Tag: Competition

Energy Storage: Opportunities and Challenges

Wind turbines in a sunny desert representing energy storage

As the dramatic consequences of climate change are starting to unfold, addressing the intermittency of low-carbon energy sources, such as solar and wind, is crucial. The obvious solution to intermittency is energy storage. However, its constraints and implications are far from trivial. Developing and facilitating energy storage is associated with technological difficulties as well as economic and regulatory problems that need to be addressed to spur investments and foster competition. With these issues in mind, the annual Energy Talk, organized by the Stockholm Institute of Transition Economics, invited three experts to discuss the challenges and opportunities of energy storage.


The intermittency of renewable energy sources poses one of the main challenges in the race against climate change. As the balance between electricity supply and demand must be maintained at all times, a critical step in decarbonizing the global energy sector is to enhance energy storage capacity to compensate for intermittent renewables.

Storage systems create opportunities for new entrants as well as established players in the wind and solar industry. But they also present challenges, particularly in terms of investment and economic impact.

Transitioning towards renewables, adopting green technologies, and developing energy storage can be particularly difficult for emerging economies. Some countries may be forced to clean a carbon-intensive power sector at the expense of economic progress.

The 2021 edition of Energy Talk – an annual seminar organized by the Stockholm Institute of Transition Economics – invited three international experts to discuss the challenges and opportunities of energy storage from a variety of academic and regulatory perspectives. This brief summarizes the main points of the discussion.

A TSO’s Perspective

Niclas Damsgaard, the Chief strategist at Svenska kraftnät, gave a brief overview of the situation from a transmission system operator’s (TSO’s) viewpoint. He highlighted several reasons for a faster, larger-scale, and more variable development of energy storage. For starters, the green transition implies that we are moving towards a power system that requires the supply of electricity to follow the demand to a much larger extent. The fact that the availability of renewable energy is not constant over time makes it crucial to save power when the need for electricity is low and discharge it when demand is high. However, the development and facilitation of energy storage will not happen overnight, and substantial measures on the demand side are also needed to ensure a more dynamic energy system. Indeed, Damsgaard emphasized that demand flexibility constitutes a necessary element in the current decarbonization process. However, with the long-run electrification of the economy (particularly driven by the transition of the transport industry), extensive energy storage will be a necessary complement to demand flexibility.

It is worth mentioning that such electrification is likely to create not only adaptation challenges but also opportunities for the energy systems. For example, the current dramatic decrease in battery costs (around 90% between 2010 and 2020) is, to a significant extent, associated with an increased adoption of electric vehicles.

However, even such a drastic decline in prices may still fall short of fully facilitating the new realities of the fast-changing energy sector. One of the new challenges is the possibility to store energy for extended periods of time, for example, to benefit from the differences in energy demand across months or seasons. Lithium-ion batteries, the dominant battery technology today, work well to store for a few hours or days, but not for longer storage, as such batteries self-discharge over time. Hence, to ensure sufficient long-term storage, more batteries would be needed and the associated cost would be too high, despite the above-mentioned price decrease. Alternative technological solutions may be necessary to resolve this problem.

Energy Storage and Market Structure

As emphasized above, energy storage facilitates the integration of renewables into the power market, reduces the overall cost of generating electricity, and limits carbon-based backup capacities required for the security of supply, creating massive gains for society. However, because the technological costs are still high, it is unclear whether the current economic environment will induce efficient storage. In particular, does the market provide optimal incentives for investment, or is there a need for regulations to ensure this?

Natalia Fabra, Professor of Economics and Head of EnergyEcoLab at Universidad Carlos III de Madrid, shared insights from her (and co-author’s) recent paper that addresses these questions. The paper studies how firms’ incentives to operate and invest in energy storage change when firms in storage and/or production have market power.

Fabra argued that storage pricing depends on how decisions about the storage investment and generation are allocated between the regulator and the firms operating in the storage and generation markets. Comparing different market structures, she showed as market power increases, the aggregate welfare and the consumer surplus decline. Still, even at the highest level of market concentration, an integrated storage-generation monopolist firm, society and consumers are better off than without energy storage.

Fabra’s model also predicts that market power is likely to result in inefficient storage investment.

If the storage market is competitive, firms maximize profits by storing energy when the prices are low and releasing when the prices are high. The free entry condition implies that there are investments in storage capacity as long as the marginal benefit of storage investment is higher than the marginal cost of adding an additional unit of storage. But this precisely reflects the societal gains from storage; so, the competitive market will replicate the regulator solution, and there are no investment distortions.

If there is market power in either generation or storage markets, or both, the investment is no longer efficient. Under market power in generation and perfectly competitive storage, power generating firms will have the incentive to supply less electricity when demand is high and thereby increase the price. As a result, the induced price volatility will inflate arbitrage profits for competitive storage firms, potentially leading to overinvestment.

If the model features a monopolist storage firm interacting with a perfectly competitive power generation market, the effect is reversed. The firm internalizes the price it either buys or sells energy, so profit maximization makes it buy and sell less energy than it would in a competitive market, in the exact same manner as the classical monopolist/monopsonist does. This underutilization of storage leads to underinvestment.

If the model considers a vertically integrated (VI) generation-storage firm with market power in both sectors, the incentives to invest are further weakened: the above-mentioned storage monopolist distortion is exacerbated as storage undermines profits from generation.

Using data on the Spanish electricity market, the study also demonstrated that investments in renewables and storage have a complementary relationship. While storage increases renewables’ profitability by reducing the energy wasted when the availability is excess, renewables increase arbitrage profits due to increased volatility in the price.

In summary, Fabra’s presentation highlighted that the benefits of storage depend significantly on the market power and the ownership structure of storage. Typically, market power in production leads to higher volatility in prices across demand levels; in turn, storage monopolist creates productive inefficiencies, two situations that ultimately translate into higher prices for consumers and a sub-optimal level of investment.

Governments aiming to facilitate the incentives to invest in the energy storage sector should therefore carefully consider the economic and regulatory context of their respective countries, while keeping in mind that an imperfect storage market is better than none at all.

The Russian Context

The last part of the event was devoted to the green transition and the energy storage issue in Eastern Europe, with a specific focus on Russia.

Alexey Khokhlov, Head of the Electric Power Sector at the Energy Center of Moscow School of Management, SKOLKOVO, gave context to Russia’s energy storage issues and prospects. While making up for 3% of global GDP, Russia stands for 10% of the worldwide energy production, which arguably makes it one of the major actors in the global power sector (Global and Russian Energy Outlook, 2016). The country has a unified power system (UPS) interconnected by seven regional facilities constituting 880 powerplants. The system is highly centralized and covers nearly the whole country except for more remote regions in the northeast of Russia, which rely on independent energy systems. The energy production of the UPS is strongly dominated by thermal (59.27%) followed by nuclear (20.60%), hydro (19.81%), wind (0.19%), and solar energy (0.13%). The corresponding ranking in capacity is similar to that of production, except the share of hydro-storage is almost twice as high as nuclear. The percentage of solar and wind of the total energy balance is insignificant

Despite the deterring factors mentioned above, Khokhlov described how the Russian energy sector is transitioning, though at a slow pace, from the traditional centralized carbon-based system towards renewables and distributed energy resources (DER). Specifically, the production of renewables has increased 12-fold over the last five years. The government is exploring the possibilities of expanding as well as integrating already existing (originally industrial) microgrids that generate, store, and load energy, independent from the main grid. These types of small-scaled facilities typically employ a mix of energy sources, although the ones currently installed in Russia are dominated by natural gas. A primary reason for utilizing such localized systems would be for Russia to improve the energy system efficiency. Conventional power systems require extra energy to transmit power across distances. Microgrids, along with other DER’s, do not only offer better opportunities to expand the production of renewables, but their ability to operate autonomously can also help mitigate the pressure on the main grid, reducing the risk for black-outs and raising the feasibility to meet large-scale electrification in the future.

Although decarbonization does not currently seem to be on the top of Russia’s priority list, their plans to decentralize the energy sector on top of the changes in global demand for fossil fuels opens up possibilities to establish a low-carbon energy sector with storage technologies. Russia is currently exploring different technological solutions to the latter. In particular, in 2021, Russia plans to unveil a state-of-the-art solid-mass gravity storage system in Novosibirisk. Other recently commissioned solutions include photovoltaic and hybrid powerplants with integrated energy storage.


There is no doubt that decarbonization of the global energy system, and the role of energy storage, are key in mitigating climate change. However, the webinar highlighted that the challenges of implementing and investing in storage are both vast and heterogenous. Adequate regulation and, potentially, further government involvement is needed to correctly shape incentives for the market participants and get the industry going.

On behalf of the Stockholm Institute of Transition Economics, we would like to thank Niclas Damsgaard, Natalia Fabra, and Alexey Khokhlov for participating in this year’s Energy TalkThe material presented at the webinar can be found here.

Disclaimer: Opinions expressed in policy briefs and other publications are those of the authors; they do not necessarily reflect those of the FREE Network and its research institutes.

Optimal Recommendation System with Competing Sellers

Person Holding Pineapple Fruit Near Red Wall Representing Optimal Recommendation System

Many e-commerce platforms that connect buyers and sellers employ recommendation systems to help customers find products and services. Such platforms seek to maximize their profits which mainly comes from a commission on sales made via the platform. This may create incentives for platforms to use a recommendation strategy that suppresses competition among sellers and keeps prices and the resulting commission high. At the same time, the huge success of platforms such as Amazon suggests that they also care about customer satisfaction. Thus, the platform has an incentive to recommend goods that are cheaper and a better match for customer’s tastes. This requires not only sufficient competition between sellers but also that sellers act to improve the fit of their product to customer needs. Since these actions are typically costly, a high commission may disincentivize sellers to undertake them, thereby negatively affecting customers. Therefore, in designing the recommendation system and deciding on commissions, the platform should carefully balance the pro-competitive customer care and anti-competitive incentives to keep high prices and profits.


When we search for a product on an e-commerce platform, such as Amazon or AliExpress, the default search outcome often contains a list of recommended products sold by vendors that are selected by the platform. The order of these sellers is, of course, not random – the platform’s decision on which sellers to recommend is strategic and there could be different forces driving such a strategy. For example, since the platform charges commission on sales, it may have an incentive to recommend the most expensive seller among those who sell similar products. At the same time, such a recommendation strategy, and high(er) prices in general, may negatively affect customer satisfaction from the marketplace and lead to a loss of its customer base. This is not in the best interest of the platform, especially if it wants to achieve long-term sustainability and growth.

The behavior of sellers adds a further layer to these considerations. Indeed, sellers are likely to adjust their pricing behavior and competitive strategies in response to a platform recommendation system.

These considerations give rise to two questions: First, how should an e-commerce platform design its recommendation system, or in other words, how does it optimally choose which sellers to recommend, which commission rate to set, etc.? Second, how does the presence of this system affect the competition and prices?

Further, a seller’s strategy may depend not only on the presence of recommendations but also on the commission rate set by the platform. Sellers usually have an option to perform costly actions in order to improve the match of their product to customers’ needs. For example, sellers may disclose more information on the characteristics of a good they are selling: spend time and money on detailed descriptions of their goods, or provide high-resolution photos. Though these actions are usually left at sellers’ discretion, they may substantially increase a customer’s satisfaction by improving the match between the purchased product and customer’s preferences.

