In this paper, we analyse the effects of transitory fiscal expansions when public debt is used as liquidity by the private sector. We conduct this analysis in an incomplete market model where agents face uninsurable idiosyncratic income risk and have limited ability to borrow against future income (i.e., markets are iliquidity constrained in the terminology of Kehoe and Levine, 2001, amongst others). Non Ricardian models of this type have on occasion been used to analyse the aggregate and welfare effects of public debt in the steady state (see Woodford, 1990; Aiyagari and McGrattan, 1998). To date, there have been surprisingly few attempts at clarifying how such economies respond to aggregate fiscal shocks. One important contribution is Heathcote (2005), who offers a quantitative assessment of the effect of tax cuts. In this paper, we attempt to characterise analytically and qualitatively the impact and dynamic effects of government spending shocks on macroeconomic aggregates.
The spending shocks of which we analyse the effects have one significant, and realistic, feature: they are at least partly financed by government bond issues in the short run, with public debt then gradually reverting to some long run target value thanks to future tax increases. Note that whether government spending is financed by taxes or debt does not matter in complete markets, Ricardian economies with lump sum taxation, because households discounted disposable income flows are identical between alternative modes of government financing. Then, under reasonable assumptions about preferences and technology, the negative wealth effects associated with transitory spending shocks lead to falls in the demand for both private consumption and leisure, which in turn produces a drop in the real wage (e.g., Baxter and King, 1993).
The deficit financing of spending shocks can, however, have very different consequences when public debt is used as private liquidity, that is, as a store of value held by agents for precautionary, or "self insurance", purposes. Starting from a situation in which liquidity is scarce (in a sense that we specify below), such policies have the side effect of increasing the stock of assets available in the economy, thereby facilitating self insurance by bond holders and effectively relaxing the borrowing constraints faced by households and firms. As we show, the liquidity effects associated with rising public debt tend to foster households private consumption demand, along with the labour demand of borrowing constrained firms. Whether and when such liquidity effects may offset wealth effects, and thus overturn the predictions of the complete markets model regarding the effects of spending shocks on private consumption and wages, is the central theme of this paper.
It is perhaps surprising that the actual impact of our fiscal experiment is still subject to so much empirical controversy. In particular, the application of different identification strategies to U.S. data has either supported the Real Business Cycle prediction of a fall in private consumption and wages following an increase in public spending (Ramey and Shapiro, 1998; Ramey, 2009), or come to the opposite conclusion that both variables actually increase after the shock (e.g., Blanchard and Perotti, 2002; Perotti, 2007), which latter is consistent with the Old Keynesian model and with a version of the New Keynesian model endowed with a sufficient number of market imperfections (Gali et al., 2007). Given this lack of consensus, our goal here is not to take any definitive position as to whether an adequate fiscal policy model should generate pro or counter cyclical responses of those variables to public spending shocks.
Rather, we use our model to illustrate that both outcomes are theoretically possible (and not implausible quantitatively), depending on the relative strengths of the liquidity and wealth effects that arise following the shock. As we show, which effect actually dominates crucially depends on how quickly the fiscal rule followed by the government ensures the reversion of public debt towards its long run target following the initial fiscal deficit. If taxes rise promptly after the increase in public spending, then public debt will not vary very much and liquidity effects will be weak; in this situation, wealth effects are likely to be dominant and private consumption and wages will fall. If, on the contrary, the slow reaction of taxes leads to a substantial growth of public debt in the short and the medium run, then liquidity effects may be strong enough to dominate wealth effects, causing private consumption and wages to rise. Overall, temporary increases in public spending are all the more effective at raising output when the simultaneous response of taxes is limited.
Download
Fiscal Policy in a Tractable Liquidity Constrained Economy
