ecr-journalЭкономическая наука современной РоссииEconomics of Contemporary Russia1609-14422618-8996Regional Public Organization for Assistance to the Development of Institutions of the Department of Economics of the Russian Academy of Sciences10.33293/1609-1442-2020-1(88)-109-127ecr-journal-461Research ArticleЭКОНОМИЧЕСКАЯ ПОЛИТИКА И ХОЗЯЙСТВЕННАЯ ПРАКТИКАECONOMICAL POLICY AND ECONOMICAL PRACTICEОбнаружение признаков горизонтального сговора при государственных закупках с использованием методов машинного обученияDetecting Indicators of Horizontal Collusion in Public Procurement with Machine Learning Methodshttps://orcid.org/0000-0001-8130-1259МолчановаГлафира ОлеговнаMolchanovaGlafira O.

Лаборатория анализа данных и отраслевой динамики ИОРИ, младший научный сотрудник, SPIN-код 2699-3804

Junior researcher at the laboratory for data analysis and industry dynamics, Institute for Industrial Economics

molchanova-go@ranepa.ruhttps://orcid.org/0000-0001-8207-1790РейАлексей ИгоревичReyAlexey I.

Лаборатория анализа данных и отраслевой динамики ИОРИ, заведующий, SPIN-код 3955-2904

Cand.Econ.Sci., Head of laboratory for data analysis and industry dynamics, Institute for Industrial Economics

rey-ai@ranepa.ruhttps://orcid.org/0000-0001-6544-5359ШагаровДмитрий ЮрьевичShagarovDmitry Yu.

Лаборатория анализа данных и отраслевой динамики ИОРИ, младший научный сотрудник

Junior researcher at the laboratory for data analysis and industry dynamics, Institute for Industrial Economics

shagarov-dy@ranepa.ruРоссийская академия народного хозяйства и государственной службы при Президенте РФ, МоскваРоссияRussian Presidential Academy of National Economy and Public Administration, MoscowRussian Federation20200504202001109127Copyright © Молчанова Г.О., Рей А.И., Шагаров Д.Ю., 20202020Молчанова Г.О., Рей А.И., Шагаров Д.Ю.Molchanova G.O., Rey A.I., Shagarov D.Y.Данная работа распространяется под лицензией Creative Commons Attribution 4.0.This work is licensed under a Creative Commons Attribution 4.0 License.https://www.ecr-journal.ru/jour/article/view/461

Совершенствование закупочных процедур и их цифровизация помогают предотвращать и выявлять картели, но в то же время приводят к появлению новых антиконкурентных схем поведения. В нашем исследовании мы концентрируем внимание на электронных аукционах, которые стали в последние годы основным способом осуществления государственных закупок в России. Тот факт, что электронные аукционы дают доступ к большему числу крупных государственных заказов, усиливает стимулы участников торгов заключать антиконкурентные соглашения. Поэтому улучшение методов выявления картелей на электронных аукционах становится особенно актуальной проблемой. Цель данной работы состояла в разработке метода для обнаружения признаков горизонтального сговора на торгах. С помощью методов машинного обучения мы тренируем классификаторы, которые предсказывают наличие или отсутствие картеля в электронных аукционах в зависимости от распределения ставок участников аукциона, времени подачи заявок, длительности проведения торгов и числа участников. Переменные для модели были отобраны на основании графиков распределений, построенных для выборки по картелям и случайной выборки. Исследование проводится на основе данных с сайта государственных закупок и информации о сговорах на торгах из дел Федеральной антимонопольной службы (ФАС). Результаты показали, что модель случайного леса наиболее точно предсказывает выявление картеля на электронном аукционе. Правильность предсказания – 84%, а полнота и точность модели – 83 и 87%. Наиболее значимыми для классификации переменными оказались уровень снижения цены, разница во времени подачи заявок и начальная (максимальная) цена контракта.

