Abstract
This paper focuses on sports-related public spending in the Member States of the European Union (EU). Based on the public procurement database of the EU (TED), a sport-related public procurement database was built and analysed. Using data from 33 countries for the period 2017–2019, the paper describes the characteristics of sport-related public procurements. The research highlights that the public database is an adequate way of making the data on public procurements available, where traditionally the latency was high. The characteristics found for the eight most active Member States include a high proportion of construction works. There is a connection between countries and the dominant type of purchasing organisations, although the involvement of central purchasing bodies is not a game-changer in this area. Higher value contracts usually lasted for longer and the length of contracts has a strong connection to the contract types. Non-negotiated types of procedures show a far higher average contract value than negotiated procedure types. When the lowest price criterion was applied, the total procurement value was significantly lower.
1 Introduction
Extensive research has been done recently on the efficiency of support and public spending connected to sports. The idea of analysing sports-related procurement arose from the publication of the European Court of Justice case C-155/19, which specifically found that the contracting authority was subject to public procurement in the field of sport. This decision gave importance to analysing sports-related EU procurements.
While earlier sport-focused papers enquired about the sources and motives of financing, public procurement research mostly aimed at legal issues. Still, little is known about sport-related public procurements and, in particular, quantitative analysis is rarely used in this field. Our paper aims to fill this gap by analysing the Tenders Electronic Daily (TED) announcement database. In addition to a general description of EU-wide sport-related public procurements, we also offer a cross-country comparison.
Our goal is to become acquainted with the characteristics of sport-related procurements in the member states of the European Union that post their public procurement notices in the Official Journal of the European Union. We produce a more in-depth analysis concerning the member states that use public procurement to the greatest extent in terms of value in order to enable a more comprehensive analysis of the individual characteristics based on the data of the Member States active in public procurement.
The analysis by Nessel and Kościółek (2020) using the DEA method, and the research by Dallmeyer et al. (2018), specifically looked at the role of the magnitude of funding, its period and consistency. These research reports only made a fleeting reference to public procurement criteria related to aid or did not mention them at all. The analyses, which expressly studied the relationship between state aid and government spending for sports purposes, dwelled much more frequently on the issue of public procurement. Typically, these analyses studied the nature of the aid from a legal point of view and they were less qualitative in nature in terms of sports in general, or football in particular (Traupel 2014; Van Rompuy – van Maren, 2016; Nicolaides 2015; Cattaneo 2018).
Irizarry (2017) took an interesting approach in a research project focusing on the analysis of a financing model and sport-related PPP, with very strong public procurement content. However, in researching the literature, we have not found any content expressly on the specificities and characteristics of sports-related public procurement spending.
This paper is to our best knowledge a first-in-kind, building on the TED data as the details of the individual announcements had to be transformed into a single consistent database to ensure a good overview. Also, the database needed considerable cleansing to sort out eventual issues in the records.
For the TED database, we built on analyses that, although not related to a sports procurement, analysed the database earlier. Svátek et al. (2014) mainly focused on the analysis of data content, Plaček (2020) investigated the specificity of delays, while Tas (2020) looked for a link between regulatory quality and competition and efficiency by analysing TED data. Fazekas and Kocsis (2020) identified red flags in the announcements in a study specifically on corruption. Similarly, the development of www.redflags.eu by Transparency International aimed to identify corruption risks. They used the standards of the country publishing the notice to determine the extent to which it is considered risky, thus developing an explicitly dynamic indicator for the database and thus making TED data analysis increasingly diverse.
The abovementioned examples show that TED data are worth analysing, but there are limitations to this analysis. Amman and Essig (2015) quote the problems of data limitations and our own experience shows that the publicly available database requires a lot of work to make it suitable for academic studies. In order to move away from focusing on the announcements only, we had to link the call for tender and the announcement of decision databases. This allowed us to investigate the specifics of public procurement at the procedural level. Muñoz-Garcia and Vila (2019) also linked microdata on selected bidders from the TED to the Orbis database in their study.
In a similarly innovative way, the so-called Common Procurement Vocabulary (CPV code), was used to select sport-related procurements. This method allowed us to investigate the specificities of the procurement procedures of a given procurement object and sector, rather than the practices of a particular contracting authority.
2 Data and methodology
Based on the experiences of the literature explored, we carry out an analysis based on TED data, which combines the aspects of the source and nature of the aid (ministry, municipality, body governed by public law, EU funding) and those of ensuring competition (solely price-based competition, publication of prior notices, single bid procedures, auctions), and other aspects that take into account various other effects of spending (support to SMEs, use of long-term methods of procurement).
During the data analysis, we decided to use both the contract notices and contract awards. Similarly, we cover both databases, as Svátek et al. (2014) concluded in their article, but with the difference that we combined the two databases and analysed the data at the procedural level. After downloading the details of public procurements for the years 2017–2019 from Tenders Electronic Daily, tender calls and result announcements had to be paired, sometimes even across years (any tender is recorded for its year of announcement). The database created contained data from 33 countries (EU28, UK, North Macedonia, Iceland, Norway, and Switzerland). We cleaned up the TED database, leaving out the “CANCELLED” procedures from the data. We used the three most important contract notices and contract award notices of the TED database: (1) contracting authorities (Annex II), (2) utilities (Annex V), (3) concession procurements (Annex XXI) of Public Procurement Directives (2014). 1
Building on the data collected, we set up two databases. The first contained each tender only once and included tender level information only. The second database collected lot (purchase entity) level information implying that each tender showed up as many times as the number of lots linked to it. The lot database was used to analyse information like the number of offers received or the characteristics of the bid winner. At this stage, we worked with all public procurement data. The database was later narrowed down to sport-related procurement items. In many cases it was not possible to eliminate data gaps in the database, therefore we did not study technical issues in several cases, which would have been interesting but filling in the data was not mandatory in several EU member states, so if the data gap was too large, we rejected the subject matter of the study.
