Factors of successful protection from pressure on business

A model for “Target light clear” target variable

Figure 8. Final model for “Target light clear” variable



The overall AIC of a model has increased, comparing with baseline model: baseline AIC equals 444.96, the AIC of final model is 432.26. In comparison with null model the residual deviance became lower and closer to the number of degrees of freedom, which signals about an improvement in model quality.

From figure8we can see which variables make a significant contribution changing the odds to enter a target class (this target variable assumes that “win” case happened when the entrepreneur was not sent to jail, and the enterprise was lost during the case).

The coefficients of a model could be interpreted as follows:

1) “Safe” activity type. In comparison with Building activity (which is treated to be one of the most dangerous, according to the previous findings [Kazun, 2015]), companies leading their activities in real estate sphere or science sphere have significantly greater chances to be lost or entrepreneur to be sent in jail. Other coefficients being fixed, the odds of companies or entrepreneurs working in real estate industry to avoid damage assumed by “light” threshold increase in 5.21 times or 420% higher than companies or businessmen working in building sphere. The odds of companies working in science research sphere increase in 4.63 times or 360% higher than for companies working in building industry.

2) Other positive factor - membership in association or self-regulatory associations. Other being fixed, the odds of “winning” increases in 1.67 times or 67% higher for organizations or entrepreneurs, having membership in association or self-regulatory organization. This is logical, because the membership gives the entrepreneur wide access to resources that provide protection in future.

3) Among negative factors, the fact that the enterprise has 15 or less employees (“Micro”) size decreases chances to win, comparing to companies having more than 15 employees. This is also may be justified by the fact that “big businesses” may have a greater access to recourses. The odds of “winning” decreases by 0.42 times or 57% lower than the same odds of “winning” for companies with number of employees greater than 15.

4) Capture case significantly decrease chances to save the enterprise or the businessman in comparing with non-capturing cases. Indeed, if the case is categorized like “capture of the enterprise” in the “Business against corruption” application, then the capture most likely already happened, and the damage was received. Other factors being fixed, the odds of “winning” decreases by 0.45 times or 55% lower if the case is characterized as “capture” case.

5) Publications about the case decreases the probability to win. This relates to the fact that not all the enterprises can allow themselves to spread information via mass media and most of the publications contain information about court cases and the fact of conviction of an entrepreneur. Other factors being fixed, the odds of “winning” decreases by 0.44 times or 56% lower if there exist any publications about the case.

6) Finally, if the application has not passed to the final stage, “Council discussion”, the chances to “win” are decreased. This can be confirmed by the fact that the majority of cases are stopped at the “Resolution” stage and do not get supportive measures. Other factors being fixed, the odds of “winning” decreases by 0.44 times or 56% lower if the application has not passed to the final stage of CPP “Business against corruption” procedure.

7)

Figure 9. ANOVA analysis for the “Target light clear” target variable model

According to ANOVA analysis, the same variables have significant coefficients.



A model for “Target light extended” target variable

Figure 10. Final model for “Target light extended” target variable

Figure 11. ANOVA analysis for the “Target light extended” target variable model



The value of AIC criterion has been decreased in comparison with baseline model: baseline model AIC 475.08 to final model AIC 458.81. The residual deviance is lower than the null deviance and closer to the number of degrees of freedom, which signalize the improvement in model quality.

Since “Target light extended” is similar to the “target light clear” variable except including extra cases for positive class. Thus, main significant variables have not changed. Coefficients changed insignificantly. The only difference is that spark stock variable appeared to be significant at 0,09 alpha level. According to the ANOVA analysis, the variable does not increase the deviance significantly.

A model for “Target strong extended” target variable

Unlike two previous variables, this target variable has stronger threshold, so it may have different significant predicts. The case is considered to be a “win” case if the enterprise and the businessman received no damage except the processual costs. The final model was the following:

Figure 12. Final model for “Target strong extended” target variable



Figure 13. ANOVA analysis for the “Target strong extended” target variable model

The AIC of the baseline model was 394.42. The AIC of the final model was 375.7.

The basic findings from this model is the following:

1) “Safe” activity code. The industry enterprise is working in still can be considered as significant variable. Just as in previous models, the reference group in this predictor is “Building” industry. Thus, “real estate” activity sphere still significantly influences chances to “win” in comparing with “Building” sector. The second activity sphere, which significantly increases chances of survival is “rural” activity sphere, which is different from the “science” sphere from the previous variables. Other coefficients being fixed, the odds of companies or entrepreneurs working in real estate industry to avoid damage assumed by “strong” threshold increase in 3.05 times or 200% higher than companies or businessmen working in building industry. The same holds for enterprises working in science industry, their chances to win increase by 3.85 times or 285% higher than companies or businessmen working in building industry.

2) The size of the enterprise still is a matter of great importance. In case with stronger threshold, however, the variable dividing sample into “Micro and Small” and “Medium and Big” enterprises worked out better. So the chances to “win” in terms of this target variable, decreases if the enterprise has under 100 employees and consider to be “Micro” or “Small” enterprise. The odds of “winning” decreases by 0.49 times or 51% lower than the same odds of “winning” for companies with number of employees greater than 100.

3) Membership in association of self-regulatory organization retained the status of significant variable, increasing the chances to survive for members. The odds of “winning” increases be 1.79 times or 79% higher for members of associations or self-regulatory organizations comparing with non-members.

4) The existence of publicly available publications about the case is still a significant variable, implying the decrease of chances to survival. Other factors being fixed, the odds of “winning” decreases by 0.51 times or 49% lower if there exist any publications about the case.

