Wednesday, 27 June 2012


Reading various (excellent) posts by Miles Kimball has got me considering the way that we think about taxation and spending. There’s been a lot of discussion over the last few years of fiscal policy as a macro policy tool for stimulating aggregate demand, but we have to remember that this isn’t the reason why governments tax, spend or even borrow money.

On a basic level, government spending – has two linked but distinct purposes. The primary objective is to correct market failure by providing services that the free market does not – things like defence at the national scale or rubbish collection at the local one. Government spending also serves a redistributional purpose. We live in a society which doesn’t think that the poor and unemployed should starve, so we pay them benefits to make sure that this doesn’t happen. There is, of course, overlap between these two aims. The NHS is a service which is also redistributive, because the poor get the same standard of care as the rich would, and they don’t have to pay for it. Similarly, if you’re being brutally logical, you can define benefits as correcting a market failure; the mass starvation of workers that might one day be part of the labour force would be, if nothing else, bad in macroeconomic terms.

You can justify government intervention from a utilitarian perspective, or you can condemn it from a libertarian one – the lack of a threat of starvation does, after all, create warped economic incentives relative to the free market baseline – but, for the purposes of this post, we’ll assume that the government has to spend money on services and benefits. Of course, that’s not the whole story. Fiscal policy has two sides, and the government also needs to raise taxes to cover the cost of these services and benefits – either now, or in the future to pay off the cost of debt.

Taxes are bad.

That’s a fairly basic principle of economics. Taxes impose a real cost on the economy. They distort economic incentives so that effort is wasted on bad things, like complicated avoidance schemes, rather than on good things, like productive work. More, they encourage people to do less work than would otherwise be the case, because they retain less of the income from doing so.* Sadly, though, taxes are a necessary cost associated with public spending, and public spending is usually a good thing (because bins need to be collected, publicly-provided health is more efficient and mass starvation is bad).

Arguably, if tax collection is necessary, then it should be done as efficiently as possible. That’s why many economists are in favour of broad-based, flat-rate taxes like VAT. If a tax applies just to one good, then less people are going to consume that good than would at the market equilibrium, which can be presumed to be socially optimum in the absence of significant externalities. This means an overall loss of welfare. However, if a tax applies to all goods, then people will make broadly similar consumption decisions that they would have made in the absence of the tax (although the relative effects of income elasticity of demand mean their consumption pattern isn’t exactly the same). The distortionary effects of the tax are therefore minimised.

Of course, externalities do exist. In fact, contrary to the beliefs of libertarians, there are quite a lot of them. This opens up an intriguing possibility. If taxes distort the economy, why not use them to distort the economy in positive ways, reducing the consumption of goods which impose a cost on society whilst raising necessary revenue at the same time? That’s exactly what Greg Mankiw’s Pigou Club, which Kimball recently declared himself a member of, advocates. If we tax socially-harmful things like tobacco, alcohol and CO2 emissions, we get to raise revenue without damaging the economy – in fact, we help the economy, helping to combat issues like smoking and global warming without having to resort to blunter, riskier tools like regulation.

There are some problems with this argument, though. The problem is that economic efficiency isn’t the only goal of taxation. Taxes also serve a redistributive purpose. One of the reasons that we pay benefits is that – up to a point – we want to reduce inequality. You can argue, as the likes of Stiglitz and Krugman do, that inequality imposes absolute costs. Looking at marginal utility of income, it probably reduces welfare, if not GDP.  But whatever the economic arguments are, the fact is that we have a progressive income tax because, like with spending, one of the purposes of taxation is redistribution. Taking from the poor to give to the poor might be more economically efficient than robbing the rich, but it misses the point of why we tax and spend. Ultimately, part of the goal of taxation is that the tax burden should be distributed not just efficiently but fairly.

That might be why the UK government’s scrappage of a planned rise in fuel duty – a Pigouvian tax – has met with so much approval. It’s really quite shocking to see the Cabinet, Tory backbenchers and Labour all in agreement on an issue. You could see this as part of a broader narrative concerning the disconnection between what economists advice and what politicians do. Quite obviously, the move does in part simply reflect political reality, because people don’t like tax increases in general, and especially not tax increases that directly and explicitly affect the prices of basic goods which they regularly buy. But that doesn’t explain why opposition to a rise in fuel duty is so vehement compared to, say, the jump in VAT that the government imposed in January 2011.Basically, it represents the government responding (or pandering, depending on how you feel about the move) to voters who have broadly accepted the need for deficit reduction but don’t think that a rise in fuel duty is fair.