When reviewing the databases, it became clear that data cleansing was inevitable. It seems that the questionnaire collecting information from the member states was sometimes filled in using the digit 9 to show the lack of data creating confusion when numeric input was asked for. Thus, in the case of the total purchase value, we had to remove numbers containing only the digit 9. To keep realistic inputs only, we also removed any total tender value above 999 999 998 euros and below 10 000 euros. As for the lots, we deleted lot values below 100 euros. We decided to create the reviewed purchase value in a separate variable, so for statistics not building on the total purchase value, we may use the entire original database.
In the end, the tender level database contained 291 032 items of which 266 480 (91.6%) included acceptable information on the total tender purchase value. The lot level database (covering exactly the same tenders) consisted of 813 115 records of which 667 351 (82.1%) contained valid information on the total lot value.
As a next step, tenders (and related lots) were hand-picked for being sport-related. We used the Common Procurement Vocabulary (CPV) codes indicating the subject matter of procurement to filter out the unambiguously sport-related subject matters of procurements (goods, services and works). The CPV codes include a number of procurement subject matters, which may be linked to sport-related procurements, but in order to study the clearly sport-related procurements, we narrowed down the scope of our study. Altogether 226 CPVs were classified as sport-related. Following the identification of the codes, we selected the notices launching and closing procedures from the TED database, whose main object was linked to a sport-related purpose. We did not take the subsidiary objects into account. We coded the data based on TED CSV Open Data Notes & Codebook (2022) Version 3.2, 2019-05-16 Altogether, 226 CPVs were classified as sport-related, Table 1 provides some examples of these.
Examples of sport-related CPV codes (including works, goods, services)
| 37415000-0 Athletics equipment 37416000-7 Leisure equipment 37420000-8 Gymnasium equipment 37421000-5 Gymnasium mats 37422000-2 Gymnastic bars or beams 37422100-3 Gymnastic bars 37422200-4 Gymnastic beams 37423000-9 Gymnastic ropes, rings or climbing accessories 37423100-0 Gymnastic ropes 37423200-1 Gymnastic rings 37423300-2 Gymnastic climbing accessories 37424000-6 Gymnastic vaulting equipment 37425000-3 Gymnastic trampolines |
45212224-2 Construction work for stadiums 45212225-9 Construction work for sports halls 45212230-7 Fitting-out work for changing rooms 45212290-5 Repair and maintenance work on sports facilities 45235200-5 Construction work for running tracks 45235210-8 Resurfacing of running tracks 45236100-1 Levelling work for various sports facilities 45236110-4 Levelling work for sports grounds 45236111-1 Golf-course resurfacing services 45236112-8 Tennis-court grading services 45236113-5 Levelling services for racetracks 45236114-2 Racecourse grading services 45236119-7 Sports field repair services 45242100-6 Construction work for water-sport facilities 45242200-7 Construction work for pleasure-boat marinas 77320000-9 Sports-field maintenance services |
Source: CPV Regulation (2007).
Of the records covering the period 2017 to 2019, 1754 (0.6%) tenders and 3109 (0.4%) lots qualified for being sport-related. Of all the 33 countries it was only Lithuania that did not report a single sport-related tender over those three years. Cyprus and Iceland recorded only one tender, while North-Macedonia and Malta had only two tenders each.
Tables 2 and 3 offer an overview of the whole selection process. As most of the countries had very few sport-related public procurements recorded over the period under study, we separated eight countries (CH, DE, ES, FR, IT, PL, SE, UK) with at least 50 sport-related tenders with valid total procurement values for analysing the country-level traits of the tenders.
Annual breakdown of the databases
| Total tenders | Tenders with valid values | Sport tenders | |
| 2017 | 92 517 | 81 384 | 684 |
| 2018 | 104 785 | 91 495 | 664 |
| 2019 | 69 178 | 59 219 | 406 |
| Total | 266 480 | 232 098 | 1 754 |
| Total lots | Lots with valid values | Sport lots | |
| 2017 | 229 854 | 217 483 | 1 133 |
| 2018 | 265 612 | 249 755 | 1 250 |
| 2019 | 171 885 | 160 537 | 726 |
| Total | 667 351 | 627 775 | 3 109 |
Source: collected by the authors.
The country level breakdown of the tender database
| Total tenders | Tenders with valid value | Sport tenders | Total tenders | Tenders with valid value | Sport tenders | ||
| AT | 3 934 | 3 850 | 21 | IT | 12 094 | 11 730 | 97 |
| BE | 3 615 | 3 428 | 13 | LI | 26 | 26 | 0 |
| BG | 8 584 | 6 541 | 16 | LT | 4 839 | 4 076 | 4 |
| CH | 7 065 | 6 861 | 59 | LU | 963 | 944 | 9 |
| CY | 314 | 298 | 1 | LV | 2 375 | 2 007 | 5 |
| CZ | 10 250 | 9 441 | 47 | MK | 775 | 687 | 2 |
| DE | 55 604 | 43 983 | 548 | MT | 628 | 601 | 2 |
| DK | 2 554 | 2 410 | 8 | NL | 5 350 | 3 374 | 56 |
| EE | 1 392 | 1 262 | 7 | NO | 3 703 | 3 450 | 27 |
| ES | 17 698 | 16 992 | 273 | PL | 44 805 | 36 464 | 63 |
| FI | 4 358 | 4 223 | 20 | PT | 3 335 | 3 125 | 8 |
| FR | 31 603 | 29 117 | 251 | RO | 3 425 | 3 065 | 15 |
| GR | 2 644 | 2 412 | 9 | SE | 12 624 | 11 135 | 68 |
| HR | 2 487 | 2 270 | 7 | SI | 2 369 | 2 177 | 5 |
| HU | 4 299 | 4 041 | 28 | SK | 1 052 | 1 029 | 6 |
| IE | 1 475 | 1 329 | 10 | UK | 10 025 | 9 544 | 68 |
| IS | 216 | 206 | 1 | Total | 266 480 | 232 098 | 1 754 |
Source: collected by the authors.
