In a comprehensive new paper disseminated by the US National Bureau of Economics Research, Harvard professor Brigitte Madrian discusses the latest thinking on using behavioural economics to optimise policy design.
There are a few items in there with direct relevance to Australia that are worth a bit of attention:
1. Orwellian language works. Madrian praises the re-branding of the UK’s dole as a “job-seeker allowance” and criticises the use of the term”work for the dole” in Australia. The evidence for the power of framing in the allocation of transfer payments is more significant than I had realised.
“Kooreman (2000) finds that the marginal propensity to consume children’s clothing is 10 times larger out of income designated as a “child benefit” than out of other income sources; in contrast, the marginal propensity to consume adult clothing is highly significant for other income sources but is negligible for income from designated child benefits. The labeling of income as a “child benefit” apparently creates in parents a moral obligation to actually spend that money on their children. Similarly, Benhassine et al. (2013) evaluate the impact on school enrollment of a labeled cash transfer program in Morocco that designated the funds for children’s education, although the funds could be used for other purposes. They find a sizeable increase in elementary school attendance by children in families who received the labeled cash transfer relative to children in control households who received nothing. They also find that a labeled cash transfer is as effective, indeed for some measures is more effective, at promoting school attendance than is a conditional cash transfer in which payments are made only if a child does in fact attend school (and is significantly less expensive to administer than a conditional cash transfer program).”
2. There’s a reason the ATO prefers to over-collect tax and deliver tax refunds each year, and it has to do with both loss aversion, and the way human minds frame problems with reference to arbitrary points.
“A natural reference point for taxpayers at the time of tax filing is whether they owe additional tax (relative to what has already been collected) or expect a refund. Engström et al. (2013) find that in Sweden, taxpayers are more aggressive about claiming deductions when they owe additional tax at the time of filing than when they expect a refund, consistent with the predictions of prospect theory. An obvious policy implication is that a tax collection strategy that relies on overwithholding followed by refunds at the time of tax filing may increase tax compliance and total taxes paid.”
3. Madrian is an economist, so she’s au fait with the merits of financial incentives. But the paper emphasises that public finances are in short supply, so it is important to look for policy tweaks that would work as well as a financial incentive.
“Levitt et al. (2012) examine the effectiveness of several different incentive schemes to motivate student performance on standardized exams. They find that giving students a trophy for meeting performance targets, at a cost of about $3 per student, has roughly the same impact on test scores as a direct financial incentive of either $10 or $20, and in some cases is more effective.”
Thinking about whether there is a behavioural economics solution that would substitute for a subsidy may also provide insight to the real reasons for a policy. For example, the government’s proposed Paid Parental Leave scheme – a subsidy for staying at home and looking after a child.
The PPL scheme is expensive, offering up to $50,000 dollars in wage matching over six months. What behavioural tweaks could make it cheaper?
The question is actually hard to answer. Is PPL designed to lift the birthrate? To keep women at home in the first 6 months of life? To return women to the workforce thereafter? To reduce demand for childcare? To support incomes? The fact that whatever the policy problem is can only be solved by this expensive PPL hints that it is in fact a policy in search of a solution.
Thomas the Think Engine 