In turn, a better fit may create a more loyal customer base for the seller, giving her more market power and increased profits. However, if the platform sets a high commission rate, sellers will have less incentive to undertake such costly actions (since the platform eats up a large share of the return to this action). This raises the questions – what is the optimal commission rate chosen by the platform, and how does the optimal commission rate affect sellers’ incentives to disclose information about their goods?

Another issue that arises here concerns the optimal precision of the recommendation system, that is, its ability to pin down customers’ tastes correctly. When the e-commerce platform deals with heterogeneous buyers, it should assess buyer’s preferences prior to making a recommendation. Although almost all research in Computer Science regarding recommendation systems focuses on how to make the precision as high as possible, I show that the highest level of precision may not be optimal from the platform’s perspective. Intuitively, this is because highly precise recommendation systems differentiate customers effectively, which in turn could give sellers local monopoly power and translate into higher prices. At the same time, an inaccurate recommendation system cannot distinguish customers with different preferences and views, which intensifies the competition by allowing sellers to compete for all customers.

In Fedchenko (2020), I address the abovementioned and other related issues on recommendation systems of e-commerce platforms. This brief summarizes the main findings of the study.

Model Description and Findings

In my model, I consider a platform that is designing a recommendation system. That is, for each seller, the platform chooses what share of customers end up receiving a recommendation to buy from this seller. This choice depends on the seller’s price, the quality of the good (if disclosed by the seller), and the buyers’ tastes. The platform also sets the commission rate it charges the sellers. I focus only on direct recommendations (i.e., the platform gives each buyer a unique recommendation). Although, in reality, platforms usually provide users with a ranking of alternatives, I assume that buyers always choose the top-ranked alternative which is equivalent to a single recommendation.

The model also assumes that a platform seeks to maximize the weighted sum of its profit (driven by commissions) and aggregate consumer surplus (motivated by the platform’s willingness to build a steady customer base). The (exogenous) weight assigned to the aggregate consumer surplus is referred to as the platform’s degree of consumer orientation (DCO). DCO is a measure of how much the platform cares about customer satisfaction and it plays an important role in determining the platform’s optimal recommendation strategy. In turn, customers have higher satisfaction if they buy a good that better fits their tastes, has higher quality, and is sold at a lower price.

Recommendation System Affects Competition

My model demonstrates that the presence of a recommendation system that charges sellers commission on sales (i.e. makes the platform have a stake in sellers’ profits) “softens” competition, and, in turn, increases prices. This effect is stronger the more a platform cares about its profits relative to customer satisfaction. The force that drives this result has already been touched upon in the introduction: if the platform has a stake in sellers’ profits, it will occasionally recommend sellers with higher prices. However, since the platform also cares about consumer surplus (which decreases if the price goes up) these high-priced recommendations will not go to all buyers, and therefore, the overall price level will not become too high. Still, the sellers are encouraged to set higher prices in this scenario, as compared to the hypothetical case in which customers know about the sellers without the platform.

Optimal Commission vs. Information Disclosure

The relationship between the commission rate and the seller’s decision on how much information to disclose is nontrivially affected by the DCO. If the DCO is high, then a higher commission rate causes sellers to disclose less information about their goods in equilibrium. If the DCO is low, the relationship is reversed: a higher commission rate increases the amount of disclosed information. This result stems from the interplay between two counteracting forces. On one hand, an increase in the commission rate decreases a seller’s return to providing disclosure, and hence, discourages sellers from making the effort to disclose. On the other hand, a higher commission rate increases the platform’s stake in the sellers’ profits and, as a result, softens competition, increases sellers’ prices and profits, and thus makes it more worthwhile for sellers to provide disclosure of their goods.

An interesting implication of this result is that for a high DCO, the optimal commission rate for a platform should be as small as possible (just enough for the platform to cover the operational cost).

Optimal Precision

Next, I show that a lower precision (i.e., ability of the recommendation system to pin down buyers’ tastes) weakens the effect of the presence of a recommendation system on competition. This happens since more imprecise recommendations effectively increase the share of “undecisive” customers and, thereby, the appeal to capture that market share. As a result, the competition for those customers will intensify.

Imprecision also affects the amount of product information sellers choose to disclose in equilibrium. However, the direction of this effect depends on the cost of disclosure: if the cost is low, a more precise recommendation system may increase the amount of disclosed information, while the result is reversed if the cost is high. The reason for that is as follows: The platform has two sources of information to infer whether a particular seller fits a certain buyer – the buyer’s preferences and the seller’s information on the quality of the product (if disclosed). If the buyer’s taste is measured imprecisely, while the seller’s information is more precise, it is optimal for the platform to focus on the latter when designing a recommendation system. This, in turn, would motivate sellers to disclose more information about their products.  In the case of low disclosure costs, this positive effect on disclosure more than offsets the direct negative effect of imprecision brought about by harsher competition and lower profits. In the case of high costs, the direct effect dominates.

I also show that some imprecision, in fact, can be optimal for the platform. Perfect precision softens the competition and results in increased prices for consumers. This negative effect on consumer satisfaction outweighs the benefits of a perfect match between seller and buyer. So, consumers prefer a certain degree of imprecision over perfect precision, which in turn, makes the platform unwilling to implement perfect precision. In other words, it is optimal to “sacrifice” some customers (i.e., not recommending them the best fitting alternative) in order to intensify the competition among sellers and, eventually, benefit all customers through lower prices.


The presence of a recommendation system on an e-commerce platform that charges sellers commissions on sales may cause softer competition and lead to higher prices and profits of sellers, as well as increased earnings for the platform. At the same time, it can sometimes be optimal for a platform to set a low commission rate since it would guarantee that sellers disclose more information about their goods which would improve the match between customers’ tastes and the goods they buy. If customer satisfaction is important for a platform, the indirect positive effect on customer satisfaction of a low commission rate, via sellers’ decisions, may outweigh the direct negative effect on the platform’s and sellers’ profits. Similarly, a recommendation system with some degree of imprecision can be beneficial for customers since it does not allow sellers to get local monopoly power. So, increasing the precision in the measurement of customers’ tastes – which seems to be the focus of many ongoing computer science studies devoted to recommendation systems, – may not actually be in the best interest of a platform.

In the modern era of digitalization, the use of e-commerce platforms is on the rise. Moreover, the ongoing COVID-19 pandemic has increased the use of such platforms even further. Understanding the implications of the strategies used by these platforms, such as recommendation systems, on prices, competition, and societal welfare is, thus, a necessary component for developing efficient regulation principles.


Disclaimer: Opinions expressed in policy briefs and other publications are those of the authors; they do not necessarily reflect those of the FREE Network and its research institutes.

Trade Induced Technological Change: Did Chinese Competition Increase Innovation in Europe?

20190428 Trade Induced Technological Change Image 01

The last 30 years has witnessed a shift of the world’s manufacturing core from Europe and North America to China. A key question is what impact this has had on manufacturing workers in other developed economies, and also on innovation, patenting, IT adoption, and productivity growth. While a rigorous data analysis on these variables for developing economies, particularly in Eastern Europe, is not yet available, this brief examines the impact of the rise of China on innovation in Western Europe, and also reviews the evidence on the impact of the rise of China generally. Recent research by Bloom, Draca, and Van Reenen (2016) found that Chinese competition induced a rise in patenting, IT adoption, and TFP by 30% of the total increase in Europe in the early 2000s. Yet, we find numerous problems with the Bloom et al. analysis, and, overall, we do not find convincing evidence that Chinese competition increased innovation in Europe.

Few events have inspired the ire of economists as much as Brexit and the rise of Donald Trump, two events seen as related as both were a seeming reaction to both globalization and slowing economic growth, particularly as some (such as Trump himself) saw the former as a key cause of the latter. Both Brexit and the trade war spawned by Trump do seem to have had negative economic effects – US equities have suffered every time the trade war has escalated, while anecdotal reports and more sophisticated economic analyses seem to suggest that Brexit has cost the UK jobs.

And yet, there is a need for policy makers and economists to hold two ideas in our heads simultaneously: Trump’s trade war and Brexit may be policy disasters, and yet globalization can create both winners and losers, even if it is clear that, generally speaking, the overall gains are likely positive and large. This is likely also true of the rise of China – one of the most dramatic events in international economics in the past 50 years. Figure 1 shows the increase in trade with China from the early 1980s to 2017, a period in which US imports from China grew from 7 to 476 billion dollars.

Figure 1. Chinese Imports (in logs, deflated)

Source: World Bank WITS

The academic literature tends to show that this impact, the rise of China, may have cost the US as much as 2.2 million jobs directly (Autor et al.), and as much as 3 million jobs once all input-output and local labor market effects are included. While approximate, these numbers are large enough for the China shock to have played a role in the initial onset of “secular stagnation” – the growth slowdown which began around 2000 for many advanced nations, including the US and Europe. In addition, Autor et al. (forthcoming) found that Chinese competition also resulted in a decline in patent growth. In the European context, however, other authors have found that although China did do some damage to certain sectors, overall, it does not appear to have been quite as damaging, particularly in Germany, which also benefitted from exporting increased machine tools to the Chinese manufacturing sector. And, in a seminal paper, Bloom, Draca, and Van Reenen (2016) find that Chinese competition actually led to an increase in patents, IT adoption, and productivity in Europe from 1996 to 2005, along accounting for nearly 30% of the increase. This is important, as it implies that without the rise of competition with China, the slowdown in European growth would have been even more pronounced than it was. It also implies that, far from being a source of stagnation, Chinese competition has been a source of strength. It also makes it more likely that the slowdown in growth since 2000 was caused by supply-side factors, such as new inventions becoming more difficult over time, as is perhaps the leading explanation among economists, notably Northwestern University business professor, Robert Gordon (2017), and also supported by others (see this VoxEU Ebook featuring a “who’s who?” among economists). It would also be evidence that contradicts the “Bernanke Hypothesis” that the former US Fed Chair first laid out in a 2005 speech at Jackson Hole, in which he suggested that international factors – particularly the savings glut and US trade deficit – were behind falling interest rates in the US. Since then, Ben Bernanke has followed up with a series of blog posts suggesting that these international factors were the cause of the initial onset of secular stagnation.

Figure 2. European Growth Relative to Trend

Source: World Bank WDI

In this brief, I present new research in which my coauthor and I test the robustness of the research finding that China had a positive impact on innovation in Europe (Campbell and Mau, 2019). We find that these findings are very sensitive to controls for time trends and other slight changes in specification. We also find that the number of patents matched to firms in the sample shrinks over the sample period (from 1996 to 2005). Overall, we conclude that, unfortunately, it is unlikely that the rise led to a significant increase in innovation in Europe, although more research is needed. Our research also sheds light on the so-called “replication crisis” currently gripping the social sciences, as researchers begin to realize that many published findings are not robust.

Trade-Induced Technical Change?

Bloom, Draca, and Van Reenen (2016) – hereafter BDV – tried to isolate the impact of the rise of China on Europe using several methods, using firm-level data for Europe. They placed each firm in a 4-digit sector, where they measured imports from China over time. First, they just looked at changes in patents, IT, and total factor productivity (TFP) at the firm level for sectors in which Chinese imports increased a lot vs. other sectors. But, because economists are always weary of the difficulty of isolating a causal relationship from non-experimental data, the authors, worrying that the sectors which saw increases in Chinese imports might differ systematically from the others, the authors also used what is called an instrumental variable. That is, they used the fact that when China joined the WTO in 2001, they also negotiated a reduction in textile quotas. Thus, BDV reason that textile sectors which had tightly binding quotas prior to removal were likely to have had fast growth in Chinese imports after China’s accession to the WTO. Thus, they end up comparing textile sectors in which the quotas were binding to sectors in which they were not binding. We went back and compared the evolution of patents in these same groups (sectors with binding textile quotas vs. not binding) below in Figure 3.