When it comes to the total sport-related (valid) tender value, Italy and France have spent the most, followed by the UK (Fig. 1). Contrary to its relatively small size and population and a modest number of recorded tenders, Hungary had outstanding spending covering 7 percent of the total tender value considered.
The distribution of sport-related total tender value across countries Source: collected by the authors.
Citation: Society and Economy 2022; 10.1556/204.2022.00011
The distribution of sport-related total tender value across countries Source: collected by the authors.
Citation: Society and Economy 2022; 10.1556/204.2022.00011
The distribution of sport-related total tender value across countries Source: collected by the authors.
Citation: Society and Economy 2022; 10.1556/204.2022.00011
The average tender value is quite different across countries (Fig. 2). Romania and Hungary had 20–40 times higher average tender values than Germany or Bulgaria. Also, these two countries have 6–8 times higher average sport-related tender values than their non-sport-related average. Due to the small number of recorded tenders in some of the countries, the cross-country paired T-statistic comparison was only performed for the eight countries with more than 50 tenders recorded. Results show that a number of significant pair-wise differences exist. At 5 percent significance the average sport procurement value for CH and DE is less than that in FR, IT, and PL (UK and SE had a huge SD across their tender values).
Average tender value for countries with at least 10 reported sport-related public procurements Note: asterix marks countries included in country-level analysis. Source: collected by the authors.
Citation: Society and Economy 2022; 10.1556/204.2022.00011
Average tender value for countries with at least 10 reported sport-related public procurements Note: asterix marks countries included in country-level analysis. Source: collected by the authors.
Citation: Society and Economy 2022; 10.1556/204.2022.00011
Average tender value for countries with at least 10 reported sport-related public procurements Note: asterix marks countries included in country-level analysis. Source: collected by the authors.
Citation: Society and Economy 2022; 10.1556/204.2022.00011
3 Most important results
We present the results in two parts. First, we focus on all the sport-related public procurements applicable in our database. Next, the paper presents the results from the comparison of the eight countries with at least 50 sport-related procurements with valid total purchase value data.
3.1 Results of the analysis of TED sport-related procurement data
Countries spent close to 29.2% of their sport-related procurements on services, 3.0% on goods and 67.8% on works (Table 4). However, paired T statistics show a significant difference between ‘supplies and services’ and ‘supplies and works’ average contract values, thus works account only for 58.5% of the purchase lots. The procurements of supplies (goods) have an average value that amounts only to one-sixth of that of works.
Sport-related procurements lots in type breakdown
| Contract type | N | Sum (euros) | Sum % | Mean | Std. Deviation |
| 1 Services | 711 | 1 326 808 497.02 | 29.20% | 1 866 116.03 | 7 932 987.66 |
| 2 Supplies | 460 | 137 545 123.32 | 3.03% | 299 011.14 | 1 032 322.51 |
| 3 Works | 1 651 | 3 079 204 934.68 | 67.77% | 1 865 054.47 | 6 378 382.55 |
| Total | 2 822 | 4 543 558 555.02 | 100.00% | 1 610 049.10 | 6 335 865.09 |
Source: collected by the authors.
As for the contracting authorities of the individual countries, most of the procurements came from regional or local authorities, and bodies governed by public law (“body governed by public law” is the contracting status of those sports market players that are mostly financed by public funds). Characteristically, latency is very high in this later group, because a list of contracting authorities including every one of them does not exist in some Member States (there is such a list, for instance, in Hungary). Interestingly, water, energy, transport and telecommunications sectors made a small number of high-value sport procurements (which could be explained by some outliers or data error), while ministries had a relatively high number of contracts with lower average values. When testing for statistically significant differences, we found that ministries' average sport purchase values were significantly lower than for regional and local authorities and other players, but significantly exceeded that of the EU agencies (Table 5).
Sport-related procurements lots by contracting body type
| N | Mean | Std. Deviation | Total Value | |
| 1 Ministry or any other national or federal authority | 160 | 674 770.41 | 2 172 497.81 | 107 963 265.57 |
| 3 Regional or local authority | 1 726 | 1 711 435.44 | 6 899 777.92 | 2 953 937 567.89 |
| 4 Water, energy, transport and telecommunications sectors | 14 | 16 453 591.40 | 27 560 237.66 | 230 350 279.57 |
| 5 European Union institution/agency | 2 | 147 660.54 | 28 423.93 | 295 321.07 |
| 6 Body governed by public law | 371 | 842 212.47 | 2 215 078.23 | 312 460 824.74 |
| 8 Other | 392 | 1 858 033.78 | 5 671 108.02 | 728 349 243.04 |
| Total | 2 665 | 1 626 024.95 | 6 444 091.47 | 4 333 356 501.88 |
Source: collected by the authors.