Improvement of procurement procedures and their digitization help prevent and identify cartels, but at the same time lead to the emergence of new anticompetitive schemes. In this paper we focus on electronic auctions, which have become the main method of public procurement in Russia in recent years. As e-auctions provide access to many big government orders; the incentives for bidders to join anti-competitive agreements are increased. Therefore, the development of methods to detect bid rigging at electronic auctions is of high practical importance. The aim of this work was to develop a method for detecting signs of horizontal collusion at an auction. We use machine learning methods to train classifiers that predict the presence or absence of cartel in electronic auctions, depending on the distribution of bidders, the time of submission of applications, the duration of the auction and the number of participants. Variables for the model were selected on the basis of distribution plots built for sample of cartels and random sample. The study is based on data from public procurement Web portal and the information about bid rigging from cases of the Federal Antimonopoly Service. The results showed that the Random forest model most accurately predicts the detection of the cartels on electronic auctions. The accuracy of the prediction is 84%, and the recall and precision of the model are 83 and 87%, respectively. The most significant variables for the classification are the level of price reduction, the difference in the time of application filing of participants and the value of the maximum starting price of contract.

выявление картелейсговоры на торгахгосударственные закупкимашинное обучениеdetecting cartelsbid riggingprocurement auctionsscreening methodsmachine learningReferencesAbrantes-Metz R.M., Froeb L. M., Geweke J. F., Taylor C. T. A variance screen for collusion // International Journal of Industrial Organization. 2006. No. 24. P. 467–486.Abrantes-Metz R.M., Froeb L.M., Geweke J.F., Taylor C.T. (2006). A variance screen for collusion. International Journal of Industrial Organization, no. 24, pp. 467–486.Abrantes-Metz R.M., Kraten M., Metz A. D., Seow G. Libor manipulation // Journal of Banking and Finance. 2012. No. 36. P. 136–150.Abrantes-Metz R.M., Kraten M., Metz A.D., Seow G. (2012). Libor manipulation. Journal of Banking and Finance, no. 36, pp. 136–150.Andreyanov P., Davidson A., Korovkin V. Corruption vs Collusion: Evidence from Russian Procurement Auctions // Technical Report. 2016. mimeo: UCLA.Andreyanov P., Davidson A., Korovkin V. (2016). Corruption vs Collusion: Evidence from Russian Procurement Auctions. Technical Report. mimeo: UCLA.Aryal G., Gabrielli M. F. Testing for collusion in asymmetric first-price auctions // International Journal of Industrial Organization. 2013. Vol. 31. P. 26–35.Aryal G., Gabrielli M.F. (2013). Testing for collusion in asymmetric first-price auctions. International Journal of Industrial Organization, vol. 31, pp. 26–35.Bajari P., Ye L. Deciding between competition and collusion // The Review of Economics and Statistics. 2003. No. 85. P. 971–989.Bajari P., Ye L. (2003). Deciding between competition and collusion. The Review of Economics and Statistics, no. 85, pp. 971–989.Balsevich A., Podkolzina E. A. Indicators of corruption in public procurement: The example of Russian regions // Higher School of Economics. 2014. Research Paper No. WP BRP 76/EC/2014. URL: Available at SSRN: https://ssrn.com/abstract=2530518 or http://dx.doi.org/10.2139/ssrn.2530518.Balsevich A., Podkolzina E.A. (2014). Indicators of corruption in public procurement: The example of Russian regions. Higher School of Economics. Research Paper No. WP BRP 76/EC/2014. URL: Available at SSRN: https://ssrn.com/abstract=2530518 or http://dx.doi.org/10.2139/ssrn.2530518.Bolotova Y., Connor J., Miller D. The Impact of collusion on price behavior: Empirical results from two recent cases // International Journal of Industrial Organization. 