Estimating Georgian Dream Bribery in GVote24

How to approach a systematic estimate

It is known that Georgian Dream engaged in extensive vote-buying for the October 26 parliamentary elections. Among sensible people this is not under discussion. The question is still what the scale of the bribery and what the total amount of spending would be. (As mentioned, I am doing a report on this that will come out tomorrow.)

My rough estimate (note: estimate, rough) is that probably more than 220.000 people received an incentive and that around 40m GEL were spent on direct vote buying, not taking into account other social spending and other favours or relief. To be clear, I am not making light of a rigged election as a speculative exercise, though it could sound that way. I am trying to get a more precise sense of what actually happened in the criminal rigging, and how it worked. I think bribery was a big part.

Here is an approach to estimate this bribery, and the model is for anyone else to test also, and to come up with other plausible estimates. At the end of this post there is a link to the spreadsheet.

To start with,

1/ put your assumption of unwilling GD voters in 2020, who needed to be induced with some money to vote for the Georgian Dream. With 100.000, my estimate for this is almost certainly on the conservative side. We already heard a lot about vote buying in that election back then. Less than this time, but still. (Incidentally, people from Giorgi Gakharia’s party who were with GD at the time might help with this number.)

2/ estimate the change in GD genuine enthusiasm since 2020. We have extensive surveys in that regard, and my assumption of -6% is one that is very generous to the Georgian Dream. Even Bidzina Ivanishvili spoke of people getting tired of his party. His own polls were showing that, but more on that later.

Konstantin Morgoshia of Alt-Info (not a friend of the opposition parties) assumed that Georgian Dream only had about 600.000 substantive supporters, only a bit more than half the people who actually voted for them. If you broadly agree with Batono Konstantin, you could put roughly -35% into this field. If for some reason you think that a country ruled by one man’s whims is a great thing (that crowd is quite active on Twitter), you can also add +10, and perhaps your story will be that all of the cash that changed hands was really donations to the Georgian Dream by happy citizens. (Upside of this model: it really works for everyone.)

My sense is that from all that we see support really did go down substantially. Does the Georgian Dream still enjoy support? Does the war vs peace narrative work for some? Absolutely, it is the one pillar of Georgian Dream that still holds. That support is not a majority, from all the surveys I could see.

3/ put in what you think the average bribe was. In the past, we heard 50-80 GEL, this time I have not once heard a sum under 100 GEL. I have repeatedly heard of 200 GEL, including in the form of vouchers. I have also heard higher sums. I am putting in 150 GEL into this field.

4/ what percentage of unwilling voters do you need to pay? Some voters you can instruct and intimidate, you do not need to offer any inducements. However, there is a limit to that, otherwise Georgian Dream may well have a revolt on its hands, and dealing with that is a lot more expensive than buying up people one-on-one before. My assumption is that you pay off about two thirds. It’s a good rule of thumb to keep the majority on your side and incriminate them a bit, too. I can also imagine that this goes higher or lower.

5/ how much do coordinators want to keep as a bonus? If people handle a lot of cash, “human temptation” (as Bidzina Ivanishvili put it) inevitably creeps in. My assumption is you have to give people at various levels a cut, or they will start taking it themselves and you do not want that aggravation. Not every coordinator takes 20%, but my assumption is that in the chain, 3-4 people handle the cash, and they know they can take 5% each for their own incidentals, so to speak.

I think that is the model. With my estimates (some of them arguably favouring the Georgian Dream, such as the low base of unwilling voters in 2020) you come to about 40m GEL in bribes. I think I did the formulas right, but perhaps there is a snag in there somewhere, any pointers gratefully received.

Here is the spreadsheet, you should be able to enter your estimates online, or download for your own use. The relevant cells are protected, so you cannot do damage.

Now you can come to radically different conclusions. Perhaps you believe there isn’t a single unwilling Georgian Dream voter. Perhaps you think 90% were intimidated. It’s also possible that you think the average bribe was higher or lower. With this model, we have a slightly more structured way of discussing this question, and estimate how many people were bribed, with how much in total.

The 40m GEL for the bribery is, of course, not the total cost of the Georgian Dream campaign: Georgian Dream were paying lots of coordinators (who were paid for many months, though many are also straight state employees), huge rallies and transport for that (easy to estimate, as we know the transport costs per passenger, roughly how many people were there, and you did not go on this trip without getting a lunch/dinner), lots of outdoor advertising (some of it almost certainly given for free), digital spending (Meta Ad Library), headquarters and advisers — but for the bribery, at least, we have some contours.

Incidentally, one trace all of this bribery is likely to leave is that there will be outgoing cash in the two banks associated with the Georgian Dream. Will there be a small inflationary surge? Hard to tell, it might be too small an effect.

Curious to hear what others come up with, and you can use the comment function or write to me. Again, the idea is to have a model to understand a terrible thing: how Georgia’s future was bought, sold and stolen in one go.