We also investigated whether the involvement of a central procurement body had any effect on the average purchase value (Table 6). The assumption is that a professional central purchasing body could be more successful in reducing prices by reaching out to more contractors or choosing a better process type than other bodies. At the same time, central bodies tend to purchase bigger amounts than individual contracting entities. We found that while contracts signed by a central body had significantly higher average purchase values for non-sport-related items than the contracts of other entities, no statistically significant differences were identified in the case of sport-related items.
Involvement of a central purchasing body in public procurements
| Sport-related | Other | ||
| Central purchasing body | N | 81 | 8 517 |
| Mean | 2 850 972 | 2 387 278 | |
| Std. Deviation | 7 487 857 | 16 512 270 | |
| Other entity | N | 1 341 | 202 967 |
| Mean | 2 490 266 | 1 725 855 | |
| Std. Deviation | 8 644 258 | 12 436 965 |
Source: collected by the authors.
The correlation between the contract duration and the total purchase value was 0.333 and significant even at a 1 percent level, which implies that higher value contracts usually lasted for longer. The length of contracts has a strong connection to the contract types (Table 7). Any paired comparison for contract types showed a significant difference. The difference in length is particularly outstanding when contrasting building and service contracts: the former hardly exceeded 1.5 years while the latter averaged to almost 4 years.
Contract length (months) for various types of sport-related contracts
| N | Mean | Std. Deviation | |
| 1 S Services | 504 | 47.95 | 61.75 |
| 2 U Supplies | 282 | 19.88 | 29.65 |
| 3 W Works | 860 | 16.89 | 41.62 |
| Total | 1 646 | 26.91 | 49.17 |
Source: collected by the authors.
At the same time, procedure types showed little difference when it came to contractual value. This is particularly interesting because for non-sport-related procurements we found significant differences across procedure types (Table 8). Paired T-tests only showed a significant difference between the average value of open procedures and competitive dialogues (COD) in the case of sport-related contracts, while for other contracts pairwise differences were significant in most of the cases. It seems that for high-value procurements, COD procedures dominated, while the lowest contractual value was linked to the most popular open procedures no matter whether the procurement was sport-related or not.
Contractual value across procedure types*
| Sport-related contracts | Other contracts | ||||||
| N | Mean | Std. Deviation | N | Mean | Std. Deviation | ||
| 1 COD | 12 | 32.71 | 30.58 | 1 COD | 619 | 21.73 | 68.80 |
| 2 NIP | 3 | 4.21 | 3.47 | 2 NIP | 66 | 2.49 | 6.94 |
| 5 NIC | 62 | 7.02 | 15.05 | 5 NIC | 13 847 | 3.84 | 25.32 |
| 6 OPE | 1 448 | 1.91 | 6.74 | 6 OPE | 209 863 | 1.44 | 9.74 |
| 7 RES | 40 | 7.35 | 11.61 | 7 RES | 5 605 | 5.38 | 24.68 |
| 8 AWP | 0 | 8 AWP | 162 | 1.02 | 2.92 | ||
| Total | 1 565 | 2.49 | 8.34 | Total | 230 162 | 1.73 | 12.43 |
Notes: AWP: award without prior publication of a contract notice; COD: competitive dialogue; NIC/NIP: negotiated with a call for competition (the abbreviation changed over the years); OPE: open; RES: restricted.
Source: collected by the authors.
However, once we categorised the procedures based on whether those were built on negotiations or nor, the difference became statistically significant (Table 9). Open and restricted procedures do not require negotiations and show a far higher average contract value than procedure types with negotiations.
Contractual value procedures with and without negotiations for sport-related procurements
| N | Mean | Std. Deviation | |
| Negotiations involved | 78 | 10 794 907.49 | 20 057 770.32 |
| No negotiations | 1 488 | 2 057 105.86 | 6 967 703.39 |
| Total | 1 566 | 2 492 322.04 | 8 339 022.73 |
Source: collected by the authors.
The Chi-square test confirmed a strong connection between procedure and contract types (Table 10). One would see a particular preference of NIC procedures for service contracts and the same for restricted procedures in the case of construction works.
Crosstab of procedure and contract types
| Type of procedure | Type of contract | ||||
| S Service |
U Supply |
W Works |
Total | ||
| COD | Count | 5.0 | 0.0 | 7.0 | 12.0 |
| Expected Count | 3.8 | 1.9 | 6.3 | 12.0 | |
| NIP | Count | 0.0 | 0.0 | 2.0 | 2.0 |
| Expected Count | 0.6 | 0.3 | 1.0 | 2.0 | |
| NEC | Count | 1.0 | 0.0 | 0.0 | 1.0 |
| Expected Count | 0.3 | 0.2 | 0.5 | 1.0 | |
| NEG | Count | 0.0 | 0.0 | 1.0 | 1.0 |
| Expected Count | 0.3 | 0.2 | 0.5 | 1.0 | |
| NIC | Count | 28.0 | 3.0 | 30.0 | 61.0 |
| Expected Count | 19.2 | 9.8 | 31.9 | 61.0 | |
| OPE | Count | 505.0 | 276.0 | 843.0 | 1 624.0 |
| Expected Count | 512.0 | 262.0 | 850.0 | 1 624.0 | |
| RES | Count | 13.0 | 4.0 | 35.0 | 52.0 |
| Expected Count | 16.4 | 8.4 | 27.2 | 52.0 | |
| Total | Count | 552.0 | 283.0 | 918.0 | 1 753.0 |
Note: NEC/NEG: negotiated with a call for competition.
Source: collected by the authors.
Accelerated procedures were not significantly different in terms of contractual value from the rest of the procedures in the case of sport-related procurements. However, the 35 accelerated procedures had an average contract value of 1.3 million euro in contrast to that of 2.5 million for the remaining 1 531 procedures suggesting the practice of choosing accelerated processes for lower value procurements also in case of sport-related items, just like in the case of other contracts, where this difference was significant and similar in proportion.
