Fiscal policy and economic growth in the short and long run: a literature review

Poylov Nikita Aleksandrovich ORCID: 0009-0001-2401-5560 Postgraduate student, Department of Economics, European University at St. Petersburg 191187, Russia, Saint Petersburg region, Saint Petersburg, Gagarinskaya str., 6/1, office A npoylov@eu.spb.ru

1. Introduction

Fiscal policy (FP) is one of the main policies of the state, through which the state achieves goals in the economy (stable economic growth, reduction of unemployment, reduction of inequality, etc.). FP affects the economy through changes in the amount of expenditures and revenues of the state. Through fiscal policy, the state smoothes fluctuations in the economy, and the criterion of its effectiveness is the property of countercyclical. In neoclassical models, it is believed that the AF can only affect the transition period (when the economy is moving towards equilibrium), but cannot affect the equilibrium. However, in endogenous growth models, AF can have an impact on economic growth in the long term. To date, the prevailing view among economists is that AF can have a long-term impact on economic growth. Nevertheless, the opinion on the channels of influence of the AF on long-term economic growth is not so clear. The optimal AF can vary significantly from country to country, depending on the conditions. However, the bulk of the literature is devoted to the impact of AF in developed countries. Recommendations based on work on developed countries may not be suitable for developing countries, which leads either to the use of suboptimal AF, or to the actions of politicians in developing countries "blindly". In addition, there are significant problems in the analysis of AF based on data from developed countries.

This paper analyzes the literature on the impact of AF on the economy in the short term and in the long term, presents a broad view of all the main channels of AF influence on the economy highlighted in the literature. The paper also attempts to compare the effects of AF for poor countries and for rich ones. First, the long-term impact of fiscal policy is considered, grouped into an analysis of unconditionally productive expenditures, conditionally productive expenditures, taxes, public debt and the size of the state. Then the short-term impact of AF is considered in terms of its cyclicity, limitations and effectiveness. The main results are summarized in the conclusion.

2. Long-term effects of AF

In the Solow model ^[1], AF has no effect on the long-term equilibrium pathway. In the model of intersecting generations (OLG) ^[2], economic growth is also set exogenously, that is, the AF cannot influence long-term economic growth. However, later works were presented, thanks to which endogenous growth models were developed and used ^[3-6], and in which AF can have a significant impact on equilibrium growth rates by influencing productivity and innovation. In the future, the Solow and OLG models were modified taking into account the fact that economic growth can be influenced by AF ^[7-12]. Empirical studies mostly confirm that AF can have a long-term impact on economic growth, for example ^[13-17].

This section analyzes the issue of the long-term impact of AF on economic growth, depending on the focus of the results of the work. This section is divided as follows: unconditionally productive expenses, conditionally productive expenses, taxes, public debt and the size of the state.

2.1. Certainly productive expenses

Government spending can be productive or unproductive. Nichkamp and Put ^[18] conduct a meta-analysis of 93 articles on the impact of AF on economic growth, productive expenses include mainly: education costs, health care costs, infrastructure costs, research and development (R&D) costs. Moreover, the costs of education and health care turn out to be certainly productive, while the rest of the costs either have some condition to be productive, or there are conflicting opinions in the literature.

For example, Barro^[5] writes that productive expenditures such as education, research, and transportation have a positive impact on growth through increased productivity and technological capabilities, while unproductive expenditures such as subsidies, social spending, and defense spending have a negative impact on economic growth.

An extensive amount of literature is devoted to the impact of human capital on economic growth. Theoretical works ^[4,19] indicate that representative individuals decide how much available time to spend on production and on the formation of human capital (self-development), while the formation of human capital is the result of only the investment of human capital. In some theoretical works ^[3,5,20], it is assumed that physical capital can also influence the development of human capital, since human capital can accumulate through production (learning by doing). An empirical article by Mankiw and co-authors ^[14] states that with the growth of human capital, income increases, which also causes an increase in human capital and, as a result, productivity is much higher.

Krebs ^[21] uses an endogenous growth model with incomplete human capital markets and possible risks for investments in human capital. Theoretical analysis has shown that reducing the risk of human capital reduces investments in physical capital, but increases in human capital, and the return on human capital is higher, since it is not risk-free, that is, redistribution towards human capital leads to an increase in output. However, agents avoid risks, which inclines them to a less risky asset (that is, to physical capital due to the ability to diversify risks), even if the expected return on a less risky asset is lower. However, if the government equalizes the risks, then agents will choose the asset whose expected return is higher, that is, a policy aimed at reducing risks to income from labor has a positive effect on growth.

