Brad de Long joins the debate over the gains from trade, responding to earlier posts by Kieran Healy and me, with a typically thoughtful and thought-provoking contribution.
I’m hoping this will be the start of a sustained debate/discussion, and I’m going to start by responding to a simple point. Brad says “1.24 percent of current consumption is nothing to be sneezed at ” and he’s right. It’s about $A 8 billion per year, which is a lot of money. For example, assuming government got about half of the gain this would be enough to restore most of the cuts made to post-secondary education over the past decade.
Brad goes on to make an argument I’m less happy with.
In the context of the Australian economy today… Gourinchas and Jeanne’s numbers say that (at a five percent per year safe real interest rate, and with a three percent per year economic growth rate) the value of international capital mobility to the Australian economy is on the order of a one-time present of some 400 billion $A.
To spell it out, with Brad’s numbers the present value of any flow that grows in line with GDP is fifty times (1/(.05-.03) its initial annual value. Multiply the initial $8 billion by fifty and you get Brad’s $400 billion.
I have two problems with this. The first is that, with such a low effective discount rate (2 per cent) a lot of these gains accrue a long way in the future (about half the PV refers to the period after 2040) and I think the impact of any given policy change is hard to predict that far into the future. For example, it may turn out that an approach taken to capital market liberalisation today turns out in 20 years time to preclude some better arrangement that would yield greater benefits.
The second is more important. Suppose, you think there are costs of capital mobility that outweigh the 1 per cent benefit. To take a really simple illustration, suppose you believe that capital mobility destroys national pride and that national pride is worth more than $8 billion per year. It doesn’t alter the argument to say that the benefit of capital mobility in PV terms is $400 billion. If your willingness to sacrifice consumption for national pride is proportional to your income, as seems reasonable, the cost of giving up your national pride can be multiplied by the same factor of fifty to get a present value greater than $400 billion.
The present value conversion is only useful if we are comparing a long-lived flow of benefits to a once-off cost, for example, the need to shift workers into more capital-intensive industries to take advantage of a capital inflow.
An issue where this kind of comparison is important is that of the cost of squeezing inflation out of an economy. Monetary hawks sometimes argue that you shouldn’t worry too much about the unemployment and loss of output associated with a very tight policy because the present value of a permanent reduction in inflation will nearly always outweigh any temporary losses.
I disagree for a couple of reasons. From my first point, I doubt the claim about permanence. The contractionary policies of the 1980s have been followed by a decade or more of low inflation, but it’s easy to see the possibility of a resurgence in inflation in a few years time, particularly in the US.
On the second point, it’s not at all clear that the economy returns to its long-term growth path after a recession. There’s clearly some rebound in the typical recovery, but if you look at an economy like New Zealand, where hawkish monetary policy produced a series of recessions in the 80s and 90s, it seems as if a fair bit of the output loss is permanent, or at least long-lived. There are similar points to be made about unemployment and hysteresis.
John Quiggin 
>>The present value conversion is only useful if we are comparing a long-lived flow of benefits to a once-off cost,
John: take an average of Australia and NZ’s growth rate over the last 10 years or so and see who comes out on top. NZ has a volatile business cycle as befits a small, open and supply-constrained economy. Take an average across the cycle and NZ’s growth rate is very respectable. What’s more, according to the IMF and OECD methodologies, the NZ economy has been growing closer to potential than Australia’s, taking an average over the last five years or so. The idea that tight monetary policy has held back the NZ economy is a widely held prejudice that does not withstand much scrutiny.
Australia has outperformed NZ over the past ten years and even more so over the past twenty years, the period to which I referred. For a source that can’t possibly be accused of bias, I refer you to Don Brash governor of the RBNZ throughout the period I’m describing. He clearly concedes that NZ growth has been lower.
Defending his performance he says “But why, some New Zealanders and even some Australians ask, has New Zealand not grown quite as strongly as Australia?” and proceeds to say that it’s structural factors, not bad monetary policy.
Brash is entitled to argue for his explanation of the gap in GDP growth. But to imply, as you do, that the gap doesn’t exist is just plain wrong.
Are we talking about GDP growth, or growth per capita? I meant the former.
I once did some simple modelling involving the composition of Probability Generating Functions. It suggested – not proved – that in an unmanaged system that was growing, risk grew at a greater exponential rate. On a large enough scale, philosophically anything can be considered “unmanaged”, as all the management comes from within the system and so the system taken as a whole is free running. Therefore the “safe” bet for entire economies over such long time spans is, that they won’t achieve those sustained growth rates because the risk will eventually trigger something that will disrupt any trend.
Of course, you tend to get that sort of result from PGFs anyway – they nearly always predict long term zero results and only offer quasi-stable stuff.
Incidentally, I believe my own earlier point about free capital mobility allowing Australian revenue streams to be bought out. Basically, that is another continuing cost with a cumulative effect, so it can be presented in the way JQ just presented his objections. I believe it is real, and not a matter of putting a subjective value on things like his example. And yes, I know it wouldn’t happen if only Australians saved more – but other things being equal they don’t, and there is a higher barrier to clear because the USA has the reserve currency.
Stephen, you get the same answer either way. Australian population growth was a bit faster so the difference is bigger with GDP than with GDP per capita.
John
I saw the Romer paper as an NBER DP, later published I believe in the Journal of Development Economics. My recollection is that Romer was talking about the welfare benefits of new goods made possible by international trade.
John
Maybe this paper says more about the deficiencies of the neoclassical model than the costs of restricting foreign investment. De Long seems to think so.
You don’t seem to have to much trouble finding fault with the neoclassical model when its apparent policy implications are those you don’t like – say, deregulating financial markets. None of the problems that come from deregulated financial markets are predicted by the neoclassical model.
I mean, deregulating domestic financial markets.
I should try to be clearer.
The basic neoclassical model implies that domestic financial markets should be deregulated. When that turns out badly, many people say, * it’s only to be expected. That model omits crucial features of bahviour in financial markets, like herd behaviour. The neoclassical model should not be used for policy judgments.*
Which is fine. But you can’t then use the same neoclassical model as authority when it turns out to have convenient policy implications, like restricting foreign investment is no big deal.
If the neoclassical model is flawed, or incomplete, then some other guidance for policy is needed.
Uncle Milton My understanding is that the argument works like this. Many influential institutions argue for the opening of capital markets (I think the fuss Malaysia’s response to the the pacific crisis provoked is good evidence of this). These are typically underpinned by neoclassical theory. Anyone arguing against such reforms or freedoms will tend to find themselves arguing about whether or not the policy would promote growth and the theory clearly says it does so you are in danger of taking on the whole theory.
If instead you can say well we are talking about a small effect and the theory is clearly only an approximation and there certainly are real risks attached to it you need not become an economic heretic.
I think the only sweepingly general objection to this argument is Prof DeLong’s point that 1 per cent may look small but over an extended period it is not. However even that has been debunked by the Daniel Davies and Prof Quiggin. To whit, a short delay would not be important, long term capitalisation is problematic and uncertain and for such figures to be fair, they would also have to capitalise the costs.
In short wide acceptance of the results of the paper ought to leave each case to be argued on its merits rather than dctated by a general theory. There is gorwth to be had but the numbers are in anybodiy’s reckoning too small to automatically trump other considerations. I don’t think anybody needs to give up their position to find this result useful. Potentially it is a vndication of neo classical economics because working through the numbers makes it less wrong than prejudice might suppose.