Electronic auctioning is very rare for sport-related contracts. While the ratio was 1.93% for other contracts, only five sport-related contracts, representing no more than 0.33%, were reported using the electronic method.
We also tested whether the contracts related to a project and/or programme financed by European Union funds had any effect on the contract value (Table 11). While both for sport-related and other procedures the differences are only significant at a 6 percent level, it is worth underlining that sport-related procedures tended to have lower contractual value when partly financed from EU funds, while the opposite was true for the rest or the procurements.
EU funding effects
| N | Mean | Std. Deviation | Sig. | ||
| Sport-related | No EU funds | 2 522 | 1 696 016.56 | 6 644 978.29 | |
| EU funds used | 237 | 876 376.90 | 2 514 215.36 | ||
| Total | 2 759 | 1 625 608.95 | 6 399 611.78 | 0.059 | |
| Other | No EU funds | 563 141 | 868 392.97 | 9 877 926.82 | |
| EU funds used | 46 664 | 958 608.31 | 8 094 535.24 | ||
| Total | 609 805 | 875 296.50 | 9 753 009.65 | 0.055 |
Source: collected by the authors.
The awarding criterion used to make the final choice among the offers had a statistically significant effect on the contractual value: the “Lowest price” method showed 40 percent lower average purchase value than the other two groups (Table 12). While unfortunately a lack of data was very frequent for this issue, the lowest price criterion seemed to radically lower the total procurement value. It is worrying that the highest contractual value was measured in the case where the purchasing entity forgot or did not want to provide information on the decision criterion applied. A pairwise T-test could not confirm a statistically significant difference between the groups “Most economically advantageous tender” and “Missing”, though. 2
Effect of awarding criterion
| N | Mean | Std. Deviation | |
| 0 Missing | 1 132 | 2 151 659.34 | 8 130 171.13 |
| 1 L Lowest price | 712 | 906 103.95 | 4 061 694.56 |
| 2 M Most economically advantageous tender | 978 | 1 495 638.21 | 5 164 427.87 |
| Total | 2 822 | 1 610 049.10 | 6 335 865.09 |
Source: collected by the authors.
As Table 13 describes, lots won by SMEs were of significantly lower value. Other firms typically gain contracts of 93.2 percent higher total value. However, the number of offers received or the number of offers received from SMEs had no significant connection to contractual value. Thus, it is not that SMEs would prefer smaller contracts or would not try to win bigger value procurements, they are just less successful with those.
Contrasting sport-related contract lots with SME winners and other winners
| N | Mean | Std. Deviation | |
| No SME | 1 120 | 2 270 739.94 | 8 221 952.58 |
| SME winning | 1 702 | 1 175 281.92 | 4 650 536.16 |
| Total | 2 822 | 1 610 049.10 | 6 335 865.09 |
Source: collected by the authors.
3.2 Results of comparing countries with the most sport-related procurements
Our cross-country analysis targeted the same variables that we reviewed in section 3.1. Our Chi-square test results show that there is a connection between countries and the dominant type of purchasing organisations (Table 14.) In Switzerland and Sweden, regional and local authorities dominate the sport-related procurements, while in Italy and Poland “other” organisations have a role more significant than elsewhere. However, in the UK “bodies governed by public law” play a more significant role than in the other countries.
Dominance of purchasing organisations across countries
| 1 Ministry or any other national or federal authority, including their regional of local subdivisions | 3 Regional or local authority | 4 Water, energy, transport and telecommunications sectors | 6 Body governed by public law | 8 Other | Total | ||
| CH | Count | 2.0 | 57.0 | 0.0 | 0.0 | 0.0 | 59.0 |
| Expected Count | 1.5 | 38.1 | 0.6 | 6.2 | 12.7 | 59.0 | |
| DE | Count | 7.0 | 372.0 | 2.0 | 51.0 | 116.0 | 548.0 |
| Expected Count | 13.8 | 353.7 | 5.4 | 57.2 | 117.9 | 548.0 | |
| ES | Count | 6.0 | 176.0 | 4.0 | 22.0 | 65.0 | 273.0 |
| Expected Count | 6.9 | 176.2 | 2.7 | 28.5 | 58.7 | 273.0 | |
| FR | Count | 12.0 | 154.0 | 3.0 | 37.0 | 45.0 | 251.0 |
| Expected Count | 6.3 | 162.0 | 2.5 | 26.2 | 54.0 | 251.0 | |
| IT | Count | 4.0 | 54.0 | 4.0 | 9.0 | 26.0 | 97.0 |
| Expected Count | 2.4 | 62.6 | 1.0 | 10.1 | 20.9 | 97.0 | |
| PL | Count | 1.0 | 13.0 | 1.0 | 8.0 | 40.0 | 63.0 |
| Expected Count | 1.6 | 40.7 | 0.6 | 6.6 | 13.6 | 63.0 | |
| SE | Count | 2.0 | 60.0 | 0.0 | 0.0 | 6.0 | 68.0 |
| Expected Count | 1.7 | 43.9 | 0.7 | 7.1 | 14.6 | 68.0 | |
| UK | Count | 2.0 | 35.0 | 0.0 | 22.0 | 9.0 | 68.0 |
| Expected Count | 1.7 | 43.9 | 0.7 | 7.1 | 14.6 | 68.0 | |
| Total | Count | 36.0 | 921.0 | 14.0 | 149.0 | 307.0 | 1 427.0 |
| Expected Count | 36.0 | 921.0 | 14.0 | 149.0 | 307.0 | 1 427.0 |
Source: collected by the authors.
