No, nothing about the election here. Instead something about the Georgian retail gasoline market, which according to some is not so competitive. Case in point is this article on an opposition (soon government) leaning news outlet that alleges price fixing in the Georgian retail gasoline market. The article is based on a recent study by Transparency Georgia. A study with some interesting data, but apparently it was all too much for a clueless (or partisan) journalist. But let’s discuss the study itself.
Transparency Georgia finds that:
- The retail gasoline market in Georgia is dominated by five companies, and industry concentration has increased since the Rose Revolution
- There are significant discrepancies between the gasoline imports reported by the Georgian revenue service, and the gasoline exports reported by Georgia’s trade partners.
- There are little to no price differences across companies, and retail gasoline prices change more or less simultaneously.
- The price transmission mechanism is asymmetric. Price increases on the world market lead to price increases in Georgia. Vice versa, price decreases on the world market do not lead to price decreases in Georgia.*
- Between 2006 and 2011 retail gasoline prices in Georgia increased significantly.
All this is seen by Transparency Georgia as evidence in support of the hypothesis that firms in the Georgian retail gasoline market are cooperating and fixing prices.
And indeed it looks and sounds fishy. Or maybe not:
- A small number of firms (an oligopoly) indeed raises the probability of cartel formation. But at the same time a small number of firms is also perfectly compatible with competitive market outcomes. This is in particular true for a market in a homogenous good with transparent prices and small to non-existent search costs for consumers. The small number of firms tells us little to nothing about the likelihood of price fixing in Georgia. Similarly, the consolidation of the industry since the Rose Revolution is at least as consistent with competition driving out unprofitable companies as it is with the existence of a cartel. In particular as a small number of firms seem to be the usual market outcome. For example, Georgia has as many major companies in the retail gasoline market as has Germany (Aral, Esso), despite the German market being larger by magnitudes.
- Discrepancies in trade statistics are nothing new, and in fact to be expected for a country that is also a transit hub.
- This is exactly what we would expect in a competitive market of a homogenous good with little to no search costs to consumers: Firms charge the same price for the homogenous good and change their price almost simultaneously in response to changes in costs.
- That prices rise faster than they fall is observed on many different markets, and in particular in the retail gasoline market. But this observation has little to say about the likelihood of a cartel, as these asymmetries can be explained by switching costs and the local market power of gas stations, or inventory adjustments.
- I have difficulties understanding why the Transparency Georgia study associates the price increases over the 2006-2011 with the consolidation of the industry but fails to even mention the evolution of world oil prices. Which perfectly explain the movement of gasoline prices in Georgia.
It might well be the case that prices on the Georgian retail gasoline market are fixed or that a cartel exists. But the data presented in the study is not evidence for it, if not against it. It’s thus puzzling to see Transparency Georgia calling for an investigation and intervention by authorities, and for this study to be cited as if it is serious evidence in various news outlets. Hopefully no antitrust and competition policies will be based on it.
* I have doubts about this result. The time series is rather short and nothing in the study suggests that this result is based on a rigorous analysis of the price transmission mechanism or a statistical test for asymmetry.
Comments
Leaving the poor "journalist" alone, I had a very similar impression of a previously published Transparency International report on competition the Georgian pharmaceutical sector (http://transparency.ge/en/post/press-release/pharmaceutical-market). The report contained some valuable factual observations about the market structure (company shares, vertical integration) but then engaged in rather shaky value judgments and interpretations. One that I "loved" the most is this:
"While these companies have played an important role in the pharmaceutical sector’s development in Georgia, their market dominance leads to high markups for medicines which explains the high expenditures on pharmaceuticals. The markup for medicines in Georgia is far above the average markup for medicines than in European countries."
How do we know that a markup is high or low? What benchmark is to be used? Should we compare Georgia's tiny pharmaceuticals market with that of the European Union?
Finding evidence for collusion and price fixing may be as hard as to find a black cat in a dark room. Not known for its economics expertise, Transparency International-Georgia seems to be successful in accomplishing this feat even in cases when the black cat is not there.
