How much poorer is New Zealand than Australia?
Over at Anti-dismal, Paul Walker links to a paper by NZIER on New Zealand incomes relative to Australia.
In the paper, NZIER states that:
the average living standards of New Zealanders in 2007 were 24% lower than those of Australians (or equivalently, relative to living standards in New Zealand, Australia’s were
32% higher)
However, I am not convinced – not yet anyway. Here’s why:
GDP per capita and average income
The NZIER report relies on the result that average income is the same as GDP per capita. However, in order to make such a claim you have to be clear about what you are stating.
In the case of GDP per capita, we are effectively looking at a “mean” income in the economy as the average.
However, I would have thought that, when looking at average living standards, we would want to look at how the average person fares in each country – which is not what the mean really tells us in this case. Given outliers, and the fact that the distribution of income in the two countries will be quite different, we should be looking at medians.
Furthermore, as we are interested in living standards, shouldn’t we be comparing the actual living standards of individuals, rather than the volume of measured activity in the economy? At best, these two results should not differ – however, why use an imprecise measure when the actual measure is available?
I haven’t got the numbers on me, but I am sure I’ve read that the average person in NZ is better off than from Tasmania (potentially on the Standard – nope on Kiwiblogblog, I commented here as there results were using incomparable data series) – which is why I am so suspicious of this result.
But income is lower
Yes, I do not doubt that for a second. But why should our average incomes be the same as those in Australia. They have different resources, and an economy that has a very different composition – is there any actual reason why we should expect the “average wage” to be the same between the two countries.
We do not bemoan the fact that skilled labour is paid more than unskilled labour – why are we complaining if a country with greater resources per person has a higher GDP per person than a country with a smaller endowment? Ultimately, given the composition of the economies, labour should move around until wage rates are equalised, and any difference will just be a result of the different specialisations associated with different nations.
But the paper also justifies looking at the issue because of migration
When we framed the idea of using GDP per capita appropriatly it became obvious how inappropriate this measure of income growth is in a large number of circumstances (Note: I wonder why they did not use MEDIAN HOUSEHOLD income, which helps account for outliers and the demographic issues). It is especially inappropriate in in the case that the paper uses to justify looking into the income growth, the case of rising “outward migration”.
In such a case, the people we want to specifically look at are the people that are at risk of leaving – the ones who can get a significantly larger wage in Australia as compared to New Zealand. We are not even interested in national income in this case – we are interested in how the compensation of employees (especially in the industries with a large wage gap) differs between these regions – and why.
If we had to look at an aggregate figure, we should look at the MEDIAN average weekly wage instead of GDP per capita – as that allows us to remove “outliers”, thereby giving us a better representation of average earnings from working in either NZ or Aussie.
If we want to pose useful policy recommendations surrounding this issue we should look at these specific industries and try to understand why labour productivity, and therefore wage growth lags in these areas. The question then should be – how does the regulatory environment influence productivity and in these industries and in what ways.
Instead NZIER, after looking at aggregate evidence that does not illustrate the central issue they are discussing, blankly states that if we increased labour productivity we could increase GDP per capita – this is the policy recommendation equivalent telling people that if we add 2 and 2 we get 4. They also tell us that “freer markets” and “better regulation” would help – a bit fuzzy for my liking, given that they haven’t justified any area for these policy to be implemented in?
In my opinion the paper not only doesn’t add value on the discussion about Trans-Tasman migration, it frames the issue in such a way that it appears to exaggerate the problem. I’m disappointed.
You might tell me that they only did this in order to illustrate that there was an issue. I would buy that if they had used AVERAGE household income instead – which is freely available!
But we need to catch up or people will keep leaving – it is as simple as that!
Really, is it as simple as that? I have a few questions here:
- If it is in the individuals interest to move away, what is the problem – it will always be in someones interest to move away, we don’t have to be scared of migration,
- Could the difference be the rest of ACTUAL fundamental differences between our two economies – if Australia has more resources/capital per person then a outflow of NZers will actually solve this “difference”,
- Isn’t the real goal to maximise social “happiness” – as long as we are doing the best we can with what we have what is the issue? Migration numbers are not the ultimate goal of policy are they?
Cross-country comparisons of GDP per capita are notoriously useless – which is why the OECD, IMF, and the World Bank do so many of them 😉 . The reason these organisations use GDP per cap is because of the unavailability of better statistics in many of the countries they want to compare. However, in the NZ-Aus case there is piles of FREE information available.
I am sure that many, many, people will venomously disagree with me – so lets hear it 🙂
Note: I am not defending government policy – however, I don’t see the provision of misleading statistics as a good way to start a debate on policy.
Update: There is something else I forgot to mention. A small open economy that experiences a increase in their terms of trade would see real GDP underestimate the countries income. Why? Higher prices for exports do not increase GDP (they increase the GDP deflator), however they should see an increase in consumption. Now if we buy a big chunk of our consumption goods overseas, then the increase in the terms of trade will see us increase consumption and increase imports. GDP=Consumption+I+G+Exports-Imports. In the extreme case a terms of trade increase will lead to no extra GDP.
Now, an increase in the terms of trade and the corresponding increase in consumption from imports is an increase in real household income (and consumption). The more export oriented the market, the greater this impact will be, and as a result GDP per capita will tend to underestimate incomes in the case where there has been a big increase in the terms of trade – which is what we have experienced. This is because the “income” of our country has risen, even though the volume of production has not changed (it is a price effect after all).
This is another way that GDP per capita is misleading for a country like NZ 😉
to the extent that the resources extracted from ozzie soil and sold overseas at a profit is greater than the profits extracted from the foriegn owned nz banks etc i would agree you have a point.
Unfortunately you dont address the large body of anecdotal but real evidence of semi skilled and skilled labour that can earn a much higher net of tax income in australia.
“Unfortunately you dont address the large body of anecdotal but real evidence of semi skilled and skilled labour that can earn a much higher net of tax income in australia.”
Agreed – I do not cover that, as I am not trying to say that NZers are not poorer than Australians in systematic ways.
I am saying that the above report does not make a very good case for what it is trying to show – namely that living standards are not comparably lower in NZ. Nor does it show that there is a gap that is causing migration (which is due to policy or institutional issues, rather than resource endowments).
It is a lot harder to make a good report than to say why a report is bad. All I am doing is saying why I thought the report was misleading and not particularly useful.
So to be clear, I am not saying that the opposite of what the report says is true – I am just disputing the way the report went about saying it.
If someone magically asked me to do a peer reviewed report on either the migration issue or living standard to would use different data sources than the ones used above, because of the above criticisms. With this counterfactual in my head, I made the above criticisms. However, I do not know what the actual physical results of my counterfactual would have been – as I have not written (or seen) any such reports.
I suspect that NZIER is going to come out with something a little more solid on this issue at some point – I think this was only a warming up of the issue before they got into some serious analysis. They are economists after all 😛
What are the numbers on median incomes, by the way?
I’ll agree with you that medians are more relevant for assessing general well-being, but doesn’t differences in averages tell you something important about likely growth rates? Suppose two countries had identical median incomes, but the average income in the second were much higher. The second country has a mass up at the right tail, leaving the rest about the same, for now. But that right tail mass funds investment and longer-term growth, no? Or do they just sit on mattresses stuffed with money?
“What are the numbers on median incomes, by the way?”
Good question. I might have a hunt around when I get home tonight.
“but doesn’t differences in averages tell you something important about likely growth rates”
I’m not really convinced about the growth argument. Just because a country has a greater annual production does not imply that annual production grows more quickly than other countries – in fact the general belief is that national “convergence” will lead to the opposite impact.
Fundamentally, growth is a function of increases in productivity, technology, and resources. Unless we believe that a more “unequal” distribution of income leads to increased productivity we cannot fundamentally state this (eg if one countries tax system is more progressive than the other countries). The distribution of income is not just the result of policy – it is also the result of the initial distribution of resources in the economy, and as a result we have to make further assumptions before we can relate the distribution of income to growth rates.
I fully accept that, intra-nationally, there is a trade-off between income distribution and growth. However, I am not convinced that we can successfully compare this.
Convergence, sure, but only if institutional quality across the two states is roughly comparable. Don’t most convergence studies correct for institutional quality? If it’s just easier to generate wealth in country X than in country Y because of institutional quality and country X has more rich folks, it’s not necessarily going to be the case that they send their money to country Y in pursuit of the higher marginal product of capital there available.
“Convergence, sure, but only if institutional quality across the two states is roughly comparable.”
True – also old endogenous growth theory destroyed the idea of convergence, pretty much on the realisation that most countries has constant, or even increasing, returns to scale 😛
“If it’s just easier to generate wealth in country X than in country Y because of institutional quality and country X has more rich folks, it’s not necessarily going to be the case that they send their money to country Y in pursuit of the higher marginal product of capital there available.”
Indeed – but that would explain why GDP itself is higher, not why growth rates are higher. Fundamentally, the MPK should diminish over time until they are equalised between countries – it is then growth in MPK that is of concern for economic growth isn’t it? If this is the case, we should be interested in factors that drive productivity etc for sure, but those factors will be country specific – making any international comparisons (of the form provided by NZIER) useless.
Point is that MPk is lower in countries with poor institutions even though capital-labour ratios are lower: the same institutional deficiencies causing lower growth also are associated with there being little point in sending over capital.
All this is somewhat pin-dancing though as it’s not plausible that the degree of difference between NZ and AUS institutions would be sufficient to generate that kind of difference in the first place.
I agree that a lower MPK must rely on some sort of institutional deficiency (or a bias against investing in some country) – however the K/L ratio depends on the aggregate production function, which is definitely different between Aus and NZ, and does not have to equalise even in the face of perfect capital mobility.
However, unless one of us is stating that the MPK is a function of a countries GDP per capita, how does this relate back to whether a country with a higher average income, but the same median income as another country has a higher rate of growth – it definitely has a level of activity, I’m just can’t understand the growth rate argument yet 🙂
Higher ex ante MPk in the country with higher average income, which is what generates the mass in the right tail of the income distribution to begin with. If you do it up right, you even get capital moving from the poor country to the rich country in search of the higher returns!
“Higher ex ante MPk in the country with higher average income, which is what generates the mass in the right tail of the income distribution to begin with. If you do it up right, you even get capital moving from the poor country to the rich country in search of the higher returns!”
But in the steady state the MPK must be equal between the countries – which implies that this effect is only transitory doesn’t it? It explains why total production is higher in one country, sure – but it does not tell us that growth in one country will exceed the other once this higher production level is reached. This relies in overall growth in productivity, which is effectively a random walk with a trend.
However, I think we are in agreement that if two countries have different resource endowments, GDP per capita will differ (stemming from your discussion of “ex-ante” MPK) – even in the absence of policy failure. As a result, looking at international comparisons of GDP per capita as an objective measure relative living standards in countries is a dodgy way of discussing the success/or requirement of further policy.
http://www.marginalrevolution.com/marginalrevolution/2007/12/why-is-new-zeal.html
Might be of some interest..
Has to equilibrate in steady state, sure. But so long as institutional quality here continues to have relative downwards surprises, it can take a while to get to steady state.
Here’s a question for you then: do you seriously think we’re ever in steady state or even close thereto? I tend to think it an ever-moving target.
Thanks for the link Owen – definitely of interest!
“Here’s a question for you then: do you seriously think we’re ever in steady state or even close thereto? I tend to think it an ever-moving target.”
Indeed I agree we are constantly moving, mainly because the steady state is always moving – overall though the steady state is a magnet be are being pulled towards.
An interesting question is “what moves the steady state” – fundamentally technology and institutional issues have to play a large role in this I would suspect – and this movement in the steady state is what truly drives growth.
Sure it would be nice to have used more than just average GDP per capita, but you haven’t cited anything that would convince me that this is a biased estimate of the income gap between the two countries. Australia’s income distribution is less unequal IIRC. And comparing median income just tells you how much richer one person in Australia is than one person in New Zealand, so hardly very descriptive either – at least average GDP gives some idea of the resources available to society as a whole.
Given Australia’s lower employment rate, their output per worker is even higher than GDP per capita comparison would suggest.
“Sure it would be nice to have used more than just average GDP per capita, but you haven’t cited anything that would convince me that this is a biased estimate of the income gap between the two countries.”
Completely fair point – however, this is an issue that should have been covered in the NZIER report during the initial research methinks.
“Australia’s income distribution is less unequal IIRC.”
Indeed it is – which should actually make it spin the other way. A very interesting fact.
“And comparing median income just tells you how much richer one person in Australia is than one person in New Zealand, so hardly very descriptive either”
But it should help compensate for the differences in the income distribution – ultimately I would have preferred research that at least attempted comparing like for like before making policy conclusions.
Ultimately, I took issue with the fact that NZIER took GDP per capita as a given income measure, and then went on discussing “income-gaps” in the knowledge of these technical issues associated with using that measure.
If they can show the same thing with average median household wage I’d be more interested – if they can then go on and show that wage gaps exist in certain industries I would be even more interested (I have actually seen this sort of work from them before – which was part of the reason I was disappointed with this).
This release was supposed to be “public good work” – it was supposed to add to the body of intellectual capital in the country. I just don’t think it stood up to those standards.
Note: I am not saying that NZ is not significantly poorer than Aussie – I am just saying that this study wasn’t a very good way of analysing any such discrepancy.