Spending on education can stimulate economic growth through the accumulation of human capital. In theory, Zagler and Durnecker ^[22] note that the effect of education echoes through generations, since educated parents contribute to education for their descendants, and the cost of obtaining another, more in-demand, qualification affects the growth of human capital, since obtaining more in-demand skills or updating outdated skills increases labor productivity.

An interesting view of education is presented in a theoretical article ^[7], which concludes in an OLG model with a socialization parameter. Education has a positive effect on growth through social cohesion, that is, education standardizes people's understanding, erasing religious and cultural differences, as a result, communication is simplified. Also, cohesion increases political stability and, as a result, economic stability.

Barro ^[13] uses a cross-sectional model for 118 countries (1960-1985) and establishes that the higher the human capital, the lower the population growth rate. School enrollment has a positive relationship with the ratio of investment to GDP, and a negative relationship with fertility, that is, higher primary school enrollment has a positive effect on growth rates ^[13]. It is also noted that an increase in the birth rate may reduce the growth rate, as consumption increases and savings decrease due to an increase in the proportion of children among the population. The empirical work ^[23] for 88 countries (1960-1996) also uses a cross-sectional model, but the authors use Bayesian estimation, which allows us to include many regressors. This work also notes the positive impact of primary education on economic growth rates (however, the model has a negative coefficient for higher education coverage, which the authors ignore). Education spending has a stronger impact on growth if education in the country is initially high, Benos concludes ^[24] using panel data on Eurozone countries and a fixed-effects model (1990-2006).

Spending on health care, as well as spending on education, stimulates economic growth through human capital. An increase in life expectancy, a decrease in the time and resources spent on sick leave and an improvement in the well-being of employees contribute to an increase in the supply of labor (not to mention the fact that employers are more likely to hire healthy workers), and a healthier population is more likely to receive education, thereby activating the education channel, they note in ^[22].

Vail ^[15] analyzes precisely the impact of health on economic growth. The author cites the following possible channels of health influence on growth:

– healthy workers work more and better; high life expectancy stimulates education, as education pays off over time;

– healthy students are less likely to miss classes and their cognitive abilities are higher, which improves the quality of education at a given level of education;

– high life expectancy encourages saving for retirement, which increases the savings rate;

– physical capital per employee can grow due to the growth of the labor applied by a healthy employee, which increases the marginal product of capital;

– a decrease in child mortality leads to a decrease in population growth, which has a positive effect on economic growth, since an increase in population growth has a positive effect on the attitude of children to adults, the amount of physical capital and land per employee decreases, which ultimately reduces economic growth.

In an empirical paper, Easterly and Rebelo ^[25] analyze panel data from 1970 to 1988 for 118 countries and separately from 1870 to 1988 for 28 developed countries and conclude that health and education have a positive impact on economic growth. Hartwig ^[26] uses the Granger causality methodology on panel data (1970-2005) and concludes that education is much more important than health for economic growth, however, only OECD countries are analyzed in this work. Iqbal and Daly ^[27] use a dynamic panel to analyze low- and middle-income countries (1996-2010), and the authors conclude that health is much more important for economic growth in low- and middle-income countries, since high-tech processes are not widespread in them. Kartseva and Kuznetsova ^[28] based on Russian data (2012-2019), they find a positive impact of health on income, especially for groups of the population who have a low level of education.

2.2. Conditionally productive expenses

As already mentioned, productive expenses include mainly: education costs, health care costs, infrastructure costs, R&D costs. The literature also notes that spending on social protection has a positive impact on economic growth ^[24,29]. Unproductive expenses mainly include defense expenses ^[18].

In empirical work ^[24], spending on social protection has a positive impact on economic growth, and the positive impact increases with increasing incomes of the population. Spending on social protection can distort the supply of labor, thereby affecting economic growth ^[18].  There are two possible channels of influence, let's consider them using the example of unemployment benefits, as Nitschkamp and Put did ^[18]:

1. higher expenses – higher unemployment benefits – lower labor supply – lower growth rates;

2. or vice versa, higher expenses – higher unemployment benefits – more opportunities to retrain or not take the first job that comes to hand, getting a job in a place more suitable for skills – higher labor efficiency – higher growth rates.

Empirical results are more supportive of the second channel ^[24,29], that is, spending on social protection has a positive impact.