Rather than to bash Transparency International (by far the most established Georgian democracy watchdog), the main point of my comment is this: the Georgian think-tank community could benefit from greater specialization and a system of professional peer review.
In favor of TI survey,I agree with the fact that "price transmission mechanism" seems to be really asymmetric in Georgia. View is that, recent researchers agree about the general trend that price increases on the global oil market don't not lead to the price increases on the domestic markets anymore, or influence seems to be reduced in developed countries as well as in developing countries. On the contrary of this argument, Governmental authorities in Georgia, while are being accused in protection of monopolistic activities and therefore in high mark-ups on the domestic market,try to throw public attention toward global markets and give general explanations about the price changes on the foreign markets. Actually, this was the concern, which pushed me to write my master work
Looking forward to TI's response! Have you asked them?
Otherwise agree with your point that you need specialization for think tanks to be on top of all the details.
It seems to me though that think tanks in Georgia are already quite specialized, what means that there is little scrutiny and little dialogue.
Regarding this TI study there is the problem that the author was completely unfamiliar with the relevant literature. But it's not just that as there are blunders in the study that even a non-economist could easily have avoided. As for example comparing retail gasoline prices in different countries without accounting for the fact that taxes differ from country to country.
Not sure what exactly went wrong here. I would wholeheartedly agree with that peer review is important.
Thank you very much for your interest in the TI report and let me provide my feedback on your remarks.
I would suggest the trends observed and discussed in the report be analyzed jointly, not one by one, as it has a crucial importance to fully understand our concerns that competition in the Georgian gasoline market might be restricted. This report in question wasn't meant as an academic paper by TI board itself, but rather for general public. For this reason, the report presents simple terminology as much as possible. A quick outlook of the line of our reasoning below may help you to make main findings of our report more apparent.
As the report has it, an essential price parallelism is observed among various companies operating in the gasoline market and there are repeated episodes of parallel conduct. Academic literatures suggest the different explanations for the phenomenon of ‘price parallelism’: some being the product of anti-competitive behavior, and others being the product of innocent behavior by companies. At the same time, an academic literature shows that choosing the practice of same prices (or ,price parallelism’) will unambiguously facilitate collusion. Indeed it has been found that there might be circumstances when firms might have an incentive to coordinate on the same price categories, as a way to make collusion more likely. Among the arguments, that help explain why committing to the same prices favors collusion and why price parallelism might help firms both to reach a collusive agreement as well as to enforce it, are: it allows firms to achieve the highest collusive profits and it makes deviation less profitable. Moreover, same prices may help firms solve coordination problems and may make the market more transparent for producers, enhancing observability of rivals’ actions (or facilitating reciprocal monitoring), and thus making collusion easier to sustain. Yet, one must take into account that same prices may also facilitate price comparisons by consumers, thereby increasing transparency on the consumer side and making the market more competitive. Some academic papers also explored that if the latter effects dominates and goods are differentiated, improved transparency on the consumer side makes collusion harder to sustain, i.e. firms may prefer to commit to different prices in order to put some friction back into the market and ,obfuscate’ consumers. While, in almost homogeneous markets, the effect vanishes and changes in transparency have almost no effect on collusion.
On the other hand, the literature acknowledges the possibility that in some circumstances price parallelism might simply be the natural outcome of normal competitive interactions among firms. For instance, firms experiencing similar demand and technological conditions will tend to use the same prices or common exogenous shocks may lead firms to increase prices proportionally. Further, there are other explanations too, such are history, habits, etc.
To the best of our knowledge, in the real world, a possibility that price parallelism may be a result of coordination among companies who want to better sustain a collusive outcome cannot be excluded. Moreover, economic theory has not said the last word on the issue of whether price transparency tends to make collusion more likely (by allowing rival firms to better monitor each other, and thus policing a collusive outcome), or competition more aggressive (by allowing consumers to better compare price offers, and thus giving firms higher incentives for aggressive price behavior).