During the period under study, radical differences emerged in contract types of sport-related public procurements (Table 15). While in Switzerland and Germany construction works gave the overwhelming majority of the contracts, Spain, UK, and Italy purchased more services than the average. The UK is also outstanding in supplies procurements, just like Sweden, and Poland.
Contract types across countries
| 1 S Services | 2 U Supplies | 3 W Works | Total | ||
| CH | Count | 5.0 | 2.0 | 52.0 | 59.0 |
| Expected Count | 19.0 | 8.1 | 31.8 | 59.0 | |
| DE | Count | 31.0 | 32.0 | 485.0 | 548.0 |
| Expected Count | 176.7 | 75.7 | 295.7 | 548.0 | |
| ES | Count | 218.0 | 37.0 | 18.0 | 273.0 |
| Expected Count | 88.0 | 37.7 | 147.3 | 273.0 | |
| FR | Count | 64.0 | 34.0 | 153.0 | 251.0 |
| Expected Count | 80.9 | 34.7 | 135.4 | 251.0 | |
| IT | Count | 67.0 | 9.0 | 21.0 | 97.0 |
| Expected Count | 31.3 | 13.4 | 52.3 | 97.0 | |
| PL | Count | 24.0 | 20.0 | 19.0 | 63.0 |
| Expected Count | 20.3 | 8.7 | 34.0 | 63.0 | |
| SE | Count | 12.0 | 37.0 | 19.0 | 68.0 |
| Expected Count | 21.9 | 9.4 | 36.7 | 68.0 | |
| UK | Count | 39.0 | 26.0 | 3.0 | 68.0 |
| Expected Count | 21.9 | 9.4 | 36.7 | 68.0 | |
| Total | Count | 460.0 | 197.0 | 770.0 | 1 427.0 |
| Expected Count | 460.0 | 197.0 | 770.0 | 1 427.0 |
Source: collected by the authors.
The preferred procedure types also showed significant dispersion (Table 16). From Switzerland, only open procedures were reported, and with a few exceptions, the same was true for Poland, Sweden and Spain. However, the competitive dialogue (COD), the negotiated (NIC) and the restricted procedures (RES) were more popular than the average in the UK and France.
Procedure types across countries
| 1 COD | 2 INP | 5 NIC | 6 OPE | 7 RES | 8 AWP | Total | ||
| CH | Count | 0.0 | 0.0 | 0.0 | 59.0 | 0.0 | 0.0 | 59.0 |
| Expected Count | 0.4 | 0.1 | 2.2 | 55.4 | 1.0 | 0.0 | 59.0 | |
| DE | Count | 0.0 | 0.0 | 19.0 | 525.0 | 4.0 | 0.0 | 548.0 |
| Expected Count | 3.5 | 0.8 | 20.0 | 514.2 | 9.2 | 0.4 | 548.0 | |
| ES | Count | 0.0 | 0.0 | 3.0 | 268.0 | 2.0 | 0.0 | 273.0 |
| Expected Count | 1.7 | 0.4 | 10.0 | 256.2 | 4.6 | 0.2 | 273.0 | |
| FR | Count | 6.0 | 0.0 | 18.0 | 217.0 | 8.0 | 1.0 | 250.0 |
| Expected Count | 1.6 | 0.4 | 9.1 | 234.6 | 4.2 | 0.2 | 250.0 | |
| IT | Count | 0.0 | 2.0 | 2.0 | 90.0 | 3.0 | 0.0 | 97.0 |
| Expected Count | 0.6 | 0.1 | 3.5 | 91.0 | 1.6 | 0.1 | 97.0 | |
| PL | Count | 0.0 | 0.0 | 0.0 | 62.0 | 1.0 | 0.0 | 63.0 |
| Expected Count | 0.4 | 0.1 | 2.3 | 59.1 | 1.1 | 0.0 | 63.0 | |
| SE | Count | 0.0 | 0.0 | 0.0 | 68.0 | 0.0 | 0.0 | 68.0 |
| Expected Count | 0.4 | 0.1 | 2.5 | 63.8 | 1.1 | 0.0 | 68.0 | |
| UK | Count | 3.0 | 0.0 | 10.0 | 49.0 | 6.0 | 0.0 | 68.0 |
| Expected Count | 0.4 | 0.1 | 2.5 | 63.8 | 1.1 | 0.0 | 68.0 | |
| Total | Count | 9.0 | 2.0 | 52.0 | 1 338.0 | 24.0 | 1.0 | 1 426.0 |
| Expected Count | 9.0 | 2.0 | 52.0 | 1 338.0 | 24.0 | 1.0 | 1 426.0 |
Source: collected by the authors.
Switzerland, Spain, Poland, and Sweden reported the near-exclusivity of processes without negotiations, while the UK, France, and Germany used negotiation-based procedures more extensively (Table 17). However, we could not identify statistically significant differences in the use of accelerated procedures and electronic auctions across countries.
Processes with and without negotiations across countries
| No negotiations | Negotiations involved | Total | ||
| CH | Count | 59.0 | 0.0 | 59.0 |
| Expected Count | 56.4 | 2.6 | 59.0 | |
| DE | Count | 529.0 | 19.0 | 548.0 |
| Expected Count | 523.4 | 24.6 | 548.0 | |
| ES | Count | 270.0 | 3.0 | 273.0 |
| Expected Count | 260.8 | 12.2 | 273.0 | |
| FR | Count | 226.0 | 25.0 | 251.0 |
| Expected Count | 239.7 | 11.3 | 251.0 | |
| IT | Count | 93.0 | 4.0 | 97.0 |
| Expected Count | 92.6 | 4.4 | 97.0 | |
| PL | Count | 63.0 | 0.0 | 63.0 |
| Expected Count | 60.2 | 2.8 | 63.0 | |
| SE | Count | 68.0 | 0.0 | 68.0 |
| Expected Count | 65.0 | 3.0 | 68.0 | |
| UK | Count | 55.0 | 13.0 | 68.0 |
| Expected Count | 65.0 | 3.0 | 68.0 | |
| Total | Count | 1 363.0 | 64.0 | 1 427.0 |
| Expected Count | 1 363.0 | 64.0 | 1 427.0 |
Source: collected by the authors.