5) As in previous variables, the characteristic of the case is significant. However, in this model the “administrative barriers” variable is significant, which can be explained by the change of positive and negative classes allocation (in previous models “capture” variable was significant variable instead). Other factors being fixed, the odds of “winning” increases by 2.09 times or 109% higher if the case is characterized as “administrative barriers” case.

6) As it can be seen from the figure12, the target with stronger threshold performed better not on the “Business against corruption” “stage” variables, but on the reaction, “BAC” provided the application with. In this model two “reaction” variables were significant: reaction not passed because of the applicants' actions reasons and reaction consultation. Thus, other predictors being fixed, the odds of winning case are decreased by 0.43 times or 57% lower than the application have not passed with reasons, committed by the applicant (loss of interest, not responding to «BAC» anymore etc.). The second one, however, had only 10 observations positive class, so it was dropped.

A model for “Is working” target variable

The baseline model AIC equals 629.88. The final model AIC equals 604.46.

The number of insignificant variables were remained in order to illustrate the results of the analysis.

Figure 14. Final model for “is working” target variable

Figure 15. ANOVA analysis for “is working” target variable model

In terms of significant predictors, “Is working” target variable indicating whether the enterprise survived until the present day is closure to the target variable with stronger threshold.

1) First, similar to “Target strong extended” target variable model, “Business against corruption” target variables performed better. “Reaction not passed by the applicant” variable is significant. Thus, other predictors being fixed, the odds of winning case are decreased by 0.49 times or 51% lower than the application have not passed with reasons, committed by the applicant (loss of interest, not responding to «BAC» anymore etc.).

2) Second, “Safe” type of activity remains an important feature: as with all models, the activity in real estate sphere increases the chances of survival, so as activity in rural activities. Other coefficients being fixed, the odds of companies or entrepreneurs working in real estate industry to survive in long term increase in 6.64 times or 560% higher than companies or businessmen working in building industry. The same holds for rural industry (increase by 3.98 times of 298% higher), manufacturing (increase by 1.76 times of 76% higher) and other categories, which includes combination of enterprises from culture, sport, education, energy, health, hotel, water supply and administrative services industries (increase by 3.25 times of 225% higher).

3) Third, Same for the all variables, membership in association or self-regulatory organization increases chances of survival. The odds of “winning” increases be 3.02 times or 202% higher for members of associations or self-regulatory organizations comparing with non-members.

4) As in previous model for target with strong threshold of survival, the characteristics of the case - administrative barriers play a significant role by increasing chances of survival. The “capture” characteristic of the case is not significant in terms of variance, and close to be treated as significant at 0,09 alpha level. Other factors being fixed, the odds of “winning” increases by 1.82 times or 82% higher if the case is characterized as “administrative barriers” case.

5) The survival of the enterprise in the long run can be increased in case if the entrepreneur is a member of political party. Other factors being fixed, the odds of “winning” increases by 1.98 times or 98% higher if the entrepreneur is a former or current member of a political party (in accordance with application creation period).

6) The last significant predictor is the age of the company. The older the company at the moment of application creation, the higher chances it has to survive in the long run. Thus, other factors being fixed, 1 year increase in the age of enterprise at the moment of application creation increases the odds to win by 1.03 times or 3%.

Using predictors, obtained on the previous step, the final decision tree model was built. The whole model main goal was to generalize the obtained knowledge and model the logic, which can lead to an increase or decrease in the chances of an enterprise to survive in the long run [See R-studio code in Attachment 5for more].

Different combinations of predictors were tried to build a solid model with good predictive characteristics for “is working” target variable.

The best model was achieved with three predictors: OKVED code, membership in association of self-regulatory organization and size (divided by “Micro” and other enterprises).

Figure 16. Decision tree model for “is working” target variable



As the decision tree assessment, train-test performance was made. The data was divided 80% by 20%. The following results for the test sample were obtained:

Table 6

Confusion matrix with 0.5 threshold on decision tree model scores predicted for test sample

Predicted	Reference
	Negative - 0	Positive - 1
Negative - 0	40	22
Positive - 1	8	26

Sensitivity: 0.54.

Specificity: 0.83.

Accuracy: 0,68.

The decision about whether to improve model scores or not depends on the error of what type would be less desirable. As the extension of the analysis, I tried to level Sensitivity and Specificity. So, at the next step, probabilities were calibrated. The new threshold was chosen by F1 score maximization (being a harmonic mean of precision and recall, F1 score can be used as metric for Sensitivity and Specificity calibration).

The value that maximizes F1 score was 0.35 probability. The following results were obtained.

Table 7

Confusion matrix with 0.37 threshold
on decision tree model scores predicted for test sample

Predicted	Reference
	Negative - 0	Positive - 1
Negative - 0	36	17
Positive - 1	12	31

Sensitivity: 0.646

Specificity: 0.75 Accuracy: 0.697

Apart from the decrease in specificity, there is an increase in value of sensitivity. The predicted values become more balanced and the overall accuracy have increased.