Figure 3. Patent Growth in China-Competing Sectors (Quota Group) vs. Other Sectors

Notes: The vertical red lines are dates when textile quotas were removed. The blue line shows the evolution of patents in the sectors without binding quotas (non-competing sectors), and the red line is the evolution of patents in the China-competing sectors. The dotted lines are 2 standard deviation error bounds.

What is immediately obvious in Figure 3 is that patents are declining rapidly over the whole period in both groups. The overall level of patents was falling in both groups for the full period. There is a 95.8% decline in patenting for the China-competing group, vs. a 96.2% decline for firms in the non-competing (“No quota”) group. By 2005, average patents per firm are close to zero in both groups (.04 in the China-competing sectors vs. .11 in the others). However, in the “No quota” group, the initial level of patents – close to three per firm per year – was much larger than in the quota group. Since patents are falling rapidly in both groups but bounded by zero, the level of the fall in patents in the non-quota group is larger, but one can easily see that much of this decline happens before quotas are removed. If we control for simple time trends, the effect goes away. Also, given the tendency of patents to decline, we can also remove the correlation between Chinese competition and patent growth in some specifications by simply controlling for the lagged level of patents. The overall declining share of patents in the BDV data also raises questions about data selection issues, as patents granted in the BDV data in the later years were a smaller share of the total patents actually granted in reality.

BDV also look at the impact of the rise of China on IT adoption. However, here they proxied IT adoption by computers per worker, but they did not collect enough data to control for pre-trends properly in the data, so we cannot be sure whether this correlation is causal or not. (For what it is worth, on the data we do have, from 2000 to 2007, including trends in the data renders the apparent correlation between Chinese import growth and computers-per-worker insignificant.)

Lastly, BDV look at the impact of the rise of China on TFP growth. Here, unlike before, we find that their measure is robust across various estimation methodologies. However, when we look at changes in a commonly used alternative measure of productivity, value-added per worker, instead of TFP (as TFP needs to be calculated using strong assumptions about the functional form of technology), we find no impact (see Figure 4 below).

Figure 4. Value-Added per worker Growth: China-competing sectors vs. others

Figure 4 above compares the evolution of value-added per worker in the most China-competing sectors vs. the others. Trends look similar for firms in either group of sectors (China-competing or otherwise), and we do not find a correlation. We also do not find that Chinese competition led to an increase in profits, nor an increase in sales per worker (in fact, we found a significant decrease in most specifications).


All in all, we find that the BDV findings suggesting that the rise of China had a large impact on innovation in Europe is not robust. However, in most specifications, we also don’t find a negative impact as did Autor et al. (forthcoming) for the US. This might have to do with data quality, although it does seem to be closer to other work, such as Dauth et al. (2014), which suggests that the rise of China had a smaller impact in Germany than in the US.

We also felt it was a bit alarming that a simple plot of  the trends in patents for China-competing and not-competing sectors was enough to seriously question the conclusions of BDV, as their paper was published in the Review of Economic Studies, a top 5 journal in academic economics. If influential articles published in the most fancy journals can exhibit such mistakes, this underscores the extent which the profession of economics may suffer from many published “false-positive” results. The reasons why this could be the case are obvious: researchers are under pressure to find significant results, as top journals don’t often publish null results, and replication is exceedingly rare in a field in which one needs to make friends to publish. However, there are signs that replication is becoming more mainstream, and as it does, we can certainly hope that voters around the world will turn back to science.


  • Autor, D., D. Dorn, G. H. Hanson, G. Pisano, and P. Shu. Forthcoming. Foreign Competition and Domestic Innovation: Evidence from US Patents. Forthcoming: AEJ:Insights.
  • Bloom, N., M. Draca, and J. Van Reenen. 2016. “Trade Induced Technical Change? The Impact of Chinese Imports on Innovation, IT and Productivity.” The Review of Economic Studies 83 (1): 87–117.
  • Campbell, Douglas and Mau, Karsten. 2019.. Trade Induced Technological Change: Did Chinese Competition Increase Innovation in Europe?”, mimeo
  • Dauth, W., S. Findeisen, and J. Suedekum. 2014. “The Rise of the East and the Far East: German Labor Markets and Trade Integration.” Journal of the European Economic Association 12 (6): 1643–1675.
  • Gordon, R.J., 2017. The rise and fall of American growth: The US standard of living since the civil war (Vol. 70). Princeton University Press.

Disclaimer: Opinions expressed in policy briefs and other publications are those of the authors; they do not necessarily reflect those of the FREE Network and its research institutes.

Too High or Too Low? The Pros and Cons of Regulating the Reserve Price in Public Procurement in Russia

Too High or Too Low FREE Policy Brief Image

In theory, an optimally set reserve price leads to an optimal outcome in all standard auctions. In reality, however, it is difficult to identify the optimal reserve price. In public procurement auctions, a higher reserve price may lead to a higher competition for the contract, because more suppliers will find the contract profitable. Thereby a higher reserve price may lead to lower prices. But on the other hand, if competition in the market is already quite low or the risk of collusion is high, a higher reserve price will just lead to higher contract prices. The controlling bodies in Russia become suspicious when the reserve price in public procurement auctions is too low because they are afraid it is a sign of collusion between the procurer and the seller. Indeed it may be the case that the reserve price is set low to exclude other sellers from competing, thus acting against efficiency. Using data on public procurement of gasoline in 11 Russian regions in 2011-2013, we show that a higher reserve price did not lead to lower contract prices, and that low competition in the private market was a major obstacle to efficiency.

Why is the reserve price important?

The reserve price is widely discussed in the auction and procurement literature. Standard auction theory says that an optimally set public reserve price results in the optimal outcome in all standard forms of auctions with risk-neutral agents and independent private values (Myerson, 1981). But practice is far from pure theory. The procurer does not have all information to set the optimal price and this leads to losses in social welfare (Klemperer, 2004; Dimitri et al, 2006).

There are several concerns for a practitioner here. First, there is the question of whether the reserve price should be known to everybody in advance (Dimitri et al, 2006; Brisset et al, 2015; Eklof&Lunander, 2003). Second, the reserve price influences the entry decision and competition for the contract (Klemperer, 2004; Krishna, 2009; Wang, 2016). Generally, a higher reserve price may lead to higher competition for the contract, because more suppliers will find that contract profitable, which may in turn lead to lower prices. But on the other hand, if there is a high probability of collusion in the market, a higher reserve price will just lead to higher contract prices due to coordinated behavior of the potential sellers (Wang, 2016). The procurer could use lower reserve prices to decrease gains from collusion (Krishna, 2009), but in a corrupt environment a lower reserve price is treated as an instrument to restrict entry for the favor of preferred bidders in exchange for bribe (Guide to Combating Corruption & Fraud in Development Projects).

Hence, there are various arguments for and against setting the reserve price in public procurement auctions higher rather than lower. We are interested in showing which of them hold true in practice, or in other words, do higher reserve prices lead to lower contract prices in public procurement auctions?

In Russian public procurement, the reserve price in an auction is set by the procurer and is visible to everybody. Moreover, before April 2011, procurers were free to set the reserve price, and they could easily set it unreasonably high, and then share the surplus with a seller. Starting from April 2011, procurers are obliged to prove that the reserve price is set at a reasonable level. In the explanations to the Law from Ministries of Economy and Finance, there are recommendations to set the reserve price higher rather than lower. Regulators are much more afraid of corruption than a high price of the contract.

Using data on public procurement of gasoline in 11 Russian regions in 2011-2013, we show that a higher reserve price did not lead to lower contract prices, and that low competition in the private market was a major obstacle to efficiency.

How does public procurement of gasoline work in Russia?

To make it clear how auctions in Russia are held, we will now present some details on Russian public procurement of gasoline.

First, the public procurement law is the same for all Russian regions. Second, the detailed information on public procurements – including calls for bids, chosen procedure, auction protocols, and supporting technical documentation – is published online at a unified website. If the reserve price is below 500000 rubles, public buyers of gasoline may choose between sealed-bid “paper” auctions and open-bid electronic auctions. If the reserve price is above 500000 rubles, they should use open electronic auctions.

To set the reserve price, a procurer may ask a few firms to provide estimates of an expected price of the contract at which they would agree to sign the contract. Alternatively, procurers may search for price information on the Internet or in other open sources on prices of goods and services (some gasoline stations publish its prices online, e.g.). The reserve price may then be calculated on the basis of these prices.

The procedures start when a procurer publishes the call for bids, stating basic characteristics of the contract and the reserve price. In sealed-bid auctions the bidders send their price quotations and the supporting documents. The bids are opened simultaneously, and the lowest bid (or the earliest bid in case where two or more equal prices are announced) wins. Open-bid auctions are conducted in two stages. By the first deadline all perspective bidders should provide a statement of interest, including the supporting documents and in some cases monetary deposits. Procurer may assess the statements of interest and exclude the firms that do not meet the basic requirements at the bidding stage. At the second stage, the preselected bidders show up at the auction and make descending open bids. The lowest bid wins the contract.

Data, empirical strategy, and results

To figure out whether higher reserve prices lead to lower contract prices in public procurement auctions, we used data on public procurement procedures available at a unified official public procurement website. In particular, we collected data on all public purchases of gasoline with octane number 92 at the regional level in 2011-2013 in 11 regions of Russia (1559 observations).

Among the characteristics of the procurement procedures, the most important are the number of bidders, the type of the procurement procedure (sealed-bid or open), and the characteristics of the contract (the volume and duration). We also take into account the number of price quotations the procurer uses to calculate the reserve price and some other characteristics of the purchase: the number of procurers in centralized purchase, the number of purchases of gasoline this procurer made in 2011-2013, and whether the procurer requested some special conditions from the seller (e.g., that the seller should have a network of gasoline stations).

The information allowed identifying:

  • Procurements with only one bidder;
  • Procurements with no or a very small price decrease as a result of the auction;
  • Procurements with a reserve price higher than the market price;
  • Procurements with a reserve price higher than the maximum of the price quotations.

Using these new variables, we test whether the probability that there is only one bidder (which is not what a regulating body would wish to see) correlates with auction characteristics and the fact that the reserve price is higher than the maximum of the price quotations (implying it is unreasonably and probably inefficiently high).

Table 1 Regression results

Probabilty Probabilty
VARIABLES One bidder Discount = 0
one bidder 1.074***
open auction 0.749*** -0.191*
(0.0973) (0.115)
volume 8.62e-06*** -3.03e-06
(2.51e-06) (1.96e-06)
duration 0.000828* 0.00209***
(0.000454) (0.000610)
number of price quotations -0.126*** 0.386***
(0.0436) (0.0539)
reserve price is higher than max of quotations -0.469*** 0.560***
(0.108) (0.136)
Constant -0.264** -0.694***
(0.126) (0.160)
Observations 931 932

Standard errors in parentheses

*** p<0.01, ** p<0.05, * p<0.1

We also test whether the probability that there is no price decrease during the auction (discount equals zero) correlates with the auction characteristics and the fact that the reserve price was higher than the maximum of the price quotations or just higher than the market price.