The state can encourage investments in physical capital. Capital accumulation through production implies that the return on investment in physical capital is greater than just the return on capital (due to positive externalities, for example, increasing the skills of workers when working on machine tools), it is noted in theoretical works ^[30,31]. Gilles ^[32] concludes in the model of endogenous growth taking into account social returns that subsidies for investment or to increase interest income for investors (for example, a premium from the state for investments) increase the growth rate. Castro and co-authors ^[8] conclude in the OLG model with possible risks for investors from opportunistic behavior of firms that if there is high capital mobility in a country, then improving investor protection leads to growth, since the interest rate of the world market is used, but if low capital mobility, then improving protection reduces growth due to an increase in the interest rate (as firms disclose real risks, therefore, the risk premium increases).

The empirical work ^[33] uses the MNC method based on US data (1947-1985) and concludes that investing in equipment has a significant impact on economic growth, namely, that increased investment in equipment leads to accelerated economic growth, and if externalities are taken into account, the return will be even higher (i.e. social return from investments). However, in empirical work ^[34] on the cross-sectional model and on the model with fixed effects, Granger causality (1965-1985) is used, the authors conclude that there is no evidence that capital accumulation or investment in equipment contributes to economic growth, but, on the contrary, economic growth contributes to capital accumulation. Nonneman and Vanhudt ^[35] use a cross-sectional model for OECD countries (1960-1985) and find that physical capital is insignificant in general (although the authors do not discuss this result in any way). 

Infrastructure development reduces transportation costs, negotiation costs, etc. The larger the population, the more benefits infrastructure development brings, it is noted in the works ^[18,22]. Zagler and Durnecker ^[22] use an endogenous growth model with a variety of final and intermediate goods and with the number of people employed in R&D and conclude that infrastructure costs can be unproductive, since the need for infrastructure depends on many factors. That is, spending on infrastructure can be unproductive unnecessarily. Ajemoglu and co-authors ^[17] find, after analyzing the binomial model for US data (1836-1900), a positive relationship between economic growth and innovation and conclude that in order to develop innovation, the state needs to develop infrastructure related to innovation (in this work, the number of offices for issuing patents was analyzed). Ponomarev ^[36] analyzes the reports of 230,000 companies in Russia and the level of infrastructure development (2011-2017) and concludes that infrastructure has a positive effect on overall productivity. Macheret ^[37] concludes that infrastructure is one of the key factors of economic development.

R&D expenditures stimulate economic growth through an increase in human and physical capital, as noted in theoretical works on endogenous growth models ^[22,38]. However, Morales ^[38] concludes that not all R&D expenditures stimulate long-term economic growth, since if the state develops the R&D sector, which can be developed with the help of private investment, then such government spending does not stimulate economic growth. Jones ^[39] uses a semi-endogenous growth model, in which investments in R&D affect only the transition state from one equilibrium to another, but do not affect economic growth in the long term, since R&D can only affect labor efficiency, that is, as a result, economic growth depends on the amount of labor in the economy. In theoretical work with duopoly ^[40], the author comes to the conclusion that the optimal amount of subsidies for R&D negatively depends on the amount of income tax and positively on the complexity of income tax evasion.

Nichkamp and Put ^[18], using meta-analysis, come to the conclusion that most of the literature considers defense spending to be unproductive expenses. However, as the authors write in ^[18,41], the build-up of military power carries benefits in the form of the fact that the conflict did not occur due to the fact that the opposite side has insufficient military power to foment conflict, that is, benefits from peace. In addition, part of the defense spending can go to technology development.

2.3. Taxes

As Nitschkamp and Put point out ^[18], all types of taxes have a negative impact on economic growth. Nevertheless, taxes are considered as costs of society, which should be less than the benefits of society from government spending. At the same time, taxes on externalities (taxes on environmental pollution, deforestation, etc.) can have a positive effect on society ^[22].

In many theoretical studies, income tax is excluded from the analysis, since very often one of the key assumptions of the models is perfect competition, that is, profit is zero at all time intervals and income tax does not affect economic growth in any way. Nevertheless, in the works devoted to the impact of various types of taxes on economic growth, income tax is considered the most distorting and the most detrimental to economic growth. In empirical work ^[42], income tax has a negative impact on economic growth through a decrease in the attractiveness of investments. The income tax reduces the net profit of enterprises, thereby reducing the return on investment in these enterprises, which reduces the incentives for investment, Johansson and co-authors note ^[42]. A decrease in profit reduces the valuation of the enterprise, which worsens credit conditions, that is, leads to high costs for servicing new loans and, accordingly, an additional decrease in profits. And in addition, a decrease in return on investment reduces household disposable income, which may additionally reduce the share of savings ^[22]. Of the possible positive effects of income tax, Trostel ^[43] notes that at the same time, the attractiveness of human capital relative to physical capital increases, and, accordingly, the incentives for investing in human capital increase.