While most of the countries hardly used any EU funds for sport-related public purchases, France and Poland had several such procurements (Table 18). In the case of the non-EU member Switzerland, this was not an option.
The use of EU funds in sport-related public procurements across countries
| Only local funds used | EU funds used | Total | ||
| DE | Count | 592.0 | 12.0 | 604.0 |
| Expected Count | 552.3 | 51.7 | 604.0 | |
| ES | Count | 315.0 | 8.0 | 323.0 |
| Expected Count | 295.3 | 27.7 | 323.0 | |
| FR | Count | 968.0 | 163.0 | 1 131.0 |
| Expected Count | 1 034.1 | 96.9 | 1 131.0 | |
| IT | Count | 107.0 | 3.0 | 110.0 |
| Expected Count | 100.6 | 9.4 | 110.0 | |
| PL | Count | 92.0 | 22.0 | 114.0 |
| Expected Count | 104.2 | 9.8 | 114.0 | |
| SE | Count | 73.0 | 0.0 | 73.0 |
| Expected Count | 66.7 | 6.3 | 73.0 | |
| UK | Count | 95.0 | 2.0 | 97.0 |
| Expected Count | 88.7 | 8.3 | 97.0 | |
| Total | Count | 2 242.0 | 210.0 | 2 452.0 |
| Expected Count | 2 242.0 | 210.0 | 2 452.0 |
Source: collected by the authors.
Finally, for the different decision-making methods, the reporting discipline was very heterogeneous (Table 19). While Germany performed very well, the UK, Italy, and Sweden had a huge proportion of missing data. When we considered the procurements with known decision-making methods only, it became clear that the “lowest price” method is preferred in Germany and Sweden, while the rest of the countries had a strong preference for the “most economically advantageous tender” method.
The use of different decision-making methods across countries
| 0 Missing | 1 L Lowest price | 2 M Most economically advantageous tender | Total | ||
| DE | Count | 84.0 | 486.0 | 38.0 | 608.0 |
| Expected Count | 253.9 | 129.4 | 224.7 | 608.0 | |
| ES | Count | 171.0 | 14.0 | 187.0 | 372.0 |
| Expected Count | 155.4 | 79.2 | 137.5 | 372.0 | |
| FR | Count | 557.0 | 6.0 | 568.0 | 1 131.0 |
| Expected Count | 472.4 | 240.7 | 417.9 | 1 131.0 | |
| IT | Count | 71.0 | 8.0 | 38.0 | 117.0 |
| Expected Count | 48.9 | 24.9 | 43.2 | 117.0 | |
| PL | Count | 33.0 | 6.0 | 76.0 | 115.0 |
| Expected Count | 48.0 | 24.5 | 42.5 | 115.0 | |
| SE | Count | 58.0 | 15.0 | 0.0 | 73.0 |
| Expected Count | 30.5 | 15.5 | 27.0 | 73.0 | |
| UK | Count | 76.0 | 0.0 | 22.0 | 98.0 |
| Expected Count | 40.9 | 20.9 | 36.2 | 98.0 | |
| Total | Count | 1 050.0 | 535.0 | 929.0 | 2 514.0 |
| Expected Count | 1 050.0 | 535.0 | 929.0 | 2 514.0 |
Source: collected by the authors.
4 Conclusions and implications
Our research project focused on the study of sport-related public procurement in 33 European countries for the years 2017–2019. To date, there has not been an expressly public procurement type antecedent to this research project. This was the first attempt to identify sport-related procurements and, after building our own database, to explore the characteristics of procurement.
Based on the total sport-related tenders, Italy and France have spent the most, followed by the UK. Contrary to its relatively small size and population, and a modest number of recorded tenders, Hungary had outstanding spending accounting for 7 percent of the total tender value considered. Romania and Hungary had 20-40 times higher average tender values than Germany or Bulgaria. Also, these two countries have 6–8 times higher average sport-related tender values than their non-sport-related average.
One of the most interesting facts is that the largest share of public money is spent on construction. In other words, public support is typically in the form of construction activities in the countries surveyed.
Importantly, in addition to regional and local governments, bodies governed by public law are the majority of contracting authorities. In our experience, this is an area of sport where it is extremely difficult to establish the existence of a public procurement obligation and therefore the level of evasion of public procurement rules is relatively high.
For example, the perception of a body governed by public law depends on the type and extent of the source of funding defined by the economic operator. Much depends on the willingness of the stakeholder to take risks and the likelihood of scrutiny when deciding on its own procurement status.
The involvement of central purchasing bodies is not a game-changer in this area. The use of negotiated procedures is less widespread (mostly used for services), and in this respect, sports procurement follows the characteristics of non-sport-related procurement.
Higher value contracts usually lasted for longer and the length of contracts has a strong connection to the contract types. The difference in length is particularly outstanding when contrasting building and service contracts: the former hardly exceeded 1.5 years while the latter averaged almost 4.0 years.