2008. No. 26. P. 1290–1307.Bolotova Y., Connor J., Miller D. (2008). The Impact of collusion on price behavior: Empirical results from two recent cases. International Journal of Industrial Organization, no. 26, pp. 1290–1307.Chawla N. V., Bowyer K. W., Hall L. O., Kegelmeyer W. P. SMOTE: Synthetic minority over-sampling technique // Journal of Artificial Intelligence Research. 2002. Vol. 16. P. 321–357.Chawla N.V., Bowyer K.W., Hall L.O., Kegelmeyer W.P. (2002). SMOTE: Synthetic minority over-sampling technique. Journal of Artificial Intelligence Research, vol. 16, pp. 321–357.Chotibhongs R., Arditi D. Analysis of collusive bidding behavior // Construction Management and Economics. 2012. No. 30. P. 221–231.Chotibhongs R., Arditi D. (2012). Analysis of collusive bidding behavior. Construction Management and Economics, no. 30, pp. 221–231.Connor J. M. Global price fixing: Our customers are the enemy. Boston: Kluwer Academic Publishers, 2001.Connor J.M. (2001). Global price fixing: Our customers are the enemy. Boston: Kluwer Academic Publishers.Grout P., Sonderegger S. Structural approaches to cartel detection. In: E. C-D, Atanasiu I. (eds.). European Competition Law Annual: 2006. Enforcement of Prohibition of Cartels. Hart Publishing, 2007. P. 83–103.Grout P., Sonderegger S. (2007). Structural approaches to cartel detection. In: E. C-D, Atanasiu I. (eds.). European Competition Law Annual: 2006. Enforcement of Prohibition of Cartels. Hart Publishing, pp. 83–103.Harrington J. E. Detecting cartels. In: Handbook of Antitrust Economics, Buccirossi P. (ed.). Cambridge: The MIT Press, 2008.Harrington J.E. (2008). Detecting cartels. In: Handbook of Antitrust Economics, Buccirossi P. (ed.). Cambridge: The MIT Press.Huber M., Imhof D. Machine learning with screens for detecting bid-rigging cartels // International Journal of Industrial Organization. 2019. Vol. 65. P. 277–301.Huber M., Imhof D. (2019). Machine learning with screens for detecting bid-rigging cartels. International Journal of Industrial Organization, vol. 65, pp. 277–301.Imhof D., Karagoek Y., Rutz S. Screening for bid rigging, does it work? // Journal of Competition Law and Economics. 2018. No. 14 (2). P. 235–261.Imhof D., Karagoek Y., Rutz S. (2018). Screening for bid rigging, does it work? Journal of Competition Law and Economics, no. 14 (2), pp. 235–261.Ivanov D. I., Nesterov A. S. Identifying bid leakage in procurement auctions: Machine learning approach. 2019. arXiv:1903.00261 [econGN].Ivanov D.I., Nesterov A.S. (2019). Identifying bid leakage in procurement auctions: Machine learning approach. arXiv:1903.00261 [econGN].Jakobsson M. Bid rigging in Swedish procurement auctions. Working Paper. Mimeo: Uppsala University. Department of Economics (Sweden), 2007.Jakobsson M. (2007). Bid rigging in Swedish procurement auctions. Working Paper. Mimeo: Uppsala University. Department of Economics (Sweden).Levenstein M. C., Suslow V. Y., Oswald L. J. Contemporary international cartels and developing countries: Economic effects and implications for competition policy // Antitrust Law Journal. 2004. No. 71. P. 801–852.Jakobsson M. (2007). Bid rigging in Swedish procurement auctions. Working Paper. Mimeo: Uppsala University. Department of Economics (Sweden).Morozov I., Podkolzina E. Collusion detection in procurement auctions // SSRN Electronic Journal. 2013. 