3.4 Data analysis conclusion

In conclusion, the data analysis section provided the logic behind data analysis and key results obtained at this step. After data exploration and data preparation steps, four logistic regression models were built in order to find relationships in data, which are constant regardless to the chosen threshold. Despite four different target variables with different thresholds were tested, the stable combination of predictors, which influence odds of “winning” the case was obtained. Among factors, increasing the odds of “winning” are membership in industry association or self-regulatory organization, the characteristic of the case as “imposing administrative barriers”. Membership in political party and the age of the enterprise appeared to increase the odds of survival in long run. “Business against corruption” measure appeared to be significant only during the case, but on the long run do not influence the odds of survival. On the other hand, small enterprises appeared to be more vulnerable and have less odds to “win” regardless of the threshold chosen. Facing the case, characterized as “capture” the odds to “win” a case become lower. The appearance of case publications in mass media resources can be seen as an indicator of probable case loss, since the publications for cases in this sample are often used as announcement of court decision and rarely used for drawing a public attention to the case. Finally, the survival odds differ on the industry the enterprise works in, since OKVED code appeared to be a significant variable in all models. Thus, working in one industry can be significantly increase the odds of the enterprise on survival, comparing to other industries.



Conclusion

Recommendations on business-saving practices

The present paper is devoted to the study of business - government relations problem, namely situations in which the interests of both parties intersect and are contradictory to each other. Such occasions may occur in different types of social, political and economic environment and the outcome of the case is highly dependent the balance of power in business - government relations. The lack of government - business relations equilibrium may provide extra threats to the business owners. Previous findings in literature show, that since the state is considered to be strong, which implies the state capability to imply its' interests over such spheres as economics and politics despite the interests of other parties (in this case - business), the individual actors, which belong to government party, may not follow the interests of the whole governmental organizations [North et al, 2009] and take the advantage of their institutional position in order to achieve their personal interests and extract rent [McChesney, 1987]. In opposite situation, business may impose its' interest regardless the position of official government.

Previous studies of business - government relations in Post-Soviet Russia have shown the balance of power between government and business has never been achieved since the signs of “weak state” and “wild capitalism” may be found in late 1990 - early 2000 and the formation of “strong state” can be found in period since late 2000 to the present day [Yakovlev, 2005; Yakovlev, 2015, Volkov, 2012]. Although it would be wrong to say that it is impossible to create a successful business in Russia (and there exists a whole set of successful enterprises founded in Russia in Post-Soviet times), but both “weak” and “strong” state positions provide extra threats to business owners, lowering business creating incentives in general.

However, state - business relations should not be treated as one-side game, since private organizations can protect their property rights themselves by uniting into independent alliances and imposing costs on government agents [Markus, 2012]. In fact, entrepreneurs may use wide range of property rights protection strategies, which are can be developed and evolved through time [Szakonyi, 2018; Markus 2012; Acemoglu, 2016, and others.]. These strategies, however, are highly context-dependent and varies according to the present state-business relations and the resources available to the businessmen. The strategies themselves, consist of single facts like the fact of establishing interpersonal connection or achieving a membership in organization, which influence the chances of the entrepreneur to get the favorable outcome.

The scope of this study was to identify these facts, or factors, that may influence the chances of minimizing the damage received by the entrepreneur and the enterprise in disputes involving state agents on the opposite side. Stating the main goal of the study in this way, the exploratory research design has been chosen, implying all the possible data should be included on the analysis in order to find the existing relationships.

The public field, however, is usually lack of information of such cases because of being treated by the business community as “illegal” and “controversial” very often. Thus, the “black box” framework for problem modelling was applied. The central idea was to study not the process of dispute itself but consider all preliminary factors and compare them to outcomes, finding the characteristics of both enterprise and entrepreneur, “winners” are likely to have.

As for the database, the Center of public procedures “Business against corruption” application dataset was taken. The idea was to find existing relationships in a set of data, where the estimated concentration of cases of interest is maximized instead of “blind search” of random sample. The information about the enterprise and the entrepreneur was enriched by publicly available materials from Spark Interfax database, mass media publications and Central election commission of the Russian Federation.

After the data collection procedure was performed, four target variables were created, since it was accepted, that the definition of “winning” is conditional. Thus, target variable with “light” threshold was created (the enterprise should not be closed of the entrepreneur should not been sent to jail), target variable with “strong” threshold was created (the entrepreneur or the enterprise should not receive any damage except for processual costs) and different cases, based on the information accessibility were tested. These variables were supposed to show how factors enterprise and the entrepreneur have and the size of the damage received correlate with each other. The final and the most objective target variable was formulated out of current enterprise status, obtained from Spark Interfax database. The target variable in this sense was supposed to reveal the combinations of factors enterprises survived in the long run have.

Since all four target variables were constructed as binary variables, logistic regression was chosen as the main method of the analysis. The separate logistic regression model was built for each target variable. After building the baseline model, two more alternative models were built, using manual variables sorting and R-studio built-in AIC maximization algorithm. Comparing these two approaches, the final models for each target variable were built. Finally, based on previous findings a decision tree model for currently “is working” target variable was built and validated by train - test performance.

As a final product of a research, it is possible to build a set of recommendations, business could follow in order reduce case losses. Despite the chosen threshold, the membership in association or industry self-regulatory organizations increases the chances to reduce costs both in short and long runs. This statement also can be supported by previously formulated family of defense strategies, based on creation of personalized and depersonalized organizations [Yakovlev 2015]. Apparently, membership in such organizations may provide entrepreneur with additional resources, association tend to have and promote entrepreneurs' interest in much more effective way, including appealing to key political actors and other non-personal organizations, such as CPP «Business against corruption».

Speaking of «Business against corruption» instance as something different from other associations and SRO's, its' supportive measures appeared to be significant only in case of target variable with “light” threshold, which implied to sentence to the entrepreneur was provided or the enterprise has not been lost during the case. Thus, “Business against corruption” supportive measures better be evaluated as measures, lowering the possible damage, but not reducing it at all, which is true, because the measures provided by CPP “Business against corruption” is often used as a “last call” measure, when the main damage is received. It is important to admit, that CPP “Business against corruption” measures appeared to be significant in scope of concrete case, meaning that the survival of the enterprise in the long run may not be guaranteed.