Indeed, if one ask the broader question of whether one can ever find an infringement of anti-trust laws by simply looking at parallel conduct - the answer is that this is possible. But, high standards of proof is required. In particular, in order to further bolster the inference of collusion, an additional economic circumstantial evidence (known as ‘plus factors’) is needed. Plus factors focus on what modern economic knowledge interprets coordinated behavior of oligopolistic companies. Scholars have also recognized that certain industry structures, firm histories, and market environments are conducive to collusion. These include but are not limited to Industry characteristics (product homogeneity, readily observed price adjustments, formal and/or informal entry barriers, high concentration, stable and inelastic demand, institutional factors etc.) that are conducive to successful coordination. All these ,plus factors’ were observed in Georgian market and are included in our report. Moreover, gasoline market structure is oligopolistic. It’s a textbook knowledge that collusion most often takes place within the small number of market players, or otherwise stated - market structure of oligopoly may favor anticompetitive conduct.
However, formalization of the above-mentioned arguments and making of final assessment on the genuine rationale of observed price parallelism is a competence and a task for the Competition Authority: Why gasoline companies on the Georgian market tend to choose same prices? is this a non-collusive motive that might lead different gasoline market players to come up with the same prices through pure non-cooperative behavior or it is the product of anti-competitive behavior? the Competition Authority should address these issues by providing competent judgment. TI also proposed to assess an impact of consolidated market structure (leaving other things equal) in gasoline price increase in Georgia. TI’s mission is to identify these trends which may be harmful to consumers’ interests and call for attention of public policy makers. I’m just wondering as to why you might disagree to our call to Competition Authority for further scrutiny and why the Competition Authority shouldn’t check whether the equilibrium established in the gasoline market is a natural outcome.
I also find groundless to compare a number of major players on the German market with that of a Georgian market. You are missing one important point here. This is a case when comparing numbers only do not make sense. In particular, in Germany the abuse of market power is prevented and protection of competition is enforced by The Federal Cartel Office (Bundeskartellamt) on the national level and by the EU Competition Commission on the EU level, in addition. Whileas, Georgia has no competition policy/law/effective agency over the last 7 years. Don’t you think this institutional factor could adversely have influenced competition on the Georgian market and the gasoline market can be anti-competitive?
You pointed to insufficient information and asked to provide more evidence. You must know though how hard it is in Georgia for NGO to obtain any statistical information for two reasons mainly: GeoStat (statistics authority) is producing completely unfit information for academic purposes and Georgian companies are not supportive to any academic study as well. Our report includes data which was possible to obtain in Georgia during the time frame of the study. It doesn’t include an econometric techniques. We do only provide simple correlation statistics which shows evident trend of parallel conduct by gasoline companies. Georgian experts also confirm that the same trends are observed over the recent years. I’d be glad if you can produce alternative study that would arguably demonstrate that Georgian gasoline market is highly competitive.
As for your remark ,,…there are blunders in the study that even a non-economist could easily have avoided. As for example comparing retail gasoline prices in different countries without accounting for the fact that taxes differ from country to country’’. This graph is included in the introductory part, just to give information to general public about gasoline prices of our close neighboring countries and doesn’t constitute the grounds for our main findings.
Regarding the discrepancies in trade statistics: we are aware of the fact that the Georgia isn’t the only country which underreports import statistics. However it may not be an argument against checking, identifying or looking for the answers on the questions concerning the problem. Moreover, there are unpaid taxes on customs and missing budget revenues behind it. I can’t find a rationale behind your argument concerning the underreporting of the Italian ,API’ gasoline import numbers and Georgia being transit hub for ,API’ gasoline transitting to neighboring countries provided economic geographics of the region. Consequently, we had to guess that all ,API’ gasoline with country of destination - Georgia from Italian customs have been imported to Georgia. Moreover, ‘API Super’ gasoline is sold at the Georgian gas stations, while according to the information officially provided by the Italian customs and the Revenue Service of Georgia, no import of ,Super’ gasoline was registered from Italy to Georgia in last years. It’s a violation of consumer rights, illegal use of trademark and criminal liability according to Criminal Code of Georgia. Unfortunately, gasoline companies were unsupportive and did not supply any comments on the trade statistics discrepancies.
