Vitalijs JASCISENS soutiendra sa thèse de doctorat en Sciences économiques le mercredi 04/07/2018 à 14h00, Salle MF 323 (Manufacture des Tabacs), sur le sujet « Three Essays on Tax Evasion and Public Procurement ».
Directeur de thèse: Stéphane STRAUB
Le jury est composé de :
- Professeur Stéphane Saussier, Professeur, Université Paris 1 Panthéon Sorbonne
- Professeur Giancarlo Spagnolo, Université de Rome Tor Vergata et Stockholm Institute for Transition Economics
- Professeur Matteo Bobba, Toulouse School of Economics & Université Toulouse 1 Capitole
- Professeur Stéphane Straub, Toulouse School of Economics & Université of Toulouse 1 Capitole
Résumé(en anglais):
This thesis consists of three independent papers on tax evasion and public procurement. The common denominator across them is the use of high-quality administrative data which in combination with econometrics of program evaluation allows to uncover previously understudied or overlooked effects in areas of tax evasion and public procurement.
In the first chapter (co-authored with Anna Zasova) we study whether in an environment characterised by high payroll tax evasion the government should tie social security benefits to reported earnings. To answer this question, we use administrative data from Latvia covering the entire working population over a 15-year period from 1996 to 2010 and study generous parental benefits, which depend on the reported wage in the time period before the childbirth. Our analysis delivers three principal results. First, we observe a sharp increase in the wage during the time period taken into account to calculate parental benefits. Depending on the specification, we conclude that during this period the wage on average increases by 5.4%-7.5%. Second, obtained effects are highly heterogeneous. We find that the wage growth is much higher in small firms, where it is presumably easier to sustain collusion between employees and employers. Finally, we demonstrate that legalisation of wages is temporary and lasts only until the end of the period taken into account to calculate parental benefits. Hence, looking from the tax policy perspective the result of tying parental benefits to reported wages is the net loss to government’s finances. Our back-of-the-envelope calculations suggest that the net fiscal loss coming from foregone tax revenues in 2005 amounted to 0.5%-0.6% of GDP.
In the second chapter, I first study how discretionary powers given to government officials affect outcomes in the market for public procurement. Next, I analyse whether market forces can undo perverse effects of bureaucratic discretion. The focus of this chapter is on one of the most crucial features of the auction design - the reserve price, which is the maximum per unit price the government is willing to pay in a given procurement auction. In this paper, I use data on drug procurement auctions in Russia. Obtained results show that reserve prices are at least 8% too high as compared to the optimum: buyers could lower them by this amount, enjoying a one to one decrease in the final price with no increase in the probability of trade not happening. The second set of results indicates that entry can solve the problem of inflated reserve prices: an additional bidder causes prices to decrease by around 8% to 9%. This effect is highly nonlinear: having more than one bidder versus one bidder causes prices to fall by about 15% to 18%. Therefore my results suggest that even in an environment where reserve prices are inflated the legislator can undo these effects by ensuring that there are at least 2 participants in a procurement auction.
Finally, in the third chapter of this thesis (co-authored with Anna Zasova), we explore whether in the presence of informal labour markets governments can implement a substantial increase in the tax rate. Additionally, we also study whether there is a path - dependency of bad policies. More precisely, we try to understand whether after implementing bad policies the government can restore the old equilibria by merely putting old rules in place. To answer these questions, we exploit a unique natural experiment, which reduced the retirement benefits of working pensioners by 70% and those of non-working pensioners by 10%. Our analysis shows that this led to at least six percentage points higher exit probability of working pensioners out of the private sector employment with respect to the control group as compared to the same difference in the public sector. We interpret the excess difference as evidence of the reform inducing labour market transitions between the private sector and informality. Obtained effects are highly heterogeneous with respect to the firm size. We find that switches to informality in firms with less than five employees are about 24 percentage points more likely as compared to firms with more than fifty employees. To understand whether the government can restore the old equilibria by merely putting old rules in place we study what happens once the reform is abolished. Our results show that conditional on exit people exiting the private sector are at least ten percentage points more likely to return to the same employer after the reform is abolished as com- pared to people working in the public sector. Thus an individual who had responded to the reform by switching to informality was more likely to re-appear in the labour market as compared to an individual who had actually quit employment. Hence our results on returns provide additional support to the interpretation of the obtained effect on exits as shifts to informality.
