.1 (2009) before sliding back to 3.8 this year.

On the losing side of the 5-year trends were Armenia, Kyrgyzstan, Turkmenistan, Russia and Uzbekistan, the last of which managed to drop 0.6 pts. This is not as dramatic a figure as Georgia's +1.5, but in the last 1 pt it becomes more difficult to score lower - as evidenced by Somalia's 2010 CPI of 1.1. It is worth noting that in 2005 Georgia and Uzbekistan were separated by only a 0.1 score; 5 years later, they are a full 2.2 pts apart. Ironically, that is approximately equivalent to the CPI gap between China and Taiwan (5.8)

Due to the amount of data, it rapidly became clear that looking at it all at once would just present a jumble, so I first divided it into 2 groups - the Caucasus (AZ, AM, GE) and Central Asia (KZ, KG, UZ, TM, TJ). In the interest of considering influence, I also included Russia in both groups, and added China to the Central Asia group. Admittedly, both categorizations are reductive - a more full consideration of 'regional influence' would have to include all the countries listed at the top of this post (Turkey, Afghanistan, Iran, etc.), but the point is to illuminate, not obfuscate.

The resulting data, when graphed over time, presents the following observations/ surprises:

Caucasus

1. For the 2nd year in a row, Russia ranks worse than all other Caucasus countries in regard to public perception of corruption. Admittedly, this Russian figure includes mostly non-Caucasian Russia, though it would be interesting to find a regional CPI, say for Chechnya.
2. After remarkable progress since 2005, Georgia showed evidence of a marked downturn in 2010. It is worth pointing out that progress in reducing corruption perception has slowed, and now reversed, since the 2008 August war with Russia.
3. Starting in 2008, Azerbaijan finally, and robustly, ended its CPI slide. The Azeri benefits from increased energy revenues may be salving the continued presence of corruption, or Azeri officials may be cutting down on overt corruption in an attempt to mollify and/or attract more foreign investment.
4. Armenia, which has historically had a very stable CPI score, continued its slow decline for the 3rd straight year, cause undetermined.
5. As a region, the 3 countries of the Caucasus present no single, clear trend. Georgia is up, then down; Azerbaijan is down, now up; Armenia slides slightly downward.

Central Asia

1. China far exceeds all Central Asian states in having a much better corruption perception index. As with Russia in the Caucasus, it would be interesting to see a breakdown of similar studies for the bordering Xinjiang province, but those numbers are not available.
2. Russia as a whole, which ranked worse than all Caucasian states, scores as well or better than all Central Asian states except Kazakhstan (which it did out-score in 2007).
3. Kazakhstan, like Georgia, is on a vibrant corruption reduction streak. While still achieving levels that are far from laudatory, progress should be commended. The likelihood that such efforts are a reflection of international pressure, especially regarding Kazakhstan's OSCE chairmanship, seem high.
4. As a whole, the regional trend is downward, and seems to be taking its cues very much from Russia.
5. The "good" news? In Central Asia, all year-to-year movements - positive and negative - are very, very small, usually just 0.1 or 0.2 points. While this means there is rarely dramatic improvement in the public's perception of corruption, it also means that, for the most part, the public view corruption as more-or-less the same year-by-year, with a slight negative trend. For countries with by-and-large low per capita GDP and varying degrees of semi-authoritarian governance, the stereotype of a public that is disappointed, but feels unable to influence change, is one logical explanation for the long-term perception of high, but not increasing, corruption.

Finally, I'd like to put all the numbers in one space with the two elephants in the room (Russia and China) compared to the averages of the regions I've described as the Caucasus and Central Asia. The Transparency International surveys specifically work within individual countries, so combining regional numbers is dubious, but presents a reasonable, perhaps helpful result. China, for all its faults, continues to provide a positive example of corruption minimalization (or at least hiding), while the Russian government is clearly following policies that maintain or decrease their citizens' confidence in its appropriate operation. The Caucasus (indirectly) seem to be following the Chinese trajectory, while Central Asia is apparently content to echo that of Russia.

As inarticulate, and potentially inaccurate, as it may be, this is the Perception conveyed by a study of the 2005-2010 Corruption Perception Index.