Johansson and co-authors ^[42] analyze OECD countries on panel data (1975-2006). Also in this work, the authors analyze the rest of the countries using descriptive statistics and conduct an extensive literature analysis on the impact of taxes on economic growth. Therefore, we will discuss the conclusions on income tax in this work separately. The authors conclude that a stable and predictable tax system can attract more investments. In addition, the stability and efficiency of the tax system attract multinational corporations, which has a positive effect on growth. Having a single income tax rate for all corporations improves the quality of investments, as distortions in asset selection disappear. The authors also conclude that low income taxes only for young and small firms may not have a significant effect on growth, since such firms often have negative or too small profits, and high income taxes, on the contrary, negatively affect firms that already have income and are trying to expand, since the higher the income tax, the slower the expansion and, consequently, the slower the growth rate. On top of that, low taxes can encourage foreign investment.

Engen and Skinner ^[44] note that income taxes have a negative impact, as expected profits decrease, which reduces venture capital investments in sectors with high risks, but with high potential returns. In addition, any taxes can distort investments, since they will be directed from the high-tax sector to the low-tax sector, according to the same principle, labor can be sent from one sector to another.

In the book on the impact of public debt on economic growth, Greiner and Finke ^{[45, p. 166]} write that a progressive tax reduces the rate of economic growth through a reduction in investment and through a decrease in productive expenditures relative to output, since a progressive tax reduces the income of wealthy households, which reduces their investments and income before taxation. Alesina and Rodrik ^[46] analyze a model of endogenous growth in which agents separate income from labor and income from capital, and conclude that transfers in favor of workers who offer their labor inelastically leads to a decrease in growth. Reducing the marginal personal income tax leads to an increase in productivity through entrepreneurship, since with the same risks, entrepreneurs receive large profits, which increases productivity ^[42]. Reducing the personal income tax limit stimulates education, because the more educated population receives a higher salary, and if incomes after education grow slightly, then there are fewer incentives to get education, which reduces human capital and reduces growth rates ^[47]. Zagler and Durnecker ^[22] suggest that the main negative impact of personal income tax is through a decrease in the purchasing power of the population, thereby a large share of income is spent on consumption and, accordingly, a smaller share on investments. It is also suggested in ^[22] that the growth of personal income tax leads to a decrease in the relative cost of recreation, which reduces the supply of labor. However, Zagler and Durnecker ^[22] do not add that a decrease in income from reducing the number of additional hours of work ultimately reduces disposable income and, accordingly, savings, which additionally negatively affects economic growth).

Stiglitz ^[48] notes that increasing the progressivity of the tax system leads to a reduction in inequality. Reducing inequality increases economic growth in the long and short term, as noted in theoretical works ^[49,50]. In ^[48], it is proposed to combat inequality by increasing taxes on the rental of land and real estate, thereby stimulating investment in more productive sectors of the economy. At the same time, Johansson and co-authors ^[42] conclude that the least distorting tax is the real estate tax. Borissov and Hashimzade ^[50], using theoretical analysis, come to the conclusion that there is no compromise between inequality and economic growth in the sense that with the help of tax reforms it is possible to reduce inequality first, and then the economy can move to a point of a new equilibrium path with the absence of inequality, where economic growth is higher than under any in the form of inequality. The increase in VAT increases inequality, since the population with low incomes spends a much larger share of their income on consumption than the population with high incomes. To offset the inequality caused by VAT, Gale ^[51] proposes the introduction of progressive VAT. (Thus, one of the possible channels for VAT to influence economic growth in the long term may be inequality.)

2.4. Government debt

Barro ^[5] in his empirical and theoretical work establishes that expenditures leading to budget deficits have a negative impact on growth, since they raise interest rates and reduce private investment. On the contrary, expenditures that require higher taxes on unproductive activities such as consumption or recreation can have a positive impact on economic growth ^[5].

In the theoretical work ^[12], the authors, based on the Solow model, establish that the level of public debt can affect growth rates through its impact on inequality. Yakita ^[9] in the OLG-modified endogenous growth model finds that balanced long-term growth is possible if the national debt never exceeds a critical value, and the critical values for the level of public debt depend on the amount of capital in the economy. Teles and Mussolini ^[10] establish in the OLG model and panel data for 74 countries (1997-2004) that, on average, the national debt limit, exceeding which reduces the growth rate of the economy, is about 22% of GDP.

