The FINANCIAL — Over the last decade, the United National Movement made impressive strides toward a more stable and safe society.
They successfully uprooted entrenched criminal organizations and police corruption through uncompromising reforms, and oversaw one of the fastest decreases in crime in modern history. Outgoing President Saakashvili is rightfully lauded for these impressive ends, but many of those patting him on the back have also told him that his means, especially the rubber stamp machine known as the Georgian judiciary, were not things to be proud of.
Over the last year the new government has been made moves to undo some of what UNM did, perhaps most notably by releasing thousands of inmates from Georgia’s prison system. The argument for these releases seems sound, if the execution was not: when conviction rates are in the impossible-to-believe range of 99%, there are certainly innocents in jail. Moreover, some of those released are considered by the new government to have been political prisoners.
Those in opposition had another valid argument, one worth considering: releasing thousands of prisoners without careful consideration has put murderers and robbers back on the streets, free to recommit their crimes. Has Georgia lost its proud position as a public safety prototype, or was all this concern for nothing?
Survey and Results
For the previous three years, Georgian Opinion Research Business International, the Georgian member of Gallup International, worked with the EU and the Georgian Ministry of Justice to conduct victimization surveys in Georgia. These surveys allowed us to draw a more accurate picture of crime than official statistics, as underreporting crimes is very common in Georgia.
What we discovered from 2010 to 2012 is that crime in Georgia, including crimes that went unreported, dropped incredibly over the last decade. In fact, in 2011 our victimization survey showed lower crime rates in Tbilisi than in every European capital city. Saakashvili’s draconian punishments and “judiciary” did in fact succeed in curtailing crime. It should naturally follow, then, that easing these punishments would raise crime rates.
Therefore, we included a shortened version of the victimization questionnaire in our recent omnibus survey. This included about a dozen questions from the original survey, meant as a sort of canary: margins are higher with a smaller sample and the questions are not as in-depth, but any major increase in crime would easily be seen in the data. To my complete and absolute surprise, the data show no such thing.
Victimization Rates in Georgia (2012, 2013) and Europe (2005)
GE 2012 GE 2013 Germany Denmark Poland
Assaulted 0.8 0.4 12.1 9.6 8.7
Robbed 0.2 0.7 1.8 3.0 5.1
Burgled 1.6 1.9 3.5 10.9 6.1
Car Theft 0.2 0.4 2.0 4.8 2.6
The percentage of respondents who said they were robbed, burgled, or had a car stolen is indeed slightly higher this year than last. While you might want to take this as evidence of an actual increase, I’m obligated to remind you that in statistics we can only get close: these results fall well within the margin of error, even when we adjust for an assumed response distribution (a concept which I discuss at the end of this column). Put simply: if crime rates did increase, it was a small enough change to not be verifiable with a single survey.
Preferred punishment for a second TV theft
2012 2013
Fine 20% 14%
Prison 22% 32%
Community service 37% 37%
Suspended prison sentence 15% 14%
In fact, the only way Georgians’ criminal environment has changed is a slight increase in the desire for imprisonment. A simple question we used to gauge punitive attitudes was as follows: “Take for instance the case of a 21-year old man who is found guilty of breaking into someone's home for the second time. This time he has taken a TV. Which of the following sentences do you consider the most appropriate for such a case?” Last year, about as many Georgians said a fine (20%) was appropriate as did prison time (22%). This year, a substantially greater number would like to see prison for the repeat thief than a fine (32% vs 14%). A plurality both years supported community service (37%).
If you’re skeptical and think that crime has indeed increased in such a way that it’s hidden between the margins, I’ve included victimization rates for a few other countries (using the same questionnaire). While the European data is a bit old, it can give you a good idea of just how large an increase in crime Georgia would have to see to reach European levels. So even if there has been a slight jump in crime, it has been nowhere close to the scenarios that were put forth in opposition to prisoner release.
A note on analysis
In order to conduct analysis on such tiny portions of a society, we cannot use standard methods to calculate margins of error. Most simple error calculations assume that a populations’ true response distribution will be 50%, which for our sample would give us margins of 3%. Margins such as these wouldn’t do us much good, because if crime increased substantially it would still fall within the margin of error.
However, when we have a good idea of what the actual response distribution is, we can adjust our calculations to give us smaller margins of error without affecting confidence intervals. We have some very good and consistent data from the past three years that tells us approximately what portion of society was victim to a given crime. With the exception of burglary, every crime was below the 1% mark in 2012. For ease of understanding, I’ve calculated the margins for these data as though each distribution was 1%.
Using this technique, we have a margin of error of .6%, small enough to make some better discussion. For the punishment question, the margin of error is the standard +/- 3%. Visit our website at gorbi.com for more articles.
Discussion about this post