Table 1 shows that a higher reserve price does not lead to a higher competition, but leads to higher probability of the situation that there will be no price decrease at all. Hence, there is no evidence that setting the reserve price at a higher level will attract more bidders and result in lower contract prices.


Auctions are viewed as one of the best ways to achieve lower prices. But in reality there are many factors that make this questionable. In this policy brief, we focus on the regulation of the reserve prices in public procurement. Is it reasonable to recommend procurers to set high reserve prices? We look at a specific market with high entry barriers, and a relatively low number of suppliers active on the public procurement market. Such markets face high collusion risk.  We show that high prices do not attract more bidders and auctions with reserve prices set higher than all quotations end up with no price decreases during the auction. A big share of auctions (46% in our data set) in Russia is inconsistent (only one bidder comes to bid), and in such an environment high reserve prices can only increase government spending. It is more reasonable to follow the ideas mentioned by Krishna (2009) and use low reserve prices to decrease contract prices and, thus, gains of suppliers from colluding behavior, even if it happens. Our study shows that general recommendations that do not take into account market specifics could not help procurers achieve efficient results.


  • Krishna, Vijay. Auction theory. Academic press, 2009, ch.11.
  • Klemperer, Paul. “Auctions: Theory and Practice.” Princeton University Press, 2004, ch.1,3,4.
  • Dimitri, Nicola, Gustavo Piga, and Giancarlo Spagnolo, eds. Handbook of procurement. Cambridge University Press, 2006, ch.11.
  • Myerson, Roger B. “Optimal auction design.” Mathematics of operations research 6.1 (1981): 58-73.
  • Brisset, Karine, François Cochard, and Julie Le Gallo. “Secret versus public reserve price in an “outcry” English procurement auction: Experimental results.” International Journal of Production Economics 169 (2015): 285-298.
  • Eklöf, Matias, and Anders Lunander. “Open outcry auctions with secret reserve prices: an empirical application to executive auctions of tenant owner’s apartments in Sweden.” Journal of Econometrics 114.2 (2003): 243-260.
  • Wang, Hong. “Information Acquisition Versus Information Manipulation in Multi-period Procurement Markets.” Information Economics and Policy (2016).
  • Guide to Combating Corruption & Fraud in Development Projects, http://guide.iacrc.org/

Higher Competition in the Domestic Market – A Way to Boost Aggregate Productivity

20170227 Higher Competition in the Domestic Market - FREE Policy Brief

Competition is a good thing not only because of lower prices and larger variety. Higher competition in the domestic market also shifts necessary labour and capital resources from less productive domestic-oriented firms to export-oriented productivity champions. Such firms will make better use of production factors and generate larger output. Thus, simply increasing the level of competition in the domestic market can boost the aggregate productivity of a country.

The aggregate productivity of a country can be boosted even without changing the productivity of individual enterprises. This can be achieved by improving the allocation of resources – the redistribution of labour and capital towards more productive firms. These firms will make better use of production factors and generate larger output. But how can one affect the allocation of resources? Economic theory says that allocation depends on the productivity of individual firms: more productive enterprises attract more labour and capital. However, there exists another factor behind allocation: distortions.

Distortions affect the allocation of resources

A model developed by Hsieh and Klenow (2009) – one of the most popular frameworks to study the allocation of resources – has a very important and realistic feature: it acknowledges that firms are not treated equally. Some firms may face lower supply of banking loans ending with higher capital costs. Other firms could confront with trade unions and higher wages. Tax rates may also differ across firms. These are all examples of distortions. Firms facing larger distortions are forced to underuse respective production factor, while firms that enjoy more favourable conditions tend to overuse capital and labour, generating more output.

While it is virtually impossible to imagine an economy without any distortions (the one where all firms face the same taxes, costs of labour, capital etc.), not all distortions damage the allocation of resources. Only distortions to productive firms create misallocation of resources by shifting labour and capital towards unproductive firms. Thus, removal of such distortions can improve the efficiency of allocation and raise the aggregate output of the country.

According to Hsieh and Klenow (2009) the distortions faced by every individual firm can be quantified from the balance sheets and profit/loss data. For example, observing lower-than-usual ratio of capital to intermediate inputs (comparing with other enterprises in a narrowly defined industry) indicates a capital distortion, possibly related with limited access to banking loans. Similarly, lower-than-usual share of wages in total production costs implies high labour distortions. Finally, the size of the distortion can be detected as a case of abnormally low share of intermediate inputs in total output, and signals about the restrictions to total output (e.g. due to higher taxes for large enterprises).

Misallocation of resources is small in Latvia

In my recent research (see Benkovskis, 2015), I use anonymised firm-level dataset for 2007–2013 and apply the Hsieh and Klenow (2009) model to study the allocation of resources in Latvia – a unique example of a small and open economy facing extreme structural shifts during the financial crisis. According to my estimates, the negative contribution of misallocation to aggregate productivity was close to 27% in 2013 (see Figure 1). In other words, it suggests that actual aggregate productivity could be boosted by 27% if all distortions were removed!

This may seem large but in fact 27% is a comparatively low figure. Hsieh and Klenow (2009) argue that full liberalisation would boost aggregate manufacturing productivity by 86–115% in China, 100–128% in India, and 30–43% in the US. Dias et al. (2015) show that removing distortions would lead to a 30% gain in output of Portugal in 2011. Thus, misallocation of resources is relatively small in Latvia. Even more important: the misallocation of resources decreased after the crisis in Latvia (contrary to the case of Portugal), adding more than 10 percentage points to aggregate productivity growth between 2010 and 2013.

Figure 1. Contribution from misallocation of resources to aggregate total factor productivity, %

Source: Benkovskis (2015). Note: shows the contribution of misallocation comparing with the counterfactual case of no distortions.

The finding that allocation of resources improved after the crisis is interesting per se, but uncovering the reasons behind the improvement is even more important. Figure 1 provides a decomposition, which shows that labour distortions are minor in Latvia due to high flexibility of labour market (in line with recent findings by Braukša and Fadejeva, 2016). The capital distortions, while being minor in 2007–2008, increased afterwards, pointing to some credit supply constraints faced by the highly productive enterprises after the financial crisis. However, by far largest contribution comes from the misallocation of intermediate inputs – the turnover of the most productive firms face some constraints. And it was the ease of constraints to turnover for the most productive firms that determined the improvements in aggregate productivity since 2010.

The level of competition matters for misallocation

My research stresses the importance of the competition level on the market, since higher competition serves as a natural constraint for the firm to increase its turnover. What if the most productive Latvia’s firms systematically come up against higher competition? I found that indeed this is the case. First, recent results by Fadejeva and Krasnopjorovs (2015) show that Latvia’s domestic market has lower competition level comparing with external markets. Second, it is widely acknowledged that exporters tend to be more productive comparing with domestically oriented firms (see e.g. Bertou et al., 2015, who report positive export premiums for EU countries, while Benkovskis and Tkačevs, 2015, find higher productivity of exporters in Latvia). Thus, Latvia’s productive export-oriented firms are subject to higher competition and cannot enlarge their turnover as easy as other entities. This shifts labour and capital towards small and less productive firms working solely on domestic market, creating the misallocation of resources.

The domestic competition factor can also explain the improving allocation of resources after 2010. The study by Fadejeva and Krasnopojorovs (2015) reveals that the competition gap between domestic and foreign markets narrowed after the financial crisis (see Table 1). Namely, life was too easy on the local Latvia’s market during the boom time, allowing unproductive firms to survive and drain away resources from more productive firms. But conditions became tougher after the crisis (although the competition level still remained lower than abroad). We can view this as a “cleansing effect of the crisis”: some of the least productive domestic oriented firms went bankrupt (or decreased their turnover), freeing the necessary capital and labour resources for productive exporters.

Table 1: Change in the competitive pressure on main product in domestic and foreign markets compared to the situation before 2008, %

Domestic market Foreign market
2008–2009 2010–2013 2008–2009 2010–2013
Strong decrease 2.9 2.2 0.9 1.0
Moderate decrease 11.8 3.8 7.6 5.9
Unchanged 33.8 24.7 45.7 51.5
Moderate increase 30.0 28.1 25.2 19.7
Strong increase 18.7 38.5 11.2 8.8
Does not apply 2.8 2.8 9.4 13.1

Source: Fadejeva and Krasnopjorovs (2015), Table A.102. Notes: based on the sample of 557 Latvia’s firms; results are weighted to represent firm population.


This research has an important policy conclusions applicable to any country that seeks to increase aggregate productivity. The competition level in the domestic market is important not only for consumers, who enjoy lower prices and higher variety. Higher competition in the domestic market also shifts necessary resources from less productive domestic-oriented firms to export-oriented productivity champions.




Does Product Market Competition Cause Capital Constraints?

FREE Policy Brief Image

At the very center of Schumpeter’s (1934, 1942) notion of creative destruction is firms’ access to bank capital, which helps to fund the innovation in competitive product markets that drives out less productive firms in favor of those with more profitable ideas. However, competition is a two-edged sword and may result in firms being unable to fund all of their otherwise economically profitable investments. Using unique survey data from 58 countries, Bergbrant, Hunter, and Kelly (2016) find that product market competition increases capital constraints and has a greater effect than banking sector competition. Further, we show that quantity-of-capital constraints negatively impact firm growth.

Capital and creative destruction

At the very center of Schumpeter’s (1934, 1942) notion of creative destruction is firms’ access to bank capital, which helps to fund the innovation in competitive product markets that drives out less productive firms in favor of those with more profitable ideas. While product market competition may be the fundamental driver of the innovation envisioned by Schumpeter, it may also impede access to the very source of capital that is supposed to fund that innovation. More intense product market competition can affect firms’ ability to finance their projects either by increasing the price of financing or by inducing capital constraints, whereby firms are unable to obtain the quantity of capital needed to fund all their positive net present value projects.

Recent research has focused on the price side of financing, showing that product market competition increases the cost of equity (Hou and Robinson, 2006) and the cost of debt (Valta, 2012). In this brief we examine the quantity side of financing; that is, whether product market competition increases capital constraints.

Isn’t it obvious that competition causes capital constraints?

Actually, no. There is a familiar argument that firms are reluctant to disclose commercially valuable information when competitors are more likely to exploit this information. Theory predicts that it is not optimal for creditors to respond to the resulting asymmetric information by raising interest rates; instead, restricting capital is more appropriate (Stiglitz and Weiss, 1981). However, competition may have the very opposite effect because a competitive environment lowers owners’ cost of monitoring and measuring managerial performance. Theory and recent empirical tests indicate that lower cost of monitoring managers induces greater disclosure by owners.

Whether or not product market competition makes banks restrict the supply of loans is arguably more important than whether it influences the cost of debt. Greenwald, Stiglitz, and Weiss (1984) show that firms’ investment behavior is not particularly sensitive to the interest rates they pay, consistent with the notion that increases in the cost of debt may reduce investment, but only at the margin; i.e., projects change from generating economic profits to generating economic losses (net present value changes from positive to negative). By contrast, increased capital constraints can lead to underinvestment by forcing firms to abandon projects which generate economic profits (net present values are positive), thus hindering investment and preventing firm innovation and growth (see Harford and Uysal, 2014).

What does the research tell us?

