Wednesday, 27 February 2013

The Case for NGDP Targeting

I have half a post plugging NGDP targeting over at Lib Dem Voice.

The Times is reporting that the Treasury is setting up an internal unit to look into the wisdom of revising the Bank of England’s ‘flexible’ 2% CPI inflation target. Mark Carney, the incoming Governor of the Bank who will replace Mervyn King in June, mentioned the merits of adopting NGDP (National Gross Domestic Product) level targeting in a speech last December, launching a frenzy of speculation. 
Considering the enormous effect expectations of future monetary policy changes have had in places such as Japan, it may be that the talk of NGDP targeting is partly behind the substantial rise in the FTSE and fall in the pound since around the New Year. Recently, however, both he and George Osborne have seemed to be rowing back on their support for their policy. Therefore, it’s worth reviewing – as the Treasury is – the case for listening to Vince Cable and targeting the level of nominal GDP instead of inflation. 
When discussing NGDP targeting, too often politicians and commentators see it as some kind of policy of ‘targeting growth.’ To clarify, NGDP targeting requires the Bank to make no assumptions about the rate of real output growth. A better way to think about it is targeting the amount of nominal spending in the economy. 
There are several advantages to targeting nominal spending as opposed to rises in the price level. Perhaps the most major reason is that NGDP is a better indicator of demand in the economy. If there are negative ‘shocks’ to the real economy – a rise in oil prices, say – then it will raise inflation, but that will not be an indicator that the economy is overheating. It would not be appropriate for the Bank of England to tighten monetary policy, reducing import prices by ensuring that people are too poor to afford them. Because a real shock would raise prices but reduce real growth, NGDP would be unaffected, allowing the central bank to ‘look through’ real shocks and avoid tightening monetary policy too much in response to sudden bursts of inflation that are outside their control. 
In 2008, for example, the Bank would have been able to see past commodity price rises and the realisation that perhaps we weren’t as productive as we thought we were, and realise that demand was collapsing. They might have prevented the banking crisis from becoming systemic and leaching through into the wider economy. 
Similarly, in the 1970s the Bank would have been able to see past the supply side effects of the oil crisis and see that accelerating NGDP growth was also pushing up prices and causing a wage-price spiral. The worst of stagflation might have been prevented. The story, of course, cuts both ways – if there are positive supply shocks, the Bank would avoid easing and overheating the economy. 
There’s a fairly strong case that in the early noughties, falling import prices thanks to cheap Chinese production and the strong pound led monetary policy to be too easy, pushing NGDP growth above trend and helping to inflate the housing bubble. NGDP targeting, then, isn’t just a recipe for monetary easing – it would promote stability at all points of the business cycle.
Another advantage of nominal GDP targeting is that it’s much easier to measure than inflation. Inflation requires you to estimate how much prices are rising by. There are several different measures, and all of them give wildly different results, each with wildly different implications for the required stance of monetary policy. The Consumer Price Index that the Bank currently targets is going to be 0.3% higher for the next three years because of the increase in tuition fees, which is not a rise in the price of higher education at all but rather a change in who pays for it. The VAT rise in 2011 created similar distortions that arguably caused the Bank to miss the derailing of the recovery. 
Measuring nominal spending is comparatively easy. Concerns about revisions to the data are overblown because headline NGDP growth would be only one of the factors affecting the Bank’s stance – inflation expectations, asset prices, unemployment and other indicators would also be used to estimate the future path of nominal GDP and therefore the appropriate policy stance. There are also ways of adapting the policy regime to compensate. 
The advantages of targeting NGDP rather than inflation are numerous and sceptics tend to overstate their case. It’s far easier for the government to tighten fiscal policy and deal with the deficit when monetary policy is co-operative; austerity may be hurting the economy right now, but there’s no reason why it has to be that way. Even absent level targeting, which I’ll address in a future post, nominal GDP is the key to a more stable monetary framework which would dramatically reduce the likelihood of future recessions.

A Parable

Imagine an alternate Britain in which the number of licensed taxi drivers is limited to 10000. People wait years, decades even, to get a taxi licence. The process is horribly bureaucratic and costly.

Taxi rides are hugely expensive because of the low number of taxis allowed. Only the rich can afford them. There are a large number of illegal taxis, which are costly and dangerous.

A politician proposes allowing more taxi drivers. They are hounded out of office by the popular press.

"Allowing more taxi drivers will push down taxi drivers' wages! Why should we make existing taxi drivers worse off just so the rich can have cheaper taxi rides?"

"The real problem is illegal taxi drivers! We need to clamp down on illegal taxis, not allow more legal ones!"

"If we allow more taxis, some of those drivers will commit crimes! If anything, we don't vet taxi drivers thoroughly enough!"

"We already have enough taxis to pick up most of the people who want a taxi. Why do we need more?"

Does anyone think we should cap the number of licensed taxis and minicabs?

