Fiscal Policy Research In Stochastic Endogenous Growth Models

This paper investigates fiscal policy problems within stochastic models of endoge-nous growth with externalities by improved continuous-time stochastic optimizationmethods-the "bogus double-state variables" method, the "additional utility" valuefunction method and the stochastic Hamilton function method, analyzes growth ef-fect and welfare effect of fiscal policy, and discusses optimal fiscal policy.This paper combines portfolio theory and economic growth theory, and introducesgovernment expenditure into utility function to analyze effects of output risk and fiscalpolicy on the propoensity to consumption, the economic growth rate, individual wel-fare and portfolio selection, and discuss equilibrium asset pricing within a stochasticmodel of endogenous growth with external utility by the "bogus double-state variables"method. It is shown that the consumption tax is not only neutral for growth, but alsoneutral for welfare. The relationship of capital's risk and bonds' risk depends on gov-ernment budget. Bonds are more riskier than capital with goverenment budget deficits,while capital is riskier than bonds with government budget surpluses.Since government expenditure is ineluctablly accompanied with congestion, thispaper investigates government consumption expenditure and production expenditurewith relative congestion. For government consumption expenditure, this paper dis-cusses the optimal government scale and the effects of government expenditure's changeson economic growth and welfare by the "bogus double-state variables" method. Syn-chronously, the optimal tax policy is derived, and the performance of each kind oftax and the financing policy of government expenditure are analyzed. It is indicatedthat reducing the optimal size of government to raise the mean growth rate can notcertainly improve individual welfare. Only when government expenditure exceeds itsoptimal level, reducing government expenditure can improve individual welfare; whengovernment expenditure is lower than its optimal level, reducing government expendi-ture may be worsen for individual welfare. For government production expenditure,this paper analyzes growth effect and welfare effect of fiscal policy and discusses opti-mal fiscal policy by the "additional utility" value function method. It is indicated that growth-maximizing and welfare-maximizing government expenditure policies do notnecessarily coincide and their standing in conflict or coincide depends on the degree ofrisk aversion of the individual. Regardless of consumption expenditure or productionexpenditure, in the presence of relative congestion the Chamley-Judd optimal tax rule,which the long-run capital income tax is zero, is no longer ture.For the flaw of the conventional expected utility function which does not allowindependent variations in intertemporal substitutability and risk aversion, using thenon-expected utility function which can disentangle intertemporal substitution fromrisk aversion, this paper indicates whether growth-maximizing and welfare-maximizinggovernment expenditure policies being in coincidence or not depends upon the in-tertemporal substitution elasticity, not the risk aversion coefficient. Synchronously, itis shown that precautionary savings are not only ralated to the degree of risk aversionof the individual, but also related to the intertemporal substitutablity on consump-tion. By the explicit introduction of inelastic labor supply, capital income risk andlabor income risk are clearly distinguished, and it is found that capital income riskand labor income risk are major determinants within the individual's intertemporaldecision-making, economic growth and the design of the optimal tax policy. In ad-dition, the separation between capital income tax and labor income tax makes theoptimal tax policy much flexible.The market distortion which the private capital return rate is less than the socialcapital return rate, generated by the human capital externality, can be corrected bygovernment transfer expenditure-capital subsidy ("negative tax revenues"). By thestochastic Hamilton function method, this paper investigates the following two cases:income subsidies financed by consumption tax and government bonds, and wealth sub-sidy financed by labor income tax and government bonds. It is found that a permanentincome subsidy can improve the ecnomic growth rate, while a transitory income subsidywill reduce the economic growth rate. The effects of income subsidies on individualwelfare depend on the level of the economic growth rate. When the economic growthrate is suboptimally high, the permenant income subsidy will reduce individual welfareand the transitory income subsidy will enhance individual welfare; when the economic growth rate is suboptimally low, the permenant income subsidy will enhance individualwelfare and the transitory income subsidy will reduce individual welfare. Whether thesingle permenant income subsidy or the single transitory income subsidy falls short ofthe flat-rate income subsidy. Wealth subsidy can enhance economic growth withoutdoubt, while the effects of wealth subsidy on portfolio selection and the propoensityto consumption depend on the degree of risk aversion of the individual. Similarly toincome subsidies, the effect of wealth subsidy on individual welfare also depends onthe level of the economic growth rate. When the economic growth rate is suboptimallyhigh, wealth subsidy will reduce individual welfare; when the economic growth rate issuboptimally low, wealth subsidy will enhance individual welfare. Additionally, takingthe labor output rate as a benchmark, the wealth subsidy rate varies with governmentfinancing policy.Extending to an open economy, this paper investigates fiscal policy problems whengovernment expenditure is taken as public consumption and public production in asmall open economy. For government consumption expenditure, by the "additionalutility" value function method, this paper analyzes economic stability problems andoptimal fiscal policy, and compares with a closed economy. It is found that an openeconomy is more stabile than a closed economy, and the government scale in an openeconomy is larger than that in a closed economy. Differing from a closed economy,in an open economy completely eliminating the fluctuation of growth is not alwaysoptimal. When a country is a net creditor, the high growth of which completely elimi-nates the fluctuation will accompany with the welfare loss. For government productionexpenditure, by the "bogus double-state variables" method, this paper analyzes theeffect of fiscal policy, discusses the optimal tax policy, and compares with a closedeconomy. It is found that more government stochastic intervention can reduce theeconomic volatility in a closed economy, while in an open economy this is uncertain.When a country is a net creditor, more openness accompanies with a larger govern-ment scale. For government consumption expenditure, the optimal tax policy in anopen economy is equivalent to that in a closed economy. But for government produc-tion expenditure, the optimal tax policy in an open economy is incompletely equivalent to that in a closed economy. Under open conditions the optimal consumption tax inthe government production expenditure model is related to the portflio selction.