Recent research by Bergbrant, Hunter, and Kelly (2016) uses survey data obtained from the World Bank’s World Business Environment Survey, conducted among non-financial firms from around the world. Capital constraints are the response to a question about the extent of the obstacle to operations and growth posed by capital constraints that managers and owners rank from 1 (No Obstacle) to 4 (Major Obstacle). Competition is represented by an index constructed from eight individual forms of competition reported by firms.

The empirical evidence indicates that the intensity of product market competition significantly increases capital constraints. Table 1 shows the marginal effects of a change in the intensity of competition on capital constraints. For instance, the first row shows that a small (instantaneous rate of) increase in product market competition leads to an increase in the likelihood that capital constraints are a “major obstacle” (4 on a four- point scale) at a rate of 18.9%. Similar results hold when competition is assessed at a one-standard-deviation (3rd row) increase or when competition changes from 0 to 1 on a version of our competition index which ranges from 0 to 1 (5th row).

Table 1: Effect of competition on capital constraints

For a change of:


No obstacle


Minor obstacle


Mod. obstacle


Major obstacle


Marginal -0.147 -0.052 0.010 0.189
p-value (0.000) (0.000) (0.062) (0.000)
+SD -0.042 -0.017 0.000 0.059
p-value (0.000) (0.000) (0.925) (0.000)
0 to 1 -0.145 -0.059 0.008 0.196
p-value (0.000) (0.000) (0.165) (0.000)

Note: The table reports the marginal effects “for a change of” product market competition of varying amounts on firms responding that capital constraints pose one of the four levels of “obstacle” for their operations.

The above results are qualitatively similar when the competition index is replaced by any one of its eight individual components. In addition, competition increases not only a measure of general capital constraints, as employed in the above analysis, but also specific forms of capital constraints. These include the credit constraints that firms experience when, as a precondition for lending, banks require that borrowers have special connections in the banking sector, pledge collateral, satisfy banks’ bureaucratic need for business documents, and pay bribes to corrupt bank officials. Further, the evidence is not unique to domestic bank capital as more intense product market competition also impedes firms’ access to nonbank equity, foreign bank capital, special export financing, and lease financing.

To further validate our main result we account for two well-established strands of research that contend that banking sector competitiveness is among the most important determinants of access to credit and that banking sector structure can also affect the competiveness of non-financial firms’ industries. The evidence reported in Table 2 shows that while (one of three measures of) banking sector competition and the degree of bank freedom affect capital constraints, in general the regulatory structure of the banking sector does not. More important, our main finding is unchanged when controlling for banking sector structure. Finally, it is important to note that in all our models we control for any cost-of-debt (higher-interest-rate) effects.

Table 2: Accounting for banking sector structure

Competition (10 separate models) +ve signif.
Lerner bank competition index +ve, signif.
Bank concentration ratio insignif.
Boone indicator of banking sector insignif.
private credit as a fraction of GDP insignif.
restrictions on nonbank activities insignif.
fraction of bank applications denied insignif.
bank freedom from gov’t interference -ve, signif.
existence of a credit registry insignif.
foreign bank share of banking system insignif.
government share of banking system insignif.

Note: We augment our main model with the above banking sector variables, one at a time, to determine their impact on the significance (signif. or insignif.) of product market competition.

Capital constraints hurt firms’ growth and so we expect our measure of capital constraints to be negatively associated with growth. We confirm this in the data, after controlling for the direct impact of competition on growth. We also find that the quantity-of-capital effect has a greater impact on expected firm growth than the cost-of-capital effect.


Our research indicates that the intensity of product market competition increases capital constraints even in the presence of controls for banking sector competition. Our work suggests several policy recommendations. First, the implementation of a product-market competition policy, for instance by several Central and Eastern European countries in the 1990s (Fingleton et al., 1996; Dutz and Vagliasindi, 2000), should contemplate the possibility that such action is likely to have negative externalities for firms’ access to capital. Second, banking sector reforms aimed at creating a more competitive banking system in order to improve access to capital should not be pursued in isolation and should take into consideration the existing competitiveness of the product market. Third, given that the quantity-of-capital effect has a greater impact on firm growth than the cost-of-capital effect, policymakers should exert at least as much effort in easing quantity constraints as they do to reduce the cost of capital.


  • Bergbrant, M.; D. Hunter; and P. Kelly, 2016. “Product Market Competition, Capital Constraints and Firm Growth”. Available at SSRN: https://ssrn.com/abstract=2594218.
  • Dutz, M. A.; and M. Vagliasindi, 2000. “Competition policy implementation in transition economies: An empirical assessment”. European Economic Review 44, 762-772. Fingleton, J.; E. Fox; D. Neven; and P. Seabright, 1996. “Competition policy and the transformation of central and eastern Europe”. Working paper. CEPR, London.
  • Greenwald, B.; J.E. Stiglitz; and A. Weiss, 1984. “Informational imperfections in the capital market and macroeconomic fluctuations”. American Economic Review 74(2), 194-199.
  • Harford, J.; and V. B. Uysal, 2014. “Bond market access and investment”. Journal of Financial Economics 112, 147-163.
  • Hou, K.; and D. Robinson, 2006. “Industry concentration and average stock returns”. Journal of Finance 61, 1927-1956.
  • Schumpeter, J.A., 1934. “The Theory of Economic Development”. Harvard University Press, Cambridge, MA.
  • Schumpeter, J. A., 1942. “Capitalism, Socialism and Democracy”. Harper and Brothers, New York, NY.
  • Stiglitz, J.; and A. Weiss, 1981. “Credit rationing in markets with imperfect information”. Amer. Econ. Review 71, 393-410.Valta, P., 2012. “Competition and the cost of debt”. Journal of Financial Economics, 105(3), 661-682.

The Issue of Repeat Cartel Offences

Image for FREE Policy Brief "The Issue of Repeat Cartel Offences"

Leniency policies have become an important antitrust tool but it is not clear whether they have effectively prevented recidivism or whether firms have learned to collude under, and even make strategic use of them. If “recidivism” is really an industry-level phenomenon, the appropriate policy measures are very different from what is necessary if individual firms, having been detected and punished for colluding, engage in the behavior again. Following Levenstein et al. (2015), this brief discusses the recidivism question as one about post-cartel behavior, i.e. the set of policies required to assure that effective competition emerges post-cartel breakup.

Measuring Recidivism

Cartels are one of the main concerns of the European Commission (EC) and the US Department of Justice (DOJ) and so, the US and EU Leniency Programmes (LPs) were designed, in 1978 and 1996 respectively, as a device for the deterrence and dissolution of collusive agreements (see Marvão and Spagnolo (2015a) for an in-depth review on the available evidence of the effects of LPs).

In the analysis of cartel formation, recidivism is an important issue. In the set of 510 cartel members fined by the EC in 1998-2014, Marvão (2015) identifies 89 “multiple offenders” (firms fined for collusion more than once), 10 “repeat offenders” (firms which initiate a cartel after being investigated for another cartel), and 5 recidivists following the definition from Werden et al. (2011): firms which initiate a cartel after being fined for another cartel.

The DOJ dataset compiled by Levenstein and Suslow (2015), spanning 1961-2013, preliminarily finds 113 “multiple offenders” but only 14 “repeat offenders”. Of these 14 firms, 5 that had been previously indicted were caught in the 1990s, but none was indicted again by the DOJ in the 2000s.

Although the number of (discovered) “true recidivists” is not zero, it is less than 1% in these two samples (EU, US). Recidivism seems to arise when there are lapses in enforcement; not surprisingly, some firms take advantage of these lapses to return to old behaviors. Designing policies that are able to prevent recidivism requires understanding whether this is an industry or firm-level phenomenon.

Industry Recidivism

Levenstein et al. (2015) use the above-mentioned EU and US datasets to show that collusion occurs in virtually all sectors of the economy, but with discernable patterns.

In the US, construction and chemicals are frequently cartelized (pre and post leniency). There are a large number of cartels in local markets in some industries, such as retail gasoline stations and dealers and ready-mix concrete. While collusion in these local markets is frequently uncovered, it is not necessarily amongst the same firms.

In the EU, chemicals and transport cartels are also frequent areas of collusive activity (although cartels that are strictly within national boundaries and prosecuted by national competition authorities are not included in the sample).

The authors show that there is a large share of repeat and multiple offenders in chemicals and a surprisingly high proportion of repeat offenders in the manufacture of transport and electrical equipment. The highest proportion of multiple offenders is found in pharmaceuticals and refined petroleum products. The transportation and storage market is a sector with a high incidence of collusion (83 convicted cartel members), but no repeat offenders.

While the determinants of cartel activity are varied and endogenous, some correlations with industry-driven recidivism can be discussed:

  1. Industry concentration. It increases the ease of tacit collusion and it should increase the likelihood of explicit collusion, but there are many cartel examples in unconcentrated industries. In some industries, it has been argued that high fixed costs make competition unstable, so that, absent collusion, firms price below long-run marginal cost and are unable to cover fixed costs (Pirrong, 1992).
  2. Culture and history. Spar (1994) argues that the cooperative culture necessary for survival for diamond miners facilitated collusion as the industry matured. Policy fluctuations can also contribute to this problem, as was the case in the US during the Great Depression.
  3. Inelastic demand. This is empirically challenging to capture if the observed prices have been affected by monopoly power, thus potentially raised to a level at which demand is elastic. In many cases, the direct consumer is a producer, so the downstream cost function and competitive intensity also influence elasticity of demand for the cartelized product. Grout and Sonderegger (2005) estimate the likelihood of collusion in the US and EU and rank industries accordingly. This could be used to target competition authority resources to select industries.

Firm Recidivism

Once a cartel breaks-up, cartel members may decide to compete in the market, merge, tacitly collude, or explicitly collude again. The latter does not mean that the cartel re-forms: a firm may collude in a new industry or product line or with a new set of co-conspirators.

U.S. Steel was involved in 6 different US cartels between 1948 and 1969, with different cartel partners and in different steel products. VSL construction was similarly involved (including as a leader) in multiple US cartels across several decades with distinct, but overlapping partners.

In the EU, Akzo Nobel N.V. has been convicted for 9 cartels, which lasted between 1987 and 2007, and in which its co-conspirators were mostly overlapping – e.g. collusion with Arkema in 6 instances (although the latter changed its name during the period). Many of the other co-conspirators were also multiple offenders. While Akzo only received one fine increase for recidivism, it received 7 leniency reductions, of which 3 were full immunity.

Other EC repeat offenders are ABB and Degussa Evonik – both convicted 4 times and received full immunity twice – as well as Brugg and Sumitomo. The latter was convicted for 7 cartels, of which 5, in the automotive wire harness, were self-reported.

What may influence repeated cartel participation, at the firm level?

  1. Firm’s corporate culture. In such a case, the leadership of the organization expects managers to collude, and collusion occurs in many markets in which the firm operates. Firm norms and expectations of managerial behavior can repeatedly encourage collusion and “disregard” previous fines, as illustrated in the ADM case (Eichenwald, 2000).
  2. Firm structure. Multi-market collusion literature focuses on the ability of firms to target punishments in particular markets. Multi-market firms may also encourage the spread of collusion if they have learned to collude in one market and share their “best practices” in another. This seems to have been the case, for example, in the spread of the vitamin cartel from vitamins A and E to other vitamins (Connor, 2008). Multi-market collusion is encouraged not only by multi-product multinationals, but also multi-market relationships between what appear to be smaller firms in local markets. For example, if gas stations are owned by multi-market firms such as large oil firms or chains of stations, that may facilitate repeated collusion over time and/or across geographic locations.

