Thursday, 8 April 2010

Behavioural politics, day 2 of 30

Yesterday, the Tories won the first behavioural battle of the election campaign.

Policy of the day is national insurance. Labour has committed to a rise of 1% in the employers' national insurance rate from April 2011 and the Conservatives have said they'll cancel this increase (for readers outside the UK, national insurance is our payroll tax, and the employers' component is the part paid by the company, as opposed to the part which is deducted from the employee's salary like income tax. Currently the rate is 12.8% and will rise to 13.8% under this policy).

When the Tories announced this policy, around twenty chief executives of large companies wrote to the FT in their support. This was followed by another three on Tuesday and, yesterday, thirty more.

So how does this play out behaviourally?

I mentioned a couple of things yesterday which remain true: the Tories gain social proof from the support of senior businesspeople, and the recency bias means that the issue may have peaked too early, as the voters will be tired of hearing about this by election day and will have forgotten most of it. What other factors are relevant?

Saliency is important. The national insurance impact is fairly immediate and obvious (at least to the employers who are leading the charge) while the public spending cuts that will pay for it are diffuse, non-specific and distant. Everyone can point to the extra £10 million (or whatever) that Marks & Spencer will pay, but nobody knows which civil servant will be fired or which at-risk child will not be visited by a social worker.

Thus, framing becomes critical. This issue has been successfully framed by the Tories as "Labour's tax rise". Labour is trying but not really succeeding at reframing it as "more Tory cuts in critical public services".

As a background to the framing question, it's fascinating that neither party can get away any more with promising unfunded tax cuts or spending increases; the British polity is still obsessed with debt, and the public debt must be brought down at all costs - hence the Tories have been forced to offer spending cuts to neutralise the tax cut, instead of being allowed to claim that economic growth will pay for it.

Money illusion is a key question here, because that's what makes employer's national insurance such an attractive tax to increase. In the long-run, classical economists argue, all taxes of this kind are paid by the employee; they raise the labour cost and therefore reduce the salaries that companies offer. In the short term, however, this isn't the case at all. Employees are very unwilling to see a pay cut, so employers need to bear any raise in the short term. Therefore, employer's NI is immediately visible only to employers, which are a very small voting group (perhaps 1.5% of the population or 2-3% of likely voters).

This issue can go away quickly in a time of high inflation - it's quite easy for an employer to raise salaries by 3% against a background of 4% inflation. Or even easier, to delay the 4% rise by three months, which is exactly equivalent. This is a classic picture of money illusion - people focus on the numeric monetary amount of their wages and prices and not on purchasing power. In the meantime, the cost is borne either by company shareholders - or by employees who would otherwise have been hired, if short-run elasticity of labour demand is high.

So the economic impact is complex and changes over time, and the costs shift from employers, then to consumers through price rises, and then to employees as salaries change. This makes the outcome very uncertain, which surfaces in our probability biases - wherein we ignore low-probability outcomes and focus on what we think is the most likely result. If there is no dominant likely outcome, we may simply ignore or underweight the future altogether. This in turn contributes to the saliency of the tax and the lack of saliency of the ultimate spending cuts.

This also impacts hyperbolic discounting. We can see the costs that are immediate, but those in the future (like repayment of debt or cutting public services) are foggy and therefore feel less important. This makes it all the more impressive that the Tories have managed to turn a tax rise in 2011, which will directly fall only on employers, into an immediate electoral issue.

Yet another impact is from the idea of loss aversion. This policy, like many economic policies, involves trade-offs. But any time you trade off two things, a gain versus a loss, the loss looms larger. Typically, in most experimental measurements of loss aversion, people weight a loss about twice as heavily as an equivalent gain. This acts to create an innate conservatism because people resist changes more than they should.

A subtle issue arching over this debate is the idea of approaching rationality. Many cognitive biases shrink or disappear with deliberation, discussion and experience. This is behind the idea of deliberative democracy and numerous transparency mechanisms. This would argue that if Labour feels on the back foot over this issue, it should keep it in the public debate and explore all the details, encouraging the voters to approach a rational view, rather than (as it is probably tempted to do) try to shove it under the carpet and move onto the next thing.

But time is scarce, and most voters won't participate in the deliberation, so this may not be the optimal approach. Labour may just have to let it run out.

The Lib Dems have helped Labour a little here by coming out for the increase. This isolates the Conservatives and could help Labour to paint them as the party of self-interest. And, if the Lib Dems agree with us, Labour can say, our policy isn't just our own idiosyncratic business-bashing. This is Labour's own version of social proof. Whether the Lib Dems have done themselves any favours is another matter...

Ratings: Conservatives 8/10, Labour 5/10, Liberal Democrats 4/10.

2 comments:

Min said...

"This makes the outcome very uncertain, which surfaces in our probability biases - wherein we ignore low-probability outcomes and focus on what we think is the most likely result."

This kind of narrow focus is quite common, and, I suspect, related to conceptualization as forming prototypes. You also see it, I think, in pseudo-rationality, such as reasoning to the most likely conclusion, seeking the most likely explanation, and maximizing single values, such as likelihood or expected utility.

Phil said...
This comment has been removed by the author.