Sunday, 24 February 2013

How to Normalise Monetary Policy

How to Normalise Monetary Policy
It’s an easily observable fact that central bankers become irrationally cautious once interest rates get close to zero. Even if you believe, as I do, that unconventional policies such as QE or the MC Rule are equally effective means of conducting monetary policy, it’s clear that policymakers would rather operate within-sample, on the familiar territory of short-term interest rate pegs as the monetary policy implement. It’s no coincidence that the developed countries that have succeeded in keeping NGDP growth mostly on target – Israel and Australia – have avoided hitting the dreaded zero lower bound. There's also the concern that extremely low interest rates can sustain 'zombie companies' and prevent the structural adjustment of the economy. As the debate begins to turn to unwinding the large expansion in central bank balance sheets since 2008, it’s worth considering the various strategies that might work for countries that want to exit the zero lower bound and thus escape the scary world of unconventional monetary policy.

Fiscal Policy
Perhaps the simplest and most conventional way to get away from the ZLB would be for the fiscal authorities to step up and borrow more. Government borrowing would reduce the amount of funds available for private-sector borrowing. In normal times, this would have negative effects, raising the cost of borrowing for the private sector and so reducing private sector investment – the ‘crowding out’ effect. What the zero lower bound means, though, is that there is more money available to be lent than the private sector ‘wants’ to borrow. The market is failing to clear because interest rates cannot fall below zero. Therefore, when the government drives up the cost of private borrowing, it pushes the equilibrium interest rate up into positive territory. This allows monetary policy to act without the constraint of the zero bound – and hopefully, when the economy returns to growth, the equilibrium interest rate will stay above zero.

The problem with using fiscal policy to raise the equilibrium real interest rate is that it’s a short-term measure. The government can’t borrow and run deficits forever. The problem encountered in Japan, which policymakers in other developed countries may soon run into, is that fiscal expansion is limited by the government’s long-term budget constraint – it can’t run huge deficits forever. If the low equilibrium real interest rate is the result of a long-term secular decline, rather than resulting from short-term factors, then other policies will have to be pursued to permanently escape the zero bound.

We’re currently at the zero bound for a mixture of short and long-term reasons, and it remains to be seen whether we’re stuck permanently. It depends a lot on the long-term growth rate of the economy – if there really is a Great Stagnation in productivity then it’s far more likely that we’re stuck forever. It’s also quite hard to tell how much of the decline is demand for safe assets following the financial crisis, which might reverse if and when the economy returns to growth. Part of the long-term decline might also be because of the ‘savings glut’ in East Asia, which might be expected to reverse itself as an aging population starts to draw on those savings to sustain itself in retirement. There’s evidence that this is already increasingly happening in Japan.

Higher Inflation
If the equilibrium real interest rate cannot be raised, then it might be practical to raise expected inflation, so that the nominal interest rate consistent with that market clearing real interest rate becomes positive. Economists assume that the welfare costs of higher inflation are substantial and nonlinear, but it may be that a long-term trend inflation rate of 4-5% imposes costs which are small relative to the potentially enormous benefits of efficient monetary policy. It might be worth putting up with modestly higher inflation in order to avoid the nasty hysteresis effects and human cost of elevated unemployment associated with bad monetary policy and the zero lower bound.

However, adopting a higher target for NGDP and inflation might be politically difficult for central bankers and the governments which set their mandate. Governments will not want to be accused of a return to stagflation, or of trying to inflate away the country’s problems. Higher inflation may pass a cost-benefit test, but there are costs, and there might be better ways of escaping the zero bound.

Increasing the Marginal Productivity of Capital
The easier way to solve the problem of the zero bound would be to raise the long-run growth rate. That would increase the expected value of new investments and also expectations of future income, raising the willingness of both firms and consumers to borrow. Of course, raising long-term growth seems pretty difficult. Economic opinion is divided on the best ways to promote growth, and even those they near-unanimously agree on, such as freer trade and looser planning regulations, can be extremely difficult for politicians to implement. But long-run equilibrium real interest rates don’t depend only on productivity growth, they depend on the overall growth rate of the economy. That suggests one foolproof tactic for escaping the zero lower bound – increased immigration.

The immediate effect of an increase in immigration would be to lower the ratio of capital to workers. That increases the value of building new houses, new factories and new office blocks – which in turn pushes up the demand for – and cost of – borrowing. Even if productivity is not increased by immigration, one side-effect it would have would be to help normalise the monetary policy environment. There are a host of reasons to favour dramatically freer migration, but improving demand-side policy in the medium run is another very good one.

Wednesday, 20 February 2013

Doctors Without Boundaries

It seems that ‘medical experts’ are once again recommending the imposition of radical policies in an attempt to improve public health. This is a perennial problem – doctors see a real and growing medical problem, and then they appear in the media with some mixture of tax and regulatory policies that might help to deal with the problem. The issue is that doctors are not economists. They often recommend either policies that would either prove impossible to realistically implement – like Denmark’s recently-repealed fat tax – or policies that would deal with the problem in an extremely expensive and intrusive way. There are a few different reasons why this might be happening.