An increase in government debt due to an increase in productive government spending can lead to an increase in output, but to a certain level, since an increase in government debt leads to an increase in the equilibrium interest rate and, accordingly, to an increase in interest payments, as noted in ^[10,52]. In the theoretical work ^[11] in the OLG model, the authors conclude that, in the optimum, the state should oblige pension funds to purchase public debt with a yield below the market if this public debt is used for productive expenses. It makes sense to raise the level of government debt-to-output ratio only for the sake of productive expenses ^[10.53].

Greiner and Finke ^{[45, p. 175]} conclude in one of the models of endogenous growth that if public investments are financed with the help of public debt, then the transitional growth rates will increase, and the long-term growth rates will decrease, as the ratio of public debt to GDP increases in the long term. However, if the growth rate of the national debt is lower than the growth rate of GDP, then the transitional growth rates will increase, but the long-term growth rates will not decrease ^{[45, p. 175]}.

Despite the fact that usually an expansionary AF is associated with higher growth rates, and a restraining AF, on the contrary, fiscal consolidation is carried out in order to stimulate economic growth through improved stability, but fiscal consolidation in stable countries does not bring a positive effect ^[54]. However, in the theoretical work ^[55] in the OLG model, it is concluded that consolidation in any country contributes to the pace of economic growth. Greiner and Finke ^{[45, p. 192]} note that public debt can have an impact on economic growth if the deficit is financed through distorting taxes, as a result this will lead to a decrease in growth rates.

Consolidation through reduction of transfers and salaries is more favorable for economic growth than consolidation through tax increases, the authors conclude in empirical works ^[16,54]. If the deficit is mainly financed through domestic resources (privatization and seigniorage), then this exerts inflationary pressure, and high inflation levels negatively affect stability, that is, as a result, negatively affect economic growth, it is assumed in ^[54], in which the authors analyze panel data on poor countries (1990-2000). In theoretical works ^[56-58], the endogenous growth model includes money, which can influence the rate of economic growth through agents' preferences between consumption and leisure, and an increase in inflation leads to a decrease in employment (since agents begin to prefer free leisure, which reduces incentives to work) and, consequently, to a decrease in economic growth. Instability of government revenues and high levels of public debt can alienate potential investors and entrepreneurs (due to a possible default and subsequent crisis), which negatively affects the pace of economic growth ^[54].

Giavaci and Pagano ^[59] suggest that expectations (Ricardian equivalence in action) play a significant role for economic growth rates. When the state carries out fiscal consolidation, economic agents see that now the national debt is decreasing, and in the future they will not have to pay for it, which stimulates consumption and, accordingly, reduces the savings rate. As a result, this has a negative impact on the pace of economic growth.

2.5. The size of the State

Churchill and co-authors ^[60] analyze 87 articles using meta-analysis on the impact of the size of the state on the economy (the share of state expenditures in output, the share of taxes in output, or the share of state enterprises in the economy) and conclude that the size of the state rather negatively affects the economy. The authors of this work also use working papers, not just published articles, to avoid sample distortion. In the empirical work ^[23], the authors also note mainly the negative impact of the size of the state. Kudrin and Gurvich ^[61] suggest that the higher the share of the state in the economy, the more economic agents are guided by non-market incentives, which leads to distortions and, consequently, to a decrease in economic growth.

Since ^[60] provides an extensive overview of the influence of the state on economic growth, we will discuss this work separately. Possible negative effects of a large state are assumed (a large share of the state in the economy): crowding out private investment; large expenditures require large taxes, and taxes have a negative impact on growth; a large state can be a source of inefficiency due to rent-oriented behavior and corruption. Possible positive effects are also assumed: provision of public goods; minimization of externalities; enforcement of the law; development of human capital through health and education; building and maintaining reliable infrastructure; from a Keynesian point of view, the state increases demand, which ultimately has a positive effect on supply. A significant negative effect is present only in developed countries, and for all other countries, a large state does not have a negative impact on economic growth, but it also does not have a positive impact. Barro ^[5] also concludes that the negative impact of government spending on economic growth is more pronounced in countries with higher levels of economic development and that policies aimed at reducing government spending can have a positive impact on long-term growth. However, Churchill and co-authors ^[60] admit that the possible negative effect of the size of the state is due to the fact that the state is growing faster than output growth, and the decrease in growth rates found in the works is caused by increases in household incomes.