Part of the reason for that is that duty on tobacco, alcohol and fuel duty is seen – with good reason – to disproportionately affect the poor, because the poor spend a much larger proportion of their income on those items than the rich do, so they are hurt more when taxes go up. However, there’s another reason that I think is also important.

The issue is one of fairness within income groups. A fuel duty rise affects those who drive a lot considerably more than those who don’t drive very often. My father drives a pickup, because he is a landscape gardener. He regularly makes 100-mile round trips on the notoriously congested M25. Because of this, he spends a lot of money on fuel, and would be severely affected by a rise in duty. Remember that our concern is equity as well as efficiency. Is it fair that he should be affected more than someone else earning the same income as a secretary? Is landscape gardening somehow detrimental to social welfare – and, if not, then why are we discouraging it? Voters – and, by extension, politicians – are concerned about this, and so they are likely to oppose a rise in fuel duty more than, say, an increase in VAT that would affect everyone equally.

The reason why rises in fuel duty are so much more contentious than similar measures targeted at tobacco and alcohol are complicated. Partly it’s one of blame. If you’re an alcoholic or a smoker, then you’re engaging in a socially stigmatised activity, and (rightly or wrongly) it’s perceived as being your fault. If tax rises disproportionately affect you, then it serves you right. Also, businesses don’t smoke or drink, but they do drive, and fuel duty offers politicians an issue on which they can appease both business and consumer lobby groups at the same time.

Ultimately, while Pigouvian taxes might seem perfect on paper, they do have some very real consequences, and the cynical explanation for why they don’t get implemented isn’t the whole story. It's important to remember why we're doing what we're doing. This is, in fact, one of the reasons why I’m somewhat apprehensive about fiscal policy as a tool of demand management – it makes it difficult to pursue a desirably efficient or egalitarian fiscal policy at the same time as providing stimulus. We tax and spend a lot of money and, given that that is the case, government should always keep its priorities in mind.

*Yes, I am aware that this may be counteracted by people working more to offset their effectively lower income. Nevertheless, whether people work harder for the same income, or less hard for a suboptimal income, it’s a negative effect.

Tuesday, 12 June 2012

Rate Expectations: NGDPLT and Uncertainty

Market Monetarism has been one of the most interesting developments in macroeconomic thought coming out of this crisis, led by Scott Sumner. What he essentially advocates is the Federal Reserve targeting the trend level of nominal GDP. I’m no expert on monetary economics, which is a fiendishly complicated subject, but stabilising the path of nominal incomes seems like a good idea. It would allow for what would amount to a temporary raising of the inflation target during times of depressed output, and a lower inflation target during times of high output. Doubtless Sumner would denounce me for thinking of monetary policy in terms of inflation at all, but it is a way to frame the concept in conventional terms.

So far, so good. But the central question is whether central banks could credibly hit an NDGP level target. This is particularly pertinent at a time when, with interest rates hard up against the zero lower bound, many doubt how much traction further attempts at monetary easing would actually have. Sumner responds to this line of argument by contending that the very act of announcing the new targeting system would change the stance of monetary policy, and would go a long way towards hitting the target with limited actual intervention. Given that monetary policy, especially at the zero lower bound, revolves around expectations, that’s a reasonable point. After all, who would bet against the Fed achieving its target, when it has promised to use all of its prodigious powers to hit that target?

It’s a similar situation to a central bank fixing the exchange rate of its currency. If there is the expectation that the central bank is willing and able to hold the currency at that level, it doesn’t have to utilise its foreign reserves to push up or hold down the value of the currency. Markets are expecting the exchange rate to hit the central bank’s target – and because foreign exchange markets are so reliant on expectations, that’s exactly what happens, even though the central bank hasn’t actually intervened. The target becomes self-fulfilling.

The problem comes when markets start wondering if the central bank is really willing and able to hold its stated exchange rate value. Speculative attacks start, and the same self-reinforcing effect happens, but in reverse – because people don’t think the central bank can hit its target, it becomes unable to do so. The British ERM fiasco in 1992 was an example of this.

The same thing applies to central bank targets. If enough people think that the central bank is limited in its ability to affect the economy, then the NGDP target would become meaningless. Whether or not a sufficient number of relevant individuals believe in the Keynesian-style liquidity trap is another matter.