Past or present (according to the application creation date) membership in a political party, however, significantly increase the probability of an enterprise to survive. This fact can be seen as a part of direct political strategies and supported by previous studies, describing how entrepreneurs use political field in order to achieve their personal goals [Szakonyi, 2018; Sakaeva, 2012]. The membership in a political party influence can be treated similar to associations membership influence, that help convey the interests of the entrepreneur to key political actors with greater efficiency.

Another factor considered to be significant in defying the chances of "success is the size of the enterprise. Despite the chosen threshold, enterprises with number of employees less than 15 have lower chances to avoid damage, than bigger enterprises with number of employees greater than 100. Indeed, the size of the enterprise can be seen as an indicator of resources the enterprises or the entrepreneurs are able to use in order protect their interests.

The results of data analysis show, that chances to survive is different in different industry spheres, which is also supported by the previous studies [Kazun, 2015].

The probability to “win” varies between the characteristics of the case. Though both characteristics were not stable and change each other from model to mode, it is logically correct to say, that if case can be defined purely as “administrative barriers” the chances for the enterprise to be in “winner” group is higher because administrative barriers do not always imply dealing the critical amount damage to the enterprise and are not always can be seen as a part of an illegal strategy to attack the welfare of an entrepreneur. On the other hand, “Capture” cases, which often implies illegal raider attack on enterprise property, lower the chances to “survive”, because at the time when the case can be considered as “capture” the significant amount of the damaged have already been received.

Finally, case publications should be treated as lowering the chances of “winning” the case predictor. It is true for the cases from the sample, that publications in the mass media are rarely aimed at publicizing the information about the case or drawing public attention to the conditions, which considered to be illegal from the point of victims' defense. In most cases, publications contain information on a final court decision or information on preliminary charges.

Overall, the present paper provides the framework for further data gathering and data analysis in hard-to-reach research field. Despite all the work done, there is still a place for improvement. In order to improve the conclusions both in qualitative and quantitative way, the greater number of preliminary factors may be analyzed. Moreover, it seems like providing cases classification first and then analyzing each case category separately may provide deeper insight into the problem and achieve interesting insights into each category.



REFERENCES

1.Szakonyi, David. 2018. “Businesspeople in Elected office: Identifying Private Benefits from Firm-Level Returns”. American Political Science Review, 112 (2): 322-338.

2.Markus, Stanislav. 2012. “Secure Property As A Bottom-Up Process: Firms, Stakeholders and Predators in Weak States”. World Politics, 64(2), 242 - 277.

3.Sakaeva, Maria. 2012. “Parliament as a «Window of opportunities»: a study on the pursuit of business interests by entrepreneurs with mandates”. Economic Sociology, 3, 96-122 (in Russian).

4.Acemoglu D. et al. 2016. “The value of connections in turbulent times: Evidence from the United States”. Journal of Financial Economics, 121, 368-391.

5.Amore, Daniele and Bennedsen, Morten. 2013 “The value of local political connections in a low-corruption environment”. Journal of Financial Economics, 110(2), 387-402.

6.Frye, Timothy. 2004. “Credible commitment and Property Rights: Evidence from Russia”. American Political science review. 98 (3), 453 - 466.

7.North, Douglass C., and Robert Paul Thomas. 1973. The Rise of the Western World: A

New Economic History. New York: Cambridge University Press.

8.Frye, Timothy and Shleifer, Andrei. 1997. “The invisible hand and the grabbing hand.” American Economic Review Papers and Proceedings 87 (May): 554-559.

9.Stiglitz, Joseph. 1999. “Whither Reform.” Annual Bank Conference on Development Economics. Washington, D.C. World Bank.

10.Radaev, Vadim. 2001. “New institutional approach: construction of a research scheme”. Economic sociology3, p. 5-26.

11.Olson, Mancur. 1993. “Dictatorship, Democracy and Development”. American Political Science Review 87, 3. 567 - 576.

12.Tilly C. 1985. War making and State Making as Organized Crime. Bringing the state back in. Cambridge: Cambridge University Press.

13.North D., Wallis J., Weingast B. 2009. Violence and Social Orders. A conceptual framework for Interpreting Recorded Human History.

14.Razo, Armando. 2015. Political Economy of Crony Capitalism: Credible Commitments without Democratic Institutions.

15.Haber, Stephen. 2002. Crony Capitalism and Economic Growth in Latin America: Theory and evidence. Hoover Institution Press.

16.Libecap, Gary. 2018. Douglass C. North: Transaction Costs, Property Rights and Economic Outcomes. Hoover institution economic working papers. 1-21.

17.Civil Code of the Russian Federation. Article 2, point 1.

18.Fligstein, Neil. 1996. Markets as Politics: a Political-Cultural Approach to Market Institutions. American Sociological Review. Vol. 61. N 4. p. 656-673.

19.Licht Amir, Siegel Jordan. 2005. The social dimensions of entrepreneurship. Oxford Handbook of Entrepreneurship. Oxford: Oxford University Press, p. 1 - 44

20.Salter, Malcolm. 2014. Crony Capitalism, American Style: What are we talking about here? Edmond J. Safra Research Lab Working Papers, No. 50 Harvard University

21.Sharafutdinova, Gulnaz. 2011. Political consequences of Crony Capitalism inside Russia. University of Notre Dame Press.

22.Kang, David. 2002. Crony Capitalism: corruption and development in south Korea and the Philippines. Cambridge University press.