Policy Tools

In complementarity with LPs, Levenstein et al. (2015) discuss additional (possibly) effective post-cartel policies, aimed at preventing firm-driven recidivism.

  1. Company Fines and Leniency. Theoretical research has emphasized the aptitude of well-designed and well-run LPs to improve cartel detection and deterrence (for a survey, see Spagnolo, 2008). However, Marvão and Spagnolo (2015b) note the generosity of the current EU LP: the average LP reduction is 45% and leniency is granted to 52% of convicted cartel members. In addition, Marvão (2015) shows that repeat offenders appear to receive larger EC leniency reductions, which suggests that firms can learn the “rules of the game”, colluding repeatedly and reporting the cartel to reduce their penalties. As such, fines need to be tougher and recidivism needs to be dealt with differently.
  2. Individual Accountability. Senior management in EU cartels does not seem to suffer from their participation in cartels. For example, Robert Koehler became CEO of SGL Carbon in 2012, after being convicted in 1999 of price-fixing in the graphite electrodes cartel. Imposing tougher sanctions, such as individual prison sentences or disqualification of senior executives from employment in their sector or role, may prevent repeated collusive behaviors (in new firms) and thus, increase deterrence levels.
  3. Follow-On Damages. Private damage suits may increase deterrence. In the US, private litigation plays a major role in the enforcement of antitrust law. Conversely, access to private damages is relatively new in the EU. A recently adopted EU Directive on damages (11/2014) prevents the use of LP statements in subsequent damage actions. However, Buccirossi et al. (2015) show that the effectiveness of damage actions can be improved if the civil liability of the immunity recipient is minimized and claimants receive full access to all evidence collected by the competition authority. Access to previous cartel decisions, for a given firm, will increase the amount of available information and can increase the likelihood and/or amount of successful damage claims.
  4. Consent Decrees. These impose conditions on the behavior of convicted firms (e.g. maximum price, and transparency). If these are violated, the authorities intervene, thus lowering the cost of prosecuting recidivists. In the US, decrees were routinely used by the DOJ in the 1960s and 1970s, but the practice was abandoned due to concerns of effectiveness and large costs. More recently, in September 2007, the Brazilian Administrative Council for Economic Defense enacted a resolution that allows for the use of consent decrees with the aim to settle cartel investigations. Two have already been executed.

If recidivism is industry-driven, its prevention may require a different set of tools, including those below, to complement leniency.

  1. Structural Remedies. Competition authorities have repeatedly permitted mergers among former cartel members, often without review, let alone structural intervention. Davies et al. (2014) examine mergers among former cartel conspirators and conclude that only 29% of the mergers were investigated by the EC. Remedies such as disclosure, divestiture of assets, selling minority shares in competitors, or licensure of intellectual property to competitors may change the nature of competition in the market and make collusion more difficult (see Marx & Zhou, 2015 regarding post-cartel mergers). This is particularly relevant if recidivism is industry-driven.
  2. Monitoring and screening. Some antitrust authorities have implemented monitoring and screening techniques to identify anticompetitive behavior in a given industry. These initiatives involve the analysis or monitoring of the characteristics of products or market structures that are thought to be more prone to collusion (mostly due to repeated offenses). Some examples are watch lists (e.g. Australia, UK, Chile), price observatories (e.g. Belgium, Spain, France), statistical screens (e.g. US FTC, Korea FTC), gasoline retail in Brazil and public procurement in Sweden (see Abrantes-Metz (2013) for further details on screens).


While literal recidivism, i.e. the formation of a cartel after having been convicted of illegal collusion, appears to be rarely detected in the EU and US, there remain policy gaps closing which could improve competition post-cartel.

A variety of post-cartel policies should be explored for their ability to increase the likelihood that workable competition, rather than tacit collusion or single firm dominance, will emerge. These reduce the reliance of competition authorities on leniency-driven self-reports, which will in turn make leniency more effective and less amenable to strategic use by firms determined to collude.



  • Abrantes-Metz, Rosa (2013). “Proactive vs Reactive Anti-Cartel Policy: The Role of Empirical Screens.” Available at SSRN: http://ssrn.com/abstract=2284740.
  • Buccirossi, Paulo, Catarina Marvão, and Giancarlo Spagnolo (2015). “Leniency and Damages,” CEPR Working Paper DP 10682.
  • Connor, John M. (2008). Global Price Fixing, 2nd ed. Berlin: Springer.
  • Eichenwald, Kurt (2000). The Informant. New York: Random House.
  • Grout, Paul and Silvia Sonderegger (2005) “Predicting Cartels,” Office of Fair Trading, Economic Discussion Paper.
  • Levenstein, M., Marvão, C., Suslow, V., 2015. Serial Collusion in Context: Repeat Offenses by Firm or by Industry? OECD Global Forum on Competition. DAF/COMP/GF(10/2015)
  • Levenstein, Margaret C., and Valerie Y. Suslow (2015). “Price Fixing Hits Home: An Empirical Study of U.S. Price-Fixing Conspiracies,” working paper.
  • Marvão, C., 2015. The EU Leniency Programme and Recidivism. Review of Industrial Organization, 48(1), 1-27
  • Marvão, Catarina and Giancarlo Spagnolo (2015a). “What do we know about the effectiveness of leniency policies? A survey of the empirical and experimental evidence,” in Beaton-Wells, C and C Tran (eds.), Anti-Cartel Enforcement in a Contemporary Age: The Leniency Religion, Hart Publishing.
  • Marvão, Catarina and Giancarlo Spagnolo (2015b). “Pros and Cons of Leniency, Damages and Screens”. Competition Law and Policy Debate (forthcoming)
  • Marx, Leslie M., and Jun Zhou (2015). “The Dynamics of Mergers among (Ex) Co-Conspirators in the Shadow of Cartel Enforcement,” working paper.
  • Pirrong, Stephen Craig (1992). “An application of core theory to the analysis of ocean shipping markets” Journal of Law and Economics, 35(1): 89-131.
  • Spar, Debora (1994). The Cooperative Edge: The Internal Politics of International Cartels, Ithaca: Cornell University Press.
  • van Driel, Hugo (2000). “Collusion in Transport: Group Effects in a Historical Perspective.” Journal of Economic Behavior and Organization, 41(4): 385–404.
  • Werden, Gregory, Scott Hammond, and Belinda Barnett (2011). “Recidivism Eliminated: Cartel Enforcement in the United States since 1999,” Georgetown Global Antitrust Enforcement Symposium, Washington DC, Sept. 22, 2011.

Decentralization Reform in Ukraine

Decentralization Reform in Ukraine Policy Brief Image

The current Ukrainian political system, which is a highly centralized “winner-take-all” system, is one of the main causes of the recent mass street protests. A decentralization reform is needed to make the system more stable by providing people with more impact on policy making, and increasing accountability of the government. A decentralization reform would reduce paternalistic expectations and provide people with more opportunities to take responsibility for public policy design in their region. In addition, it would improve the quality of national politics by introducing more competition and allowing successful regional politics to spread to the national level. However, as all reforms, decentralization bears some risks. This policy brief discusses the benefits and risks of such reform, suggests some ways of mitigation of the risks, and the procedure for reform development.

“In decentralized systems, problems can be solved early and when they are small. And when there are terrible failures in economic management—a bankrupt county, a state ill-prepared for its pension obligations—these do not necessarily bring the national economy to its knees.” / Nassim Taleb

In their path-breaking article Roger Myerson and Tymofiy Mylovanov argue that the underlying reason for the Ukrainian street protests in 2004 and 2014 is a fundamental flaw in the country’s Constitution, namely, the design of its government system. Currently, it is basically a “winner-take-all” system, where a winner of the national elections gains almost a dictator’s power, and then tries to prolong his stay in office with all means.

Such a system – where almost all the power is concentrated in the hands of the central government, and where local authorities, even the elected ones, have very little room for their own decisions – resembles an inverted pyramid and is therefore unstable. A natural way to stabilize the system is to put the pyramid on its foundation – i.e. to provide people with more impact on (and responsibility for!) both local and central government policy.

However, the Ukrainian government has announced a decentralization reform, and has already adopted a Decentralization Concept, which defines the main goals and milestones of the reform. According to the Concept, the legislative base for the decentralization should be developed by the end of 2014. However, it is clear that these plans are unrealistic. This, since on top of Constitutional changes, the reform implies changes to the administrative structure of the country, a redistribution of responsibilities between different levels of local government, and changes to the Tax Code, the Budget Code, and to several other documents. Such a scope of reforms is hardly attainable within the planned timeframe.

So far, the President’s office has developed changes to the Constitution, and the Cabinet of Ministers has drafted changes to the Budget Code. However, both documents miss the main point of the reform – empowering of people (rather than simply delegating some responsibilities from central to local governments). Instead, the drafted law on changes to the Constitution empowers the President, and the drafted changes to the Budget Code are an attempt of the central government to get rid of its “headaches” (e.g. ecological or social housing programs) while at the same time consolidating “electorally valuable” spheres, such as education and healthcare. This Draft Law proposes transferring some revenue sources from central to local levels, and at the same time to extract a part of the revenues that currently belong to local budgets to the central budget. A more detailed analysis of the proposed changes is provided in this article.

To my mind, the main impediment to the decentralization reform is a lack of a systemic approach. The Decentralization Concept does not provide a clear reform path, and changes to the legislation proposed so far look like pieces of a puzzle that do not fit together.

I suggest that the decentralization reform should be developed together with the administrative reform and proceed according to the following algorithm:

  1. Define functions of the state and distribute them between different levels of government according to a subsidiarity principle; i.e. a function should be transferred to the lowest government level capable of implementing it.
  2. Estimate the volume of funds needed to implement these functions.
  3. Assign sufficient revenue sources to local governments.
  4. If a community is too small to generate a sufficient revenue flow, merge several communities and repeat steps 3-4, keeping the distance between the center of such a united community and its most remote settlement below a defined limit.
  5. Establish feedback mechanisms through which people in a community could control the authorities and impact their decision-making. These mechanisms are not only elections, but also, more importantly, permanent between-elections activities, such as public hearings/discussions of drafts of local government decisions.
  6. Use a few communities as pilots and thus find out potential strengths and weaknesses of the proposed reform and make necessary corrections.

The outcome of this algorithm should be a logically connected package of legislative changes rather than a bunch of separate documents.

The development of this reform should be as transparent as possible, and accompanied by wide information and education campaigns about the opportunities that decentralization will provide, and the ways to use these opportunities. These information campaigns are necessary because many Ukrainians now think that decentralization (or federalization) is pushed by the Russian president in order to split Ukraine into parts.

As with all reforms, the decentralization has its potential benefits and risks, which should be accounted for. Fortunately, there exists both a wide academic literature and international experience on this issue.

The economic literature, both theoretical and empirical, does not unambiguously show that “decentralization is good”. Rather, a success of decentralization depends on a number of other factors, such as the presence of democracy (Inman, 2008) and a sufficient accountability of the government (both local and central).

In itself, decentralization does not lead to higher economic growth (e.g. the review of Feld et al, 2013). However, when accompanied by other growth-enhancing reforms, decentralization can positively impact a country’s economic development (Bardhan 2002).

