Government spending outside public procurement

Each year, governments in the EU purchase goods and services worth over 2 trillion EUR. However, only 17% of this goes through tenders following EU rules that ensure basic competition and transparency. The other 83% may be regulated by national procurement laws. As we will show, even that is typically not the case. We will argue that the majority of government purchases happen outside of any legal rules, lacking even the most basic transparency.

Let’s start with a look at basic data. Using national accounts and the European procurement journal (TED), the European Commission publishes estimates of money spent on procurement following EU regulation relative to all government spending. Figure 1 shows vast differences across countries — while in Germany, only 5.7% of tenders get published in TED, in Latvia its 59.1 % (measured by volume). For whole EU it’s the above mentioned 17.3 % tenders falling under the procurement directive.

Figure 1: Share of government spending regulated by EU procurement directive (source: [1] — table 7, own calculation)

Yes, the EU procurement directive is important. But maybe not thatimportant. We see that it regulates only a fraction of government spending. But, it also sets direction for national legislations dealing with the rest. However, only some of the countries follow such direction— as we will see shortly.

East — West division

From the media, it appears that public procurement is one of the largest areas for the government with the highest risk of nepotism, wasted money and corruption. Stories from Central and Eastern Europe show how this affects people’s lives: from antiseptics in Romanian hospitals to Czech roads that lead nowhere.

However, this is quite an unfair picture. Yes, large contracts sometimes bring scandals and media attention — but is this really where most public money gets wasted? Isn’t it rather in the exceptions where corruption and inefficiency thrive? Unluckily for the EU, these “exceptions” form 83% of its procurement market. The worst records are surprisingly not in Eastern Europe, but in Germany and the Netherlands, where under 10% of tenders apparently abide by EU rules. Figure 2 shows, how such pattern divides EU-15 from the new member states.

Figure 2: Share of government spending regulated by EU procurement directive (source: [1] — table 7, own calculation)

So, where does the 83% of the money go?

The most common reason for a contract falling outside the scope of EU procurement directive is its size. “Below threshold” tenders consist of goods and services below 135 000 € and construction below 5 225 000 €. Why is it then that purchases in richer countries seem to fall more often below the thresholds?

The answer is unclear — either these countries actually prefer smaller projects, or they are better at dodging EU legislation. Either way, the consequences are that their procurement markets are more closed and that national legislation can play a bigger role. The reader is free to speculate about the motives of these countries, we will not do so.

So, do we have a surprisingly small EU regulation? No need to panic yet. The 83% of “exceptions” is where the national legislation comes in. Below EU threshold, each country can set its own rules to ensure efficient spending. But do they?

Of course. Like in the case of the EU, most of the countries have their thresholds above which national rules apply. Below this national threshold, there is no binding legislation and the government units are free to spend the money as they desire — and again, the highest corruption and waste risks apply.

How are the national thresholds set? Our colleagues at the DIGIWHIST project have done an extensive mapping of national legislations. Of course, there are many subtle legal differences which are hard to quantitatively compare, but the thresholds for application of the law are relatively universal.Figure 3 shows thresholds for purchases of goods, above which at least some regulations apply — typically mandatory publication or competition.

Figure 3: National procurement thresholds in 2015, goods purchases by central government. Source: [2], own calculation. The revised data were kindly provided by website authors.

We found five countries at the edge of the EU threshold of 134 000 € — basically passing on regulating tenders below the EU directive scope. Surprisingly, these seem to intersect with countries with the lowest share of EU regulated purchases Germany, the Netherlands, as well as Luxembourg and Austria.

Is this a coincidence? Or does weak regulation below thresholds provide incentives to squeeze the contracts? Its quite intuitive — if German tender under 134K € has little associated duties, German authorities will try hard to get below this threshold. Authorities in other countries, where duties are nearly the same — below or above threshold, logically don’t care about squeezing below thresholds. The larger the difference in legal duties, the more authorities try to stay in the safe, cozy below-threshold world. This is also suggested by pioneering Czech research [3]. In our next article we will use tool called to look deeper into Czech below-threshold spending, and try to find out how exactly this squeezing might work.

Volume of unregulated purchases

Let us end with computing one final number: a rough estimate of government purchases unregulated by both national and EU regulations, ensuring elementary transparency and competition. The high volume of such purchases, again, does bring higher risks of nepotism, corruption or simply wasteful behavior (now methodological notes follow, if not interested, feel free to skip to Figure 4).

To estimate this volume, we will assume that contract volumes are evenly distributed. Such simplification might cause bias— as documented for instance in [3], authorities do naturally squeeze more than proportional volume below the threshold. That is why our estimate will be conservative — true volumes below the national threshold are potentially higher than reported.

With such this assumption we can easily use numbers from Figure 3 to get a rough estimate of purchases falling outside the scope of the national law. For instance in Poland that would be:

30.000 (Polish threshold) / 134.000 (EU threshold)= 0.21 = 21 % purchases outside national law.

Apparently, for Germany, same figure would give 134.000 /134.000 = 100 %.

Using such estimates, we can combine them with earlier figures on volumes above the threshold, to get an overall distribution of spendings:

Figure 4: Government purchases by regulation type. Own calculation based on [1], [2],[4], reflecting national accounts and legislation in 2015.

Finally, when using national accounts [4] in the same fashion as [1], to get the absolute volume of government spendings, we can transform our computed percentages to figures in €, country by country (we drop Sweden and Denmark for lack of threshold data). After doing the math, we get that 1004 out of total 1907 bln. € fall below national thresholds, arriving at our final conclusion:

We roughly estimate, that at least 52 % of government purchases in EU is not regulated by EU procurement directive nor national legislation.

The main portion of such purchases takes place in Germany, Netherlands, Italy, Austria, and Belgium. We hope to stir debate, whether this implies lack of transparency, corruption risk, protectionism from single European market, or it is just perfectly normal. Before talking about problems in the public procurement, we might first want to name the big elephant in the room — the purchases happening outside the public procurement.

Continue reading: Anatomy of “small scale” purchases —the Czech case


[1] Public Procurement Indicators 2015 — DG GROW G4 — Innovative and e-Procurement, link

[2] European Public Accounability Mechanisms, procurement legislative mapping in 2015, link

[3] Manipulation of Procurement Contracts: Evidence from the Introduction of Discretionary Thresholds. Ján Palguta; Filip Pertold. American Economic Journal, 2017

[4] Eurostat, Government revenue, expenditure and main aggregates — gov_10a_main, 2015