It could be an example of salience bias. Doctors are on the frontline of the war against obesity. They treat people who are dying of obesity-related conditions, or have their quality of life reduced by debilitating diseases like diabetes. What they don’t see are the costs of their policy recommendations. They miss the costs of implementing and enforcing the taxes and regulations they favour and the possibility that those taxes will be avoided and regulations ignored. They do not consider the regressive impact of taxes on products such as food and tobacco. They see the people dying of heart disease, not the people whose freedoms are curtailed by regulations, the businesses who get tangled up in excessive red tape.

From a more charitable perspective, medical experts may simply be trying to shift the Overton Window. Taxes, especially regressive ones like alcohol duty that consumers perceive as hitting them directly in the wallet, tend to be unpopular. The smoking ban has got less controversial as the number of smokers has fallen, but there are still vocal, concentrated minorities who want the right to smoke in pubs and who tend to be a lot more interested in supporting it than apathetic non-smokers are in opposing it. By advocating radical policies to try and improve public health, doctors increase the perceived political feasibility of moderate, reasonable taxes and regulations. For this reason, it might be optimal for them to advocate extreme government interventions even if the intervention itself would not be optimal.

Doctors’ paternalistic interventionism can be extremely annoying. But abrasive libertarians can be even more so. 

Monday, 4 February 2013

On Pandemics, Immigrants and Wages

Between 1347 and 1351, an outbreak of what was probably bubonic plague swept Europe. Estimates vary, but the disease probably killed 30-60% of Europe’s population, with the death toll in some areas of Italy, Spain and France reaching as high as 80%.

One interesting economic consequence of this tragedy may have been a large rise in wages. There were fewer workers being supported by the same amount of agricultural land as before. This was effectively a rise in the ratio of land to labour, which raised the marginal productivity of each worker, and therefore wages.* In some sense, therefore, the people that survived the plague greatly benefited from the reduction in population. The flipside of this, of course, is that increases in population reduced average living standards. The value of land and therefore the leverage of the landowners would go up, allowing them to exploit the workers more, and with decreasing marginal returns to labour, output per worker would fall.

This scenario has interesting parallels with immigration. Opponents of immigration suggest that higher immigration will have the reverse effect of the Black Death. The supply of labour will go up, the ratio of capital to worker and therefore productivity and wages will fall, and people will be worse off. The immigration literature, on the other hand, does not show much evidence of these effects.

This is because the world we live in is very different to the world of the 14th century. Back then, almost everyone worked in the agricultural sector. Particularly in Europe, virtually all of the capital stock was in the form of land.** Obviously, it’s (almost) impossible to make more land if we have more workers, so adding more workers in this scenario results in a falling capital-to-worker ratio. But today, most capital is not land or even natural resources. Most capital takes the form of office blocks and machinery and educated workers. Increasingly, capital is crystallised labour, not a fixed and finite resource. What that means is that when the marginal productivity of capital rises, the rising price of capital provides a signal to the market that it should make more. If the number of workers doubles, we can simply build twice as many schools and office blocks and factories, and (in the long run) everyone is almost as productive as before.

Then there is the basic Smithian economies of scale argument. There was very little specialisation in medieval Europe, because most people were farmers. Adding more farmers just meant more people doing the same job. But today’s industries have many opportunities for specialisation, so the benefits to larger, deeper markets are substantial.

The world of the Black Death also had many fewer spillover effects from immigration. Economies of scale were low. Serfs’ productivity wasn’t dependent on the productivity of the other serfs on the farm. Now, however, we live in a world that looks much more like Kremer’s ‘O-Ring’ modelof development, where worker productivity depends on the success of their colleagues. The implication of this model is that grouping large numbers of people together in one place agricultural labourer’s productivity does not necessarily affect another labourer’s productivity all that much. It’s notable that in sectors like academia and technology, incumbents are very supportive of immigration, even though immigrants would compete with those workers. This is likely because workers in these high-skill industries complement each other and make everyone more productive. More academics also means more people working on the technological discoveries that accelerate the development of humanity.

Population density also has different implications than it once did. Too many people clustered together in the 14th Century meant breeding grounds for squalor and disease. These days, density means reduced impacts on the environment, increased social and economic opportunities, and (most importantly) more great restaurants.

‘It’s a global market’ is often a weasel phrase deployed to assert that ‘this time is different,’ painting reasonable economic concerns as antiquated scaremongering. But the world really has changed. It’s not the Malthusian dystopia it was prior to the industrial revolution. Worryingly often, proponents of immigration restrictions treat it as though it were one.

*Under serfdom, marginal productivity might not have had much effect on wages. The shortage of labourers, however, meant that there was less incentive for landowners to enforce the system, and far more incentive to compete with each other to attract workers with higher wages. So serfdom began to break down during this period. The Black Death had a lot of fascinating historical and political implications outside of the economic story I’m focusing on today.

**The implication of this is that big exogenous population declines would have had less of an effect in places like Egypt and China, where farming methods relied much more on capital-intensive methods of irrigation. I’d be very interested to see data on this.

PS: Happy #ImmigrationTweetDay, all!