3. Short-term effects of AF

In ^[62], the short-term and long-term effects of fiscal policy are analyzed in one model, combining the neoclassical approach with the endogenous growth model. AF affects the economy in the short and long term, but the main impact is in the short term ^[62]. The state intervenes in the economy in order to reduce the volatility of economic cycles. AF is one of the main tools of the state to achieve this goal, and AF is effective if it is countercyclical, however, not all countries have the opportunity to conduct countercyclical AF.

3.1. The cyclical nature of the FP

FP is effective if it is countercyclical, thereby maintaining financial stability and stable economic growth, but some developing countries, on the contrary, conduct a pro-cyclical FP.

Talvi and Weich ^[63] conduct an analysis of 56 countries for the cyclical nature of AF and a theoretical analysis of the cyclical nature of AF in a model with perfect foresight. The authors suggest that the implementation of a pro-cyclical FP is mainly due to poorly developed political institutions and strong fluctuations in government revenues. The fact is that during periods of economic growth, there is an increase in incomes and an increase in state revenues, which leads to inefficient budget waste due to corruption in states with weak political institutions, and sometimes even such expenses negatively affect economic growth. As a result, during periods of economic growth, the state conducts stimulating fiscal policy through tax cuts and spending increases (spending increases are associated with political pressure), whereas during periods of economic downturn, tax rates are increased to keep government revenues at the same level, or expenses are reduced to meet the reduced budget. Unfortunately, the work does not reflect in any way the possible inefficiency of tax collection in developing countries, that is, during periods of crisis, the share of the shadow economy may also grow, which may additionally reduce tax revenues. In the paper, the authors conclude that the inability to carry out stimulating AF during periods of recession is also due to the fact that a country may have limited access to the debt market. An increase in government spending usually leads to an increase in inflation. Inflation reduces the purchasing power of the population, thus in developing countries, purchasing power decreases due to rising inflation during crisis periods and increases during periods of growth, which is also a kind of procyclical component. And it is precisely such a component that the government must fight with the means of countercyclical OP.

Aldama and Creel ^[64] analyze the cyclical nature of the FP of OECD countries (1996-2017) using panel data and find that in many OECD countries the discretionary part of the FP is procyclical due to political pressure, the countercyclical component in such countries is mainly automatic stabilizers (unemployment benefits, pensions, etc.). Empirically In ^[65], the authors find that the oil-producing countries (1970-2007) are characterized by a procyclical AF.

During the COVID-19 period, many states used unprecedented measures to support the population and businesses. Hale and co-authors ^[66] analyze OP measures in the ten largest countries during the COVID-19 period (from February 2020 to September 2021) using a dynamic panel and conclude that the most pro-inflationary measures are those that provide funds to consumers. The least inflationary measures are those that ease the debt burden for firms, that is, instead of giving money to firms, it is better, if inflation is undesirable, to ease the debt burden, and even better to do it only for firms ^[66]. However, for firms in crisis periods, the most appropriate tool may be an increased depreciation premium, since in unfavorable periods profits decrease and may become negative (that is, a reduction in income tax does not affect in any way), whereas an increased depreciation premium reduces the tax base in future periods and forms more positive expectations about future profits ^[67]. Fiscal incentives also have a positive effect on consumer expectations and sentiment, which are also pro-inflationary factors ^[66]. If the state seeks to stabilize both the main macroeconomic indicators and public finances, the state should choose an unlimited FP in combination with a two-mandate monetary policy, which is based on unemployment and inflation ^[68].

3.2. Limitations of the FP

Conducting an ill-considered expansionary FP can lead to budget destabilization, which is why some countries set fiscal rules. Nevertheless, in times of crisis, many countries have access to the debt market in some form, which allows them to increase spending during the crisis without raising taxes, which leads to an increase in public debt. Due to the decrease in the effectiveness of MP, many countries began to actively pursue stimulating OP, especially during crisis periods (unprecedented expenditures during COVID-19), and the ratio of public debt to GDP increased. Taxes distort the economy, so their increase is undesirable, besides, tax increases are a complex political process. Thus, some countries have nothing left but to increase the level of public debt in order to compensate for the lack of investment in the economy in human capital and in high technologies, otherwise the economy may become uncompetitive ^[69]. However, the FP is limited by the level of taxes collected and the level of public debt. By itself, the level of public debt is not so important, interest payments on it are important ^[53]. If interest payments on the national debt reach critical values, then there is a risk that the level of taxes collected will be insufficient to cover the cost of servicing the national debt, which provokes instability (especially if the probability of default increases critically).