23.Gel'man, Vladimir. 2000. The Dictatorship of Law in Russia: neither dictatorship nor Rule of Law. European University in St. Petersburg. PONARS Policy Memo 146.

24.Barros, Robert. 2003. Dictatorship and the Rule of Law: Rules and Military Power in Pinochet's Chile. Democracy and the Rule of Law. Cambridge University Press, pp.188-220.

25.Ivanovic, Zeljko. 2017. The political Economy of Crony Capitalism: A Case Study of the Collapse of the Largest Croatian Conglomerate. Croatian Political Science Review, Vol. 54, No. 4, pp. 40-60

26.Yakovlev A. À. Evolution of Business State Interaction in Russia: from State Capture Y14 to Business Capture? [Text] : working paper WP1/2005/02 / A. A. Yakovlev: State University -- Higher School of Economics. -- Moscow: Publ. house of SU HSE, 2005. -- 40 p

27.Volkov, Vadim. 2012. Violent entrepreneurship, XXI century: economic sociology analysis. St. Petersburg: European University Press. P.352

28.Sameeksha, Desai. Acs, Zoltan. Weitzel, Utz. 2010. A model of destructive entrepreneurship. Working paper N 2010/34. World Institute for Development Economics Research. P.16.

29.Yakovlev, Andrei. Sobolev, Anton. Kazun, Anton. 2014. Means of production versus means of coercion: can Russian business limit the violence of a predatory state? Post-Soviet Affairs, 171-194

30.McChesney S. 1987. Fred Rent Extraction and Rent Creation in the Economic Theory of Regulation. The Journal of Legal Studies, pp. 101-118

31.Djankov, Simeon. 2015. Russia's Economy under Putin: From Crony Capitalism to State Capitalism. Peterson Institute for International Economics. PB15-18. P.1-8.

32.Kosals, Leonid. 2006. Interim outcome of the Russian transition: clan capitalism. Discussion Paper # 610. Kyoto University.

33.An interview with founder of “DO-DO Pizza” after drugs distribution accusation. [Internet-source]. Link: youtube.com/watch?v=ZPhy9WGPLMU&t=15s

34.Business Ombudsmen's annual report to the president. [Internet-source]. Link to all reports: doklad.ombudsmanbiz.ru.

35.Violent pressure defense news articles. [Internet-sources].

a.Putin instructed to create a digital platform for appealing business to pressure. 27.02.2019. Link: tass.ru/ekonomika/6163536.

b.Bastrykin ordered to create a hotline SK for reporting pressure on business. 22.02.2019. Link: meduza.io/news/2019/02/22/bastrykin-velel-sozdat-goryachuyu-liniyu-sk-dlya-soobscheniy-o-davlenii-na-biznes.

c.Vladimir Putin stood up for business again. 20.02.2019. Link: vedomosti.ru/business/articles/2019/02/20/794715-putin-biznes

36.Kadushin, Charles. 2011. Understanding Social Networks: Theories, Concepts and Findings. Oxford University press.

37.Keller, Franziska. 2015. Networks of power. A social network analysis of the Chinese Communist elite, 1982-2012. New-York University.

38.Kang, David. 2002. Transaction costs and crony capitalism in East Asia. Working paper ¹02-11. Tuck school of Business at Dartmouth.

39.Marques II I., Levina I. A., Kazun A., Yakovlev A. A. Calling the Cavalry: Firm-level Investment in the Face of Decentralized Expropriation // Journal of Institutional Economics. 2020

40.Levina, Irina; Kisunko, Gregory; Marques, Israel I.; Yakovlev, Andrei. 2016. Uncertainty as a factor in investment decisions : the case of the Russian Federation's regions (English). Policy Research working paper; no. WPS 7806. Washington, D.C. : World Bank Group.

41.Kazun A. Violent Corporate Raiding in Russia: Preconditions and Protective Factors // Demokratizatsiya: The Journal of Post-Soviet Democratization. 2015. Vol. 23. No. 4. P. 459-484.

42.Decree of the President of the Russian Federation of May 13, 2000 No. 849 (redaction on December 13, 2018) "On the Plenipotentiary Representative of the President of the Russian Federation in the Federal District". Web-source. Link: http://www.consultant.ru/document/cons_doc_LAW_23329/. Date of access: 2020-05-20.

43.“On the adoption and enforcement of the All-Russian Classifier of Economic Activities (OKVED2) OK 029-2014 (NACE Rev. 2) and the All-Russian Classifier of Products by Economic Type activities (OKPD2) OK 034-2014 (CPA 2008)" “Rosstandart” order of January 31, 2014 No. 14-st (redaction on October 16, 2018). Web-source. Link: http://www.consultant.ru/document/cons_doc_LAW_163268/. Date of access: 2020-05-20.

44.Spark risk indicator. Web-source. Link: http://www.spark-interfax.ru/ru/features/indexes. Date of access: 2020-05-20.



Attachment 1

The table of variables included in final dataset (including variables created during the analysis)

Ïîäîáíûå äîêóìåíòû

• Ìîäåëü ðàáîòû êîìïàíèè ÎÎÎ "Àêêîðä"

  Õàðàêòåðèñòèêà ïðîãðàììíîé ñðåäû Business Studio 3.6. Äåìîíñòðàöèîííàÿ áàçà íà ïðèìåðå ïîêóïêè è äîñòàâêè îôèñíîé ìåáåëè. Ñîäåðæàíèå, âðåìåííàÿ è ëîãè÷åñêàÿ î÷åðåäíîñòü îïåðàöèé áèçíåñ-ïðîöåññà êîìïàíèè "Àêêîðä" ã. Ðîñòîâ-íà-Äîíó; îáëàñòè ìîäåëèðîâàíèÿ.
  
