The aim of the paper is to explain why, instead of the necessity of the pension reform and a superior performance of the funded system, Russia experiences quite moderate size of the reform and a slow progress in the transformation to the multipillar pension system. There are several advantages of the Funded system in comparison with PAYG that made it very attractive in the present conditions of ageing of the population. They are a higher return on the contributions, a more active capital market, an increase in savings and investment (although there is a lot of discussion about it).The flaws in the design of Russian pension system inherited from the USSR worsen the performance of pension provision especially in the difficult transition period.The reform was urgent since Russian pension system faces a very serious solvency problem. A demographic crisis, a weak compliance and tax evasion reduce contribution inflow, while loose eligibility rights and a lot of privilege beneficiaries increase the outflow. It was decided to introduce a multipillar system. The introduction of the funded pillar was stipulated by the following advantages: a closer perceived link between contributions and benefits, a higher return on contributions and an undistorted individual labour supply.It is proved in the paper that the reconstruction contributes to the increase in tax compliance and restores the solvency of pension system. It is especially important in the conditions of the scarcity of other sources of financing of the reform (e.g. a state debt, a reduction of the benefits, or an increase of the social tax rate). However, as long as the scale of the reform is determined through the interaction between the self-interest elective government and individuals, the outcome of the reform is not Pareto-efficient. In the paper the interaction is analysed by means of Stackelberg game. The inefficiency of the outcome means that the scale of the reform (or, alternatively, the size of the funded pillar) and the revealed part are not as high as they could be. Thus the reform doesn’t reach one of its main goals – the increase in tax compliance. It seriously undermines the ability of the new pension system to give sufficient provision to the elderly. There are several conditions in Russia which stipulate the inefficient outcome:1. The surplus of revealed part of income caused by the reform is a main source for financing of the reconstruction of the pension system. External sources are limited by the debt-burden budget. An increase of the social tax rates is limited by their high level and its negative effect on the tax compliance. A contraction of eligibility rules as a source has been already exhausted. A further reduction of the benefits is not possible because of their low level. 2. Because of the necessary institutional and economical arrangements and the obligation to provide current pensioners the costs of the reconstruction are increasing with the scale of the reform. Gradually, the corresponding surplus becomes insufficient to cover the costs. 3. The scale of the reform is determined by an elective government. Because of the self-interests to retain the power, the government puts higher weight on current costs of transition than on the long-run gains of the reform. The possible reason can be an inability of the government to persuade citizens to tighten the living conditions today for the better pension provision in the future. The situation is peculiar for Russia, where there is a lack of trust of the population to such kind of promises.4. The government has no other tools (in addition to the very introduction of the funded pillar) to elicit an increase in the compliance, i.e. tools to constraint utility level of agent. The change in power structure in favour of the government can eliminate Pareto-inefficiency. It is worth to note that such a change in power structure can be unfavourable for individuals. Because if the government can fix/constraint their utility levels, it may have a power to reduce the utility further down and made individuals worse off. So even though the outcome is Pareto-inefficient it can be more favourable for individuals to have some inefficiency with the freedom, rather than to be squeezed by the self-interest government in the Pareto-efficient way.As long as the highlighted conditions exist and supplement each other the attempts to accelerate reform are unlikely to be successful. This result corresponds to the way the current pension reform in Russia goes. The reform has a moderate scale and causes just a slight reduction in tax evasion.The paper considers such instruments as a stricter punishment for tax evasion or a more even distribution of the social tax among employees and employers. These instruments reduce the tax evasion and increase the scale of reform. But still, they do not eliminate Pareto-inefficiency. The paper is structured as follows.Chapter 1 describes the origin of the problem. It emphasises the circumstances that made the reform of Russian pension system urgent.Chapter 2 analyses the path of the reform by means of the game-theory approach. The special attention is devoted to the reasons of the moderate size and slow progress of the reconstruction. Chapter 3 introduces additional policy instruments: a punishment for the tax evasion and a more even distribution of the social tax rate between an employer and employees. Afterwards I take a look on the current reforms launched in the end of 2001. Conclusion part summarizes the main findings of the paper.