Furman ^[53] formulates a "new view" on AF, the main statements of which are as follows:

– stimulating FP perfectly complements countercyclical MP by increasing inflation and lowering real interest rates;

– during periods when there is an effective lower bound (ELB) problem, that is, when interest rates cannot fall below without causing a massive outflow of funds from banks, government spending can stimulate investment, as it can stimulate economic growth;

– government spending (especially investments) can be self-sustaining, as they lead to economic growth, increased inflation and a decrease in the real rate;

– an expansionary FP has a positive effect on nominal output, that is, the ratio of government debt to nominal output may not change or even decrease, besides, an increase in nominal output leads to an increase in taxes collected, which has a positive effect on the budget balance;

– the debt crisis is not always caused by an increase in public debt, in most cases the debt crisis is preceded by a bubble or a black swan, which are the reason for the increase in public debt.

The review paper ^[70] concludes that the Eurozone relies much more heavily on MP, ignoring OP for the benefit of stability, and this contributes to the fact that the Eurozone is much more likely to face the ELB problem, and growth rates are lower than in the United States, where OP is used much more actively.

In the theoretical Gemmel and co-authors ^[62] assess the impact of an increase in government spending and taxes simultaneously (an increase in spending per unit together with an increase in taxes per unit), and conclude that the impact is rather zero, that is, the effects cancel each other, as in the theoretical work ^[71].

Fiscal consolidation is mainly aimed at maintaining stability by increasing tax revenues and reducing government spending. Fiscal consolidation is carried out according to plan, regardless of cycles, set by Aldama and Creel ^[64]. If there is an ELB problem, and consolidation is necessary, then a VAT increase may be appropriate. VAT is the only "restraining" tool of the Federal Tax Service, which has a positive effect on prices. VAT is gradually increasing its share of government revenues in many countries, whereas for some poor countries VAT is the main source of government revenue, as it replaces duties under free trade agreements, according to empirical works ^[25,72,73]. Imported goods are much easier to tax than locally produced goods, as a result, in countries where the state is unable to collect taxes in full, imported goods are subject to VAT, and locally produced goods are not always subject to VAT, which leads to a decrease in the competitiveness of imported goods ^[63,72,73]. Alavuotunki and co-authors ^[73] conclude that openness has a negative impact on government revenues when introducing VAT, which is at odds with the results of Keane and Lockwood ^[72]. The work ^[73] concludes that, most likely, the differences with the work ^[72] are due to the fact that VAT does not fully compensate for the loss of government revenues from the removal of trade duties.

VAT, like personal income tax, reduces the purchasing power of the population, thereby spending most of the income on consumption and less on investment, however, VAT, unlike personal income tax, makes consumption less attractive compared to investments, which can have a positive impact on the savings rate. At the same time, enterprises need to purchase at a higher price, this is true for both capital and raw materials. Enterprises for whose goods demand is most elastic are likely to lose profits, since they will not be able to shift all VAT-related costs to consumers. However, the results of an empirical study in the Eurozone countries for 70 product groups (1999-2013) ^[74] show that the transfer of VAT changes to prices is mainly one-to-one, that is, the demand for goods is mostly inelastic. Nevertheless, prices react much faster to VAT increases than to decreases in most countries and for most goods and services ^[75]. Moreover, price reductions can occur over several years, and increases within a few weeks, according to Benzarti and co-authors ^[75], who analyze the VAT experiment in Finland (2005-2015), as well as 427 firms in the Eurozone (1996-2015). That is, it is necessary to take into account the possible asymmetry of responses from a decrease and VAT increases.

3.3. Assessment of the impact of AF

The problem of assessing AF is a possible inverse relationship between AF and economic growth ^[54]. When growth slows down, the ratio of government spending to GDP increases with government spending unchanged. Blanchard and Perotti ^[76] presented a way to assess the impact of AF on the economy in the short term. When assessing the impact of AF on the economy in the short term, multipliers are used, and the analysis is mainly related to costs, since, as a rule, costs increase during crises. Expenses are analyzed by groups of government expenditures, and it is much more difficult to make such an analysis with taxes, since taxes are much less likely to change. There are two approaches to taxes: changing the marginal tax rate or changing tax charges. At the same time, the assessment of the effect of AF in dynamic stochastic general equilibrium (DSGE) models is more often analyzed. Polbin and Andreev ^[77] in a review paper on estimates of the effect of AF in DSGE models show that the main conclusions from other models are confirmed in DSGE models.

