کارآفرینی و یارانه های دولتی: تجزیه و تحلیل تعادل عمومی

This paper quantitatively studies the effects of government credit subsidies. We find that current credit assistance programs in the form of interest subsidies exert strong effects on the allocation of credit to targeted entrepreneurs, but at the cost of non-targeted entrepreneurs. Total entrepreneurial activities are reduced and large output loss is incurred. The paper also examines several alternative credit programs. Our analysis suggests that income subsidy programs and programs that specifically target poor and capable entrepreneurs are more effective in promoting entrepreneurial activity and improving total output. These findings are based on a model with a large number of infinitely lived agents whose saving behavior and occupational choice are influenced by precautionary saving motives and borrowing constraints. Government credit subsidies, on the benefit side, enhance the liquidity of agents by providing an additional means of smoothing consumption and by effectively loosening borrowing constraints. On the cost side, these subsidies and resulting taxes crowd out capital and have adverse incentive effects.

The U.S. government plays a central role in promoting entrepreneurship in the United States. Every year, billions of dollars are put into the business community through the Small Business Administration (SBA).1 The level, time path, and type of government subsidy are obviously important issues in credit policy. In this paper, we analyze the quantitative effects of government subsidies on business credit allocation and economic efficiency, and contrast the results of the current program with alternative programs that use different credit instruments or different targeting rules. The analysis is conducted with a model parameterized to match various features of the U.S. economy. Under our parameterization, we find that the current interest subsidy has large allocational effects. This program greatly increases entrepreneurial activity of the targeted group, albeit at the expense of the non-targeted group. The net result is a reduction in the rate of total entrepreneurship. The efficiency cost is also high: output is about 0.62 percent lower than that under the optimal subsidy rate and 0.52 percent lower than that of the zero subsidy case. Finally, our analysis of alternative credit programs indicate that income subsidies and programs that target specifically poor and capable entrepreneurs are more efficient in promoting entrepreneurship and total output than the current practice. The belief that capital markets do not provide adequate funds for businesses, particularly new businesses, is the main rationale for government assistance programs. Recent empirical studies lend support to this belief. Evans and Jovanovic (1989), Evans and Leighton (1989), Holtz-Eakin et al. (1994), and Blanchflower and Oswald (1998), among others, find that a lack of wealth affects people's ability to become self-employed, even after accounting for the possible correlation between entrepreneurial ability and wealth. In a more recent study, Bond and Townsend (1996) reported the results of a survey of financial activity in a low-income, primarily Mexican neighborhood in Chicago and found that borrowing is not an important source of finance for business set-ups. In their sample, only 11.5 percent of business-owners financed their start-up with a bank loan, while 50 percent of the respondents financed their start-up entirely out of their own funds. Other authors (see, for example, Carpenter et al., 1994) find evidence of significant capital market imperfections even for publicly traded manufacturing companies. A relevant model for our purpose, therefore, is one with capital constraints, and one where agents in the economy endogenously make their occupational decisions. The model used here consists of a large number of infinitely lived agents whose saving behavior and occupational choice are influenced by precautionary saving motives and borrowing constraints. A subset of the agents receive subsidies from the government on their loan payments if they choose to carry out their projects. These subsidies are financed by a proportional income tax (Quadrini 2000 first used such a model without government intervention to study the influence of entrepreneurship on wealth distribution and social mobility). This model incorporates an additional role for the government subsidization of entrepreneurship that standard models do not include, and captures trade-offs between the benefits and costs of varying the amount of the subsidy. On the benefit side, the government subsidy enhances the liquidity of agents by providing an additional means of smoothing consumption in addition to claims to capital, as well as by effectively loosening borrowing constraints. When the interest rate is raised, the government subsidy makes assets both less costly to hold and more effective in smoothing consumption. On the cost side, the government subsidy, and the resulting taxes which they imply, have adverse incentive effects. In particular, government subsidies crowd out capital and entrepreneurship of non-targeted agents via high interest rates. In this respect, this paper resembles that of Aiyagari and McGrattan (1998) on the optimum quantity of government debt. The difference between different subsidy programs lies in the different incentive effects these programs have on agents’ saving and entrepreneurial behavior. In terms of the literature, there are a number of papers that investigate the effects of government credit policy on investment and entrepreneurial activity. Mankiew (1986), de Meza and Webb (1988), Smith and Stutzer (1989), Gale (1991), Innes (1991), and Li (1998) provide theoretical models of the effects of credit subsidies in markets with imperfect information. With the exception of Li (1998), this literature has largely ignored the relationship between wealth and entrepreneurial activity. Moreover, in most of these studies, the role of government relies on its superior ability to offer cross subsidizing contracts to borrower relative to private agents (see Lacker, 1994 for a discussion). Gale (1991) is the only paper that quantitatively evaluates the economic effects of federal credit programs. The paper's analysis, however, is static and conducted in a partial equilibrium setting. Hence, it does not capture the potential liquidity provision role of government subsidies. The model we consider is a variant of the deterministic growth model modified to include a large number of individuals subject to uninsured, idiosyncratic shocks to their labor productivity and entrepreneurial ability. Although there is no aggregate uncertainty in these models, there is individual uncertainty due to the absence of insurance markets. This is the feature that generates precautionary saving. This tendency to save, together with the borrowing constraint, give rise to the positive correlation between the probability of becoming an entrepreneur and individuals’ ownership of assets. Further, there is ex-post heterogeneity among individuals. In the steady state, a distribution of individuals over asset holdings and occupational choice emerges. One virtue possessed by this type of set-up is that, in addition to generating many plausible empirical implications, it is quite useful for quantitative analysis. As Aiyagari (1994), Huggett (1993) and Quadrini (2000) argued, this class of models generates a stationary distribution of wealth whose degree of concentration can account for the inequality observed in the U.S. economy. The plan of the paper is as follows. In Section 2, we present the theoretical model and define the equilibrium. In Section 3, we describe how we choose the parameter values for the benchmark model. In Section 4, we report the results and the robustness of our results for some alternative parameter values. In Section 5, we analyze the effects of alternative subsidy programs. Section 6 concludes.

Whether there exists a role for government in imperfect capital markets is open to debate, both theoretically and empirically. This paper draws on the existing literature and quantitatively analyzes the effects of the interest subsidy program currently in place in a dynamic general equilibrium model. Our analysis suggests that although there is still a role for the government in the business credit market, that role is more limited and somewhat different in nature than current policy would suggest. Specifically, the current subsidy program has large allocational effects, it exerts a powerful influence on the allocation of credit to the targeted group. The rate of entrepreneurship within the targeted group is greatly improved, but at the cost of the non-targeted entrepreneurs. As a result, the fraction of all agents engaging in entrepreneurial activities is lower. Moreover, the current interest subsidy creates a large efficiency loss. Total output is lower than under both the optimal subsidy case and the zero subsidy case. Our study of alternative credit subsidy programs also indicates that programs that are less distortionary, such as income subsidies and programs in which the government provides appropriate marginal incentives, appear to be more efficient than the current subsidy. In light of our results, a question that arises is: why then have interest subsidies been so popular? It may be that differences in government budgetary accounting allow guarantees to be passed more easily since, prior to 1992, interest subsidies did not appear in the budget until a payment was made.8 Furthermore, in practice, it may be difficult to implement a rule that targets less wealthy entrepreneurs since verifying their actual asset holdings is costly. This is perhaps why many credit subsidies have chosen to target business or owner characteristics that are easier to monitor. These characteristics include location of the business, and the race and gender of the business owner. For now, this paper makes no attempt to resolve these issues.