When the Commerce Commission was looking into retail grocery competition, my shop urged the Commission to check on something important.
As best we could tell, it would be effectively impossible for a new large-format supermarket chain to open in New Zealand. Cities had not zoned enough space for full-sized grocery stores. So, anyone wanting to open a set of stores up and down the country would have enormous difficulty finding viable sites with the right zoning.
If the government was worried about a perceived lack of competition in retail groceries, and cities had de facto banned competition, solutions could be straightforward. Just legalise new supermarkets and see what happens. If new grocers enter, the problem is solved. If none enter despite its being easy, maybe there weren’t super profits to be had in the first place.
But we did not have the Geographic Information Systems expertise at hand to prove just how big the problem really was.
That was about five years ago. And how the world has changed in the interim. Not in terms of groceries – although the new fast-track process for consenting new supermarkets is an improvement, and Auckland now has a Costco.
It is now possible for an economist and a few frontier AI models to do the work that looked well beyond our limits not all that long ago.
Last week, The Initiative launched an experimental website project.
I asked Anthropic’s Claude Fable 5 to scrape through the district plans for New Zealand’s major cities and figure out where it would be easy to get permission to build a new supermarket, where it would be a challenge, and where it would be next to impossible.
I wanted it to build me the maps showing everything.
Fable classified roughly 250 zones across nine plans, plus hundreds of precinct documents. It built the maps for Auckland, Hamilton, Wellington (and environs), Christchurch and Dunedin.
The tools are superb, but far from perfect. I found a lot of things that Fable missed, catching some because I now have a fair bit of background in zoning relating to supermarkets, and others because I have lived in Wellington and Christchurch and know some of the terrain.
I enlisted the assistance of some urban economists and urban planning friends to cast their eyes over it. We caught a few more errors and omissions. Other frontier models – ChatGPT 5.6 and Perplexity Computer, provided adversarial review.
Fable rebuilt the maps about a dozen times during the review process.
The maps are now live at supermarketzoning.nzinitiative.org.nz. If you are reading the online version of the column, one of the maps, current as of 16 July, is likely in front of you. The ones on the Initiative’s website will update and improve over time.
A few things stand out.
First, most downtowns make large supermarkets a ‘permitted’ activity. However, those are the places where land prices are highest, traffic is most difficult and a large-footprint supermarket may have a hard time being viable simply because of land cost. Smaller metro-style supermarkets are more common downtown. Further away from downtown, there are fewer and smaller areas where supermarkets are allowed.
Second, the rules very explicitly forbid more competition in a whole lot of places. We found thirteen areas that cap the number of supermarkets. Other ones restrict their total floor area. Entrants will often have to prove that their new supermarket does not harm the ‘vitality and viability’ of the existing centres around incumbent supermarkets.
If this kind of geographic market segmentation were done by a quiet handshake among the larger retailers, they’d be in court on charges of anticompetitive conduct. But planners can do it with impunity.
Finally, the rules would be almost impossible for any potential new entrant to navigate without the help of local ‘fixers’. Different councils have different rules for supermarkets and define them differently. Even incumbents have difficulty navigating the system. Regulatory complexity creates a substantial barrier to entry.
The government’s new fast-track process for new supermarkets is an improvement. It is not perfect and hopefully resource management reform will solve the underlying problem. Urban planning must view competitive urban land markets, and competition more generally, as a benefit to be sought by the system rather than a harm to be mitigated by it.
Our work here is explicitly provisional. I am sure that there are still improvements to be made. We have a feedback form on the site. If you find something that needs to be fixed, let us know. Your suggestion may make it into the next rebuild of the website. We also provide all the workings so that your own preferred AI agent can test it out for you.
And perhaps the Commerce Commission can adopt this kind of approach in its future work. Before considering other options for encouraging competition, could the Commission first make sure that town planning isn’t the underlying problem?
Dr Eric Crampton is Chief Economist at the New Zealand Initiative. This article was sourced HERE
But we did not have the Geographic Information Systems expertise at hand to prove just how big the problem really was.
That was about five years ago. And how the world has changed in the interim. Not in terms of groceries – although the new fast-track process for consenting new supermarkets is an improvement, and Auckland now has a Costco.
It is now possible for an economist and a few frontier AI models to do the work that looked well beyond our limits not all that long ago.
Last week, The Initiative launched an experimental website project.
I asked Anthropic’s Claude Fable 5 to scrape through the district plans for New Zealand’s major cities and figure out where it would be easy to get permission to build a new supermarket, where it would be a challenge, and where it would be next to impossible.
I wanted it to build me the maps showing everything.
Fable classified roughly 250 zones across nine plans, plus hundreds of precinct documents. It built the maps for Auckland, Hamilton, Wellington (and environs), Christchurch and Dunedin.
The tools are superb, but far from perfect. I found a lot of things that Fable missed, catching some because I now have a fair bit of background in zoning relating to supermarkets, and others because I have lived in Wellington and Christchurch and know some of the terrain.
I enlisted the assistance of some urban economists and urban planning friends to cast their eyes over it. We caught a few more errors and omissions. Other frontier models – ChatGPT 5.6 and Perplexity Computer, provided adversarial review.
Fable rebuilt the maps about a dozen times during the review process.
The maps are now live at supermarketzoning.nzinitiative.org.nz. If you are reading the online version of the column, one of the maps, current as of 16 July, is likely in front of you. The ones on the Initiative’s website will update and improve over time.
A few things stand out.
First, most downtowns make large supermarkets a ‘permitted’ activity. However, those are the places where land prices are highest, traffic is most difficult and a large-footprint supermarket may have a hard time being viable simply because of land cost. Smaller metro-style supermarkets are more common downtown. Further away from downtown, there are fewer and smaller areas where supermarkets are allowed.
Second, the rules very explicitly forbid more competition in a whole lot of places. We found thirteen areas that cap the number of supermarkets. Other ones restrict their total floor area. Entrants will often have to prove that their new supermarket does not harm the ‘vitality and viability’ of the existing centres around incumbent supermarkets.
If this kind of geographic market segmentation were done by a quiet handshake among the larger retailers, they’d be in court on charges of anticompetitive conduct. But planners can do it with impunity.
Finally, the rules would be almost impossible for any potential new entrant to navigate without the help of local ‘fixers’. Different councils have different rules for supermarkets and define them differently. Even incumbents have difficulty navigating the system. Regulatory complexity creates a substantial barrier to entry.
The government’s new fast-track process for new supermarkets is an improvement. It is not perfect and hopefully resource management reform will solve the underlying problem. Urban planning must view competitive urban land markets, and competition more generally, as a benefit to be sought by the system rather than a harm to be mitigated by it.
Our work here is explicitly provisional. I am sure that there are still improvements to be made. We have a feedback form on the site. If you find something that needs to be fixed, let us know. Your suggestion may make it into the next rebuild of the website. We also provide all the workings so that your own preferred AI agent can test it out for you.
And perhaps the Commerce Commission can adopt this kind of approach in its future work. Before considering other options for encouraging competition, could the Commission first make sure that town planning isn’t the underlying problem?
Dr Eric Crampton is Chief Economist at the New Zealand Initiative. This article was sourced HERE
No comments:
Post a Comment
Thank you for joining the discussion. Breaking Views welcomes respectful contributions that enrich the debate. Please ensure your comments are not defamatory, derogatory or disruptive. We appreciate your cooperation.