It seems that for high-value procurements, competitive dialogue procedures dominate, while the lowest contractual value is linked to the most popular open procedures, no matter whether the procurement is sport-related or not. Procedure types that require negotiations showed a far lower average contract value than procedure types without negotiations.
SMEs are more successful for lower values in sports procurement. Nevertheless, the number of offers received or the number of SME offers received had no significant connection to the contractual value.
When contrasting individual countries, various significant differences came to light. In sport-related procurements, Switzerland and Sweden stand out in the percentage of local government purchases. In Switzerland and Germany, construction works make up the overwhelming majority of the contracts, while Spain, the UK, and Italy purchased more services than the average. The UK is also outstanding in supplies procurements, just like Sweden, and Poland. Predominantly open procedures were reported from Switzerland, Poland, Sweden and Spain, however, in the UK and France, competitive dialogue, restrictive procedures and negotiated procedures with call for competition are more popular than the average.
Switzerland, Spain, Poland, and Sweden reported nearly exclusively processes without negotiations, while the UK, France, and Germany used negotiation-based procedures more extensively. However, we found no differences in the use of accelerated procedures and electronic auctions across countries. Only France and Poland tend to use EU funds for sport-related procurements regularly. When it comes to deciding on the best offer, the EU legislation explicitly lays down the obligation to choose additional evaluation criteria instead of the lowest price, although the lowest price criterion seemed to radically lower the total procurement value. Still, Germany and Sweden prefer the “lowest price” method, while the rest of the countries favour the “most economically advantageous tender” method.
Our analysis is constrained by the fact that the data subject to the national regimes, i.e., below the EU threshold, are not comparable and available in an appropriate structure. Results are also limited by the lack of valid data in the case of some non-obligatory TED variables. Data quality may also differ across countries leading to biases.
Acknowledgement
This work was created on the basis of an NKFI commission under project K 137794.
References
Amann, M. – Essig, M. (2015): Public Procurement of Innovation: Empirical Evidence from Eu Public Authorities on Barriers for the Promotion of Innovation. Innovation: The European Journal of Social Science Research 28(3): 282–292.
Cattaneo, A. (2018): State Aid and Sport. In: Anderson, J. – Parrish, R. – García, B. (eds): Research Handbook on EU Sports Law and Policy. Edward Elgar Publishing.
CPV Regulation (2007): Commission Regulation (EC) No 213/2008 of 28 November 2007 amending Regulation (EC) No 2195/2002 of the European Parliament and of the Council on the Common Procurement Vocabulary (CPV) and Directives 2004/17/EC and 2004/18/EC of the European Parliament and of the Council on public procurement procedures, as regards the revision of the CPV.
Dallmeyer, S. – Wicker, P. – Breuer, C. (2018): The Relationship Between Sport-Related Government Spending and Sport and Exercise Participation: The Role of Funding Size, Period, and Consistency. International Journal of Health Promotion and Education 56(4–5): 237–247.
Fazekas, M. – Kocsis, G. (2020): Uncovering High-Level Corruption: Cross-National Objective Corruption Risk Indicators using Public Procurement Data. British Journal of Political Science 50(1): 155–164.
Irizarry, A. M. H. (2017): If You Build it, They Will Relocate: Public Private Partnerships in Sport Stadium Financing. Public Contract Law Journal 46(4): 853–872.
Muñoz-Garcia, C. – Vila, J. (2019): Value Creation in the International Public Procurement Market: In Search of Springbok Firms. Journal of Business Research 101: 516–521.
Nessel, K. – Kościółek, S. (2020): The Total Sporting Arms Race: Benchmarking the Efficiency of Public Expenditure on Sports in EU Countries. European Sport Management Quarterly. https://doi.org/10.1080/16184742.2020.1833956.
Nicolaides, P. (2015): A Critical Analysis of the Application of State Aid Rules to Sport. Maastricht Journal of European and Comparative Law 22(2): 209–223.
Plaček, M. – Ochrana, F. – Schmidt, M. – Nemec, J. – Půček, M. (2020): The Factors Causing Delays in Public Procurement: the Czech Republic Versus the UK. Public Money & Management 40(2): 131–139.
Svátek, V. – Mynarz, J. – Węcel, K. – Klímek, J. – Knap, T. – Nečaský, M. (2014): Linked Open Data for Public Procurement. In: Auer, S. – Bryl, V. – Tramp, S. (eds): Linked Open Data--Creating Knowledge Out of Interlinked Data. Cham: Springer, pp. 196–213.
Tas, B. K. O. (2020): Effect of Public Procurement Regulation on Competition and Cost-Effectiveness. Journal of Regulatory Economics 58(1): 59–77.
TED CSV Open Data Notes & Codebook (2022): TED CSV Open Data Notes & Codebook. Version 3.2, 2019-05-16.
TED database : Tenders Electronic Daily (2017–2019) www.ted.europa.eu, accessed 17/05/2022.
Traupel, T. (2014): Football and State Aid: Really the Greatest Pastime in the World. European State Aid Law Quarterly .2014(3): 414–425.
Van Rompuy, B. – van Maren, O. (2016): EU Control of State Aid to Professional Sport: Why Now? In: Duval, A. , Van Rompuy, B. (Eds.), The Legacy of Bosman. TMC Asser Press, The Hague, pp. 153–185.
Directive 2014/24/EU on public procurement (repealing Directive 2004/18/EC); Directive 2014/25/EU on procurement by entities operating in the water, energy, transport and postal services; Directive 2014/23/EU on concessions.
Two different awarding criteria can be indicated in the contract notice, despite the fact that the public procurement legislation provides for more than one possibility. Therefore, the data analysis allowed us to draw conclusions on these two criteria.