Thank you for your detailed response. Will respond in detail in private. But for the moment I think it is very important in these kind of studies to move beyond undergraduate textbooks and to refer to the literature. Which is extensive (and inconclusive) when it comes to retail gasoline markets. The other point is that given the very ambigious evidence (at best) one should be careful with judgements and conclusions. It's maybe just a matter of language but the study makes it look that hard evidence has been found, and that thus an investigation is justified.
Natia, your last paragraph contains VERY strong statements (violation of property/consumer rights, tax evasion...) without proper justification. I certainly don't know what was provided be Italian Customs but you can easily check those numbers on Eurostat and see that since appearance of API Euro 5 we only import "Premium" gasoline from Italy and export quantity in ton exactly match the imports published by Geostat. www.tradamap.org includes all those numbers. I not protecting Wissol here (actually, I prefer fuel from other retailer), but in general, why do you think that such a well-positioned conpany would spend so much money on a single brand marketing and then sell other product instead taking all the risks potential consequences?
I would suggest to read TI blog post:
http://transparency.ge/en/blog/wissol-has-stopped-selling-api-super-branded-gasoline
I think we as scholars do not have much to argue on. However TI has been a pioneer who have tried to take on with the problem as the competition authority lacked competence to do so. Despite not being academic paper we did our best to take into account the most recent academic views in the study. Yet, being prevented from the access to data (oil companies were unsupportive and TI lacked power to obtain data) we have focused on trends only, deferring the privilege of the full-fledged study to the Competition Authority. Our report has highlighted an importance of powerful Competition Authority in line of the EU recommendations and international standards.
We presented arguments from recent academic literature, such are: observed ,price parallelism’ and ,plus factors’(described as in report as well as above) which provides grounds for legitimate assumption that there MAY BE attempts by market players to coordinate.
The crucial factor in the story is that in Georgia the competition authority was completely ineffective during last years. So, taking all these factors into consideration, the assumption that the Georgian market in terms of competition could have a problem, has more than the reasonable foundation. Because of the lack of transparency in the gasoline market, because of observed trends that according to international practice might rise concerns in terms of competition, TI proposed to competition agency to undertake a competition assessment that you disagree regrettably.
Anyway, the fact is that the topic in question isn’t the area where there is much space for absolute (fixed) concepts. Even in judicial analysis of the factors to justify an anticompetitive conduct, outcomes are largely dependent on courts’ unarticulated intuitions about the likely cause of observed price parallelism. The final judgments always are made on the basis of case by case studies by the competent bodies having full access to information.
is there truth to the claim that prices have been falling in Georgia after the elections, in pharmaceuticals, after Ivanishvili's public claim that there was price fixing?
I don't know about pharmaceuticals, but I noticed that at some point there was an 8 tetri spread for Euro Regular at Wissol and Socar. One week later this price spread has all but vanished, but it's still an interesting anomaly. Or so it seems. One explanation is that there is a cartel, another that Wissol wanted to earn some political brownie points, and another would be based on the different cost structures or inventory strategies of Wissol and Socar. But as I argued in my blog post I doubt that one discriminate between these theories based on the price data.
I bought Diovan by Novartis at PSP for my parents just for 43 GEL, before it was 75 GEL. I have asked GPC they said yes, prices are cheaper now.
But when it comes to Diovan specifically a likely explanation is the expiration of the patent, see this Forbes article.
We buy Nolaxen, Diovan, Cerebrovin, Detralex... No visible changes in price since October 2012.
I just asked my friends about that, some of them mentioned: plavix, nimesil among .
Also Aversi representative said in recent TV Report by 9channel that Novask now is offered for 60 and now for 12 GEL. Imigran 80 and now 12 GEL.
http://www.youtube.com/watch?v=12IOEbMjhOc
But they did not mention if new prices are offered by manufacturers or importers.
P.S. Sorry if names are misspelled.