W.K.Y. Fung et al /Journal of Development Economics 61(2000)111-13 system actually subsidizes the government investment projects and the loss-incur ring state-owned enterprises (vi) Since the state banking system is employed as the main vehicle to transfer resources from the non-state sector to the state sector, the money supply process is endogenously determined by the governments budget requirement 3. The model In our analysis, we will incorporate into our model all of the six institutional features of the Chinese economy mentioned above. The model constructed in this paper has the following three sets of attributes. First, in order to examine the impact of government policy changes on the output growth, we construct an overlapping generations model featuring endogenous growth. Externalities or learning by doing are introduced to the production technologies, which makes sustained growth feasible through a combination of increasing social returns and diminishing private returns to capital. As such, economic growth is driven by the augmentation of the capital stock in the econom Second. to evaluate how allocation of financial resources affects the level of capital investment, production and thus the growth rate of output, our model pecifies an economic system with two sectors: a real sector consisting of the markets for a consumption good, a capital good, and labor services, and a financial ctor consisting of the markets for money, government bonds, and bank deposits and loans. Money is introduced into the model by assuming that there are cash-in-advance constraints on the purchases of the consumption and capital roods These bank loans are"hidden deficits"of the government(McKinnon, 1994). Using a two-dimen- onal moral hazard model, Zuo and Sun(1996) show that there are incentive-based reasons for the low interest rate policy in China. ccording to Feltenstein and Farhadian (1987), in China, changes in the money supply for the period of 1954-1983 can be explained by: (i the government deficit, (ii) procurement payments to the farmers, and(iii) the wage bill of govemment and state enterprises. As Brandt and Zhu(1995)argue when rapid economic growth is generated in the non-state sector, to equalize the benefits of growth, the govemment uses the financial system as the vehicle to transfer resources from the non-state sector to the state sector. As a result, the money supply process is endogenously determined by the government transfer requirement. nat the individuals of the economy are three-period lived overlapping generations, we can simplify their optimization problems and obtain a tractable framework for dynamic general equilibrium analysis. In addition, in contrast to a standard infinitely lived representative agent model an overlapping generations model incorporates the heterogeneity among individuals and allows the govemment's choice between and bond financing of its budget deficit and debt repayment to have real effects on the (1987), and Bencivenga and Smith ver externalities"considered by Romer(1986), Boyd and Prescott These are similar to the116 M.K.Y. Fung et al.rJournal of DeÕelopment Economics 61 2000 111–135 ( ) system actually subsidizes the government investment projects and the loss-incur￾ring state-owned enterprises. 12 Ž . vi Since the state banking system is employed as the main vehicle to transfer resources from the non-state sector to the state sector, the money supply process is endogenously determined by the government’s budget requirement. 13 3. The model In our analysis, we will incorporate into our model all of the six institutional features of the Chinese economy mentioned above. The model constructed in this paper has the following three sets of attributes. First, in order to examine the impact of government policy changes on the output growth, we construct an overlapping generations model featuring endogenous growth. 14 Externalities or learning by doing are introduced to the production technologies, which makes sustained growth feasible through a combination of increasing social returns and diminishing private returns to capital. 15 As such, economic growth is driven by the augmentation of the capital stock in the economy. Second, to evaluate how allocation of financial resources affects the level of capital investment, production and thus the growth rate of output, our model specifies an economic system with two sectors: a real sector consisting of the markets for a consumption good, a capital good, and labor services; and a financial sector consisting of the markets for money, government bonds, and bank deposits and loans. Money is introduced into the model by assuming that there are cash-in-advance constraints on the purchases of the consumption and capital goods. 12 These bank loans are ‘‘hidden deficits’’ of the government McKinnon, 1994 . Using a two-dimen- Ž . sional moral hazard model, Zuo and Sun 1996 show that there are incentive-based reasons for the low Ž . interest rate policy in China. 13 According to Feltenstein and Farhadian 1987 , in China, changes in the money supply for the Ž . period of 1954–1983 can be explained by: i the government deficit, ii procurement payments to the Ž. Ž . farmers, and iii the wage bill of government and state enterprises. As Brandt and Zhu 1995 argue, Ž. Ž . when rapid economic growth is generated in the non-state sector, to equalize the benefits of growth, the government uses the financial system as the vehicle to transfer resources from the non-state sector to the state sector. As a result, the money supply process is endogenously determined by the government’s transfer requirement. 14 By assuming that the individuals of the economy are three-period lived overlapping generations, we can simplify their optimization problems and obtain a tractable framework for dynamic general equilibrium analysis. In addition, in contrast to a standard infinitely lived representative agent model, an overlapping generations model incorporates the heterogeneity among individuals and allows the government’s choice between money and bond financing of its budget deficit and debt repayment to have real effects on the economy. 15 These are similar to the ‘‘spillover externalities’’ considered by Romer 1986 , Boyd and Prescott Ž . Ž. Ž. 1987 , and Bencivenga and Smith 1992