  êóðñîâàÿ ðàáîòà [1,3 M], äîáàâëåí 01.06.2014
  

• Ïðèìåíåíèå ÏÏÏ Business Studio äëÿ èìèòàöèîííîãî ìîäåëèðîâàíèÿ ÔÑÀ áèçíåñ-ïðîöåññà ïðåäïðèÿòèÿ

  Ïðîöåññ èíòåãðàöèè òåõíè÷åñêèõ è ïðîãðàììíûõ ñðåäñòâ âî âñå àñïåêòû äåÿòåëüíîñòè ïðåäïðèÿòèÿ. Ãðóïïà êîíòðîëëèíãà è åå çàäà÷è. Îïåðàöèîííî-îðèåíòèðîâàííûé ðàñ÷åò ñåáåñòîèìîñòè ïðîäóêòà (óñëóãè). Îïðåäåëåíèå âðåìåíè âûïîëíåíèÿ è ñòîèìîñòè ïðîöåññà.
  
  ðåôåðàò [547,5 K], äîáàâëåí 14.09.2010
  

• Ðèñê è äîõîäíîñòü

  Ìîäåëü îöåíêè äîëãîñðî÷íûõ àêòèâîâ (Capital Asset Pricing Model, ÑÀÐÌ). Îöåíêà äîõîäíîñòè è ðèñêà íà îñíîâå èñòîðè÷åñêèõ äàííûõ. Âûáîð îïòèìàëüíîãî ïîðòôåëÿ èç ðèñêîâàííûõ àêòèâîâ. Ðèñê è íåîïðåäåëåííîñòü äåíåæíûõ ïîòîêîâ. Ðàñ÷åò áåòà-êîýôôèöèåíòà.
  
  ïðåçåíòàöèÿ [104,1 K], äîáàâëåí 30.07.2013
  

• Analytical assessment of favorable ratio between thickness of knife and grate of auger meat chopper aimed to improve minced meat quality and reduce cutting pair wear rate

  Mathematical model of the grinding grating bending process under the action of a meat product load parabolically decreasing along the radius. Determination of the deflection of a knife blade under the action of a parabolic load of the food medium.
  
  ñòàòüÿ [1,3 M], äîáàâëåí 20.10.2022
  

• Expanding Finnish fertilizer and soils market

  Mission, aims and potential of company. Analysis of the opportunities and threats of international business. Description of the factors that characterize the business opportunities in Finland. The business plan of the penetration to market of Finland.
  
  êóðñîâàÿ ðàáîòà [128,3 K], äîáàâëåí 04.06.2013
  

• The business cycles as a form of economic development

  Definition and stages of business cycles, their causes and the characteristic of kinds. Types and a continuity of business cycles. Kondratyev's wave. A role of cycles in stabilization of a policy of the state. Great depression as an economic crisis.
  
  ðåôåðàò [130,5 K], äîáàâëåí 20.03.2011
  

• Procedure of preparation business-plan

  Business plans are an important test of clarity of thinking and clarity of the business. Reasons for writing a business plan. Market trends and the market niche for product. Business concept, market analysis. Company organization, financial plan.
  
  ðåôåðàò [59,4 K], äîáàâëåí 15.09.2012
  

• Imperatives of international cooperation for economic diplomacy of Ukraine

  Support of business entities on the part of specialized agencies of the state on world markets. Interconnection of economic diplomacy of Ukraine in international cooperation with influence on the results of foreign economic activity of the country.
  
  ñòàòüÿ [30,1 K], äîáàâëåí 19.09.2017
  

• Change and Continuity in Contemporary Business

  Impact of globalization on the way organizations conduct their businesses overseas, in the light of increased outsourcing. The strategies adopted by General Electric. Offshore Outsourcing Business Models. Factors for affect the success of the outsourcing.
  
  ðåôåðàò [32,3 K], äîáàâëåí 13.10.2011
  

• Automation of business processes in hospitals

  Technical and economic characteristics of medical institutions. Development of an automation project. Justification of the methods of calculating cost-effectiveness. General information about health and organization safety. Providing electrical safety.
  
  äèïëîìíàÿ ðàáîòà [3,7 M], äîáàâëåí 14.05.2014
  

• Modern Means of Business Communication

  Business as a combination of types of activities: production, distribution and sale, obtaining economic profit. Basic types and functions of banks. The principle of equilibrium prices and financial management. The use of accounting in the organization.
  
  êîíòðîëüíàÿ ðàáîòà [17,8 K], äîáàâëåí 31.01.2011
  

• Socio-economic development of Ukraine

  Prospects for reformation of economic and legal mechanisms of subsoil use in Ukraine. Application of cyclically oriented forecasting: modern approaches to business management. Preconditions and perspectives of Ukrainian energy market development.
  
  ñòàòüÿ [770,0 K], äîáàâëåí 26.05.2015
  

• Growth theory

  The concept of economic growth and development. Growth factors: extensive, intensive, the growth of the educational and professional level of personnel, improve the management of production. The factors of production: labor, capital and technology.
  
  ïðåçåíòàöèÿ [2,3 M], äîáàâëåí 21.07.2013
  

• Risk in an entrepreneurial activity

  Entrepreneurial risk: the origins and essence. The classification of business risk. Economic characteristic of entrepreneurial risks an example of joint-stock company "Kazakhtelecom". The basic ways of the risks reduction. Methods for reducing the risks.
  