10.2139/ssrn.2221809.Levenstein M.C., Suslow V.Y., Oswald L.J. (2004). Contemporary international cartels and developing countries: Economic effects and implications for competition policy. Antitrust Law Journal, no. 71, pp. 801–852.OECD, 2016. Fighting bid rigging in public procurement: Report on implementing the OECD Recommendation. URL: http://www.oecd.org/daf/competition/Fighting-bid-rigging-in-public-procurement-report‑2016.pdf.Levenstein M.C., Suslow V.Y., Oswald L.J. (2004). Contemporary international cartels and developing countries: Economic effects and implications for competition policy. Antitrust Law Journal, no. 71, pp. 801–852.Pesendorfer M. A study of collusion in first-price auction // The Review of Economic Studies. 2000. No. 67. P. 381–411.Morozov I., Podkolzina E. (2013). Collusion detection in procurement auctions. SSRN Electronic Journal. 10.2139/ssrn.2221809.Porter R. H., Zona J. D. Detection of bid rigging in procurement auctions // The Journal of Political Economy. 1993. No. 101. P. 518–538.Morozov I., Podkolzina E. (2013). Collusion detection in procurement auctions. SSRN Electronic Journal. 10.2139/ssrn.2221809.Porter R. H., Zona J. D. Ohio school milk markets: An analysis of bidding // RAND Journal of Economics. 1999. No. 30. P. 263–288.OECD (2016). Fighting bid rigging in public procurement: Report on implementing the OECD Recommendation. URL: http://www.oecd.org/daf/competition/Fighting-bid-rigging-in-public-procurement-report‑2016.pdf.Symeonidis G. In which industries is collusion more likely? Evidence from the UK // Journal of Industrial Economics. 2003. No. 51. P. 45–74.OECD (2016). Fighting bid rigging in public procurement: Report on implementing the OECD Recommendation. URL: http://www.oecd.org/daf/competition/Fighting-bid-rigging-in-public-procurement-report‑2016.pdf.Pesendorfer M. (2000). A study of collusion in first-price auction. The Review of Economic Studies, no. 67, pp. 381–411.Pesendorfer M. (2000). A study of collusion in first-price auction. The Review of Economic Studies, no. 67, pp. 381–411.Pesendorfer M. (2000). A study of collusion in first-price auction. The Review of Economic Studies, no. 67, pp. 381–411.Pesendorfer M. (2000). A study of collusion in first-price auction. The Review of Economic Studies, no. 67, pp. 381–411.Porter R.H., Zona J.D. (1993). Detection of bid rigging in procurement auctions. The Journal of Political Economy, no. 101, pp. 518–538.Porter R.H., Zona J.D. (1993). Detection of bid rigging in procurement auctions. The Journal of Political Economy, no. 101, pp. 518–538.Porter R.H., Zona J.D. (1993). Detection of bid rigging in procurement auctions. The Journal of Political Economy, no. 101, pp. 518–538.Porter R.H., Zona J.D. (1993). Detection of bid rigging in procurement auctions. The Journal of Political Economy, no. 101, pp. 518–538.Porter R.H., Zona J.D. (1999). Ohio school milk markets: An analysis of bidding. RAND Journal of Economics, no. 30, pp. 263–288.Porter R.H., Zona J.D. (1999). Ohio school milk markets: An analysis of bidding. RAND Journal of Economics, no. 30, pp. 263–288.Porter R.H., Zona J.D. (1999). Ohio school milk markets: An analysis of bidding. RAND Journal of Economics, no. 30, pp. 263–288.Porter R.H., Zona J.D. (1999). Ohio school milk markets: An analysis of bidding. RAND Journal of Economics, no. 30, pp. 263–288.Symeonidis G. (2003). In which industries is collusion more likely? Evidence from the UK. Journal of Industrial Economics, no. 51, pp. 45–74.Symeonidis G. (2003). In which industries is collusion more likely? Evidence from the UK. Journal of Industrial Economics, no. 51, pp. 45–74.Symeonidis G. (2003). In which industries is collusion more likely? Evidence from the UK. Journal of Industrial Economics, no. 51, pp. 45–74.Symeonidis G. (2003). In which industries is collusion more likely? Evidence from the UK. Journal of Industrial Economics, no. 51, pp. 45–74.

The authors declare that there are no conflicts of interest present.