In the beginning of 2014 several online portals tried to estimate the total amount of money that Közgép SA., owned by former close friend of the Hungarian Prime Minister, Viktor Orbán, won in public procurement in 2013. The four sources (Portfolio.hu, Átlátszó, K-Monitor, Hír24) calculated four different sums. There was more than half a billion Euro difference between the highest and the lowest estimates. Given that public procurement records are open to the public, it is evidently worthwhile to investigate these calculations and their different outcomes. The CRCB has compared the estimates of Átlátszó and K-Monitor using the www.kozpenzkereso.eu data. We seek to explain why Átlátszó estimated 1,454 billion, www.kozpenzkereso.eu 1,228 billion, and K-Monitor 1,207 billion euro. The attached brief analysis shows that there are exclusively technical reasons for this. The contracting authorities do not always comply with the law governing the publication of contract details. For example, the results of a public procurement contract published in the Official Journal of the European Union do not consistently appear again in the Public Procurement Bulletin, as required by law. Hence duplications cannot be filtered automatically. The winners of public procurements are not indicated in the same way in every announcement, and it is challenging to identify the members of the applying consortiums, since they do not have any unique identifiers. In addition, the misspelling of names only makes the situation more complicated. If all of the data were available in a well-structured, ordered and downloadable form in the Public Procurement Database, problems like these would not be present. Improved data quality and the enforcement of existing laws on publishing procurement contract details are essential to improving the public’s ability to hold the government accountable.
 Calculated by the average HUF-EUR exchange rate in 2013. Source: http://demo.mtieco.hu/Pages/hist.aspx?sub=1&CategoryId=19&lang=hun&menuid=f86de2d3-4f83-4952-852e-14a09776a2f2&year=2013&histtype=3
The construction of modern highways is a massive undertaking involving large sums of money and significant work. In Hungary these projects represent a significant portion of public spending and generate some of the largest outflows of funds from state coffers to the bank accounts of the companies winning tenders. Given the high complexity and cost of these projects, they are especially at risk for corruption. It is therefore worthwhile to study this part of the public procurement market closely, and to monitor how corruption risks change over time.
In the accompanying chart, we plot the average CRI (Corruption Risk Index) score per contract related to highway construction each year from 2009 to 2012. The CRI is calculated based on a collection of elementary objective indicators serving as red flags in public procurement contracting aggregating to a value between 0 and 1. The highway-related data set consists of 264 contracts. Scores above 0.2 indicate medium corruption risk presence, while .27 and above can be considered high. According to these benchmarks, Hungarian highway construction public contracts face medium to high corruption risks, with a worsening trend. In 2012 the average CRI score was nearly 0.3.
The favoritism of individuals or groups by ideally impartial government bodies can be difficult to measure empirically. When it comes to the topic of favoritism in public procurement, we can ask the following: are contracts awarded by the state on political merits? Do firms with connections to ruling political parties and their members have a decisive advantage over their connectionless competitors? In the Hungarian case, the data represented in the included chart suggests that the answer to both questions is a resounding yes.
The chart tracks two groups of companies involved in public procurement in Hungary. The red group consists of the top 30 companies by public procurement market share in 2009. The graph plots their collective market share in the following years. In 2009 they won around 30% of all procurement contracts by value. The orange group consists of the top 30 companies by public procurement market share in 2011. In 2011 these firms also held 30% of the market. In the meantime, the old market leaders in the red group saw their collective market share fall to below 10%. Like a reflection in a mirror, the orange group’s market share in 2009 was also below 10%.
What happened in the mean time? In 2010 Hungary witnessed a change in government. The numbers indicate a large change in the group of public procurement market winners. Succinctly: the old winners lost twenty percent of the entire market to the new winners within a year of a change in government. More egregious examples can be found in specific public procurement submarkets. In construction, the shift between the leading groups before and after was nearly 50% of the total market.
In our opinion this is clear evidence that political favoritism is a key driver of the Hungarian public procurement market. The CRCB’s formal paper on this topic is available on our webpage.
It is very important to study the change in concentration of high corruption risks in terms of the connections between participants in Hungarian public procurement before and after Prime Minister Viktor Orbán’s election in 2010. In other words: in what way did networks of corruption in public procurement change with the change in government? More fundamentally: does the network viewed as a whole look different? The kozpenzkereso.eu database can help us answer these questions.
In the first graphic we have drawn two networks of public procurement issuers and contract winners involved in high corruption risk procurement and the connections between them. The networks on the left (a) and right (b) are constructed from data from before and after the change of government in 2010, respectively.
The colored points represent either issuers or winners of public procurement contracts. Red actors have high corruption risk scores for a majority of the contracts in which they were involved, while the black participants have high corruption risk scores for all of their procurement dealings. The size of the points is proportional to the amount of money paid out, in the case of issuers, or won, in the case of bidders, in the given time period. Two entities are connected by an edge if they were involved in a successful procurement together.
We can see from the diagram that following the change in government the amount of participants whose contracts always signaled high corruption risk, colored black, decreased. On the other hand, the number of participants with a majority of high corruption risk dealings, colored red, increased. The overall volume of contracts and money represented by red points increased. It is also apparent that the network as a whole has become increasingly centralized. While before 2010 the network of corrupt actors was more spread out, with several isolated or weakly-connected subnetworks, since the change in government the network has become denser and more centered.