Both the literature and experience of other countries suggest the following major risks of decentralization:

  1. Decentralization may increase corruption at the local level. If a local official is not accountable to a higher-level government, she may try to extract some rent from her position. This risk can be reduced by a high transparency of the government and working mechanisms of control of citizens over officials.

Indeed, Lessmann and Markwardt (2009) show that decentralization lowers corruption in countries with high levels of freedom of the press, and is harmful for countries where monitoring of the government is not efficient. Besides, Fan, Lin and Treisman (2009) find that “giving local governments a larger stake in locally generated income can reduce their bribe extraction”, i.e. for decentralization to lower corruption, the institutional setup should encourage local officials to create a favorable business environment in their regions.

  1. Decentralization may intensify secessionist movements. To lower this risk, the largest volume of responsibilities should be transferred to the lowest (community) level. It is rather easy for separatists to buy support of oblast-level officials and get control over an entire oblast. It would be much harder for them to buy every community head in an oblast. Moreover, getting control over an oblast, even rayon by rayon, let alone by community, is practically infeasible.
  2. Decentralization enhances initial inequality between regions – so the central government has to step in by providing subsidies/subventions to less developed regions (Cai and Treisman, 2005).

At the same time, the “bonuses” of decentralization are worth taking the risks:

  1. Reduction of tensions between the regions. In the Ukrainian situation, this implies removing grounds for mutual accusations that “one region feeds other regions” or “one region rules the entire country”. If a party that wins a majority in the national elections does not have extensive power over the daily life of people, they can more easily accept the fact this is not the party they voted for.
  2. Improvement of the national politics by increasing competition between local officials, and between local and central officials. As we know, competition typically increases the quality of a product. Political competition is no exception. As Myerson (2006) notes, “by creating more opportunities for politicians to build reputation as responsible democratic leaders, a federal [decentralized] system can effectively offer an insurance policy against general failure of democracy”. Thus, democracy and decentralization strengthen each other.
  3. More efficient government. On average, policy decisions will be made closer to their final beneficiaries and hence, will be more fitted to the needs of a certain community. At the same time, all levels of government will work more efficiently.

Decentralization does not imply a weakening of the central government. Rather, it frees its institutions from an unnecessary workload allowing them to concentrate on more strategic tasks, such as:

  • protecting people’s rights by establishing a working judicial and security (police and army) systems;
  • forming a strategic vision and general directions of the country’s development;
  • protecting the country’s interests on the international level.

To make sure that decentralization does not result in feudalization, local officials should be controlled not only by local citizens but also by the central government (law enforcement); strong country-wide political parties would also help to hold the country together.


A decentralization of the Ukrainian political system is currently in the very focus of political, public and research debate.

However, this reform is not likely to be an easy one. The prerequisites for successful decentralization include functioning democratic mechanisms – fair elections, a free press and a strong civil society – resulting in government accountability. Also, for the decentralization reform to succeed, it needs to be coherently bundled with a range of political and administrative reforms (such as the development of a functioning judicial system, deregulation, reduction of rent-seeking opportunities etc.), and development and implementation of such a package is challenging and time-consuming.

At the same time, a wisely designed decentralization process will be highly beneficial for Ukraine, both politically and economically. It will strengthen democracy (by increasing people’s participation) and improve the quality of national politics by introducing more competition into the political system. It is also likely to significantly contribute to economic growth and prosperity, and these benefits make the decentralization reform in Ukraine a challenge worth undertaking despite of all the costs and risks.



  • Bardhan, Pranab (2002). “Decentralization of Governance and Development,” Journal of Economic Perspectives, American Economic Association, vol. 16(4), pp. 185-205
  • Brancati, Dawn (2006). Decentralization: Fueling the Fire or Dampening the Flames of Ethnic Conflict and Secessionism? International Organization. Vol.60, issue 03, pp. 651-685
  • Cai, Hongbin and Daniel Treisman (2005). Does competition for capital discipline governments? Decentralization, globalization and public policy. The American Economic Review, Vol. 95, No. 3, Jun.2005
  • Cai, Hongbin and Daniel Treisman (2009). Political decentralization and policy experimentation. Quarterly Journal of Political Science. Vol 4. Issue 1.
  • Deiwiks, Christa, Cederman, Lars-Erik und Kristian S. Gleditsch (2012). Inequality and Conflict in Federations. Journal of Peace Research. March 2012 vol. 49 no. 2, pp. 289-304
  • Enikolopov, Ruben and Ekaterina Zhuravskaya (2007). Decentralization and political institutions. Journal of Public Economics, No. 91, pp. 2261–2290
  • Fan, C. Simon, Lin, Chen and Daniel Treisman (2009). Political decentralization and corruption: Evidence from around the world. Journal of Public Economics. Vol.: 93 (2009)
    Issue: 1-2, pp: 14-34
  • Inman, Robert P. (2008). Federalism’s Values and the Value of Federalism. NBER Working Paper 13735. http://www.nber.org/papers/w13735
  • Lars P. Feld, Baskaran, Thushyanthan and Jan Schnellenbach (2013). Fiscal Federalism, Decentralization and Economic Growth: A Meta-Analysis. Public Finance Review 41 (4), 421-445
  • Lessmann, Christian and Gunther Markwardt (2009). One Size Fits All? Decentralization, Corruption, and the Monitoring of Bureaucrats, CESIFO Working Paper No. 2662, Cat. 2: Public Choice.
  • Myerson, Roger B. (2006). Federalism and Incentives for Success of Democracy. Quarterly Journal of Political Science, 2006, 1: 3–23
  • Treisman, Daniel (2006). Fiscal decentralization, governance, and economic performance: a reconsideration. Economics and Politics, July 2006, 18, 2, pp. 219-35.

Reputation for Quality and Entry in Procurement: Is there a Trade-Off?

20140224 Reputation for Quality and Entry Image 01

How much weight should be given to past performance indicators when selecting contractors? Does a large weight assigned to suppliers’ previous performance deter entry by new, innovative suppliers that have no track records for the very reason that they are new? If yes, how should we take this into account when designing procurements for firms or governments? This note describes recent research that sheds light on these questions crucial for every government and organization.


How should past performance be accounted for when selecting (public or private) contractors? On one hand, giving large weight to suppliers’ previous performance in assigning contracts may improve incentives in procurement; on the other hand, it may deter entry by new, innovative suppliers for the very reason that they are new. Are there ways to structure procurement rules and procedures to minimize or eliminate these costs? How can well performing suppliers be rewarded, but not at the expense of losing the most innovative start-ups, that could pose important positive externalities on the buyer and on society overall? These questions are important ones and every procurement manager, in the private and public sector, should know how to answer, or at least how to think about, them. Unfortunately, if one looks at the leading management and operations textbooks, or at public procurement textbooks, it is hard to find a line that could help in making these crucial decisions. The only procurement book that at least mentions these crucial questions, to our knowledge, is the Handbook of Procurement (2006; see Ch. 18, by Dellarocas et al.). However, even that handbook falls short in providing evidence-based or – more generally – research-based guidance for these questions. This is the case because there is practically no research dealing with these everyday problems. One recent exception is an experimental study recently undertaken by Butler, Carbone, Conzo and Spagnolo (2013) that will be discussed in depth in the reminder of this brief.

Public Policy Relevance  

Before getting into the results of this recent study, let me provide some background information that will give an idea of the relevance of these questions for public policy.

Public procurement currently accounts for between 15% and 20% of the GDP in developed countries (see http://cordis.europa.eu/fp7/ict/pcp/key_en.html).  In 2011, the total public procurement market in the EU – i.e. the purchases of goods, services and public works by governments and public utilities – reached a size of approximately €2,500 billion, corresponding to 19 percent of GDP (see e.g. http://ec.europa.eu/internal_market/publicprocurement/docs/modernising_rules/public-procurement-indicators-2011_en.pdf). As Table 1 below shows, despite year-to-year fluctuations, there has been an overall increase in procurement expenditures relative to 2007 levels, both absolutely and proportionately. The increasing shift in focus in EU innovation policies, from “push-based” mechanisms like R&D subsidies/tax breaks towards demand-led “pull” mechanisms, like Pre-Commercial Procurement, is likely to further increase the volume of this market.

Table 1: Total EU procurement expenditure on works, goods and services
In EUR billion











As a % of GDP











Source: DG MARKT. 2011. Public Procurement Indicators.

The enormous size of this market notwithstanding, we know relatively little about whether, when, and how buyers should use reputational indicators based on past performance in selecting among sellers, and whether the use of such indicators necessarily reduces the ability of new sellers—i.e., sellers with no history of past performance—to enter the market.

It is well known that reputational mechanisms that reward past performance are important governance tools that complement (and sometimes substitute for) contracts in private transactions (Calzolari and Spagnolo 2009). Private buyers, however, are typically only concerned about the price and quality of the good they buy. Regulators in charge of public procurement, instead, are usually also concerned that the public procurement process is transparent and open for obvious accountability reasons. The need to prevent favoritism and corruption has led lawmakers around the world to ensure that open and transparent auctions where bidders are treated equally—even when in some crucial dimensions they have very different track records—are used whenever possible.

The trend in the US

The costs of limiting discretion to ensure public buyers’ accountability – such as the possibly large cost of not allowing reputational forces to complement incomplete procurement contracts – were stressed by Kelman (1990), who pushed for a deep reform of the US procurement system when he was the head of public procurement during the Clinton administration. The reform was targeted at reducing the rigidity of procurement rules in the Federal Acquisition Regulations and allowing public buyers to adopt more flexible purchasing practices common in the private sector, including giving more weight to suppliers’ past performance. Since the Federal Acquisitions Streamlining Act of 1994, US federal departments and agencies are expected to record past contractors’ performance evaluations and share them through common platforms for use in future contractor selection.

However, the US Senate recently expressed the apparently widespread concern that past performance-based selection criteria could hinder new and small businesses’ ability to enter and compete effectively, leading to an intriguing, but inconclusive report by the General Accountability Office.

This is not to say that US regulators were not concerned with the ability of small and medium enterprises (SME) – sometimes the most innovative part of the economy – to enter public procurement markets. In the US, this long-held concern led to large programs like the Small Business Act, with its rules limiting the bundling of public demand in very large procurements and establishing the Small Business Agency, and the ‘set aside’ (procurement only open to SMEs) common in many types of procurement auctions. However, the worry that past-performance based selection may contribute to the exclusion of novel and smaller firms only arose in the last couple of years.

The (opposite) trend in the EU

The European Union has instead been moving precisely in the opposite direction. An important concern driving procurement regulation in Europe since the Treaty of Rome has been helping the process of common market integration by increasing cross-border procurement, i.e., the amount of goods and services each EU Member State buys from contractors based in other states. The EU Procurement Directives that coordinate public procurement regulation in the various European states have been limiting the use of past-performance information in the process of selecting among offers—a feature that came under broad attack during the 2011 consultation for the revision of the EU Directives (see Replies to the Consultation on the 2011 EU Green Book on Public Procurement regulation). The EU regulators appear to have been always convinced that using reputational indicators as a criteria for selecting contractors leads to manipulations in favour of local incumbents, at the expense of cross-border procurement and market integration.  (see e.g., http://ec.europa.eu/internal_market/publicprocurement/modernising_rules/consultations/index_en.htm).

Only very recently – now that the US is moving towards reconsidering the effects, and possibly limiting the use of past-performance indicators – has the EU started to move (again in the opposite direction to the US) towards leaving more space to these indicators, perhaps as a reaction to the comments they received in recent consultations.