  êóðñîâàÿ ðàáîòà [374,8 K], äîáàâëåí 07.05.2013
  

• Business the plan as strategic object of the company

  The essence, structure, îbjectives and functions of business plan. The process’s essence of the bank’s business plan realization. Sequential decision and early implementation stages of projects. Widely spread mistakes and ways for their improvement.
  
  êóðñîâàÿ ðàáîòà [67,0 K], äîáàâëåí 18.12.2011
  

• Drug trafficking - a global challenge of the XXI century

  Description situation of the drugs in the world. Factors and tendencies of development of drugs business. Analysis kinds of drugs, their stages of manufacture and territory of sale. Interrelation of drugs business with other global problems of mankind.
  
  êóðñîâàÿ ðàáîòà [38,9 K], äîáàâëåí 13.09.2010
  

• Entrepreneurial project Individual Project

  Executive summary. Progect objectives. Keys to success. Progect opportunity. The analysis. Market segmentation. Competitors and competitive advantages. Target market segment strategy. Market trends and growth. The proposition. The business model.
  
  áèçíåñ-ïëàí [2,0 M], äîáàâëåí 20.09.2008
  

• Presentation to Creditors

  Financial position of the "BTA Bank", prospects, business strategy, management plans and objectives. Forward-looking statements, risks, uncertainties and other factors that may cause actual results of operations; strategy and business environment.
  
  ïðåçåíòàöèÿ [510,7 K], äîáàâëåí 17.02.2013
  

• Comprehensive analysis of the economic activity of the company "Turboatom"

  Directions of activity of enterprise. The organizational structure of the management. Valuation of fixed and current assets. Analysis of the structure of costs and business income. Proposals to improve the financial and economic situation of the company.
  
  êóðñîâàÿ ðàáîòà [1,3 M], äîáàâëåí 29.10.2014
  

• Transformation of business: shift to the subscription economy

  Software as a Service, a form of cloud computing service model of software users. SaaS subscription model: key features, market drivers and constraints. Impact of SaaS subscription services business in the economy and society in Russia and abroad.
  
  äèïëîìíàÿ ðàáîòà [483,8 K], äîáàâëåí 23.10.2016
  

• ãëàâíàÿ
• ðóáðèêè
• ïî àëôàâèòó
• âåðíóòüñÿ â íà÷àëî ñòðàíèöû
• âåðíóòüñÿ ê íà÷àëó òåêñòà
• âåðíóòüñÿ ê ïîäîáíûì ðàáîòàì

