سیاست های پولی و آزادسازی مالی : موردی از میزان مصرف انگلستان

This paper outlines work done recently to investigate the impact of financial liberalization on monetary policy effectiveness. We look at the case of United Kingdom consumption behavior, in particular the savings ratio. Over the 1990s this sector has been heavily influenced by comprehensive deregulation of the U.K. financial services industry which has caused stable long-run relationships for the U.K. consumption function to shift significantly. We analyze the relationship between financial liberalization and consumption by evaluating and extending a forward-looking model in which the degree of financial deregulation influences the behavior of two groups of consumers: those constrained to consume from current income only and those able to borrow on the basis of expected future income flows. We show that specifications incorporating such features capture recent U.K. behavior reasonably well and pick up the main shifts in the savings ratio. We also show that changes in the extent of financial regulation cause the transmission of monetary policy to alter in important ways which have implications for the way in which monetary policy should be implemented.

One of the principal ways in which monetary policy can affect the macroeconomy is through its influence on personal consumption. Over the last two decades, personal consumption in most industrial countries has been subject to considerable change. For example, in the United States, the savings ratio has shown a pronounced downward trend, and in the United Kingdom the savings ratio has shown a distinct cycle. As a result, most recent attempts to model consumption behavior in the U.K. have hit two major *We would like to thank Julia Darby, Brian Henry, John Ireland, John Mueflbauer, James NLxon and Andrew Sentance for comments and correspondence. Financial support from ESRC grant no Ll16251013, "Macroeconomie modeling and policy analysis in a changing world," at the Centre for Economic Forecasting, London Business School, London, United Kingdom, is gratefully acknowledged. All errors remain the responsibility of the authors.hurdles: they have failed to pick up tile fall in the savings ratio during tile consumption boom of the late 1980s, and they have failed to pick up the recovery in saving during the early 1990s. Relationships whieh appeared to be stable began to shift, and there was a tendency for "long-run, equilibrium" models to underprediet during the 1980s boom and overprediet in subsequent periods. This phenomenon has inspired a number of studies, such as Attanasio (1997), Attanasio and Weber (1994), Caporale, Sentanee and Williams (1997) and Miles (1997), which highlight the role of the comprehensive deregulation of U.K. financial services during the 1980s and 1990s and the impact this had on housing and financial wealth as major factors in the behavior of the U.K. savings ratio. An alternative view suggests that in many respects the failure of many consumption speeifieations arises beeause they have relied upon baekwardlooking error correction models (ECMs) as representations of the consumption process. For the U.K. case, Chureh, Smith and Wallis (1994a, 1994b) show that within this framework eointegrating relationships representing long-run equilibria can be established quite straightforwardly, but stable short-run dynamic structures are more elusive. The most successful formulations have included additional terms in the dynamic strueture such as the unemployment rate or housing market turnover to piek up changes in finaneial or real activity. However, even in these eases there still tend to be problems in the lag response of the models, and it is clear that no single additional term has been able to provide an unambiguous solution. The limitations of ECM specifications can also be explained in terms of the exposure of the underlying structural forms to the Lueas (1976) critique. Blake and Westaway (1993), for example, show that in a framework where an ECM is solved out from its microeeonomie foundations there is evidence that instability may arise both from the exogenous ineolne process and from institutional shifts in the model's "deep" struetural parameters. They argue that this type of insight can only be gained when the final estimated equations are based on a clear theoretical framework. This is rarely the case in ECM specifications, which tend to use the statistieal fit of the model as the main desiderata. Using eonsumer theory in the design of aggregate functions is one of the ways in which a solution to statistical drawbaeks might be found. Muellbauer (1994), Muellbaner and Lattimore (1994) and Attanasio (1997) provide surveys of some of the current thinking in this area, whieh has highlighted inter alia the role of forward-looking income expectations and financial (eredit) eonstraints in agent consumption allocations. With this in mind, this paper outlines a model with an explicit role for financial variables, such as asset portfolios, to investigate whether these play a role in accounting for consumer behavior and whether they affect the impact of interest ratepolicy. It seems likely a priori that the effect of monetary policy and its transmission to aggregate demand will alter as the financial environment evolyes. The absence of financial variables and forward-looking behavior in many ECM models means that consumer anticipation of changes in credit conditions or in monetary policy cannot in general be picked up. This clearly represents a severe limitation of this approach. In this paper we address these problems by using a particular forwardlooking consumption function, that of Darby and Ireland (1994), in which the degree of financial deregulation influences the consumption of two groups of agents: those constrained to consume from current income only and those able to borrow on the basis of expected future income flows. It is often argued that the comprehensive deregulation of the financial sector in the U.K. during the 1980s has allowed previously constrained consumers to borrow to finance extra consumption. By modeling financial change explicitly as part of the consumption process we are able to avoid the main source of structural instability exhibited by other specifications. We are also able to illustrate how changes in the financial environment affeet the transmission and impact of monetary policy on aggregate consumption patterns. The paper is organized as follows. Sections 1 and 9. outline the modeling strategy and the results of the estimation. Section 3 looks at the performanee of the model in response to a selection of policy shocks over two distinct regimes. Finally, Section 4 offers a summary and some conclusions

In this paper we have looked at the problem of modeling consumption in an environment subject to considerable innovation. We have noted that standard ways o£ modeling consumption in the United Kingdom have been unable to cope with the comprehensive deregulation of the U.K. financial sector that took place during the 1990s. These changes have caused stable specifications previously believed to represent long-run equilibrium relationships to breakdown and so have created huge uncertainty about their predictions for the impact of monetary policy. We look at an alternative way of modeling consumption that admits two types of agents: those constrained to consume out of current income only, and those able to borrow on the basis of expected future income flows. The weight of each type of behavior in aggregate consumption is considered to depend on the degree of financial liberalization in the economy, which varies over time due to exogenous changes in the financial environment. The model therefore allows us to consider how the impact of monetary policy changes in the presence of financial innovation. We show that models which have an explicit role for this type of evolution can pick up U.K. consumption patterns well and can shed light on the transmission and impact o£ interest rate policy. The easier is access to credit within an economy the greater is the propensity for intertemporal substitution by agents and so the greater is the sensitivity of aggregate demand to monetary policy. In the United Kingdom easier access to credit pre-1990 made interest rate policy more effective, and the contraction or tightening of credit conditions after 1990 reduced its impact markedly.