by Peter Ireland
No. 3, January 2005 - June 2005

Motivation for the Research
With inflation seeming to have stabilized since the 1990s to levels associated with the period before 1960, monetary economists and central bankers have recently rediscovered some of the special problems that can arise under conditions of price stability, chief among them the problems associated with the liquidity trap.

In this paper, the author studies the behavior of the economy and the efficacy of monetary policy under zero nominal interest rates.

Research Approach
Krugman and Svensson have separately reconsidered the idea of the liquidity trap, using state-of-the-art monetary models in which optimizing agents have rational expectations. Absent from these new models of the liquidity trap is the idea of the real balance effect, whereby a change in real balances impacts household wealth and thus affects consumption and output, enabling the central bank to influence the economy even after the nominal interest rate hits its lower bound.

By introducing population growth, as modeled by Weil, to the Krugman framework, the author establishes assumptions under which the real balance effect reappears.

Key Findings

• A real balance effect fails to appear in Krugman’s and Svensson’s models because these models, which feature a single, infinitely lived representative agent, depict economic environments in which government-issued money is not a component of aggregate private-sector wealth. This result stems from a version of the Ricardian equivalence theorem, which also states that in the same models, government-issued bonds are not a component of private-sector wealth.
  
  
• With a growing population, households alive in the present pay only a fraction of the taxes to be levied in the future when the government chooses to contract the money supply. Money becomes net wealth; consequently, an operative real balance effect gives the central bank control over the price level even when the nominal interest rate equals zero; that is, the liquidity trap disappears.
  
  
• The same distributional effects that give rise to the real balance effect can also work to make many agents much worse off under a zero nominal rate than they are when the nominal interest rate is positive.
  
  
• Paradoxically, a zero nominal interest rate is something to be achieved in the models in which the liquidity trap survives; with the introduction of the real balance effect, a zero nominal interest rate becomes something to be avoided.

Implications
The findings suggest that the principal dangers posed by deflationary policies have little to do with zero nominal interest rates per se and even less to do with the Keynesian liquidity trap. Rather, both the problems and their ultimate solutions lie in the mechanics through which deflationary policies are implemented.

Full text of Working Paper 05-3