Monetary Policy Transmission through the Consumption-Wealth Channel. (Session 2: The Macroeconomic Environment)

Abstract

INTRODUCTION Asset market values react to economic news and policy changes, and consumers react to changes in asset market values. The consumption-wealth channel of monetary policy spells out this mechanism: changes in monetary policy affect asset values, which in turn affect consumer spending on nondurable goods and services. This paper attempts to quantify these linked effects. After sketching the evolution of research on the wealth channel, we turn to evidence of its size from a number of large-scale econometric models. We then estimate a small, structural vector autoregression (VAR) under some identifying assumptions to provide our preferred estimates of the wealth channel. We find that evidence of an important wealth channel for monetary policy is scant. However, there is some role for a consumption-wealth channel in the large-scale models--with considerable variation as to its size. Finally, the structural VAR framework is found to show little or no sign of a consumption-wealth channel. EVOLUTION OF THOUGHT ON THE WEALTH EFFECT The wealth channel has deep roots in the literature on monetary policy and economic stabilization, reaching back at least to the earliest literature stimulated by Keynes' General Theory. Early on, Gottfried Haberler and A.C. Pigou noted that changes in consumer spending generated by countercyclical changes in the value of the money stock could help provide an automatic stabilizing force to an economy subject to inflationary and deflationary forces (see the discussion in Gilbert [1982]). Subsequent work, notably by Modigliani (1944, 1963) and Patinkin (1965), elucidated the conditions needed in the money, goods, and labor markets through which this real balance could stabilize the economy at full employment. (1) Other work of Modigliani and collaborators, in particular, expanded this theoretical literature on the balance effect into a full-blown analysis of the impact of wealth changes induced by monetary policy (as opposed to the passive change in the balances examined in the earliest literature). Modigliani's life-cycle model of consumer spending emphasized the critical role of household wealth in determining spending on nondurable goods and services (Brumberg and Modigliani 1954, 1980; Ando and Modigliani 1963). Parallel to this work on the life-cycle model of consumption was the development of the Federal Reserve Board's econometric model of the U.S. economy, which specifically traced the connection between changes in monetary policy instruments and changes in asset markets (de Leeuw and Gramlich 1969; Modigliani 1971). (2) Point estimates from this model (Modigliani 1971) suggest that roughly one-half of the impact of monetary policy changes on economic activity through time periods of policy interest could be attributed to changes in spending arising from policy-induced changes in stock market values. (3) Aside from the quantitative evidence, it is possible that the qualitative importance of the wealth channel for policy analysis grew during the 1970s and 1980s. In the 1960s and early 1970s, some of the most obvious impacts of monetary policy were evident through monetary policy's effects on thrift deposit inflows, mortgage lending, and homebuilding. This was an artifact of the Regulation Q ceilings on deposit rates and the illiquidity of mortgages. With the removal of deposit ceilings and the development of secondary mortgage markets, this mortgage credit channel faded as a dramatic sign of monetary policy effects. However, if the traditional econometric model evidence of a strong consumption-wealth channel remained intact, there would be obvious effects (and guideposts along the way, in the form of changes in long-term rates and stock market values) of policy changes. In the absence of particularly large monetary policy changes in recent years, specific discussion of the wealth channel of monetary transmission has diminished. …

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 36 citations exist upstream.