Auerbach and Gorodnichenko ^[78], for the first time, attempt to estimate multipliers in crisis and normal periods in a single SVAR model for US data (1947-2008), they follow the approach from ^[76]. The main conclusion in ^[78] is that in times of crisis, defense spending has the greatest impact. However, Rami ^[79] criticizes the approach ^[76] because of its instability to changes in output during crisis periods, which is why the multipliers turned out to be larger than they actually are. Rami ^[79] evaluates multipliers on the same data in three different ways (method ^[76], method ^[80] and method ^[81]) and gets three significantly different results, while the method from ^[76] shows incredibly high multipliers, while the other methods show multipliers below one. In ^[78], the multipliers are also significantly larger than one, which does not agree with the conclusions from DSGE models, in which the multiplier can hardly exceed one, provided that there is an ELB problem and a crisis ^[77]. In the standard RBC models (on which the standard DSGE models are based), it is assumed that consumption decreases from an increase in government spending, while VAR models show the opposite, according to an empirical analysis in the VAR model based on US data (1954-1999) in ^[82]. Hale and co-authors ^[66] also find that consumption reacts positively to an increase in government spending. In times of crisis, consumption is extremely important for the economy, that is, the higher the consumption, the easier the crisis passes. Thus, the conclusions from DSGE models about low multipliers may be related to the assumption of a reaction of consumption to an increase in government spending.

The problem is that in the scientific community there is no clear understanding of what exactly are AF shocks, which channels of the transmission mechanism are involved and are the main ones, and, finally, the most important problem is that there are no generally accepted ways to measure AF multipliers, therefore, conclusions of work on multipliers should be approached carefully ^[79]. However, an incorrect estimate of the size of multipliers does not mean that you need to stop evaluating multipliers using traditional methods. The main thing is that the multipliers of government expenditure groups evaluated in one work can be compared with each other and identify the best and worst. However, the exact dimensions of the multipliers undoubtedly contribute to a more optimal AF.

4. Conclusion

This paper analyzes the literature on the impact of AF on economic growth in the short and long term.

FP can indeed influence long-term economic growth rates, and the overwhelming majority of studies confirm this. Basically, the theoretical analysis of the impact of AF on economic growth is carried out using endogenous growth models, but modified Solow and OLG models are also used. The main models for empirical analysis of the effectiveness of fiscal policy include: a model with fixed effects on panel data, a dynamic panel and a cross-sectional model (in older works).

The main lever of the AF's influence on long-term economic growth is the impact on the development of human capital. Government investments in education and health are the most powerful tool of the Federal Budget for a positive impact on economic growth. The state can also invest in R&D, equipment, infrastructure, social protection and defense, but the positive effect of such measures does not always occur.

Taxes affect economic growth by influencing the incentives of economic agents to use the resources available to them. The most disastrous tax for economic growth is the income tax, and the real estate tax is the least distorting. A progressive tax system, on the one hand, reduces inequality, and a number of authors argue that reducing inequality leads to an increase in growth rates. On the other hand, many works assert precisely the negative impact of the progressive tax system on economic growth, so there is no consensus on this issue in the literature.

The level of public debt has a negative impact on economic growth due to interest payments related to unproductive expenses. However, some studies still note the negative impact of the level of public debt on economic growth. Rather, the size of the State has a negative impact on economic growth, with a particularly significant negative effect observed for high-income countries.

AF has a significant impact on the economy in the short term. In developed countries, AF is rather countercyclical, but discretionary AF can be procyclical due to political pressure. In poor countries, OP is more likely to be pro-cyclical due to limited access to the debt market.

Building up government debt to cover unproductive expenses is undesirable, but increasing government debt for productive purposes can be self-sufficient. Fiscal consolidation can have a positive impact on economic stability. The least negative consolidation for the economy is consolidation through cost reduction. If it is necessary to consolidate through taxes, then the VAT tax may be the most appropriate, however, it is necessary to take into account the possible asymmetry of changes in the VAT rate.

The assessment of the effect of AF in the short term can be greatly overestimated, since the standard approach ^[76] used in many studies can greatly overestimate the multipliers. Thus, the conclusions of the work on the evaluation of multipliers must be approached with caution.

Very little literature is devoted to the impact of VAT on economic growth in the long term. A little more work is devoted to the impact of VAT on the economy in principle. States may be able to change VAT much more actively in order to conduct a more confident countercyclical FP. However, even with the problem of an effective lower limit, VAT has remained without due attention in the literature.

Very few studies provide an empirical assessment of the impact of fiscal policy on economic growth in the short and long term. Meanwhile, the conclusions obtained for developed countries may be completely different than for developing countries.