Finally, to have an idea of the lack of research-based knowledge guiding policy in this field, note that in some cases EU regulation already acknowledges the crucial importance of past-performance based reputation for some types of procurement. For example, the European Research Council (ERC) provides funding to top researchers in Europe, who are selected through peer review, and the track record of the researchers is usually the main awarding criterion. ERC funding is distributed almost exclusively based on reputation criteria in order to support the best and the brightest. Other European instruments for the procurement of research, such as the FET-OPEN program, are based on a strictly enforced, completely anonymous evaluation instead, without obvious reasons justifying this opposite approach. On the dedicated homepage of these programs it states: “The anonymity policy applied to short proposals has changed and is strictly applied. The part B of a short STREP proposal may not include the name of any organization involved in the consortium nor any other information that could identify an applicant. Furthermore, strictly no bibliographic references are permitted.”

Reputation and Entry in Procurement

In a recent research paper (Butler, Carbone, Conzo and Spagnolo, 2013) we have been trying to fill at least part of this knowledge gap and offer some initial evidence-based guidelines for future policy.

The study

We build a simple model of repeated procurement with limited enforcement and potential entry and implement it in the laboratory. We focus on reputation as an incentive system to limit moral hazard in the quality dimension as well as on the effect of reputation on selection through entry. We assume that some costly-to-produce quality dimension of supply, although observable to the parties, is too costly to verify for a court to be governed through explicit contracting and is therefore left to reputational governance. We make the additional assumption that there is a potential entrant firm that is more efficient than all incumbent firms. In this context, we study how quality, price, entry and welfare change with the introduction of a simple and transparent reputational mechanism. This mechanism rewards an incumbent firm that chooses to provide (costly) high-quality production with a bid subsidy in the subsequent procurement auction, and may also award a bid subsidy (of varying size) to an entrant with no history of production.

Note that in the case of public procurement and of firms’ vendor rating systems, we are talking about reputational mechanisms based on public rules, known and accepted by suppliers. Formal mechanisms and rules give commitment power to the buyer and can be designed in many different ways. A common mistake is to assume that reputational mechanisms must be designed along the line of the eBay feedback system, in which new sellers start with “zero reputation”. However, a buyer with some commitment power concerning the rules for information aggregation and diffusion and for selecting suppliers may well award a positive rating to new entrants—e.g. the maximum possible rating, or the average rating in the market, putting entrants at less of a disadvantage—and ensure that this is taken into account by the scoring rule that selects the contractor, even if the contractor has never before interacted with the buyer. Indeed, private corporations often have vendor rating systems in which all suppliers start off with the same maximal reputational capital—a given number of points—and then lose points when performing poorly and are suspended for some time if their reputational capital falls below a certain low threshold. This type of vendor rating system creates an advantage for new suppliers, most likely stimulating rather than hindering entry, suggesting that it is possible to design a reputational mechanism in public procurement that simultaneously sustains quality and entry.

The results

First of all, the study shows that concerns about reputation-based selection hindering entry are justified: naively introducing a “standard” reputational mechanism in which only good past performance is rewarded with a bid subsidy in the following procurement auction increases quality provision, but it also significantly reduces entry.

In contrast to this first result, the study goes on to show that properly designed reputational mechanisms in which new entrants, with no history of past performance, are awarded a moderate or high reputation score—as is often done in the private sector—actually foster rather than hinder entry while, at the same time, delivering a substantial increase in high quality goods provision.

The third important result of this study is that the total cost to buyers (buyer’s transfer) does not increase when a reputational mechanism is introduced, even though (costly) quality provision increases. The introduction of bid subsidies for good past performance appears to benefit the buyer/taxpayer by increasing competition for incumbency, driving winning bids down sufficiently to offset the potential increase in procurement costs due to bid subsidies.


Considered together, the findings in Butler et al. (2013) suggest that there need not be a trade-off between reputation and entry in procurement, and that the debates both in the EU and the US are rather misplaced. The results suggest that the dual goals of providing incentives for quality provision and increasing entry and cross-border procurement – in the EU or elsewhere – are achievable through an appropriately designed reputational mechanism. Policy makers should therefore probably stop quarrelling about whether a generic past-performance based reputational mechanism should be introduced, and instead focus on how such a mechanism should be designed.


  • Butler Jeff, Carbone Enrica, Conzo Pierluigi and Giancarlo Spagnolo. 2013. “Reputation and Enty in Procurement.” CEPR Discussion Paper No. 9651.
  • Calzolari, Giacomo and Giancarlo Spagnolo. 2009. “Relational Contracts and Competitive Screening.”  CEPR Discussion paper No. 7434.
  • European Commission. 2011. Green Paper on the modernisation of EU public procurement policy Towards a more efficient European Procurement Market. Available at http://ec.europa.eu/internal_market/consultations/index_en.htm
  • European Commission. 2011. Green Paper on the modernisation of EU public procurement policy Towards a more efficient European Procurement Market. Synthesis of Replies.  Available at http://ec.europa.eu/internal_market/consultations/docs/2011/public_procurement/synthesis_document_en.pdf
  • Government Accountability Office. 2011. (GAO-12-102R, Oct. 18, 2011). Prior Experience and Past Performance as Evaluation Criteria in the Award of Federal Construction Contracts. Available at  http://www.gao.gov/products/GAO-12-102R
  • Kelman, Steven. 1990.  Procurement and Public Management: The Fear of Discretion and the Quality of Government Performance: American Enterprise Institute Press.
  • Yukins, Christoher. 2008. “Are IDIQS Inefficient? Sharing Lessons with European Framework Contracting.” Public Contract Law Journal,  37(3): 545-568.

Can Public Enforcement of Competition Policy Increase Distortions in the Economy?

High office buildings facing sky representing Institutions and Services Trade

Authors: Vasiliki Bageri, University of Athens, Yannis Katsoulacos, Univeristy of Athens,  and Giancarlo Spagnolo, SITE.

Competition law has recently been introduced in a large number of developed and emerging economies. Most of these countries adopted the common practice of basing antitrust fines on affected commerce rather than on collusive profits, and in some countries caps on fines have been introduced based on total firm sales rather than on affected commerce. Based on recent research, this policy brief explains how a number of large distortions are connected to these policies, which may facilitate competition authorities in their everyday job but at the high risk of harming the consumer and distorting industrial development. We conclude by discussing the possibility to depart from these distortive rules-of-thumb opened by recent advancements in data availability and econometric techniques, as well as by the considerable experience matured in estimating collusive profits when calculating damages in private antitrust litigation.

Competition policy has become a prominent policy in many developing economies, from Brazil to India. Indeed, the available evidence suggests that in countries where law enforcement institutions are sufficiently effective, a well designed and enforced competition policy can significantly improve total and labor productivity growth.

It is already well known that the private enforcement of competition policy can give rise to large distortions: since competition law is enforced by Judges and not by economist, it is easy for firms to strategically use the possibility to sue under the provision of competition law to protect their market position rather than the law being used to protect competition.

It is somewhat less known that a poor public enforcement of Competition Law by publicly funded competition authorities can also end up worsening market distortions rather than curing them. In the reminder of this policy brief we explain why, according to recent research, a mild and suboptimal enforcement of antitrust provisions – in the sense of fines that are too low to deter unlawful conduct (horizontal agreements and cartels in particular) and fines which are based on firm revenue rather than on the extra profits generated by the unlawful conduct, could significantly harm social welfare, even if we abstract from the direct cost the public enforcement of competition law imply for society.

Current Practice in Setting Fines

A very important tool for the effective enforcement of Competition Law is the penalties imposed on violators by regulators and courts. In this policy brief, we uncover a number of distortions that current penalty policies generate, we explain how their size is affected by market characteristics such as the elasticity of demand, and quantify them based on market data.

In contrast to what economic theory predicts, in most jurisdictions, Competition Authorities (CAs), but also courts where in charge, use rules-of-thumbs to set penalties that – although well established in legal tradition and in sentencing guidelines and possibly easy to apply – are hard to justify and interpret in logical economic terms. Thus, antitrust penalties are based on affected commerce rather than on collusive profits, and caps on penalties are often introduced based on total firm sales rather than on affected commerce.

A First Well Known Distortion Due to Legal Practice

A first and obvious distortive effect of penalty caps linked to total (worldwide) firm revenue is that specialized firms which are active mostly in their core market expect lower penalties than more diversified firms that are also active in several other markets than the relevant one. This distortion – why for God’s sake should diversified firms active on many markets face higher penalties than more narrowly focused firms? – could in principle induce firms that are at risk of antitrust legal action to inefficiently under-diversify or split their business to reduce their legal liability.

In a recent paper published in the Economic Journal, we examine two other, less obvious, distortions that occur when the volume of affected commerce is used as a base to calculate antitrust penalties.

A Second Distortion: Poorly Enforced Competition Law May Increase Welfare Losses from Monopoly Power

If expected penalties are not sufficient to deter the cartel, which seems to be the norm given the number of cartels that CAs continue to discover, penalties based on revenue rather than on collusive profits induce firms to increase cartel prices above the monopoly level that they would have set if penalties were based on collusive profits. Intuitively, this would be done in order to reduce revenues and thus the penalty. However, this exacerbates the harm caused by the cartel relative to a monopolized situation with similar penalties related to profits, or even relative to a situation with no penalties due to the distortive effects of the higher price and, in comparison to a situation with no penalties, the presence of antitrust enforcement costs.

A Third Distortion: Firms at the Bottom of the Value Chain May Pay a Multiple of the Fine Paid by Firms at the Top for an Identical Infringement

Firms with a high revenue/profit ratio, e.g. firms at the end of a vertical production chain, expect larger penalties relative to the same collusive profits that firms with a lower revenue/profit ratio would get. Our empirically based simulations suggest that the welfare losses produced by these distortions can be very large, and that they may generate penalties differing by over a factor of 20 for firms that instead should have faced the same penalty.

Note that this third distortion takes place also when at least for some industries fines are sufficiently high to deter cartels. This distortion means that competition is only enforced in industries that happen to be in the lower end of the production chain, and not in industries where the lack of competition is producing larger social costs. Note also that our estimation is based only on observed fines, i.e. on fines paid by cartels that are not deterred. Since cartels tend to be deterred by higher fines, this suggest that if we could take into account the fines that would have been paid by those cartels that were deterred (if any), the size of the estimated distortion would likely increase!

Concluding remarks

We argue that if one wants to implement a policy, one must be ready to do it well otherwise it may be better to not do it at all. This is particularly relevant for countries with weaker institutional environments where it is likely that political and institutional constraints will not allow for a sufficiently independent and forceful enforcement of the Competition Law.

It is worth noting that – in particular in the US but also increasingly so in the EU – the rules-of-thumb discussed above do not produce any saving in enforcement costs because the prescribed cap on fines requires courts to calculate firms’ collusive profits anyway. Furthermore, the distortions we identified are not substitutes where either one or the other is present. Instead, they are all simultaneously present and add to one another in terms of poor enforcement.

Where there are sufficient resources to allow for a proper implementation and where enforcement of Competition Law is available, developments in economics and econometrics make it possible to estimate illegal profits from antitrust infringements with reasonable precision, as regularly done to assess damages. It is time to change these distortive rules-of-thumb that make revenue so central for calculating penalties, if the only thing the distortions give us is savings in the costs of data collection and illegal profit estimation.