Thursday, 5 March 2009

Quantitative easing and psychology

Well, it's (almost) official. The Bank of England today is likely to start a formal policy of quantitative easing, buying British government bonds from the private sector in order to increase the money supply.

Stephanie Flanders has a useful post outlining the issues and questions, in which she references (but does not really answer) the most important question: what effect will it have on behaviour? This is a question that can be considered using three theory bases:
  • macroeconomic models
  • microeconomic incentives
  • psychology, behavioural theory and cognitive biases
The macroeconomic models in the current situation give a direction but not a magnitude: QE should increase the money supply and stimulate demand. We are so far outside normal conditions that it's impossible to gauge from the models how strong the boost will be.

Understanding the microeconomic incentives also gets us only so far. More money will be in private actors' accounts; it won't be earning any real return (indeed there is probably a negative real return); therefore there's a slightly greater incentive to spend or invest it than there was before. But is the additional incentive great enough to overcome the disincentives to spending?

An interesting argument here is the assertion that "there is no demand for debt". Some people say that the banks have plenty of money to lend (as evidenced by their central bank reserves for instance) but that companies and individuals want to save not borrow. I think this is too simple, and would point to another complicating factor: the market for debt is not a perfectly efficient one. A deal only comes together with the participation of several parties: equity holders and management in the borrowing company; a manager of some kind at the bank; a credit committee; and whoever is responsible for either the bank's own capital levels, government support for it, or investors in the corporate bond markets. In a fluid market this should not matter and the market price of debt will simply guide the decisions of all parties. But I believe the corporate debt market is not currently acting in a fluid way (and I have a bit of recent experience of this, though it's anecdotal). All of this makes the effect on incentives harder to predict.

The third area for theory is to explore the cognitive responses of the holders of cash.

Flanders touches on this but only briefly. But it has a disproportionate effect on the transmission of money and on the effectiveness of a given amount of QE. If people are more inclined to spend and lend, the multiplier effect will be much bigger.

The effect on decision-making depends on some substantial cognitive biases. Chief amongst these are framing and mental accounting.

Framing means the subconscious cues that influence how a decision is perceived. If a person receives cues that put them in a mood to be more risk-taking or want to consume more, cash in their account is more likely to find its way out. If the cues point towards conserving cash or being worried about the future, they are more likely to save it.

Mental accounting means that people are influenced by whether they consider money to be income, capital, permanent or temporary. If people receive a £500 one-off bonus in August (with an annual pay review due in January), they behave differently than if they receive a £100 increase in monthly salary in August - which rationally has the same effect.

I am writing a longer article on cognitive biases and their effect on propensity to save or consume, so I will leave this here for now. The key message for policy is: don't leave this as a technical exercise in the markets - try to support it with communication that influences behaviour in the way you want.

No comments: