Vox recently published a list of “surprising maps” that documented counterintuitive or little-known facts about the world. Number 15 on the list was this map, which shows that 50% of US GDP is produced in a mere 23 urban areas:

US economic activity split in half

This map shows the economic importance of cities, which are tremendously productive precisely because they concentrate a lot of skills, ideas, and capital in a small area. But surely things are different in New Zealand due to our much more agriculturally-based economy?

I was curious about this, so I got some data from Statistics New Zealand to see how economic activity is distributed in NZ. First, I got the Regional GDP statistics for 2013, which break down the economy at a regional council area. Second, because some regional councils are much bigger than the cities they contain – think Canterbury versus Christchurch – I used 2013 Census data on employment at a detailed geographic level to proportionately allocate out regional GDP based on the share of regional employment in a particular area (see footnote 1). Finally, I grouped the data up by main urban area – essentially, city boundaries plus satellite towns like Pukekohe and Rolleston.

The results are shown in this map. The majority of New Zealand’s economy is located in its three main cities – Auckland, Wellington, and Christchurch. There is more economic activity happening within these three orange blotches than outside of it:

City share of NZ GDP map v2

All together, 56% of New Zealand’s GDP is produced in a mere 0.9% of its total land area. While rural and urban economies are interdependent – they sell goods and services to each other – this data highlights the degree to which New Zealand’s economy is now an urban economy. This won’t change any time soon – if we are to generate new sources of wealth in the future, it will happen in the cities.

I also took a look at the GDP produced in New Zealand’s smaller cities. Overall, the 15 largest urban areas in New Zealand are home to over three-quarters of our national economy. The full results are shown in the following table:

City share of NZ GDP table v1

One other interesting fact drawn from this table is that Auckland’s economy outweighs, by a large margin, the non-urban economy (“Rest of NZ”).

Footnote 1: This is a fairly simple approach to allocating out regional GDP that ignores the effects of:

  • Differences in industrial structure between cities and adjacent rural areas – cities tend to be home to higher-productivity industries such as professional services and manufacturing
  • Within-industry differences in productivity that depend upon location – after controlling for firm characteristics, businesses located in denser areas tend to be more productive than similar firms in other areas.

It’s certainly possible to account for these well-documented effects. In my day job, I develop methods to take a much more detailed look at New Zealand’s economic geography. (But I prefer to get paid for that rather than doing it for free on a Sunday morning!) Not accounting for differences in industry structure and agglomeration economies means that I have probably underestimated the share of New Zealand’s economy produced in cities.

Share this

63 comments

  1. New Plymouth is an interesting outlier as the only smaller centre with share of GDP higher than share of Employment. Are we to assume Oil and Gas pays well? (yes I couldn’t get around that horrid pun)

  2. GDP is a pretty crappy marco-economic tool. In this case, relatively high economic activity is just a reflection of population density so its not really telling us much. Also, some time ago the metrics that can be included in GDP calculations changed, and loans etc are now included. So if the bank lends me $500K for a home, then that is included in GDP, so economic activity is also conflated by property markets/bubbles etc.

        1. So you are saying if Fonterra relocated its head office from Auckland to Hamilton we would suddenly become the least productive city/region in the country (per capita)?

        2. I think what Patrick is saying is that if have have numbers presented as fact or without an obvious statement of opinion, they should be supported by evidence or a reference to where the number came from.

          A lot of the public debate doesn’t seem to be well supported by well researched evidence, particularly motorway/roading projects (Puhoi-Warkworth I’m looking at you), so expecting supporting evidence of claims seems to me to be reasonable.

        3. All I am saying is that Fonterra turns over ~ $20bn annually with the majority of that being for the export market. If all of that was included in Auckland’s GDP by virtue of the head office then it would constitute a very high percentage of the city’s GDP.

        4. James if you read the post carefully you’ll see that Peter clearly states that there is no head office capture effect in the data he is using; his last line states:

          ‘Not accounting for differences in industry structure and agglomeration economies means that I have probably underestimated the share of New Zealand’s economy produced in cities.’

        5. My original post was intended as a reply to Bryan at October 13, 2014 at 1:07 pm. I didn’t realise it was in the wrong place until now, which is why it might not have made any sense!

        6. James, you’re correct about Fonterra’s turnover, but its contribution to GDP is much smaller (the entire agricultural sector in NZ has GDP of around $11 billion), and it won’t be allocated to Auckland except to the extent that value is created there. It’ll really just be the head office functions that are counted in Auckland.
          Whether you look at regional GDP, employment, or incomes, you’ll get fairly similar results to what Patrick has shown here.

    1. “relatively high economic activity is just a reflection of population density” – So agglomeration benefits from large cities produce lots of economic activity?

      Yes, that would be the point of the post and all the other posts about why cities are so important to growing the economy.

      1. No, not really. Its saying that increased GDP relative to other areas is simply a reflection of density. Density in of itself doesn’t contribute to increased economic activity, it simply pools it.

  3. Do StatsNZ allocate the GDP generated by Fonterra to the individual processing plants or the company head office, i.e. is it sophisticated enough to apportion the GDB between the farm, the processing plant, and head office? Ditto for companies with regional plants or offices (e.g. is Steven’s home loan allocated to the GDP of his home’s region, or to the bank’s region)?

  4. How is GDP per location identified in this survey? For example, where does Fonterra’s contribution to GDP sit? In Auckland, where the registered head office is, or it spread throughout the dairying regions? It would seem very hard to take that second approach, and the first approach is clearly nonsense, so I suspect this data is probably fairly meaningless.

    1. It’s the latter apparently, allocated to the region of primary production. Interesting to see if Auckland gets anything from that, presumably the head office does add some value to the product!

  5. This poses a number of questions. The first being what is actually measured by GDP. Insurance, banking etc aren’t actually generating any income, they’re just passing existing wealth around surely. What is the measure of “new” economic activity, e.g. activity that has increased the overall wealth in NZ?

    1. To paraphrase Bill Clinton from a comment a few years back, “Its the Economy, stupid”

      No seriously, GDP is the measure of NZ Incs economic activity, and thats economic as in money/”Economists” not economic as in “thrifty” or “wise”.

      Yes Banking and Finance in insurance don’t “increase our exports” which is what you are conflating GDP is a measure with. But they do contribute to the overall output of NZ in many ways.
      For instance, banking and insurance lets us build more houses and sell exports to our overseas markets more easily than they would have if we had to do things the old fashioned way without them.
      [Can you for instance imagine your bank lending you a mortgage on your house and land if you couldn’t also insure it against fire/slips etc?]

      Many people consider the emergence of these third tier economic activities – like banking, finance and insurance as the hallmarks of a developed economy i.e. first world.

      Yep, GDP is not a perfect measure, and comes with a lot of downside effects – like ruining your environment in an unsustainable manor by chopping down all your trees for exporting as wood chips, burning all your coal/oil reserves as fast as you can or polluting all your rivers, or even buying lots of SUVs from overseas and then fuelling them up with imported fueld – all these cause your GDP to go up in the short term, and then the eventual spending of tonnes more money on cleaning up the mess, e.g. dirty dairying (or CO2 pollution or replanting all those now denuded hills with more trees) also increases your GDP even further.

      Now, GDP doesn’t attempt to assign “goodness” or “badness” to your GDP components and any $ of GDP is just as good as the next one as far as GDP is concerned.

      But as a simple measure of where “stuff happens” in your economy, its as good a measure as anything else we have, and has a nice simple yardstick for comparative purposes the $.

      Of course, equating a high GDP with “sustainability” is a different thing altogether, which is why GDP is flawed in that context.

      And as Bill English said NZ as a country can’t get rich buying and selling houses to each other, but we sure could put our GDP through the roof if each year, as a nation we did exactly that – sold and then bought houses off each other.

  6. In reply to comments from JamesS, Bryan, and nick1234 questioning the Regional GDP statistics:
    First off, this is why I provided links to the underlying data. You can follow the link to the Regional GDP release and read a description of their methodology.
    Second, Fonterra’s entire output is _not_ allocated entirely to the location of their head office. That would be ridiculous. Here’s how the process works:
    * Statistics NZ surveys a representative sample of all NZ firms, including just about every large firm, in their Annual Enterprise Survey. It aggregates and weights these results at a national level to produce the _national_ GDP statistics, which don’t care about the precise location of all activities.
    * In order to produce regional GDP statistics, some additional adjustments need to be made. For companies that operate only in a single location, it’s easy to identify where economic production is taking place. For multi-plant firms, Statistics NZ uses data on employment at each plant (collected from pay-as-you-earn tax filings) to allocate out the firm’s production to multiple locations.
    * For example, if a manufacturer employs 100 people in Auckland and 50 people in Canterbury, then 2/3 of its contribution to GDP would be allocated to the Auckland region. Getting back to Fonterra – the company has a head office in Auckland but production facilities spread around throughout other regions. This means that a relatively small share of Fonterra’s contribution to GDP would be attributed to Auckland.
    Finally, and this is probably a bit petty of me, but when statistics challenge your assumptions about the world, the appropriate response is _not_ to question the numbers, it’s to question your prior assumptions! Stats NZ is bloody good at their work – they don’t make a lot of elementary errors – and regular users of their products aren’t idiots either 😉

    1. There is a flaw in using staff numbers in that it doesn’t reflect productivity gains. Say a firm has 100 staff in HO and 100 operating a plant. Output split 50:50. Now the plant is automated so it can be run by 50 staff. Now the split is 67:33. This increases the HO contribution to GNP and reduces the plants. Stats must be aware of this but not sure how they adjust. When I worked for multiple site firms, some of the Stats returns were site specific (rather than legal entity) so this may be how they split by region.

      1. Agreed! As the empirical evidence on agglomeration suggests that firms and plants located in larger or denser urban areas tend to be more productive than firms elsewhere, this probably means that the Regional GDP numbers _underestimate_ the size of NZ’s urban economies to a degree.

  7. “but when statistics challenge your assumptions about the world, the appropriate response is _not_ to question the numbers, it’s to question your prior assumptions!”
    But really! Where does this talk of believing the numbers and not one’s prejudices leave poor old Gerry Brownlee? He just “knows” that everyone wants more roads. Facts (or statistics) have no place in this thinking -or lack of thinking.

    1. Well he’s not Transport Minister any more, and perhaps that self-belief in “his hunches” over everyone else’s numbers is maybe a reason he got moved along?

      1. GregN – re your comment “NZ as a country can’t get rich buying and selling houses to each other, but we sure could put our GDP through the roof if each year, as a nation we did exactly that – sold and then bought houses off each other.” – isn’t that just exactly what we are doing? And then Key and English crowing about a rockstar economy, when actually it’s just people paying stupid prices for a villa in Grey Lynn?

        1. I can’t comment on the Grey Lynn Villa example but I know they said the same thing 30 or even 20 years back when Grey Lynn villas, Herne Bay Villas, Remuera Villas and anything even a tin shack in the GZ were all selling for “stupid prices” back then too, oddly enough, whatever they paid then, I’m sure its going to cost 10 times more $ to buy the same thing now as it did even if its now 20-30 years more run down.

          And I’m equally sure there is a lot of people who paid “stupid prices” too for similar old villas back in the day – except they were located in the provincial centres – they thought they could rent them out for a bomb, only I guess they’re not crowing about the money they make in the local newspapers these days – probably finding most of those renters are now standing empty as the renting population there has moved on when the jobs, and local newspapers there, did too.

          I think the stats on stupid house prices merely show that when it comes to housing demand its all about location, location, location, and if you have 50% of your GDP production concentrated in places (cities) where less than 50% of the countries house/apartment stock is located, then prices in those cities for those houses etc can only go up until you get a better balance in the economy.

          Most economists call that sort of thing a “shortage”. Well the current government actually calls it a “crisis” but, like a “War on Terror”, you know what they mean.
          So maybe the 50% GDP located in 3 cities, explains pretty much why all those places (‘cept Wellington) has a “housing crisis”?

          Maybe its really more a GDP crisis – too much GDP in too few locations? Not enough Of the “Dairy White Gold” being spread around the regions as the Fed Farmers would have you believe?

          Anyway we can’t all buy the same villa in Grey Lynn (or anywhere else) at the same time, so each can only be bought and sold once per sale.

          But Dairy or similar commodity futures? Yep plenty of those to go around for everyone to buy and sell to each other as much as they like – makes for a real simple zero sum game, but with lower commissions than houses.

          And forget Commodities, maybe we need a housing futures market, so we stop speculating on and building “real” houses, and simply speculate on paper houses?

          Or maybe Lotto should stop giving out cash prizes and start offering real houses in Auckland as better deal than any cash prize
          – that way you don’t lose 25% of the cash prize value by the time you’ve found your house and its gone up in value by that much since you started looking.
          So that way you get a house you can live in, while you find the next one, and sell up and move on.

  8. I guess I should have known this already but I didn’t as I have never looked at regional GDP before. The weakness with GDP is it captures all spending on (legal) final goods so that includes putting Christchurch back up etc. The regional data is based on production rather than spending data (you can also assess GDP from Income). All three methods should give the same answer. What I dont understand is how they deal with the government sector (G) in the production data, transfers will be ignored but how do they allocate G? Maybe it doesn’t matter because the government doesnt produce anything? The good thing about using production survey data for the regional GDP is is should allocate to the region where the goods were produced and avoid the head office issue people have raised. But I was certainly surprised by the maps.

      1. I’m also a little surprised that in Wellington, there is a blip up the coast – it’s past Porirua (where lots of work gets done), so I guess it must be Waikanae or Paraparaumu – which is an area predominantly featuring old retired people growing old and wrinkly in the sun. I wouldn’t have thought that there was that much actual GDP being generated there – perhaps I’m wrong. Maybe it is that if they have a job, it is still in Wellington, and they are taking the money back home? Or am i just being picky? Still: very interesting.

        1. Excellent question! If you look at Christchurch and Auckland I’ve done something similar and included satellite towns with a fair share of commuting into the city as they’re likely to be part of the overall urban economy.

          Basically, I’ve assumed that a supermarket in Paekakariki is more similar to a supermarket in Lower Hutt than one in Dannevirke. The former two primarily serve urban workers.

  9. I don’t care. The TV says that the real backbones of our economy are Fonterra’s truck drivers in shorties so that’s what I believe.

  10. Obviously a huge chunk of Auckland’s GDP is derived from services not real production that would earn foreign exchange. A dozen flat screen TVs sold in Auckland don’t add anything to the nations wealth but it sits there as part of Aucklands GDP. How about a post explaining which parts of the country earn foreign currency.

    1. Try telling people who work in service industries that their income doesn’t count! The economics of consumption is not irrelevant. Obviously we have to sell to the outside world to buy things from it, but that doesn’t somehow make trading with each other irrelevant. If we have trouble selling to the world then imports will become more expensive [or rather our dollar will likely fall], and while clearly most of our exports are primary produce the only way to diversify from that is to, wait for it, diversify from it: ie build up skills and experience in those less material fields of services. This idea that nothing of value happens in cities is exactly the silly idea that this post is, in part addressing.

      1. And a well balanced economy has a large part of its GDP coming from “internal” demand, not from just selling commodity products overseas to anyone who will buy your products from you one day, and from your competitor countries the next – just as long as you’ll accept their price.

        All the focus on exports as good “GDP” does is allow us to have the forex to buy the commodity products we sold last year back as finished goods at “The Warehouse”, or from a “Grab one” deal site.
        Better to have the skills and expertise and capital, to make the stuff here in the first place, that way we get to have the skills, the jobs, the products, and the taxes and avoid the need to sell yet more crap overseas to buy a different lot of crap back.

    2. Having personally spent 8 months lat year commuting from Auckland to Sydney to sell services to Australian organisations, you may need to rethink the concept that services don’t add anything to GDP. Considering the faces regularly saw in Koru at both ends, I doubt I’m the only one.

    3. Indeed, I sell ‘services’ to Asia and Australia every day from my desk in central Auckland.

      Anyway I don’t buy the argument that only foreign trade counts for anything. Take that argument a little further to the limits of the world and all our international trade isn’t doing anything for the wealth of the world because we’re not selling anything to the Martians.

    4. From an economic perspective, it’s bizarre to say that services “aren’t real production”. If urban service workers aren’t productive, why are people paying for their services?

      Supply and demand: how does it work?

    5. So who’s buying the flat screen TV’s? possibly someone working in an industry that supports the parts of the economy that export (and of course exports aren’t just primary products). A classic case people love to harp on about is finance and insurance sectors which are primarily based in Auckland. Without the banks and the lending they provide farmers wouldn’t have been able to invest in new stock or equipment and without that they wouldn’t be able to get the production levels they do. Without the insurance sector they could be stuffed in the case of floods or other disasters.

      1. It’s interesting that until the deregulation of the 1980s, most banks and insurance companies were based in Wellington, and there were many regional banks and insurance companies. Since then, we’ve seen the mergers and acquisitions and subsequent relocations to Auckland, many of which were justified on the grounds that “you had to have your HO in the biggest market”. Although Alan Hubbard proved the regions could run a finance company just as badly as any Aucklander – South Canterbury Finance’s $1.6b was 19% of the $8.5b of investor funds lost since 2006. http://www.sharkpatrol.co.nz/failed-finance-companies.htm

  11. GDP is about the output of both goods and services. Tell a hairdresser their work doesn’t count for national output or income. Just to put my two pennies in I see lmitations with this drive for a lower dollar and to be an export-led economy. It may improve the balance of payments but I wonder if it doesn’ t lower our standard of living. We’re not supposed to buy imports because they lead to a worsening of the current account and a lower valued dollar helps exports but increases our household consumption spending costs (cost of buying imports).

    How about more products being produced in NZ so we don’t have to rely on imported goods? In the U.S. they realise their economy does better when consumption spending increases. They don’t rely on exports. Lets focus on producing great value-added products for both export and domestic consumption.

  12. When a factory in provincial NZ with a head office in Auckland earns money, I don’t consider that “Auckland generated”.

    Sure you’re not totally misrepresenting the NZ economy here?

    NZ’s economy is very much in provincial NZ. Men in denim build the country, men in suits take the credit!

    1. Seriously read the post Geoff or some of Peter’s comments above. The GDP figures are based on where the production happens, not where the head office is.

      The whole point of this post seems to be to highlight so well is that the idea most of NZs economy is based in the regions is simply a myth.

    2. sure youre not completely missing the point of the post? urban nz generates far more wealth than rural nz. and its not caused by head offices – instead its caused by a whole bunch of relatively productive people producing goods and services that people value. like transport blogs.

      1. When a financial services company in Auckland generates X billions of dollars, that money comes from around NZ. It is not generated from within that business in Auckland itself. Money is created by manufacturing something and selling it. IMO the real source of the value is at the point of manufacture, not the point of sale.

        1. I’m guessing you believe Fonterra’s wealth is generated in their head office, and has nothing to with the farmers, Peter!

          You would be wrong.

        2. Geoff, in the chart above Foterras wealth is allocated to the regions of production, not the head office.

      2. The data seem to indicate that GDP is almost directly proportional to numbers of persons in employment. 24.5% of those in employment are outside the 15 largest urban areas and they generate 23.2% of the GDP. I find that higher than I would have expected and certainly doesn’t indicate that urban NZ has a substantial productivity premium. Have I missed something? Should I question the methodology (in spite of the warnings)?

        1. MFD raises an interesting point. Perhaps there are lots of employed but unproductive poor people in Auckland dragging the average down?

        2. Where do most of the beneficiaries live? Are their a higher percentage in Auckland than in the regions? Or minimum wage workers – I suspect not too many cleaners in rural NZ?

          How does rural productivity compare to urban productivity – GDP per dairy farmer compared to GDP per Fonterra factory worker or GDP per Fonterra HO office worker? Does the Stats NZ data allow you to drill down to discover the “value add” at each level? Ditto for tourism – can it separate out a tour company’s GDP “value add” to the regions where the tours operate? If I book a coach tour of the South Island, where does the GDP get counted?

          I expect the good folk at Stats NZ have been asking all these questions for decades. 😉

        3. I am assuming all unemployed/sick/parental/retired beneficiaries aren’t included in the stats. I expect more minimum wage workers in the cities which could hide the productivity of others in the city.

          I think the picture is great though, a different way to look at our economy.

        4. Ari, I’m not sure about the NZ stats, but in the US, it was worked out that for every dollar spent on unemployment benefits, the economy grew by $1.61, which is apparently a higher level of economic growth than generated by some employed folk. Has to do with the rate of spending, which for unemployed people is faster (they spend it as soon as they receive it). Some economists say that having a high unemployment rate during recovery from recession is good for the economy, as money is pumped back in to the economy much sooner than it would if the unemployment rate were low.

        5. MFD do you think it is possible that there is not enough share traders in those regions. Or possibly they need to have a few more people like Mr Key to trade in currency speculation.

  13. Is the problem in the rural sector is all those farming wives as unpaid labourers? Surely anything that is not paid for is not measurable as part of GDP. What about those farmers that are burying carbon instead of buying fertiliser, they are not being loyal NZ’ers by not paying ofr their added soil fertility and soil health.

  14. The Auckland economy is riding on the backs of the Ag and Hort sector, we all know this. And yes we expect more population equals more growth – means absolutely nothing. If we stopped increasing the population into Auckland, the housing would crash and so would all related middle men services the realters,insurance,traders,investors. And is left is insurance for farms, lawyers, accountants, and businessmen for fonterra, and educators for there children.

    Stop population growth in NZ and the regions with thrive, and Auckland will be seen for what it is, a nothing city full of middlemen who rely on property booms and Agriculture.

Leave a Reply

Your email address will not be published. Required fields are marked *