Variable name	Variable format	Variable source	Interpretation
DB_ID	Integer		Database id.
BAC_ID	Integer	Initial «BAC» data	«Business against corruption» base id.
Applicant	String	Initial «BAC» data	Full name of the applicant.
Applicant position	String	Spark data	Position of the applicant in accordance to the indicated company.
Victim	String	Initial «BAC» data, Mass media publications	Full name of an entrepreneur considered to be a victim (in accordance with the application).
Victim position	String	Initial «BAC» data, Mass media publications, Spark data	Position of an entrepreneur considered to be a victim (in accordance with the application).
Company Name	String	Initial «BAC» data	The name of a company considered to be damaged (in accordance with the application).
Company reg num	Integer	Spark data	Company registration number - used for data parsing.
INN	Integer	Spark data	Individual tax number - used for data parsing.
Region name spark	String	Spark data	The name of the region company registered in.
Region code spark	Integer	Spark data	The code of the region company registered in.
Federal districts	String	Originally based on “region code spark” variable, grouped by Presidents' decree information [Decree of the President of the Russian Federation, 2000].	Grouping region code by federal districts: Caucasus, Central, Far East, Moscow, Moscow region, North West, Saint Petersburg, Siberian, South, Urals, Volga.
Largest federal districts	String	Originally based on Spark data, grouped by Presidents' decree information [Decree of the President of the Russian Federation, 2000].	Grouping of Federal district variable in order to get bigger categories by the closest regions: Saint Petersburg + North West = North West, South + Caucasus = South, Far East + Siberia = Far Siberian, Others see “Federal Districts”.
Legal form short	String	Spark base	Legal form of the enterprise, short form.
Legal form full	String	Spark base	Legal form of the enterprise, full form.
Property form	String	Spark data	Property form of the enterprise, full form.
Public administration code	String	Spark data	All-Russian classifier of public authorities and administration, assigned to the enterprises to be statistically revised.
Spark web site	Integer	Spark data	Whether enterprise has a website, according to Spark.
Registration date	Date	Spark data	Enterprise registration date.
Company age until 2020	Integer	Spark data	Number of years passed from the date of registration to 2020-05-01, not including closing date.
Company age including liquidation	Integer	Spark data	Number of years passed from the date of registration to 2020-05-01, including closing date.
Liquidation date	Date	Spark data	Liquidation date (if any).
Age until application date	Integer	Spark data, initial «BAC» data	Number of years passed from the month of registration to the month of application creation.
OKVED code	Integer	Spark data	OKVED industry affiliation code.
OKVED code meaning	String	Originally by Spark, grouped by [“Rosstandart” order of January 31, 2014]	OKVED industry affiliation code grouping description.
Macro OKVED code	String	Originally by Spark, grouped by [“Rosstandart” order of January 31, 2014]	OKVED industry affiliation macro categories grouping.
Macro OKVED code group	String	Originally by Spark, grouped by [“Rosstandart” order of January 31, 2014]	Macro OKVED macro categories grouping, every category with < 15 observations moved to `other categories.
Spark stock ticket	Integer	Spark data	Whether the company's shares are publicly traded.
Spark risk indicator	String	Spark data	Spark risk indicator[Spark risk indicator], for working enterprise only.
Spark credit limit	Integer	Spark data	Spark credit limit indicator [Spark risk indicator], for working enterprise only.
N employees upperbound	Integer	Spark data	Upper bound of spark number of employees' interval, if any.
N employees added	Integer	Spark data	Upper bound of spark number of employees' interval, every missing value replaced by 5.
Category by size with missing data	String	Spark data, categorized by [On Small and Medium-Sized Enterprises Development in Russian Federation, 2007] Federal law number of employee's criterion only	Categorization of the enterprise size based on “n employees upper bound” variable, including missing data. The rules: if ¹ of employees <= 15, “Micro”, If ¹ of employees >= 16 and <= 100, “Small”, If ¹ of employees >= 101 and <= 250, “Medium”, else “Big”.
Category by size Micro and else	String	Spark data, categorized by [On Small and Medium-Sized Enterprises Development in Russian Federation, 2007] Federal law number of employee's criterion only	Grouping of “Category by size with missing data» variable by “Micro” sized enterprises (the biggest category) and enterprises of other sizes (“Small”, “Medium” and “Big” enterprises).
Category by size Micro-Small enterprises and other enterprises	String	Spark data, categorized by [On Small and Medium-Sized Enterprises Development in Russian Federation, 2007] Federal law number of employee's criterion only	Grouping of “Category by size with missing data» variable by “Micro” and “Small” sized enterprises and enterprises of other sizes (“Medium” and “Big” enterprises).
Authorized capital	Integer	Spark data	Authorized capital of the enterprise, if any.
Authorized capital group	String	Spark data	Grouping of “Authorized capital” variable on three equally sized groups: Enterprises with authorized capital “Under 10k”, from 10k to 210k “under 210k” and enterprises with authorized capital “over 210k”.
Pure profit 2010	Integer	Spark data	Pure profit, given year.
Pure profit 2011
Pure profit 2012
Pure profit 2013
Pure profit 2014
Pure profit 2015
Pure profit 2016
Pure profit 2017
Pure profit 2018
Enterprise status	String	Spark data	Status of the enterprise in May 2020 (is working, liquidated, bankruptcy, etc.).
Is working	Integer	Spark data	Flag whether the enterprise is working in May 2020.
Target light clear	Integer	Spark data (court cases), mass media publications	Target variable with “light” threshold: in order to “win” (1) enterprise should not be lost and the entrepreneur should not be sent to jail. “Clear” - only well-defined cases included. Case is considered to be “clear” when the documentû (publications or court cases) about the outcome of the case were found and contained an unambiguous conclusion about the results of the case.
Target light extended	Integer	Spark data (court cases), mass media publications	Target variable with “light” threshold: in order to “win” (1) enterprise should not be lost and the entrepreneur should not be sent to jail. “Extended” - aside from “clear” cases being included, extra group of “positive” cases included in the analysis, when the outcome of the case was not found to be documented, but the enterprise is working, and the “victim” assigned to the managerial position.
Target strong extended	Integer	Spark data (court cases), mass media publications	Target with “strong” threshold: in order to “win” (1) both enterprise and the entrepreneur should not receive any damage except for processual costs. “Extended” - aside from “clear” cases being included, extra group of “positive” cases included in the analysis, when the outcome of the case was not found to be documented, but the enterprise is working, and the “victim” assigned to the managerial position.
Administrative position	Integer	Spark data, Publications, CIC	Whether the entrepreneur held a position in an administrative institution before the application.
Administrative connections	Integer	Spark data, Publications	Whether free access sources have information on the entrepreneur's relations with persons holding administrative positions.
In political party	Integer	Publications, CIC, Spark data	Where free access sources have information on entrepreneur's membership in political party.
Political party	String	Publications, CIC, Spark data	The name of political party, if any.
In association or SRO	Integer	Publications, Spark data	Whether the entrepreneur or the organization is a member of an associations and self-regulatory organization.
Association SRO name	String	Publications, Spark data	Association or SRO name, if any.
Case publications	Integer	Publications	If any mass media resources have publications about the case.
Application topic	String	«BAC» data	Description of the case of the application.
Criminal prosecution	Integer	«BAC» data	Whether described case of the application includes criminal prosecution of the entrepreneur.
Capture	Integer	«BAC» data	Whether described case of the application includes capture of the enterprise.
Corruption	Integer	«BAC» data	Whether described case of the application includes signs of the corruption.
Barriers	Integer	«BAC» data	Whether described case of the application includes signs of administrative barriers.
Application date	Date	«BAC» data	Application date.
Application year	Integer	«BAC» data	Application year.
Have court case	Integer	«BAC» data	The associated court case has been found.
Is guilty	Integer	«BAC» data	Whether the entrepreneur was recognized as guilty in associated court case.
Reviewed by BAC	Integer	«BAC» data	Whether the application passed all preliminary stages of «BAC» applications review procedure.
Max BAC stage	Integer	«BAC» data	Maximal stage of «BAC» applications review procedure the application ever passed to.
Max BAC stage, grouped	Integer	«BAC» data	Grouping of “Max BAC stage”, stages from 1 to 3 are grouped as “Information collection” category, stages from 4 to 5 are grouped as “Resolution” category, stages from 6 to 7 grouped as “Council discussion” category.
Supported by BAC public council	Integer	«BAC» data	Whether the application was supported by «BAC» public council.